Mathematical Modeling Libro

International Perspectives on the Teaching and
Learning of Mathematical Modelling
Gloria Ann Stillman

Gabriele Kaiser
Christine Erna Lampen Editors
Mathematical
Modelling
Education and
Sense-making
International Perspectives on the Teaching
and Learning of Mathematical Modelling
Editorial Board IPTL
Series editors
Gabriele Kaiser, University of Hamburg, Germany
Gloria Ann Stillman, Australian Catholic University, Australia
Editorial board
Maria Salett Biembengut, Pontifical Catholic University of Rio Grande do
Sul - PUCRS, Brazil
Werner Blum, University of Kassel, Germany
Helen Doerr, Syracuse University, USA
Peter Galbraith, University of Queensland, Australia
Toshikazu Ikeda, Yokohoma National University, Japan
Mogens Niss, Roskilde University, Denmark
Jinxing Xie, Tsinghua University, China
This book series will publish various books from different theoretical perspectives
around the world focusing on Teaching and Learning of Mathematical Modelling at
Secondary and Tertiary level. Fully refereed books arising from the biennial
conference called ICTMA, organised by the ICMI affiliated Study Group ICTMA
(International Community of Teachers of Mathematical Modelling and Applications)
will also be published in this series. These books display the worldwide state-of-
the-art in this field and will be of interest for a wider audience than the conference
participants.
ICTMA is a worldwide unique group, in which not only mathematics educators
aiming for education at school level are included but also applied mathematicians
interested in teaching and learning modelling at tertiary level are represented.
ICTMA discusses all aspects related to Teaching and Learning of Mathematical
Modelling at Secondary and Tertiary Level, e.g. usage of technology in modelling,
psychological aspects of modelling and its teaching, modelling competencies,
modelling examples and courses, teacher education and teacher education courses.
More information about this series at http://www.springer.com/series/10093

Gloria Ann Stillman • Gabriele Kaiser
Christine Erna Lampen
Editors
Mathematical Modelling
Education and Sense-making
Editors
Gloria Ann Stillman Gabriele Kaiser
Australian Catholic University University of Hamburg
Ballarat, VIC, Australia Hamburg, Germany
Christine Erna Lampen

Stellenbosch University
Stellenbosch, South Africa
ISSN 2211-4920 ISSN 2211-4939 (electronic)

International Perspectives on the Teaching and Learning of Mathematical Modelling
ISBN 978-3-030-37672-7 ISBN 978-3-030-37673-4 (eBook)
https://doi.org/10.1007/978-3-030-37673-4
© Springer Nature Switzerland AG 2020

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the
editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.
This Springer imprint is published by the registered company Springer Nature Switzerland AG.
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Series Preface
Applications and modelling and their learning and teaching in schools and universi-
ties have become a prominent topic in the last decades in view of the growing
worldwide relevance of the usage of mathematics in science, technology and every-
day life. There is consensus that modelling should play an important role in math-
ematics education, and the situation in schools and universities is slowly changing
to include real-world aspects, frequently with modelling as real world problem solv-
ing, in several educational jurisdictions. Given the worldwide impending shortage
of students who are interested in mathematics and science, it is essential to discuss
accelerating possible changes of mathematics education in school and tertiary edu-
cation towards the inclusion of real world examples and the competencies to use
mathematics to solve real world problems.
This innovative book series established by Springer, “International Perspectives
on the Teaching and Learning of Mathematical Modelling”, aims at promoting aca-
demic discussion on the teaching and learning of mathematical modelling at various
educational levels all over the world. The series will publish books from different
theoretical perspectives from around the world dealing with Teaching and Learning
of Mathematical Modelling in Schooling and at Tertiary level. This series will also
enable the International Community of Teachers of Mathematical Modelling and
Applications (ICTMA), an International Commission on Mathematical Instruction
affiliated Study Group, to publish books arising from its biennial conference series.
ICTMA is a unique worldwide group where not only mathematics educators deal-
ing with education at school level are included but also applied mathematicians
interested in teaching and learning modelling at tertiary level are represented as
well. Five of these books published by Springer have already appeared.
The planned books will display the worldwide state-of-the-art in this field, most
recent educational research results and new theoretical developments and will be of
interest for a wide audience. Themes dealt with in the books will be teaching and
learning of mathematical modelling in schooling and at tertiary level including the
usage of technology in modelling, psychological, social and cultural aspects of
modelling and its teaching, modelling competencies, curricular aspects, modelling
examples and courses, teacher education and teacher education courses. The book
v
vi Series Preface
series aims to support the discussion on mathematical modelling and its teaching
internationally and will promote the teaching and learning of mathematical model-
ling and research of this field all over the world in schools and universities.
The series is supported by an editorial board of internationally well-known
scholars, who bring in their long experience in the field as well as their expertise to
this series. The members of the editorial board are: Maria Salett Biembengut
(Brazil), Werner Blum (Germany), Helen Doerr (USA), Peter Galbraith (Australia),
Toshikazu Ikeda (Japan), Mogens Niss (Denmark), and Jinxing Xie (China).
We hope this book series will inspire readers in the present and the future to
promote the teaching and learning of mathematical modelling all over the world.
Gloria Ann Stillman Gabriele Kaiser

Ballarat, Australia Hamburg, Germany
Contents
1 Helena Wessels: Her “Life Story”�� 1

Dirk Wessels
2 Sense-Making in Mathematical Modelling
and Applications Educational Research and Practice�� 15
Gloria Ann Stillman, Gabriele Kaiser, and Christine Erna Lampen
Part I Innovative Approaches in Modelling Educational

Research and Teaching
3 Connections of Science Capital and the Teaching
and Learning of Mathematical Modelling: An Introduction �� 33
Pauline Vos, Paul Hernandez-Martinez, and Peter Frejd
4 Teachers’ Possibilities to Generate Science Capital
for Modelling�� 39
Peter Frejd
5 Science Capital, Habitus, and Mathematical Modelling
Practices in the Field of University Education�� 51
Paul Hernandez-Martinez
6 On Science Museums, Science Capital, and the Public
Understanding of Mathematical Modelling�� 63
Pauline Vos
7 Investigating the Complexity of Student Modelling
by Coordinating a Diversity of Research Perspectives �� 75
Corey Brady, Jeffrey A. McLean, Hyunyi Jung, Aran W. Glancy,
and Ángeles Domínguez
vii
viii Contents
8 Mapping Shifting Discourses in Classroom Talk

Through Indexical Discourse Analysis: From the World
of Mathematics to the World of Work�� 85
Stephen T. Lewis and Azita Manouchehri
9 Joy of Mathematical Modelling: A Forgotten Perspective? �� 95
Yuriy Rogovchenko, Olov Viirman, and Stephanie Treffert-Thomas
10 Re-signifying Function Concept: A Mixed Methods Study
to Understand the Contributions of the Dialogic Approach
of Ethnomodelling�� 107
Milton Rosa, Daniel Clark Orey,
and Diego Pereira de Oliveira Cortes
Part II Research into, or Evaluation of, Teaching Practice
in Mathematical Modelling Education
11 Sense-making with the Mathematical Modelling Process:
Developing a Framework for Faculty Practice�� 119
Todd Abel, Mary Elizabeth Searcy, and Tracie Mc Lemore Salinas
12 Mathematical Modelling with a Solution Plan: An Intervention
Study about the Development of Grade 9 Students’ Modelling
Competencies�� 129
Catharina Beckschulte
13 A Case Study of Tensions and Challenges Arising as a Swedish
Upper Secondary Teacher Designs and Implements a Model
Development Sequence on Statistics�� 139
Jonas Bergman Ärlebäck
14 Year 6 Students’ Gradual Identification of Mathematical
Models of Average Speed when Making Sense of ‘Walking’�� 151
Takashi Kawakami, Shigekazu Komeda, and Akihiko Saeki
15 A Time-Based Measurement of the Intensity of Difficulties
in the Modelling Process�� 163
Heiner Klock and Hans-Stefan Siller
16 Deepening and Expanding Mathematical Models of Speed
in Relation to Walking: The Case of Year 8 Students �� 175
Shigekazu Komeda, Takashi Kawakami, Masafumi Kaneko,
and Takashi Yamaguchi
17 Facilitating Modelling Activities in a Grade 5 Classroom�� 187
Azita Manouchehri, Mehmet Bekdemir, and Xiangquan Yao
18 Mathematical Modelling as a Learning Environment
to Transform a Street Activity into a Sport Practice�� 199
Daniel Clark Orey, Milton Rosa, and Rogério Braga Soares
Contents ix
19 Concurrent Development of Engineering Technician

and Mathematical Modelling Competencies �� 209
Lidamari de Villiers and Dirk Wessels
Part III Pedagogical Issues for Teachers and Teacher Educators

Using Mathematical Modelling and Applications
20 Sense-making in Modelling Tasks: What Can We Learn
from Other Domains?�� 223
Piera Biccard
21 The Deployment of Mathematics in Areas Other than Those
Normally Associated with Mathematical Modelling
and the Applications of Mathematics �� 237
Cyril Julie
22 Rising to the Challenge: Promoting Mathematical Modelling
as Real-World Problem Solving�� 253
Peter Galbraith, Derek Holton, and Ross Turner
23 Mathematics Trails and Learning Barriers�� 265
Iwan Gurjanow and Matthias Ludwig
24 Validating with the Use of Dynamic Geometry Software�� 277
Corinna Hankeln
25 Using a Mathematical Modelling Activity to Assist Students
to Make Sense of a Limit Theorem in Trigonometry�� 287
Toshikazu Ikeda and Max Stephens
26 Exploring Habits and Habitus of Biomedical Students
with Modelling Tasks �� 299
Hanti Kotze
27 Metacognitive Strategies in Group Work in Mathematical
Modelling Activities – The Students’ Perspective �� 311
Alexandra Krüger, Katrin Vorhölter, and Gabriele Kaiser
28 Fostering Students’ Construction of Meaningfulness
of Mathematics with Mathematical Modelling Problems�� 323
Katrin Vorhölter and Björn Schwarz
29 Teachers’ Perspectives on Students’ Metacognitive Strategies
during Mathematical Modelling Processes – A Case Study�� 335
Lisa Wendt, Katrin Vorhölter, and Gabriele Kaiser
x Contents
Part IV Assessment of Mathematical Modelling in Schools

30 Cognitive Diagnostic Modelling for Mathematical Modelling
Assessment�� 349
Cigdem Alagoz and Celil Ekici
31 Occurrences of Mathematical Modelling Competencies
in the Nationally Set Examination for Mathematical Literacy
in South Africa�� 361
Mbulelo Bali, Cyril Julie, and Monde Mbekwa
32 Expectations for Challenge in Modelling and Its Assessment�� 371
Jill P. Brown
33 Modelling Tasks in Central Examinations Based
on the Example of Austria �� 383
Hans-Stefan Siller and Gilbert Greefrath
Part V Applicability at Different Levels of Schooling and University

34 Characterising Modelling Competency in Students’ Projects:
Experiences from a Natural Science Bachelor Program�� 395
Morten Blomhøj
35 From Royaumont to Lyon: Applications and Modelling
During the 1960s�� 407
Dirk De Bock and Bert Zwaneveld
36 Interactive Case Practice Teaching on Mathematical
Modelling Course �� 419
Xiaojun Duan, Dan Wang, and Mengda Wu
37 Pre-service Teachers’ Sense-making of Mathematical
Modelling through a Design-Based Research Strategy�� 431
Rina Durandt and Geoffrey V. Lautenbach
38 Mathematical Modelling: A ‘Growing Tree’ for Creative
and Flexible Thinking in Pre-service Mathematics Teachers�� 443
Rajendran Govender
39 The Design of a Successful Teacher Training to Promote
Interdisciplinary STEM Modelling Projects �� 455
Jean-Marie Lantau, Martin Bracke, Wolfgang Bock, and
Patrick Capraro
40 Using Mathematical Modelling in the Teaching and Learning
of Mathematical Literacy: Perspectives on the South African
Context�� 467
Duncan Mhakure
Contents xi
41 A Lecturer’s Learning Goals for Teaching Mathematical

Modelling�� 479
Kerri Spooner
Part VI Examples of Modelling and Applications in Practice

42 Embodied Phenomenology in Mathematical Modelling
of Sailing for Integrated STEM Learning �� 493
Celil Ekici and Cigdem Alagoz
43 Algebra Students Build Stock/Flow Models to Study
Non-linear, Dynamic Feedback System Problems�� 505
Diana M. Fisher
Refereeing Process �� 515
Index�� 517
Chapter 1
Helena Wessels: Her “Life Story”
Dirk Wessels
Abstract The life of Dr. Helena Wessels – natural researcher, born teacher, and
brilliant lecturer in mathematics education – tragically and prematurely ended when
her and her husband’s family home in the Strand burnt down on the night of 24
February 2018. At the age of 62 years, she was reaching for several ideals in her
academic life: A very important one for her was to improve her own teaching to the
fullest. She worked diligently on this. After a presentation, or after a series of les-
sons, she studied and carefully considered all elements of feedback. She continually
strove for excellence. Each aspect of criticism was critically and reflectively ana-
lyzed. Toward the end, she was keenly interested in comments about the use and
integration of technology in her presentations and in her teaching of mathematics
education. It came as no surprise that she received the highest award from
Stellenbosch University (SU) for quality teaching. Her recommendation by SU was
accepted by the Higher Education Learning and Teaching Association of South
Africa – the HELTASA group awarded the Teaching Advancement at Universities
(TAU) Fellowship to her for a 3-year period. This prestigious award was very spe-
cial to her, and she enjoyed and thrived on the challenging level of collaborative
project work and research with colleagues from other universities in South Africa.
Her appointment as lecturer in Foundation Phase Mathematics at Stellenbosch
University in January 2009 was really the beginning of a successful, industrious
second life for Helena. In her earlier career, she excelled as a music and mathemat-
ics primary school teacher. This privilege and opportunity meant a lot to her, and she
dedicated her heart and life to it. Later, as an emerging scholar and researcher, all
her research was done with her own learners and students. Research projects were
often designed and chosen with national and international colleagues who shared
It was a great privilege to write this dedication to Helena Wessels, my late wife. It was a very dif-
ficult task, and I would not have been able to accomplish this without the sterling help of Piera
Biccard, Sonja Strydom, and Ené de Bruin – three people who knew Helena well and loved her
dearly. Thank you for your friendship. Dirk Wessels.
D. Wessels (*)
Department of Curriculum Studies, Stellenbosch University, Stellenbosch, South Africa
e-mail: dirkwessels@sun.ac.za
© Springer Nature Switzerland AG 2020 1

G. A. Stillman et al. (eds.), Mathematical Modelling Education
and Sense-making, International Perspectives on the Teaching and Learning
of Mathematical Modelling, https://doi.org/10.1007/978-3-030-37673-4_1
2 D. Wessels
her interest and passion – the teaching and learning of statistics education and math-
ematics education. Her love for music, statistics, and mathematics education were
driven by the need to foster and develop an audiences’ enjoyment or students’
understanding. She was dedicated to her students, and her energy went into her
preparation to teach in such a way that they would see and understand the close con-
nection between theory and practice through the problem-centered teaching
approach. She worked energetically and enthusiastically until they were able to
implement this in their own classrooms. Her meticulous preparation for the lectures
the day after student lesson presentations kept her up for hours, and sometimes she
did with no sleep at all. She had an innate specific and special ability to elucidate
how theory and practice were integrated – this became apparent in her theoretical
analyses of the problem-centered teaching approach, but also in her foundational
views about the social responsibility programs she coordinated in the Faculty of
Education. She was also able to integrate these projects with her own teaching and
her own research. Her husband Dirk, also a professor in mathematics education, was
her biggest supporter and admirer. Helena, the person with her passion and energy,
will be missed by friends, colleagues, and students.
Keywords Born teacher · Statistics and mathematics education · Reflective

practitioner · Improving own teaching · Dedication as lecturer · Love for her
students · TAU award winner
1.1 Biography
Helena Margaretha Henning (15 January 1956–24 February 2018) was born in
Zastron in the Free State province of South Africa. She and her older sister, Hester,
were the only daughters to Nicolas and Idelette Henning. The Henning family was
a caring and progressive family and moved from a farm in Zastron to Groblersdal
where Helena grew up, and later to Pretoria. Helena grew up in a peaceful and lov-
ing Christian home. She was married to Neels Müller (first marriage) with whom
she had three children: Idelette and twin boys, Helgard and Henning. After her
divorce from Neels Müller, she married Dirk Wessels in 2004 who had three chil-
dren, Heleen, Dirkie and Herman. Helena had a very good relationship with the
three Wessels children. The marriage of Dirk and Helena brought these two families
together, and from then on, they were seen as one family. Dirk and Helena also
referred to each other’s children as their own children. The couple initially resided
in Pretoria but later settled in Strand, Western Cape province, when Helena was
appointed lecturer in Mathematics Education (Foundation Phase) at Stellenbosch
University in Stellenbosch.
1 Helena Wessels: Her “Life Story” 3
1.2 Tertiary Education and Becoming a Teacher of Note
Helena (Fig.1.1) first began a Bachelor of Science at the University of Pretoria but
after a tour to Europe she returned and enrolled in BMus in Music Education with
Mathematics III (1979). She only wanted to study music, which was the most
important thing in her life. Her love for music resulted in her completing a BMus
(Hons) in 1991. Initially, she taught Mathematics (Grades 8–12) and General
Science (Grades 8–9) for 4 years after the completion of her undergraduate degree.
After the birth of her children, she decided to stay at home, to ensure the three tal-
ented children received a good home upbringing. It is during this time that she
started her private music school (Muzikid for children aged 4–9) with Pat Michel.
They used music therapy to help develop school readiness in children. The Suzuki
method was followed where children learned to compose music and create music
with support from Helena. In 1992–1993, she served as lecturer in the Department
of Music at the University of Pretoria. She completed her MMus Ed in 1994 and Dip
MMus, Music Ed. During this time, the Music Department decided to recite an
operetta, The Merry Widow from Franz Lehar. Helena was chosen to sing the lead-
ing soprano role of Hanna.
In 1993, she was appointed at Lynnwood Primary School to teach mathematics
and music while she was also responsible for the school choir, an unusual but per-
fect combination for her. After her appointment, she became a master teacher of the
school for 12 years, which was the highest education award the Gauteng Education
Department (GDE) issued. As a teacher, she prepared a poster (Müller 2001) for
PME (Utrecht) in 2001 which focused on data handling in primary school statistics.
This was a project with her own children in Grade 5. During this time, she continued
Fig. 1.1 Dr. Helena Wessels (15 January 1956–24 February 2018)
4 D. Wessels
her studies and completed her master’s degree in School Music at the University of
Pretoria (UP). The Senior School Choir of Lynnwood Primary School performed
nationally and won numerous awards. A culture of positive participation and enjoy-
ment is what Helena always strived for in whatever she was involved. For 6 con-
secutive years, she organized a weeklong choir camp at Horseshoe Guest Farm in
Normandien District near Newcastle in KwaZulu-Natal Province. Participation and
professionalism were exceeded above expectation in performances with, among
others, the well-known Drakensberg Boys Choir, KwaZulu-Natal Youth Orchestra,
and other well-known singers and large fundraising projects locally in the Newcastle
District. She stood out as a choirmaster and through this and her own love of music
made important contributions toward music and the discipline and learning of math-
ematics. In 2006, she obtained her DEd degree in Didactics (Mathematics Education)
from the Department of Further Teacher Education, University of South Africa
(Unisa). Her husband Dirk Wessels (Unisa) and Hercules Nieuwoudt (North-West
University (NWU)) supervised her mathematics education study. The focus of the
study was statistics education, and in 2006 she was the first PhD/DEd student in
South Africa with a doctorate in statistics education.
1.3 Helena the Devout Christian
Helena grew up in a Christian home and lived as a dedicated disciple of Jesus Christ.
Everyone who knew her testified about it. Rina Scott-Wilson of Australia, a joint
former PhD student of Helena and Dirk, said the following in a letter to Dirk: “It
was always so clear to me that God delighted in her and her ways and honored her
before men. This is rare, Prof. Wessels. Not many have been granted the quality of
life and the experiences that God worked in her life, often through her. Consider that
there is NOTHING that comes to us, except that which God works in through a
grace, then you have to admit that God has accomplished amazing things through
her, blessing others like myself and Piera.”
For Helena as a dedicated mother who always cared for her children, it was not
uncommon for her to stay up late waiting for the twin boys over the weekends to
come home. Her children’s birthdays were highlights. Detailed planning and execu-
tion with original decorations was never too much effort for her. Piera Biccard, one
of Dirk’s doctoral students, said the following at Helena’s funeral: “She had a wealth
of inner strength which she gave freely to those around. A lasting impression about
Helena is that she was always, always a lady. Beautiful inside and outside. Perhaps
in time I may forget our conversations… but will never forget the way she made me
feel.” Helena was always willing to give someone else an opportunity. As a third-
year student in the 1970s, she was asked at Tant Kotie Women’s Residence at the
University of Pretoria to make herself available as Primaria. However, a close friend
also applied for the position, and Helena decided to withdraw in order to give her
friend the opportunity.
1.4 Interest in Higher Education
At other times, she immediately said yes to a challenge or opportunity. For example,
it was the case with an opportunity to go to the University of Wisconsin-Madison
for 4 months as a PhD student in 2001 (see Fig. 1.2) to participate in an NSF-NRF
Exchange project. Norman Webb, senior researcher of the Wisconsin Center for
Educational Research who worked there with Elizabeth Fennema, Tom Romberg,
and Tom Carpenter, said about Helena: “I knew her to always be in good spirits.
When she lived in Madison she took full advantage of her experience and became
active in so many ways including the university choir (sang as soprano in the Elijah
concert by Mendelsohn), the church, attend MEd classes (Eric Knuth and others),
and whatever Madison has to offer.”
In the 1990s, she was introduced to the problem-centered approach (PCA) to
teaching and learning mathematics that prepared her for the visit to Wisconsin as it
was considered the strong base of the problem-centered approach in the
USA. Fennema and Carpenter set up the world-renowned Cognitively Guided
Instruction approach that was very close to the problem-centered approach of South
Africa under the leadership of Piet Human, Hanlie Murray, and Alwyn Olivier. The
problem-centered approach fitted Helena’s teaching style like a glove. She empha-
sized the advancement of mathematical thinking, and the school’s results showed it.
She trained colleagues in it, and later, the Gauteng Education Department asked her
Fig. 1.2 Lunch at Elizabeth Fennema’s house in Madison, December 2001: left to right – David
Mogare, Owen Fennema, Elizabeth Fennema, Helena Wessels, Peter Hewson, and Dirk Wessels
6 D. Wessels
to train others at other schools. She became a carrier of the problem-centered

approach signature pedagogy.
Piera Biccard, a colleague at Unisa, once asked her why Lynnwood Ridge
Primary School’s mathematics was so good: “I remember in our earlier discussions,
asking Helena about how she revolutionized mathematics teaching and learning at
Lynnwood Primary School. She spoke with such confidence and conviction about
what she did there that it was difficult not to notice both her logic and passion about
the problem-centred approach. Helena was unwavering amidst some strong opposi-
tion about the approach. Gerrit Stols, a friend and mathematics education colleague
from the University of Pretoria, said the following about her academic career: “She
was quickly known by local and international colleagues in Mathematics and
Statistics Education through her role at Stellenbosch University, her articles, her
conference presentations and her research.” Her desire to enter tertiary education
was so strong that it led her to resign from Lynnwood School in 2007 and choose a
postdoctoral fellowship at North-West University with the theme “The Effective
Teaching and Learning of School Statistics.” She started two projects in the Delmas
district on statistics education with practicing teachers. She then published three
articles on these two projects which began her research and academic career.
In 2009, Helena was appointed to the Department of Curriculum Studies at
Stellenbosch University where she was responsible for teaching Mathematics
Education for second to fourth year BEd students in the Foundation Phase (FP). She
was mentored and worked closely with Hanlie Murray, who was a world leader in
the problem-centered approach to education for young children. Hanlie had no trou-
ble spending time and energy on this young, capable, and eager teacher. Helena
became a strong advocate of problem-centered mathematics education at
Stellenbosch University. In subsequent years, the problem-centered approach was
expanded to include mathematical modelling. Her research and academic focus
shifted to the modelling community’s activities (e.g., Wessels 2011). Helena and
colleagues attended international congresses which resulted in numerous academic
papers and research reports (e.g., Scott-Wilson et al. (2017) and Wessels (2017)).
Helena became a known name in the International Community of Teachers of
Mathematical Modelling and Applications (ICTMA).
Helena believed wholeheartedly that the theory and practice of teaching and
learning Foundation Phase Mathematics should be closely aligned. Her desire was
that students should master the problem-centered approach and apply it success-
fully. Her students affirmed that the problem-centered teaching-learning approach is
the most appropriate way to teach and learn mathematics both theoretically and
practically. Experts also considered it the ideal way to encourage children to do
mathematics as mathematicians would do. Her point of departure was with Zalman
Usiskin, the director of the University of Chicago School Mathematics Project, and
Werner Blum, University of Kassel: always start your mathematics lesson with a
problem. Dan Canada, mathematics education, Eastern Washington University,
USA, wrote about Helena: “Professionally, what a keen mind she had, lending her
sharp intellect to the most inspiring thoughts on our shared interests of the teaching
and learning of Mathematics. As I visited her University, I could see the experience
and diligence that she had on her students. And it was clear to me that her students
knew Helena.”
She believed that the didactic practice to be developed with students was not just
to be able to name and understand the practice taught but also to establish a problem-
centered practice in their classrooms. Despite numerous challenges, her students
spoke with respect and belief about the problem-centered approach. Her emphasis
on learner contributions and learner development when applying the problem-
centered approach has helped keep students on track. She believed in the devolution
of authority, the empowerment of the young, and the dissemination of expertise,
energy, and strengths. She was convinced that when you trust people with a task and
they believe they can do it, individuals would not disappoint but that they would suc-
ceed. Annually, class sizes grew which resulted in an alternative model in order to
pay effective attention to each student or group as she could no longer give individ-
ual attention to each student or group. She had to develop models to still maintain
effective education without doing it all individually, thus developing her own version
of lesson study that was modelled on the Japanese model. Working with Geoff Wake
from Nottingham University (UK) enabled her to do so as she visited Nottingham
twice and went with him to Japan. The microteaching model and the practical educa-
tion model she developed worked well – she empowered students because they soon
began to deliver high-quality work in their groups. She tied this in with her own
assessment model. Students followed a prescribed lesson model based on the prob-
lem-centered teaching-learning approach. She encouraged and empowered students
so that they entered the class with greater confidence in their first year of teaching.
Helena strongly believed in the exposure of fourth-year students to real teaching
practices. Thursday mornings for the first semester BEd IV students would see them
travel in groups to underprivileged primary schools in the outskirts of Stellenbosch.
The aim of this project was that students would help and support teachers to develop
the mathematics skills (number sense) of learners in R-3. This became known as the
Count-Me-In project (CMI project). The Rupert Foundation, which mainly supports
schools in Franschhoek, donates annually large amounts to enable this project. This
project is still continuing and has now become a Department of Curriculum Studies
project. The Count-Me-In project also gained an international partner in 2016 when
Prof. Catherine Paolucci of the State University of New York (New Paltz Campus)
brought USA students to host a 2-week program with fourth-year students from
Stellenbosch University in Franschhoek during the July holidays. Fourteen students
had to work in pairs and prepare lessons and assess them. The cooperative and con-
textual enrichment of both groups was a highlight of this project. From the project
with Catherine Paolucci, Helena and Catherine published an article in the USA
outlining their work in the International Mathematics Enrichment Project (IMEP),
setting a good example to address cooperative and contextual challenges between
collaborating international groups.
Helena and husband Dirk annually mentored a group of fourth-year students.
This took place at the Wessels’ home. Dirk’s contributions focused on the nature,
8 D. Wessels
history, and philosophy of mathematics and attitude and beliefs thereof. The eve-
ning ended with a South African “braai” and fellowship. These mentors (volunteers)
had to take groups of third-year students and prepare them for their microteaching
lesson. This happened at the beginning of the year and ended at the end of the year
with a reflection session. The mentor team affirmed that the method used to mentor
a group of BEd students was complex but worthwhile. It allowed fellow students to
unlock their own potential, plan, consolidate, reanalyze, and then achieve success
during their microteaching session. Prof. Christa van der Walt, Vice Dean research
at Stellenbosch University Faculty of Education, commented: “What struck me is
the way she has involved her students and the very close ties between her teaching
and research. I think that’s what made her such a good lecturer. For example, I think
of the ‘Count Me In’ project and the way she handled school practice.” Ené de
Bruin, who was part of such a team in 2014, described Helena’s performance as a
lecturer and her leadership in Foundation Phase Mathematics education as
“phenomenal.”
Helena was always focused on providing excellent education. Her preparation
was thorough and comprehensive. Sleep was not a priority for her. She gave and
demanded only the best, attended all possible courses offered by Stellenbosch
University for lecturers, and tested new programs and tools. Her lectures became
dynamically interactive and technology-rich, which was enjoyed and applied by her
students themselves. Annually, the students rated her highly because of her dynamic
and progressive approach. She also participated in Stellenbosch University’s Centre
for Teaching and Learning conferences, which was reflected in articles on her expe-
riences and outcomes. The latest trends in Foundation Phase Mathematics education
in the world, such as noticing, modelling, lesson study, and the integration of tech-
nology, had been introduced into her classes and where she has done research
(Wessels 2018). Helena initiated a new degree, that is BEd FP Hons. She helped
with the design and development of the two mathematics education modules.
Mathematics education and literacy are the two main pillars of the degree.
1.5 Helena and Dirk: The Team 1
In 2015, Helena (project leader) and Dirk (sub-project leader) bid to host
ICTMA18 in Cape Town in 2017 presenting successfully to ICTMA’s Executive
Committee. A local organizing committee was set up and the big job started. With
great dedication, she and her team worked very hard and made the conference a
great success (Fig. 1.3).
She then also became the lead editor, until her untimely death, of the Springer
book publication arising after the conference with conference delegates being able
to propose chapters that were then rigorously reviewed and if accepted will appear
in this book to be published early in 2020. Piera Biccard noted: “Helena had an
incredible ability to multi-task. Recently, she was the chairperson of a steering com-
mittee that hosted an international conference. Everything ran like clockwork and
had the special Helena touch.”
Fig. 1.3 Helena at the podium in July 2017 (ICTMA18), handing thank you gifts to Erna Lampen
with Dirk watching on
1.6 Awards and Recognition
Helena received numerous awards. She was the best presenter (low technology) in
Afrikaans at the SA Akademie vir Wetenskap en Kuns’s (The South African
Academy for Science and Art) Graduate Research School based on a presentation
from her DEd study. As a lecturer, the Faculty of Education nominated her for three
consecutive years for the Rector’s Award for Excellence in Teaching which she
received from 2012 to 2014. In 2016, she received the Stellenbosch University
Teaching Fellowship for research entitled “Researching and Improving Technology-
Enriched Teaching and Learning in Higher Education” with a grant for 3 years and
received, after nomination by Stellenbosch University, the Higher Education
Learning and Teaching Association of South Africa Teaching Fellowship Award in
2017, which is known as the Teaching Advancement at Universities (TAU)
Fellowship, the highest teaching award at the national level. The representatives of
the respective universities meet four times a year for a week and work on a joint
project. It was awarded for 2 years. Karin Cattell-Holden, senior advisor of the
Centre for Teaching and Learning, said that the requirements for acquiring a TAU
Fellowship are very high. All these seemingly small and noteworthy recognitions
meant a lot to Helena and motivated her to further improve as a scholar and teacher.
10 D. Wessels
1.6.1 Helena the Scholar
Helena had a strong love for statistics education – after all, her DEd dissertation
dealt with this. Of the 21 publications from her pen, 6 dealt with statistics education.
Her research approach and focus dealt with “the effective teaching and learning of
mathematics and statistics education with a strong emphasis on teacher education.”
She described her own preferences for teaching as follows: “teaching for under-
standing is necessary for better mathematics results and this in turn becomes pos-
sible when teachers notice learners’ mathematical thinking. The development of
reflective practices is one of the cornerstones in teaching for understanding and
preservice teachers (PSTs) need well-structured and focused opportunities for
repeated systematic reflection over time to learn to notice and make sense of learner
thinking and reasoning.” In 2002, she attended the International Conference for the
Teaching of Statistics (ICOTS6) in the Cape with another poster session. In 2003,
there were three presentations – two at the Psychology of Mathematics Education
(PME) conference in Hawaii. In 2004 and 2005, her focus shifted to the develop-
ment and assessment of spatial skills in children. In 2005, she attended the
SAARMSTE congress in Windhoek and talked about data modelling, and later that
year in Australia with another statistical theme (Fig. 1.4).
At these PME and ICOTS conferences, her path crossed with international spe-
cialists in statistics education, such as Reading, Gal, Chick, Callingham, Pfannkuch,
Wild, Shaughnessy, English, Batanero, Pegg, Makar, Canada, Hatfield, Ben-Zvi,
Fig. 1.4 Statistics Education Round Table Conference, Monterrey, Mexico 2008. (Left to right)
Helen Chick, Helena Wessels, Rosemary Callingham, Jane Watson, and Janet Ainley
and Watson. In 2006, she attended the ICOTS7 Congress in Salvador, Brazil, and
the PME-North America Chapter (PME-NA28) Congress in Merida, Mexico, and in
July 2008 the Joint Study ICMI/IASE Round Table Conference in Monterrey,
Mexico. During this period, she increased her focus on modelling, whereby the
modelling tasks of her research were based on the mathematical as well as the sta-
tistical fields. With Helena’s postdoctoral appointment, she focused in her research
project on teachers’ professional development. In the years following her appoint-
ment at Stellenbosch University, she began to focus more on teaching, especially
research on improving her own teaching.
As Helena’s involvement increased in the field of mathematics education, the
need arose to focus on specific aspects of the field. The problem-centered approach
and modelling were always there, as was the professional development of her stu-
dents in various fields. Other areas of interest included creativity, noticing, reflec-
tion, metacognition, a better and deeper understanding of mathematics, lesson
study, mathematical thinking and reasoning, the role of technology in mathematics
education, the ideas that learners make, choosing and constructing modelling prob-
lems, and the choices that students and teachers make. She published in all these
areas. A task bank of modelling problems for teachers at all levels known as math-
ematical modelling problems (MMPs) or mathematics eliciting activities (MEAs)
was also part of her research agenda. Helena met globally with co-workers and
friends who shared her interest and passion, including Iddo Gal (University of
Haifa, Israel), Dan Canada (Eastern Washington University, Cheney, USA), Chris
Reading (University of New England, Armidale, Australia), Hans-Stefan Siller
(University of Wuerzburg), Gabriele Kaiser (University of Hamburg, Germany),
Catherine Paolucci (Ireland and SUNY’s New Paltz Campus, USA), Norman Webb
(UWM, USA), Maxine Pfannkuch (University of Auckland, NZ), Joachim Engel
(University of Ludwigsburg, Germany), Geoff Wake and Malcolm Swan
(Nottingham, UK), Keiichi Nishimura (Tokyo Gakugei University, Japan), Chris
Haines (City University, London), Dani Ben-Zvi (University of Haifa, Israel), and
others. Helena spent 3 weeks in Hamburg as a guest of Gabriele Kaiser resulting
from an exchange program between the two universities (Fig. 1.5).
Helena received 16 grants for research and projects – approximately ZAR1.89 mil-
lion - in the 10 years she worked at Stellenbosch University. The projects went to
the classroom and from there to the research agenda. She wanted to use every
opportunity for research. In this way, she could work more with national and inter-
national colleagues. Helena and Sonja Strydom, a senior advisor at the Centre for
Learning Technologies, were working on a strong project to investigate the integra-
tion of learning technologies into Foundation Phase Mathematics teacher training.
The article based on that work will appear in 2020. Sonja described Helena: “I
always say there is a handful of colleagues in Higher Education who have made a
lasting impression on me because of their expertise, humility, enthusiasm and the
way they care for others. This is often more non-verbal – their way of communicat-
ing with students and colleagues, as well as the willingness to learn and grow.
Helena was one of those people.”
12 D. Wessels
Fig. 1.5 Helena as the guest of Gabriele Kaiser in Hamburg – exchange program. Left to right:
Gabriele Kaiser (University of Hamburg), Hilda Borko (Stanford University), and Helena Wessels
(Stellenbosch University)
1.6.2 Dedication from Around the World
From requests at national and international levels to commemorate Helena’s life and
academic work, it was decided to publish a special edition of South African Journal
of Childhood Education published by SARAECE on Helena’s work and research.
This will be published with Catherine Paolucci as guest editor. Lyn English from
Queensland University in Australia asks the following: “Helena was a wonderful
colleague and same with Dirk. I was looking forward to catching up with them at
the next ICTMA. What a great loss to our community. I wonder if it might be pos-
sible to have an award in her name for young researchers or HDR students or some-
thing like that?” Norman Webb commented further: “Helena contributed in so many
ways. The mathematics education field will really miss her. She was truly taking off
in getting grants, in her teaching, and being a significant member of the mathemat-
ics education community. In our last meeting, I was impressed by her vision for
improvement and all she was doing to attain that vision.”
1.7 Helena and Dirk: The Team 2
Her husband, Dirk, mentored and partnered with Helena, was always there for her
and supported her throughout her academic career. He had been a full professor of
Mathematics Education at the University of South Africa’s Faculty of Education for
some time. As they shared a study, there were many conversations and debates that
certainly helped shape both of their thinking. When he congratulated her on her
achievements and shared her pride with her, she always said: “If it wasn’t for you, it
would never have happened.” He responded, “I was just there to point and some-
times open the door – you walked the road yourself.” Dirk’s approach was to paint
the bigger picture and then unfold the thinking process between the two of them.
There was an ongoing critical awareness of quality, honesty, proper planning, and
thorough delivery. Her integrity ensured the consistency of implementing the right
principles during the execution of the plan or project was always to be in awe of.
This was followed by a thorough and honest assessment. It was the continuous fea-
ture of all the projects she was involved with. Dan Canada said of the relationship
between Dirk and Helena: “Personally, I think what touched me most was the bond
between Helena and Dirk that made a deep impression on me, and that memory
abides with me even after the passing of years. How blessed are we, whose lives
were touched by Helena!” A friend and schoolmate of Dirk, Izak du Plessis, said
once: “You know, when I think of the two of you, I always think of you as a team.”
Piera Biccard testifies to the following of Helena: “In many ways, Helena and Dirk
became what I call ‘my academic parents.’ Although Prof Dirk was my supervisor,
Helena played an instrumental role in shaping my thinking like an academic. As my
academic mother, she played a very special, nurturing and supportive role. Often, in
my discussions with him, he would ask her to come to where we were working and
help us sort something out. She always listened and gave careful consideration to
the problem. Her response was always calm and well thought out. In these discus-
sions, I could see the deep respect they had for each other as academics. It is through
Helena that my new academic research project took focus. I don’t think she knew
how much I valued her opinions and I wish I had told her.”
In Memory
Helena, you will be missed!
References
Müller, H. M. (2001). Peer assessment of data handling (Grade 5’s) (poster). In M. Van den Heuvel-
Panhuizen (Ed.), Proceedings of PME 25 (Vol. 1, p. 409). Freudenthal Institute, University of
Utrecht, The Netherlands: PME.
Scott-Wilson, R., Wessels, D. C. J., Wessels, H., & Swart, E. (2017). Hidden benefits of modelling
for students with disabilities. In G. A. Stillman, W. Blum, & G. Kaiser (Eds.), Mathematical
modelling and applications: Crossing and researching boundaries in mathematics education
(pp. 455–465). Cham: Springer.
Wessels, H. (2011). Using a modelling task to elicit reasoning about data. In A. Rogerson &
L. Paditz (Eds.), The Mathematics Education into the 21st Century Project. Turning Dreams
into Reality: Transformations and Paradigm Shifts in Mathematics Education. Proceedings of
the 11th International Conference. Grahamstown: Rhodes University. http://directorymathsed.
net/download/
14 D. Wessels
Wessels, H. (2017). Exploring aspects of creativity in mathematical modelling. In G. A. Stillman,

W. Blum, & G. Kaiser (Eds.), Mathematical modelling and applications: Crossing and
researching boundaries in mathematics education (pp. 491–501). Cham: Springer.
Wessels, H. M. (2018). Noticing in pre-service teacher education: Research lessons as a context
for reflection on learners’ mathematical reasoning and sense-making. In G. Kaiser, H. Forgasz,
M. Graven, A. Kuzniak, E. Simmt, & B. Xu (Eds.), Invited lectures from the 13th international
congress on mathematical education (ICME-13 monographs) (pp. 731–748). Cham: Springer.
Chapter 2
Sense-Making in Mathematical Modelling
and Applications Educational Research
and Practice
Gloria Ann Stillman, Gabriele Kaiser, and Christine Erna Lampen
Abstract This latest contribution from members of the ICTMA research commu-
nity focusses on how students, teachers, tertiary educators, and researchers make
sense of mathematical modelling and applications educational research and prac-
tice. Innovative approaches in modelling educational research and research into,
and evaluation of, teaching practice are showcased; pedagogical issues, assessment,
and applicability at different levels of education relating to modelling and applica-
tions are investigated; and examples of modelling and applications in educational
practice are provided.
Keywords Assessment · Mathematical modelling · Modelling educational

research · Pedagogical issues · Real-world contexts · Teaching practice ·
Sense-making
2.1 Introduction
According to the Concise Oxford English Dictionary (Stevenson and Waite 2011,
p. 1311), to make sense of something we “find meaning or coherence in” it. The
English phrase seems to have its origins in Middle English (c.1150–c.1470)
deriving from the Latin “sensus, the faculty of feeling, thought or meaning”.
G. A. Stillman (*)
School of Education Victoria, Catherine of Siena Centre, Australian Catholic University,
Ballarat, VIC, Australia
e-mail: gloria.stillman@acu.edu.au
G. Kaiser
Faculty of Education, University of Hamburg, Hamburg, Germany
C. E. Lampen
Faculty of Education, Stellenbosch University, Stellenbosch, South Africa

16 G. A. Stillman et al.
However, as Mercer (2013) points out, “the words we hear or read do not simply
activate a mental dictionary, they generate dialogic responses in our own minds as
we use our existing knowledge to make sense of them” (p. 502). Lev Vygotsky
(1934/1986), paraphrasing Paulhan (1928), does most justice to the nature of the
term as used in educational contexts where English is the dominant language, when
he writes that “the sense of a word … is the sum of all the psychological events
aroused in our consciousness by the word. It is a dynamic, fluid, complex whole
which has several zones of unequal stability”. On the other hand, “meaning is only
one of the zones of sense, the most stable and precise zone”. He continues:
A word acquires its sense from the context in which it appears; in different contexts, it
changes its sense. Meaning remains stable throughout the changes of sense. The dictionary
meaning of a word is no more than a stone in the edifice of sense, no more than a potential-
ity that finds diversified realization in speech. (p. 245)
Thus, the notion of sense-making involves more than dictionary meanings of what
is being said or read, or what is happening to us or around us even in situations such
as engaging in mathematical modelling or finding the solution of a mathematical
application.
Sense-making theories have arisen in several fields such as human-computer
interaction, the theory of organisations, communications and education from the
1970s onwards being of a similar time period when interest in real-world modelling
and applications began to gain traction in mathematics education and research (Niss
et al. 2007). Recent uses of the term sense-making in mathematics education are
often associated with National Council of Teachers of Mathematics (NCTM) pro-
nouncements in the USA where sense-making is defined as “developing under-
standing of a situation, context, or concept by connecting it with existing knowledge”
(NCTM 2009, p. 1). The pervasive nature of sense-making in the mathematics
classroom is captured by Kaezer and Menon (2015/6) who claim that “reasoning
and sense making are thinking habits that should be inherent in all mathematical
activity” (p. 344). Sense-making either as extracting sense from a real-world situa-
tion to be modelled or giving sense to a real-world context that might be described
in text in a real-world modelling or application task presented to students would
appear to be pivotal to be engaged in by modellers in educational contexts and thus
the focus of research.
This volume, which includes a selection of chapters arising from presentations
by their authors at the 18th International Conference on the Teaching of Mathematical
Modelling and Applications (ICTM18) in Cape Town, South Africa, in 2017,
addresses the theme of Sense-Making in Mathematical Modelling and Applications
Educational Research and Practice. Six parts follow where innovative approaches
in modelling educational research and research into and evaluation of teaching prac-
tice are showcased, and pedagogical issues, assessment, applicability at different
levels of education and examples of modelling and applications in educational prac-
tice are investigated.
2 Sense-Making in Mathematical Modelling and Applications Educational Research… 17
2.2 I nnovative Approaches in Modelling Educational

Research and Teaching
New approaches arise in teaching and modelling educational research as both teach-
ers and researchers try to make sense of what occurs when mathematical modelling
or solving mathematical applications is the focus of teaching and learning both in
and outside the classroom. The collective chapters in the first part of this book are
indicative of the broad spectrum of innovation this brings in both research and
teaching.
The theoretical concept of science capital is introduced by Vos, Hernandez-
Martinez, and Frejd in their chapter. Science capital is a set of resources that offer
people advantages within scientific contexts. The underlying sociological theory of
Bourdieu (1986) explains how inequities are caused by some people having better
access to science capital than others. Science capital is used as an analytical device
in the studies described in the ensuing three chapters, connecting mathematical
modelling education to science capital.
The first of these chapters is by Frejd who explores mathematics teachers’ pos-
sibilities to accumulate science capital, with the aim of identifying teachers’ poten-
tials and constraints regarding their use of realistic modelling activities within
teaching. Analysis of three teacher interviews showed scientific preferences and
intentions for establishing a teaching practice connected to the world beyond school
were high, but obstacles to progress were lack of financial resources and social net-
works of people working with science. In the following chapter, Hernandez-
Martinez analyses two second-year undergraduate engineering students’ narratives
of engagement with a mathematics course aimed at developing mathematical mod-
elling competences. Both students came to university with enough science capital to
expect they would be successful in their mathematics courses and were confident in
their mathematical knowledge and skills, but their narratives revealed the different
value that they attributed to mathematics. This was reflected in their engagement
with the mathematical modelling aspects of the course. Bourdieu’s theory of prac-
tice (1972) was used to understand the complexities of social practices in higher
education and to explain the different positions that these two students took regard-
ing mathematical modelling. Finally, Vos explores students’ opportunities to learn
informally about mathematical modelling (e.g. by visiting museums). In addition to
science capital, Vos’s study drew on the notion of public understanding of mathe-
matical modelling and explored how these are mediated in two science museums.
The research method entailed observations of displays, artefacts, and visitors. One
science museum completely detached mathematics from its use value, while the
other offered histories of how people used mathematics to solve societal problems.
Vos then makes recommendations for design and research of environments for
informal learning about mathematical modelling.
Brady, McLean, Jung, Glancy, and Domínguez mine the complexity of student
modelling in classrooms by taking an innovative approach to analysis which
coordinates a diversity of research perspectives to make sense of what was happen-

ing as students discussed and worked on a model-eliciting activity (Doerr and Lesh
2011). They present a collaborative, post hoc analysis of an episode in the work of
a group of USA middle-school students as they engaged in a model-eliciting activ-
ity that the researchers regarded as pivotal to the modelling. Central features of this
collaborative analysis were the identification and elaboration of three perspectives:
representations, mathematical uncertainty, and idea development. These features
resulted from iterative interpretations of video data by research team members.
Coordination of these perspectives highlighted a cyclical pattern in the students’
progression through the modelling activity, as they encountered and managed math-
ematical uncertainty and expressed their developing ideas with physical materials
and representations. This work is viewed by the research team as a researcher-level
modelling effort, to synthesize independent analytical approaches, develop shared
perspectives, and construct an explanatory account of the flow of students’ collec-
tive thinking.
Mapping shifting discourses in classroom talk through indexical discourse anal-
ysis is the focus for Lewis and Manouchehri. They outline how indexical discourse
analysis coupled with an ethnomodelling framework (Rosa and Orey 2015) can be
used to analyse mathematical modelling interactions by making visible the interre-
lationship between the ideologies of the real-world and mathematics. They illustrate
this position by sharing classroom-based data of students’ ideas whilst considering
business profitability problems. They investigated how focusing on indexical terms
within classroom talk aid a researcher in detecting subtle shifts in discourse that
represent different discursive worlds, namely, the world of mathematics and the
world of work (cf. Psycharis and Potari 2017). In this context, the teacher’s scaf-
folds shifted students’ work from solving an application task into a mathematical
modelling inquiry.
The relevance of including an affective perspective in mathematical modelling
educational research is argued by Rogovchenko, Viirman, and Treffert-Thomas in their
chapter dealing with the joy of modelling. Schukajlow et al. (2018) have previously
pointed out the paucity of research on affective aspects in modelling research.
Rogovchenko et al. emphasise the importance of an affective perspective for the
teaching and learning of mathematical modelling at all levels, but especially at univer-
sity level. A survey of mathematics lecturers’ views on mathematical modelling
(Treffert-Thomas et al. 2017), several follow-up interviews, and a review of literature
on mathematical modelling relating to enjoyment, pleasure, and appreciation support
this conclusion. Findings indicated that many practitioners hold strong views on the
importance of enjoyment in doing and teaching mathematical modelling.
The main objective of the research reported by Rosa, Orey, and Cortes was to
show how dialogical approaches of ethnomodelling contribute to the resignification
of function concepts. This investigation was developed in collaboration with a pub-
lic school and a local farmer’s market in Brazil. The participants were 38 high
school students who interacted with a farmer-vendor at the market and observed his
practices. Data were collected and analysed through a mixed methods approach.
Findings confirm ethnomodelling provides an integrative approach to school
mathematics when both etic and emic mathematical knowledge are considered in a
dialogical manner (Rosa and Orey 2017).
2.3 esearch into, or Evaluation of, Teaching Practice

R
in Mathematical Modelling Education
Researching teaching practice in mathematical modelling education and evaluating

such teaching are impetuses for deepening our research base and establishing
research evidence for bringing new ideas into future classroom practice or to show
where there are current gaps.
Abel, Searcy, and Salinas report how Weick’s (2001) sense-making methodology
was used by an educational collective – secondary and community college teachers,
mathematicians, and mathematics educators – in the USA to come to terms with
engagement with mathematical modelling within the classroom. The participants’
perspectives and discussion points illuminated aspects of the mathematical model-
ling process that are difficult to reconcile, such as dealing with ambiguity and
uncertainty and distinctions between interpretation and computation. The group
produced a framework for faculty practice that described characteristics of mathe-
matical modelling tasks and student work at three different levels within four stages
(formulate, compute, interpret, validate) based on the Common Core State Standards
for Mathematics (National Governors Association Center for Best Practices &
Council of Chief State School Officers 2010) representation of a mathematical
modelling cycle.
Beckschulte reports on an intervention study that investigated the influence of
modelling with a five-step solution plan on the development of student modelling
competencies. The study used a pre-post-follow-up design with 29 German second-
ary classes. Evaluation of the study involved using Item Response Theory with an
instrument focussing on assessing modelling subcompetencies. The solution plan
had an impact on the development of students’ competencies in interpreting math-
ematical results in real situations, replicating the results of Schukajlow et al. (2015).
However, the use of the solution plan did not influence the development of compe-
tencies in simplifying, mathematising, and validating. Post hoc analysis revealed a
statistical trend of a more sustainable development by students who worked with the
solution plan compared to those who did not for the subcompetencies simplifying
and interpreting. Beckschulte was thus encouraged to suggest that this strategic
instrument could lead to developing long-term competencies over time.
An upper secondary teacher trying to transform her teaching by designing and
implementing a model development sequence (Lesh et al. 2003) on statistics in a
Swedish year 11 mathematics class is the focus of the study by Ärlebäck. Cultural
Historical Activity Theory (Engeström 1987) is used to analyse lesson plans, reflec-
tion notes, recordings of planning meetings, and student data, to identify tensions
and challenges arising as the teacher tried to adopt this new modelling perspective
on teaching and learning mathematics. The results illustrated the need to renegotiate
appropriate classroom norms and facilitate shifts in power, ownership, and respon-
sibilities in the mathematics classroom if a modelling approach is to effectively
impact and change the teaching and learning of mathematics using mathematical
modelling.
Kawakami, Komeda, and Saeki illustrate the extent to which year 6 Japanese
students identified a straight-line time–distance graph as a mathematical model of
average speed when making sense of walking situations using graphing calculators
and motion detectors. The results demonstrated the types of models that students
identified explicitly, the phases of identifying such mathematical models, and essen-
tial actions needed for this identification. The approach of identifying mathematical
models, through sense-making of embodied actions and technologies, can highlight
students’ gradual mathematisation differentiating the phenomenal, virtual techno-
logical, and mathematical conceptual worlds.
Klock and Siller present an observation study which investigated the solution
processes of two groups of German students at a technical high school (18–20-year-
olds) working on a modelling task. While solving the task, the students’ key activity
was the analysis of a mathematical model given in the form of a simulation. The
study focused on difficulties in the modelling process, especially on the intensity of
difficulties which were identified by video recording and time taken. The aim was
to identify the most persistent difficulties in the solution process. For that purpose,
a method consisting of a time-based measurement was used. A coding scheme for
identifying difficulties in the modelling process based on previous studies was cre-
ated and used for the evaluation of the solution processes. Measuring the time stu-
dents spent overcoming their difficulties enabled judgements about the intensities of
these difficulties.
Komeda, Kawakami, Kaneko, and Yamaguchi explored the potential for mathe-
matical modelling to be used as a didactical means for supporting student learning
of speed and related mathematical concepts. This study explored modelling for
learning in the secondary grades to deepen and expand mathematical models of
speed in relation to walking by using motion detectors and graphing calculators, as
a part of the project Modelling a walk. Year 8 Japanese students’ activities in teach-
ing experiments were analysed. Students were able to find a way of thinking leading
to instantaneous speed in deepening and expanding the mathematical model of
speed, while associating this with movement, graphing, and computation.
The essential role of the teacher in a teaching/learning environment where the
focus is on real-life situations and the complex nature of teaching practice in such a
setting are highlighted in the work reported by Manouchehri, Bekdemir and Yao.
Ten hours of whole group discussions in a USA Grade 5 classroom were analysed
to identify teacher interventions aimed at facilitating student work on modelling
tasks and identifying the type of mathematical actions those interventions lead to. In
navigating such challenging situations, two types of teacher interventions were foun-
dational in advancing students’ work: (1) presenting a follow-up question that
demanded greater attention to precision and (2) asking for connections among inter-
pretations, ideas, and representations. Teacher interventions impacted students’
reasoning, structuring, and mathematizing processes; however, the teacher’s sug-

gestions for scrutinizing/validating a particular solution or consolidating conflicting
arguments were not always taken up by students as tensions perceived by the teacher
were not always shared by students.
Orey, Rosa and Soares apply mathematical modelling as a learning environment
to cultivate students’ own interest in a leisure activity for educational purposes. A
Brazilian street activity, carrinhos de rolimã (roller carts), is the stimulus for stu-
dents to discuss and interrogate mathematical ideas used in the construction of the
carts in a Youth and Adult Education night school program. Teaching practice in this
environment used mathematical modelling as an opportunity to deepen students’
understanding and broaden their critical reflections on this everyday activity in their
own context and community. Orey et al. conclude that when systems are borrowed
from reality, students begin to study and analyse them symbolically, systematically,
analytically, and critically in order to make informed decisions based less on emo-
tion and more on data.
Competency gaps and opportunities to acquire such competencies have been a
contentious issue in several fields for many years but particularly engineering
(Alpers 2017). In their chapter, de Villiers and Wessels explain how mathematical
modelling and engineering technician competencies can be developed concurrently
using modelling tasks. Twelve first year civil engineering technician students study-
ing at a South African University of Technology participated in a one semester,
design-based, research study. Results from use of a sequence of six modelling tasks
showed slow but consistent progress in both engineering technician and mathemati-
cal modelling competence development.
2.4 edagogical Issues for Teachers and Teacher Educators

P
Using Mathematical Modelling and Applications
Pedagogical issues for teachers and teacher educators using mathematical model-
ling and applications in their teaching are always areas of engagement and study for
the research community. The first two chapters in this part of the book arose from
two of the plenaries that were presented in Cape Town.
Biccard’s chapter leads, drawing on scholarly conceptions of sense-making from
domains outside of mathematics education, to outline ways that mathematical mod-
elling tasks may meet some of the criteria underpinning these different conceptions.
Myths around sense-making from human and machine cognition (Klein et al. 2006)
as well as Weick’s (2001) seven elements of sense-making in the context of organi-
zation science are presented to extend the discussion to sense-making within math-
ematical modelling. An analogy is drawn between ideas around sense-making from
these domains and sense-making in mathematical modelling activities. Although
the discussions about sense-making from these diverse domains do not result in a
perfect fit with the processes of mathematical modelling, Biccard suggests the con-
cepts may extend our thinking about sense-making and provide us with vocabulary
to discuss sense-making in mathematical modelling more deeply.
The extra-mathematical domains that feature prominently for mathematical

modelling and the applications of mathematics in ICTMA deliberations are the
physical, biological, economic, environmental, social, and engineering sciences.
Julie’s chapter focuses on the deployment of mathematics in domains other than
these. Sources such as novels, stories, monographs, journalistic-like commentaries
on mathematics, and anecdotal accounts on mathematics by ordinary people were
mined to ascertain the uses mathematics is put to in such sources. These uses
included backgrounding storylines through encryption and decryption; depictions
of cleverness, dissent/avoidance, and madness; rendering political commentary; and
entertainment through humour. Julie suggests that students should be afforded
opportunities to put mathematics to use in a similar manner.
The International Mathematical Modelling Challenge (IM2C) adds a new incen-
tive for applications of mathematics to problems beyond the classroom. In this
chapter, Galbraith, Holton, and Turner describe the design of support material to
serve two audiences: teachers intending to introduce modelling into curricula inde-
pendent of IM2 Challenge involvement and for those entering teams in the Challenge.
The identification and application of design principles for these two audiences had
to consider commonalities but also serve their differing interests. This involved
addressing issues that continue to engage the international community, such as the
purpose of modelling cycles, competencies, and mentoring of student modellers.
Mathematics trails are a popular way to implement outdoor education outside
mathematics classrooms. The MathCityMap project combines the idea of maths
trail with the possibilities of modern mobile devices. In a study by Gurjanow and
Ludwig, year 9 students completed a maths trail with the MathCityMap application,
enabling the researchers to observe their difficulties with measuring and handling
quantities. Analyses of student notes and calculations revealed that over one third
had poor measurement sense. As these competencies are prerequisites to success-
fully completing a maths trail as well as for modelling (Hagena 2015), this issue
needs to be addressed prior to engaging students in such activities.
A difficult step in the modelling process is to validate results, models, or assump-
tions. Using Dynamic Geometry Software (DGS) such as GeoGebra to do model-
ling could foster students’ modelling subcompetencies. The test instrument
development to evaluate doing this is described elsewhere (Hankeln, Adamak and
Greefrath 2019). Hankeln presents results from a quantitative study in which the
development of validating in 15 classes working with GeoGebra throughout a mod-
elling project was compared to development in another 15 classes who worked only
with paper and pencil. Results showed both groups significantly and lastingly
improved their validating competencies, but there was no significant difference
between the results of the two groups when they were tested using tasks without
digital tools. This study complements other work using DGS and mathematising
(Greefrath et al. 2018).
Ikeda and Stephens investigated whether mathematical modelling, interpreted as
a series of interactive translations, can also be effective to assist students to con-
struct mathematical knowledge. They describe a series of possible teaching activi-
ties in which the theorem of a limit of a trigonometric function can be approached
through solving a real-world problem involving ways to determine the shortest air
route between two cities. They establish that these transformations have the poten-
tial to build up mathematical concepts. These findings are intended to form a frame-
work for subsequent experimental teaching.
The aim of the chapter by Kotze is to explore the potentialities of mathematical
modelling tasks that underscore South African first year biomedical technology stu-
dents’ habits and habitus. When students were engaged in modelling tasks for the
first time, ingrained habits – imposed by traditional teaching and learning
approaches – were challenged affecting students’ future expectations and aspira-
tions, that is, Bourdieu’s (1990) habitus. Three questionnaires were employed to
analyse students’ perceptions of four modelling tasks, spread over a semester. The
novel modelling environment was an impetus for change; however, modification of
habits and a refinement of habitus were marginally demonstrated.
Metacognitive competency can be described as a subcompetence of mathemati-
cal modelling competence. A study by Krüger, Vorhölter, and Kaiser aimed at
investigating students’ perception of metacognition in modelling processes.
Fourteen student groups were interviewed at the beginning and at the end of a teach-
ing unit following their teachers’ attending related professional learning. One group
is the focus for the chapter. By the end of the study, this group used cognitive strate-
gies and skills extensively showing a sense of direction in the formulation phase of
the modelling process as an indicator of metacognitive strategies.
It is often argued that working on modelling problems in school increases stu-
dents’ motivation to learn mathematics as they see how often mathematics is used
in society and everyday life. Vorhölter and Schwarz present a theoretical framework
for analysing students’ personal meaningfulness of mathematics with regard to fac-
tors influencing this and consequences. They present the results of a qualitative
study with 15 German year 10 students that showed which characteristics of model-
ling problems and of working with modelling problems in class had a direct or an
indirect influence on these students’ construction of personal meaningfulness.
The usage of metacognitive strategies can help students avoid or overcome cog-
nitive barriers during modelling processes according to Wendt, Vorhölter, and
Kaiser. Teachers need to act on a meta-meta level and monitor their students’ usage
of metacognitive strategies to impart relevant metacognitive strategies and support
students during the modelling process. Therefore, the teachers’ perception of stu-
dents’ metacognitive activities during modelling processes is crucial. A qualita-
tively oriented intervention study was used by Wendt et al. to examine this topic.
The results reported focus on one teacher, who showed a broadening of her percep-
tion of the students’ usage of metacognition in mathematical modelling processes.
2.5 Assessment of Mathematical Modelling in Schools
Modes of assessment in mathematical modelling fall into two categories: (1) a

holistic approach, where the modeller works (individually or in groups) on a com-
plete modelling problem and is assessed on this as a whole, and (2) an atomistic
approach, where the modeller is asked to demonstrate separate competencies (e.g.,

validating) needed for only part of the modelling cycle.
Alagoz and Ekici aimed to evaluate the assessment of mathematical modelling
holistically where its multidimensional nature was retained but by using a noncom-
pensatory approach where a low proficiency in one dimension is not compensated
for by high proficiency in another. They used a cognitive diagnostic model which
allowed them to show that a multidimensional noncompensatory psychometric
model for scoring mathematical modelling performance provided a statistical model
that fits to the assumptions of the mathematical modelling assessment. Furthermore,
this approach enabled them to demonstrate that distinct attribute mastery profiles
for learners (in this case teachers in a professional development on modelling and
assessment) could be identified and fine grained individualized feedback provided.
Bali, Julie, and Mbekwa argue that, as mathematical modelling competency has
a direct relationship with Mathematical Literacy, the development of mathematical
modelling competencies should be accorded priority in Mathematical Literacy in
South Africa. Furthermore, they demonstrate a strong relationship between argu-
ments for incorporating mathematical modelling and mathematical applications
into everyday mathematics classrooms and reasons for the introduction of
Mathematical Literacy into the South African school curriculum. Schemes of mark-
ing of five consecutive, nationally set, end of year 12 examinations for Mathematical
Literacy were analysed to gauge the level of prioritisation of assessment of mathe-
matical modelling competencies in these examinations. Although producing a
mathematical solution from a mathematical model gained high priority and there
was some evidence of mathematisation, there was no formulation of a problem
statement nor validation of a model in any of the marking schemes. Bali, Julie, and
Mbekwa then point to the implications of this misalignment between competencies
and the aims of Mathematical Literacy for the actual competencies assessed in
examinations.
Brown reports a document analysis of key curriculum documents in two states of
Australia to assess level of challenge expected in modelling tasks in the intended
curriculum. Mathematical modelling tasks are part of the assessment in the final
year of schooling in both states. Analysis found that information provided to teach-
ers in Queensland was broader and more substantive including with respect to sam-
ple assessment instruments and their implementation and assessment. A framework
of complexity used to analyse sample tasks found the indicative levels of expecta-
tions with respect to degree of challenge were lower in Victoria than in Queensland.
Brown concurs with Blum (2015) that the cognitive level of tasks in assessment
mediates student performance.
Tasks in central examinations are also the subject of the analysis conducted by
Siller and Greefrath who investigated the modelling character of the written school
leaving central examinations in Austria. The criteria used were authenticity (factual
and of the mathematics used), relevance to current lives of students, openness of
solution approach, and subcompetencies of modelling targeted. In addition, they
examined the relationship between modelling character and success rate as an indi-
cator of the difficulty of these tasks. They found that the tasks were related to reality
in the main but were not very authentic for the mathematics used, closed, and not
immediately relevant to the students’ lives. From a modelling subcompetencies per-
spective other than working mathematically, only a few examples of interpreting
and mathematising were present, and there were no examples of simplifying and
validating. With respect to interpreting, there was no direct relationship between
modelling character and task difficulty. These results echo previous similar findings
(e.g. Julie 2015; Vos 2013) in other countries.
2.6 pplicability at Different Levels of Schooling

A
and University
The demonstration of applicability of teaching through modelling and applications

at different levels of schooling and at the tertiary level is an on-going interest of
many researchers.
Blomhøj investigated modelling competency in problem-oriented project work in
the interdisciplinary natural science bachelor program at Roskilde University,
Denmark. His analyses focused on student project reports in the area of infectious
diseases modelled by the standard SIR (Susceptible-Infectious-Recovered) com-
partment model and extensions of this. Student modelling competency was anal-
ysed with respect to the dimensions: degree of coverage of the modelling process
and the technical level of the modelling and use of mathematics. Students’ reflec-
tions related to the particular modelling subprocesses and reflections concerning the
role and function of the model and its results in a context of application were exam-
ined. A framework including the two dimensions and the two types of reflections is
suggested as a way to characterise progress in students’ mathematical modelling
competency in similar problem-oriented project work.
At the Royaumont Seminar in 1959, the New Math reform was officially
launched. In the decade following, many mathematics educators were involved in
actions to facilitate the implementation of this New Math reform. A deep knowledge
and understanding of the structures of modern mathematics was seen as a prerequi-
site to substantial applications (Niss et al. 2007), but in actual classroom practices,
the applied side of mathematics was often neglected. Even in Royaumont, there
were alternative voices arguing for taking the role of applications seriously. De Bock
and Zwaneveld investigate these arguments for integrating applications in mathe-
matics education, as well as the kind of applications envisaged, at the Royaumont
Seminar and in the next decade.
The design of a particular kind of interactive case practice teaching in a tertiary
mathematical modelling course at a Chinese University is outlined by Duan, Wang,
and Wu. The process of interactive case creation and problem solving between stu-
dents and teacher aims at promoting students’ creativity, initiative, and interest in
learning, as well as efficient teaching. The stages of the interactive case practice
teaching process are interactive case creation, subject activity, open learning, and
feedback and evaluation. The students in these stages experience the whole teaching
process including problem creation, problem analysis, problem solving, and knowl-
edge application. In order to demonstrate the interactive case teaching process in
practice, the case of the sensitive question survey is presented.
Modelling and the teaching of modelling are often difficult for teachers, but both
can be learnt through the development of mathematical, modelling, and facilitator
competencies of preservice teachers during their formal education (Borromeo Ferri
and Blum 2010). Durandt and Lautenbach explored how South African preservice
teachers translated the modelling cycle and developed particular competencies over
time by undertaking a series of modelling activities as modellers themselves and as
teachers of modelling. Qualitative findings not only revealed how these preservice
teachers improved in their sense-making of the mathematical modelling process but
also uncovered some shortcomings related to preparation for teaching modelling.
A qualitative case study by Govender investigated the extent to which South
African preservice mathematics teachers were competent in invoking and using
modelling processes to solve a real practical problem: finding the height of a tree.
Analysis of the written reports and presentations of student groups showed how they
were not only able to experience the modelling process but also creatively invoked
and applied necessary skills and abilities to enable them to move from one step to
the next. In particular, they were able to make assumptions, set up a mathematical
model through mathematizing, use their mathematical knowledge to solve the
model, interpret and critically reflect on their solutions to see they made sense.
According to Govender, the kinds of novel models built and used by two groups in
creating nonstandard solutions suggested the tree modelling task stimulated creative
and flexible thinking.
Lantau, Bracke, Bock, and Capraro consider how to successfully design a teacher
training focusing on the implementation of interdisciplinary STEM modelling proj-
ects based on a Segway. To do this, some deeper knowledge about physical, techni-
cal, and mathematical concepts is needed. Despite this, four interdisciplinary STEM
modelling projects in which grade 12 students gained insight into the control of a
Segway have been implemented following 2-day teacher trainings. The design of
the teacher training as well as the resulting modelling projects are described. The
aim was to identify successful factors of the teacher training.
Mhakure provides a theoretical perspective on how the South African
Mathematical Literacy curriculum differs from the definition of Mathematical
Literacy given by the Organisation for Economic Cooperation and Development
(2013). Through unpacking this definition, and the South African Mathematical
Literacy curriculum definition, he demonstrates with examples that mathematical
modelling is the missing link in the South African Mathematical Literacy curricu-
lum context. Furthermore, he argues that this difference is significant with respect
to students’ acquisition of mathematical skills and competencies needed in the
workplace and for active citizenship. He also argues that the teaching of Mathematical
Literacy should focus on the development of key mathematical and statistical skills.
Spooner investigated the student learning goals held by a university first year
lecturer in a mathematical modelling course. Hypothetical learning trajectories
were adopted as a framework for the study. Data were collected from interviews and
teaching observations. Preliminary findings showed the lecturer had clear student
learning goals in place for teaching mathematical modelling to solve real-world
problems. These included allowing students to experience modelling techniques
and processes concurrently and providing opportunities for students to model. The
centrality of the lecturer in students learning to model is discussed.
2.7 Examples of Modelling and Applications in Practice
Finally, in this last part examples of modelling and applications in educational prac-
tice are demonstrated firstly in the context of sailing and secondly in terms of the
prediction by Thomas Malthus in 1798 that hopes for continued human social hap-
piness must be in vain, as population will always tend to exceed the growth of pro-
duction of food (Malthus 1798).
Mathematical modelling is under-investigated in complex learning environments
where embodied knowledge such as sailing is a natural and essential part of learn-
ing. Ekici and Alagoz examined how the participants, young islander sailors, used
mathematical modelling as an ongoing process in building and testing models to
make sense of their sailing practices. They demonstrate how a culturally responsive
modelling practice (Ladson-Billings 1995) can build on learners’ experiences, their
knowledge, and existing practices as young sailors. Participants engaged in experi-
mental and active mathematical modelling of sailing in its own lifeworld in this
integrated STEM learning context. Participants’ living/lived embodied knowledge
of sailing was used in the modelling of tacking to travel towards the wind. Ekici and
Alagoz demonstrate and discuss how an embodied phenomenological approach
works as learners engage in modelling phases from understanding the tacking prob-
lem to validating their models.
Analysing personal, local, national, and global nonlinear dynamic feedback sys-
tems is central to addressing many current complex problems. Fisher worked with
a relatively broad ability cross-section of secondary mathematics students in both
algebra classes and year-long modelling classes for 20 years in which students built
system dynamics (stock/flow) models to study such complex problems, at an intro-
ductory level. Fisher provides an example of one such algebra lesson, the Malthus
problem, providing a partial modelling sequence, identification of the software used
by the students, and questions posed to students.
2.8 Conclusion
As with other books in the ICTMA series, this volume addresses many of the issues
and associated research questions that continue to engage the international commu-
nity including development and assessment of modelling competence and compe-
tencies in both technology-free and digital technology environments; student,
teacher (preservice and in-service), teacher educator, and tertiary lecturer construc-
tion of personal meaningfulness and sense-making through mathematical model-
ling; affect; and the use of culturally inclusive practices such as ethnomodelling and
culturally relevant pedagogy. In addition, new constructs such as science capital and
public understanding of mathematical modelling and new techniques or approaches
to research such as a collaborative analysis incorporating different research perspec-
tives from a cross-institutional collaborator network, indexical discourse analysis,
schemes for analysing the modelling character of examination papers, sense-making
methodology, and an embodied phenomenological approach are incorporated.
South African researchers and educators are also given a significant voice among
the chapters here.
References
Alpers, B. (2017). The mathematical modelling competencies required for solving engineering
statics assignments. In G.A. Stillman, W. Blum, & G. Kaiser (Eds.), Mathematical model-
ling and applications: Crossing and researching boundaries in mathematics education
Blum, W. (2015). Quality teaching of mathematical modelling: What do we know, what can we
do? In S. J. Cho (Ed.), Proceedings of ICME 12 (pp. 73–96). Cham: Springer Open.
Borromeo Ferri, R., & Blum, W. (2010). Mathematical modelling in teacher education-experiences
from a modelling seminar. In V. Durand-Guerrier, S. Soury-Lavergne, & F. Arzarello (Eds.),
Proceedings of CERME 6 (pp. 2046–2055). Lyon: Institut national de recherche pédagogique.
Bourdieu, P. (1972). Outline of a theory of practice. Cambridge: Cambridge University Press.
Bourdieu, P. (1986). The forms of capital. In J. Richardson (Ed.), Handbook of theory and research
for the sociology of education (pp. 241–258). Westport: Greenwood.
Bourdieu, P. (1990). In other words, essays towards a reflexive sociology. Stanford: Stanford
University Press.
Doerr, H. M., & Lesh, R. (2011). Models and modelling perspectives on teaching and learn-
ing mathematics in the twenty-first century. In G. Kaiser, W. Blum, R. Borromeo Ferri, &
G. Stillman (Eds.), Trends in teaching and learning of mathematical modelling (pp. 247–268).
Dordrecht: Springer.
Engeström, Y. (1987). Learning by expanding: An activity-theoretical approach to developmental
research. Orienta-Konsultit: Helsinki.
Greefrath, G., Hertleif, C., & Siller, H.-S. (2018). Mathematical modelling with digital tools—
A quantitative study on mathematising with dynamic geometry software. ZDM Mathematics
Education, 50(1–2), 233–244.
Hagena, M. (2015). Improving mathematical modelling by fostering measurement sense: An
intervention study with pre-service mathematics teachers. In G. A. Stillman, W. Blum, &
M. S. Biembengut (Eds.), Mathematical modelling in education research and practice:
Cultural, social and cognitive influences (pp. 185–194). Cham: Springer.
Hankeln, C., Adamak, G., & Greefrath, G. (2019). Assessing sub-competencies of mathematical
modelling – Development of a new test instrument. In G. A. Stillman & J. P. Brwon (Eds.),
Lines of inquiry in mathematical modelling research in education, ICME-13 monographs
(pp. 143–160). Cham: Springer Open.
Julie, C. (2015). Learners’ dealing with a financial application-like problem in a high-stakes
school-leaving mathematics examination. In G. A. Stillman, W. Blum, & M. S. Biembengut
(Eds.), Mathematical modelling in education research and practice: Cultural, social and cog-
nitive influences (pp. 477–486). Cham: Springer.
Keazer, L. M., & Menon, R. S. (2015/6)). Reasoning and sense making begins with the teacher.
Mathematics Teacher, 109(5), 342–349.
Klein, G., Moon, B., & Hoffer, R. R. (2006). Making sense of sensemaking 1: Alternative perspec-
tives. IEEE Intelligent Systems, 21(4), 70–73.
Ladson-Billings, G. (1995). Toward a theory of culturally relevant pedagogy. American Research
Journal, 32(3), 465–491.
Lesh, R. A., Cramer, K., Doerr, H. M., Post, T., & Zawojewski, J. S. (2003). Model development
sequences. In R. A. Lesh & H. M. Doerr (Eds.), Beyond constructivism (pp. 35–58). Mahwah:
Erlbaum.
Malthus, T. (1798). An essay on the principle of population as it affects the future improvement of society,
with remarks on the speculations of Mr. Godwin, M. Condorcet, and other writers. Available from
https://oll.libertyfund.org/titles/malthus-an-essay-on-the-principle-of-population-1798-1st-ed
Mercer, N. (2013). The social brain, language, and goal-directed collective thinking: A social con-
ception of cognition and its implications for understanding how we think, teach, and learn.
Educational Psychologist, 48(3), 148–168.
National Governors Association Center for Best Practices & Council of Chief State School
Officers. (2010). Common core state standards for mathematics. Washington, DC: Author.
NCTM. (2009). Executive summary. In Focus in high school mathematics: Reasoning and sense
making. Reston: National Council of Teachers of Mathematics.
Niss, M., Blum, W., & Galbraith, P. (2007). Introduction. In W. Blum, P. L. Galbraith, H.-W. Henn,
& M. Niss (Eds.), Modelling and applications in mathematics education: The 14th ICMI study
(pp. 3–32). New York: Springer.
Organisation for Economic Co-operation and Development (OECD). (2013). PISA 2012 assess-
ment and analytical framework: Mathematics, reading, science, problem solving and financial
literacy. Paris: OECD Publishing. https://doi.org/10.1787/9789264190511-en.
Paulhan, F. (1928). Qu’est-ce le sens de mots? Journal de Psychologie, 25, 289–329.
Psycharis, G., & Potari, D. (2017). Mathematics teachers’ learning at the boundaries of teaching
and workplace. In G. A. Stillman, W. Blum, & G. Kaiser (Eds.), Mathematical modelling and
applications: Crossing and researching boundaries in mathematics education (pp. 301–312).
Cham: Springer.
Rosa, M., & Orey, D. (2015). Ethnomodelling as a research lens on ethnomathematics and mod-
elling. In G. A. Stillman, G. Kaiser, W. Blum, & J. P. Brown (Eds.), Teaching mathematical
modelling: Connecting to research and practice (pp. 117–127). Dordrecht: Springer.
Rosa, M., & Orey, D. (2017). Ethnomodelling as the mathematization of cultural practices. In
G. A. Stillman, W. Blum, & G. Kaiser (Eds.), Mathematical modelling and applications:
Crossing and researching boundaries in mathematics education (pp. 153–163). Cham:
Springer.
Schukajlow, S., Kolter, J., & Blum, W. (2015). Scaffolding mathematical modelling with a solution
plan. ZDM Mathematics Education, 47(7), 1241–1254.
Schukajlow, S., Kaiser, G., & Stillman, G. (2018). Empirical research on teaching and learn-
ing of mathematical modelling: A survey on the current state-of-the-art. ZDM Mathematics
Education, 50(1–2), 5–18.
Stevenson, A., & Waite, M. (2011). Concise Oxford English dictionary (12th ed.). Oxford: Oxford
University Press.
Treffert-Thomas, S., Viirman, O., Hernandez-Martinez, P., & Rogovchenko, Y. (2017). Mathematics
lecturers’ views on the teaching of mathematical modelling. Nordic Studies in Mathematics
Education, 22(4), 121–145.
Vos, P. (2013). Assessment of modelling in mathematics examination papers: Ready-made mod-
els and reproductive mathematising. In G. A. Stillman, G. Kaiser, W. Blum, & J. P. Brown
(Eds.), Teaching mathematical modelling: Connecting to research and practice (pp. 479–488).
Vygotsky, L. (1934/1986). Thought and language. (A. Kozulin, Ed. & Trans.). Cambridge, MA:
The MIT Press.
Weick, K. E. (2001). Making sense of the organization. Maldon: Wiley-Blackwell.
Part I
Innovative Approaches in Modelling
Educational Research and Teaching
Chapter 3
Connections of Science Capital
and the Teaching and Learning
of Mathematical Modelling:
An Introduction
Pauline Vos, Paul Hernandez-Martinez, and Peter Frejd
Abstract This chapter is an introduction to the theoretical concept of science capi-

tal. It also serves as an introduction to the next three chapters of this book. These
ensuing chapters all connect mathematical modelling education to science capital.
In short, science capital is a set of resources that offer people advantages within
scientific contexts. For example, a friend who works in research can be a resource
to better understand the gist of science. Not all people have such friends, and the
underlying sociological theory of Bourdieu explains how inequities are caused by
some people having better access to science capital than others. In this chapter, we
explain how the concept of science capital can be used in research on mathematical
modelling education. This analytic use of science capital will be enacted in the stud-
ies described in the ensuing three chapters.
Keywords Bourdieu (theory of) · Science capital · Theoretical perspectives · Use

value (of mathematics) · Exchange value (of mathematics) · Socio-cultural
environment (of mathematical modelling)
P. Vos (*)
Department of Mathematics, University of Agder, Kristiansand, Norway
e-mail: pauline.vos@uia.no
P. Hernandez-Martinez
Department of Mathematics, Swinburne University of Technology, Hawthorn, VIC, Australia
e-mail: phernandezmartinez@swin.edu.au
P. Frejd
Department of Mathematics, Linköping University, Linköping, Sweden
e-mail: peter.frejd@liu.se

34 P. Vos et al.
3.1 Introduction
In the past decade, researchers of mathematical modelling education increasingly

explicate theories for their research. To name just a few, Barbosa and Pereira (2013)
used Basil Bernstein’s discourse theory for analysing arenas of conflict between
teachers and researchers in designing modelling tasks, Rosa and Orey (2013) used
Ubiratan d’Ambrosio’s theory of ethnomathematics to propose principles for ethno-
modelling in the classroom, and Ärlebäck and Frejd (2013) used Anna Sfard’s theory
of commognition to analyse students’ communication while working on modelling
tasks. In this chapter, we introduce a concept, which emerged from yet another the-
ory. First, we explain the rationale and background and thereafter elaborate on how
the concept can be useful in research of mathematical modelling education.
The overall aim to use this concept in our research is to shed more light on the
wider context of mathematical modelling education and what effect this wider con-
text has on the teaching and learning of mathematical modelling. We want to
advance research on mathematical modelling education beyond the study of class-
rooms, teacher training, examinations, or curriculum resources, all of which can be
linked to educational institutions (schools, universities, ministries, textbook pub-
lishers, etc.). Mathematical modelling is an activity, in which problems from real
life are solved using mathematical tools. In this activity resides a duality, namely,
that (1) school-based modelling activities offer a window on the world beyond
school and (2) the world beyond school offers a window on the use of school-based
mathematics in practical situations. This duality makes mathematical modelling
typically an activity that is affected by the world beyond the educational institu-
tions. For example, teachers and students can have experiences with mathematical
modelling in informal fields, such as at home or through newspapers and other
media. They can know people who work in engineering and who can inform them
about the gist of mathematical modelling in engineering. When they bring this
knowledge into the classroom, they can have an advantage over those who do not
have such acquaintances. Therefore, there is a need to study aspects of mathemati-
cal modelling education that transcend the institutional doors. Examples of research
questions could be the following: To what extent do students see themselves using
mathematics in their prospective professional lives? To what extent do teachers
identify with mathematical modelling researchers? How are teachers informed
about the twenty-first-century approaches to modelling, such as agent-based model-
ling (Bonabeau 2002)? How do newspapers inform their readers about mathemati-
cal modelling? Such research questions ask for a scope on the wider socio-cultural
environment of mathematical modelling education and will need theories that assist
in conceptualising more than the inner-institutional aspects of mathematical model-
ling education. To explore this, we selected a theory from sociology; we chose a
conceptual construct (to be explained later) and agreed to apply it in three different
studies. These are described in the ensuing three chapters of this book.
Research on the socio-cultural contexts of mathematical modelling education is
justified because there is evidence that not all students have access to develop
3 Connections of Science Capital and the Teaching and Learning of Mathematical… 35
modelling competencies to the same extent. First, there are differences between
countries, curricula, schools, and teachers that affect students’ performances. For
example, differences between countries can entail cultural differences. For instance,
Djepaxhija et al. (2016) found cultural effects in the responses of Albanian students
to a PISA problem. On the question which pizza is better value for money?, the
students did not interpret “value” in a capitalist-economic meaning (the amount of
substance for a unit amount of money) but rather gave “value” a social meaning
(food is to be shared), a nutritional meaning (food is to fill your stomach), and an
environmental meaning (food should not be thrown away). These students’ cultural
background led them to interpret the task context differently from what was intended
and increased their likelihood of failing on this PISA task.
Students’ mathematical modelling competencies are also affected by socio-
economic factors. Cooper and Dunne (1998) showed that lower class students in
England had more difficulties in handling modelling tasks than middle class stu-
dents, because the contexts (e.g. about tennis) were less familiar to them. Thus,
students’ performances on modelling tasks are affected by factors outside the edu-
cational institutions. Even if all students were taught mathematical modelling
through the same tasks by well-trained teachers, their social and cultural back-
grounds would affect not only their knowledge of “how the game of modelling is
played” but also the “quality of cards” that they hold.
In this chapter, we focus on one particular socio-cultural aspect that affects math-
ematical modelling dispositions. The first researchers to address this aspect were
Archer et al. (2015). They studied inequities within education, in particular, the
unequal participation of some ethnic groups in science education. They showed that
students are more likely to do well in science-related inquiry tasks and to pursue
science careers, if they have access to scientific resources through out-of-school
life. Examples of such resources are family members who are scientists, visits to
science museums, or a home environment that values scientific documentaries over
cartoon movies. Archer et al. (2015) coined the term science capital to capture such
resources and framed this term within the theory of the French sociologist Pierre
Bourdieu. Below, we will explain parts of this theory.
3.2 Bourdieu and Science Capital
Pierre Bourdieu (1930–2002) was a French sociologist and intellectual, who studied
the dynamics of power in society, in particular, the subtle ways in which power is
transferred and social order maintained across generations. He developed a theory,
which explains how certain people have privileges and better know how to “climb
the ladders” in society. Bourdieu studied why a democratic country like France was
ruled by a small group of politicians, most of whom had the same elitist back-
grounds. He observed that it was not just economic capital that enabled social
mobility; power was also maintained within the same circles through, for example,
“good” diplomas and the “right” social contacts.
36 P. Vos et al.
Bourdieu (1984) did, among others, a sociological study of taste, showing that
certain preferences of food, music, and art are taught to children, which they then
internalize as members of their social class (wine or beer, classical or pop music,
paintings or graffiti). Unconsciously, children then develop a love for and aversion
towards certain objects, and subsequently, this taste preference supports people in
recognising and relating to “their own” and distinguishing between social classes.
Important concepts that Bourdieu (1986, see also Bourdieu and Wacquant 1992)
introduced were social capital and cultural capital. Social capital consists of dura-
ble acquaintances, such as an old boys’ network (connections among former stu-
dents of male-only private schools) and people with whom one has goodwill (so you
can ask them a favour). Cultural capital consists of educational qualifications (diplo-
mas, academic titles, etc.), dispositions (knowing conventions, etiquette, speaking a
certain language), and cultural artefacts (e.g. owning books). Both social capital and
cultural capital are valuable resources that people can accumulate, just like eco-
nomic capital, and that generate profits and privileges. According to Bourdieu
(1986), one type of capital can help to accumulate another. For example, economic
capital can build cultural capital: rich people can hire a tutor for improving their
children’s education. Or social capital can build cultural capital: if you know some-
one who can help you get an internship, this will improve your vocational training.
Bourdieu also described symbolic capital, which is regarded as having much
exchange value for developing honour, prestige, or recognition (e.g. royals, war
veterans, Olympic medallists, or former ministers).
The above-described concepts were constructs developed by Bourdieu to analyse
conservative forces in society. These concepts and Bourdieu’s underlying theory
have been used in mathematics education (e.g. Nolan 2012; Noyes 2009; Williams
and Choudry 2016), but not yet in mathematical modelling education research.
Using Bourdieusian language, Archer et al. (2015) defined another concept: sci-
ence capital. It captures the resources that offer advantages within scientific con-
texts. They wrote:
Our first iteration of a theoretical model of science capital combines the following: scien-
tific forms of cultural capital (scientific literacy; science dispositions, symbolic forms of
knowledge about the transferability of science qualifications), science-related behaviours
and practices (e.g., science media consumption; visiting informal science learning environ-
ments, such as science museums), science-related forms of social capital (e.g., parental
scientific knowledge; talking to others about science). (p. 929)
We add that science capital also encompasses economic capital, for example,
having financial resources to visit a science museum. We take science capital as
being related to inquiry and research, primarily in the natural sciences (biology,
physics, etc.) but also in the social sciences or humanities. For example, knowledge
about research methods in historical research (text comparison between historical
documents, chemical analyses of artefacts at archaeological sites) is included in sci-
ence capital. Also, we follow Black and Hernandez-Martinez (2016) in claiming
that knowledge of, and about, mathematics and mathematical modelling is part of
science capital. They also pointed at the distinction between the exchange value of
science capital (as a means to accumulate capital (e.g. qualifications) for its own
3 Connections of Science Capital and the Teaching and Learning of Mathematical… 37
sake) and the use value of science capital (as a means to engage with scientific
activities).
We want to caution that science capital is not the same as scientific capital, a
term used by Bourdieu (2004, p. 34) to refer to “… a particular kind of symbolic
capital, a capital based on knowledge and recognition”. Scientific capital is, for
example, expressed through academic titles, which symbolize competence and dis-
tinction, based on peer recognition, associated with personal academic success.
Scientific capital entails more grants, bigger projects, better students, and other
privileges.
3.3 cience Capital as a Tool in Research on Mathematical

S
Modelling Education
Archer et al. (2015) observed that science capital is unevenly spread across societal
groups and students with more science capital are more likely to enjoy and do well
in science subjects. It is yet unexplored, whether science capital can also give teach-
ers and students advantages in mathematical modelling teaching or learning.
Therefore, we contend that the concept of science capital and the theory of Bourdieu
can be useful in research on mathematical modelling education. The three ensuing
chapters of this book describe studies, in which science capital is studied in relation
to mathematical modelling education. In the first contribution, Frejd (Chap. 4, this
volume) describes three teachers and their experiences and possibilities to accumu-
late science capital for teaching mathematical modelling. The teachers state they
learn much from talking with people working in science, or when they visit, for
example, laboratories, which they thereafter can use in their teaching. In the second
contribution, Hernandez-Martinez (Chap. 5, this volume) describes the cases of
Scott and John, two engineering students who both consider themselves good at
mathematics, and their parents have provided them with resources that allowed
them to succeed in their education. In Bourdieusian terms, both have much science
capital. However, their attitudes and approaches to a teaching practice where math-
ematical modelling is central differs considerably. These two cases show the com-
plex interrelation between capital, identity, and the institutional cultures of the field
of education. In the third contribution, Vos (Chap. 6, this volume) describes how
visitors to science museums can accumulate science capital through the mathemati-
cal activities and exhibits there. In one museum, mathematics is presented through
puzzles and soap bubbles, as a subject with no use value; in the other museums, the
usefulness of mathematical models is stressed (e.g. mathematical models are used
for aircraft safety, protection from floods, etc.). The three studies open a window on
the wider context of mathematical modelling education, covering teachers’ profes-
sionalization, students’ perspectives on their future, and arenas for informal learn-
ing. The studies show that the concept of science capital is a powerful analytical tool
for research on socio-cultural aspects of mathematical modelling education. Other
questions that could be studied from this Bourdieusian perspective could deal with,
38 P. Vos et al.
for example, students’ home support for doing mathematical modelling tasks, the
role of modelling researchers in the alumni networks of mathematics teachers, the
sources that inspire designers of mathematical modelling tasks, or the way that
mathematical modelling is displayed in newspapers or popular television series.
References
Archer, L., Dawson, E., DeWitt, J., Seakins, A., & Wong, B. (2015). “Science capital”: A con-
ceptual, methodological, and empirical argument for extending bourdieusian notions of capi-
tal beyond the arts. Journal of Research in Science Teaching, 52(7), 922–948. https://doi.
org/10.1002/tea.21227.
Ärlebäck, J. B., & Frejd, P. (2013). Modelling from the perspective of commognition – An emerg-
ing framework. In G. A. Stillman, G. Kaiser, W. Blum, & J. P. Brown (Eds.), Teaching math-
ematical modelling: Connecting to research and practice (pp. 47–56). Dordrecht: Springer.
Barbosa, J. C., & Pereira, A. M. (2013). Conflicts in designing tasks at collaborative groups. In
C. Margolinas (Ed.), Task design in mathematics education: ICMI study 22 (pp. 541–548).
Oxford: ICME.
Black, L., & Hernandez-Martinez, P. (2016). Re-thinking science capital: The role of ‘capital’and
‘identity’in mediating students’ engagement with mathematically demanding programmes at
university. Teaching Mathematics and its Applications, 35, 131–143. https://doi.org/10.1093/
teamat/hrw016.
Bonabeau, E. (2002). Agent-based modeling: Methods and techniques for simulating human sys-
tems. Proceedings of the National Academy of Sciences, 99(suppl 3), 7280–7287.
Bourdieu, P. (1984). Distinction: A social critique of the judgment of taste (G. Nee, Trans.).
Cambridge, MA: Harvard University Press.
Bourdieu, P. (2004). Science of science and reflexivity. Cambridge: Polity Press.
Bourdieu, P., & Wacquant, L. J. (1992). An invitation to reflexive sociology. Chicago: University
of Chicago Press.
Cooper, B., & Dunne, M. (1998). Anyone for tennis? Social class differences in children's responses
to national curriculum mathematics testing. The Sociological Review, 46(1), 115–148. https://
doi.org/10.1111/1467-954X.00092.
Djepaxhija, B., Vos, P., & Fuglestad, A. B. (2016, July 24–31). The Pisa scoring guidelines and
their prescriptions for interpreting a problem situation. Paper presented at ICME13 in Topic
Study Group 39 (Large-Scale Assessment), Hamburg.
Nolan, K. (2012). Dispositions in the field: Viewing mathematics teacher education through the
lens of Bourdieu’s social field theory. Educational Studies in Mathematics, 80(1), 201–215.
https://doi.org/10.1007/s10649-011-9355-9.
Noyes, A. (2009). Participation in mathematics: What is the problem? Improving Schools, 12,
277–288. https://doi.org/10.1177/1365480209342682.
Rosa, M., & Orey, D. C. (2013). Ethnomodelling as a methodology for ethnomathematics. In
G. A. Stillman, G. Kaiser, W. Blum, & J. P. Brown (Eds.), Teaching mathematical modelling:
Connecting to research and practice (pp. 77–88). Dordrecht: Springer.
Williams, J., & Choudry, S. (2016). Mathematics capital in the educational field: Bourdieu and
beyond. Research in Mathematics Education, 18(1), 3–21. https://doi.org/10.1080/1479480
2.2016.1141113.
Chapter 4
Teachers’ Possibilities to Generate Science
Capital for Modelling
Peter Frejd
Abstract This chapter adapts the concept, science capital, to explore mathematics
teachers’ possibilities to accumulate science capital, with the aim to identify teach-
ers’ potentials and constraints regarding their use of realistic modelling activities
within mathematics education. An analysis of transcripts indicated that three inter-
viewed teachers’ scientific preferences and intentions for establishing a teaching
practice connected to the world beyond school are high. While the interviewees
presented examples of how modelling activities can be realised in the education
system, obstacles for progress are also identified in the analysis referring to lacking
financial resources and social network of people working with science.
Keywords Bourdieu (theory of) · Mathematical modelling · Mathematics teachers

· Science capital · Theoretical perspectives · Use value (of mathematics)
4.1 Introduction
Arguments have been raised for a paradigm shift within science and mathematics
education to reorganise the dominant teaching practice (Black and Hernandez-
Martinez 2016). Instead of aiming for high scores on tests (exchange value), one
should develop a teaching practice aiming for critical awareness of how science and
mathematics come into play at different levels of society (use value) (Black and
Hernandez-Martinez 2016). One such practice focusing on the use value of mathe-
matics and science is the teaching of mathematical modelling. While mathematical
modelling is a teaching goal in many curricula worldwide (Borromeo Ferri 2013), it
has not (yet) gained momentum everywhere in day-to-day teaching and learning of
P. Frejd (*)
e-mail: peter.frejd@liu.se

40 P. Frejd
mathematics (Frejd 2014). Thus, for the paradigm shift mentioned to become more
than an unreachable utopia, it requires that teachers themselves are given opportuni-
ties to develop knowledge related to teaching modelling, with possible influences on
their teaching practice.
The aim of this study is to explore teachers’ experiences and possibilities to
accumulate science capital. Research results suggest a variation of how teachers
conceptualise the principles of science capital and how they operationalise it in their
teaching practice (King et al. 2015). Part of the variability is explained by teachers’
beliefs of science as relevant to students’ everyday life and teachers’ aspiration to
improve the teaching practice of science (King et al. 2015), which are factors related
to teachers’ own accumulated science capital. Based on these results and the fact
that mathematical modelling is a key component in science focusing on the use
value of mathematics (Black and Hernandez-Martinez 2016), an analysis of highly
qualified science and mathematics teachers about their possibilities to develop sci-
ence capital will be useful for discussing opportunities and challenges for the devel-
opment of a teaching practice, where mathematical modelling becomes a central
component. The following research questions are posed to address this aim:
RQ1: What aspects of science capital do teachers accumulate in and out of school
practice?
RQ2: How can differences in possibilities to accumulate science capital be charac-
terised and explained?
4.2 cience Capital for Mathematics Teachers and Its

S
Relation to Mathematical Modelling
Archer et al. (2015) draw on Bourdieu’s (1986) concept of capital for their develop-
ment of the notion of science capital (for further reading, see Vos et al. 2020).
Science capital has been used in several empirical studies (e.g. Archer et al. 2015;
Black and Hernandez-Martinez 2016) exploring students’ access to and use of sci-
ence capital. Some studies also investigate teachers’ conceptualizations of science
capital and their implementation of the concept in their teaching (King et al. 2015).
In this chapter, I adapt the notion of science capital for the purpose of making it
applicable for an analysis of teachers’ possibilities to accumulate science capital,
thereby laying a foundation for a discussion about teachers’ potentials for develop-
ing teaching practices of real-life modelling. Table 4.1 characterises the notion of
science capital for mathematics teachers.
All aspects in Table 4.1, except for 6 and 10, are based on Archer et al. (2015,
pp. 929–932), but the descriptions are slightly adapted to refocus on the science
capital of the teachers themselves and their teaching practice. The sixth aspect con-
cerns teachers’ intentions for science teaching, and the last aspect (10) relates to
Bourdieu’s (1986) concept of economic capital and comprises the financial resources
that a teacher possesses for professional development.
4 Teachers’ Possibilities to Generate Science Capital for Modelling 41
Table 4.1 Science capital for mathematics teachers

Descriptions of aspects of science capital for
Form Aspects of science capital mathematics teachers:
Cultural 1. Scientific literacy A teacher’s knowledge of science and how
capital science works. This also includes his or her
confidence in teaching science
2. Scientific preferences A teacher’s attitudes, values, and dispositions
of science as relevant to students’ everyday life
3. Symbolic knowledge of A teacher’s understanding of the benefits of
transferability of science being scientifically qualified, even beyond the
qualifications needs of his/her teaching career
Behaviours 4. Consumption of science A teacher’s use of scientific media for his or
and practices media her professional development and as a resource
for teaching
5. Out-of-school science A teacher’s participation in informal science
learning context learning/teaching contexts (e.g. visits to science
museums and workplaces)
6.Intentions for science A teacher’s aspiration to continue and improve
teaching the teaching practice of science and modelling
Social capital 7. Science identity Science-related family backgrounds and the
extent to which a teacher is recognised by
others as being scientific
8. Knowing people working A teacher’s social circle of people with
with science science-related work
9. Talking to others about A teacher’s conversations about science in and
science out of school with key people in his/her life
(e.g. colleagues, friends)
Economic 10. Economic resources for Available financial resources for improving
science science education (e.g. money for professional
development)
Adapted from Archer et al. (2015)
As indicated in the introduction, there is a link between mathematical modelling

and science capital, which Black and Hernandez-Martinez (2016) address as the use
value (a term used in the theory of Marxian economics) of mathematics. Economic
models, weather forecasts, new technology, etcetera are products of mathematical
modelling work (see e.g. Frejd and Bergsten 2016) and influence our daily lives at
both an individual and a societal level. In Table 4.1, modelling can be embedded in
science, in particular, in the categories of scientific literacy, scientific preferences,
and intentions for science teaching. Scientific literacy includes teachers’ knowledge
about the role of mathematics in the process of developing a model for phenomena,
in the evaluation and communication of its outcomes. Scientific preferences and
intentions for science teaching concern teachers’ attitudes and willingness for teach-
ing science, which includes mathematical modelling.
42 P. Frejd
4.3 Method
The empirical data for this chapter consist of transcribed interviews from three
mathematics and science teachers (Abraham, Bert, and Cathy: pseudonyms) work-
ing in three different upper secondary schools in Sweden. The interviews were con-
ducted during spring 2016 as a part of a larger study by the author. The aim of the
larger study is to investigate how professional education of upper secondary math-
ematics teachers can be developed, organised, and sustained within the local school
organisations and the current educational system, so that it strengthens the relation
to industry and society as well as facilitates the students’ learning of mathematical
modelling. The three teachers, all known to the author, were chosen based on the
following criteria: their working experiences (more than 10 years of teaching), their
high levels of teaching science, and a high science identity (recognised by the
author). The interview questions were organised around the teachers’ background,
their experiences of professional development in mathematics education, their
experiences of working collaboratively with agents from out of school, and their
attitudes to such short- and long-term activities that strengthen the relation between
industry and mathematics education at school. The interviews, lasting between 30
and 60 min, were carried out drawing on the interview questions in a flexible man-
ner in a semi-structured interview format (Robson 2002). The transcripts were ana-
lysed based on the 10 aspects in Table 4.1 to identify teachers’ accumulated science
capital (RQ1). In order to characterise possible differences in the participants’ accu-
mulation of science capital (RQ2), the transcripts were compared and ranked
according to a relative scale of high, medium, and low accumulation of science
capital. This ranking was based on the participants’ relative emphasis on the aspects
of science capital.
4.4 Results
The aspects of scientific capital accumulated by the teachers and how they differ
between the teachers are analysed in Sects. 4.4.1, 4.4.2 and 4.4.3. The results are
summarised in Sect. 4.4.4.
4.4.1 The Case of Abraham
Abraham has 18 years of teaching experience from different upper secondary

schools and programs, and he has taught almost all courses in mathematics and
physics. He is now head of mathematics in an upper secondary school in a middle-
size town and was recently in charge of running a government initiated professional
development course (Skolverket 2012) for mathematics teachers in the same town.
In addition, he has authored upper secondary school physics textbooks.
During his career, Abraham has gained scientific literacy through several differ-
ent out-of-school science learning activities. He has participated in professional
development courses in Sweden and abroad (the USA and Switzerland), in exchange
programs with other European countries and in science days and teachers’ confer-
ences. He has also visited science museums and some scientific workplaces. While
Abraham, in the interview, was reflecting on his varied professional development
participation, he highlighted the importance of talking to others about science if one
wants to make a change in classroom practice:
It has happened during my career that you attend these teachers’ conferences and they are
incredibly fruitful, however, compared to other types of professional development it is a
‘push’ for that individual at that point. But if you want something to happen a week after
for more than the person visiting the conference, then you need to have someone to talk to,
such as a gathering to discuss the issue.
He also described talking to other colleagues as one of the two most important
ways to gain scientific literacy and said “very much of the best professional devel-
opment happens informally in the teachers’ working room”. In particular, he
stressed that the contribution of colleagues is important for developing an atmo-
sphere, productive for professional development. The other important component of
gaining scientific literacy refers to “curiosity”.
That you are curious, that is you inform yourself and find things out. It has become so much
easier since [the] Internet has grown. You have a full library at home at 2 am and you have
10 helpful librarians that helps you.
The quotation above also indicates that Abraham consumes scientific literature.
He uses scientific literature both to prepare his teaching and to write textbooks in
physics. Abraham was appointed head of mathematics of his school district, partly
due to his professional curriculum vitae and partly in recognition of his well-devel-
oped science identity by his principal.
Cooperation with industry is not prioritised according to Abraham, who said:
…that is one of the things I’m bad at, that I wish I could do better. I think I’ve improved
some since I moved to […], but it still is not good.
The teachers at his school have no regular activities related to out-of-school prac-
tice, even though they have visited a power plant and a university. He has several
explanations for the lack of connections to the workplace. Some of them relate to
knowing people working with science, lack of time to prepare visits, as well as the
lack of funding, that is, lack of economic resources for science.
One year we had to take money from physics [to be able to go to a power plant 200 km
away] and that money could have been used for buying a few voltmeters. Voltmeters are
permanent…So are students’ experiences. However, from a school perspective this point of
view is ignored.
In his career, however, Abraham has been able to obtain financial support for
visits to Berkeley Laboratories in USA, to CERN in Switzerland, and also to Iceland
and Great Britain. His current workplace has “an amount of money that you may
use [for such purpose]. It is not a lot of money every year, but usually there is
enough”.
44 P. Frejd
Abraham has high intentions for science teaching to gain out-of-school science
learning activities for his students, so that they do not feel that “mathematics is just
something we do in school”. He stresses the use value of mathematics and suggests
the following design:
You start with a workplace visit yourself, to get into their work situation. Then you involve
the students in some form. An industry program … send students to do real world problems.
That includes estimations and the use of mathematics and physics. Such problems do prob-
ably appear in many workplace situations and one could let students work with such prob-
lems as case studies.
However, he adds that such activities are time-consuming for the students but even
more so for the teacher to prepare.
4.4.2 The Case of Bert
Bert has 16 years of teaching experience from two different upper secondary
schools. Similarly to Abraham, Bert has taught almost all courses in mathematics
and physics and has been in charge of running the same professional development
course. He has had the position of head of department and recently moved to another
school. Due to his professional curriculum vitae and his career, he was appointed to
run a professional development course for other teachers about computer-aided
laboratory activities in physics.
Compared to Abraham, Bert has had fewer opportunities to gain scientific liter-
acy in his career. He said he had been to some conferences in mathematics educa-
tion and entered a course in discrete mathematics, but “other than that nothing, nor
courses connected to didactics or methods for teaching mathematics”.
However, he has experiences of out-of-school science learning related to phys-
ics, such as professional development in computer-aided laboratory activities and
medical physics. He has also visited some workplaces and attended science days.
The teachers at his school can choose between four types of professional develop-
ment: learning study, research circle (reading and discussing research literature),
ICT in natural science (computer-aided laboratory activities), and a fourth type he
could not recall. From evaluating the professional development course initiated by
the government, he emphasised talking to colleagues about science and mathemat-
ics as an important part of professional development. He said that “many found it
useful to sit down [in a group of teachers] and discuss mathematics for 1 to 1.5
hours every week”.
The teachers at his school also do classroom visits to help, and learn from, each
other to gain scientific literacy. He mentioned that he uses ideas from colleagues in
his teaching. Bert has not gained the same economic resources as Abraham in his
career, but he indicated that there might be opportunities to increase his economic
resources for science.
If I would like to take a course at the university in relation to my subjects, I would probably
get financial support to do so and maybe some reduction of teaching time to complete the
course, because it may take more time than the time we get for professional development.
He argued that he consumes scientific literature “to prepare and keep up to date
with the subject, such as new ideas”. Bert has positive scientific preferences for con-
necting to out-of-school practice focusing on the use value of mathematics and
wants to improve his teaching (Intentions for science teaching):
I think students would appreciate to see the application of mathematics, in physics they may
see it more. To see the usefulness of the mathematics they study and will study would be a
benefit to them.
Time and experience (e.g. he did not know people working with science) together
with the demands of the national course tests in mathematics were mentioned as
constraints for developing his teaching practice, in particular, with respect to model-
ling. Nevertheless, he suggested the following activity:
You may follow them [workers] for a period of time. First invite them to the classroom to
make a start, followed by a workplace visit and maybe the students can do something by
themselves a day at the company. Let them see that this is a part of the course, not just
something fun, and that this activity may be useful in the long run.
His interview answers indicate that his scientific identity mainly refers to his
qualifications of professional development and his recognition by the principal and
colleagues, but also that his wife, who is a nurse, contributes to it with regard to
medical applications of physics.
4.4.3 The Case of Cathy
Cathy has 12 years of teaching experience at upper secondary school and has a PhD
in cell biology. She used her PhD certificate to be appointed as a “lecturer” (a posi-
tion requiring a PhD) at her upper secondary school, which signals high science
identity. She has taught almost all courses in mathematics, chemistry, and biology
in different programs. Together with representatives from the workplace and a uni-
versity, she is also a member of the committee for school cooperation at this univer-
sity. The aim of the committee is to improve the attitudes of students in grades 7–12
towards science and mathematics. To achieve this goal, the committee arranges
activities for students and teachers such as science days and visits to the university.
Her postgraduate education to a PhD and her involvement in this committee indicate
that her knowledge of people working with science is more extensive than was the
case for the other two participating teachers. Her career also shows ways to improve
a teacher’s science identity. However, she regrets that her scientific connections to
industry have a minor influence on her teaching:
46 P. Frejd
Unfortunately I would say that it is very, very, very little of that [cooperation with agents
from out of school]. Sometimes problems from [a teacher journal] may be related to real
life. However, I think it is rare.
As with the other participants, Cathy would “absolutely be interested” in coopera-

tion with science-related workplaces. Her scientific preferences and intentions for
science teaching for connecting her teaching with out-of-school learning are high
and particularly illustrated by her design of a noncompulsory course in science,
focusing on cell biology, with four workplace visits to the university and hospitals.
She also stressed the use value of mathematics and the necessity of developing the
students’ science identity by giving them opportunities to meet, or to cooperate
with, people using mathematics in their work:
[There is a need for] students to see, and to some extent teachers, the use of mathematics in
the workplace. The classical question: What does an engineer do? They need to be
acquainted with and hear that they do mathematics.
She suggested activities such as inviting people from workplaces to give lectures in
the schools and increasing the number of workplace visits. Cathy has not had the
same opportunities as the other participants to gain economic resources for science
for professional development from her school:
No, we have not got any resources for professional development, nothing… very meagre
funds… you may get economic funding for buying literature…but the school administra-
tion does not pursue any professional development…nothing within subjects.
4.4.4 Summary of Results
To answer RQ1, the identified aspects of scientific capital that the interviewed
teachers have accumulated through work within upper secondary school and other
experiences of out-of-school practice are summarised and ranked in Table 4.2. This
table shows that all the participants have accumulated high scientific preferences for
teaching mathematics connected to science and they emphasised the use value of
mathematics relevant for students’ everyday life. The three interviewed teachers
have high intentions for science teaching to improve their teaching practice. To
improve their scientific literacy, they all are large consumers of scientific media, and
they all emphasise talking to others about science as a central part of their
Table 4.2 Aspects of science capital identified for mathematics teachers: H = high;
M = medium; L = low
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
Aspects S-liter. S-pref. Symb. Cons. Out of. Int. S-iden. Know. Talk. Econ.
Abraham H H M H H H H L H H
Bert M H M H M H M L H M
Cathy L H H H L H H M H L
professional development. However, none of the participants mentioned that they

had science-related discussions with family and friends except Bert.
Concerning RQ2, there seems to be a clear connection between the possibilities
to accumulate scientific literacy, out-of-school learning experiences, and economic
resources for science, since columns (1), (5), and (10) are identical. This is not sur-
prising, because professional development for scientific capital and out-of-school
activities often require schools to spend economic resources. Cathy has used her
PhD degree as cultural exchange value (symbolic knowledge of transferability of
science qualifications) to be appointed as a lecturer, whereas Abraham and Bert
have used other accumulated qualifications to become heads of mathematics at their
schools. The circumstance that Cathy has a PhD signals that her science identity is
high and also may explain that she knows more people working with science than
Abraham and Bert do. That Abraham has written textbooks for physics indicates
that his science identity is high. Since Bert has neither written textbooks nor earned
a PhD, his accumulated science identity is ranked as medium.
4.5 Discussion and Conclusion
This empirical case study shows, in answering RQ1, that the interviewed teachers
all have high scientific preferences and intentions for science teaching for improv-
ing teaching practice in mathematics and its connections to out-of-school practice
emphasising the use value of mathematics. These preferences and attitudes are rea-
sonable prerequisites for a paradigm shift within mathematics education and for a
successful introduction of realistic modelling activities in the classroom. However,
in order to develop a teaching practice focusing on such realistic modelling activi-
ties, both professional development regarding mathematical modelling (scientific
literacy on modelling) and out-of-school activities are required. In line with
Skolverket (2013), and in answering RQ2, this study shows that these three teachers
in different schools have different possibilities for such a professional development
due to economic resources. As described by Bourdieu (1986), “the different types of
capital can be derived from economic capital, but only at the cost of a more or less
great effort of transformation, which is needed to produce the type of power effec-
tive in the field in question” (p. 252), indicating that economic resources are the
foundation for a change of practice. Providing teachers with more economic
resources would facilitate their professional development and could give them more
time to design, implement, and evaluate modelling activities. The heads of schools
and policy makers are agents that come into play regarding economic funding, but
how these agents should be convinced to set aside money for stimulating teachers’
science capital for modelling is a different issue. One possible solution may be to
use the social capital of knowing people in science with high symbolic capital to
make an impact on education. Since the teachers in this study are positive for chang-
ing their practice under the right circumstances, they may also put some pressure on
their principals to increase their available economic capital.
48 P. Frejd
The findings are not able to be generalised beyond this particular sample.
However, the answers of the teachers in this study suggest how a collaboration with
the workplace might be included in the Swedish education system and elsewhere. In
order to let students also accumulate science identity, both Abraham and Bert sug-
gested that students and teachers together visit workplaces and identify realistic
modelling problems to be solved by the students. This suggestion seems to be in line
with other suggestions found in international research literature regarding identity
making for professional modelling, such as role play activities and extracurricular
activities (e.g. Frejd 2017; Vos 2015). The use of communication, talking to others
about science, is a social aspect of science capital highlighted by the teachers in this
study as a central part of professional development in science and mathematics. One
of the most important aspects of the work of professional mathematical modellers is
communication (Frejd and Bergsten 2016), which implies that communication
should be an important part also in teaching.
This chapter defined science capital for mathematics teachers with the aim to
analyse teachers’ possibilities to accumulate science capital, in order to discuss
potentials for developing the teaching practice of modelling. To what extent teach-
ers’ science capital as defined here is adequate for uncovering the teachers’ poten-
tials and constraints for implementing realistic modelling in their science and
mathematics teaching is a matter of discussion. On the one hand, the analysis
informs about aspects from the world beyond schools that affect teachers on how to
teach modelling and how differences in accumulated capital may cause inequities in
modelling education. It also resulted in some insights about components for chang-
ing teaching practice and some examples of ways to collaborate with the workplace.
On the other hand, as argued by Bourdieu (1986), the different forms of capital are
intertwined and not mutually independent which makes a categorisation of tran-
scripts in terms of teachers’ capital nontrivial. The implicit description of modelling
as a key component in teachers’ science capital also influences the robustness of the
discussion. For example, some of the aspects of teachers’ science capital, such as
the symbolic knowledge of transferability of science qualifications, seem to focus
on social capital (Bourdieu 1986) rather than connecting to mathematical model-
ling. However, this chapter reports only a preliminary small case study. Further
research is needed to explore teachers’ possibilities for gaining science capital for
mathematical modelling.
References
Archer, L., Dawson, E., DeWitt, J., Seakins, A., & Wong, B. (2015). “Science capital”: A concep-
tual, methodological, and empirical argument for extending Bourdieusian notions of capital
beyond the arts. Journal of Research in Science Teaching, 52(7), 922–948.
Black, L., & Hernandez-Martinez, P. (2016). Re-thinking science capital: The role of ‘capital’ and
‘identity’in mediating students’ engagement with mathematically demanding programmes at
university. Teaching Mathematics and its Applications, 35, 131–143.
Borromeo Ferri, R. (2013). Mathematical modelling in European education. Journal of

Mathematics Education at Teachers College, 4(2), 18–24.
Frejd, P. (2014). Modes of mathematical modelling. An analysis of how modelling is used and inter-
preted in and out of school settings. Doctoral dissertation. Linköping: Linköpings universitet.
Frejd, P. (2017). Mathematical modelling as a professional activity – Lessons for the classroom.
In G. A. Stillman, W. Blum, & G. Kaiser (Eds.), Mathematical modelling and applications:
Crossing and researching boundaries in mathematical modelling education. New York:
Springer.
Frejd, P., & Bergsten, C. (2016). Mathematical modelling as a professional task. Educational
Studies in Mathematics, 91(1), 11–35.
King, H., Nomikou, E., Archer, L., & Regan, E. (2015). Teachers’ understanding and operation-
alisation of ‘science capital’. International Journal of Science Education, 37(18), 2987–3014.
Robson, C. (2002). Real world research: A resource for social scientists and practitioner-
researchers (2nd ed.). Oxford: Blackwell.
Skolverket. (2012). Matematiklyftet. [Mathematics lift]. Stockholm: Skolverket. Available from:
http://matematiklyftet.skolverket.se
Skolverket. (2013). Resource allocation to schools by municipalities: A summary of two reports
from the Swedish National Agency for Education. Stockholm: Skolverket. Available from:
www.skolverket.se/publikationer?id=2961.
Vos, P. (2015). Authenticity in extra-curricular mathematics activities: Researching authenticity
as a social construct. In G. A. Stillman, W. Blum, & M. S. Biembengut (Eds.), Mathematical
modelling in education, research and practice: Cultural, social and cognitive influences
Vos, P., Hernandez-Martinez, P., & Frejd, P. (2020). Connections of science capital and the teach-
ing and learning of mathematical modelling: An introduction. In G. A. Stillman, G. Kaiser, &
E. Lampen (Eds.), Mathematical modelling education and sense-making (pp. 33–38). Cham:
Springer.
Chapter 5
Science Capital, Habitus,
and Mathematical Modelling Practices
in the Field of University Education
Paul Hernandez-Martinez
Abstract This chapter analyses two second-year undergraduate engineering stu-

dents’ narratives of engagement with a mathematics course aimed at developing
mathematical modelling competences. Both of these students have come to univer-
sity with enough science capital to expect they will be successful in their mathemat-
ics courses and are confident in their mathematical knowledge and skills, but their
narratives revealed the different value that they attribute to mathematics, which is
reflected in their engagement with the mathematical modelling aspects of the course.
Bourdieu’s Theory of Practice is used to understand the complexities of social prac-
tices in the field of higher education and to explain the different positions that these
two students take regarding mathematical modelling.
Keywords University mathematics · Engineering mathematics · Mathematical

modelling · Narratives of mathematics · Science capital · Bourdieu’s Theory of
Practice
5.1 Introduction
In the introduction to this sequence of chapters (Vos et al., Chap. 3, this volume), we
expressed our aim to study aspects of mathematical modelling education that tran-
scend the classroom doors. For this purpose, we use some of the concepts developed
by the French sociologist Pierre Bourdieu and, in particular, Archer et al.’s (2015)
notion of science capital which they define as science-related skills and resources
that individuals acquire – usually at home or in out-of-school activities – and that
P. Hernandez-Martinez (*)
Department of Mathematics, Swinburne University of Technology, Hawthorn, VIC, Australia
e-mail: phernandezmartinez@swin.edu.au

52 P. Hernandez-Martinez
can generate forms of social advantage in an educational setting. This term has
become important to understand participation and engagement in science, technol-
ogy, engineering, and mathematics (STEM) subjects at school and university (see,
e.g. New Scientist 2017).
However, elsewhere Black and Hernandez-Martinez (2016) have argued that the
notion of science capital needs to be reconceptualised in order to fully account for
the dialectic relationship between individuals and the social field where they partici-
pate. In other words, we hypothesise that having science capital is not enough to
guarantee success in a STEM career or that access to this kind of capital will not
always produce the same engagement with or learning in mathematics or, in the
case of this chapter, mathematical modelling at university. To explore such hypoth-
esis and to try to theorise science capital more fully, I use Bourdieu’s Theory of
Practice (Bourdieu 1972), which explains the complex interplay between (science)
capital, and other important Bourdeusian concepts such as habitus and field. In this
chapter, I analyse the narratives of two second-year undergraduate engineering stu-
dents who have come to university with a good amount of science capital and are
confident in their mathematical knowledge and skills. They were part of a mathe-
matics course designed to develop mathematical modelling competences, and given
their backgrounds, the expectation was that both students would engage in similar
ways and achieve alike in this course. Therefore, this chapter will answer the fol-
lowing research questions: How are two engineering students, who have come to
university with similar science capital, positioned by a practice designed to develop
students’ mathematical modelling competences? How can differences and similari-
ties in these positions be explained from a Bourdeusian perspective?
5.2 Bourdieu’s Theory of Practice
In his Outline of a Theory of Practice, Bourdieu (1972) aimed to reconcile the dual-
ism of structure versus agency, that is, the structuralist perspective that gives pri-
macy to the rules and patterns of society (i.e. the social structures) as dictating
human thoughts and behaviours versus the voluntarism or functionalist perspective
that gives primacy to the individuals as free to make choices and determine their
lives (i.e. the free will of individuals). To overcome this dualism, Bourdieu explained
practice as the complex interplay of his main concepts of capital, habitus, and field:
“(Habitus × Capital) + Field = Practice” (Bourdieu 1984, p. 101). I will briefly
explain each of these concepts.
The habitus consists of a person’s thoughts, beliefs, interests, understandings,
histories, and ways of acting – a socialised body that demonstrates the ways in
which not only is the body in the social world but also the ways in which the social
world is in the body (Bourdieu 1998). Bourdieu (1984, p.170) defined the habitus as
“a structuring structure, which organises practices and the perception of practices”
or, in other words, the habitus predisposes individuals towards certain behaviours
5 Science Capital, Habitus, and Mathematical Modelling Practices in the Field… 53
that are expected of “people like us” (Bourdieu 1990, p. 78). For example, some
students might come from homes where their parents value education and encour-
age certain discipline and a scientific way of seeing the world. When these students
get to school, their habitus is so attuned to the practices of the classroom that they
have a “feel for the game”, acting “like fish in water” (Bourdieu and Wacquant
1992, p. 127) and having an advantage over their peers whose habitus might not be
in tune with certain school practices (see e.g. Willis 1977).
The field is the arena of social interaction where people take positions and com-
pete for power and privilege as they attempt to maintain or change their position in
the field. The rules of the field – called by Bourdieu the “structured structures”
(Bourdieu 1984) – are not formalised but rather tacit in nature (Wacquant 2011) and
are quite arbitrary, reflecting the values of the dominant culture (Bourdieu and
Passeron 1990). For example, in the field of (mathematics) education, it is quite
common to believe that mathematical proficiency is due to natural ability rather
than effort, citing the “genius” of famous mathematicians. Therefore, when students
achieve a degree of success by hard work, they are seen as having worked too hard,
almost unnaturally.
Each person entering a field brings with them a certain amount of resources that
they can use to compete for positions and power. These resources, or capital, can
take different forms: economic, cultural, social, etcetera. The forms of capital that
are valued in a field can be exchanged for access to more resources or positions of
privilege (e.g. a first class degree can be “exchanged” for a high paid job). Capital
can be objectified (e.g. in mathematical instruments), embodied in the habitus (e.g.
being a hard worker or having modelling competences), or institutionalised (e.g.
having a STEM degree from a top class university). Science capital is science-
related resources such as achieving a mathematics secondary school certificate that
allows a student to enter a STEM degree or having parents/relatives with a science
degree and being used to listening to scientific ideas at home.
Bourdieu’s Theory of Practice emphasises that to understand a practice, for
example, that of learning mathematics in an engineering degree, one needs to look
carefully at the interplay between the rules of the field of university education (i.e.
what is possible to do), the capital that is valued in that particular field (e.g. being
mathematically “gifted”), and the habitus of each individual student that predis-
poses him/her to think and act in particular ways (e.g. knowing what is required to
succeed in an exam).
5.3 Methodology
The data for this chapter comes from semi-structured, in-depth interviews with two
second-year undergraduate engineering students from a research-intensive univer-
sity in England, hereafter called Scott and John (pseudonyms). Both students had
recently completed a mathematics for engineering course as part of their degree.
Scott obtained an A∗ grade, with a mark very close to 100, while John obtained a C
grade, with a mark in the middle 50s. The minimum passing grade in this university
(typical of most English universities) is D or 40 marks. The course was a mixture of
typical lectures (i.e. exposition of the content of the syllabus by the lecturer in front
of the class) with more innovative aspects that were designed to develop students’
mathematical modelling competences (Maaß 2006), as well as meta-knowledge
about modelling that is “the background knowledge about the nature of modelling,
how it is conducted and why mathematics can be applied in real situations” (Brown
and Stillman 2017, p. 357), and other social modelling competences such as col-
laborative work, argumentation, and reporting (Galbraith et al. 2007; Kaiser 2007).
These more innovative practices consisted of (1) mathematical modelling tasks
related to the topic of the week (i.e. the mathematics that had to be used was more
or less known from the outset), where students had to discuss and solve these tasks
in groups of four and present a written report of their solution (e.g. finding the
dimensions of a box that maximises a fixed volume, or modelling a slip casting
process with a differential equation), (2) students had to look for engineering exam-
ples of the topics seen during the week and explain these to the class, and (3) a
coursework video project consisting of a 6-min video where groups of four students
explained an engineering application of one topic seen in the course (e.g. Newton’s
Law of Cooling and how it is used in the cooling systems of a coal power plant).
This assessment counted for 15% of the total course marks. The other marks were
gained through two in-class timed tests (12.5% each) and the final timed examina-
tion (60%).
Scott and John volunteered to be interviewed, and their narratives provided a
remarkable difference in how they interacted with the mathematical modelling
aspects of the practice despite both having come with similar amounts of science
capital and a positive attitude towards mathematics. This situation provided rich and
interesting evidence of how their habitus and capital interacted differently with the
field and its structures.
The interviewer asked the students about their background, including familial
background and experiences with mathematics, their school experiences with math-
ematics and other mathematically related subjects, how they chose their degree,
their experience of their first year at university, and their opinions about the second-
year mathematics course, in particular the three innovative practices described above.
To analyse the interviews, narrative analysis was used. This methodology focuses
on the ways people make and use stories to interpret the world. It views these sto-
ries – or narratives – as interpretative devices through which people represent them-
selves and their worlds to themselves and others (Lawler 2002). For the purpose of
this chapter, the narratives that were created out of the interviews focused on the
students’ science capital before coming to university and their subsequent experi-
ences of mathematics during their second-year course, through the lenses of the
relationships between capital, habitus, and field. This analytic approach is well-
matched with Bourdieu’s Theory of Practice as “narratives construct a social world
that, in Durkheim’s terms, has exteriority and constraint. They are constitutive of
the realities they depict. Stories not only generate social realities, but become hard-
ened into institutional structures that then perpetuate and enforce them” (Bruner
1997, p. 137). In other words, narratives are the vehicle through which individuals
express the way they see the world and act on it (the “structuring structures” or habi-
tus), but at the same time they are a reflection of the rules of society (the “structured
structures” or the rules of the field). In the following, I present Scott’s and John’s
narratives of their experiences with mathematics and how they come to interact with
the practice of learning mathematics in an engineering degree during a university
course aimed at developing mathematical modelling competences.
5.4 Results
5.4.1 Scott’s Narrative
Scott comes from a family where both his parents are engineers. His relationship
with mathematics throughout his schooling has always been good. He described it
as follows:
Scott: Maths has always come pretty easily for me. (…) I read a couple of books
about Statistics. One of them was bought for me and then after that I went and
bought the other one. (…) Both my parents are engineers and they are very good
at maths.
When asked why he chose engineering as a degree, he explained that his best
subjects at school were physics and mathematics, and therefore engineering was a
“natural” option. However, it is reasonable to suggest that Scott’s parents also had
an influence in this choice, being both engineers. It is also clear from what he said
that he acknowledges the relationship between engineering and being “good at
maths”. His interest in reading Statistics books shows his mathematical capital
being embodied which is then confirmed by his mathematics habitus where mathe-
matics comes “pretty easily” for him. He described in more detail what he likes
about mathematics in the following way:
Scott: I do find it quite satisfying when in a piece of coursework or something…
when there’s a really long, complex thing I spent really long time getting through it,
I really worry that at any point I could’ve made mistakes, and I get to the end and
get an answer I’m pretty sure is right… and that kind of satisfaction looking back
just over the train wreck of workings, and then that leading to the nice neat answer.
It’s quite satisfying.
It is interesting to note that what he considers to be satisfying about mathematics
relates to long procedures with a “nice neat answer” at the end. It seems his math-
ematics habitus has been shaped by cultural views of mathematics as being “black
and white” or “having one correct answer” (Williams et al. 2009).
When asked to comment on the mathematical modelling tasks, Scott acknowl-

edged the value of knowing how mathematics could be applied to engineering but
sharply differentiated this knowledge from the need to get good marks in the exam:
Scott: In the teaching of this module, there was a differential equation of a box…
you were building a box and you… for a certain surface area of cardboard you had
the maximum volume, and so you had to create a set of equations, then create a dif-
ferential equation to find… how the volume relates to the dimensions and then you
solve it to find the maximum volume for surface area, and that was nice because…
again, you feel like ‘oh, this is how you would do it if you were in the situation’, but
I think that would probably be frustrating in the exam when you are just trying to
answer the question and get the marks.
When asked to comment particularly about the real-world examples that the lec-
turer requested from students, his narrative returned to this separation between
knowing how the mathematics can be applied and the needs of the exam.
Scott: I understand why he (the lecturer) tried it but I think… there were prob-
lems with that. I think throughout the whole module he was trying to give us this
understanding of real world applications of maths and so, he wanted us to e-mail
him examples of things that you were learning and then walked through how it’s
relevant and things like that… and I think the problem is while [it] is useful and
important to understand the real world applications, that at the end of the day…
most of the people… well, pretty much people care about passing the module and
get good grades and earn a degree, and in some cases trying to give an understand-
ing of the engineering applications detracted from actually just making sure you’re
learning how to do it.
Here we notice how Scott separates procedural knowledge (“learning how to do
it”) from the more relational understanding that the lecturer wanted to teach (i.e.
how mathematical concepts relate to the engineering profession). This was again
brought back by Scott when asked about the coursework video project:
Scott: The maths video project was over a month’s long of doing a 6-min video
about differential equations and I don’t feel it helped my maths knowledge in any
way. It felt unnecessary and taking up time that could’ve been used for actually
learning the content and so while I think it is important and useful to have the real
world applications I think is important not to… to remember that the students… just
at the end of the day care about the grades because while they want to get better at
what they’re doing, if they fail the module because you spent half of the lectures
talking about the real world applications and not the maths and then that wasn’t
relevant in the exam… it’s kind of… they rather just learn how to do the maths.
It is evident that Scott attributes great value to the exam and getting good grades
as a way to earn a degree. This contrasts with the value he ascribes to the mathemati-
cal modelling activities, the real-world examples, and even the coursework video
project that, although he acknowledges their potential use, he considers of lim-

ited worth.
5.4.2 John’s Narrative
John had a very different upbringing from Scott, as he described:

John: I was home educated and my father taught me maths at Primary level. At
GCSE (Secondary school) level was the first time maths really became important
for me, I was learning how to solve quadratics… and for some reason I just loved
that, it was great, I actually felt in love with it. (…) My mom does a lot of
(self-motivated) research just to make sure we get the best. She’s the one that’s been
pushing all this time to get us into university, I’m really proud of that.
None of John’s parents went to university, but his home education meant that
they were the main source of his educational capital. After John finished his second-
ary education, he went on to Sixth Form College to do a vocational pre-university
qualification in engineering (called BTEC) and took the pre-university academic
mathematics qualification (called A level) on the side – self-studied. It is evident
that John’s parents, in particular his mother, invested heavily in the development of
their children’s capital, even joining a “group of people who are home educated that
go to trips to museums and do projects and stuff”. In fact, John described his family
as always the first among their circle of friends to “try something, going to univer-
sity and stuff”. This clearly shows John’s accumulation of science capital. He has
developed a mathematics habitus where he “just loves doing it, it just makes me
really happy”, even though he knows that such enthusiastic attitude is not always
seen positively by his peers. His positive attitude towards mathematics showed
clearly when asked to comment on the modelling tasks:
John: I really liked the fact that he (the lecturer) did something different, he did
the group work, so every tutorial it had a rather complex problem and this problem
would have to do with mathematical modelling. And then we’d have to work
together in like, a group of four, and just work together trying to work out the model
and then solve the model. And it was stretchy, because it wasn’t like... it was harder
than our course so we had to really think about it. It was just different and engaging.
(…) Just teaching us to think about the problem in a modelling way.
For John, the modelling tasks provided him a challenge that “stretched” his mind
and made him “really think about it”. Furthermore, he found this engaging for the
sake of it. He showed a similar attitude with respect to the real-world examples.
He said:
John: In the beginning he (the lecturer) asked us to give him… research exam-
ples every week about the relevancy of the topic we just covered, and that
was really nice. I mean, some people in the class came up with really good
examples and it was good to see how like tide [sic] in with industry and
engineering as well… it was definitely very good.
Int: Did it help you?
John: Yes, it showed us where… it tied in with our general course but also, I’m
sure that having known these examples, like knowing that it could be
applied to like… uhm Newton’s law of cooling and stuff and going into
more technical modules I think it will tie back very nicely to this module.
It is important to notice the different views of mathematical modelling between
Scott and John. While Scott appreciated the modelling tasks and the real-world
examples, he could not see how these could bring any benefit to his learning, which
had to do mainly with an exchange value of mathematics (i.e. “exchanging” grades
for a university degree). For John, these activities provided a challenge and a useful
vehicle to tie in the mathematics with the “more technical modules” in his degree.
For him, the mathematics that he learned through these activities had a use value to
his engineering studies. Even the coursework video project provided him with a
valuable learning experience, as he expressed:
John: The group video was a very different way of assessing stuff. The class was
not too happy… cos basically we… they don’t think that maths should have a group
project… like I said, people have their own set ways of learning maths. (…) I kind
of enjoyed it although we did have a lot of deadlines in the last week and I know it
was very low on people’s priorities so they didn’t put much work as they could’ve
done. (…) I got a lot from it, we did a lot of like… general group work, working
with people, I learned how to put together a massive video in PowerPoint, we did
like a time-lapsed video of ice cream melting just to demonstrate Newton’s law of
cooling and… we did like interesting stuff and we did a lot of research into engi-
neering examples, how the concepts related to engineering, so it was good from that
point of view and expanded our knowledge in a lot of things.
Here it is worth noticing John’s use of “we” and “they” in describing his views
on the video project. His peers seem to be “not too happy” about it, and even though
he felt pressurised by deadlines, he “got a lot from it” by learning new skills that he
considers valuable in expanding his knowledge. This way of viewing things con-
trasts with that of Scott, in which the exam counts for almost everything.
5.5 Discussion and Conclusions
Scott and John came to university with similar good amounts of science capital and
strong mathematical knowledge and skills, such that it would have been expected
that they would do similarly well in their second-year mathematics course. However,
Scott obtained a much better grade than John (an A∗ as opposed to a C), and their
narratives in relation to the mathematical modelling components of their course dif-
fered in substantial ways. How can this difference be explained?
Using Bourdieu, one would explain that this difference is not about the cognitive
abilities of individuals, or only due to the capital that they bring into the field of
university education. Instead, the explanation is that one should look at the com-
plexity of the interactions between the rules of the field, what capital is valued in
this particular field, and, more important, how students negotiate their habitus in this
context, in order to understand why individuals do what they do and who gets to be
successful or not in a practice.
Hence, it seems that the rules of the field of mathematics education in this uni-
versity (and probably a considerable number of universities around the world) are
such that they benefit those students that, like Scott, are geared towards seeing
mathematics for its exchange value and disadvantage those students that, like John,
see mathematics mainly for its use value. Students like Scott have a better “feel for
the game”; they act “like fish in water” (Bourdieu and Wacquant 1992, p. 127). In
the course described in this chapter, where part of the practice emphasised the
development of modelling competences and the use value of mathematics through
relevant applications, Scott and “most of the people” in the course were “not too
happy” because some of the rules that they have learned to play were changed. The
mathematical modelling activities promoted values of understanding mathematical
concepts through their use in the engineering profession. These activities were not
primarily aimed at learning as preparation for a test. The modelling practice was
more in line with John’s habitus of enjoying the mathematics and seeing its use in
other parts of his degree, a more vocational disposition. If the field had valued the
aims of mathematical modelling, John would have been better positioned to become
a more successful student. But given that substantial parts of the course adhered to
“traditional” rules with the majority of the marks gained through the typical forms
of timed written examinations and the coursework video project only representing
15% of the total marks, Scott’s habitus was much better attuned than John’s to this
practice.
The results of this chapter point to an issue that many scholars have written
about: the performativity system under which the educational sector operates (Ball
2003) and which sees schools and universities disproportionately preoccupied by
league tables, number of student applications, number of graduates, etcetera at the
expense of other equally important variables such as deep learning, enjoyment of
learning, dispositions, etcetera. This might explain in part why mathematical mod-
elling practices are rare in the mathematics education field and why practices such
as teaching to the test, chalk-and-talk teaching, and drilling and memorising
(Popham 2001) are prevalent. These practices are particularly predominant in math-
ematics, which serves as the gatekeeper of scientifically oriented disciplines and
therefore has a special value in capitalist industrialist societies. Pais (2013) appro-
priately argues that:
The discourse of mathematics as important for everyday use conceals its real importance as
a testing and grading device. What is seen as direct property of object mathematics – its
utility – is indeed the result of the place mathematics occupies within the structure of capi-
talist economics. (p. 17)
One can conclude that if the educational field does not reward the use value of math-
ematics – that is, there is no time to prepare students for the exam and at the same
time teach them how mathematics is relevant for the real-world, for example,
through mathematical modelling – then students like John will likely struggle to
make their habitus resonate with the field. It is important to remember at this point
that, as Williams and Choudry (2016, p. 4) say, “the education system is actually a
well-tuned structure functioning to reproduce inequality and a classed society from
generation to generation”. The “game” of examinations is meant to produce suc-
cessful students at the expense of others who will eventually accept their fate of “not
being good at mathematics”. Furthermore, while one should not forget this fact,
knowing it makes it possible to change the system so that more students like John
can make their capital worth and become more successful at the educational game.
References
Archer, L., Dawson, E., DeWitt, J., Seakins, A., & Wong, B. (2015). “Science capital”: A con-
ceptual, methodological, and empirical argument for extending Bourdieusian notions of
capital beyond the arts. Journal of Research in Science Teaching, 52(7), 922–948. https://doi.
org/10.1002/tea.21227.
Ball, S. (2003). The teacher’s soul and the terrors of performativity. Journal of Education Policy,
18(2), 215–228. https://doi.org/10.1080/0268093022000043065.
Black, L., & Hernandez-Martinez, P. (2016). Re-thinking science capital: The role of ‘capi-
tal’ and ‘identity’ in mediating students’ engagement with mathematically demanding pro-
grammes at university. Teaching Mathematics and Its Applications, 35(3), 131–143. https://
doi.org/10.1093/teamat/hrw016.
Bourdieu, P. (1972). Outline of a theory of practice. Cambridge: Cambridge University Press.
Bourdieu, P. (1984). Distinction: A social critique of the judgement of taste. Cambridge, MA:
Harvard University Press.
Bourdieu, P. (1990). The logic of practice. Cambridge: Polity Press.
Bourdieu, P. (1998). Practical reason. Cambridge: Polity Press.
Bourdieu, P., & Passeron, J.-C. (1990). Reproduction in education, society and culture.
London: Sage.
Bourdieu, P., & Wacquant, L. (1992). An invitation to reflexive sociology. Chicago: The University
of Chicago Press.
Brown, J. P., & Stillman, G. A. (2017). Developing the roots of modelling concep-
tions: ‘Mathematical modelling is the life of the world’. International Journal of
Mathematical Education in Science and Technology, 48(3), 353–373. https://doi.org/10.108
0/0020739X.2016.1245875.
Bruner, J. (1997). Celebrating divergence: Piaget and Vygotsky. Human Development, 40(2), 63–73.
Galbraith, P., Stillman, G., Brown, J., & Edwards, I. (2007). Facilitating middle secondary model-
ling competencies. In C. Haines, P. Galbraith, W. Blum, & S. Khan (Eds.), Mathematical mod-
elling: Education, engineering and economics (pp. 130–140). Chichester: Horwood.
Kaiser, G. (2007). Modelling and modelling competencies in school. In C. Haines, P. Galbraith,
W. Blum, & S. Khan (Eds.), Mathematical modelling: Education, engineering and economics
(pp. 110–119). Chichester: Horwood.
Lawler, S. (2002). Narrative in social research. In T. May (Ed.), Qualitative research in action
(pp. 242–258). London: Sage.
Maaß, K. (2006). What are modelling competencies? ZDM Mathematics Education, 38(2),
113–142. https://doi.org/10.1007/BF02655885.
New Scientist. (2017). How much is your science capital worth? Retrieved from https://www.
newscientist.com/article/mg23531410-100-how-much-is-your-science-capital-worth/
Pais, A. (2013). An ideology critique of the use-value of mathematics. Educational Studies in
Mathematics, 84(1), 15–34. https://doi.org/10.1007/s10649-013-9484-4.
Popham, W. J. (2001). Teaching to the test? Educational Leadership, 58(6), 16–20.
Wacquant, L. (2011). Habitus as topic and tool: Reflections on becoming a prizefighter. Qualitative
Research in Psychology, 8(1), 81–92. https://doi.org/10.1080/14780887.2010.544176.
Williams, J., & Choudry, S. (2016). Mathematics capital in the educational field: Bourdieu and
beyond. Research in Mathematics Education, 18(1), 3–21. https://doi.org/10.1080/1479480
2.2016.1141113.
Williams, J., Black, L., Hernandez-Martinez, P., Davis, P., Pampaka, M., & Wake, G. (2009).
Repertoires of aspirations, narratives of identity and cultural models of mathematics in prac-
tice. In M. Cesar & K. Kumpulainen (Eds.), Social interactions in multicultural settings
(pp. 39–69). Rotterdam: Sense Publishers.
Willis, P. (1977). Learning to labour: How working class kids get working class jobs. Farnborough:
Saxon House.
Chapter 6
On Science Museums, Science Capital,
and the Public Understanding
of Mathematical Modelling
Pauline Vos
Abstract Students’ opportunities to learn informally (e.g. by watching documen-

taries, visiting museums) explain socio-economic inequities in school performances.
To explore informal learning about mathematical modelling, I studied two science
museums, as these are environments typically visited by middle-class families. I
framed the study by using the notions science capital and the public understanding
of mathematical modelling (PUMM) and explored how these are mediated in sci-
ence museums. The research method entailed observations of displays, artefacts,
and visitors. One science museum completely detached mathematics from its use
value, whereas the other offered histories of how people used mathematics to solve
society’s problems. This leads to recommendations for the design of, and research
on, environments for informal learning about mathematical modelling.
Keywords Bourdieu (theory of) · Informal learning (about mathematical

modelling) · Parental support (for mathematical modelling) · Meta-knowledge
(about mathematical modelling) · Public image (of mathematics and mathematical
modelling) · Public understanding (of mathematics and mathematical modelling) ·
Science capital · Science museums · Use value (of mathematics)
6.1 Introduction
The Organisation for Economic Cooperation and Development (OECD) reports

there are inequities that affect mathematics scores on the PISA-test (OECD 2014).
The report shows that students, whose parents are higher professionals, outperform
students whose parents are workers in elementary occupations. This phenomenon is
observed in all participating countries in PISA. Such socio-economic inequities are
P. Vos (*)
Department of Mathematics, University of Agder, Kristiansand, Norway
e-mail: pauline.vos@uia.no

64 P. Vos
caused by differences in the educational environment, as students whose parents are

higher professionals are more likely to attend schools with more resources and
higher qualified teachers. Yet, even if all students would learn within the same edu-
cational environment (same curriculum, same teacher, same tasks, etc.), the socio-
economic background plays a role in their uptake of, and interest in, science and
mathematics. Richer families can use their economic resources to create advan-
tages, for example, by paying for tutors. Also, the social background can empower
students as Archer et al. (2013) showed: within white, middle-class home environ-
ments, students are more likely to know people who work in science-related jobs
and learn from them about how science works. These acquaintances can be role
models in their ways of reasoning, explaining phenomena, or questioning causali-
ties. Also, in such home environments, students are more likely to be encouraged to
watch documentaries or visit science museums. Thus, socio-economic inequities in
science and mathematics scores are, among others, related to students’ informal
(out-of-school) learning opportunities.
Archer et al.’s (2013) research focused on the resources that can help less privi-
leged students to succeed in areas of the natural sciences. Black and Hernandez-
Martinez (2016) pointed out that Archer and colleagues did not refer to mathematics
and mathematical modelling. Yet, mathematics is a vital tool for solving many prob-
lems, not only in the natural sciences but also in the social sciences, business, and
so forth. Competencies in mathematical modelling assist students to succeed in
non-mathematical disciplines and in future professions. Therefore, educational
authorities throughout the world now advise schools and teachers to include math-
ematical modelling activities in their curricula (Kaiser 2014). Nevertheless, the per-
tinent issue of socio-economic inequity in mathematical modelling performances
remains. This inequity has been observed by OECD (2014) cited at the beginning of
this chapter, which is based on parental occupations and their relation to the scores
on a mathematics test, in which competencies were tested regarding the application
of mathematics in real-life situations.
To explore socio-economic inequities in mathematical modelling education, I
conducted a study on opportunities for students’ informal (outside-school) learning
of mathematical modelling. Informal learning comprises students’ learning at home
or with peers, when shopping, travelling, participating in sports, watching videos,
and so forth. Informal learning is unsystematic and unstandardised (Marsick 2009);
it happens ad hoc, not guided by explicit goals or a curriculum; it is social and
context-bound. Informal learning is erratic and happens in informal settings.
Therefore, it is hard to capture by scientific research methods. Current research on
informal learning consists of (1) studies on workplace-related learning; (2) studies
through questionnaires, in which students report on their hobbies, pets, and fre-
quency of reading science magazines and having assistance with homework (e.g.
Lin and Schunn 2016); and (3) studies on designed environments for informal learn-
ing, such as zoos and museums (e.g. Borun et al. 1996; Van Schijndel and Raijmakers
2016). In the latter, the researchers study how visitors interact with artefacts, infor-
mation, routing, and each other and what knowledge and dispositions they take
home from it. In the current study, I followed this line of research and focused on
6 On Science Museums, Science Capital, and the Public Understanding… 65
the messages that visitors of science museums get about mathematical modelling. I
chose to study science museums, because these are known to increase inequity
between students: a visit to a (science) museum is a typical white, middle-class lei-
sure activity (Archer et al. 2013). Also, when schools organise an excursion to a
science museum, it means that the school is situated in an affluent country and has
financial resources for the excursion.
6.2 Theoretical Frame
This study was framed by two concepts. The first concept is science capital, which
is based on the sociological theory of Pierre Bourdieu. For a more detailed explana-
tion of this theory, see Vos, Hernandez-Martinez, and Frejd (Chap. 3, this volume).
Here, it suffices to say that science capital is an extension of social capital (the
social network of people who will assist when asked for help) and cultural capital
(diplomas, knowledge of etiquette, access to information, etc.) (Bourdieu and
Wacquant 1992). Both social and cultural capital are valuable resources that people
own and accumulate, just like economic capital, and that can generate profits and
privileges. Extending the Bourdieusian theory, Archer et al. (2013, 2015) defined
science capital as the resources that offer advantages within scientific contexts, such
as science dispositions, science media consumption, parental scientific knowledge,
and so forth. Science capital can be accumulated in schools, but also out of school.
A typical example of science capital that some have, and others not, is a relative
who works in a science-related job and who can tell how science works. Another
example of how science capital can be accumulated is through a family visit to a
science museum. Archer et al. (2013, 2015) established that students with more sci-
ence capital are more likely to enter professions with science components (research,
engineering, etc.). In the present study, I took resources pertaining to mathematics
and mathematical modelling as being an integral part of science capital. This means
that cognitive and metacognitive modelling competencies (Galbraith et al. 2007;
Kaiser 2007, 2014), affect, and interest (Black and Hernandez-Martinez 2016;
Schukajlow et al. 2012) are included. It remains to be noted that science capital is a
conceptual construct for analytic research.
The second concept used in the present study is the public understanding of
mathematical modelling (PUMM). PUMM is an adaptation of the public under-
standing of science (PUS), which is the understanding, awareness, and engagement
of the general public of scientific knowledge and organisation (Bauer et al. 2007).
Many researchers of PUS work in Institutes for Communication Studies, and they
study how groups of people (e.g. shoppers in a supermarket, fishers in a coastal
region, readers of a certain newspaper) understand the complexities of science,
technology, and innovation and how they choose to use or disregard that knowledge
(e.g. Dash 2015).
PUMM is a similar construct as the public image of mathematics which is the
general public’s knowledge of, and about, mathematics. This public image of
66 P. Vos
mathematics is shaped on the one hand by traditional mathematics education with

meaningless and repetitive tasks, with alienating symbols, and so forth. On the other
hand, the public image of mathematics is shaped by dialectics of modern society’s
simultaneous mathematisation and de-mathematisation (Gellert and Jablonka 2007;
Keitel 2006). The mathematisation of society consists of an increased use of math-
ematics virtually anywhere, whereby mathematics is considered as value-free and
useful for establishing truths and making decisions. Simultaneously, there is a de-
mathematisation process, which is the process whereby mathematics becomes
increasingly invisible, being black boxed in technological devices. Thus, the public
image of mathematics has been studied, but PUMM has not been studied yet. There
are no studies yet on whether the general public knows the term “mathematical mod-
elling” at all, or whether certain groups of people have meta-knowledge of mathe-
matical modelling, which Brown and Stillman (2017) defined as “the background
knowledge (…) about the nature of modelling, how it is conducted and why mathe-
matics can be applied in real situations” (p. 357). Neither do we know, whether the
general public has experienced the usefulness of mathematics through mathematical
modelling activities. The current study is the first to explore PUMM, in particular
PUMM among middle-class families because of its focus on science museums.
In this study, science capital and PUMM are used as complementary analytic
concepts. They will be helpful in the following way. Science capital is the set of
resources that enable an individual to advance in science contexts (including in
mathematical modelling contexts). It is an individual person’s asset, whereby some
have more of it than others. Thus, science capital is at the microlevel, and it is used
to analyse differences between individuals. In contrast, PUMM is a social asset,
used to describe a certain knowledge at the macrolevel, being present in the public
domain. Science capital and PUMM are related, yet different. PUMM can hinder or
support individuals in their accumulation of science capital. Conversely, if many
people accumulate much science capital related to mathematical modelling, this
will improve PUMM.
The aim of the study was to explore both science capital related to mathematical
modelling and PUMM without striving for exhaustive descriptions. By focusing on
the informal learning regarding mathematical modelling that can occur in science
museums, I would be able to capture aspects of both. The research questions were:
What possible science capital related to mathematical modelling can students (in
the school going age) accumulate in science museums? What possible PUMM can
science museums generate?
6.3 Methods
To answer the research questions, I selected the science museums/centres on the

following criteria: (1) to keep cohesion in the study, the visits were to take place
within a few months; (2) to have a fresh eye, I should not have visited the museums
before; and (3) to limit bias, I should not have prior professional engagement with
them (as consultant). This resulted in visits to the Experimentarium in Copenhagen
(Denmark) and the London Science Museum (UK). On both occasions, I went
together with a teacher trainer (respectively, a mathematics teacher trainer and a
primary school teacher trainer). Generally, a visit to a science museum will take
more than a day, and one undertakes a spontaneous routing. To stay concentrated on
the research, I limited our visits to two hours and to only those museum sections
tagged with the keyword mathematics on the official website (the search word
“model∗” led to a section on wax modelling).
Similar to the research approaches of other science museum researchers (e.g.
Borun et al. 1996; Van Schijndel and Raijmakers 2016), I made field notes, photo-
graphs, and short videos. I registered the environment, the information on displays
and in videos, and the activities for visitors. The visitors present during our stay
were considered as a sample of convenience. Without disturbing them, I observed
their activities and the time they stayed. I did not ask their permission, since I did
not breach their privacy. Additionally, I asked my colleague to read the displays,
watch the videos, and engage in the activities as if it was a regular visit. Afterwards,
we discussed the visit and I wrote a report. I analysed these from two angles. First,
I used the concept of science capital related to mathematical modelling, to identify
resources, which can offer advantages to an individual in future science endeavours.
Second, I used the concept of PUMM to identify messages to the general public on
understanding mathematical modelling.
6.4 Results
In this section, I will first report on the observations made in the Experimentarium
in Copenhagen (Denmark) and then of those made in the London Science
Museum (UK).
When we visited the Experimentarium in Copenhagen in June 2017, we were
surprised by its brand-new architecture, design of artefacts, and activities. In fact,
this centre had opened only 5 months earlier (in January 2017). At the moment of
our visit, on a Friday morning, the visitors consisted mainly of students aged
8–14 years old. Most came in a school excursion accompanied by a few teachers,
which resulted in groups of 2–5 students roaming the centre unaccompanied by an
adult. According to their website www.experimentarium.dk, there were two sec-
tions tagged as being about mathematics: Bubblearium and The Solver (see Fig. 6.1).
In the section Bubblearium, the visitors were invited to create soap bubbles,
which was exciting as judged from the visitors’ noise. Particularly attractive were
the rings to create a cylinder around a person. However, after having made a few
soap shapes, and watching others make these, all students left this section. Texts on
the wall and three bubble machines (two of which were out of order) were designed
with the intent to make students explore combined bubbles and the reflection in
bubble surfaces. However, few students did this, and only if induced by an adult. In
68 P. Vos
Fig. 6.1 Experimentarium Copenhagen, sections Bubblearium (left, © Pauline Vos) and The
Solver (right, © Pauline Vos)
this section, the students’ activities were haphazard and aimless, also because the
shapes were not stable and disappeared after a few seconds into the air. The maxi-
mum time that visitors spent here was three min.
The second section in the Experimentarium tagged as mathematical was The
Solver. In the middle was a labyrinth painted on the floor, on which one could walk.
This labyrinth was surrounded by tables on which there were physical puzzles con-
sisting of wooden or plastic pieces (spheres, blocks, and other shapes). These were
to be put together (tiled, stacked) or to be separated from one another. All objects
were attractively colourful and tangible. In the time span of an hour, few visitors
came to this section; it was not noisy at all. The students, who entered this section,
tried one or two puzzles and left after a few unsuccessful attempts within two min-
utes. Only two boys stayed longer than half an hour, seated at one table being fully
absorbed in a puzzle.
Both sections in the Experimentarium in Copenhagen connected to mathematical
shapes, aimed at showing that these shapes can create wonder and inquiry. When
analysing in light of science capital related to mathematical modelling, in this sci-
ence centre visitors can learn to enjoy mathematics. However, the exciting artefacts
and activities showed a type of mathematics that has no use value for solving prob-
lems in real life. We, as visitors with a background in mathematics, were able to
recognise the mathematics behind the puzzles and bubbles, but there were no indi-
cations that the students could. The sections connected neither to traditional math-
ematics education nor to mathematical modelling. As such, the sections assisted
students in accumulating a certain science capital, but no capital related to mathe-
matical modelling. Analysing the sections in light of PUMM, we can but observe
that they did not show the usefulness of mathematics and they propagate a public
image of mathematics as detached from real life. By including sections tagged as
mathematics, the Experimentarium clearly intends to make mathematics visible to

its visitors, but its implementation keeps mathematical modelling invisible and
makes PUMM void.
The second science museum visited in this study was the London Science
Museum (UK). We were there on a Friday morning in May 2017. Only one section,
the Winton Gallery, was tagged as being about mathematics according to the web-
site (www.sciencemuseum.org.uk). It was designed by architect Zaha Hadid, who
studied mathematics before turning to architecture. This section opened in December
2016. When we entered it, we were caught by violet curls hanging from the ceiling,
which surrounded an antique aircraft, described as an authentic 1929 Handley Page
Gugnunc. Displays and a video explained that the violet curls show the air flow
around the aircraft in flight (see Fig. 6.2). It was also explained that aircraft engi-
neers needed the Navier-Stokes equations to better understand the dangers of flying
and that mathematics is needed to make flying an aircraft safer.
Surrounding the airplane, there were thematic exhibits telling histories of how
people used mathematics for social purposes. For example, there was the story of
Florence Nightingale and how she visualised mortality statistics in the Crimean War
(1853–1856) to convince political and military leaders that lack of hygiene killed
more soldiers than the enemy. Another example was the story of the flooding disas-
ter of 1953 in England and how Winston Churchill asked mathematicians for better
weather and tide predictions. Behind glass was the authentic tide prediction machine
made by Lord Kelvin, which yielded the tides a year in advance after four hours of
cranking its handle. Another authentic, antique object displayed in the Winton
Fig. 6.2 Curved shapes showing airflow around an aircraft (© Zaha Hadid Architects)
70 P. Vos
Gallery enabled conversion of weights in international trading. It was a cabinet with

96 drawers, each holding the weights from a place overseas.
The Winton Gallery informed us, as visitors, through narrative displays, videos,
and artefacts in glass showcases, of which the authenticity was clarified through
texts explaining their source (Vos 2015). The histories were about people who used
mathematics to solve real-life problems; the terms “model” and “modelling” were
explicitly used. Visitors were shown formulae, yet did not need to do mathematics.
There was much to read and nothing to be touched. Unlike the Experimentarium in
Copenhagen, there were no tangible objects or exciting activities. During our visit,
I observed only adults in the Winton Gallery, and none stayed longer than 15 min.
Although the museum attracted younger visitors, they rather went to the flight simu-
lators elsewhere in the museum.
When looking through the lens of science capital related to mathematical model-
ling, the Winton Gallery offered role models of women and men who solved real-
life problems by using mathematics. The term “mathematical model” was explicitly
used; the texts offered insights into the purposes of mathematical modelling. Thus,
the science capital that students can accumulate here is similar to what they can
gather from science documentaries related to mathematical modelling: this science
capital can increase an individual’s dispositions to, and knowledge about, mathe-
matical modelling through raising curiosity and interest. However, modelling as an
activity remains vague, as the visitors cannot experience modelling activities them-
selves. Also, the objects were untouchable and the explanations were rather verbal;
both can be considered less attractive for students, whom we want to become com-
petent modellers. Looking through the lens of PUMM, the Winton Gallery con-
veyed a message that mathematics is important because of its use value. We were
told histories of how women and men struggled with the creation of mathematical
models to solve problems that mattered to their society. As such, this gallery can add
to PUMM that mathematical modelling is a human activity and it serves social pur-
poses, such as keeping flight passengers safe and reducing death toll in wars.
6.5 Conclusion, Discussion, and Recommendation
The current study opened a window on informal learning of mathematical model-

ling, inspired by research that observed that informal learning is a source for socio-
economic inequities in mathematical modelling education. Students who have
access to out-of-school resources pertaining to mathematical modelling, for exam-
ple, through parental support, will be more likely to succeed in it. To study informal
learning, I explored two science museums, because these are considered typical
environments for informal learning, in particular by middle-class families (Archer
et al. 2013, 2015). To frame the study, I used two concepts: (1) the concept of sci-
ence capital to analyse out-of-school resources for students that can give them
advantages when it comes to mathematical modelling and (2) the public under-
standing of mathematical modelling (PUMM), which captures how the general
public conceptualises mathematical modelling. The research questions were: what

possible science capital related to mathematical modelling can students accumulate
in science museums? What possible PUMM can science museums generate?
When looking through the lens of science capital related to mathematical model-
ling, the study yielded a mixed picture. On the one hand, a science museum can be
like the Experimentarium in Copenhagen (Denmark) and focus on offering activi-
ties on bubbles (attractive to many, but only for a short time) and puzzles (only
attractive to a few) that connect to an esoteric mathematics detached from real life.
In this case, the use value of mathematics for solving real-life problems was not
aimed for and thus ignored. Such a science museum offers many experiences to the
middle-class children coming there, but no science capital related to mathematical
modelling. On the other hand, a science museum can be like the London Science
Museum (UK) and include a section that strongly focuses on the use value of math-
ematics for solving social problems. This science museum offered histories of
mathematical modellers and offered insights into the purposes of mathematical
modelling, thereby enhancing science capital (science dispositions and knowledge).
It remains to be noted that the science museum of the first kind offered kinaesthetic
experiences and was entertaining to young students, whereas the second merely
presented factual knowledge, required a lot of reading effort, and did not offer
appealing activities to young students. Thus, in neither of the museums, students
would be able to accumulate science capital related to mathematical modelling,
although in the London Museum it was accessible to visitors receptive to narratives
and antique objects. This answers the first research question.
When looking through the lens of PUMM, the study yielded a different picture.
We saw that a science museum like the Experimentarium can choose to offer attrac-
tive activities that people with a mathematical background will recognise as being
connected to mathematics, yet which a general, non-specialised public will only
experience as shapes for entertainment and detached from real life. So, the PUMM
of middle-class families will not be enhanced. On the other hand, a science museum
can include a section like the Winton Gallery with a strong focus on the use value of
mathematics for solving social problems and offer meta-knowledge about mathe-
matical modelling through personal stories and authentic artefacts. It adds to the
PUMM of middle-class visitors by explicitly using the terms mathematical models
and modelling and that these serve humankind. Thus, one science museum may not
add to PUMM at all, whereas another can. However, neither of the visited science
museums gave visitors the opportunity to experience the usefulness of mathematics
by engaging in modelling activities, and so the potential PUMM observed in this
study was of the meta-knowledge type. This answers the second research question.
A number of issues arise from this study. A first issue is the difference between
propagating mathematics versus propagating mathematical modelling. A science
museum can focus on mathematical modelling as a human activity for solving social
problems, as done in a narrative way in the Winton Gallery in the London Science
Museum. In contrast, a science museum can also propagate mathematics as detached
from real life and add to the invisibility of mathematical modelling. Second, both
museums contributed to the process of de-mathematisation (Gellert and Jablonka
72 P. Vos
2007) with sections on kinematics, commerce, and other themes that contained a lot
of invisible mathematical modelling. Third, it remains an open question how muse-
ums intending to explicate mathematical modelling can include tangible artefacts
and exciting activities, in which young visitors can actively engage.
As science museums assist students to accumulate science capital, and as they
increase inequity between students, we need to consider (1) making them more
accessible to less privileged students and (2) how to connect these better to mathe-
matical modelling. To decrease inequity, it is pertinent to establish such institutions
in less affluent countries, ask governments to reduce entrance fees, assist lower-
class parents to understand their importance, and so forth. Also, the resources from
science museums could become more universally available through digital media.
As for the point to better connect science museums to mathematical modelling for
students in the school-going age, we may learn from both museums in this study.
The Experimentarium was more successful in offering excitement and inquiry
activities, whereas the Winton Gallery was more successful in showing the useful-
ness of mathematics and even used the term mathematical modelling. Therefore, it
is recommended to carry out further research into (1) how science museums/centres
can combine excitement and inquiry activities without detaching mathematics from
real life and its use value for solving real-life problems, (2) what other environments
enhance students’ informal learning about mathematical modelling (and thus their
science capital), and (3) how PUMM can further be studied and enhanced.
References
Archer, L., Osborne, J., DeWitt, J., Dillon, J., Wong, B., & Willis, B. (2013). ASPIRES: Young
people’s science and career aspirations, age 10–14. London: King’s College London.
Archer, L., Dawson, E., DeWitt, J., Seakins, A., & Wong, B. (2015). “Science capital”: A concep-
tual, methodological, and empirical argument for extending Bourdieusian notions of capital
beyond the arts. Journal of Research in Science Teaching, 52(7), 922–948.
Bauer, M. W., Allum, N., & Miller, S. (2007). What can we learn from 25 years of PUS survey
research? Liberating and expanding the agenda. Public Understanding of Science, 16(1), 79–95.
Black, L., & Hernandez-Martinez, P. (2016). Re-thinking science capital: The role of ‘capital’ and
‘identity’ in mediating students’ engagement with mathematically demanding programmes at
university. Teaching Mathematics and Its Applications, 35, 131–143.
Borun, M., Chambers, M., & Cleghorn, A. (1996). Families are learning in science museums.
Curator: The Museum Journal, 39(2), 123–138.
Bourdieu, P., & Wacquant, L. J. (1992). An invitation to reflexive sociology. Chicago: University
of Chicago Press.
Brown, J. P., & Stillman, G. A. (2017). Developing the roots of modelling conceptions:
‘Mathematical modelling is the life of the world’. International Journal of Mathematical
Education in Science and Technology, 48(3), 353–373.
Dash, B. (2015). Public understanding of cyclone warning in India: Can wind be predicted? Public
Understanding of Science, 24(8), 970–987.
Galbraith, P., Stillman, G., Brown, J., & Edwards, I. (2007). Facilitating middle secondary model-
ling competencies. In C. Haines, P. Galbraith, W. Blum, & S. Khan (Eds.), Mathematical mod-
elling: Education, engineering and economics (pp. 130–140). Chichester: Horwood.
Gellert, U., & Jablonka, E. (2007). Mathematization and demathematization: Social, philosophi-
cal and educational ramifications. Rotterdam: Sense.
Kaiser, G. (2014). Mathematical modelling and applications in education. In S. Lerman (Ed.),
Encyclopedia of mathematics education (pp. 396–404). Dordrecht: Springer.
Keitel, C. (2006). Mathematics, knowledge and political power. In J. Maaß & W. Schlöglmann
(Eds.), New mathematics education research and practice (pp. 11–22). Rotterdam: Sense.
Lin, P. Y., & Schunn, C. D. (2016). The dimensions and impact of informal science learning expe-
riences on middle schoolers’ attitudes in science. International Journal of Science Education,
38(17), 2551–2572.
Marsick, V. J. (2009). Toward a unifying framework to support informal learning theory, research
and practice. Journal of Workplace Learning, 21(4), 265–275.
OECD. (2014). Do parents’ occupations have an impact on student performance? PISA in Focus, 36.
Schukajlow, S., Leiss, D., Pekrun, R., Blum, W., Müller, M., & Messner, R. (2012). Teaching
methods for modelling problems and students’ task-specific enjoyment, value, interest and self-
efficacy expectations. Educational Studies in Mathematics, 79(2), 215–237.
Van Schijndel, T. J., & Raijmakers, M. E. (2016). Parent explanation and preschoolers’ exploratory
behavior and learning in a shadow exhibition. Science Education, 100(1), 153–178.
Vos, P. (2015). Authenticity in extra-curricular mathematics activities; researching authenticity
as a social construct. In G. A. Stillman, W. Blum, & M. S. Biembengut (Eds.), Mathematical
modelling in education research and practice: Cultural, social and cognitive influences
Chapter 7
Investigating the Complexity of Student
Modelling by Coordinating a Diversity
of Research Perspectives
Corey Brady, Jeffrey A. McLean, Hyunyi Jung, Aran W. Glancy,

and Ángeles Domínguez
Abstract In this chapter, we present a collaborative, post hoc analysis of a pivotal

episode in the work of one group of USA middle-school students as they engaged
in a model-eliciting activity. A central feature of this collaborative analysis was the
identification and elaboration of three perspectives through iterative interpretation
of video data: representations, mathematical uncertainty, and idea development.
Coordinating these perspectives highlighted a cyclical pattern in the students’ pro-
gression through the modelling activity, as they encountered and managed mathe-
matical uncertainty and expressed their developing ideas with physical materials
and representations. We view this work as a researcher-level modelling effort, to
synthesise independent analytical approaches, develop shared perspectives, and
construct an explanatory account of the flow of students’ collective thinking as they
solved a rich modelling task.
C. Brady (*)
Department of Teaching and Learning, Vanderbilt University, Nashville, TN, USA
e-mail: corey.brady@vanderbilt.edu
J. A. McLean
Statistics and Operations Research, University of North Carolina at Chapel Hill,
Chapel Hill, NC, USA
e-mail: mclean@unc.edu
H. Jung
Department of Mathematical and Statistical Sciences, Marquette University,
Milwaukee, WI, USA
e-mail: hyunyi.jung@marquette.edu
A. W. Glancy
Engineering Education, Purdue University, West Lafayette, IN, USA
e-mail: aglancy@purdue.edu
Á. Domínguez
School of Medicine and Health Sciences, Tecnologico de Monterrey, Monterrey, Mexico
e-mail: angeles.dominguez@tec.mx

76 C. Brady et al.
Keywords Idea development · Models and modelling perspective · Model-

eliciting activity · Representations · Mathematical uncertainty
7.1 Introduction
Modelling in the context of realistic, open problems, such as model-eliciting activi-

ties, or MEAs (Lesh et al. 2000), provokes nuanced and multilayered mathematical
learning and activity (English and Gainsburg 2016; Kaiser et al. 2011; Kaiser and
Sriraman 2006). In such settings, groups of students approach realistically complex
problems, bringing to bear widely varying personal and mathematical knowledge
and insights and diverse ways of thinking (English et al. 2016). Their ideas, actions,
and interactions in such environments are often intense and multifaceted. Moreover,
as researchers of learning, we consider this complexity is an important feature,
rather than a “bug” or flaw: in fact, we design modelling activities to heighten the
diversity of thinking and the complexity of interpersonal interactions during model-
ling (English et al. 2008). Thus, the data record from such activities can involve
student discourse and interaction dynamics that are rich and challenging to analyse.
Complexity also appears in the mathematical concepts and practices that arise in
student solutions to modelling problems. Unlike less context-rich application prob-
lems, MEAs and other realistic modelling tasks demand that learners interpret real-
world situations so as to mathematise key aspects of them. This typically involves
engaging with more than one “textbook topic”, as learners fashion situation-specific
constructs to make sense of the problem (Brady et al. 2018). As interaction com-
plexity meets disciplinary complexity in realistic modelling, there thus also is a
legitimate concern that any single research perspective may fail to do justice to the
phenomena and experiences of learning in these settings.
Assembling large-scale, multitier collaborative projects to study modelling is a
suitable response to this complexity challenge, and this strategy has demonstrated
its value (Lesh et al. 2008). However, orchestrating and funding research of this
nature can be an enormous enterprise. To conduct exploratory design-based research,
or DBR (Cobb et al. 2003), that can lead to such studies, an approach is needed that
can allow researchers who do not yet have experience with multitier design research
in modelling to assemble perspectives in ways that provoke and illuminate emergent
researchable questions about the thinking and idea development of groups of learn-
ers engaged in modelling tasks.
In this chapter, we describe ongoing work to support the emergent development
of collaborative research on modelling that integrates diverse researcher-level per-
spectives. We asked: (a) How can researchers bootstrap collaborative research by
engaging in analyses of pre-existing data? and (b) How does this variety of research
perspectives illuminate the complexity of students’ thinking?
7 Investigating the Complexity of Student Modelling by Coordinating a Diversity… 77
7.2 Theoretical Framework
Our exploratory analysis focused on three broad perspectives, namely, representa-

tions, mathematical uncertainty, and idea development, that arose from video data
that are sampled in this chapter. These perspectives were not firmly established in
advance; rather they emerged in the authors’ interaction with the particular data and
then were aligned with the literature. Nevertheless, to orient the reader, in this sec-
tion we provide an overview of each perspective ahead of the analysis.
7.2.1 Representations
The representations perspective draws on the centrality of externalising ideas in the

models and modelling perspective (Lesh and Kelly 2000). A representation is
broadly defined as “a configuration of some kind that, as a whole or part by part,
corresponds to, is referentially associated with, stands for, symbolizes, interacts in
a special manner with, or otherwise represents something else” (Palmer 1977,
p. 398). Building on this definition, Goldin and Kaput (1996) distinguished internal
representations from external ones: the former refers to mental configurations in
learners, and the latter refers to observable configurations, including words, graphs,
symbols, and tools. Similarly, Lesh and Kelly (2000) offered examples of “repre-
sentation systems”, which are “systems of written symbols, systems of graphic
images, or systems involving concrete manipulatives, experience-based metaphors,
or spoken language” (p. 210).
7.2.2 Mathematical Uncertainty
Our uncertainty perspective is focused on learners as they confront the conceptual

challenge of creating a new mathematical model. Models are conceptual systems
that humans construct and use to make sense of the world around them and to oper-
ate effectively in, and on, that world. A model thus involves both a way of seeing the
world (an interpretive system) and a set of operations (possibly implicit) that can be
brought to bear on the world. When encountering a situation where ready-to-hand
models break down and we must construct a new one, we are thrown into a state of
uncertainty (cf Heidegger 1927/1962). In such situations, groups of learners repur-
pose elements of mental constructs in a collaborative bricolage. The group typically
brings several (possibly competing) candidate model fragments to bear, which sur-
face in their discussion as different ways of seeing and thinking. In the cyclical
modelling process characteristic of MEAs, limitations in models can provoke a
78 C. Brady et al.
sequence of moments of breakdown, driving model construction, as well as iterative

refinements or syntheses of the group’s models, leading to more sophisticated solu-
tions to the problem. In this perspective, we hypothesised that learners’ recognition
of this breakdown process might show itself in discourse and actions, in ways that
we tentatively called “mathematical uncertainty”. Similar uses of uncertainty in the
literature (e.g. Engle 2012; Zaslavsky 2005) and their compatibility with our
approach encouraged us to solidify this perspective.
7.2.3 Idea Development
The perspective we labelled idea development is concerned with tracking the math-
ematical concepts that students identify as resources for a problem and how these
ideas evolve in interaction during problem solving. This perspective was built on
two main premises: (1) conceptual understanding can be described and understood
through the Lesh translation model (Lesh and Doerr 2003), where rich concepts can
manifest themselves in a variety of representations (real world, symbolic, pictorial,
language based, etc.) and where understanding is marked by one’s ability to trans-
late between representations of the concept both within and between representa-
tional modes (cf., Moore et al. 2013), and (2) significant development of conceptual
models can occur when those models are placed in contact with the real world (Lesh
and Harel 2003).
7.3 Methodology
Our goal was to explore the potential to use preexisting modelling data to provoke
interactions between researchers’ perspectives on modelling and to identify the
grounds for pursuing future collaborative research. As such, we describe here our
methods at three levels: first, the structure of the MEA that was the basis for our
research; second, the means by which we collected the student-level data that we
used to stimulate exploratory analyses; and third, the procedure we used to orches-
trate researcher level engagement with these data.
7.3.1 Description of the Problem
The Pelican Colony problem (Moore et al. 2015; Pompei 2010) focuses on a popu-
lation of pelicans in North America. This bird species was considered endangered,
and the US Fish and Wildlife Service aimed to protect the birds to prevent their
becoming extinct. A client letter explains that the service does not have the resources
to protect all pelicans and must determine which pelican breeding grounds will be
the most effective target for their conservation efforts. Teams of students are asked
to aid the US Fish and Wildlife Service in selecting suitable pelican colonies by
developing a procedure for estimating the number of pelican nests in a breeding
ground. The US Fish and Wildlife Service provides maps of two irregularly shaped
breeding grounds and corresponding aerial photographs of the pelicans in small
subregions of each colony. Students are asked to write a letter to the US Fish and
Wildlife Service describing their procedure for estimating the number of pelican
nests in a colony, including how they have tested their procedure, giving estimates
for the number of pelicans in each of two colonies, declaring assumptions that they
have made, and making any requests for additional information that might help to
improve their procedure.
7.3.2 Student Level
The video data we analysed came from the EngrTEAMS project (NSF #1238140)
and involved sixth-grade students at a suburban middle school in Midwestern
USA. According to state-wide standardised tests, 43% of students in the school
were considered proficient in mathematics, and 54% were proficient in reading at
the time of the study. The class that participated in the Pelican Colony MEA was a
regular-track sixth-grade mathematics class, and the MEA was implemented near
the end of the academic school year. The teacher in the room had over 10 years of
experience teaching middle-school mathematics; however, this was only the second
year in which she had added MEAs to her curriculum. This addition to the curricu-
lum was supported by one of the authors of this chapter (Glancy). The analysis
sampled in this chapter focuses on one group of students chosen from the entire
class of participants.
7.3.3 Researcher Level
We independently viewed the video footage with an open attitude towards determin-
ing which of our research perspectives or lenses might be most productive, recon-
vening afterwards for an interpretive discussion. At the end of this process, we
narrowed our attention to two student groups’ work on the MEA, and we chose to
focus on three analytic perspectives: representation, mathematical uncertainty, and
idea development. Each researcher “adopted” a perspective, so that one or more
researchers were associated with each perspective.
In the next viewing of the two groups’ work, researchers independently coded
the data, using the utterance as a unit of analysis in student discourse and action
with materials as a unit of analysis in nonverbal activity. We identified moments in
the video record where our perspective simply applied as a code, as well as moments
where it appeared to be pivotal in making sense of the students’ problem-solving
80 C. Brady et al.
work. After this coding pass, we used the video timestamps to merge our coding in
a shared spreadsheet. We then met to discuss the merged artefact and identify epi-
sodes in the students’ modelling work, defined in terms of the flow of their problem
solving, the density of codes of different kinds, and the occurrences of moments
marked as pivotal by a perspective.
The next phase of work involved reflexive interaction between the definitions of
the perspectives and each researcher’s evolving account of the learner’s activity
anchored in a single perspective. Our interpretations of the data and our articulation
of the perspectives’ definitions thus contributed mutually to each other’s contribu-
tion. At this point, we selected a single episode in a single student group’s modelling
work as a focal passage. We prepared a “brief” from each perspective and met to
negotiate a collaborative interpretation of these learners’ work in that passage over-
all. We identified a rich set of links between the facets of student behaviour high-
lighted by each perspective, which gave rise to enriched models of the ways that the
students’ ways of thinking developed over the episodes, through collaboration,
material interaction, and reflection on mathematical uncertainties that arose in
their work.
7.4 Analysis
In this section, we provide an analysis of one of the student groups’ work at a pivotal
point in their problem solving—one of several episodes highlighted by our coding
of the video data according to our three perspectives.
At the beginning of this episode, a team of four students attempted to compare
the size of two colonies and appeared to want to use direct comparison of their
shapes. To do this, they made a representational choice, using transparencies and
markers to capture the outline of the colonies on the photographs and superimpose
them (Fig. 7.1). However, on overlaying them, they found that the comparison was
not definitive.
Fig. 7.1 Traced and

superimposed pelican
colony sitemaps (lighter
is green)
S2: How are we supposed to tell which one is bigger, because this one is longer
but this one is like…well…?
S2: I think the green one [Colony A] is bigger.
S1: What?
S2: I think that Colony A is bigger. But this one is just longer. ‘Cause this one
is bigger.
S1: You’ve got to measure them up to the areas that would fit.
S2: There’s something [inaudible] It says…
S1: So each one is 10 feet at the bottom. 10 metres.
S3: Colony B is bigger but…
S2: I think Colony A is bigger.
S1: I mean Colony B is longer, but …
S1: I don’t know which one because, even though Colony B is longer, it could
still be bigger.
S2: Well yeah, but it’s longer. But this one is bigger and then … I’m not sure.
I just think that Colony A is bigger.
S1: I don’t know which one because, even though Colony B is longer, it could
still be bigger.
S2: Well yeah, but it’s longer. But this one is bigger and then … I’m not sure.
I just think that Colony A is bigger.
T: Okay, so what is the plan?
S3: Well, we’re going to use the rice for birds, but we’re trying to figure out,
we’re just matching them up to the colonies and stuff, to see which one
would be bigger. But colony B is longer, but um, colony um A is taller.
S2: So I think that colony A would be bigger because it’s like more spread out,
but right here it’s not. Wait actually, I think Colony B is because this is
more like, close in, and then this is all like spread out and then [inaudible]
T: Okay, but how will you convince if one is bigger than the other? How will
you know? How will you know that?
S2: I think this one is small, it like, is smaller right here and it’s shorter. And
this is longer and it’s not spread out there.
T: Okay, but then the question is how many nests are there. It’s not just about
which area is bigger but how many nests are there, so what are you to do?
From an uncertainty perspective, it appears the students began to work under the
assumption that the area or physical extent of the colonies could be directly com-
pared in a straightforward way; but when neither outline encompassed the other, a
variety of different perceptions competed in their conception of “bigger”. In rapid
succession, the students alternately noticed and attended to the lengths and heights
of the two colony outlines.
Thus, whereas initially “bigger” was treated as an unproblematic and single
property of a shape, in this excerpt they began to become aware of the need to coor-
dinate different “size” dimensions. Finding their path blocked, they reframed the
problem and pushed forward the development of new ideas.
82 C. Brady et al.
Fig. 7.2 Rice-covered

pelican colony sitemap
Perhaps provoked to reflect on their process by the teacher’s arrival and question,
the team’s conversations shifted from pure assessments of length and height—an
approach to area only appropriate for regular, even rectangular, shapes—to ideas
more appropriate for investigating the area of irregular shapes (e.g. comparing cross-
sectional heights of the colonies at different points). At this point in their work, the
teacher then highlighted a key feature of the problem statement, cueing the group to
focus on the number of pelican nests in the colonies, rather than (only) their area.
Faced with uncertainty about both the feasibility of their area comparison work and
its relevance to the problem, the students shifted their approach. At this point, they
also began to act on a significant representational choice, using grains of rice as a
tool to denote pelican nests (Fig. 7.2). The students did not (yet) take into account
the relative density of the nests in the two colonies; rather, they seemed to investigate
how many nests could fit in the space occupied by the colony. This created continu-
ity between their initial approach and their new direction. Covering their colony
contours with rice, they then attempted to estimate the number of grains. Their ideas
about how to compare sizes of colonies as quantities thus transitioned from calculat-
ing the area of a colony’s shape to counting the number of nests/grains that fit inside.
As a collaborative researcher group, we brought our diverse ways of thinking about

modelling to bear on preexisting video data of groups of students engaged in a mod-
elling task. As we worked to make sense of the video record, several coherent ana-
lytical perspectives emerged, which were related to our independent research
approaches but also responsive to what we were collectively seeing in the data. We
view this as a researcher-level modelling process, analogous in some ways to the
work that the groups of students were doing as they solved the Pelican Colony
problem. We discuss three ways that our perspectives expressed and shaped our
emerging understanding of students’ modelling work.
First, our perspectives interacted to guide the selection of episodes and our ulti-
mate focus on the data presented here. As we discussed our independent analyses of
video data, node points in students’ modelling work where multiple themes con-
verged became salient. In our early discussions, we considered additional perspec-
tives, related to teacher facilitation and to group dynamics, but we did not find they
identified points of resonance in conjunction with the other three perspectives.
Second, in organising our insights about students’ thoughts and actions, these
perspectives structured our analysis of a pivotal episode in one group’s work towards
a solution to the problem. Here, identifying relationships between our perspectives
offered a model of the students’ flow of thought. As a result, idea development took
centre stage and structured the “story” of the student work that emerged. (This view
of their work is different from, for example, a narrative foregrounding of students’
collaboration dynamics.) Our initial analysis revealed a cyclical pattern in the
group’s progression through the activity, which was highlighted by our attention to
the flow and evolution of ideas. Instances of mathematical uncertainty became
salient as interruptions in the flow of the group’s work, with representations mediat-
ing changes in the ways of thinking that were dominant in the group’s discourse.
These transitions also corresponded with pivotal moments of idea development.
Third, reflecting on our analyses of students’ interaction allowed us to identify
tentative models describing more general conceptual relations between them. For
instance, we realised that adopting an idea development perspective led us to attend
to connections students made between mathematical concepts in their talk and their
actions with representational media and materials. The mathematical uncertainty
perspective helped us to document key instances of “indecision” and “crisis” that
sparked changes in students’ strategies and led to new interactions with the materi-
als and the problem context. Additionally, the representations perspective high-
lighted for us how both of these aspects of learners’ modelling behaviour were
mediated by materials and representations as tools that permitted them to invent
measurement techniques and transfer measurements (e.g. the use of transparency
sheets and grains of rice, above).
References
Brady, C., Eames, C., & Lesh, R. (2018). The student experience of model development activi-
ties: Going beyond correctness to meet a client’s needs. In S. Schukajlow & W. Blum (Eds.),
Evaluierte Lernumgebungen zum Modellieren (pp. 73–92). Wiesbaden: Springer Spektrum.
Cobb, P., Confrey, J., DiSessa, A., Lehrer, R., & Schauble, L. (2003). Design experiments in edu-
cational research. Educational Researcher, 32(1), 9–13.
Engle, R. A. (2012). The productive disciplinary engagement framework: Origins, key concepts,
and developments. In D. Y. Dai (Ed.), Design research on learning and thinking in educational
settings (pp. 170–209). New York: Routledge.
English, L. D., & Gainsburg, J. (2016). Problem solving in a 21st-century mathematics curriculum.
In L. D. English & D. Kirshner (Eds.), Handbook of international research in mathematics
education (3rd ed., pp. 313–335). New York: Taylor & Francis.
84 C. Brady et al.
English, L. D., Jones, G. A., Bartolini Bussi, M. G., Lesh, R., Sriraman, B., & Tirosh, D. (2008).
Moving forward in international mathematics education research. In L. D. English (Ed.),
Handbook of international research in mathematics education: Directions for the 21st century
(2nd ed., pp. 872–905). New York: Routledge.
English, L. D., Ärlebäck, J. B., & Mousoulides, N. (2016). Reflections on progress in mathematical
modelling research. In A. Gutiérrez, G. C. Leder, & P. Boero (Eds.), The second handbook of
research on the psychology of mathematics education (pp. 383–413). Rotterdam: Sense.
Goldin, G., & Kaput, J. (1996). A joint perspective on the idea of representation in learning and
doing mathematics. In L. Steffe, P. Nesher, P. Cobb, G. Goldin, & B. Greer (Eds.), Theories of
mathematical learning (pp. 397–430). Hillsdale, NJ: Lawrence Erlbaum.
Heidegger, M. (1962). Being and time (J. Macquarrie & E. Robinson, trans.). New York: Harper &
Row. (Original work published 1927.)
Kaiser, G., & Sriraman, B. (2006). A global survey of international perspectives on modelling in
mathematics education. ZDM Mathematics Education, 38(3), 302–310.
Kaiser, G., Blum, W., Borromreo Ferri, R., & Stillman, G. (2011). Trends in teaching and learn-
ing of mathematical modelling—preface. In G. Kaiser, W. Blum, R. Borromreo Ferri, &
Lesh, R., & Doerr, H. M. (2003). Foundations of models and modelling perspective on math-
ematics teaching, learning, and problem solving. In R. Lesh & H. M. Doerr (Eds.), Beyond
constructivism: Models and modelling perspectives on mathematics problem solving, learning,
and teaching (pp. 3–33). Mahwah: Lawrence Erlbaum.
Lesh, R., & Harel, G. (2003). Problem solving, modeling, and local conceptual development.
Mathematical Thinking and Learning, 5(2–3), 157–189.
Lesh, R., & Kelly, A. (2000). Multitiered teaching experiments. In A. Kelly & R. Lesh (Eds.),
Research design in mathematics and science education (pp. 197–230). Mahwah: Lawrence
Erlbaum.
Lesh, R., Hoover, M., Hole, B., Kelly, A., & Post, T. (2000). Principles for developing thought
revealing activities for students and teachers. In A. Kelly & R. Lesh (Eds.), Handbook of
research design in mathematics and science education (pp. 591–645). Mahwah: Lawrence
Erlbaum.
Lesh, R., Kelly, A., & Yoon, C. (2008). Multi-tier design experiments in mathematics, science,
and technology education. In A. E. Kelly, J. Y. Baek, & R. A. Lesh (Eds.), Handbook of design
research methods in education: Innovations in science, technology, engineering, and math-
ematics (pp. 131–148). New York: Routledge.
Moore, T. J., Miller, R. L., Lesh, R. A., Stohlmann, M. S., & Kim, Y. R. (2013). Modeling in engi-
neering: The role of representational fluency in students’ conceptual understanding. Journal of
Engineering Education, 102(1), 141–178.
Moore, T. J., Doerr, H. M., Glancy, A. W., & Ntow, F. D. (2015). Preserving pelicans with models
that make sense. Mathematics Teaching in the Middle School, 20(6), 358–364.
Palmer, S. E. (1977). Fundamental aspects of cognitive representation. In E. Rosch & B. B. Lloyd
(Eds.), Cognition and categorization. Hillsdale: Lawrence Erlbaum.
Pompei, V. (2010). Pelican colonies model-eliciting activity. University of Minnesota. Available
at: https://unlvcoe.org/meas/wp-content/uploads/2013/01/Pelican-Colonies-MEA-Teacher-
Materials.pdf
Zaslavsky, O. (2005). Seizing the opportunity to create uncertainty in learning mathematics.
Educational Studies in Mathematics, 60(3), 297–321.
Chapter 8
Mapping Shifting Discourses in Classroom
Talk Through Indexical Discourse
Analysis: From the World of Mathematics
to the World of Work
Stephen T. Lewis and Azita Manouchehri
Abstract In this paper, we outline how indexical discourse analysis coupled with
an ethnomodelling framework can be used to analyse mathematical modelling inter-
actions by making visible the interrelationship between the ideologies of the real
world and mathematics. We illustrate this position by sharing classroom-based data
of students’ ideas while considering business profitability problems. We investigate:
How does focusing on indexical terms within classroom talk aid a researcher in
detecting subtle shifts in discourse that represent different discursive worlds? We
highlight how the teacher’s particular scaffolds shifted students’ work from solving
an application task into a mathematical modelling inquiry.
Keywords Discourse analysis · Ethnomodelling · Indexicality · Intuition ·

Mathematising · Scaffolding
8.1 Introduction
With recent calls advocating for examination of classroom discourse that supports
student learning in mathematical modelling contexts (Cai et al. 2014), the need for
conceptualising robust analytical tools appropriate for analysing educational dis-
course becomes even more apparent (Hennessy et al. 2016, p. 18). Hennessey and
colleagues (2016) posited that absence of suitable analytical tools hinders the
researchers’ ability to robustly make sense of teacher-learner interactions and cap-
ture shared functions of dialogue (p. 17). We argue, however, that dialogic analysis
as a singular tool is insufficient for capturing the complex nature of interactions
S. T. Lewis (*) · A. Manouchehri

College of Education and Human Ecology, The Ohio State University, Columbus, OH, USA
e-mail: slewis@pastfoundation.org; Manouchehri.1@osu.edu

86 S. T. Lewis and A. Manouchehri
around mathematics, in particular in revealing how and when learners’ mathemati-

cal conceptions might have aligned to their prior experiences and intuitions. Since
mathematical modelling involves transitions between real-world contexts and math-
ematics (Blum and Leiss 2007; Niss et al. 2007; Pollak 1979), coding schemes
become particularly useful if they allow the researchers to unfold how problem
solvers reconcile these worlds and how prior contextual experiences might be
reflected in students’ responses and their thinking. Currently, there exists a broad
array of tools in linguistic discourse analysis that could be brought to bear on math-
ematical contexts and afford avenues for advancing researchers’ analysis. In this
chapter, we elaborate on how indexical discourse analysis (Silverstein 1992) can
serve this function in revealing mathematical modelling interactions and elucidating
their nature. Our research was motivated by the following question: How does
focusing on indexical terms within classroom talk aid a researcher in detecting sub-
tle shifts in discourse that represent different discursive worlds?
8.2 Background Literature
At its core, mathematical modelling is mediated through the process of mathematis-

ing (Cobb et al. 1997). Rosa and Orey (2010) describe mathematising as a process
of producing mathematics by cultural groups to organise, analyse, comprehend,
understand, and solve problems in the context of their daily lives. Rosa and Orey
further argue that these culturally produced tools offer different methods for inter-
preting and representing problems and, hence, discovery of relations and regulari-
ties in the problem context in various forms. In this view, mathematisation affords
the means to transition contextually located problems into mathematical ideas
(p. 20). These mathematical ideas need not match the conventional mathematical
forms endorsed and promoted by western perspectives, but rather take shape as a
cultural product of that society (Martin 1997). This perspective warrants the need to
develop specific conceptual tools and analytical techniques appropriate for examin-
ing interactions within the cultural context of group activities. We consider the con-
ceptual domain of ethnomodelling as a means for developing an understanding of
cultural practices in mathematical ways and discourse analysis (Bloome et al. 2010)
as a methodological tool for their analysis.
8.2.1 Ethnomodelling
Orey and Rosa (2015) describe ethnomodelling as “the study of mathematical ideas
and procedures elaborated by members of distinct cultural groups” (p. 367). This
process involves examination of local mathematical practices in diverse situations
stemming from the daily lives of group members. The intent of this process, accord-
ing to Orey and Rosa, is to develop an understanding of the interdynamics of these
8 Mapping Shifting Discourses in Classroom Talk Through Indexical Discourse… 87
local practices as well as their transference across other local systems (p. 368). An
ethnomodelling framework sanctions the researcher to rely on his/her mathematical
lens to discover, interpret, elaborate on, and report these local systems in recogni-
sable forms to the community at large.
With such a stance, the researcher enters the research site with the intent to
develop a glocal or mid-level view of the mathematical practices of members of a
cultural unit as they engage in solving authentic problems relevant to their lives
(Orey and Rosa 2015). This glocal view is mediated through the interaction of emic
or local practices that occur and the etic view of the researcher informed by their
background, perspectives, and experience. The first major component of navigating
this research process is discovering the emic systems, where a researcher examines
the interactions that occur within this cultural unit not for the purpose of imposition
of their view, but rather to identify patterns that emerge in local problem solving.
The result is the emergence of an emic ethnomathematical knowledge which is
refined further through dialogic inquiry and translation of ideas mediated between
the researcher and the members of the group (Lewis and Manouchehri under review;
Lewis 2018). These emic ethnomodels can be elaborated by pairing them alongside
mathematical practices exhibited by other groups, such as academic mathematics.
That is to say that through this ethnomodelling process, emic practices can be war-
ranted as mathematical in their nature and represented in recognisable ways to the
research community. The process of elaboration foundationally relies on an ethno-
methodological framework. To examine these emic systems, we turn to the field of
discourse analysis as it offers a focus on the interactive components of the system
through a detailed examination of the language in use (Bloome et al. 2010).
8.2.2 Discourse Analysis and Indexicality
Wortham and Reyes (2015) describe discourse analysis as “a research method that
provides systematic evidence about social processes through the detailed examina-
tion of speech, writing and other signs” (p. 1). The authors indicate further that
social events contain multiple components worth noting such as the participants, the
message communicated over the course of an interaction, and finally the setting or
contexts that the event occurs in. Furthermore, through these characteristics, social
consequences are elicited (p. 10).
Bloome et al. (2010) argue that one key component in the social construction of
meaning is identifying and recognising intercontextual links between events.
Moreover, in order for intercontextual links to be established, the participants must
acknowledge and recognise this connection (Bloome et al. 2010, p. 44). To ensure
that knowledge of connections has indeed occurred, the researchers would need to
validate findings through reengagement with participants through dialogue. These
axiomatic principles of discourse analysis parallel the translation and elaboration
processes stressed in ethnomodelling research.
In establishing that discourse analysis can be useful in elucidating intercontex-

tual links, in our research we contemplated specific tools that would allow us to
highlight learners’ perceptions and the teacher’s responses to those perceptions in
mathematical modelling contexts. The construct of indexicality (Silverstein 1992)
was distinguished as a helpful tool for our scholarly endeavour.
Silverstein (1992) describes indexicals as signs that point to a presupposed con-
text in which they occur or to some entailed potential context. Silverstein further
indicates that these indexicals, through discursive interaction, afford the interac-
tional participants the ability to make sense of these interactions. Central to this
metapragmatic process of construing meaning within interactions is considering the
means by which language becomes a discursive object. Through indexicality, differ-
ent ideologies and systems of relationships that are constructed become visible
between the words used and those contexts that they are located in (Silverstein 1992).
Silverstein (1992) outlines three key components of the indexical process: first
that all semiotic signs have a significant part of their meaning in how they are placed
in the world. For example, in a discussion that takes place during a mathematics
class, one could reasonably assume aspects of that discussion relate to mathematical
ideas. Second, Silverstein indicates that meaning of discursive interactions is war-
ranted through the use of these metapragmatic functions. Finally, the key to deter-
mining meaning of a discursive interaction is determining where the metapragmatic
functions are located.
In determining indexicals of merit, Scollon and Scollon (2003) outline that those
most frequently examined are personal pronouns (I, we, you, etc.), demonstratives
(this, that), deictics (here, there, now), and tense, but further establish that all words
can act as indexicals in some capacity (p. 31). Additionally, Scollon and Scollon
propose that it is through study of these indexical processes that researchers can
establish how these semiotic signs point to the underlying meaning of an interaction
and further how different ideologies are constructed within.
Viewing situations considering Silverstein’s construct of indexicality, we recog-
nise that these presupposed contexts could apply directly to mathematical model-
ling. Since modelling tasks demand a transition between the two, often dichotomous,
ideological worlds of mathematics and real life (Pollak 1979), there is benefit to
being able to examine how these two worlds are signalled within a discussion.
Viewing interactions through the lens of indexicality demands the researcher to
listen carefully to how students interpret and respond to both the tasks that they are
engaged in as well as how their teacher engages them during this process, so as to
determine how they draw on contextual and mathematical components when solv-
ing problems. We illustrate the utility of indexicality in capturing these delicate
features of mathematical modelling, in particular how learners take up the notion of
real-world business by analysing data from a classroom discussion where learners
and their teacher were engaged in solving problems related to profit, cost, and
revenue.
8.3 Context and Background
Our programme of research aims to analyse interactions in mathematical modelling

contexts and focuses primarily on teacher scaffolding mechanisms through reflexive
discussion (Qualley 1997) and how students interpret and respond to tasks
(Manouchehri and Lewis 2017). The data for this study comes from a microethno-
graphic study (Bloome et al. 2010) of an 11th grade precalculus class at a private
academy in the Midwestern United States that occurred during the 2016–2017 aca-
demic year. The goal of this study was to examine teacher-learner interactions in
mathematical modelling contexts and analyse those interactions considering the
teachers’ receptivity to student ideas and conceptions. The teacher of this class was
in her 11th year of instruction and had a focus on implementation of mathematical
modelling tasks on a regular basis over the course of the year. Our rationale for
selecting this research site was that this teacher wholeheartedly endorsed the math-
ematical modelling process and viewed it as a crucial component of her instruction.
This provided the benefit of observing a class devoted to mathematical modelling,
which Niss et al. (2007) argue is rare outside of specialised university programmes.
Our analysis consisted of transcribing video and audio recordings of classroom
interactions by constructing message units of utterances (Bloome et al. 2010;
Wortham and Reyes 2015). We then conducted an indexical analysis (Silverstein
1992; Wortham and Reyes 2015) and focused on identifying how indexicals were
invoking different discursive worlds, how those discursive worlds were interacting
with each other, and their relationship to the mathematical modelling process. The
indexical analysis process entailed mapping out how particular indexicals were sig-
nalling meaning across the discussion. We used a colour coding system to highlight
whether particular indexicals, for example, here, referenced specific mathematical
objects such as a graph or aspects of a real-world context such as a business. Using
this method, we were able to trace how a discussion shifted between solving a math-
ematical application problem and mathematical modelling. We exemplify our pro-
cess below in the context of a discussion related to profit, cost, and revenue models
in an application problem.
8.4 n Indexical Analysis of Profit, Cost,

A
and Revenue Models
This event occurred during a lesson on solving simultaneous nonlinear functions.

The Profit, Cost and Revenue Problem below was assigned to students to complete
the night prior to the lesson as a homework problem. During the initial phase of the
lesson, our cooperating teacher asked her students for any questions that they had
the night before, and students unanimously indicated that they needed support on
the task. This discussion was structured by the teacher asking her students for their
ideas on how to proceed and opted to allow them to find the solution using any
means they deemed appropriate. Our teacher indicated to us that she selected this
task in particular as a homework problem because she felt her students would be
familiar with the context as many of the students’ parents were business owners or
had experiences with profit, cost, and revenue.
Profit, Cost, and Revenue Task

Economists for Smith Brothers Incorporated find that the company profit (P)
is equal to the revenue (R) minus the cost (C). Determine when Smith Brothers
Incorporated yields a profit if R(x) = 0.0125x2 + 412x and
C(x) = 12,225 + 0.00135x3 (Demana et al. 2011).
As the teacher facilitated a discussion of this problem, her students offered ways
that the solution might be determined. Students opted to construct a graph of the
cost and revenue functions and determined that a solution would occur for a positive
value of profit between 30 and 542 customers. Following this discussion, the teacher
then initiated a brief intercontextual link between the solution and real-world busi-
ness by inviting in student perceptions of whether this model is realised by actual
business. The short transcript below outlines the discussion that progressed from
this point. Following this teacher prompting, we observed that her students began
critiquing the proposed models against their experiences and knowledge of real-
world business and proposed solutions that would alter the conventional models to
allow for the business to be profitable.
Teacher: Okay, so keep talking to me Adam, just for a second. Why did you say
anything outside of this isn’t going to work? (Referring to the region
of intersection)
Adam: Um, because it’s above the revenue so the cost is too high.
Teacher: So the cost is too high here? (Pointing to the region of intersection)
Like literally in real-world terms, not math terms, the cost is too high
for this business model to work?
Adam: Well that’s what the graph says.
Mira: That’s a bit risky for a business.
Class: (Laughs)
Adam: Well ‘cause the more customers you have the more you’d have to pay;
you’d have to hire more people.
Mira: Exactly, but if you go hire that many more people your business is
gonna go bankrupt.
Amber: In the equation it says that revenue has to be greater than the cost, so
if the cost is higher on the graph than the revenue, then you’re going
to have a negative profit.
Teacher: Does that make sense Ced?
Ced: Me?
Teacher: Does that make sense what they just said?
Ced: I think probably this is a good model, but it’s not perfect because real-
istically the more, you’re not going to lose money with each increas-
ing custom uh customer.
Adam: Mmhm.
Ced: Uh…so it’s probably a good model for between 30 and 542, it’s just
not perfect after that.
Teacher: Hmm, what does that probably say about their current business?
Ced: It’s somewhere between 30 and 542.
Teacher: Yeah, ‘cause if the model starts to get a little goofy as it gets higher,
maybe you haven’t just quite…
Mira: But if they have more profits, couldn’t they raise a bunch of money
whatever their selling is, and therefore make you have a profit then?
Teacher: Maybe.
Mira: ‘Cause if people are buying it, then you want to get more money so
you could raise it.
In focusing how different indexical signs point to meaning and construct ideolo-
gies, this event showcases a transition from the world of mathematics to the world
of profit potential and business sustainability. Notice that early on the solution to the
overarching problem that drove this discussion is being indexed through pronouns
such as it’s above the revenue (referring to the graph) on line 3, demonstratives such
as anything outside of this (indicating the region of intersection) on line 2, and deic-
tics such as the cost is too high here (indicating the region outside of the graphical
intersection) on line 4.
However, when the teacher introduces the notion of real-world terms in line 4, a
shift in the metapragmatic function through the word work occurs. Prior, in the con-
text of the discussion, the term work referred to a valid solution of the application
problem. On line 5, when the teacher said “Like literally in real-world terms, not
math terms, the cost is too high for this business model to work?”, here work seems
to be indexing the sustainability of the business. Mira picks up on this notion in line
8 when she introduces the notion of riskiness using the demonstrative that (referring
to a model that won’t sustain growth). As risk analysis was not a part of the problem
at hand, we notice that the conception of a valid solution now refers differently to
the context of business than it had prior, insomuch that before the associated task
was determining a solution to the application problem presented, while now the
business has been authenticated in deeming it fallible (Manouchehri and Lewis
2017). We see further evidence that this notion of riskiness was picked up in line 20
when Ced establishes a dichotomy by claiming that it is a good model but not per-
fect realistically. In this utterance, we observe Ced critiquing the profit, revenue, and
cost models presented in the application problem against their utility in actual busi-
ness. When Ced references a good model (lines 20 and 24), the validity of the solu-
tion to the application task is being acknowledged, which is further exemplified
when he elaborates that it could be useful for between 30 and 542 customers (line
24). This he distinguishes from the model as not perfect realistically (line 21), where
he is weighing the context of a business operating with a higher level of demand.
As the discussion further unfolds, we see Mira propose raising the price per unit
of the item being sold (lines 30–31) as she interprets a high level of demand for the
product (lines 33–34). According to her view, the realised business having more
profits indexes a demand for the product near or above 542 customers. This is re-
articulated in line 33, when Mira indicates that people are buying this item. Mira
then offers in lines 30–31 as well as reaffirms in line 34 that the price per unit could
be increased to generate a larger earning potential. We see Mira’s utterances as
interpretation of mathematical results against the real context, which Niss et al.
(2007) indicate exemplifies modelling tasks. We further interpret Mira’s proposition
of raising the price per unit as an act of refinement of the original model correspond-
ing to the refinement and re-engagement phases of the modelling process.
Our own etic lens of mathematics and economics did indeed shape how we inter-
preted Mira’s suggestions, relating it to ways of increasing the profit of a business
using profit margin analysis assuming a high demand for the product (Cafferky and
Wentworth 2014), and that this etic view did influence what we considered as wor-
thy of further inquiry with the participants. Motivated by our etic view of economics
and mathematics, we sought to validate that Mira was indeed proposing raising the
price per unit of the product being sold giving Mira an opportunity to translate this
statement to our research team. In a post-interview, Mira verified that, based on her
understanding of business and drawing on her family background and interest in
pursuing this as a potential career after high school, she felt that intuitively raising
the price should expand this profit region.
We note further that this dialogic shift transitioned the task from an application
context to a modelling situation and that this transition was initiated through teacher
prompting, which we view as a mechanism of reflexive teaching (Qualley 1997).
Finally, we argue that this discussion transpired because the teacher positioned her
students to reference their realistic considerations and experiences of the business
context, allowing the task to be authenticated by the students, and afforded the prop-
osition of refinement by the group. Our ethnomodelling lens sharpened our particu-
lar attention to the teacher’s stance, knowledge of her thinking, and the resulting
classroom interactions. The indexical analysis afforded us the opportunity to notice
that mathematical modelling had indeed occurred.
Hennessy et al. (2016) advocate for the importance of researchers to be able to

“access levels of dialogicality through carrying out fine-grained analysis of what
teachers and students actually do and say when engaged in dialogic interactions
across educational contexts” (p. 18). This merits the ability to not only understand
structural characteristics of how the discussion progresses but also how the discus-
sion indexes student conceptions of mathematics. A coordination of ethnomodelling
with indexical analysis afforded us access to the import behind what students shared
and how the teacher was receptive to those ideas. We were able to document not
only how these interactions transformed all participating members but also the
instructional event itself.
Ethnomodelling demands the researchers to be receptive to infinitesimally small
and typically unnoticed exchanges during interactions and to acknowledge them as
mathematical in their nature. Through indexical analysis, we were able to recognise
and distinguish between the two ideologies of real-world business and mathematics.
In particular, we observe that these indexical links create the thread that transitioned
the discussion from application to modelling. However, this thread is just one small
tie in the grand tapestry of the context of teaching and learning in which mathemati-
cal modelling is present. To examine the holistic process of how these social model-
ling practices are developed and nurtured among learners in educational settings,
historical and contextual background needs to also be considered.
In our analysis, we focused on only one aspect of the classroom event, the dia-
logic exchanges, and as Wortham and Reyes (2015) point out, other components to
consider such as setting, context, and the actors themselves must be considered
when exploring interactions. As such, further coding systems capable of incorporat-
ing these components should be sought in order to offer a rich conceptualisation of
modelling events and further identify instances of modelling that may have gone
unnoticed in the past.
While the task itself was presented as an application problem, it was the teacher’s
vision (Goodwin 1994) of sustained development of modelling capacity that
prompted her invitation of real-world business. Supporting teachers in recognising
those instances when modelling can be thoughtfully addressed becomes paramount
when treating modelling as a sustained and long-term goal for instruction.
References
Bloome, D., Carter, S. P., Christian, B. M., Otto, S., & Shuart-Faris, N. (2010). Discourse analysis
and the study of classroom language and literacy events: A microethnographic perspective.
New York: Routledge.
Blum, W., & Leiss, D. (2007). How do students and teachers deal with mathematical modelling
problems? The example “Filling up”. In C. Haines, P. Galbraith, W. Blum, & S. Khan (Eds.),
Mathematical modelling (ICTMA 12): Education, engineering and economics (pp. 222–231).
Chichester: Horwood.
Cafferky, M., & Wentworth, J. (2014). Breakeven analysis: The definitive guide to cost-volume-
profit analysis. New York: Business Expert Press, LLC.
Cai, J., Cirillo, M., Pelesko, L., Borromeo Ferri, R., Borba, M., Geiger, V., Stillman, G., English,
L. D., Wake, G., Kaiser, G., & Kwon, O. (2014). Mathematical modelling in school education:
Mathematical, cognitive, curricular, instructional and teacher education perspectives [Research
Forum]. In P. Linljedahl, C. Nicol, S. Oesterle, & D. Allan (Eds.), Proceedings of the joint
meeting of PME 38 and PME-NA 36 (Vol. 1, pp. 145–172). Vancouver: PME/PME-NA.
Cobb, P., Boufi, A., McClain, K., & Whitenack, J. (1997). Reflective discourse and collective
reflection. Journal for Research in Mathematics Education, 28(3), 258–277.
Demana, F., Waits, B., Foley, G. D., & Kennedy, D. (2011). Precalculus: Graphical numerical,
algebraic (8th ed.). Boston: Pearson Education.
Goodwin, C. (1994). Professional vision. American Anthropologist, 96(3), 606–633.
Hennessy, S., Rojas-Drummond, S., Higham, R., Márquez, A. M., Maine, F., Ríos, R. M., et al.
(2016). Developing a coding scheme for analysing classroom dialogue across educational
contexts. Learning, Culture and Social Interaction, 9, 16–44. https://doi.org/10.1016/j.
lcsi.2015.12.001.
Lewis, S. (2018). Theorizing teaching practices in mathematical modeling contexts through the
examination of teacher scaffolding. (Electronic Thesis or Dissertation). Retrieved from https://
etd.ohiolink.edu/.
Lewis, S. T., & Manouchehri, M. (under review). The construction and application of an ethno-
modeling cycle. Journal of Mathematics and Culture.
Manouchehri, A., & Lewis, S. T. (2017). Reconciling intuitions and conventional knowledge: The
challenge of teaching and learning mathematical modelling. In G. A. Stillman, W. Blum, &
G. Kaiser (Eds.), Mathematical modelling and applications: Crossing and researching bound-
aries in mathematical modelling (pp. 107–116). Cham: Springer.
Martin, B. (1997). Mathematics and social interests. Ethnomathematics: Challenging Eurocentrism
in mathematics education. In A. B. Powell & M. Frankenstein (Eds.), Ethnomathematics:
Challenging Eurocentrism in mathematics education (pp. 155–172). Albany: State University
of New York Press.
& M. Niss (Eds.), Modelling and applications in mathematics education (pp. 3–32). New York:
Springer.
Orey, D. C., & Rosa, M. (2015). Three approaches in the research field of ethnomodeling: emic
(local), etic (global), and dialogical (glocal)/Três abordagens no campo de pesquisa da etno-
modelagem: êmica (local), ética (global) e dialógica (glocal). Revista Latinoamericana de
Etnomatemática, 8(2), 364–380.
Pollak, H. O. (1979). The interaction between mathematics and other school subjects. In
International Commission on Mathematical Instruction (Ed.), New trends in mathematics
teaching (Vol. 4, pp. 232–248). Paris: United Nations Educational, Scientific and Cultural
Organisation.
Qualley, D. J. (1997). Turns of thought: Teaching composition as reflexive inquiry. Portsmouth:
Boynton/Cook.
Rosa, M., & Orey, D. C. (2010). Ethnomodeling as a pedagogical tool for the ethnomathematics
program. Revista Latinoamericana de Etnomatemática, 3(2), 14–23.
Scollon, R., & Scollon, S. W. (2003). Discourses in place: Language in the material world.
London: Routledge.
Silverstein, M. (1992). Metapragmatic discourse and metapragmatic function. In J. Lucy (Ed.),
Reflexive language (pp. 33–58). Cambridge: Cambridge University Press.
Wortham, S., & Reyes, A. (2015). Discourse analysis: Beyond the speech event. New York:
Routledge.
Chapter 9
Joy of Mathematical Modelling:
A Forgotten Perspective?
Yuriy Rogovchenko, Olov Viirman, and Stephanie Treffert-Thomas
Abstract We argue the relevance of including an affective perspective in mathe-

matical modelling education research and emphasise its importance for the teaching
and learning of mathematical modelling at all levels, especially at university. Our
argument is supported by a recent survey of mathematics lecturers’ views on math-
ematical modelling, several follow-up interviews, and a review of literature on
mathematical modelling that relates to enjoyment, pleasure, and appreciation.
Findings from the survey and the follow-up interviews indicate that there is a group
of practitioners who hold strong views on the importance of enjoyment in doing and
teaching mathematical modelling.
Keywords Affective domain · Emotions · Enjoyment · Mathematical modelling ·

University education
9.1 Introduction
Connections between beauty and mathematics can be traced to antiquity with

Aristotle asserting that “the mathematical sciences particularly exhibit order, sym-
metry, and limitation; and these are the greatest forms of the beautiful” (Metaphysics,
XIII, 3.107b). Since then many mathematicians have distinguished beauty as a par-
ticularly appealing feature of mathematics. Poincaré related the joy of doing math-
ematics to its aesthetical aspects:
Y. Rogovchenko (*) · O. Viirman

Department of Mathematical Sciences, University of Agder, Kristiansand, Norway
e-mail: yuriy.rogovchenko@uia.no; olov.viirman@hig.se
S. Treffert-Thomas
Mathematics Education Centre, Loughborough University, Loughborough, UK
e-mail: s.thomas@lboro.ac.uk

96 Y. Rogovchenko et al.
Mathematics has a triple end. It should furnish an instrument for the study of nature.
Furthermore it has a philosophic end, and, I venture to say, an end aesthetic. [Mathematicians]
admire the delicate harmony of number and of forms; they are amazed when a new discov-
ery discloses for them an unlocked for perspective; and the joy they thus experience, has it
not the aesthetic character although the senses take no part in it? Only the privileged few are
called to enjoy it fully, it is true; but is it not the same with all the noblest arts? (as cited in
Gaither and Cavazos-Gaither 2012, p. 1316)
For Hilton, however, the joy of mathematics comes from its great power and
perfection:
We mathematicians should be envied the joy which mathematics brings us. This joy comes
to us through our awareness of the great power and subtlety of a fertile mathematical idea,
and also through our sense that such an idea has been expressed in a form capable of perfec-
tion. (Hilton 1992, p. 276)
Affective and aesthetic aspects have traditionally been associated with pure mathe-
matics. Davis and Hersh point towards the assumed “inferiority” of applied mathe-
matics and mathematical modelling:
The reputed superiority of mind over matter finds mathematical expression in the claim that
mathematics is at once the noblest and purest form of thought, that it derives from pure
mind […] and there is a pervasive unspoken sentiment that there is something ugly about
applications. (Davis and Hersh 1981, p. 85)
Members of our research team find doing and teaching mathematical modelling
highly enjoyable. Hence, preparing a questionnaire to survey mathematics lectur-
ers’ views on the teaching of mathematical modelling in universities in England and
Norway (Treffert-Thomas et al. 2017), we decided to include items related to the
enjoyment of mathematical modelling. Feedback during a pilot phase highlighted
their relevance, and we included a number of statements on the affective perspective
of mathematical modelling in the final version of the questionnaire. During data
analysis, we found that a group of respondents emphasised and foregrounded affec-
tive aspects in relation to the aims of mathematical modelling in their professional
work (p. 130) and in teaching (pp. 131–132). To the best of our knowledge, the
affective or “enjoyment” perspective has not been explored much in the mathemati-
cal modelling literature (see Schukajlow et al. 2018). In this chapter, we present
data and results from the questionnaire study and from ongoing follow-up inter-
views with lecturers alongside a review of mathematical modelling literature to
argue the relevance of including an affective (“enjoyment”) perspective in research
on the teaching and learning of mathematical modelling at all levels, especially at
the university level.
9 Joy of Mathematical Modelling: A Forgotten Perspective? 97
9.2 Affective Aspects of Mathematics Teaching and Learning
The need for a re-evaluation of the role of the affective domain in the learning of
mathematics arose when research based on the cognitive perspective left unan-
swered many questions regarding students’ difficulties with mathematics. McLeod
(1992) argued for the inclusion of the affective domain in research into cognition,
and his article is often seen as a turning point in mathematics education in this
respect. Leron and Hazzan (1997) also assert that there has been a “strong emphasis
on cognitive aspects, and consequent neglect of affective and social factors” (p. 266),
and that in situations where students experience confusion and loss of meaning dur-
ing solution of mathematical problems “affective and social factors are as much part
of the student’s thinking and behaviour as the cognitive factors” (Leron and Hazzan,
p. 265).
Nowadays mathematics education researchers widely acknowledge that emo-
tions are intrinsic to cognition. Op’T Eynde et al. (2006) argue that emotions are
“very much part of problem solving” in mathematics classrooms (p. 204). Evans
(2000) stresses that “affect and emotion [are] inseparable from thinking, including
mathematical thinking” (p. 228). Even though emotions may not necessarily inter-
fere with mathematical thinking, they are viewed “in terms of charges of feeling
attached to ideas and thus related to the cognitive” (Evans, p. 230, emphasis in
original). Rodd (2010) relates pleasure and success in learning to “students [being]
attracted to mathematics principally because of their prior success in the subject or
because of their pleasure in engaging in mathematical activity, and these are
linked” (p. 4).
Defining enjoyment in terms of pleasure and absence of pressure, Lim et al.
(2009) explored university students’ views on mathematical modelling concluding
that although attitudes towards mathematics did not significantly change, students
“seem to enjoy mathematics more than they had done before the modelling project”
(p. 449). This agrees with a study by Schukajlow et al. (2012) who observed that for
secondary school students, “enjoyment and interest really profit from the divergent
solution structure of modelling problems” (p. 231).
Sinclair (2004) pointed at parallels between mathematicians and students argu-
ing that “students may, in fact, share some aesthetic tendencies with mathemati-
cians, but may not know how to use them in the context of mathematical inquiry”
(p. 282). Koichu et al. (2017) noted that the gap between how mathematicians feel
about beauty and how students do was “smaller than previously believed” (p. 54),
echoing Hadamard’s (1945) views that “between the work of the student who tries
to solve a problem in geometry or algebra and a work of invention, one can say that
there is only a difference of degree, a difference of level, both works being of a simi-
lar nature” (p. 104). Hilton (1992) describes a highly emotional mathematicians’
thrill of discovery, creation, and acquisition of new knowledge:
It is my claim based on my own experience and observation over many years that, in math-
ematics, the thrill of real understanding is comparable with the thrill of original discovery.
Mathematicians take genuine pleasure in the triumphs of others and, in a sense, recreate the
act of discovery in mastering the intricacies of another’s thought. (Hilton 1992, p. 278)
We believe that students experiencing positive emotions such as joy or aesthetic

pleasure are better motivated to learn mathematics through a change in disposition,
as put forth by Hernandez-Martinez and Pampaka (2017) in their research with
secondary school students:
Emotional moments can change the motives of engagement of an individual in a certain
activity, such as when we hear a highly inspirational, emotional speech, when a teacher
introduces an “out of the ordinary” activity that “connects” with our interests, or when
someone experiences a “Eureka moment” that is accompanied by an emotional state of hap-
piness. (Hernandez-Martinez and Pampaka 2017, p. 192)
Categorising international research perspectives on mathematical modelling, Kaiser

and Sriraman (2006) distinguished several perspectives: realistic (applied), episte-
mological (theoretical), sociocritical (emancipatory), contextual, and educational,
with the last divided into didactical and conceptual modelling. Buchholtz (2013)
wrote that although such a classification was useful, its strong systematising simpli-
fication “might not reflect all research approaches appropriately” (p. 103). In par-
ticular, it did not distinguish between didactical approaches and research perspectives
and was revised soon after (Kaiser et al. 2007). However, affective aspects are not
covered by this classification, and some of the findings in our recent study (Treffert-
Thomas et al. 2017) encourage us to argue for the need of including this perspective
in the research on mathematical modelling.
9.3 Data Collection
First, we present data from a survey of Norwegian1 university mathematics lectur-

ers’ views on the aims of mathematical modelling in research and teaching (Treffert-
Thomas et al. 2017), focusing on aspects relevant to the current chapter. The survey
contained two questionnaire items that listed enjoyment of mathematical modelling
among the statements from which respondents were asked to select and then rank
three statements they most agreed with.
The first item, on lecturers’ views on the aims of mathematical modelling in their
professional work, was based on the realistic and epistemological perspectives in
Kaiser and Sriraman (2006) augmented with the enjoyment perspective (three state-
ments for each perspective) (see Fig. 9.1). The second item explored lecturers’
views on the aims of teaching mathematical modelling with the enjoyment category
added to the full classification of Kaiser and Sriraman (2006), this time two state-
ments for each category (Fig. 9.2). The categorisation of the statements was not
communicated to respondents. Of 498 surveys sent out in Norway, we received 119
replies, a response rate of 24%. In presenting our data analyses relevant to this
chapter, a selection of at least two statements pertaining to a particular perspective
1
We also surveyed English lecturers where results and findings were similar. However, the data
collection was delayed so that the analyses could not be included in Treffert-Thomas et al. (2017).
The aim of practicing mathematical modelling (in your professional capacity) is
to describe, explain and/or predict reality.

to solve real -life, authentic problems.
to increase our understanding of the real world.
to gain mathematical insights.
to advance mathematical theory.
to solve problems purely within (pure) mathematics.
for interest and/or enjoyment.
to foster creativity.
to address/seek out challenges.
None of the above.
Fig. 9.1 The aim of practising mathematical modelling
The aim of teaching modelling is for students ….
to solve real-life problems.

to increase their understanding of the real world.
to gain mathematical insights.
to advance their understanding of mathematical theory.
to become more critically aware as citizens.
to gain conceptual understandings of mathematics.
to have/add structure in their learning of mathematics.
to encourage critical reflection and discussion of the standing of mathematics
in the world.
to elicit models from solving word problems.
to learn to apply models and modelling to different situations.
to enjoy learning mathematics.
to become interested in mathematics.
Fig. 9.2 The aim of teaching mathematical modelling
was considered as the indication of lecturer’s agreement with that particular per-
spective, as was giving a statement the highest ranking.
Second, we include preliminary results from the follow-up interviews with
eleven lecturers, five in Norway and six in England. All completed the questionnaire
indicating willingness to be interviewed to explore their views on mathematical
modelling in more detail. Lecturers were contacted by email; semi-structured inter-
views were conducted in person or via Skype. Interviews were audio-recorded, and
further detailed analyses are planned. The following four interview questions were
communicated to the interviewees in advance:
1. What motivates you to use mathematical modelling in your teaching?

2. How do you explain (the meaning of) mathematical modelling to your students?
How do you understand it yourself?
3. Could you provide an example of using mathematical modelling in your
teaching?
4. Are there any changes you intend to make, or that you would like to make next
time you teach mathematical modelling?
Third, we searched for literature relating to emotions and mathematical modelling.
We conducted a search of databases – ERIC, JSTOR, and Google Scholar – and
articles in collections of major academic publishers (Springer, Elsevier, and Taylor
and Francis) using the search terms “joy”, “enjoyment”, and “pleasure” in combina-
tion with “mathematical modelling” and spelling “modelling” with one and two “l”.
Most citations collected from articles and books were authors’ opinions termed by
Treffert-Thomas and Jaworski (2015) as professional literature. Being juxtaposed
with research literature understood as research conducted in accordance with recog-
nised methodological principles, this emphasises the scarcity of research in the area.
9.4 indings from Questionnaire Study, Interviews,

F
and the Literature
First, we present findings from the questionnaire study that explicitly relate to the
enjoyment perspective. Answering the question about the aims of mathematical
modelling in their professional work, ten respondents (11%) included at least one of
three statements related to enjoyment. Of these, three respondents selected two
enjoyment statements, and five respondents ranked one of the enjoyment statements
highest, indicating that the enjoyment perspective is important in their professional
practice of mathematical modelling. Answering the question about the aims of
teaching mathematical modelling, most respondents selected statements from dif-
ferent categories (53%). Eighteen respondents (22%) included at least one of the
enjoyment statements. Of these, three respondents selected both enjoyment state-
ments, and four ranked one of the enjoyment statements highest, suggesting that the
enjoyment perspective is very important in their teaching of mathematical modelling.
Hence, while most of the respondents did not rank the “enjoyment” statements
particularly highly, quite a few did include one of the statements among the three
chosen. In addition, there were respondents who ranked the enjoyment perspective
very highly. In fact, two respondents selected two or more enjoyment statements
among the three they had ranked on both questionnaire items, indicating that they
rated them as very important. A further two respondents on each questionnaire item
assigned their highest ranking to a statement related to the enjoyment perspective.
This clearly identifies a small group of respondents for whom this perspective was
crucial and indicative of their views on practising and teaching mathematical

modelling.
Second, we present our initial findings from the interviews. Many lecturers
expressed enthusiasm for mathematical modelling demonstrated in their engage-
ment with our interview questions. Some clearly believed that their research and
teaching were interrelated, so that using mathematical modelling in their own prac-
tice impacted directly on teaching, as expressed by one of the respondents:
I am a mathematical modeller for my research so everything I do from a research point of
view is mathematical modelling. I am a very interdisciplinary mathematician. […]
Everything I do is going out talking to other disciplines, finding interesting problems, using
modelling to write down the problem that I want to solve and then going off and using the
maths and solving it. It is my research area so of course I use modelling for my teaching.
(L4B, 00.28) (emphasis added)
One interviewee described changes made in the design of an applied mathematics

course to include mathematical modelling and problem-solving in order to engage
students who had avoided optional applied mathematics courses during their studies
in favour of “pure” mathematics:
Historically this course was quite boring, let’s say ten years ago, because it did sort of prob-
lems which are usually called static problems, ladders on inclines and slip, some very bor-
ing things and students were really put off by this. So [students] had this course and after
this they thought that applied mathematics is really not interesting at all, it’s just some
boring mechanics that doesn’t contain anything that they would want to engage with.
(L6B, 00.43)
This course is actually distracting students from what applied mathematics should be, so
hence why I say fun. And what we then did over the years, the course was changed.
(L6B, 01.26)
At times, the interviewer focused on the enjoyment perspective asking, for example,
“Is modelling fun for you or your students?” We cite responses from three lectur-
ers below.
Probably a bit of both. Because obviously it is enjoyable teaching something that you your-
self find interesting and exciting and this then shows. (L6B, 29.15)
I think probably I do find all of teaching quite fun but what I want to give to the students is
an idea that the maths is fun. (L3B, 34.19)
If you are able to bring about engagement in the way that [students] are coming and both
asking you and their fellow students, when you experience that they are actively engaging
with the problems, then this brings about enjoyment, both in me, but also I can see that the
motivation is also rising in the students. (L1A, 18.28)
While these three lecturers pay particular attention to the enjoyment of teaching and
doing mathematics and mathematical modelling, the next one focuses on the solu-
tion of modelling tasks. Here the lecturer describes the “fun” of mathematical mod-
elling in terms of obtaining a solution to a difficult task.
I remember what I thought was huge fun when I managed to solve something difficult. But
that was almost never fun initially. […] I think the enjoyment [for students] would come
primarily from the sensation of having solved a difficult task on [their] own, more or less.
Of course, it is an advantage if the problem in itself is self-motivating, if it’s something that
sounds interesting, but I think the main enjoyment would come from actually solving it.
(L3A, 25.20)
Since respondents self-selected for interview, it is likely that those with a strong
interest in mathematical modelling have come forward. However, the strength of the
statements suggests that for this group of lecturers affective aspects of mathematical
modelling significantly influence its use in teaching.
Third, we present results of a literature search. While it is easier to find citations
on beauty and joy of mathematics from mathematicians’ reflections on their prac-
tice, a similar search for mathematical modelling proved much more difficult with
far fewer “hits”. However, there were several strong statements in the context of
modelling that surprised us. In relation to mathematical modelling, quite often utili-
tarian aspects prevail. For example, Dym (2004) directly points to practical aspects
of modelling; this echoes pragmatic views of the lecturer L3A (see quotation above):
Since the modeling of devices and phenomena is essential to both engineering and science,
engineers and scientists have very practical reasons for doing mathematical modeling. In
addition, engineers, scientists, and mathematicians want to experience the sheer joy of for-
mulating and solving mathematical problems. (Dym 2004, p. 4)
On a related note, Lucas (1999) sees the joy of mathematical modelling rather prag-
matically, primarily in communicating the results and their potential
implementation:
One must check whether the model is reasonable. [...] Avoiding this aspect of modeling is
to miss an important element of problem solving, as well as lose an opportunity to be
involved with the joy of seeing the results communicated to others and perhaps imple-
mented in society. (Lucas 1999, pp. 8-9)
Admitting the importance of the practical side of mathematical modelling, Aris

(1997) fairly asks: “Astarita2 suggested that utilitarian ends create their own aes-
thetic, but is not the process guided by it, too?” (p. 2453). Comparing two formulas,
an ugly one, which exactly fits the data, and a more elegant one, which fits the data
far less well, he argues: “Are there any situations in which beauty is taken as a cri-
terion of truth? […] I am inclined to take a middle road and say that aesthetic con-
siderations are indeed dominant, but the judgment based on them needs to be taken
tentatively” (p. 2453). Talking about the role of aesthetic considerations in engi-
neering, Aris concludes:
It is generally conceded that mathematicians recognize that elegance has a legitimate role
in mathematics. Relating this to the engineers’ craft of mathematical modelling brings in
the power and beauty of the dimensionless parameter. Though beauty may not be certain
guide to truth, the exercise of mathematical modelling as a craft certainly leads to the enjoy-
ment of it. (Aris 1997, p. 2447)
Our claim about enjoyment of mathematical modelling would have remained with
little support had we not come across passionate confessions by Aris (1999) who
summarised his long life with mathematical modelling:
2
Giuseppe Astarita, an Italian architect and engineer of the late-Baroque or Rococo period.
Looking back on 40 years of mathematical modeling, my chief reaction is to say what fun
it has been and continues to be. For the exercise of a craft to the best of one’s ability is great
fun as well as being a great joy, and, although joy is a peculiar possession and the province
of friends, fun is its public aspect and may be shared promiscuously. (Aris 1999, p. 419)
One of the enjoyable sides of mathematical modelling is finding solutions to open-

ended problems and, associated with this process, mathematical discoveries. Pollak
(2015) points out that “One of the joys of creating a mathematical model for a real-
world situation is that you can never be sure what mathematics you are going to get
into” (p. 273). Koponen and Tala (2014) emphasise how important it is for students
to solve real problems in order to experience the freedom of creative discovery simi-
lar to that enjoyed by professional mathematicians:
In mathematics and IT lessons, students may engage more easily in studying the dynamics
of models and modelling in the virtual or mathematical world without striving for a direct
one-to-one relationship with the physical world, thus enjoying more freedom to explore
theoretical ideas. Practicing scientists enjoy such freedom, so why not permit the same
freedom and joy of invention in teaching and schooling. (Koponen and Tala 2014, p. 1165)
The significance of enjoyment of mathematical modelling and real-world problem

solving and its impact on students’ performance has been emphasised by Heilio
(2011) who saw it as “an ideal way to build interest and enthusiasm” (p. 487).
Students’ self-reported enjoyment was positively related to performance (Schukajlow
et al. 2012) because they “want to experience joy in mathematical activities and feel
challenged by these” (Vorhölter et al. 2014). We conclude the literature review with
a challenging invitation of Pollak (2015):
A number of people have written books entitled something like “The Joy of Mathematics.”
I should like to see a book entitled “The Joy of Mathematical Modeling” with fifty to a
hundred examples, taken mostly from everyday human experience. The joy I have had in
my life of doing and teaching mathematical modeling should be transmitted. Who will join
me? (p. 275)
This selection of quotations demonstrates an enthusiastic and emotional approach

of a group of mathematicians to practising and teaching mathematical modelling
which, in our opinion, substantiates the need for further research linking affective
aspects with teaching and learning of mathematical modelling.
9.5 Concluding Remarks
With this chapter, we put forward a case for considering the enjoyment and pleasure
from engaging in mathematical modelling tasks as a direction requiring further
research, especially with relevance to university education. We base our claim on
three different types of evidence that, in our view, give support to our call for more
research in this area. We also promote curriculum changes in response to current
discussions.
From our questionnaire study, analysis of two items directly mentioning enjoy-
ment as one of the aims of practising and teaching mathematical modelling points
towards a group of university lecturers for whom affective aspects are important in
their professional activity. Furthermore, in follow-up interviews to the survey, the
significance of enjoyment, “fun” and pleasure, and their impact on practising and
teaching of mathematical modelling has been confirmed by a small number of
mathematics lecturers. Since Kaiser and Sriraman (2006) conducted a classification
of research perspectives, they could not have included an affective category as this
category was not present in the research literature. We believe that our results pro-
vide support for augmenting the existing categorisation with the affective perspec-
tive. In addition, the literature search demonstrates that, while there is a large body
of papers and books discussing the beauty and joy of mathematics, there is also a
small but non-negligible amount of publications on the joy of mathematical
modelling.
Finally, there is widely acknowledged research evidence of the impact of positive
emotions on students’ learning (e.g. Hernandez-Martinez and Pampaka 2017; Lim
et al. 2009; Schukajlow et al. 2012; Vorhölter et al. 2014). Cited research demon-
strates that pedagogical practices that are designed to make positive emotions such
as joy and pleasure a part of teaching contribute to students’ better learning. This is
especially important for teaching and learning of mathematical modelling where
students solve open-ended, real-world problems and may experience the whole
spectrum of emotions – from curiosity and puzzlement to frustration and despair to
pleasure and satisfaction (Goldin 2000). This is also mentioned by interviewees in
our study, who emphasise the joy that might result from problem solving. Further
research is needed relating the affective perspective with the design of mathematical
modelling tasks. Enjoyment and pleasure should be reflected in new approaches to
teaching mathematical modelling and are important for student learning, promoting
university students’ mathematical development and curiosity, and better preparing
them for real-world challenges.
Acknowledgements Research presented in this chapter was supported by a MatRIC (www.mat-

ric.no) small research award, project number 170804. The authors wholeheartedly express their
gratitude to Dr. Paul Hernandez-Martinez of Swinburne University of Technology who equally
contributed to this chapter and whose name, unfortunately, could not appear on the title page due
to ICTMA’s chapter limit.
References
Aris, R. (1997). Reflections on Keats’ equation. Chemical Engineering Science, 52(15), 2447–2455.
Aris, R. (1999). Mathematical modeling. A chemical engineer’s perspective. San Diego:
Academic Press.
Aristotle. Metaphysics. Retrieved from http://classics.mit.edu/Aristotle/metaphysics.1.i.html.
Accessed 1 Sept 2017.
Buchholtz, N. (2013). The eyes to see: Theoretical lenses for mathematical modelling research. In
Davis, P. J., & Hersh, R. (1981). The mathematical experience. Boston: Birkhäuser.
Dym, C. (2004). Principles of mathematical modeling (2nd ed.). San Diego: Elsevier.
Evans, J. (2000). Adults’ mathematical thinking and emotions: A study of numerate practice.
London: RoutledgeFalmer.
Gaither, C. C., & Cavazos-Gaither, A. E. (Eds.). (2012). Gaither’s dictionary of scientific quota-
tions. New York: Springer.
Goldin, G. (2000). Affective pathways and representation in mathematical problem solving.
Mathematical Thinking and Learning, 2(3), 209–219.
Hadamard, J. (1945). The psychology of invention in the mathematical field. New York: Dover
Publications.
Heilio, M. (2011). Modelling and the educational challenge in industrial mathematics. In G. Kaiser,
W. Blum, R. Borromeo Ferri, & G. Stillman (Eds.), Trends in teaching and learning of math-
ematical modelling (pp. 479–488). Dordrecht: Springer.
Hernandez-Martinez, P., & Pampaka, M. (2017). “I did use to like maths...”: Emotional changes
towards mathematics during secondary school education. In U. X. Eligio (Ed.), Understanding
emotions in mathematical thinking and learning (pp. 187–220). London: Elsevier.
Hilton, P. (1992). The joy of mathematics. The College Mathematics Journal, 23(4), 274–281.
Kaiser, G., Sriraman, B., Blomhøj, M., & Garcia, F. J. (2007). Modelling and applications –
Differentiating perspectives and delineating commonalities. In D. Pitta-Pantazi & G. Philippou
(Eds.), Proceedings of the fifth congress of the European Society for Research in Mathematics
Education (pp. 2035–2041). Larnaca: ERME & University of Cyprus.
Koichu, B., Katz, E., & Berman, A. (2017). Stimulating student aesthetic response to mathemati-
cal problems by means of manipulating the extent of surprise. The Journal of Mathematical
Behavior, 46, 42–57.
Koponen, I. T., & Tala, S. (2014). Generative modelling in physics and in physics education: From
aspects of research practices to suggestions for education. In M. R. Matthews (Ed.), International
handbook of research in history, philosophy and science teaching (pp. 1143–1169). Dordrecht:
Springer.
Leron, U., & Hazzan, O. (1997). The world according to Johnny: A coping perspective in math-
ematics education. Educational Studies in Mathematics, 32(3), 265–292.
Lim, L. L., Tso, T.-Y., & Lin, F. L. (2009). Assessing science students’ attitudes to mathematics:
A case study on a modelling project with mathematical software. International Journal of
Mathematical Education in Science and Technology, 40(4), 441–453.
Lucas, W. F. (1999). The impact and benefits of mathematical modeling. In D. R. Shier &
K. T. Wallenius (Eds.), Applied mathematical modeling: A multidisciplinary approach
(pp. 1–26). Boca Raton: CRC Press.
McLeod, D. B. (1992). Research on affect in mathematics education: A reconceptualization. In
D. A. Grouws (Ed.), Handbook of research on mathematics teaching and learning: A project
of the National Council of Teachers of Mathematics (pp. 575–596). New York: Macmillan.
Op’T Eynde, P., De Corte, E., & Verschaffel, L. (2006). “Accepting emotional complexity”:
A socio-constructivist perspective on the role of emotions in the mathematics classroom.
Pollak, H. (2015). The place of mathematical modeling in mathematics education. In G. A. Stillman,
W. Blum, & M. S. Biembengut (Eds.), Mathematical modelling in education research and
practice: Cultural, social and cognitive influences (pp. 265–276). Cham: Springer.
Rodd, M. (2010). Hot and abstract: emotion and learning in undergraduate mathematics. Retrieved
from http://users.math.uoc.gr/~ictm2/Proceedings/pap203.pdf. Accessed 1 Sept 2017.
Schukajlow, S., Leiss, D., Pekrun, R., Blum, W., Müller, M., & Messner, R. (2012). Teaching
methods for modelling problems and students’ task-specific enjoyment, value, interest and self-
efficacy expectations. Educational Studies in Mathematics, 79(2), 215–237.
Schukajlow, S., Kaiser, G., & Stillman, G. (2018). Empirical research on teaching and learn-
ing of mathematical modelling: A survey on the current state-of-the-art. ZDM Mathematics
Education, 50(1–2), 5–18.
Sinclair, N. (2004). The roles of the aesthetic in mathematical inquiry. Mathematical Thinking and
Learning, 6(3), 261–284.
Treffert-Thomas, S., & Jaworski, B. (2015). Developing mathematics teaching: What can we learn
from the literature? In M. Grove, T. Croft, J. Kyle, & D. Lawson (Eds.), Transitions in under-
graduate mathematics education (pp. 259–276). Edgbaston: The University of Birmingham.
Treffert-Thomas, S., Viirman, O., Hernandez-Martinez, P., & Rogovchenko, Y. (2017). Mathematics
lecturers’ views on the teaching of mathematical modelling. Nordic Studies in Mathematics
Education, 22(4), 121–145.
Vorhölter, K., Kaiser, G., & Borromeo Ferri, R. (2014). Modelling in mathematics classroom
instruction: An innovative approach for transforming mathematics education. In Y. Li,
E. A. Silver, & S. Li (Eds.), Transforming mathematics instruction: Multiple approaches and
practices (pp. 21–36). Cham: Springer.
Chapter 10
Re-signifying Function Concept: A Mixed
Methods Study to Understand
the Contributions of the Dialogic Approach
of Ethnomodelling
Milton Rosa, Daniel Clark Orey, and Diego Pereira de Oliveira Cortes
Abstract This investigation was developed in collaboration with a public school

and a local farmer’s market in Brazil. The main objective was to show how dialogi-
cal approaches of ethnomodelling contribute to the re-signification of function con-
cepts with 38 high school students during their interaction with a farmer-vendor and
his labour practices. The research question investigated was: Using a dialogical
approach, what is the contribution that ethnomodelling offers to the re-signification
of function concepts to students enrolled in a public high school in the metropolitan
region of Belo Horizonte? Data were collected and analysed through a mixed meth-
ods approach. The main result from the study confirms ethnomodelling provides an
integrative approach to school mathematics when it considers both etic and emic
mathematical knowledge in a dialogical way.
Keywords Dialogic approach · Emic approach · Ethnomodelling · Etic approach ·

Function concept · Mixed methods study · Re-signifying
10.1 Initial Considerations
This chapter discusses the use of ethnomodelling in the mathematics curriculum. In

this context, ethnomodelling has emerged as a pedagogical action that relates to the
research fields of ethnomathematics and mathematical modelling. Therefore, mod-
elling provides a valuable pedagogical approach appropriate to an ethnomathemat-
ics programme because it contextualises mathematical knowledge developed
locally. Ethnomodelling is the translation of local mathematical ideas, procedures,
M. Rosa (*) · D. C. Orey · D. P. de Oliveira Cortes

Departamento de Educação Matemática – DEEMA, Universidade Federal de Ouro Preto –
UFOP, Ouro Preto, Brazil

108 M. Rosa et al.
and practices in which the prefix ethno is related to the specific mathematical knowl-
edge possessed by the members of distinct cultural groups (D’Ambrosio 2006).
Thus, it is necessary to start with the sociocultural context, reality, and interests of
the students and not by enforcing a set of external values and decontextualised cur-
ricular activities without meaning. This can be considered as a way to mediate the
function knowledge that is culturally rooted in diverse mathematical procedures
related to the labour practices of, say, a farmer-vendor at a market and its connec-
tions to the mathematical knowledge practised by the students in a high school
environment. This approach helps students to understand the importance and the
role of mathematics in society (Rosa and Orey 2017). Consequently, supported by
the theoretical foundations found in ethnomathematics, modelling, and ethnomod-
elling, the following research question was investigated: Using a dialogical
approach, what is the contribution that ethnomodelling offers to the process of re-
signification of the function concept to students enrolled in the second year in a
public high school in the metropolitan region of Belo Horizonte?
10.2 Ethnomodelling: Emic, Etic, and Dialogical Approaches
Ethnomodelling is considered as one form of practical application of ethnomathe-

matics that adds cultural perspectives to mathematical modelling concepts by study-
ing mathematical phenomena within a culture. This process applies both
ethnomathematics and mathematical modelling in order to allow us to perceive real-
ity by using different lenses, which gives us insight into mathematical knowledge
developed in holistic ways (Rosa and Orey 2013).
The linking of modelling and culture is appropriate for an effective holistic
understanding of the development of mathematical ideas, procedures, and practices.
Hence, it becomes important for us to understand the unique roles of local (emic),
global (etic), and dialogical approaches in relation to the development of mathemat-
ical knowledge through modelling.
Emic knowledge is a form of local knowledge unique to given cultural groups or
societies. It is developed overtime in accordance with perceptions and understand-
ings deemed appropriate by the insider’s culture. Emic mathematical knowledge is
dynamic and continually shaped by internal inventivity, creativity, and experimenta-
tion as well as influenced by contact with other external mathematical knowledge
systems (Rosa and Orey 2013). It is us about us.
The emic approach investigates how local people think and reason and problem
solve; how they perceive, explain, and categorise the world; and how they develop
rules for their behaviour. Thus, members of distinct cultures are their own judges of
the validity of the descriptions of their cultural ideas, notions, procedures, and prac-
tices. These individuals are considered as culturally specific (Sue and Sue 2003).
10 Re-signifying Function Concept: A Mixed Methods Study to Understand… 109
Emic knowledge may contrast with etic knowledge and be related to the global
system generated by academic institutions. It is developed according to the concep-
tual schemes and categories that are regarded as meaningful and appropriate by the
community of scientific observers and researchers. It provides a framework for
determining the influences of outsiders’ knowledge on the development of local
mathematical practices (Rosa and Orey 2013). It is them (researchers, educators,
outsiders) about us.
Educators and researchers who apply etic approaches generally look for univer-
sal or culture-free concepts and theories. They search for variables and constructs
common to all cultures, which are compared in order to discover how they might
differ from, or are similar to, each other. These practices are equated with objective
explanations of sociocultural phenomena from external points of view. They are
considered culturally universal (Sue and Sue 2003).
Dialogical knowledge helps us to create synergistic spaces of interdependent,
reflexive, and co-arising relationships between global and local processes. Hence, it
is important that global mathematical ideas, procedures, and practices adapt them-
selves to local cultures and vice versa. This contact of local knowledge with other
external knowledge systems provokes cultural dynamism. In this dynamic, the
members of distinct cultural groups identify and decode local knowledge that has
been acquired from generation to generation, accumulating it and transmitting it. At
the same time, contact with academic knowledge helps them to establish relation-
ships and comparisons between these two types of knowledge (D’Ambrosio 2006).
The dialogical approach enables the translation between emic and etic knowl-
edge systems. In this cultural dynamism, these systems are used to describe, explain,
understand, and comprehend knowledge generated, accumulated, transmitted, and
diffused and internationalised by people from other cultures. It involves a process of
negotiating mathematical meanings expressed between local and global contexts
through translation in the ethnomodelling process (Rosa and Orey 2017).
Therefore, ethnomodelling applies emic and etic approaches through the devel-
opment of a dialogical process of teaching and learning mathematical content
related to functions. A mathematical curriculum based on both emic and etic
approaches provides a more complete understanding of the knowledge of mathe-
matical practices (functions) developed by members of distinct cultural groups,
such as farmer-vendors.
We argue that the ethnomodelling process, including emic knowledge, is con-
cerned with the differences that make mathematical ideas, procedures, and practices
unique from the insiders’ viewpoint. Thus, emic ethnomodels are grounded in what
matters in the mathematical world of those being modelled. However, many ethno-
models are etic in the sense that they are built on data gleaned from the outsiders’
viewpoint. Therefore, etic ethnomodels represent how the modellers think the world
works through systems taken from reality while emic ethnomodels represent how
people who live in such worlds think these systems work in their own reality.
110 M. Rosa et al.
10.3 Re-signifying Function Concept
Re-signification is associated with the understanding that concepts can be perceived

in different ways. This notion is related to the understanding of how individuals re-
elaborate a particular meaning in accordance to the influence of emotional, physi-
cal, intellectual, cultural, and historical contexts (Bandler and Grinder 1982). It also
can be considered as a conceptual reformulation aimed at changing a definition in
relation to phenomena that are experienced by individuals so that they can under-
stand it through different paradigms, thus changing its meaning (Watzlawick et al.
1974) by applying a translational process between the real-world and academic
mathematics (Eglash et al. 2006).
In this process, vom Hofe et al. (2005) argued that translation is considered as the
main mathematical activity because its concepts, procedures, or practices represent
real-life contexts by interpreting its solution in real-world situations. Hence, they
affirmed that essential reasons of some serious problems of mathematical under-
standing are caused by conflicts concerning the intuitive level that are based on the
fact that often mathematical concepts and symbols are filled by students with a
totally different meaning from what is intended by the teachers.
In this study, the re-signification of the function concept was considered as the
insertion of new viewpoints to its meaning. This approach provided an opportunity
for students to produce new knowledge with regard to this mathematical content
through ethnomodelling because the study of functions was directed towards the
understanding of contextualised phenomena that enabled students to re-signify the
function concept according to their understanding of the labour practices of a farmer-
vendor. Thus, it is important that students understand the difference between func-
tion studied in schools with the ideas and notions used in everyday life because this
process enables them to acknowledge and value different mathematical practices.
According to the Brazilian Curricular Parameters for Mathematics (Brasil
1998), it is necessary for teachers to provide students with opportunities to work
with phenomena that deal with the contextualisation of the function concept by con-
structing descriptive models of phenomena that occur in everyday life. Thus, this
approach allows students to make connections with local and academic mathematics
through ethnomodelling by developing the re-signification of mathematical concepts.
10.4 Research Design and Methodological Procedures
This research was conducted in a public school in the metropolitan region of Belo
Horizonte, Minas Gerais, in Brazil, and in a local farmers’ market. The main objec-
tive of this study was to show how the dialogical approach of ethnomodelling con-
tributed to the process of re-signification of the function concept. The data from this
study came from 38 students, aged from 15 to 17 years old, in the second year of high
school, during their interaction with a local farmer-vendor and his labour practices.
Complementing the overall objective of this study, the specific objectives were as
follows: (a) describe how emic, etic, and dialogical approaches of ethnomodelling
are found in meetings between the students and the farmer-vendor and (b) verify
how the mathematical practices of the farmer-vendor can be used in the classrooms
in the development of the educational action of ethnomodelling through the elabora-
tion of ethnomodels by the students.
A mixed methods approach was applied in order to “build on the synergy and
strength that exists between quantitative and qualitative research methods in order
to understand a phenomenon more fully than is possible using either quantitative or
qualitative methods alone” (Gay et al. 2006, p. 490). Both quantitative and qualita-
tive data were collected and analysed by using the methodology of a mixed methods
study by applying the QUAN + QUAL design in order to capitalise on the strengths
of each approach (Creswell and Plano Clark 2007). By mixing the datasets, the
researchers were able to provide a better understanding of the problem under study
than if either dataset had been used alone (Tashakkori and Teddlie 1998).
In order to answer the research question, three blocks of activities (Emic
Mathematical Activities, Transitory Emic-Etic Mathematical Activities, and
Dialogical Mathematical Activities), two questionnaires (initial and final), a semi-
structured interview with the farmer-vendor, and the fieldnotes of the researcher-
teacher were used as data collection instruments in this study. In addition to these
data collection instruments, students visited the farmer-market and then participated
in a seminar with the farmer-vendor. This seminar helped to clarify students’ doubts,
questionings, and queries with regard to the sales practices developed by the farmer-
vendor. The excursion to the farmer-market and the seminar held in the classroom
helped students to establish conjectures and develop their sense of re-signification
of function concepts that were manifested in the labour practices of the farmer-vendor.
In this study, triangulation of design was used concurrently in order to provide
equal priority to quantitative and qualitative data analysis. Thus, the researchers
applied triangulation to ensure validity and reduce the risk of bias and limit the
conclusions of the study by using three different sources for data collection. For
example, Patton (2002) argued that triangulation strengthens the results of a study
by combining different methods, which includes both quantitative and qualitative
research approaches.
10.5 Results and Discussions
After the conclusion of the analytical phase, qualitative data were quantified, which
enabled the interpretation of the results through the elaboration of three thematic
categories: School Environment (Etic), Out of School Environment (Emic), and
Dialogical Environment. The data analysis of these categories revealed that quanti-
tative and qualitative findings complemented each other when the research ques-
tions were addressed and analysed.
112 M. Rosa et al.
The interpretation of the categories showed that during the daily market it is pos-
sible to recognize the development of local mathematical practices that can be mod-
elled by elaborating ethnomodels through the process of ethnomodelling. For
example, mathematical practices developed by the farmer provided pedagogical
resources that involved rapid mental calculations used to solve problems and situa-
tions related to discounts, profits, and losses, this also included the notion of propor-
tional thinking related to functions.
In this study, the market was considered as the multicultural setting in which the
farmers and the students could share and make use of mathematical knowledge that
shaped the transdisciplinarity of this out-of-school environment. The results showed
that the daily life of the market is impregnated with mathematical procedures and
practices rooted in the culture of the farmer-vendor, which are related to quantifica-
tion, measurement, classification, comparison, and model that can be studied during
the conducting of the ethnomodelling process.
An important contribution of ethnomodelling to the process of re-signification of
academic function concepts was to provide an analysis of the local (emic) strategies
applied by the farmer-vendor as well as the academic techniques (etic) employed by
the students in their own cultural context that constitutes ambiences of effective
exchange of mathematical knowledge through the elaboration of ethnomodels. The
farmer-vendor developed an emic ethnomodel by mathematising the calculation of
the price of his products:
Let’s assume that you buy a 10 kg box of tomatoes for 40 reais,1 and the kilogram is sold at
4 reais, thus each 100 grams cost 40 cents, then you cannot sell it at that price because we
have expenses like gas, transportation, employees, packaging, etcetera. Thus, I sell each
kilogram of tomatoes by 5 or 6 reais because it should be more expensive since you do not
go to the market to buy the products and sell them at the same price. In this case, I increased
the price by 25 or 50 percent. Sometimes, I need to sell my products, for example, at 100 or
60 percent more, depending on the price I buy them and the expenses I have. This system is
used to determine the price of any of the products I sell. For example, if I buy a product for
80 or 100 reais each box, then the price of the kilogram should be 16 and 12.80 reais [60%
Mark up] or 20 or 16 reais [100% Mark up].
This emic ethnomodel is in accordance with the perceptions and understandings

deemed appropriate by the farmer-vendor culture in this context. Thus, the primary
goal of an emic approach is a descriptive idiographic orientation of mathematical
phenomena because it places an emphasis on the uniqueness of each mathematical
idea, procedure, or practice developed by the members of diverse cultural groups
(Rosa and Orey, 2013).
In this context, an etic approach provided cross-cultural contrasts and compara-
tive perspectives by using aspects of academic mathematics that translated the
farmer-vendor’s practices in order to create connections and new understandings
related to how individuals from a different cultural background use mathematical
thinking. This approach is necessary to comprehend and explain this particular
1
The Brazilian real or reais (R$) is the official currency of Brazil, which is subdivided into
100 cents.
If CP(m) = 40, then SP(m) = v . m, where 5 ≤ v ≤ 6 CP = Cost Price

SP = Sale Price
If CP(m) = 80, then SP(m) = v . m, where 12 ≤ v ≤ 16
M = Mass (kg) of the product
If CP(m) = 100, then SP(m) = v . m, where 16 ≤ v ≤ 20 V = Variation of price including expenses and
charges
Fig. 10.1 Possible dialogical ethnomodel. (Source: Authors’ personal file)
mathematical practice as a whole from the point of view of the students (outsiders).
For example, students developed an etic ethnomodel, which is an approximation of
the emic ethnomodel used by the farmer-vendor:
A product, whose cost price is 40 reais, has a sale price between R$ 5.00 and R$ 6.00.
Another product, whose cost price is 80 reais, has a sale price between R$ 12.00 and R$
16.00. And a third product whose cost price is 100 reais, has a sale price between R$ 16.00
and R$ 20.00. However, it is important to note that these sales prices may be increased by
other costs related to the market’s expenses.
The interpretation of these results shows that the determination of these prices,
besides being related to the quantity of products purchased, is also bounded to emic
constructs developed by the unique labour experiences of the farmer-vendor.
Figure 10.1 shows the elaboration of a dialogical ethnomodel by the students, which
represents the sale process developed by the farmer-vendor.
Another emic ethnomodel with rhetorical characteristics was observed by the
students in their visit to the farmer’s market with respect to the mental calculations
developed by the farmer-vendor when he solved the following problem: Based on
the price of 8 reais a kilogram, how would you calculate the price of a product that
weighed 1 kilogram and 450 grams? Thus, the farmer-vendor explained that “First,
I put 8 reais aside because I am not very good at doing some calculations. Then, I
know that 200 grams will be 1.60 reais, and twice 1.60 will be 3.20 reais, then I find
the price of 50 grams, which is 40 cents, and the value will be 3.60 reais. Therefore,
the price is 8 plus 3.60 that is 11.60 reais. That is it.”
Although the farmer-vendor did not develop a formal knowledge regarding the
study of functions and its main mathematical characteristics, his experience and
observation of the world around him enabled him to use this concept implicitly in
his work at the farmer’s market. This context allowed students to develop a dialogi-
cal ethnomodel through the mathematisation of this labour practice developed by
the farmer-vendor such as SP(v) = v. m or SP(v) = 8 + v. m, in which 0.1 ≤ v ≤ 0.45.
In this context, Amit and Neria (2010) argued that some mathematical abilities and
the generalisation processes related to it can be revealed in pattern problems. Thus,
linear patterns related to functions can be expressed as ax + b. They can also be
expressed in figurative or numerical representations. These functional patterns are
found in real-life contexts since they are dependent on sociocultural environments.
Students also mathematized the price of some horticultural products sold by the
farmer-vendor at the market by using a functional representation such as “(u) =
1.50(u)” to determine the price of chives, coriander, and parsley and “(u) = 2.50(u)
114 M. Rosa et al.
and 2(u) = 2.50 (2u) – 1.00 = 4.00” to determine the price of lettuce. In this mathe-
matisation process, students defined that variable u represents the lettuce unit and
that 2u represents two units of this product. This approach shows the students’
attempt to translate the promotion offered by the farmer-vendor to his customers in
which two lettuce units are sold at a discount of one real, for example, one lettuce
unit costs 2.50 reais and two lettuce units cost 4 reais.
We have come to use the term translation to describe the process of modelling
local cultural systems (emic) that enables for global academic representations (etic)
(Rosa and Orey 2013). An effective application of ethnomathematics makes use of
modelling in order to establish relations between local conceptual frameworks
(emic) and the mathematical concepts embedded in global practices. This act of
translation as applied in this process is best referred to as ethnomodelling where
mathematical “knowledge can be seen as arising from emic rather than etic origins”
(Eglash et al. 2006, p. 349).
This process of mathematisation used by students shows the existence of dialogi-
cal connections between the knowledge developed in the school environment and
those perceived in the non-school environment of the market. Hence, Rosa and Orey
(2015) argue that the connections between these two contexts should be prioritized
in schools, since they can reduce the gaps between theoretical and practical mathe-
matical knowledge.
In this way, the ethnomodelling allowed for the reconceptualisation and applica-
tion of the function concept through the elaboration of mathematical activities origi-
nating in the sociocultural context of the school community by applying the
ethnomodelling process in the mathematics curriculum. This approach enabled the
dialogical development between the ideas, procedures, and mathematical practices
intrinsic to the labour practices of farmer-vendor (emic approach) and school math-
ematical contents (etic approach) with the use of problem situations that emerged
from the context of a fair dialogical approach.
Rosa and Orey (2015), for example, found similar results in an ethnomodelling
study in which students developed ethnomodels derived from curves along a school
wall to verify if they were related to exponential, parabolic, or catenary functions.
In order to make the necessary arguments to support their conjectures, ethnomodels
were elaborated, analysed, and discussed in the classroom through a methodologi-
cal procedure in which some curves were randomly selected from the wall. After
examining the ethnomodels, they came to the conclusion that the curves on the wall
built during slavery were approximations of catenary curves.
The interpretation of the re-signification of function concepts became evident
because students were able to transcend the concepts of function presented in the
etic context of academic mathematics. This transcendence was characterized by the
development of emerging dialogic ethnomodels, since their emic elaboration was
unrelated to the etic approach. For example, Bandler and Grinder (1982) argued that
the act of re-signifying is associated with the understanding that members of dis-
tinct cultural groups perceive contents and situations in different ways.
10.6 Final Considerations
When researchers investigate knowledge possessed by the members of distinct cul-

tural groups, they are able to find unique mathematical ideas, procedures, and prac-
tices that may be considered as ethnomathematics, which are used to express the
relationship between culture and mathematics (D’Ambrosio 2006) through ethno-
modelling. In the context of this study, we argue that ethnomodelling added cultural
perspectives to the modelling process.
In the process of ethnomodelling, the promotion of dialogue (re-signification of
function concept) between emerging knowledge (farmer-vendor) and existing math-
ematics (function concept) is important to enable the approximation of this knowl-
edge through the proposition of contextualized mathematical activities. In this
cultural dynamism, local knowledge has interacted dialogically with the knowledge
consolidated by the academy, developing a reciprocal relationship between these
two approaches.
Therefore, it is important that members of distinct cultural groups are able to re-
interpret the world, re-design or re-signify mathematical concepts, and understand
different points of view in order to produce internal descriptions of mathematical
knowledge. In order for this process to be successful, it is important to elaborate
ethnomodels in order to link the theory into practice and include a dialogical
approach in the mathematics curriculum.
In conclusion, emic, etic, and dialogical approaches of an ethnomodelling are
essential in the process of (re)designing pedagogical tools that are integrative, par-
ticipative, and relevant and use self or community as an object of learning mathe-
matics. This approach can be developed through the elaboration of activities based
on the ideas, procedures, and mathematical practices developed by the members of
distinct cultural groups through ethnomodelling.
References
Amit, M., & Neria, D. (2010). Assessing a modeling process of linear pattern task. In R. Lesh,
P. L. Galbraith, C. R. Haines, & A. Hurford (Eds.), Modeling students’ mathematical modeling
competencies (pp. 155–166). New York: Springer.
Bandler, R., & Grinder, J. (1982). Reframing – NLP and the transformation of meaning. Boulder:
Real People Press.
Brasil. (1998). Parâmetros curriculares nacionais: mathematics [National curricular parameters:
Mathematics]. Brasília: MEC/SEF.
Creswell, J. W., & Plano Clark, V. L. (2007). Designing and conducting mixed-methods research.
Thousand Oaks: Sage.
D’Ambrosio, U. (2006). Ethnomathematics: Link between tradition and modernity.
Rotterdam: Sense.
Eglash, R., Bennett, A., O’Donnell, C., Jennings, S., & Cintorino, M. (2006). Culturally situ-
ated designed tools: Ethnocomputing from field site to classroom. American Anthropologist,
108(2), 347–362.
116 M. Rosa et al.
Gay, L. R., Mills, G. E., & Airasian, P. W. (2006). Educational research: Competencies for analy-
sis and applications. Upper Saddle River: Merrill/Prentice Hall.
Patton, M. Q. (2002). Qualitative research and evaluation methods. Thousand Oaks: Sage.
Rosa, M., & Orey, D. C. (2013). Ethnomodelling as a research lens on ethnomathematics and
modelling. In G. A. Stillman, G. Kaiser, W. Blum, & J. P. Brown (Eds.), Teaching mathematical
Rosa, M., & Orey, D. C. (2015). Modelling the wall: The mathematics of the curves on the wall
of Colégio Arquidiocesano in Ouro Preto. In G. A. Stillman, W. Blum, & M. S. Biembengut
(Eds.), Mathematical modelling in education research and practice: Cultural, social, and cog-
Rosa, M., & Orey, D. C. (2017). Ethnomodelling as the mathematization of cultural practices.
In G. A. Stillman, W. Blum, & G. Kaiser (Eds.), Mathematical modelling and applications:
Crossing and researching boundaries in mathematics education (pp. 133–142). Cham:
Springer.
Sue, D. W., & Sue, D. (2003). Counseling the culturally diverse: Theory and practice.
New York: Wiley.
Tashakkori, A., & Teddlie, C. (1998). Mixed methodology: Combining qualitative and quantitative
approaches. Thousand Oaks: Sage.
vom Hofe, R., Kleine, M., Blum, W., & Pekrun, R. (2005). The effect of mental models (“grund-
vorstellungen”) for the development of mathematical competencies. First results of the longitu-
dinal study Palma. In Proceedings of the Fourth Congress of the European Society for Research
in Mathematics Education. Working Group 1: The role of metaphors and images in the learning
and understanding of mathematics (pp. 142–151). Sant Feliu de Guíxols: CERME 4.
Watzlawick, P., Weakland, J. H., & Fisch, R. (1974). Change: Principles of problem formation and
problem resolution. New York, NY: Norton.
Part II
Research into, or Evaluation of, Teaching
Practice in Mathematical Modelling
Education
Chapter 11
Sense-making with the Mathematical
Modelling Process: Developing
a Framework for Faculty Practice
Todd Abel, Mary Elizabeth Searcy, and Tracie Mc Lemore Salinas
Abstract The context of sense-making is used to frame the activities of a study

group of secondary school teachers, post-secondary faculty, mathematicians, and
mathematics educators working to characterise engagement with mathematical
modelling within the classroom. The perspectives of the participants and points of
discussion illuminate aspects of the mathematical modelling process that are diffi-
cult to reconcile, such as dealing with ambiguity and uncertainty and distinctions
between interpretation and computation. The group organised their experiences and
knowledge into a framework for faculty practice that describes characteristics of
mathematical modelling tasks and student work at three different levels within four
stages of a mathematical modelling process.
Keywords Faculty practice · Framework · Mathematical modelling · Sense-

making · Teaching
11.1 Background and Context
In his pivotal work on sense-making, Karl Weick (1995) argued that sense-making
is about “placement of items into frameworks, comprehending, redressing surprise,
constructing meaning, interacting in pursuit of mutual understanding, and pattern-
ing” (p. 6). Sense-making involves organising data and dealing with complexity and
chaos (Weick et al. 2005) and is an ongoing, retrospective, social process that is
grounded in an individual’s sense of identity. Talk and actions occur throughout the
T. Abel (*)
Department of Mathematics, University of Central Arkansas, Conway, AR, USA
e-mail: tabel1@uca.edu
M. E. Searcy · T. M. L. Salinas
Department of Mathematical Sciences, Appalachian State University, Boone, NC, USA
e-mail: searcyme@appstate.edu; salinastm@appstate.edu

120 T. Abel et al.
process; however, “acting is an indistinguishable part of the swarm of flux until talk
brackets it and gives it some meaning” (p. 412). During sense-making, certain cues
are noticed, and experiences are labelled and categorised. Often individuals engage
in sense-making when an ongoing expected activity is interrupted by some external
stimuli (Weick 1995). The introduction of mathematical modelling standards within
the United States’ Common Core State Standards for Mathematics (CCSSM)
(National Governors Association Center for Best Practices and Council of Chief
State School Officers 2010) was such an interruption for many secondary teachers’
expectations of mathematics teaching. Similar interruptions are experienced by
other teachers when mathematical modelling is integrated into curricula in other
nations as well. As in the case of the classroom teachers described in this chapter,
these interruptions often lead to occasions of ambiguity and uncertainty (Weick
1995). Ambiguity happens when an ongoing experience “supports several different
interpretations at the same time” (pp. 91–92). Teachers affected by the CCSSM
modelling mandates certainly faced ambiguity when teaching mathematical model-
ling, in that they often must do so based on limited experience with it (Goos 2014;
Manouchehri et al. 2017). The result is uncertainty, which is not so much about
interpretations as it is about what happens next in the experience and “implies a lack
of predictability, of structure, of information” (Rogers 2003, p. 6).
Pirolli and Russell (2011) stated that some tasks “are sufficiently complex to
require multiple perspectives and talents to understand and make sense of the data
available” (p. 5). This chapter describes a process of collective sense-making
wherein secondary teachers, with community college and university faculty, dis-
cussed and reconciled their varied experiences and understandings of mathematical
modelling. Within such sense-making, information is filtered and meaning becomes
a series of approximations – dealing more with plausibility than accuracy (Weick
et al. 2005). The perspective on modelling within this chapter emphasises student
ability to apply mathematics to real-world situations as described by Pollak (Gould
et al. 2012). However, modelling is treated not just as content to be learned but also
as a vehicle for learning mathematics (Julie and Mudaly 2007). The Realistic
Mathematics Education perspective (Freudenthal 1968) of guided reinvention, and
particularly Gravemeijer’s (1999) emergent modelling approach to instructional
design, informed the consideration of mathematical modelling as the latter and
guided the construction of mathematical modelling tasks to focus on the discussion
described below. These mathematical modelling perspectives acknowledge that sev-
eral different interpretations may be simultaneously supported by a context and that
next steps may not be immediately clear. Hence, within these experiences con-
structed to address classroom teachers’ ambiguity and uncertainty in teaching math-
ematical modelling, they “continued redrafting an emerging story” (Weick et al.
2005, p. 412) as they struggled to confront and characterise the ambiguity and
uncertainty inherent in the modelling process itself.
11 Sense-making with the Mathematical Modelling Process: Developing a Framework… 121
11.1.1 The Mathematical Modelling Study Group
The participant group began as part of the Appalachian Mathematics Partnership

(AMP), a grant-funded professional development project for secondary and middle-
grade teachers. A smaller teacher learning group focused on mathematical model-
ling developed within the project and was eventually supported by a mini-grant
through the North Carolina Community College System. As a result, the group drew
in mathematicians, mathematics educators, university teaching faculty, and com-
munity college faculty to collaborate on sense-making activities related to teaching
and learning mathematical modelling. This chapter will refer to that collection of
individuals as the mathematical modelling study group. Throughout the year during
which this group worked together, a total of six classroom teachers (CT), three com-
munity college faculty (CCF), three university teaching faculty (UTF), two univer-
sity mathematics educators (UME), and five university applied mathematicians
(UAM) participated in the collaborative work. While all these individuals contrib-
uted at some point, not all were present for all activities.
11.2 Engaging in the Sense-Making Process
11.2.1 Foundations of the Sense-Making Process
The genesis of the process was with a teacher learning group consisting of one uni-
versity mathematics educator and six classroom teachers. The group sought to bet-
ter understand mathematical modelling and its implementation within classrooms.
Two documents used during the AMP project, within which this initial group of
teachers originated, were particularly influential. Guidelines for Assessment and
Instruction in Statistics Education (GAISE) is a report on statistics education from
the American Statistical Association (Franklin et al. 2007) that considers the statisti-
cal process in four stages. An accompanying framework breaks student engagement
in each stage into three sequential levels describing increasing depth of understand-
ing and sophistication of methods. Similarly, Hull et al. (2012), in their Standards
for Mathematical Practices Proficiency Matrix, outline initial, intermediate, and
advanced descriptions of student mathematical practice within each of the standards
for mathematical practice found in CCSSM. For standard 4 (model with mathemat-
ics), the initial description includes the use of models and the translation of the
solution. Subsequent levels incorporate an explanation of the representation (inter-
mediate) and the use of a variety of models in demonstrating a solution (advanced).
The structure of sequential levels became a central framing device for the group’s
sense-making on mathematical modelling.
122 T. Abel et al.
11.2.2 Representation of the Mathematical Modelling Process
Before the full study group coalesced around the teacher learning group, teachers
considered multiple interpretations of the mathematical modelling process to make
sense of its various descriptions. The teachers were conscious of the dual perspec-
tives on modelling both as content and vehicle for learning other content (Julie and
Mudaly 2007), particularly within CCSSM. Bliss et al. (2014) imply a movement
between real-world scenarios and mathematics being used to describe them. The
notion of modelling as a process moving between “real-world” situations and math-
ematical representations (Gould et al. 2012) is similarly emphasised in many other
process diagram representations of mathematical modelling (Bliss et al. 2014; Blum
2011; Giordano et al. 2008; Zbiek and Conner 2006). All process diagrams are nec-
essarily simplifications of the process and therefore emphasise different aspects of
it (Cirillo et al. 2016). With this in mind, teacher participants ultimately chose to
base their sense-making around the CCSSM-described process (see Fig. 11.1). This
simplistic representation of a complex process shows movement in a single direc-
tion around the cycle. The choice of the CCSSM cycle representation emphasises
the pragmatism of the teachers’ approach to mathematical modelling. Though they
recognised that other representations might be richer or more nuanced, those advan-
tages were outweighed by the fact that classroom content was required to align with
CCSSM standards. The teacher learning group was thus most concerned with prac-
tical implementation of mathematical modelling within their own particular
classrooms.
11.2.3 Initial Framework Development
After the initial teacher learning group negotiated and selected representations of
mathematical modelling, a facilitated follow-up meeting focused on adding detail to
the framework by proposing descriptions for each level within each phase. These
sense-making conversations were generated and focused by engagement with a col-
laborative modelling task. Each stage was initially conceived by participants in
Fig. 11.1 CCSSM representation of the mathematical modelling cycle (p. 72). (Used under Public
License. © Copyright 2010, National Governors Association Center for Best Practices and Council
of Chief State School Officers. All rights reserved)
small groups, with resulting characterisations vetted by the larger group. The pro-
cess led to a first iteration of the modelling framework.
While discussing the formulate stage, the group focused in part on what informa-
tion students would or should be offered by an enacted task:
UTF 1: I think that [information that students identify as possibly useful that
may not actually be] helps them practice how to discard what’s not
important and keep what is important. I think that’s good.
CT 1: So part of the modelling process is to think about what I can use or not
and to adapt what is or isn’t useful.
UTF 1: I think we have to give them time to fail and to think about this. I like
when they fail and we can turn this [the failure] into next steps.
Moreover, levels within the formulate stage were differentiated based in part on how
well-defined the initial modelling question is. One secondary teacher gave a descrip-
tion of the formulate phase:
CT 2: Level [A]: I told you what to do and how to do it. Level [B]: I told you
what to do but not how to do it. Level [C]: I didn’t tell you what to do,
and I didn’t tell you how to do it.
The small group’s initial characterisation of the compute phase also emphasised
the degree to which students were required to independently engage with the model.
One teacher participant, in discussing an example of matrices for geometric trans-
formations, clarified “if they memorise the matrix, then that is level A, but if they
have to create it, that is level B [or C]”. The group proposed that the amount of guid-
ance from the task or instructor might distinguish between B and C.
Distinctions between interpret and validate phases were difficult for the group to
capture. This may be, in part, because, as one participant noted: “we spend too
much time in the formulate and compute phases”. Beginning with a prompt to iden-
tify what activity constituted “interpret” within the sample problem, the following
exchange captures the difficulty in distinguishing between the two phases:
CCF 1: You have to interpret all the different things you’ve been doing, inter-
preting the questions you’ve asked at the beginning.
UAM 1: [After some other discussion]: Two questions you can ask – one I think
of as interpret and one as validate. For interpret: Did you answer cor-
rectly? Did you answer the question you asked? And for validate: Did
you ask the correct question in the first place?
The university mathematician’s comment guided the subsequent discussion around
interpretation versus validation. Other distinctions in language use became appar-
ent. For example, while exploring the formulate stage, it became clear that there
were different uses of the word “formula”, which in turn translated into differences
in the understanding of what constituted a model.
CT 1: Up to calculus, many of the models are formula-driven.
UTF 1: I think of algorithms as formula-driven.
124 T. Abel et al.
CT 1: But not all algorithms are formulas.

UAM 1: And I think of it as a mathematical structure. It could be a formula; it
could be a computational algorithm.
CT 1: If you can condense it into a single function or system, then it is a for-
mula. If you cross into computing algorithm, then that to me is some-
thing different.
In negotiating the distinctions in definitions, the group was engaging in the pro-
cess of developing meaning through a series of approximations (Weick et al. 2005).
In fact, the focus on language prompted some participants to acknowledge the
impact their own choice of language had on the sense-making process, with one
university faculty member commenting, “I also worried from the other meeting
notes that my [personal] language from that meeting might have been too [promi-
nent in the descriptions]. It might have introduced terms like ‘sensitivity’ that teach-
ers don’t understand.” Collaborative sense-making was thus challenged by the
complexity of the topic but also by the difficulty of clear communication among
the group.
The discussion as a whole highlighted the group’s attention to mathematical
modelling as a practice (content to be learned) with which students develop facility
as they learn content (modelling as a vehicle). In addition, a tension is developed
between descriptions of mathematical modelling tasks as intended or enacted and
descriptions of student work resulting from engaging in modelling tasks. Ultimately,
the group decided to provide two sets of characteristics for each level within each
phase: characteristics of the task or activity and characteristics of student work.
Thus, the group’s sense-making framework shifted to accommodate additional
understanding. The skeleton of this iteration of the framework is shown in Fig. 11.2,
and initial descriptions for each level were added.
Fig. 11.2 Structure of the framework after first mathematical modelling study group meeting
11.2.4 Framework Refinement
After the initial meeting of a subset of the mathematical modelling study group, the
framework draft was presented to a group of mathematicians, most of whom con-
duct research in and teach coursework on mathematical modelling. The group
reviewed the emerging framework, comparing their own definitions of the stages of
modelling with those crafted by the previous group. The question of how to distin-
guish between interpretation and validation remained difficult, with one university
mathematician claiming that “most people see them as the same thing”. The group
ultimately characterised the validate phase as focusing on the model itself, rather
than on the answers provided by the model. In particular, by considering mathemat-
ical modelling as moving between a “real” world and a mathematical realm, the
group characterised interpret as occurring within the mathematical realm, while
validate occurs within a recontextualised problem. In fact, they came to see a higher
level interpret task as prompting movement towards validate. This group also
emphasised flexible movement between stages, as shown in Fig. 11.3. In the reality
of mathematical modelling, one might move from one phase to any other phase,
more flexibly moving between the real and mathematical realms, for instance.
Later, in a professional development session, a different group of K-12, commu-
nity college, and university educators, led by members of the mathematical model-
ling study group, interacted with a modelling task and then reflected on their
experience with the process. They were then presented with the framework and
provided opportunities to examine it as a tool for their respective educational levels.
One theme that emerged was how to best use the framework. Though conceived as
a sense-making tool that allowed the mathematical modelling study group to recon-
cile the complexity of mathematical modelling, some secondary school mathemat-
ics teachers in this group wanted to treat it as a prescriptive rubric. Similarly, it is
tempting to treat the mathematical modelling process as an algorithm. While the
framework itself is useful for engaging in modelling in the classroom (Abel and
Salinas 2017; Abel et al. 2016), it is still a sense-making artefact, and ambiguity and
uncertainty will remain. The framework was, in that sense, similar to a mathemati-
cal model – users were more eager to embrace the explanatory nature than they were
to acknowledge or investigate the degree to which it was a simplification.
Fig. 11.3 Altered

visualisation of the
mathematical modelling
process showing more
flexible movement
between phases
126 T. Abel et al.
11.2.5 Vetting the Framework
Because sense-making is grounded in the individual’s sense of identity, participants

vetted the framework from their particular perspectives throughout the process.
These perspectives offered additional contextual factors within which the modelling
framework would be implemented. One frequent theme was the pressures that
instructors felt related to course pacing, testing, and time. Classroom teachers
referred to the pressures of “covering” their required curriculum, which implies that
all students accomplish the same outcomes. One participant explained:
CT3: I could see this as a place where there is a separation in how we do this
or how the colleges would do it. When you have particular goals, and I
am going around the classroom to see if they [the students] are there in
their groups, it’s a temptation to say, “Oh, you are not learning what you
are supposed to.”
Clearly, this approach precludes opportunities to conduct open-ended explora-
tions in which students may use different strategies and arrive at different models.
Interestingly, even the university level faculty reflected on the pressures they felt in
completing course requirements, though their perspectives are somewhat different:
UAM 1: We have things to teach, too, and sometimes we have a choice. Really it
comes down to, ‘Is this a math major?’, and they need line, slope, and
move on. Or is this a non-math major where we have the luxury of hav-
ing some space to think and explore. The irony is that the math majors
are the ones who then don’t get the exploration.
The pressures of time and testing also impacted participants’ views on vertical
alignment. While exploring the compute phase, participants questioned at what edu-
cational level students should recognise “messiness” in the compute/analyse phases.
One university-applied mathematician suggested that “the earlier the better, but the
constraints of the curriculum really hinder that”. Classroom teachers seemed to
agree, and a community college faculty member later added, “I think the vertical
alignment is kind of pointless because [consider] my students who are doing third
grade math vs. a calculus class – what should my expectations be?”, implying that
differences in mathematical content may be so disparate that they impact the nature
of mathematical modelling itself at different levels.
11.3 The Framework
The mathematical modelling study group’s sense-making converged into a frame-

work for faculty practice that expands on the CCSSM mathematical modelling
cycle as a practical tool for teachers, providing guidance for task selection and
sequencing and for predicting student responses. Similar to the GAISE framework,
the resulting construct presents two dimensions of mathematical modelling: the four
stages of the modelling process and the developmental levels that accompany them.
The developmental levels transition from Level A to Level C, and characteristics

clarify both the nature of the task that aligns with the developmental stage and the
student work that demonstrates it. Thus, the levels provide a sense of student pro-
gression in mathematical modelling fluency. Descriptions of levels within each of
the four stages enrich the CCSSM depiction by connecting it to representations that
focus more closely on cognition (Stillman et al. 2007; Zbiek and Conner 2006). The
full framework is too detailed to fit in this space but may be found online at the fol-
lowing link: http://mathsciamp.appstate.edu/node/49. The language within the
framework was refined over the course of the development process, but evidence of
the sense-making discussions described above may be seen within it.
11.4 Conclusions
Weick (2015) acknowledges that “organizing itself is 1. Always ambiguous. 2.

Always a substitution. 3. Always interrupted. 4. Always relational” (p. 120). He
points out that ambiguity is a permanent part of the process and that interpretation
will always change, while problems can occur when sense-makers assume other-
wise. The final framework for faculty practice is the result of sense-making among
mathematicians and educators at multiple levels. It organises that group’s experi-
ences, understandings, and knowledge into a framework that can be used to assess
modelling tasks, understand student engagement, and guide classroom mathemati-
cal modelling implementation. Multi-stakeholder conversations about mathematical
modelling during the sense-making process also illuminated several obstacles to
developing shared understandings and highlighted the potential benefits of this type
of collective sense-making (see also Abel and Salinas 2017). However, at the end of
the process, the extent to which ambiguity was embraced by the mathematical mod-
elling study group remains unclear. As noted earlier, some secondary teachers in the
group tended to default to using the framework as a grading rubric with clear and
unambiguous indicators. This may be evidence of the strong influence of previously
held sense-making teaching constructs that are resistant to acknowledging inherent
complexity.
The work described in this chapter therefore demonstrates the rich promises of
collective sense-making activities across different communities of interest while
illuminating the topics within the mathematical modelling process that are difficult
points to clarify. At the same time, the sense-making process described here is an
example of how sense-making within the processes of mathematical modelling and
teaching mathematical modelling is typical of sense-making in general – useful and
important for understanding but in need of constant re-working and re-definition in
response to the inherent ambiguity. Thus, the framework cannot be considered a
finished work but can help communities of interest engage with, and manage, the
continuous ambiguity present in both teaching and learning of mathematical
modelling.
128 T. Abel et al.
References
Abel, T., & Salinas, T. M. (2017). Challenges to agreeing how to teach mathematical modeling.
MAA Focus, 37(3), 16–19.
Abel, T., Baird, A., Hirst, H., & Salinas, T. M. (2016). Introducing students to the mathematical
modeling process. The Centroid, 41(2), 13–17.
Bliss, K. M., Fowler, K. R., & Galluzzo, B. J. (2014). Math modeling: Getting started and getting
solutions. Philadelphia: Society for Industrial and Applied Mathematics (SIAM).
Blum, W. (2011). Can modelling be taught and learnt? Some answers from empirical research. In
G. Kaiser, W. Blum, R. Borromeo Ferri, & G. Stillman (Eds.), Trends in teaching and learning
of mathematical modelling (pp. 15–30). New York: Springer.
Cirillo, M., Pelesko, J., Felton-Koestler, M., & Rubel, L. (2016). Perspectives on modeling in
school mathematics. In C. Hirsch & A. McDuffie (Eds.), Annual perspectives in mathematics
education: Mathematical modeling and modeling mathematics (pp. 3–16). Reston: National
Council of Teachers of Mathematics.
Franklin, C., Kader, G., Mewborn, D., Moreno, J., Peck, R., Perry, M., & Scheaffer, R. (2007).
Guidelines for assessment and instruction in statistics education (GAISE) report. Alexandria:
American Statistical Association.
Freudenthal, H. (1968). Why to teach mathematics as to be useful. Educational Studies in
Mathematics, 1(1), 3–8.
Giordano, F., Fox, W., Horton, S., & Weir, M. (2008). A first course in mathematical modeling (4th
ed.). Boston: Brooks Cole.
Goos, M. (2014). Researcher-teacher relationships and models for teaching development in math-
ematics education. ZDM Mathematics Education, 46(2), 180–201. https://doi.org/10.1007/
s11858-013-0556-9.
Gould, H., Murray, D., & Sanfratello, Y. (Eds.). (2012). Mathematical modeling handbook.
Bedford: The Consortium for Mathematics and it Applications.
Gravemeijer, K. (1999). How emergent models may foster the constitution of formal mathematics.
Mathematical Thinking and Learning, 1(2), 155–177.
Hull, T. H., Miles, R. H., & Balka, D. S. (2012). The common core mathematics standards:
Transforming practice through team leadership. Thousand Oaks: Corwin Press.
Julie, C., & Mudaly, V. (2007). Mathematical modelling of social issues in school mathematics in
South Africa. In W. Blum, P. Galbraith, H. Henn, & M. Niss (Eds.), Modelling and applications
in mathematics education: The 14th ICMI study (pp. 503–510). New York: Springer.
Manouchehri, A., Yao, X., & Sağlam, Y. (2017). Mathematical modeling for teaching: An explor-
atory study. In E. Galindo & J. Newton (Eds.), Proceedings of PMENA 29. Indianapolis:
Hoosier Association of Mathematics Teacher Educators.
National Governors Association Center for Best Practices & Council of Chief State School
Officers. (2010). Common core state standards for mathematics. Washington, DC: Author.
Pirolli, P., & Russell, D. M. (2011). Introduction to this special issue on sense-making. Human-
Computer Interaction, 26, 1–8.
Rogers, E. M. (2003). Diffusion of innovation (4th ed.). New York: The Free Press.
Stillman, G., Galbraith, P., Brown, J., & Edwards, I. (2007). A framework for success in imple-
menting mathematical modelling in the secondary classroom. In J. Watson & K. Beswick
(Eds.), Proceedings of MERGA30 (Vol. 2, pp. 688–707). Adelaide: MERGA.
Weick, K. E. (1995). Sense-making in organizations. Thousand Oaks: Sage.
Weick, K. E. (2015). Ambiguity as grasp: The reworking of sense. Journal of Contingencies and
Crisis Management, 23(2), 117–123.
Weick, K. E., Sutcliffe, K. M., & Obstfield, D. (2005). Organizing and the process of sense-
making. Organization Science, 16(4), 409–421.
Zbiek, R., & Conner, A. (2006). Beyond motivation: Exploring mathematical modeling as a con-
text for deepening students’ understandings of curricular mathematics. Education Studies in
Mathematics, 63, 89–112.
Chapter 12
Mathematical Modelling with a Solution
Plan: An Intervention Study about
the Development of Grade 9 Students’
Modelling Competencies
Catharina Beckschulte
Abstract A possible approach to support students during modelling activities is to

supply them with a ‘solution plan’ – a simplified modelling cycle that provides strate-
gies in each step and thus serves as a metacognitive strategic guidance. This chapter
reports on an intervention study that investigates if modelling with a newly developed
five-step solution plan influences the development of students’ modelling competen-
cies. The study was carried out in a pre-post-follow-up design with 29 classes of
German secondary schools and evaluated within the frame of item response theory
using a test instrument focussing on sub-competencies of modelling. Results regard-
ing the subcompetencies simplifying, mathematising, interpreting, and validating are
presented. The results reveal differences in the development of competencies between
students working with a solution plan and students working without such an instrument.
Keywords Interpreting · Mathematical modelling (sub-competencies) ·

Mathematising · Metacognitive strategic instrument · Modelling competencies ·
Simplifying · Solution plan · Validating
12.1 Sub-competencies of Mathematical Modelling
Mathematical modelling is part of the general mathematical competencies in

German educational standards for mathematics and thus plays an important role in
teaching and learning mathematics in Germany (for an overview of the German
discussion on modelling, see Greefrath and Vorhölter 2016). However, mathematical
C. Beckschulte (*)
Institute of Education in Mathematics and Computer Science, University of Münster,
Münster, Germany
e-mail: ca.beckschulte@gmx.de

130 C. Beckschulte
modelling is still a challenge for many students (Blum 2015). Hence, this chapter
discusses a teaching approach called ‘solution plan’ for supporting students during
modelling activities and the impact of this instrument on the development of stu-
dents’ modelling competencies.
There is a wide range of different notions on the definition of modelling compe-
tency in the international discussion. Following the Danish KOM project, mathe-
matical modelling competency is defined as being able to autonomously and
insightfully carry out all aspects of a mathematical modelling process in a certain
context (Blomhøj and Kjeldsen 2006; Niss 2004). A modelling process can be illus-
trated as a cyclical sequence to describe how students approach a modelling task
concerning different sub-processes. The ability to perform such a sub-process can
be seen as a partial competence of modelling (Kaiser 2007; Maaß 2006), so that
mathematical modelling competency includes sub-competencies. Following the
definition of Maaß (2006, p. 116), there are five different sub-competencies which
are fundamental for the study presented:
• Understanding/Simplifying: Competencies to understand the real problem and to
set up a model based on reality.
• Mathematising: Competencies to set up a mathematical model from the
real model.
• Working Mathematically: Competencies to solve mathematical questions within
this mathematical model.
• Interpreting: Competencies to interpret mathematical results in a real situation.
• Validating: Competencies to validate the solution.
Those sub-competencies can be seen as a necessary requirement for successfully
carrying out a whole modelling process. However, as we know from several empiri-
cal studies, it has been shown that there are potential ‘blockages’ for students in
each sub-process of modelling (Galbraith and Stillman 2006; Stillman 2011;
Stillman et al. 2010). Blum (2015) states that a lack of metacognitive competencies
could lead to problems as students often do not have strategies available for solving
real-world problems.
12.2 Modelling with Strategic Instruments
12.2.1 ‘Solution Plan’ as a Teaching Approach
As stated above, modelling is a difficult process for most students, and these diffi-
culties as well as ways to support students during modelling activities have been
well researched. One approach is to give students a metacognitive strategic instru-
ment named a ‘solution plan’ (see Blum 2011) which can be seen as a simplified
modelling cycle for students. Providing strategies within every step of modelling
stimulates various metacognitive activities (Blum 2011). The help given by this
12 Mathematical Modelling with a Solution Plan: An Intervention Study about… 131
instrument can be seen as general strategic assistance (Zech 1998) as it does not
give content-related advice but instead refers to general modelling methods. Thus,
it is possible to provide a solution plan within modelling activities in different con-
tent areas.
A solution plan should be a systematic scaffold for the individual learner
(Schukajlow et al. 2015). Smit et al. (2013, p. 817) describe scaffolding as a tempo-
rary support that helps students to perform a task they cannot complete by them-
selves and that is intended to bring students gradually to a state of competency in
which they can complete a similar task independently. In terms of a scaffolding
instrument, a solution plan does not attempt to overcome short-term local difficulties
but instead aims at developing long-term competencies. As soon as students know
about the solution steps and strategies in a modelling process, they might no longer
require a strategic instrument. A solution plan is not meant to be a schema that has
to be used by students but as an aid for difficulties that might occur during the mod-
elling process (Blum 2011). The goal is that students work independently on model-
ling tasks, whereas teachers support students in an adaptive way with the help of
strategy-oriented interventions related to the solution plan (Schukajlow et al. 2015).
12.2.2 Studies on the Use of Solution Plans
Many different versions of strategic instruments for supporting students during

modelling activities were developed, and there are some interesting empirical
results concerning the use of such methods in class. Maaß (2004) found, in a quali-
tative study with grade 7 and 8 students in Germany, that students who worked with
a simplified modelling cycle felt they were aided by knowing about the modelling
process and the cyclical representation. Another qualitative study using a five-step
problem-solving method – a solution plan consisting of the steps understanding the
problem, devising a plan, carrying out the plan, explaining the result, and checking
the process – showed that grade 6 students felt supported by this method
(Greefrath 2015).
In the DISUM project,1 Blum and colleagues developed a method for students
named the ‘solution plan’ (Blum 2011, p. 24). It consists of four steps, namely
understanding the task, establishing the model, using mathematics and explaining
the results, where each step is explained by bullet points. It could be shown that
grade 9 students who worked with the four-step scheme in the area of the ‘Pythagorean
Theorem’ increased their modelling competencies, whereas students who worked
without the solution plan method did not develop their competencies. Regarding
competencies in interpreting mathematical results, they found the solution plan had
an impact on the development of this sub-competency (Schukajlow et al. 2015).
1
The DISUM project (‘didactical intervention modes for mathematics teaching oriented towards
self-regulation and directed by tasks’) was founded in Kassel and is led by W. Blum, R. Messner,
and R. Pekrun.
132 C. Beckschulte
Including heuristic worked-out examples in modelling problems is another pos-

sibility to support students (see Zöttl et al. 2010). Worked-out examples are tasks
that are already solved to a certain step. Combined with a three-step scheme, namely,
understanding the task, calculating, and explaining the result, Zöttl et al. (2010)
examined participants’ modelling competencies in the KOMMA project finding
these increased significantly during the project, but long-term effects were much
smaller.
12.2.3 Development of a New Strategic Instrument
Based on existing strategic instruments and related studies, we developed a new

solution plan consisting of five steps (see Fig. 12.1) within the frame of the project
LIMo (in German: ‘Lösungsinstrumente beim Modellieren’) at the University
of Münster. One reason for developing a new strategic instrument was to use a
solution plan in a different mathematical content area, namely geometry. Another
reason was to focus on particular steps of mathematical modelling. For example,
‘Interpreting’ and ‘Checking’ are different steps in the five-step solution plan.
Fig. 12.1 Five-step solution plan of the LIMo project

Those steps are listed separately as students should become aware of the impor-
tance of each subprocess of modelling. Especially those extra-mathematical
steps like ‘Understanding and Simplifying’ and ‘Checking’ should be empha-
sised as students often do not have strategies available for those sub-processes
(Blum 2015).
As mentioned before, Blum (2015) states that a lack of metacognitive competen-
cies could lead to problems as students often do not have strategies available for
solving real-world problems. To avoid that problem, different strategies are pro-
vided in each step of the solution plan, either at a cognitive or metacognitive level.
In the first step of understanding and simplifying one of the given strategies – ‘Make
assumptions that are necessary for solving the task’ – serves as a metacognitive
strategy, whereas ‘Make a sketch’ in the step of mathematising can be seen as a
cognitive organisational strategy for setting up a mathematical model. A detailed
stepwise introduction of this plan is essential so that students know what is meant
by each step.
12.3 Research Questions
As we decided to develop a solution plan consisting of five steps to stress the impor-
tance of each step, we are especially interested in the effects of this solution plan on
a student’s sub-competencies of mathematical modelling that is in simplifying,
mathematising, interpreting, and validating. Regarding the definition of Maaß
(2006; see section 1), we left out working mathematically, as this intra-mathematical
sub-competency can be seen as not characteristic of modelling. Therefore, the
research questions of the study presented are the following:
1. Does the use of the five-step solution plan have an impact on students’ compe-
tencies in simplifying, mathematising, interpreting, and validating?
2. Is there a higher development of those sub-competencies for students who work
on modelling tasks with the solution plan than for students who model in a ‘clas-
sic way’?
3. Is the development of competencies sustainable?
Due to previous studies that showed positive results for the use of strategic instru-
ments, we assume that modelling with the five-step solution plan leads to a higher
development of competencies than in a group of students solving modelling prob-
lems without using the solution plan. Regarding the solution plan as a scaffolding
instrument, we furthermore assume that the development of competencies is more
sustainable when modelling with the strategic instrument as the solution plan aims
at developing long-term competencies.
134 C. Beckschulte
12.4 Methods
12.4.1 Sample and Design of the Study
The study was conducted in 29 grade 9 classes from 16 higher achievement schools
(the German ‘Gymnasium’) in North Rhine-Westphalia, so that 742 students at the
age of 14–15 years took part. In order to evaluate the use of a solution plan, we used
a control-group design with two different groups. As far as possible, the 29 classes
were randomly assigned to an experimental group (EG: 14 classes), which means
that these students worked with the five-step solution plan and a control group (CG:
15 classes) without the presence of this instrument. In some cases, the schools
decided in which of the groups they want to participate. Because of this restriction,
a convenience sample and not a randomised control trial was used.
The study took place during the regular mathematics lessons and consists of a
four lesson teaching unit on modelling activities in the field of geometry with each
lesson taking 45 minutes. A modelling test at three points of measurement – a pre-
test before the teaching unit, a post-test afterwards, and follow-up test about
3 months after the post-test – was used to assess the development of modelling sub-
competencies (see Sect. 12.4.3). All lessons including the testing lessons followed a
detailed plan and were given by teachers who were trained before including an
explanation of the strict procedure of the study.
12.4.2 Treatment
The students worked independently in pairs on four different modelling tasks during
the teaching unit, one in each lesson of 45 min. Progressing in difficulty and varying
the mathematical content, the first two lessons included modelling tasks in the field
of area and scaling calculations, whereas optimum location problems were the
mathematical key topic of the last two lessons. In the first lesson, a guided step-wise
task was used to initiate modelling activities. As the steps of this task were related
to the solution plan in the experimental group, the plan was introduced to those
students at the end of the first lesson. Exemplified by the task, each step was dis-
cussed and explained in class. The modelling tasks were based on existing, already
employed tasks. Despite this, pilot studies in which students worked on the tasks
were observed. Furthermore, the whole intervention was piloted within a class.
During the other three lessons, the solution plan was provided on a sheet for each
student, as well as on a poster in the classroom. To alert students to the solution
plan, students were asked to solve each modelling task on an additional ‘working
sheet’ that was structured regarding the five steps of the plan. Thus, it was possible
to discuss the task at the end of each lesson with a focus on each step. The control
group worked on the same tasks including the step-wise introductory task but with-
out a solution plan and a structured sheet. The Tower Task shows an exemplary
modelling problem students worked on during the second lesson.
Tower Task
In Cologne you find the so-called Cologne Triangle, a tower with a very spe-
cial form. It is 103.20 metres high with 29 floors and a viewing platform at the
roof top. Besides floors with conference rooms and restaurants, there are
some floors which business people can rent for offices. In the picture on the
right, you can see such an office floor as well as evaluators, stairwell, and
toilets.
How many square metres is such a floor?
To estimate the area and thus to build up a mathematical model, the given situa-
tion has to be simplified. Students need to distinguish between relevant and irrele-
vant information. Relevant for solving the task would be, for example, that the area
of the floor is to be determined as well as the given scale. A possible assumption is
that the inner circle including stairs and toilets does not belong to the renting floor.
Different geometrical forms and corresponding formulae for calculating the areas
could serve as mathematical models. The mathematical results have to be inter-
preted and validated with regard to the real situation, for example, by comparing the
area with the size of a football field or classroom.
12.4.3 Assessment and Evaluation
In order to assess different sub-competencies, we developed a new test instrument

with the help of item response theory (see Hankeln et al. 2018). Considering the
conceptualisation of Maaß (2006), items focussing on each subcompetency (i.e.
simplifying, mathematising, interpreting, and validating) were constructed. Thus, a
four-dimensional Rasch model was used regarding each sub-competency as one
latent dimension with reliabilities lying between .60 and .62. Exemplary test items
for each measured sub-competency can be found in Hankeln et al. (2018).
The tests were in a written format following a rotating design with two versions
for each point of measurement. We used a total of 32 items in the test, 16 items in a
test booklet at each point of measurement. Item response theory was chosen as an
adequate method because that way it was possible to link the different test booklets
within and between different points of measurement by using anchor items.
Furthermore, it led to parameters indicating a person’s ability at the different points
136 C. Beckschulte
of measurement. For estimating the person parameters, we used the method of

weighted likelihood estimates (WLE) serving as a basis for the statistical analysis.
All test items were either in a multiple choice or an open answer format. We coded
40% of the tests with two raters reaching a satisfying interrater reliability Cohen’s
Kappa (Cohen, 1960) for open answer items of .81 < κ < .95. Nine out of 32 items
were coded as either correct or incorrect and thus scored with 1 or 0 points, whereas
in 23 items up to 2 points could be achieved.
12.5 Results
The development of competencies based on the students’ WLE parameters, called

‘students’ performance’ in this sub-section, was evaluated using analyses of vari-
ance (ANOVA) with repeated measures. We used an alpha level of 0.05 for all sta-
tistical analyses as well as partial η2 as an effect size. Table 12.1 shows means and
standard deviation of students’ performance at all points of measurement for each
of the measured sub-competencies. Comparing scores in the pretest, the mean score
of the control group is higher than the experimental group in every dimension.
However, this is not statistically significant, as t tests show. In order to analyse the
impact of modelling with a solution plan on students’ competencies, a two-way
ANOVA was carried out.
From the analysis, we found that the use of the solution plan influences the devel-
opment of competencies in interpreting (F(2) = 3.880, p = 0.022, η2 = 0.01) with
small effect size, whereas there is no interaction effect between group and point of
measurement for the other sub-competencies. As post hoc analyses using Bonferroni
correction show, students of both groups achieved significantly higher results in the
post-test. However, this development is sustainable only for the experimental group,
as students’ performance in interpreting decreases significantly from post- to fol-
low-up test for the control group. Regarding the sub-competency of mathematising,
there are no effects at all. Students of the control group increase their performance
Table 12.1 Mean (standard deviation) of students’ performance in pre-, post-, and follow-up test
Subcompetency Group Pre-test Post-test Follow-up test
Simplifying Experimental group 1.00 (0.70) 1.08 (0.68) 1.11 (0.71)
Control group 1.10 (0.74) 1.23 (0.71) 1.18 (0.79)
Mathematising Experimental group 0.39 (0.78) 0.39 (0.75) 0.41 (0.81)
Control group 0.49 (0.72) 0.52 (0.67) 0.47 (0.72)
Interpreting Experimental group 0.32 (0.79) 0.58 (0.90) 0.62 (0.83)
Control group 0.36 (0.77) 0.75 (0.87) 0.58 (0.78)
Validating Experimental group 0.03 (0.76) 0.23 (0.84) 0.13 (0.87)
Control group 0.08 (0.76) 0.33 (0.81) 0.25 (0.89)
in simplifying, whereas there is no significant difference for the experimental group

between pre- and post-test results. However, there is a statistical trend regarding the
development from pre- to follow-up test for students who worked with the solution
plan as the increase of students’ performance is marginally significant (p = 0.07,
η2 = 0.01). Students of both groups increased their scores significantly within the
dimension of validating. However, we found no significant difference for the devel-
opment of this sub-competency between the groups.
12.6 Summary and Discussion
In order to summarise the findings, we can relate the statistical results to our research
questions. Referring to the first question, we have statistical evidence that the pre-
sented five-step solution plan has an impact on the development of students’ com-
petencies in interpreting mathematical results in real situations. This replicates the
results of Schukajlow et al. (2015). However, the results show that the use of the
solution plan did not influence the development of competencies in simplifying,
mathematising, and validating. A reason for this could be the short intervention of
only four lessons as a limitation of the study. However, if we have a deeper look at
the findings of the post hoc analysis and thus considering the second research ques-
tion, we can see differences in the development for each group regarding competen-
cies in simplifying and validating. We found a higher short-term improvement for
students of the control group compared to students of the experimental group for
simplifying, interpreting, and validating. It might be possible that, on the one hand,
modelling with a five-step solution plan is too complex for students as they might
not be able to differentiate between the steps, so that a qualitative analysis of stu-
dents’ solutions seems to be necessary. On the other hand, the combination of work-
ing sheet and solution plan could impose a restriction on students’ modelling
activities. However, if we consider the third research question, we can see a statisti-
cal trend of a more sustainable development for students who worked with the solu-
tion plan for the sub-competencies simplifying and interpreting. Thus, we assume a
tendency that this strategic instrument could lead to developing long-term compe-
tencies. This is reminiscent of the solution plan as a scaffolding instrument, as
described by Schukajlow et al. (2015), that intends bringing students gradually to a
state of competency and thus becoming superfluous as soon as students know about
the strategies presented in such a plan.
For future perspectives, there should be further qualitative analyses regarding the
use of this solution plan on the one hand. On the other hand, it seems to be necessary
to evaluate if such a strategic instrument has an impact on developing students’ mod-
elling competencies in other content areas than geometry, as it does not give content-
related advice but instead refers to general modelling methods (see Sect. 12.2).
138 C. Beckschulte
References
Blomhøj, M., & Kjeldsen, T. H. (2006). Teaching mathematical modelling through project work.
ZDM Mathematics Education, 38(2), 385–395.
of mathematical modelling (pp. 15–30). Dordrecht: Springer.
do? In G. A. Stillman, W. Blum, & M. S. Biembengut (Eds.), Mathematical modelling in edu-
cation research and practice (pp. 73–96). Cham: Springer.
Cohen, J. (1960). A coefficient of agreement for nominal scales. Educational and Psychological
Measurement, 20(1), 37–46.
Galbraith, P., & Stillman, G. (2006). A framework for identifying student blockages during transi-
tions in the modelling process. ZDM Mathematics Education, 38(2), 143–162.
Greefrath, G. (2015). Problem solving methods for mathematical modelling. In G. A. Stillman,
practice (pp. 173–183). Cham: Springer.
Greefrath, G., & Vorhölter, K. (2016). Teaching and learning mathematical modelling. Approaches
and developments from German speaking countries (ICME-13 topical surveys). Cham:
Springer.
Hankeln, C., Adamek, C., & Greefrath, G. (2018). Assessing sub-competencies of mathematical
modelling with the help of item response theory – Presentation of a new test instrument. In
G. A. Stillman & J. P. Brown (Eds.), Lines of inquiry in mathematical modelling research in
education (ICME-13 monograph series). Cham: Springer.
Kaiser, G. (2007). Modelling and modelling competencies in school. In C. R. Haines, P. Galbraith,
Maaß, K. (2004). Mathematisches Modellieren im Unterricht. Ergebnisse einer empirischen
Studie. Hildesheim: Franzbecker Verlag.
113–142.
Niss, M. (2004). Mathematical competencies and the learning of mathematics: The Danish KOM
project. In A. Gagtsis & S. Papastavridis (Eds.), 3rd Mediterranean conference on mathemati-
cal education, January 2003 (pp. 3–5). Athens: The Hellenic Mathematical Society.
Smit, J., van Eerde, H. A. A., & Bakker, A. (2013). A conceptualisation of whole-class scaffolding.
British Educational Research Journal, 39(5), 817–834.
Stillman, G. (2011). Applying metacognitive knowledge and strategies in applications and mod-
elling tasks at secondary school. In G. Kaiser, W. Blum, R. Borromeo Ferri, & G. Stillman
(Eds.), Trends in teaching and learning of mathematical modelling (pp. 165–180). Dordrecht:
Springer.
Stillman, G., Brown, J., & Galbraith, P. (2010). Identifying challenges within transition phases
of mathematical modeling activities at year 9. In R. Lesh, P. L. Galbraith, C. R. Haines, &
A. Hurford (Eds.), Modeling students’ mathematical modeling competencies (pp. 385–398).
New York: Springer.
Zech, F. (1998). Grundkurs Mathematikdidaktik. Weinheim/Basel: Beltz.
Zöttl, L., Ufer, S., & Reiss, K. (2010). Modelling with heuristic worked examples in the KOMMA
learning environment. Journal für Mathematikdidaktik, 31(1), 143–165.
Chapter 13
A Case Study of Tensions and Challenges
Arising as a Swedish Upper Secondary
Teacher Designs and Implements a Model
Development Sequence on Statistics
Jonas Bergman Ärlebäck
Abstract A study of an upper secondary teacher trying to transform her teaching

by designing and implementing a model development sequence on statistics in a
grade 11 mathematics class in Sweden is presented. Cultural-historical activity the-
ory (CHAT) is used to analyse lesson plans, reflection notes, recordings from plan-
ning meetings, and student data to identify tensions and challenges arising as she
tried to adopt this new perspective on teaching and learning mathematics. The
results illustrate the need to re-negotiate appropriate classroom norms and facilitate
shifts in power, ownership, and responsibilities in the mathematics classroom if a
modelling approach is to effectively impact and change the teaching and learning of
mathematics using mathematical modelling.
Keywords Cultural-historical activity theory · Models and modelling perspective

· Model development sequences · Statistics · Teaching · Upper secondary level ·
Upper secondary teacher
13.1 Introduction
Mathematical modelling has become more important in mathematics curriculum

documents and standards at all grade levels worldwide (Blum et al. 2007). However,
as noted by Blum (2015), the impact and degree of implementation of mathematical
modelling in everyday classrooms have been low. Reasons and arguments for why
J. B. Ärlebäck (*)
e-mail: jonas.bergman.arleback@liu.se

140 J. B. Ärlebäck
this is the case have been discussed in terms of different obstacles (Blum and Niss
1991) or barriers (Burkhardt 2006) manifested at different institutional levels.
From general educational research, we know that among the factors that influ-
ence students’ learning, teachers have been argued to be one, if not the most, impor-
tant factor facilitating students attaining the learning goals specified in curricula
(Hattie 2009; Timperley et al. 2007). This is particularly true for the teaching and
learning of mathematical modelling, and the challenges in terms of obstacles and
barriers teachers need to manage and overcome are numerous (Blum 2015). Blum
and Niss (1991) separately discuss obstacles for teachers and for learners and
instructional obstacles, whereas Burkhardt (2006) focuses on obstacles collectively
as habits, beliefs, and teaching skills as part of a systemic inertial barrier.
Some of the challenges discussed in the literature are: modelling and applica-
tions make the mathematics classroom more complex, less predictable, and much
more demanding; tensions between in-subject priorities regarding basic skills ver-
sus problem solving or pure versus applied mathematics; modelling requires more
from the teachers than just pure mathematical knowledge, and many teachers feel
uneasy and unable to deal with problems originating from subjects and disciplines
outside their field of expertise; the question is if modelling, applications, and con-
nections to other subjects belong to mathematics instruction at all; that assessing
students’ progress and achievements engaged in modelling is complex and difficult
and a skillset that most teachers do not have; and time is an issue – there is no time
or space to include applications and modelling in an already overstuffed curriculum
(Blum 2015; Blum and Niss 1991; Burkhardt 2006; Kaiser and Maaß 2007; Maaß
2005). This chapter describes how some of these obstacles and barriers surfaced and
were dealt with by a teacher implementing a modelling approach in her classroom.
The two research questions addressed in the chapter are: What does a teacher’s
attempt to design and implement a modelling perspective in her teaching look like?
and What are the tensions and challenges that arise when a teacher tries to adopt a
modelling perspective in her teaching?
13.2 Theoretical Background
The diversity of perspectives on mathematical modelling in mathematics education

research is well documented (Kaiser and Sriraman 2006). Since different perspec-
tives are grounded in disparate cultural settings and traditions, it is crucial to be as
explicit as possible about basic notions used such as mathematical model and math-
ematical modelling in each particular research setting. Hence in this section we
begin with clarifying the perspective on modelling adopted in the work presented in
the chapter. This is followed by a discussion of aspects of cultural-historical activity
theory (CHAT), the analytical lens used in analysing the data collected in the case
study reported.
13 A Case Study of Tensions and Challenges Arising as a Swedish Upper Secondary… 141
13.2.1 The Models and Modelling Perspective
The models and modelling perspective originates from an adaptation and blend of
Piagetian and Vygotskyan ideas as well as influences from Americans pragmatists
such as Dewey, Mead, and Sanders (Lesh and Doerr 2003). From this perspective,
models are conceptual systems, inherently social in nature, that are used to make
sense of situations and phenomena in a dynamic and changing world (Mousoulides
et al. 2007). In this chapter, the notion that models are externally representable con-
ceptual systems (consisting of objects, operations, relations, and interaction-
governing rules) used to describe, explain, predict, or understand some other system
(Lesh and Doerr 2003) is adopted. Further, a mathematical model focuses on the
structural characteristics of the model in question. Models develop through being
modified, extended, and revised through “multiple cycles of interpretations, descrip-
tions, conjectures, explanations and justifications that are iteratively refined and
reconstructed by the learner” (Doerr and English 2003, p. 112).
13.2.1.1 Model Development Sequences
Model development sequences (Lesh et al. 2003) offer a systematic way to design
and implement teaching and learning experiences to support students’ learning
across multiple tasks. A model development sequence consists of an initial model
eliciting activity (MEA) followed by model exploration activities (MXAs) and
model application activities (MAAs). Aimed at eliciting the students’ initial ideas
about a meaningful situation, the MEA is designed to engage students in an iterative
and cyclical process of expressing, testing, and revising their ways of thinking about
the relevant quantities and relationships involved. MXAs, on the other hand, have
students thinking about the underlying mathematical structure of their elicited
model by exploring, interpreting, and using different representations purposefully
and productively. Lastly, MAAs provide new and structurally similar contexts in
which the students have to think with their models, often resulting in further devel-
opments and adaptations of their model. It is through interacting with other students
and the teacher that students refine and develop their models by engaging in multi-
ple cycles of descriptions, interpretations, conjectures, and explanations (Lesh
et al. 2003).
13.2.2 ultural-Historical Activity Theory and the Structure

C
of Human Activity
Activity theory, or cultural-historical activity theory, grew out of the post 1917 rev-
olution of Soviet psychology founded by Vygotsky, Leont’ev, and Rubinstein
(Wertsch 1981) and can be considered as a set of basic ideas and principles, which
together form a conceptual system of a more general nature than what is normally
considered “a theory” (Kaptelinin 1996). Roth and Lee (2007) describe activity as
“an evolving, complex structure of mediated and collective human agency” (p. 198),
and Engeström (1987) brings together the Vygotskyan idea of cultural mediation of
actions, Leont’ev’s notions of activity – action – operation, and Marx’s ideas about
production, distribution, exchange, and consumption as central aspects of human
activity, to formulate the model of the structure of human activity, an activity system.
Engeström’s (1987) activity system consists of six interconnected nodes (sub-
ject, object, instrument, community, rules, division of labour) organised in a three-
level triangular shape (see Fig. 13.1 for an illustration of this framework applied to
the research setting in this study). Briefly explained, subject either refers to an indi-
vidual or a group depending on whose agency is in focus. The object is what the
activity is directed towards and on which the subject(s) uses instruments, either
internal or external as well as physical or symbolic (semiotic) mediating artefacts or
instruments (tools, language, and signs), to transform, shape, or design an outcome
of the activity. Community stresses that the activity partly also is mediated by the
communities carrying out the activity. The rules account for all norms, standards,
procedures, regulations, and conventions accepted within a community or activity.
Finally, division of labour highlights two dimensions of the dynamics of an activity;
it accounts for the actual division of labour and responsibility for achieving the
goals of the actions of the activity, and it accounts for status and power relations
within the communities.
In Engeström’s (1987) account, learning is driven by inner contradictions within
the activity system since “[w]hen inner contradictions are conscious, they become
the primary driving forces that bring about change and development within and
Fig. 13.1 Mathematics class activity system with modelling perspective and model development
sequence introduced as new instruments to transform the curriculum on statistics into new ways of
teaching and learning
between activity systems” (Roth and Lee 2007, p. 203). According to Engeström
(1987), there are four types of contradictions appearing at different levels: primary
contradictions within nodes in the activity system, secondary contradictions
between different nodes in the activity system, tertiary contradictions appearing
between the activity system and the wish to introduce new elements in, or trans-
form, the activity system, and quaternary contradictions between activity systems.
The last will not concern us here.
13.3 Method and Setting
The case study reported in this chapter comes from a 1-year project aiming at
improving 13- to 18-year-old (grades 7–12) students’ attitudes and results in math-
ematics in two Swedish municipalities. A researcher worked together with two
teachers from different schools and grades as partners (c.f. Jaworski 1999) in a co-
learning agreement (Wagner 1997). The point of departure in the project was the
possibilities and challenges the participating teachers saw in their everyday teaching
practices, leading to a formulation, planning, and implementation of a 1-year long
project with specified aims and goals. Both teachers described their teaching as
traditional centred around the use of a textbook series, teacher-led content reviews,
and students working on tasks in their textbooks. This chapter primarily reports on
the second model development sequence designed and implemented by the upper
secondary teacher in the project.
The aim of the project was to try out and investigate alternative approaches to
teaching and ways of working in the classroom that support students’ learning and
increase their engagement and interest to learn mathematics. The project drew on
the models and modelling perspective and adopted a design-based research-inspired
methodology, both introduced by the researcher, and implemented two iterations of
(1) designing teaching focusing on model development sequences, (2) implement-
ing these sequences, (3) evaluating students’ learning and experiences, and (4) giv-
ing recommendations for how to improve the sequences.
In the project initiation phase, the teachers were introduced to, worked with, and
read about (a) modelling, applications, and problem-solving generally in a series of
non-technical Swedish texts by the chapter author and (b) the models and modelling
perspective (the first chapter in Lesh and Doerr (2003) and Mousoulides et al.
(2007)), model development sequences (Lesh et al. 2003), as well as various exam-
ples of modelling tasks from the literature. In the second phase, the teachers worked
on designing two model development sequences each on different content matter.
This work was primarily done by the individual teachers, and the researcher’s role
was more passive and functioned as a resource and discussion partner who occa-
sionally contributed with alternative perspectives and ideas.
Some of the negotiated and general questions guiding the teachers’ work as a
whole and functioning as a basis for reflection were: What challenges and possibili-
ties are there with working with model development sequences in mathematics at
the upper secondary level? What are the experienced differences and similarities
between teaching based on modelling development sequences and more “tradi-
tional” ways of teaching? How do the students experience these new practices and
do they learn the prescribed content? Sub-questions guiding each implementation
were: How is the model development sequence construed? How does one focus the
activities towards a given learning goal? What mathematical knowledge do the
activities “generate” and how? What is easy and what is challenging for the stu-
dents? What do the students learn through this way of working?
13.3.1 Data Collection and Analysis
Data were collected in the form of notes and audio from planning meetings; the
teacher’s documentation of designing the model developing sequence, lessons
plans, activities, and reflection notes; students’ work on the activities (mostly writ-
ten work but also some video and audio); and students’ evaluations. The video and
audio recordings were watched and listened to multiple times and interesting epi-
sodes time-stamped and transcribed focusing on instances when aspects related to
the design of the model development sequence came to the fore, or when the teacher
or the students expressed opinions regarding the implementation of the sequence.
For the analysis, Engeström’s (1987) model of human activity was used since it
offers a framework to map the challenges and tensions that arose in the teacher’s
classroom in terms of contradictions in contrast to what her traditional classroom
normally looks like as a consequence of her attempt to use a modelling perspective.
Hence, the data were analysed in terms of a mathematics class activity system shown
in Fig. 13.1 by identifying expressed and realised contradictions between, and in,
nodes in this system induced by the model development sequence as a new teach-
ing tool.
13.4 Results
As well as illustrating the model development sequence the teacher developed,

implemented, and evaluated within the project, this section also presents the result
of the analysis in terms of primary and secondary contradictions manifested in the
mathematics class activity system (e.g. see Fig. 13.1) that surfaced during the teach-
er’s efforts to transform her teaching on statistics using the model development
sequence (i.e. introducing a tertiary contradiction).
13.4.1 Teacher’s Modelling Development Sequence
The second model development sequence the upper secondary teacher designed
focused on the content area of statistics and was centred on three modelling tasks
and spanned five lessons. The first task, The Basketball Penalty Competition, was a
MEA aimed at eliciting students’ spontaneous and prior understanding of measures
of spread. The second task, Pairing Up, was a MXA facilitating students’ under-
standing, coordination, and making connections between the notions and concepts
of median, mean, mode, variation width, bar charts, standard deviation, and box
plots. The third task, Free Diving, was a MAA where students applied regression,
knowledge about normal distributions, statistical thinking, and the content they pre-
viously learnt in the sequence. The design was grounded in the learning objectives
as described as core content in the Swedish national curriculum:
Teaching in the course should cover the following core content: …
• Statistical methods for reporting observations and data from surveys, including
regression analysis.
• Orientation and discussion of correlation and causality.
• Methods for calculating different measures of central tendency and measures of dis-
persion including standard deviation.
• Properties of normally distributed material. (Skolverket 2011, p. 20)
According to the teacher, statistics as content within the teaching of mathematics at
the Swedish upper secondary level is normally not prioritised and often delayed
until the end of the semester and done only if there is time. In the case reported here,
the planning for Mathematics course 2B and the surrounding circumstances meant
that five lessons were spent on statistics.
Figure 13.2 shows how the activities in the model development sequence were
distributed over the five lessons as well as what statistical content surfaced and was
Fig. 13.2 The sequence spanning five lessons and the content surfacing, introduced, and treated
treated. The concepts in light grey boxes arose naturally during the activities and
were brought up by the students of their own accord in whole class discussions. The
concepts in dark grey boxes also came from the students, however only through
teacher scaffolding. The concepts in black boxes were introduced externally to the
students, either by presentations made by the teacher or the students’ reading and
working with the textbook.
13.4.1.1 The Basketball Penalty Competition: The MEA
In the MEA, The Basketball Penalty Competition (see Fig. 13.3a), the students were
divided into three teams and asked to each do 10 basketball penalty shots from 3
different distances. The teams collected data from all their members, and these data
were then shared among all the students. The question the students were asked to
answer was “Which team is the best?” The students used bar charts, the total fre-
quency, mean, and mode to argue for the team they considered the best. Supported
by the teachers, the students were also able to use scatterplots and range to describe
the data. In the second lesson, the teacher introduced boxplots and standard devia-
tion and had the students apply these to their basketball data.
When planning the MEA, the teacher expressed anxiety about her design and
insecurity in how to go about implementing the activity; she wrote: “Panic for what
to do… How do I get an activity that’ll show what they [the students] know without
making the activity too easy? How do I spark good discussions?” This suggests the
tertiary contradiction (i.e. the modelling perspective) being manifested as poten-
tially secondary contradictions involving the rules (how mathematics normally is
practiced in the teacher’s classroom), the model development sequence as an instru-
ment for organising teaching, and division of labour in terms of modified roles and
Fig. 13.3 (a) The Basketball Penalty Competition activity. (b) The Pairing Up activity
responsibilities in the mathematics class activity system. After the activity, the
teacher was relieved: “The [first] lesson went better than I expected! The students
enjoyed playing basketball and both when doing this as well [as] back in the class-
room had good and lively discussion about what/when/why one has to organize data
of this type”.
Students’ reaction to the MEA was positive overall. In the evaluation, students
wrote “Rich and fun – nice to do something outside the regular classroom”; “Fun
task and a good start-up on statistics – more practical activities please!”; and
“Confusing at first, but rewarding when you started to catch on [to] the purpose with
the whole thing – playing basketball was fun!” However, for some students the
activity did not do much as indicated by the response “Fun to do something outside,
but I didn’t learn that much”.
13.4.1.2 The Pairing Up Activity: The MXA
For the MXA in the sequence, the teacher had designed sets of matching cards to be
paired up in groups of four (see Fig. 13.3b). When the teacher reflected on the
MXA, she wrote that “The activity offered opportunities to work and coordinate the
content really well” but that she and the students unfortunately “didn’t have time to
discuss [the activity] in detail in class”. In her opinion, “the students seemed to
really appreciate this activity” and “the [students’ group] discussions were lively
and the level of engagement high”. The teacher also stressed that many misconcep-
tions and difficulties related to standard deviation and boxplots surfaced during the
activity, which she attributed to the structure and organisation of the MXA. Overall
however, the teacher was more comfortable with the MXA than the MEA and found
it productive, rewarding, and enlightening for both the students and herself.
The students’ opinions about the MXA were ambiguous, and some of the stu-
dents expressed that they did not understand the purpose of the activity nor appreci-
ated it: “A bit unserious – a boring way to learn math”. Others expressed that the
timing to do the activity might have been ill chosen as, for example, the student who
wrote: “The activity felt kind of hard. It would have been better if we’d done it later
when we had the basics more nailed down”. However, many of the students’ evalu-
ations concurred with the teachers’ conclusion that her students liked the activity:
“You learned a lot from the activity – it was hard, but that was perhaps a good
thing?” and “Good, but stressful time-wise”. The teacher’s reflection and students’
comments illustrate how they gradually started to cope with and tackle the chal-
lenges in the evolving mathematics class activity system, as well as express the
general secondary contradiction present in all classrooms arising due to time
constraints.
13.4.1.3 Free Diving: The MAA
The model application activity the teacher designed was organised around the con-
text of free diving and was introduced using a YouTube clip. The activity was
divided into four parts with different foci: (a) linear regression, (b) correlation vs.
causation, (c) spread and standard deviation, and (d) interpreting normally distrib-
uted material. Much of this content was handed over to the students as homework
to read up on and learn using their textbook. Together with a shorter whole class
content discussion in lesson 5, the parts (a)–(d) functioned as a guide for the stu-
dents what to focus on. The teacher expressed that she was surprised that the work
the students handed in showed that they took the activity seriously and had spent a
lot of work on it. However, the teacher also acknowledged that the four subtasks in
the MAA “had some flaws that need to be fixed before used again in the future”. She
noted that the MAA “[was] not challenging enough for some students and too hard
for others”. Two other issues raised by the teacher were more closely connected to
the models and modelling perspective and the model development sequence. Firstly,
she noted that “Too much responsibility [was] put on the students to learn the more
difficult concepts by themselves”, indicating a secondary contradiction, centred
around time constraints, between rules and tools, division of labour in the mathe-
matics class activity system. Secondly, the teacher expressed that she felt a discon-
nect between the MEA and MXA and MAA and that the last did not quite build on
the former two to the degree that she would have liked. This can be interpreted as
both a primary contradiction (instrument) and as secondary contradictions between
rules and tools, division of labour in the sense as pointed out above in the section on
the MEA.
Traces of the teachers’ comments can also be found in the students’ evaluation of
the MAA, which generally was positive but showed a diversity of opinions. One of
the students expressed that “This activity really didn’t give that much” whilst
another said that “This was the activity on statistics that gave the most”. One student
wrote that the MAA was “Tricky, but solvable. It’s fun with something more chal-
lenging”, while another student wrote “It was good. The YouTube clip was nicely
used and was fun and made the whole thing a bit more personal – and you got a
good insight as well on free diving and statistics ☺”.
13.5 Conclusion
This chapter provides an account of an upper secondary teacher’s attempt to imple-

ment a modelling perspective in trying to realise the statistics curriculum by design-
ing and implementing a model development sequence. Although CHAT affords a
multitude of conceptualisations of the dynamics of the classroom, the use of
Engeström’s (1987) activity diagram as an analytical lens proved productive to map
out and understand the nature of the obstacles and challenges that surface in the
teacher’s mathematics class as a result of her introducing model development
sequences and a modelling perspective as tools for transforming her teaching. Much
in line with previous research on obstacles and barriers, the analysis points to the
need for re-negotiating appropriate classroom norms of what it means to learn math-
ematics and to facilitate shifting power, ownership, and responsibilities in the math-
ematics classroom if model development sequences are to effectively impact and
change the teaching and learning of mathematics using mathematical modelling.
References
do? In S. J. Cho (Ed.), The proceedings of the 12th international congress on mathematical
education (pp. 73–96). Cham: Springer.
Blum, W., & Niss, M. (1991). Applied mathematical problem solving, modelling, applications, and
links to other subjects: State, trends and issues in mathematics instruction. Educational Studies
in Mathematics, 22(1), 37–68.
Blum, W., Galbraith, P., & Niss, M. (2007). Introduction. In W. Blum, P. Galbraith, H.-W. Henn,
& M. Niss (Eds.), Modelling and applications in mathematics education. The 14th ICMI study
Burkhardt, H. (2006). Modelling in mathematics classrooms: Reflections on past developments
and the future. ZDM Mathematics Education, 38(2), 178–195.
Doerr, H. M., & English, L. D. (2003). A modeling perspective on students’ mathematical reason-
ing about data. Journal for Research in Mathematics Education, 34(2), 110–136.
Engeström, Y. (1987). Learning by expanding: An activity-theoretical approach to developmental
research. Orienta-Konsultit: Helsinki.
Hattie, J. A. C. (2009). Visible learning: A synthesis of over 800 meta-analyses relating to achieve-
ment. London: Routledge.
Jaworski, B. (1999). Mathematics teacher education, research and development. The involvement
of teachers. Journal of Mathematics Teacher Education, 2(2), 117–119.
Kaiser, G., & Maaß, K. (2007). Modelling in lower secondary mathematics classrooms – Problems
and opportunities. In W. Blum, P. Galbraith, H.-W. Henn, & M. Niss (Eds.), Modelling and
applications in mathematics education. The 14th ICMI study (pp. 99–108). New York: Springer.
Kaptelinin, V. (1996). Activity theory: Implications for human computer interaction. In
B. A. Nardi (Ed.), Context and consciousness. Activity theory and human-computer interaction
(pp. 103–116). Cambridge, MA: MIT Press.
Lesh, R. A., & Doerr, H. M. (Eds.). (2003). Beyond constructivism. Mahwah: Erlbaum.
Lesh, R. A., Cramer, K., Doerr, H. M., Post, T., & Zawojewski, J. S. (2003). Model development
sequences. In R. A. Lesh & H. M. Doerr (Eds.), Beyond constructivism (pp. 35–58). Mahwah:
Erlbaum.
Maaß, K. (2005). Barriers and opportunities for the integration of modelling in mathematics classes:
Results of an empirical study. Teaching Mathematics and Its Applications, 24(2–3), 61–74.
Mousoulides, N. G., Sriraman, B., & Christou, C. (2007). From problem solving to modeling – The
emergence of models and modelling perspectives. Nordic Studies in Mathematics Education,
12(1), 23–47.
Roth, W.-M., & Lee, Y.-J. (2007). “Vygotsky’s neglected legacy”: Cultural-historical activity the-
ory. Review of Educational Research, 77(2), 186–232.
Skolverket. (2011). Mathematics. Retrieved 05/30, 2012, from http://www.skolverket.se/
Timperley, H., Wilson, A., Barrar, H., & Fung, I. (2007). Teacher professional learning and devel-
opment. Wellington: Ministry of Education.
Wagner, J. (1997). The unavoidable intervention of educational research: A framework for recon-
sidering researcher-practitioner cooperation. Educational Researcher, 26(7), 13–22.
Wertsch, J. V. (1981). The concept of activity in Soviet psychology. In J. V. Wertsch (Ed.), The
concept of activity in Soviet psychology (pp. 3–36). Armonk: M. E. Sharp.
Chapter 14
Year 6 Students’ Gradual Identification
of Mathematical Models of Average Speed
when Making Sense of ‘Walking’
Takashi Kawakami, Shigekazu Komeda, and Akihiko Saeki
Abstract This chapter illustrates how, and to what extent, students (11–12-year
olds) identified a straight-line graph (time–distance graph) as a mathematical model
of average speed when making sense of walking situations in relation to technology.
The results demonstrated the types of models that students identified explicitly, the
phases of identifying such mathematical models, and essential actions needed for
this identification. The approach of identifying mathematical models, through
sense-making of embodied actions and technologies, can highlight students’ grad-
ual mathematisation differentiating the phenomenal world, virtual technological
world, and mathematical conceptual world.
Keywords Average speed · Embodied action · Identifying mathematical models ·

Mathematisation · Sense-making · Technology · Primary school
14.1 Introduction
The possibilities for mathematical modelling as a didactical means for supporting

student learning of mathematical concepts have been explored (e.g. Swan et al.
2007). When learning mathematics through mathematical modelling, the activities
T. Kawakami (*)
Faculty of Education, Utsunomiya University, Utsunomiya-shi, Japan
e-mail: t-kawakami@cc.utsunomiya-u.ac.jp
S. Komeda
Graduate School of Education, Saga University, Saga-shi, Japan
e-mail: komeda@cc.saga-u.ac.jp
A. Saeki
Graduate School of Education, Naruto University of Education, Naruto-shi, Japan
e-mail: asaeki@naruto-u.ac.jp

152 T. Kawakami et al.
students undertake to make sense of real-world situations, in which they abstract

and identify mathematical objects, are crucial (Doerr and Pratt 2008). Previous
research has pointed out that technology plays an important role in supporting these
activities (e.g. Arzarello et al. 2007; Doerr and Pratt 2008). This study explores
modelling for primary grades as a means of identifying and building mathematical
models of average speed, in association with embodied actions using motion detec-
tors and graphing calculators, as part of the project ‘Modelling a walk’ (e.g.
Kawakami et al. 2015; Komeda et al. 2020). This project aims to use longitudinal
learning and teaching sequences to construct and develop the concepts of average
speed and instantaneous speed in primary and secondary school grades.
14.2.1 Developing a Concept of Average Speed
In developing primary school students’ conceptions of speed, we make a clear dis-

tinction between motion as a phenomenon and speed as a quantity. According to
Piaget (1970), the concept of speed is constructed as a proportional relationship
between the distance moved and the time of movement. In the primary school math-
ematics curricula of many countries, speed is treated as average speed and inter-
preted as a ratio of distance and time. Previous literature (e.g. Thompson and
Thompson 1992) indicates that students have a problem with developing the con-
cept of average speed as a rate of distance and time, because the quantity of average
speed is less visible; therefore, the teaching of average speed tends to focus on cal-
culating the speed. According to Thompson and Thompson (1992), students go
back and forth among several phases, perceiving average speed as (1) distance, (2)
two separate quantities (distance and time), (3) two simultaneous quantities, and (4)
a rate. Some students fail to connect mathematical models of speed appropriately to
real-world situations (e.g. Ärlebäck and Doerr 2015). However, limited research
exists that develops the concept of average speed in association with bodily actions.
14.2.2 romoting Identification of Mathematical Models

P
Through Sense-making of Embodied Actions
and Technologies
Research on teaching modelling has suggested that the activity of modelling bodily
actions, using technologies, including computer simulation, promotes the develop-
ment of difficult mathematical concepts (e.g. Ärlebäck and Doerr 2015; Arzarello
et al. 2007) and offers mathematical ‘microworlds’ for students to devote them-
selves to mathematical investigations (Kaput and Roschelle 1999; Papert 1980).
14 Year 6 Students’ Gradual Identification of Mathematical Models of Average Speed… 153
Such an embodied cognition approach reveals the mechanism of mathematical

thinking, assuming that human beings construct and make sense of mathematical
objects through bodily experiences and conceptual metaphor (Núñez et al. 1999).
Arzarello and his colleagues (2007) adopted the embodied cognition approach in
modelling activities using graphing calculators and sensor kits in order to construct
and develop the meaning for mathematical objects with respect to students’ embod-
ied actions.
In the context of using technology in mathematics learning through modelling,
Doerr and Pratt (2008) elaborated the notion of exploratory modelling. Exploratory
modelling means that students make sense of mathematical models that embody the
intentions, concepts, and structures of an expert, which is inherent in technologies.
The issue, however, remains: how can students be helped to make connections
between situated understanding via technologies and the real-world phenomena that
mathematical models originate from? To address this issue, this study adopts the
embodied cognition approach in the modelling learning environment; it adapts the
exploratory modelling process (Doerr and Pratt 2008, p. 270) and elaborates on the
learning process of identifying mathematical models using technologies through
mathematical sense-making of body motion and the graphical representation shown
via technologies (Nemirovsky et al. 1998), as shown in Fig. 14.1. In the process,
how an individual interacts between the phenomenal world and the virtual techno-
logical world is emphasised. In the phenomenal world, students carry out embodied
actions, such as ‘walking’. This action is specialised to the learning objective of
calculating average speed. In the virtual technological world, time–distance graphs,
in which the meaning of the vertical and horizontal axes is not indicated for stu-
dents, are plotted on the screens of graphing calculators through motion detectors.
Essentially, the graph on the screen is a model of real-world enactment and a model
of the mathematical concept of speed. Students make sense of embodied actions,
Fig. 14.1 Conceptualisation of identifying mathematical models through mathematical sense-

making of situations
and the graph is shown on the screen, mathematically. They translate between the
two worlds by conjecturing, verifying, modifying (Geiger 2011; Saeki et al. 2004),
and controlling (Greefrath 2011). Finally, they identify the variables to determine
average speed—time and distance—and the underlying mathematical models of
average speed, straight-line graphs (time–distance graphs), and the formula of aver-
age speed, speed=distance/time.
14.3 The Study
This chapter addresses the following research question: how, and to what extent, did
students identify a straight-line graph (time–distance graph) as a mathematical
model of average speed, when making sense of walking situations in relation to
technologies? To investigate this question, the authors analysed group activities in a
series of teaching experiments on average speed, using graphing calculators and
motion detectors (Kawakami et al. 2015). Two classes of Year 6 students (11–12-year-
olds: 37 males and 37 females) in a national primary school in Japan took part in
this study. In previous schooling, the students had learnt a ratio of two quantities of
different types, graphs, and algebraic expressions of proportion. This was the first
time they had worked with graphing calculators and motion detectors, as well as
modelling. In both classes, they learned about the concept of speed as a rate, as part
of the Japanese Year 6 mathematics curriculum.
14.3.1 Design
14.3.1.1 Outline of Teaching Experiment
In both classes, the teaching experiment consisted of three lessons (45 min each). In
the first lesson, the students were divided into 12 groups. They were asked to (a)
walk back and forth in front of the detectors so as to draw designated graphs on the
screens of the calculators (Fig. 14.2) and (b) consider the meaning of the vertical
and horizontal axes in the graphs so as to help them realise that speed is determined
by two variables (i.e. distance and time). The four graphs have different features so
that students can be made aware of the relationships between bodily motion and
Fig. 14.2 Four kinds of graphs given to students. (Vertical axis: time; horizontal axis: distance)
speed, the difference of speed, and the proportional relationships between distance
and time. The teacher distributed a stopwatch and a tape measure on request. In the
middle of the lesson, he also distributed overhead projector (OHP) sheets on which
the four graphs were printed. The students made a note of the manner and speed of
walking, followed by brainstorming and summarising ideas using such keywords/
remarks as ‘speed’, ‘fast’, ‘slow’, ‘speed doesn’t change’, and ‘speed changes’. The
second lesson entailed a discussion about the meaning of the vertical and horizontal
axes in the graph, with a view to prompting the students to realise that speed was
determined by the two variables of ‘time’ and ‘distance’. The relationship between
these variables was elicited with a table and was then expressed using equations for
proportional relationships. In the third lesson, the students produced two types of
graphs that were difficult to make sense of by simply looking or acting out by walk-
ing. While thinking about which graphs represented faster walking, the students
derived the idea of rate. Finally, speed was represented using the expression of
‘speed=distance/time’.
14.3.1.2 Post-Lesson Task
In this study, the first lesson was designed as a student-led group activity; the second
and third lessons were designed as a teacher-led classroom activity. In order to
investigate how the model identified by individual students changed due to collec-
tive thinking, this study conducted post-lesson tasks after the first and the third les-
sons and compared the results. As a preliminary analysis, this chapter focuses on the
post-lesson task after the first lesson to investigate to what extent the students identi-
fied mathematical models of average speed by themselves. The post-lesson task was
‘What do Graphs A and B mean?’ The meaning of the vertical and horizontal axes
was not indicated for the students. The students wrote their answer on a sheet of
paper using both mathematical language and everyday language.
The data collection comprised students’ responses to the post-lesson task, tran-
scripts of all focus group discussions and video recordings in the first lesson, for
which students’ activities and calculator screens were coupled with lesson artefacts
and field notes. These data were analysed in two phases in order to investigate the
research question about products and process of students’ identification of mathe-
matical models of average speed. In phase 1, students’ responses to the post-lesson
task were coded and categorised according to students’ perceptions of proportion
inherent in a concept of average speed to specify their models. As the perception
progressed from qualitatively/intuitively to quantitatively/logically (e.g. Thompson
and Thompson 1992), higher codes were assigned. In phase 2, two groups were
chosen out of the original 12 as representative of both a typical successful case (a
group that identified mathematical models) and a typical unsuccessful case (a group
that did not identify mathematical models). Transcripts and videos of these group
activities were analysed to determine the characteristic phases of the identification
process (Fig. 14.1) from the following viewpoints: (a) where students focused on in
the phenomenal world (e.g. motion or speed) and (b) how students interpreted the
graphical representation shown on the screen of the graphing calculator in the vir-
tual technological world (e.g. picture or graph).
14.4 Results
This section illustrates the findings, describing the students’ responses to the post-
lesson task and demonstrating three episodes from the groups’ activities in the
first lesson.
14.4.1 tudent-Identified Mathematical Models

S
of Average Speed
Table 14.1 shows the results of 74 students’ models in the post-task. Overall, 93%
of the students (n = 69) could at least identify mathematical models of average
speed at a visual level, connecting the pace of their walking to the slope of the graph
on the screen. For code 0 in each case, all students were not able to interpret or
misinterpreted constant speed (e.g. increasing speed) and the axes of the graph (e.g.
the vertical axis as speed). Concerning code 1 in each case, all students focused on
the visual aspects of mathematical models. They recognised the difference in the
graph’s slope as being a difference in speed using everyday language (i.e. ‘slowly’
and ‘fast’). For example, one student described Graph A with the following
Table 14.1 Types of students’ models (N = 74)

Frequency
Code Characteristics Example of students’ descriptions (n) (%)
0 Inappropriate focusing Not interpreting/misinterpreting speed and the 5 (7%)
axes of the graph
1 Focusing on visual Recognising the difference in graph slope as the 27 (36%)
aspects difference of speed
2 Focusing on the Mentioning the proportion with vague, rough 40 (54%)
underlying proportion statements such as constant speed, but not
qualitatively/ intuitively explicitly mentioning proportional relation
between time and distance
3 Focusing on the Explicitly mentioning speed as a proportional 2 (3%)
underlying proportion relationship between distance moved and time
quantitatively/ logically of movement
statement: ‘first we walk faster, then we walk slower, because first the graph is
slanted, then it is gentle’. Despite proper recognition of speed, 20 students assigned
code 1 were not able to interpret or misinterpreted the axes of the graph. All students
assigned code 1 grasped how fast they walked in terms of angles of the graph on the
screen, as if they were observing pictures or figures. For code 2 in each case, all
students focused on the underlying proportion qualitatively or intuitively. They
appropriately interpreted the axes of the graph and mentioned the proportion with
vague, rough statements such as ‘proportion’, ‘constant’, and ‘same speed’, but they
did not explicitly mention the proportional relation between distance and time. Only
two students focused on the underlying proportion quantitatively or logically. They
explicitly mentioned speed as a proportional relationship between distance and
time, expressing speed by using a ratio of two quantities of different types, and so
were assigned code 3.
14.4.2 ow Students Identified Mathematical Models

H
of Average Speed
This section focused on two typical and contrasting groups’ episodes: Group A
consisted of three students assigned codes 0 or 1 (Table 14.1); Group B consisted of
four students assigned codes 2 or 3 (Table 14.1).
14.4.2.1 Episode 1: Group A’s Activities: Looking at the Screen Only
In the group activities, students in Group A looked at the screen of the graphing
calculators only; they did not observe the walking action. Then, when the graph was
plotted on the screen, they indicated: ‘The slope is slanting. Walking faster’. They
fitted only a plotted graph into the OHP sheet, conceiving the graph to be a picture
or figure without directing their attention to quantity. Throughout the group activi-
ties, they did not make quantitative use of the graphs being plotted on the screen nor
did they move to change the graphs as needed (e.g. measuring distance and time). In
the post-lesson task, all students in Group A misinterpreted the vertical axis of the
graph as speed and horizontal axis as distance.
14.4.2.2 pisode 2: Group B’s Activities: Measuring, Verifying,

E
and Controlling Iteratively
Group B consisted of four students (Yuri, Tomo, Yoshi, and Masa). In the post-
lesson task, Yuri produced a code 3 response, focusing on proportion logically. The
other students produced code 2 responses, focusing intuitively on proportion. In the
group activities, they drew Graph C correctly. Yoshi and Masa marked the tape
Fig. 14.3 (a)–(f) Students measuring, verifying, and controlling iteratively
measure with the stop position (Fig. 14.3a). They conjectured that the position was
the result of movement up to that time. In this way, they started to become aware of
quantity. Then, Masa walked, and the other students said ‘Stop’ at the time they felt
he should stop, observing the plotted graph on the screen in order to verify the posi-
tion (Fig. 14.3b). Yoshi measured and marked the position, pointing out ‘300 cm’,
and they associated the position with the scale of the graph (Fig. 14.3c). Explicitly,
they began to be conscious of the distance. Furthermore, when Masa stopped at the
marked position and they confirmed that the scale of the vertical axis meant ‘dis-
tance’, they conjectured that the scale of the horizontal axis meant ‘time’, by observ-
ing the graph (Fig. 14.3d). Then, they tried to draw Graph A. Yoshi tried to count his
walking time with a stopwatch to control the motion. Thus, they began to be con-
scious of the time. Masa walked, and Yoshi said, ‘One, second, now slow!’, count-
ing time (Fig. 14.3e). As a result, Yoshi was sure that ‘Walking 300 cm in 2 seconds!’
up to the marked position was Graph A, expressing ‘2’ with his hand (Fig. 14.3f).
They began to be conscious of both distance and time. They recorded the following
descriptions for Graph A: ‘[first] 2 seconds in 3 m’, ‘[then] 7 seconds in 2 m’, and
‘walking 3 m in 2 seconds, and stopping. Equal speed as Graph A’ on their work-
sheet. They had conceived a graph with a scale.
14.4.2.3 pisode 3: Group B’s Activities: Willingness to Express

E
Enactment with Mathematics
Through the group activities, Yuri in Group B mathematised the action of walking
using her prior mathematical knowledge of ratios of two quantities of different
types as follows: ‘[About Graph D] first walking 100 cm in 5 seconds, that is, walk-
ing 20 cm in a second, then walking 500 cm in 2 seconds; that is, walking 250 cm
in a second’. She recognised speed as a rate. After the group activities, Tomo
described her impressions about the first lesson: ‘It was very fascinating today. I
want to express a sense in something certain or numbers. I want to explore how
much speed it is’, although she did not demonstrate prior mathematical knowledge
of ratios of two quantities of different types in her worksheet. It is assumed that Yuri
and Tomo had the willingness to express enactment with mathematics. In the group
activities, Tomo explained the change of speed necessary to draw Graph D to Yuri
by gestures expressing the different slope of the graph and by movements express-
ing the different pace of walking. Based on these unsophisticated expressions, they
were conscious of the difference between qualitative enactment and quantitative
mathematics.
This chapter has addressed some aspects of students’ identification of mathematical

models of average speed when making sense of walking action and technologies.
More than 90% of the students clearly identified such mathematical models at a
visual level (see Table 14.1). Their recognition of ‘slope’ did not involve mathemat-
ical concepts of tangents, rather they used everyday language, but it will contribute
to a fundamental understanding of linear functions and graphs depicting instanta-
neous speed in a time–distance setting at higher grades (Komeda et al. 2020). On the
other hand, it appeared that it was difficult for the students to identify mathematical
models independently at a structural level, recognising average speed as a rate (see
Table 14.1), although they had previously learned about a ratio of two quantities of
different types. This result confirms the difficulty of learning the construct of speed
as a rate (e.g. Thompson and Thompson 1992), indicating that more research is
needed on the role of teachers in developing students’ conception of speed.
Three episodes from two contrasting groups’ activities demonstrated three
phases of identifying mathematical models of average speed when making sense of
walking action and technologies (Fig. 14.1): (1) separating or mixing the phenom-
enal world and the virtual technological world, (2) linking the phenomenal world
with the virtual technological world, and (3) distinguishing between the phenome-
nal world and the virtual technological world and switching to the mathematical
conceptual world. In the first phase, students focus on bodily motion only or ignore
bodily motion in the phenomenal world. In the virtual technological world, at most,
they interpret the graph on a screen as only a picture or figure. In Episode 1, Group
A students, fitting the plotted graph into an OHP sheet on the screen, played the
‘game’ only in the virtual technological world. We conjecture that they separated or
mixed real-world enactment and virtual objects (cf. Nemirovsky et al. 1998). This
resulted in their being unaware of quantity and did not lead to an elimination of their
misunderstanding of the meaning of the axes of the graph. In the second phase,
students focused on both bodily motion and speed in the phenomenal world. In the
virtual technological world, they interpreted the graph on the screen as a graph with
a scale. In Episode 2, speaking only in the phenomenal world, Group B students
were first made conscious of bodily motion and speed as distance (Fig. 14.3c).
Then, they were also made aware of speed as time (Fig. 14.3e). Last, they
focused on speed as two simultaneous quantities (Thompson and Thompson 1992).

The students were consistently aware of their bodily movement. In the virtual tech-
nological world, they found the meaning of the axes of the graph (Fig. 14.3f). This
resulted from the iterative translation between real-world enactment and virtual
objects, leading to their consciousness of the underlying proportion relationship. In
the third phase, students distinguished real-world enactment and virtual objects and
switched to the mathematical conceptual world, where theoretical objects and logi-
cal reasoning are treated. As in Yuri’s case in Episode 3, they can mathematise speed
as a rate in the mathematical conceptual world and recognise the graph plotted on
the screen as a mathematical entity. Although these phases are not always propor-
tional to types of students’ models (Table 14.1), the phases strengthen the four steps
in students’ development of sophisticated concepts of average speed (Thompson
and Thompson 1992), demonstrating that they extend to the relationship between
embodied action and concepts of average speed when using technology. In addition,
the phases provided empirical evidence of incorporating situated understanding via
technologies with the real-world phenomena that is the origin of mathematical mod-
els (Ärlebäck and Doerr 2015; Doerr and Pratt 2008).
There appear to be three essential actions necessary for the identification of
mathematical models of average speed. The first action is associating the phenom-
enal world with the virtual technological world (e.g. measuring the distance and
time). In Episode 2 (Fig. 14.3a, c), this action explicitly encouraged Group B stu-
dents to be aware of quantity. In addition, this action involved controlling their
embodied actions to reproduce the necessary graphs (Fig. 14.3e). The second action
is verifying and confirming students’ conjectures about movements and graphs. As
in Episode 2 (Fig. 14.3b, d, e), this action can be the trigger for identifying a time–
distance graph as a mathematical model of average speed. These results illustrated
the roles of technology in mathematical modelling indicated in previous literature
(e.g. Greefrath 2011). The third action is constructing embodied models as the base
of mathematical models. In Episode 3, using gesture and movement to physically
express speed or graphs on the screen were models of this kind. We speculate that,
thanks to these models, the students understood average speed with real feeling and
conceived the advantages and limitations of enactment, leading to being able to
distinguish between the phenomenal world and virtual technological world and then
switch to the mathematical conceptual world. These models can be ‘embodied’ situ-
ation models (Blum and Leiß 2007), describing a mental representation of the situ-
ation by enactment, indicating that we had better pay more attention to these
unsophisticated models in modelling education research.
Just as Doerr and her colleagues (2017) indicated that computational media in
modelling activities provide new ‘worlds’ to explore and potentially shape the world
we try to model, the approach of identifying mathematical models through sense-
making of embodied actions and technologies can highlight students’ gradual math-
ematisation from undifferentiated worlds, where the phenomenon is not differentiated
from its real-world manifestation, to differentiated worlds (i.e. the phenomenal
world vs. the virtual technological world vs. the mathematical conceptual world).
Acknowledgements The authors thank Koichi Tateishi, Atsushi Urago, and Go Ishii for their
cooperation in conducting classroom-teaching experiments. This work was supported by JSPS
KAKENHI Grant Numbers JP17K14053.
References
Ärlebäck, J. B., & Doerr, H. M. (2015). Moving beyond a single modelling activity. In
G. A. Stillman, W. Blum, & M. S. Biembengut (Eds.), Mathematical modelling in education
research and practice (pp. 293–303). Cham: Springer.
Arzarello, F., Pezzi, G., & Robutti, O. (2007). Modelling body motion: An approach to function
using measuring instruments. In W. Blum, P. L. Galbraith, H.-W. Henn, & M. Niss (Eds.),
Modelling and applications in mathematics education: The 14th ICMI study (pp. 129–136).
New York: Springer.
Blum, W., & Leiß, D. (2007). How do students and teachers deal with modelling problems? In
C. Haines, P. Galbraith, W. Blum, & S. Khan (Eds.), Mathematical modelling (ICTMA12):
Education, engineering and economics (pp. 222–231). Chichester: Horwood.
Doerr, H. M., & Pratt, D. (2008). The learning of mathematics and mathematical modelling. In
M. K. Heid & G. W. Blume (Eds.), Research on technology and the teaching and learning of
mathematics: Vol. 1. Research syntheses (pp. 259–285). Charlotte, NC: Information Age.
Doerr, H. M., Ärlebäck, J. B., & Misfeldt, M. (2017). Representations of modelling in mathemat-
ics education. In G. A. Stillman, W. Blum, & G. Kaiser (Eds.), Mathematical modelling and
applications: Crossing and researching boundaries in mathematics education (pp. 421–432).
Cham: Springer.
Geiger, V. (2011). Factors affecting teachers’ adoption of innovation practices with technology
and mathematical modelling. In G. Kaiser, W. Blum, R. Borromeo Ferri, & G. Stillman (Eds.),
Trends in teaching and learning of mathematical modelling (pp. 305–314). Dordrecht: Springer.
Greefrath, G. (2011). Using technologies: New possibilities of teaching and learning modelling –
Overview. In G. Kaiser, W. Blum, R. Borromeo Ferri, & G. Stillman (Eds.), Trends in teaching
and learning of mathematical modelling (pp. 301–304). Dordrecht: Springer.
Kaput, J., & Roschelle, J. (1999). The mathematics of change and variation from a millennial
perspective: New content, new context. In C. Hoyles, C. Morgan, & G. Woodhouse (Eds.),
Rethinking the mathematics curriculum (pp. 155–170). London: Springer.
Kawakami, T., Komeda S., Urago, A., Tateishi, K., & Ishii, G. (2015). Eliciting the concept of
speed through walking: Utilizing a graph calculator and distance sensor. In Research report of
Japan Society of Science Education (pp. 1–6). Saga: JSSE (in Japanese).
Komeda, S., Kawakami, T., Kaneko, M., & Yamaguchi, T. (2020). Deepening and expanding math-
ematical models of speed in relation to walking: The case of year 8 students. In G. A. Stillman,
G. Kaiser, & E. Lampen (Eds.), Mathematical modelling education and sense-making
Nemirovsky, R., Tierney, C., & Wright, T. (1998). Body motion and graphing. Cognition and
Instruction, 16(2), 119–172.
Núñez, R., Edwards, L. D., & Matos, J. F. (1999). Embodiment cognition as grounding for situat-
edness and context in mathematics education. Education Studies in Mathematics, 39, 45–65.
Papert, S. (1980). Mindstorms: Children, computers, and powerful ideas. New York: Basic Books.
Piaget, J. (1970). The child’s conception of movement and speed. New York: Basic Books.
Saeki, A., Ujiie, A., & Kuroki, N. (2004). Students’ analysis of the cooling rate of hot water
in a mathematical modelling process. In H.-W. Henn & W. Blum (Eds.), ICMI Study 14:
Applications and modelling in mathematics education: Pre-Conference volume (pp. 235–240).
Dortmund: University of Dortmund.
Swan, M., Turner, R., Yoon, C., & Muller, E. (2007). The roles of modelling in learning mathemat-
ics. In W. Blum, P. L. Galbraith, H.-W. Henn, & M. Niss (Eds.), Modelling and applications in
mathematics education: The 14th ICMI study (pp. 275–284). New York: Springer.
Thompson, P. W., & Thompson, A. G. (1992). Images of rate. Paper presented at the annual meet-
ing of the American Educational Research Association, San Francisco, CA.
Chapter 15
A Time-Based Measurement
of the Intensity of Difficulties
in the Modelling Process
Heiner Klock and Hans-Stefan Siller
Abstract In this chapter, we present an observation study which investigated the

solution processes of two groups of students. While solving the task, the students’
key activity was the analysis of a mathematical model given in the form of a simula-
tion. The study focused on difficulties in the modelling process, especially on the
intensity of difficulties which were identified by video recording and time taken.
The aim of this study was to identify the most persistent difficulties in the solution
process. For that purpose, a method consisting of a time-based measurement will be
presented. A coding scheme for identifying difficulties in the modelling process
based on previous studies was created and used for the evaluation of the solution
processes. Measuring the time the students spent overcoming their difficulties
allowed us to make a statement about the intensities of these difficulties.
Keywords Blockages · Difficulty · Level of intensity · Mistake · Time-based

measurement · Segregation · Simulation
15.1 Introduction
Mathematical modelling competency has to be mediated in mathematics education

(Niss and Højgaard 2011). To coach students’ modelling process adequately, a dis-
tinctive diagnostic competency is important for teachers. Identifying students’
H. Klock (*)
Department of Mathematics, University of Koblenz-Landau, Koblenz, Germany
e-mail: hklock@uni-koblenz.de
H.-S. Siller
Department of Mathematics, University of Würzburg, Würzburg, Germany
e-mail: hans-stefan.siller@mathematik.uni-wuerzburg.de

164 H. Klock and H.-S. Siller
difficulties when working on modelling problems is crucial with regard to an adap-

tive support of students (Stillman et al. 2010). Adaptive support means that it is
based on a diagnosis, adjusted in content and method, providing minimal help and
preserving students’ independence (Tropper et al. 2015). Findings concerning typi-
cal difficulties in the modelling process can therefore contribute to diagnostic com-
petencies and are relevant for the support of modelling activities.
Some problems during the modelling process are already recognised. Every
modelling step can in principle be a difficulty for students. In fact, it is important to
take a closer look at these sub-processes to provide a precise statement about diffi-
culties in modelling activities (Blum and Leiss 2007). For that reason, we want to
take previous studies into account and establish a new perspective on difficulties in
the modelling process. These studies examined difficulties using modelling tasks
covering all steps of the modelling cycle. Since our study used a task with a math-
ematical model already given, it is necessary to ensure that all steps were
also mapped.
We present a study which is based on a qualitative methodology but includes
quantitative elements. Our intention is the development of an instrument that mea-
sures the intensity of difficulties in the modelling process. The qualitative part
allows us to identify the difficulties in the modelling process, and with the help of a
quantitative time measurement, we are able to draw conclusions on the “Level of
Intensity” (Stillman et al. 2010, p. 393) of these difficulties. Thus, the method is
based on the assumption that a longer working process will occur when students
struggle in overcoming a difficulty. As the construct “Level of Intensity” has not
been quantified, we investigate two research questions:
1. Do the students run through all steps of the modelling cycle if they work on the
Segregation Task in which a mathematical model is given?
2. Does time measurement serve as a method and quantification of the level of
intensity to rank observed difficulties in terms of intensity?
15.2 Difficulties in the Modelling Process
In the past, numerous studies investigated impeding aspects in the modelling pro-
cess (e.g. Galbraith and Stillman 2006; Galbraith et al. 2007; Maaß 2004; Schaap
et al. 2011), although different terms and definitions were used to describe them.
Maaß (2004, p. 160 ff.) found that deficits occur in every modelling step. Mistakes
during the creation of the real model often appear together with errors in validating
the results. Average- to low-performance students working on complex tasks espe-
cially make mistakes in mathematising, solving the mathematical model and
15 A Time-Based Measurement of the Intensity of Difficulties in the Modelling Process 165
interpreting. Maaß summarised all mistakes in the following main categories:

Forming the real model, Forming the mathematical model, Interpreting, Validating
and in an overall category which includes aspects not referring to the model-
ling cycle.
Galbraith et al. (2007) observed blockages in transitions during the modelling
process (see also Galbraith and Stillman 2006). Blockages are characterised as
impeding progress. The authors developed a very detailed framework in which the
transition processes form categories as well. This framework was tested using addi-
tional data and finally confirmed. There is a “potential for blockages to occur when
any of these component activities have to be undertaken” (Galbraith and Stillman
2006, p. 146). By means of more data, the construct “Level of Intensity”, which
describes the “robustness of a blockage” (Stillman et al. 2010, p. 393), was devel-
oped in another study. They observed that blockages of high intensity occur during
the formation of the real and mathematical model. However, they also mentioned
that factors responsible for the variation of intensity have to be identified (Stillman
et al. 2010). In an additional study, students were asked about recognised difficulties
in the modelling process in a qualitative questionnaire. The students mentioned dif-
ficulties in different sub-processes of the modelling process as formulation, model
building, model solution, interpreting interim goals and verification (Stillman et al.
2013). Schaap et al. (2011) also investigated blockages in the modelling process,
but they focussed on the transition process of Mathematising starting at a real situ-
ation and ending at a mathematical model. They identified blockages in the sub-
processes Understanding, Structuring and Formalising.
In the studies discussed, the constructs mistake, blockage and difficulty were
used when describing impeding aspects in the modelling process. For example, a
blockage (Galbraith and Stillman 2006) may hinder the solution process, but there
is not necessarily a mistake in the sense of Maaß (2004) in this situation. Conversely,
a mistake will always lead to a kind of blockage. The word difficulty can serve as a
generic term including mistakes and blockages. Thus, we use the term difficulty in
this chapter and define it as follows: There is a difficulty in the modelling process if
the students make a mistake or if it is perceivable that they did not understand a
content of the real or mathematical world which is necessary to solve the task.
Based on the studies mentioned above, a framework was developed and used as
a coding scheme (see Table 15.1). Initially, all named mistakes, blockages and dif-
ficulties from the studies were classified. The main categories by Maaß (2004,
p. 160 ff.) served as a basis for the classification. The formulation of all sub-
categories was adapted and unified so that they all state deficits and do not overlap
in content. Two new difficulties were discovered during the analysis of the data and
were added to the category system: Fail to understand a foreign language means
having difficulties in understanding texts which are written in a foreign language.
The category Fail to understand mathematical contents means not being able to
understand the model or a mathematical text describing it.
Table 15.1 Categories of difficulties in the modelling process

Categories (modelling
steps) Subcategories (modelling subprocesses)
1. Forming a real-world 1.1 Fail to understand the context a, d
model 1.2 Fail to understand the task b, d
1.3 Fail to search for missing information c, d
1.4 Fail to identify relevant variables a, b
1.5 Fail to make meaningful and simplifying assumptions a, c
1.6 Fail to understand a foreign language e
1.7 Calculating without including the context d
2. Forming a mathematical 2.1 Fail to define variables a
model 2.2 Fail to realise dependencies between variables a
2.3 Fail to use adequate mathematical methods to mathematise a
2.4 Fail to use technologies a
2.5 Fail to understand the situation / mathematics conceptually d
2.6 Fail to understand mathematical contents e
3. Working 3.1 Fail to use adequate formulae a
mathematically 3.2 Fail to use adequate solution strategies and algorithms a, d
3.3 Fail to use technologies a
3.4 Fail to understand the situation / mathematics conceptually c
3.5 Fail to solve the model due to an excessive complexity d
3.6 Fail to convert units d
4. Interpreting 4.1 Fail to identify the correct meaning of aspects of the
mathematical results / model a, d
4.2 Fail to answer the question with the help of the mathematical
results d
5. Validating 5.1 Fail to reconcile (interim-)results with the real situation a
5.2 Fail to identify the influence of constraints / real-world aspects
on the mathematical results a
5.3 Fail to find opportunities to improve the model d
5.4 Fail to validate including all relevant aspects d
5.5 Fail to improve the model, so it fits to the real-world situation d
Note. a Galbraith and Stillman (2006), b
Schaap et al. (2011), c
Stillman et al. (2013), d
Maaß
(2004), e added by the authors
15.3 Method
The Segregation Task is about the actual problem of immigration and the potentially
involved formation of ghettos in German cities. Due to the presence of this topic in
the media, the task provides an authentic reference to reality. The challenge was not
the formation but the analysis of the mathematical model. This type of modelling
task was chosen because the mathematisation is very demanding and a suitable
simulation was available. Additionally, the model given allows a direct focus on
mathematics. It should be ensured that the students not only work with the context
but also take mathematical considerations into account when solving the task.
Segration Task
In the current political debate, the formation of ghettos in German cities due
to the number of refugees is an important issue. To prevent this, constraints
for the formation of ghettos have to be identified.
Thomas C. Schelling developed a model which describes the separation of
persons (= segregation) with a same attribute (e.g. religion, skin colour, etc.)
by means of an algorithm. This model was implemented in a computer simu-
lation and can be found at
http://nifty.stanford.edu/2014/mccown-schelling-model-segregation/.
Use this simulation to identify constraints for the formation of ghettos in
German cities.
The model given consists of a simulation based on a model developed by Thomas

C. Schelling (1971) which was originally evolved to describe a segregation process
of groups with different skin colours in large American cities. The intention was to
simulate a global behaviour which is caused by an individual attitude. The simula-
tion (see Fig. 15.1a) consists of a two-dimensional grid of grey and black cells
representing persons with different attributes. White cells represent empty spaces.
The algorithm distinguishes between satisfied and dissatisfied persons. A person is
satisfied if at least Similar % of the direct neighbours share the same attitude. Empty
cells are ignored (e.g. Similar=40%: A grey person with 2 grey, 4 black and 2 empty
cells neighbouring is dissatisfied). In the same round of the algorithm, every unsat-
isfied person moves to a random free cell. If there are not enough free cells, chance
decides which person moves. This procedure lasts until all persons are satisfied. A
segregation – a separation of two groups in areas called “ghettos” – is visible (see
Fig. 15.1b).
The simulation converges for different combinations of Similar %, the ratio
Grey/Black and the number of Empty cells (see McCown 2017). A potential
approach is the systematic variation of these parameters. As a result, students can
make a statement about the convergence of the algorithm and the number of rounds
needed to satisfy all persons. For example, with the help of self-generated data,
students can identify constraints of the segregation process.
Fig. 15.1 Simulation before the segregation process was started (a) Simulation after segregation
process has converged (b) (McCown 2017)
Two groups of male students (3 and 4 students, 18–20 years old) of a technical
high school were observed modelling the problem for 2 days (in total 11 h per
group). This type of school is largely attended by male students so that all groups
consisted of males. The groups did not interact within the regular working time and
worked on the task simultaneously. They were supervised by two secondary math-
ematics pre-service teachers per group who had to intervene adaptively on students’
request. To minimise the influence of the tutors, their main task was to observe the
students and to diagnose difficulties in the modelling process. There were no
content-related interventions by the pre-service teachers. The groups were video
recorded and their written work on the task was used for analysis.
In the analysis, we concentrated on difficulties within the stages of the modelling
cycle (see Table 15.1). In a first step and with the help of the coding system, coding
units were detected and labelled with one subcategory (1.1–5.5), initially regardless
of whether there was any difficulty. A multiple coding of the same coding unit was
not allowed. The duration of these coding units was calculated. In a second step, a
subsequent analysis of the coding units identified if there was a difficulty or not,
using the definition in Sect. 15.2. The variable # Codings (number of coding units
in a sub-category), Duration (duration of all coding units in a sub-category), #
Difficulties (number of coded units in which a difficulty appeared) and Duration
difficulties (duration of all coding units in which a difficulty appeared) were calcu-
lated to quantify the observation data. A part of the complete video material (12,7%)
was coded by the first author as well as a research assistant to determine inter-rater
correlation for the mapping of the coded units to the sub-categories in Table 15.1.
For that purpose, a definition, typical example transcripts and rules for coding were
created for every subcategory to ensure an independence of observation. The rela-
tively low agreement (Agreement .72, Cohen’s κ=.69) can be associated with the
extensive coding scheme.
15.4 Results
A total of 236 coding units were identified with the help of the coding scheme,
which encompasses 7 h 36 min 02 s of the complete video material (2 × 11 h = 22 h).
Group 1 provides 141 coding units (Min 03s, Max 28 min 40 s, Mean 02 min 09 s)
with a total time of 4 h 56 min 14 s, but group 2 contributes in 98 coding units (Min
04 s, Max 24 min 08 s, Mean 01 min 40 s) only 2 h 39 min 48 s to the overall coded
time. The considerable difference between the overall coded time of 7 h 36 min 02 s
and the total working time of 22 h can be explained by only considering difficulties
in the categories 1–5 and leaving out the creation of the final presentation. There
were also periods in which work was not carried out productively.
Table 15.2 shows the number of coding units and the total duration of all sub-
categories to which a code has been assigned. Relating to the first research question,
it can be stated that both groups have gone through all steps of the modelling pro-
cess, as at least one sub-category of each category (modelling steps in Table 15.1)
was coded. Therefore, tasks with a given model also have the potential to cover all
steps of the modelling cycle.
Both the groups differ essentially in their approach to the problem. Group 1 spent
much time on enquiry into the context (see Table 15.2, Subcategory 1.1) and tried to
program the simulation on their own (see Table 15.2, Subcategory 2.4). This led to
numerous difficulties. In the following example, a student explains his programmed
script to another team member.
S1: It is not working. Why is it not working? Look! [S1 and S2 look at the computer
together.] What is wrong with this? […] You just enter a number like, (…) a side of
the grid and then the other side of the grid is (…) It is always square. You square one
side and then you have the total number of cells. Then, you enter a percentage, for
example 10% (…) 10 by 100 is 0.1. You calculate the total number of 0.1 times and
then you have 10 empty fields. That worked. I already checked. But the total number
of persons does not work. The total number of persons is the total number of cells
minus the empty cells multiplied by the portion [e.g. for the grey persons] (…) Why
does it not work?
The situation shows that there is a mistake in the syntax of the script. As a conse-
quence, in the second step, following the definition in Sect. 15.2, this unit was coded
as difficulty. The two subcategories mentioned most differ in length from the
Table 15.2 Number of coded units and their duration per subcategory and group
Group 1 Group 2
Subcategory (Difficulties in #
Table 15.1) # Codings Duration Codings Duration
1.1 Understand the context 18 1 h 25 min 05 s 11 19 min 56 s
1.2 Understand the task 16 10 min 45 s 12 06 min 35 s
1.4 Identify relevant variables 6 09 min 52 s 2 01 min 12 s
1.5 Make assumptions 1 43 s 4 06 min 23 s
1.6 Understand foreign languages 3 45 s 1 35 s
2.1 Define variables 3 02 min 02 s 1 22 s
2.2 Realise dependencies 10 10 min 11 s 17 34 min 31 s
2.3 Use methods to mathematise 13 17 min 09 s 2 40 s
2.4 Use technology 8 1 h 03 min 56 s
2.5 Understand math. conceptually 3 04 min 07 s
2.6 Understand math. contents 17 21 min 10 s 7 15 min 33 s
3.1 Use adequate formulae 1 06 min 42 s
3.2 Use strategies and algorithms 1 08 min 12 s 10 51min 03 s
3.6 Convert units 14 01 min 33 s
4.1 Identify meaning of results 6 08 min 17 s 8 08 min 00 s
4.2 Answer the question 9 14 min 52 s 6 02 min 55 s
5.1 Reconcile with real situation 1 21 min 09 s 8 09 min 47 s
5.2 Improve the model 10 06 min 42 s 1 47 s
Fig. 15.2 Part of the working process of group 2 with corresponding coding units and non-coded
episodes
remaining ones. The group took 21 min 10 s to understand the model (see Table 15.2,
Subcategory 2.6). All other categories have a smaller time range but were entirely
relevant for solving the problem. Group 2 invested a lot of time to realise dependen-
cies between single variables of the simulation (Table 15.2, Subcategory 2.2), which
helped them to solve the task faster than group 1. Figure 15.2 shows a part of the
working process and the corresponding coding units. In the marked coding unit,
they executed the algorithm of the simulation manually for different configurations
(Table 15.2, Sub-category 3.2). The following transcript exemplifies a discussion on
the algorithm.
S1: It would be best to find the solution with the fewest moves.
S2: Yes, you have to mark all those [persons] who are dissatisfied and then you
have to think where you put them.
S1: Yes, but if I start with it, I have to see how I do it with as few moves as
possible.
S3: Yes, then you have to mark the dissatisfied ones first.
It is perceivable that the students have not fully understood the algorithm, as they
want to move the persons in such a way that they need as few moves as possible.
Since the algorithm runs randomly, the students did not understand a part of the
mathematical world. Group 2 spent only 15 min 33 s to understand the model (see
Table 15.2, Sub-category 2.6) that may have caused this difficulty.
Table 15.3 shows all coding units in which a difficulty was coded. We used work-
ing time needed to overcome a difficulty as a way to describe the “Level of Intensity”
(Stillman et al. 2010, p. 393) of a difficulty. The calculated variable Duration diffi-
culties characterises the time a group spent working on a difficulty and was used to
put the sub-categories in an order. It can be seen that in group 1 the sub-category
Fail to use technology (see Table 15.3, Sub-category 2.4), in which the students tried
to program the simulation, was by far the most time-consuming and therefore, by
our measure of intensity, carried the highest level of intensity. The sub-category Fail
to reconcile (interim-)results with the real situation (see Table 15.3, Sub-category
5.1) can also be interpreted as a relatively high intensity difficulty. For group 2, sub-
category Fail to use adequate solution strategies and algorithms (see Table 15.3,
Sub-category 3.2) was a high intensity difficulty. They misused it and took a long
time to detect their mistake. We observed another difficulty in the sub-category Fail
Table 15.3 Ranking of intensities of emerging difficulties based on the variable Duration
difficulties
Subcategory Duration
(Difficulties in # # difficulties
Ranking Table 15.1) Codings Duration Difficulties (Intensity)
Group 1 1 2.4 Use technology 8 63 min 3 32 min 41 s
56 s
2 5.1 Reconcile with 9 21 min 4 16 min 18 s
real situation 09 s
3 3.2 Use adequate 1 08 min 1 08 min 12 s
solution strat. 12 s
4 2.3 Use methods to 13 17 min 4 07 min 20 s
mathematise 09 s
5 5.2 Identify influence 1 06 min 1 06 min 42 s
of constraints 42 s
6 4.1 Identify meaning 14 08 min 7 04 min 28 s
of results 17 s
7 2.5 Understand math. 3 04 min 2 03 min 30 s
conceptually 07 s
8 2.6 Understand math. 17 21 min 3 03 min 21 s
contents 10 s
9 1.2 Understand the 16 10 min 3 01 min 50 s
task 45 s
10 2.2 Realise 10 10 min 1 00 min 49 s
dependencies 11 s
Group 2 1 3.2 Use strategies and 10 51 min 6 37 min 23 s
algorithms 03 s
2 1.2 Understand the 12 06 min 4 02 min 21 s
task 35 s
to understand the task (see Table 15.3, Sub-category 1.2), which has a relatively low
level of intensity.
It is conspicuous that group 1 had difficulties in many more subcategories than
group 2. This can either be attributed to different key activities carried out by the
students in the modelling process or to the modelling competence of the groups
themselves. Group 1 chose a much more challenging activity in programming the
simulation which led to more difficulties in the solution process than in group 2.
Therefore, it can be noted that emerging difficulties depend on students’ approach
to solve the task. The students’ modelling competence is also significant as demand-
ing activities can be mastered by high-performance students. The intensity of a dif-
ficulty should therefore always be understood in connection with the students who
try to overcome the difficulty.
15.5 Discussion
Based on the studies on mistakes (Maaß 2004), blockages (Galbraith and Stillman
2006; Galbraith et al. 2007) and difficulties (Stillman et al. 2013) in mathematical
modelling processes, we have created a category system. It was used to investigate
students’ difficulties solving the Segregation Task in which a mathematical model
was given. Against this background, the developed category system seems suitable
for investigating difficulties in modelling processes. We have determined that stu-
dents have gone through all modelling steps although a model was given in the task.
On the basis of the time-based measurement, it was possible to bring the intensities
of the observed difficulties into an order. However, investigations of further model-
ling tasks are needed to determine whether the time required to overcome a diffi-
culty is an appropriate measure of the intensity of difficulties. Highest intensity
difficulties determined the solution process not only in terms of time but also had a
very strong influence on its content-related development. In the study, the students’
solution process was in large parts based on their own programming of the simula-
tion (group 1) and the manual execution of the algorithm (group 2). Thus, a valid
measure of the two highest intensity difficulties can be assumed in this study. A
comparison of the emerging intensities between the groups does not make sense,
since the intensity always depends on attributes of the individual group.
The time to overcome a difficulty is influenced by support of the tutors. In this
study, the pre-service teachers did not intervene in terms of content, minimally and
only at students’ request. Even if the tutors have been asked to intervene minimally
in order to minimize any influence, it can never be completely excluded. Furthermore,
earlier difficulties affect later ones since there may be less time to get into other dif-
ficulties. The occurrence of difficulties and, thus also their intensity, depends to a
large extent on the students’ approach. It is therefore hard to predict difficulties of
high intensity for a task. However, the time-based measurement provides a first
pragmatic approach for quantifying the intensity of difficulties and at least enabled
us to find the most robust ones. The investigation of further solution processes of the
Segregation Task is necessary to confirm the difficulties found or to identify further
potential difficulties of high intensity.
Acknowledgments We thank the German Ministry for Education and Research for funding the
project MoSAiK (Kauertz and Siller 2016) – support code 01JA1605 – in the framework of
“Qualitätsoffensive Lehrerbildung”.
A special thanks to Pauline Vos for supporting us and discussing the idea of the
“Segregation Tasks”.
References
Blum, W., & Leiss, D. (2007). How do students and teachers deal with modelling problems? In
C. Haines, P. Galbraith, W. Blum, & S. Khan (Eds.), Mathematical modelling: Education,
engineering and economics (pp. 222–231). Chichester: Horwood.
Galbraith, G., Stillman, G., Brown, J., & Edwards, I. (2007). Facilitating middle secondary mod-
elling competencies. In C. Haines, P. Galbraith, W. Blum, & S. Khan (Eds.), Mathematical
modelling: Education, engineering and economics (pp. 130–140). Chichester: Horwood.
Kauertz, A., & Siller, H.-S. (2016). MoSAiK – Modulare Schulpraxiseinbindung als Ausgangspunkt
zur individuellen Kompetenzentwicklung [Modular integration of school practice as a starting
point for individual competence development]. Project of the University of Koblenz-Landau
in the framework of “Qualitätsoffensive Lehrerbildung”. http://mosaik.uni-koblenz-landau.de
Maaß, K. (2004). Mathematisches Modellieren im Unterricht. Ergebnisse einer empirischen Studie
[Mathematical model building in the mathematics classroom. Results of an empirical study].
Hildesheim: Franzbecker.
McCown, F. (2017). Schelling’s model of segregation [Simulation]. Searcy, AR: Computer Science
Department, Harding University. http://nifty.stanford.edu/2014/mccown-schelling-model-seg-
regation/. Accessed 13 July 2017.
Niss, M., & Højgaard, T. (2011). Competencies and mathematical learning. Ideas and inspiration
for the development of mathematics teaching and learning in Denmark. Roskilde: Roskilde
University.
Schaap, S., Vos, P., & Goedhart, M. (2011). Students overcoming blockages while building a
mathematical model: Exploring a framework. In G. Kaiser, W. Blum, R. Borromeo Ferri, &
Schelling, T. (1971). Dynamic models of segregation. Journal of Mathematical Sociology, 1,
143–186.
Stillman, G., Brown, J., & Galbraith, P. (2010). Identifying challenges within transition phases of
mathematical modelling activities at year 9. In R. Lesh, P. Galbraith, C. Haines, & A. Hurford
(Eds.), Modelling students’ mathematical modelling competencies (pp. 385–398). New York:
Springer.
Stillman, G., Brown, J., & Galbraith, P. (2013). Challenges in modelling challenges: Intents and
purposes. In G. A, Stillman, G. Kaiser, W. Blum, & J. P. Brown (Eds.), Teaching mathematical
Tropper, N., Leiss, D., & Hänze, M. (2015). Teachers’ temporary support and worked-out exam-
ples as elements of scaffolding in mathematical modeling. ZDM Mathematics Education,
47(7), 1225–1240.
Chapter 16
Deepening and Expanding Mathematical
Models of Speed in Relation to Walking:
The Case of Year 8 Students
Shigekazu Komeda, Takashi Kawakami, Masafumi Kaneko,

and Takashi Yamaguchi
Abstract This research explores the integration of mathematical modelling into

actual learning to see whether it is possible for mathematical modelling to be used
as a didactical means for supporting student learning of speed and related mathe-
matical concepts. This study explores modelling for learning in the secondary
grades to deepen and expand mathematical models of speed in relation to the act of
walking by using motion detectors and graphing calculators, as a part of the project
“Modelling a walk” in primary and secondary education. Students’ activities in
teaching experiments with Year 8 students in Japan are analysed. As a result of this
research, students were able to find a way of thinking leading to instantaneous speed
in deepening and expanding the mathematical model of speed while associating
this with movement, graph, and computation.
Keywords Mathematical models · Mathematics · Speed · Technology · Threshold

concept · Year 8 students · Walking
S. Komeda (*)
Graduate School of Education, Saga University, Saga-shi, Japan
T. Kawakami
Faculty of Education, Utsunomiya University, Utsunomiya-shi, Japan
e-mail: t-kawakami@cc.utsunomiya-u.ac.jp
M. Kaneko
Graduate School of Education, Naruto University of Education, Naruto-shi, Japan
e-mail: mkaneko@naruto-u.ac.jp
T. Yamaguchi
Junior High School Attached to Saga University, Saga-chi, Japan
e-mail: sn7848@cc.saga-u.ac.jp

176 S. Komeda et al.
16.1 Introduction
This research explores the integration of mathematical modelling into actual learn-
ing to see whether it is possible for mathematical modelling to be used as a didacti-
cal means for supporting student learning of speed and related mathematical
concepts (Swan et al. 2007). It is well known that technology is useful for grasping
mathematical concepts in elementary and secondary levels of schooling (Pead and
Ralph 2007). Many previous studies (e.g. Brown 2015) have aimed to support stu-
dents’ understanding of the mathematical concept of function, while few (e.g.
Ärlebäck and Doerr 2018) support the understanding of the mathematical concept
of speed. Therefore, with a focus on developing the concept of speed, we began the
project “modelling a walk” in primary and secondary school (e.g. Kawakami et al.
2015, 2016). The research project looks at how the model of speed was conducted
by Year 6 students (11–12 years old) and deepened and expanded by Year 8 students
(13–14 years old) in relation to the embodied function “walking” by utilising a
motion detector and graphing calculator.
In this chapter, we report on how to deepen and expand the model of speed by
analysing Year 8 students’ learning activities. After a short discussion on existing
literature regarding making sense of instantaneous speed and teaching Calculus-like
topics in the mathematics curriculum, we describe how we conducted teaching
plans for deepening and expanding the mathematical models of speed in association
with “walking” and multiple mathematical models, such as a table model, in Year 8.
Then we analyse typical students’ activities throughout the teaching practice.
Finally, we discuss the findings of the process in which students deepened and
expanded their knowledge using mathematical models of speed.
16.2 revious Research on Instantaneous Speed

P
and Calculus-Like Topics
Previous studies on making sense of instantaneous speed and teaching Calculus-like

topics (e.g. Ärlebäck and Doerr 2018) show that instantaneous speed is usually
taught alongside calculus in secondary school. In order to improve traditional
Calculus courses, technologies such as dynamic graphing and manipulation have
been utilised (Tall 2010). However, Tall (1993) noted the difficulty in understanding
the concept of limit as well as expanding the concept from average speed to instan-
taneous speed. Meyer and Land (2003) distinguished limit as a threshold concept.
Understanding a threshold concept at a basic level is important in order to deepen
understanding of advanced concepts (Perkins 2006). Some researchers (Ebersbach
and Wilkening 2007; Van Galen and Gravemeijer 2010) have already attempted to
teach the concepts of instantaneous speed and Calculus-like topics in primary
school. This research suggests that computer simulation is effective for primary
students to explore dynamic phenomena in complex problems (Van Galen and
16 Deepening and Expanding Mathematical Models of Speed in Relation to Walking… 177
Gravemeijer 2010). However, the students were only able to express their under-
standing of dynamic situations through graphs (Ärlebäck and Doerr 2018), and they
explored the concept of instantaneous speed using the idea of tangents (Beer et al.
2017). In line with previous research, the research presented in this chapter will
verify, within the Japanese school curriculum, whether students can learn for them-
selves instantaneous speed in relation to walking.
16.3 The Case Study
This project was started in 2015 in order to improve primary and secondary school
students’ ability to develop the concept of speed. We decided to gradually develop
the understanding of problems designed for this purpose, which Wittmann (2002)
referred to as a Substantial Learning Environment (SLE). SLE is a teaching material
that is a rich source of mathematical activity including the development beyond
grade level. At first, we led introductory lessons for Year 6 students to make sense
of the concept of constant speed, expressed as “speed = distance/time”. It became
obvious that the mathematical model of constant speed, in association with embod-
ied walking and using multiple models of mathematics (utilising motion detectors,
graphing calculators and other technology), was useful (Kawakami et al. 2015;
Kawakami et al. 2016). So far, we have not carried out research on deepening and
expanding the mathematical model of speed in relation to walking in our project.
In this new study, we investigated Year 8 students’ deepening and expanding
understanding from the mathematical model of constant speed to the mathematical
model of instantaneous speed in relation to “walking”. To do this, we carried out a
teaching experiment addressed in the following research questions:
(a) How did students deepen and expand their understanding of mathematical mod-
els of speed in relation to embodied “walking”?
(b) How did students articulate an idea leading to Calculus by learning to explore
instantaneous speed?
16.3.1 Speed in Japanese Curriculum
Figure 16.1 shows traditional Japanese curriculum examples which students have
already learnt before Year 8. Year 6 students learn the concept of average speed as a
ratio, and Year 8 students learn the rate of change and straight-line graphs during the
unit on linear functions. In the unit, students are invited to compare two straight
graphs (time-distance-graph) representing different speeds. They also learn to esti-
mate how fast they walk and to calculate constant speed partially from a given graph
(see Fig. 16.1).
Ms. Ikeda walked from her house to her

uncle's house. She stopped on the way at
a cake shop.
How far was it from her house to her
uncle’s house?
How long did it take her to get from her
house to her uncle’s house?
What was her speed from her house to
the cake shop?
Where was she in 70 minutes?
Fig. 16.1 A linear function problem in a Year 7 Japanese textbook
The following content shows traditional Japanese curriculum which students

will learn after Year 8. In Year 9, they learn about average speed in the unit regarding
quadratic functions as follows: A train does not move at a constant speed; therefore,
it is appropriate to find its average speed. For example, a train goes 135,500 m in
108 min from Asakusa to Toubunikkou. Students find that the average speed is
1255 m/minute. In Year 11, they learn instantaneous speed in the unit about Calculus
and Physics. In Japanese traditional curriculum, they learn the concept of speed step
by step. However, they learn about it without connecting it with embodied action.
The problem, therefore, is associating the real world with mathematical models. As
a result, students have insufficient understanding of the concept of speed as noted in
the literature as occurring elsewhere as well (Ärlebäck and Doerr 2018).
16.3.2 Research Design
To investigate these research questions, we analysed students’ activities during

teaching experiments with four classes of Year 8 students (13–14 years old) in
Japan. All classes comprised approximately 40 students. These students received
two mathematics lessons.
Major activities in each lesson were as follows:
Lesson 1: Reproduction of the given graphs (see Fig. 16.3) by using graphing cal-
culators and motion detectors. Consideration of the meaning of the vertical axis
and horizontal axis in the graphs.
Lesson 2: Examination of which is faster by comparing the two kinds of graphs
(straight-line graph and curved graph). Finding of the moment when the speed of
the two kinds of graphs becomes equal by using video, table, and graph.
We collected data: video recordings which were coupled with students’ activities
and calculator screens, interviews with students after lessons, lesson artefacts, and
field notes. The data were coded and examined for patterns and trends using an
analytical framework relating different mathematical models of speed. We analysed
their knowledge in association with embodied “walking” and multiple mathemati-

cal models by using a new framework (see Fig. 16.2) which was based on the previ-
ous framework of linking multiple models (Doerr and Pratt 2008; Kawakami et al.,
2016; Lesh and Doerr 2003).
The frameworks included the following mathematical models used in coding:
• Physical Model: Models showed a sense of “speed” with students using voice
and motion.
• Graph Model: Models used a graph of time (horizontal axis) and distance (verti-
cal axis), showing the state of the walking motion.
• Table Model: Models used a table of time and distance regarding the walk-
ing motion.
• Expression and Value Model: Model used expressions and values (distance, time)
for the state of the walking motion.
16.3.3 Teaching Practice
At the introduction of the first lesson, a teacher (the fourth author) explained to the
students how to use the graphing calculator and the motion detector. The 40 students
were divided into 10 groups. Then all students created graphs by using graphing
calculator and motion detector in each group.
The students then reproduced the given graphs (see Fig. 16.3) by using graphing
calculators and motion detectors. The teacher did not teach the meanings of the
vertical and horizontal axis until this time. Finally, the students discussed the mean-
ings of the vertical axis and the horizontal axis in the graph.
Fig. 16.2 Framework to analyse Year 8 student activities

Fig. 16.3 (a), (b) Two kinds of graphs given to students
At the introduction of the second lesson, the students reviewed the previous les-
son. Students confirmed how to walk graph (a) and (b) in Fig. 16.3. Then the teacher
asked the students a key question: “Which is faster, the motion shown by graph (a)
or (b)?” Many students answered that graph (a) is faster because the movement
shown in graph (b) apparently looks slow. The teacher showed videos to the stu-
dents which could simultaneously reproduce graph (a) and graph (b) afterwards.
Some students responded that it was the “same speed” because the start time and
the goal time were nearly simultaneous. Others could not answer and were confused.
The teacher, therefore, asked, “Which is faster at the beginning, and which is faster
at the end?” Some students answered, “(a) is faster at the beginning; (b) is faster at
the end”. The teacher confirmed to the students that “the speeds of (a) and (b) are
reversed halfway through. So, at that moment, the speeds of (a) and (b) were equal”.
The teacher asked the students a second key question, “When does the speed
become equal?” The students thought for a while but could not answer. Since the
students had learnt only constant speed up until then, this question triggered their
thinking about instantaneous speed. Then the teacher asked, “How can you solve it?
What kind of information do you need?” Some students answered, “Watch the video
footage, compare the graphs, and we need numeric values to understand the distance
and the time”. The teacher then showed the students a table representing the time and
distance of graph (b). The table represented the distance every 0.1 s from 0 to 9 s.
The students tried to solve the problem. Almost all students could calculate the
speed of graph (a). They calculated the speed of graph (a) as equal to 5 m/9 s because
the motion represented in (a) was moving at a constant speed. They then began
individually to find the moment at which the speeds are equal. After solving in their
own way, they discussed in each group. At the end of the second lesson, some of the
groups explained how to solve the problem and wrote an impression of the lesson.
16.4 Analysis
We analysed the collected data in two phases. In the first phase, we analysed group
activities based on the new framework and classified them. In another phase, we
analysed successful students who could solve instantaneous speed in various ways.
We describe the analysis results for a class typical of the four classes. Our analysis
is based on the framework in Fig. 16.2. Students’ answers were divided into three
categories, as there were three kinds of typical answers for the second key question:
those associated with the graph model, the physical model, and the computational
model initially. Out of the ten groups in the class, one associated their answer with
the physical model, four related to the graph model, and five connected their answer
to the computational model. We analysed three successful students’ work which
illustrates these categories in the following sub-sections.
16.4.1 Physical Approach
Student A related his answer to the physical model at first. He expressed his idea in
class as: “the speeds become equal when the distance between (a) and (b) was the
longest”. After the second lesson, we interviewed him, and he reasoned, “The dis-
tance between (a) and (b) gradually goes away and becomes the maximum. After
that, it gradually approaches. When the distance goes away, (a) is faster than (b).
And when the distance approaches, (b) is faster than (a). That is, the speed becomes
equal when the distance is the longest” (see Fig. 16.4b). We inferred that his idea
was based on his experiences in another subject. For example, he had raced with
other students in the physical education class. Firstly, he associated walking with
the physical model (see Fig. 16.4a①). Secondly, he searched for the place where the
speed became equal by overlapping graph (a) with graph (b) and by measuring the
length with a ruler (see Fig. 16.4b). He thus connected the physical model with the
graph model (see Fig. 16.4a②). Thirdly, he read the x-coordinate of the point. He
linked the graph model with the expression and value model (see Fig. 16.4a③). His
idea is important and leads to Calculus as follows:
Graph (a) is y = (5/9)∗x. Graph (b) is assumed to be y = f(x). F(x) = (5/9)∗x− f(x).
F′(x) = 5/9−f'(x). If F(x) has extreme (maximum) at x = x0, F′(x0) = 0. Therefore,
5/9-f′(x0) = 0, f′ (x0) = 5/9. Because f′(x0) means the slope of graph (b), the slope of
graph (b) at x = t becomes 5/9. As a result, when the distance between (a) and (b) is
the longest, the two speeds become equal.
Fig. 16.4 (a) Analysis of student A’s activity (b) worksheet of student A
Fig. 16.5 (a) Analysis of student B’s activity (b) worksheet of student B
16.4.2 Graphical Approach
Student B first related his answer to the graph model. He overlapped graph (a) with
graph (b) and drew some lines parallel with graph (a) below (see Fig. 16.5b). He
finished drawing the lines when a line parallel with graph (a) became a tangent to
graph (b). He answered, “As the slope of the graph is speed, I drew some parallel
lines below it (graph (a))” (see Fig. 16.5b).
We inferred that he came up with this answer because he had learnt about the
slopes of graphs of linear functions, as we mentioned before. He associated walking
with the graph model (see Fig. 16.5a①). We further inferred that his idea was that he
gradually narrowed the section until the speed became equal. This idea leads to the
mean value theorem. The mean value theorem roughly states that for a given dif-
ferentiable function f between two endpoints, there is at least one point at which the
tangent to f is parallel to the secant through its endpoints. Second, he read the
x-coordinate of the tangent point. He linked the graph model with the expression
and value model (see Fig. 16.5a②). However, he did not stop thinking about this
second key question.
16.4.3 Computational Approach
Student C first focused on the values in the table. He calculated the speed from 0 s
to 0.1 s, 0.1~0.2, 0.2~0.3, … by using the table (see Fig. 16.6a①). We inferred he
could calculate the speed at every 0.1 s because he had learned to calculate constant
speed partially from the given graph, as mentioned above. However, he stopped
halfway because the calculated speed was not close to the speed of graph (a). He had
calculated without relating it to the walking. Second, he compared the graphs of (a)
and (b) after a while. He muttered, “the slopes around 5 seconds look equal” (see
Fig. 16.6a②). We surmised that he partially visualised the graph and compared
slopes. Third, he resumed his calculation from 5 s to 5.1 s, 5.1~5.2, … (see
Fig. 16.6b). He came up with an estimate by linking walking with the graph model
Fig. 16.6 (a) Analysis of student C’s activity (b) worksheet of student C
and read numerical values from the table model (see Fig. 16.6a③). Fourth, he
resumed his calculations and solved for when the speeds became equal (see
Fig. 16.6a④). He summarised his comments after the lesson as: “It is important to
calculate by omitting the unnecessary parts. I can make an estimate by looking at
the graph”. Otherwise, he felt limited to the mechanical calculation. He found a
rational way of thinking by combining the real world with mathematics. His learn-
ing process shows the importance of linking the real world and mathematical models.
With respect to the first research question, students elicited the instantaneous speed
from mathematical models that they had already learnt. Then they investigated the
instantaneous speed by using those models. Before considering the two key ques-
tions, the students only thought about problems using constant speed. However,
these questions provided opportunities to think about instantaneous speed. The stu-
dents then explored those questions in relation to mathematical models that they had
already learnt, as well as to their real-world experience. Students answered the sec-
ond question by observing the videos of the students who could simultaneously
reproduce graph (a) and graph (b) (Physical Approach), by considering a parallel
tangent (Graphical Approach) as mentioned in the study of Beer et al. (2017) and by
calculating the speed at narrowed sections of time (Computational Approach). As a
result, they deepened and expanded their understanding of the mathematical model
of speed.
For the second research question, students were able to articulate an idea leading
to Calculus based on mathematical models that they had already learnt, combined
with their experience. Student A, for example, answered the second key question
with the idea that the speed became equal when the distance between graphs (a) and
(b) was the longest. This idea leads to the extremum, as mentioned above (see Sect.
16.4.1). Student B drew parallel lines with graph (a) to narrow the section that leads
to the mean value theorem.
This study illustrates that Year 8 students could deepen and expand mathematical
models of speed in relation to embodied “walking” by utilising motion detectors
and graphing calculators. Students also derived ideas that lead to Calculus in the
process of exploring instantaneous speed. Furthermore, this research has longitudi-
nal potential. Year 6 students developed the model of constant speed in relation to
walking (Kawakami et al. 2015, 2016). They compared two speeds and expressed
two straight lines with different slopes in a graph. This subsequent study with Year
8 students shows a deepening and expanding of the mathematical model of speed
with relation to walking occurred when exploring the moment at which speeds
expressed by a straight line and a curve become equal. At the next stage, we expect
students in upper secondary grades to solve for the moment at which speeds
expressed by two kinds of curves become equal by using Calculus. Use of
Wittmann’s (2002) idea of gradually developing problems has proved fruitful in
these studies at sequential schooling levels so far.
References
Ärlebäck, J. B., & Doerr, H. M. (2018). Students’ interpretations and reasoning about phenomena
with negative rates of change throughout a model development sequence. ZDM Mathematics
Education, 50(1/2), 187–200.
Beer, D. H., Gravemeijer, K., & Eijck, M. V. (2017). A proposed local theory for teaching instan-
taneous speed in grade five. The Mathematics Enthusiast, 14(1–24), 435–468.
Brown, J. (2015). Visualisation tactics for solving real world tasks. In G. A. Stillman, W. Blum,
& M. S. Biembengut (Eds.), Mathematical modelling in education, research and practice:
Cultural, social and cognitive influences (pp. 431–442). Cham: Springer.
Doerr, H. M., & Pratt, D. (2008). The learning of mathematics and mathematical modelling. In
mathematics: Vol. 1. Research syntheses (pp. 259–285). Charlotte, NC: Information Age.
Ebersbach, M., & Wilkening, F. (2007). Children’s intuitive mathematics: The development of
knowledge about nonlinear growth. Child Development, 78(1), 296–308.
Kawakami, T., Komeda S., Urago, A., Tateishi, K., & Ishii, G. (2015). Eliciting the concept of
speed through walking: Utilizing a graph calculator and distance sensor. In Research report of
Japan Society of Science Education (pp. 1–6). Saga: JSSE (in Japanese).
Kawakami, T., Komeda, S., Saeki, A., & Tateishi, K. (2016). How year 6 students mathematize
the act of walking in the introductory teaching of “speed”: A perspective of linking models. In
Proceedings of the 40th annual meeting of Japan Society of Science Education (pp. 63–66).
Ooita: JSSE (in Japanese).
Lesh, R., & Doerr, H. (2003). Foundations of a models and modelling perspective on mathematics
teaching, learning, and problem solving. In R. Lesh & H. Doerr (Eds.), Beyond constructivism:
A models and modeling perspective (pp. 3–33). Mahwah, NJ: Erlbaum.
Meyer, J., & Land, R. (2003). Threshold concepts and troublesome knowledge: linkages to ways
of thinking and practising within the disciplines. In C. Rust (Ed.), Improving student learning:
Improving student learning theory and practice – Ten years on (pp. 412–424). Oxford: OCSLD.
Pead, D., & Ralph, B. (with Muller, E.) (2007). Uses of technologies in learning mathematics
through modelling. In W. Blum, P. L. Galbraith, H.-W. Henn, & M. Niss (Eds.), Modelling
and applications in mathematics education: The 14th ICMI study (pp. 309–318). New York:
Springer.
Perkins, D. (2006). Constructivism and troublesome knowledge. In J. Meyer & R. Land (Eds.),
Overcoming barriers to student understanding: Threshold concepts and troublesome knowl-
edge (pp. 33–47). London: Routledge.
Swan, M., Turner, R., & Yoon, C. (with Muller, E.) (2007). The roles of modelling in learning
mathematics. In W. Blum, P. L. Galbraith, H. -W. Henn, & M. Niss (Eds.), Modelling and appli-
cations in mathematics education: The 14th ICMI study (pp. 275–284). New York: Springer.
Tall, D. (1993). Students’ difficulties in calculus. In Proceedings of working group 3 on students
difficulties in calculus, ICME-7 (pp. 13–28). Québec: ICME.
Tall, D. (2010). A sensible approach to the calculus. In Plenary at The National and International
Meeting on the Teaching of Calculus. Puebla: The National and International Meeting on the
Teaching of Calculus.
Van Galen, F., & Gravemeijer, K. (2010). Dynamische grafieken op de basisschool. Ververs
foundation.
Wittmann, E. C. (2002). Developing mathematics education in a systemic process. Educational
Chapter 17
Facilitating Modelling Activities in a Grade
5 Classroom
Azita Manouchehri, Mehmet Bekdemir, and Xiangquan Yao
Abstract Ten hours of whole group discussions in a US Grade 5 classroom were

analysed to identify teacher interventions aimed to stabilize student work on model-
ling tasks and the type of mathematical actions those interventions provoked.
Analysis of group interactions highlights that teacher interventions impacted stu-
dents’ reasoning, structuring, and mathematizing processes. These processes cor-
responded to the stages of the modelling cycle, including interpreting, structuring,
working mathematically, and validating ideas.
Keywords Data · Facilitating · Classroom interactions · Grade 5 · Rate of change ·

Teaching experiment · Teacher interventions
17.1 Introduction
There is consensus that teacher interventions are key in facilitating learners’ math-
ematical practices both in general and in modelling-based contexts in particular
(Brown 2017). Blum and Borromeo Ferri (2009) coined the construct of strategic
A. Manouchehri (*)
e-mail: Manouchehri.1@osu.edu
M. Bekdemir
Department of Mathematics Education, School of Education, Erzincan University,
Erzincan, Turkey
e-mail: mbekdemir@erzincan.edu.tr
X. Yao
College of Education, Pennsylvania State University-College Park, University Park, PA, USA
e-mail: yao.298@buckeyemail.osu

188 A. Manouchehri et al.
interventions to characterize those teacher moves that strike a balance between

guiding students’ modelling work and maximizing their independence. Creating
this balance is neither simple nor straightforward as recognized by the international
community of researchers of mathematical modelling. There is consensus that
teaching practices in modelling contexts can be far more complicated due to the
open-ended nature of tasks and unpredictability of students’ ideas. Calls have been
made for the need to invest greater scholarly efforts towards understanding the pro-
cess of teaching mathematical modelling, its associated outcomes, and how produc-
tive discussions around modelling activities can be facilitated in real classroom
settings (Cai et al. 2014).
In this work, we examined classroom interactions and one teacher’s attempts at
facilitating students’ mathematical modelling around one task over the course of a
unit of instruction that spanned approximately 10 h across nine class sessions. The
unit of instruction was developed during the experiment and was the result of stu-
dents’ reaction to an opening modelling activity, which asked them to study histori-
cal Olympic Games results and to predict the distribution of 2020 Olympic medals
according to the country and events. One broad question guided our inquiry: What
is the nature of teacher interventions in the course of students’ modelling process
and what impact do these interventions have on learners’ modelling practices?
17.2 Background Literature
Research has established that certain teacher practices promote meaningful, math-
ematical learning. Among many, these practices include encouraging active partici-
pation, building on students’ ideas to advance their thinking, stressing building and
justifying solutions, and creating a collaborative learning community that fosters
conjecturing and alternative ways of thinking (Mueller et al. 2014). Towers (1998),
in her synthesis of a substantial body of research on types of teacher interventions
in mathematics classrooms, identified 12 common approaches. She posited that in
places where instruction focused on fostering conceptual understanding and inquiry,
more prominent types included reinforcing—stressing an idea; nudging; inviting—
suggesting a new avenue of exploration; rug-pulling—introducing a puzzling idea;
and retreating—allowing the student to think further. Within the genre of mathemat-
ical modelling, relative to the teacher activities in the course of interactions, some
reports have noted the presence of all or a subset of such interventions within whole
or small group settings (de Oliveira and Barbosa 2013). This body of work in math-
ematical modelling highlights the role of the teacher as an active listener (Doerr
2006) and interpreter of students’ thinking (Brown 2017) and points at particular
strategies teachers used including inviting argumentation, collaboration, and sup-
porting autonomy (Anghileri 2006; Blum and Leiβ 2007; de Oliveira and Barbosa
2013). There is consensus that in both inquiry-based and modelling-based instruc-
tional media the teacher plays a pivotal role in advancing learners’ cognition, pro-
viding feedback when necessary whilst supporting their autonomy (Cai et al. 2014).
17 Facilitating Modelling Activities in a Grade 5 Classroom 189
The challenge associated with successful implementation of modelling-based

instruction, as identified by Blum and Leiβ (2007), is securing a balance between
minimizing guidance and maximizing students’ independent mathematical work.
Though intuitively sensible, greater clarity is needed on this type of teaching and
interventional types that assist in establishing such a balance. Our research aimed to
address this need.
17.3 Methodology
Data for this chapter comes from a portion of a teaching experiment (Steffe and
Ulrich 2013) conducted in one Grade 5 classroom in the USA during one academic
semester (approximately 3 months). This setting served as pilot work for a larger
study in which the goal is to construct detailed case studies of student-teacher inter-
actions as modelling contexts are used in different grade levels in three different
countries.
Our research is shaped by our view of teaching as the process of interacting with
students in an effort to first understand their mathematics, their models, and inter-
pretations and to then examine analytically how to proceed with instruction in ways
that students’ models are utilized. These views match the axioms of a teaching
experiment methodology. A teaching experiment involves a sequence of episodes
including a teacher/researcher, one or more students, an observer of the teaching
episodes, and a method of recording what transpires during the episodes (Steffe and
Ulrich 2013). Teaching in a teaching experiment involves learning how to interact
with students; the teacher does not enter the classroom with the intent to impose a
particular mode of thinking on students but uses interactions to better understand
students’ mathematics and constantly gauge her actions in order to offer a more
accurate account of mathematics of young students. Our analytical interactions in
this study became explicit in the type of activities we organized in response to the
students’ ideas, positioning ourselves as co-authors of the experience with the
learners.
17.3.1 Setting and Context
The research site was a Grade 5 classroom involving 24 students (17 male and 7
female). At the regular classroom teacher’s request, we commenced the teaching
experiment in January with the understanding that the research team would work on
mathematical modelling tasks with students once a week for approximately 60 min
each class session. The classroom teacher served the function of a witness to the
teaching experiment and collaborated in planning activities, contemplating class-
room discussions, and analysing classroom events. At the time that we entered the
classroom, students had been learning about fractions.
To start, we shared historical data (1910–2012)1 on the distribution of medals in

three events and asked students to prepare a presentation outlining their predictions
regarding the outcomes of the 2020 Olympics. We gauged the task to be of the open
modelling type (Galbraith 1989) so as to allow us to assess what students consid-
ered important and interesting mathematical tools they used in their preparation and
factors they considered in their analysis. One class session was devoted to small
group work during which students prepared a list of patterns they had noticed about
the data and questions they had about the results. They shared their observations and
questions in the large group during the second class session. The follow-up activi-
ties were informed by how to best help students develop mathematical skills and
tools useful for their documentation and analysis of data, based on their reports. All
these activities (approximately 7) focused on three main mathematical ideas: docu-
menting change in multiple forms, graphs and graphing, and quantifying rates
of change.
17.3.2 Data Analysis
Data analysis followed a three-stage process. First, an interactional map was con-
structed for each session to capture the group’s exchanges, depict turns in discourse
by the contributing members, and highlight episodes of “action-reaction” during the
interactional event. Since the goal of our research was to identify a typology of
teacher interventions during the whole group episodes, as a starting point in our
coding process, we relied on characterizations identified by the literature to mark
episodes of each type of intervention. Additional interventional types revealed in the
data were noted and frequency of their occurrence tallied. This phase of analysis
was augmented by a cross mapping of specific phases of the modelling cycle (Blum
and Leiß 2007) that the comments seemingly targeted or facilitated based on stu-
dents’ work. During the last phase, we noted students’ actions in reactions to the
teacher’s interventions.
17.4 Results
In reviewing the Olympic medal results, the students’ proposed over 100 comments
outlining patterns they had noticed and questions they had about data. Almost exclu-
sively students were curious to learn about the impact of socio-economic, political,
geographic climate, and finances devoted to hiring coaches and training athletes on
the number of medals secured by different countries or in different events. They
wanted to know why women were not allowed to compete until the 1920s and how
1
Data were obtained from http://www.theolympicdatabase.nl/olympic/medals/countries.
their training may have been different from that of male competitors. The issues the
students raised and their questions offered a fruitful ground for conducting a larger
interdisciplinary modelling (Blum and Niss 1991) where boundaries between math-
ematics and other disciplinary domains are removed. However, we opted to focus
students’ work on mathematical analysis. Despite this, students were encouraged to
select one or two questions from the list they had produced and to research those
questions online, an activity which several students had already begun to do.
Students suggested that they could use fractions and did so in ranking the number of
medals gained by different countries, different events, or according to gender. They
were encouraged to use other techniques including graphs. Whole group discussion
of their initial predictions revealed intense interest in comparing speed of the com-
petitors in various events and how male and female groups rivalled in different cat-
egories. They believed that, if competitors’ performance was expected, based on
current data, to improve over time, distribution of medals would be affected by such
changes. Students were particularly interested in exploring: How much faster is the
faster group? Which gender group is getting faster? How much faster are runners
compared to swimmers? How fast or slow is the gap closing between the groups?
What are some factors that might contribute to the decrease in the fastest times? In
comparing the women’s world record in the long course with men’s, which group is
getting faster in time? This set of questions shaped the follow-up activities used in
the subsequent sessions with a focus on facilitating the development of mathemati-
cal tools that would allow students to explore their own questions, which eventually
led to learners’ exploration of the concept of rate of change, slope, and connections
between instantaneous and cumulative graphs. Table 17.1 outlines the sequence of
activities used.
17.4.1 Student-Teacher Interactions
Two independent coders coded close to 10,000 utterances (not turns) as exhibited
during the whole group discussions of tasks. Table 17.2 summarizes the frequency
of different types of teacher interventions, students’ reactions to these interventions,
and modelling phases that the teacher’s comments tended to target and facilitate. In
addition, there were 30 prolonged pauses/silences (> 20 s) by the teacher.
Notice that the most prominent types of teacher interventions included eliciting
ideas, elaborating on students’ ideas and eliciting alternative representations, inter-
pretations and solutions, and probing and clarifying questions. These interventions
were crucial in establishing a discourse community in which students’ ideas took
centre in shaping instruction.
Prolonged episodes of teacher silence frequently followed “awe-like” expres-
sions as she attempted to decide what an appropriate follow-up move or questions
might be. Suggesting episodes were almost exclusively followed by retreating
actions. Note also that among the various teacher attempts not all were successful in
yielding the intended goal. Indeed, with the exception of eliciting explanations and
Table 17.1 Sequence of activities

Session Topics raised in class
Presentation of data on gold, silver, and bronze medals in running, Establishing trends
swimming, and triathlon events (2 sessions) Representing data
Running data (world record data on 100 m and 400 m male and Graphing data: building
female runners). Students were asked to organize the data and trend lines
report on trends (1 session) Continuous vs. discrete
graphs
Proportional reasoning
Swimming data (comparison of data on male and female records- Graphing data: distance/
freestyle and butterfly) (1 session) time graphs
Speed/time graphs
Constant vs. increasing rate
of change
Running vs. swimming (comparing data – how are they different?) Quantifying rate of change
Students compared speeds of record-holding runners and Distance/time graphs
swimmers in two distance categories and described them Speed/time graphs
numerically, graphically) (2 sessions)
Graphing events (students watched videos of 400 metre swimming Moving among different
and running episodes and documented the events on the same representations
coordinate graph) (1 session) Distance/time and velocity
time graphs
Matching graphs to stories (students were presented with Moving among different
narratives of running and swimming events of different representations
competitors in these events where a variety of functions could be Reading rate of change
used in capturing shifts in speeds if changed according to specific Connections between speed
time intervals) (1 session) time graph deriving the
distance time
Writing stories (events) to match graphs (students were presented Articulating rate of change,
with different graphs which captured occasions where change in quantifying rate of change,
movement was constant, increasing, or decreasing and asked to comparing rates of change
write narratives that matched those movements) (1 session) Formalizing steepness as
slope
elaborations, eliciting different representations, provoking generalizations, and

asking for connections among ideas, teacher prompts were not taken up by students
a majority of times. As illustrated in Table 17.2, when students were explicitly asked
to provide explanations, justifications, or generalizations, they tended to take up the
teacher’s suggestions.
In considering the relationship between the teacher’s interventions and students’
work along the modelling cycle, data revealed that over 90% of the teacher com-
ments targeted structuring, validating, and mathematizing stages. Elaborating on
students’ ideas, rephrasing students’ ideas, and restating/repeating ideas aimed to
create a common problem space for exploration and defining common representa-
tional tools. These interventions appeared crucial in advancing students’ articula-
tion of ideas in forms that were understood and referenced by others. Structuring
attempts became imperative in narrowing the range of variables and constraints
students considered which could easily hinder their mathematizing practices. A
Table 17.2 Percentages of teacher intervention types and modelling phase impacted
% of times
% of taken up by
Teacher intervention type occurrence students Modelling phase
Providing suggestions (typically 12 26.8 Interpreting,
phrased as I wonder what might mathematizing, validating
happen if…)
Asking questions about 4.8 66.7 Structuring and
connections amongst ideas simplifying
Interpreting students’ ideas 5.1 Expanding understanding
Probing/clarifying questions 30.2 90 Structuring and
simplifying
Invitations to debate 3.5 Validating, interpreting
Enculturating: introducing 4.7 Mathematizing, working
academic language, conventions, or mathematically
problem solving
Eliciting alternative representations 15 85 Working mathematically,
Eliciting explanations mathematizing, validating
Eliciting generalizations 5.9 69.2 Validating
Eliciting justification 6.2 92.7 Validating
Modelling specific practices 3.2 Working mathematically,
(setting norms) validating, structuring
Elaborating on students’ ideas 10 Expanding understanding,
Rephrasing students’ ideas structuring
Restating/repeating ideas
close inspection of the time intervals during which the teacher’s comments were of
provoking debates or eliciting justification or reasoning types revealed that over
60% of these types of interventions occurred during the first 4 weeks of the teaching
experiment. This same pattern was revealed again when we compared the frequency
of episodes of enculturating during the first 6 weeks and the last 5 weeks of the
teaching experiment. These shifts are illustrative of evolution of a modelling dispo-
sition among students. As data indicate, there was not a one-to-one correspondence
between a specific type of teacher intervention and a unique modelling phase of
students’ work. Indeed, as evident in Table 17.2, the same type of comment appeared
when facilitating students’ activities in various stages.
It is important to reveal that in the course of discussions students did not find it
particularly necessary to challenge various interpretations or representations that
their peers offered. When the teacher insisted that they may need to compare and/or
connect these ideas and representations, students complied. Episodes of dismissing
and/or ignoring the teacher’s suggestions became visible when her suggestions
required students to consider an alternative approach or representation when they
had already determined the problem resolved or intuitively obvious. These episodes
also challenged the teacher to find different ways to problematize the situation so as
to re-engage students in pursuing or formalizing ideas. She was not always success-
ful in meeting this goal. In places where students were not convinced of the
importance of the teacher’s request, they ignored her suggestions. However, when
requests addressed a common result in all groups, students tended to take on the
challenge, as depicted in the vignette described below.
17.4.2 An Illustrative Example
The vignette followed a group discussion of graphs that different groups produced
to capture the fastest runner’s (Wade Van Niekerk from South Africa) performance
and fastest swimmer’s (Sun Yang from China) performance as shown in two videos
from 400-metre swimming and running events in the 2016 Olympic games. The
three graphical representations that students produced appeared different (see
Fig. 17.1). Group 1 had generated a graph representing distance as a function of
time to record the events. Group 2 had considered time as function of distance, and
Group 3 had recorded speed as a function of time. The conversation that ensued
challenged the teacher to find effective ways of engaging students in analysis of
these graphs and their connections. Three images different groups had produced
were projected onto the screen as students explained their records. All students were
encouraged to comment on what was shared.
1. Kieren (Group 1): See with the runner it takes a few seconds for him (runner) to
get to the 10-metre point and then he runs at, covers about almost all the distance
with the same speed, and then again during the last 10–15 metres he speeds up
(points at their graph). It takes the swimmer a longer time to go to the finish
point, but Sun Yang swims at the same rate for the entire race. It only takes
him longer.
2. Jared (pointing at Kieren’s graphs): But the graph for Yang is steeper so it
means that it has to be faster.
3. Kieren: (silence)
4. Ivey (Group 2): In our graphs, you can see that the runner is faster… see? The
line for the runner is steeper.
5. Sina: I think they are the same. Like, if you rotate Kieren’s graph, it is just
like yours.
6. No comment from the group
7. Teacher: Sina, could you explain again what you mean?
Fig. 17.1 Three graphical representations produced by student groups

8. Sina: I just see the same thing only that they are reversed.
9. Jamie: You can’t just rotate graphs. I think Ivey’s is correct.
10. Teacher: But wait. I think Sina has a point here. I wonder how these two sets of
graphs might be related? I wonder if adding more detail might help.
11. Sam: Ours is different (third representation). We included the parts where the
speed changed—at the beginning it took both of them a couple of second to get
to the high speed. Their speed then increased at the end.
12. Jared: Van Niekerk’s speed stayed the same throughout the race.
13. Sam: That’s because the runner is so fast we can’t see it.
14. Jared: It looks like the graphs are parallel.
15. Tammi: But they can’t be the same since their speeds are different. Van Niekerk
is faster. It is not in the graph.
16. Sam: That’s why his graph is higher.
17. Silence (30 seconds)
18. Teacher: Any comments or questions? (prolonged silence)
19. Teacher: I am looking at these graphs, and if they are to describe the same
events, I wonder how they may be connected? I am wondering if we get some-
one from the hallway to come and read these graphs, not having seen the videos
whether they would be able to accurately describe the events you recorded. I
wonder also whether if we give a person the third set of graphs they would be
able to produce what the first and second groups tried to do which was captur-
ing the distances of the two competitors.
The vignette above is typical of the occasions when the teacher’s efforts at navi-
gating students’ practices were not particularly smooth. Note also her heavy reli-
ance on “I wonder” as a way of provoking further inquiry by students.
Our results further pronounce the indispensible role of the teacher (Blum and
Borromeo Ferri 2009) in an instructional environment where focus is on examina-
tion of real-life situations and highlights the complex nature of teaching practice in
such a setting. In our work, it became evident that the teacher’s suggestions for
scrutinizing/validating a particular solution or consolidating conflicting arguments
were not always taken up by students and for legitimate reasons—perceived ten-
sions by the teacher are not always shared by students. In navigating such challeng-
ing situations, two types of teacher interventions became foundational in advancing
students’ work: (1) presenting a follow-up question that demanded greater attention
to precision and (2) asking for connections among interpretations, ideas, and
representations.
First, it became obvious that success and failure of the teacher hinged upon her
knowledge of' developmentally appropriate ways of re-introducing concepts in
ways that accommodated students’ interests. In places where the teacher was
successful in problematizing the issue, using a different but compatible context,

students were more receptive to acknowledge concerns. This type of knowledge is
different from what Blum and Borromeo Ferri (2009) characterize as “task space”,
the potential of task for motivating various solution methods—but an understanding
of types of tasks and questions that activated learners’ cognition. Frequency of epi-
sodes of teacher’s long pauses increased when the students’ conversations around
the concept of rate of change became more intense and seemingly more difficult to
formalize in a manner that was developmentally appropriate. Second, the teacher’s
emphasis on seeking connections amongst the various representations, ideas, and
interpretations that students publicly shared was critical in advancing students’ for-
malizing and mathematizing work. In assuring that various ideas and representa-
tions that students shared were understood, the teacher relied heavily on rephrasing
and restating students’ ideas. Such a “connection seeking” stance became crucial in
helping students learn to robustly validate and formalize ideas.
We recognize that the Olympic Medal activity had the potential to move students
in mathematical directions other than those we pursued with them in the work we
reported here. We do wonder if the account of students’ work would be similar if the
teacher/researcher/research team had focused students’ modelling practices towards
the use of a different set of mathematical tools. We raise the issue of the need to go
beyond general descriptions of teaching actions and to unpack specific features of
strategic interventions in the context of content explored under investigations. This
issue is certainly important to the enterprise of teaching and learning mathematical
modelling since effective instruction relies on teachers’ knowledge of learners’ cog-
nition, associated obstacles in how concepts are conceptualized, and the impact of
those elements on ways that students interpret and mathematize contexts.
References
Anghileri, J. (2006). Scaffolding practices that enhance mathematics learning. Journal of

Mathematics Teacher Education, 9, 33–52.
Blum, W., & Borromeo Ferri, R. (2009). Mathematical modeling: Can it be taught and learnt?
Journal of Mathematical Modelling and Application, 1(1), 45–58.
Blum, W., & Leiβ, D. (2007). How do students and teachers deal with mathematical modelling
problems? The example “Filling up”. In C. Haines, P. Galbraith, W. Blum, & S. Khan (Eds.),
Blum, W., & Niss, M. (1991). Applied mathematical problem solving, modelling, applications,
and links to other subjects – State, trends and issues in mathematics instruction. Educational
Brown, R. (2017). Using collective argumentation to engage students in a primary mathematics
classroom. Mathematics Education Research Journal, 29(2), 183–199.
Cai, J., Cirillo, M., Pelesko, L., Borromeo Ferri, R., Borba, M., Geiger, V., Stillman, G., English,
L., Wake, G., Kaiser, G., & Kwon, O. (2014). Mathematical modelling in school education:
Mathematical, cognitive, curricular, instructional and teacher education perspectives [Research

Forum]. In P. Linljedahl, C. Nicol, S. Oesterle, & D. Allan (Eds.), Proceedings of the joint
meeting of PME 38 and PME-NA 36 (Vol. 1, pp. 145–172). Vancouver: PME/PME-NA.
de Oliveira, A. M. P., & Barbosa, J. C. (2013). Mathematical modelling, mathematical content and
tensions in discourses. In G. A. Stillman, G. Kaiser, W. Blum, & J. Brown (Eds.), Teaching
mathematical modelling: Connecting to research and practice (pp. 67–76). Dordrecht:
Springer.
Doerr, H. (2006). Examining the tasks of teaching when using students’ mathematical thinking.
Galbraith, P. (1989). From applications to modelling. In D. Blane & M. Evans (Eds.), Mathematical
modelling for the senior years (pp. 78–86). Melbourne: The Mathematical Association of
Victoria.
Mueller, M., Yankelewitz, D., & Maher, C. (2014). Teachers promoting student mathematical rea-
soning. Investigations in Mathematics Learning, 7(2), 10–30.
Steffe, L. P., & Ulrich, C. (2013). Constructivist teaching experiment. In S. Lerman (Ed.),
Encyclopedia of mathematics education. Berlin: Springer.
Towers, J. (1998). Telling tales. Journal of Curriculum Theorizing, 14(3), 29–35.
Chapter 18
Mathematical Modelling as a Learning
Environment to Transform a Street
Activity into a Sport Practice
Daniel Clark Orey, Milton Rosa, and Rogério Braga Soares
Abstract This chapter shares research in regard to a specific sports practice and
discusses its potential for educational contexts. A sports activity has significant
potential that can transform students’ lives. Thus, a specific street activity in Brazil
named carrinhos de rolimã (roller carts) can be linked to mathematical ideas
through the modelling process developed in an alternative learning environment.
This approach was an opportunity to deepen students’ understanding and broaden
their critical reflections on this well-known street activity in their own context and
community. From this practice, researchers propose the configuration of sporting
modalities in a modelling learning environment that may lead to a competition in
which students fairly compete among themselves.
Keywords Critical and reflective dimensions · Educational sports engineering ·

Learning environment · Mathematical modelling · Mixed methods study · Sports
practices · Street activity
18.1 Introduction
One of the main objectives of this study was to apply mathematical modelling as a
learning environment in order to contribute to the restructuring of teaching prac-
tices. In this context, one of the greatest challenges for teachers is to link theory with
practice in their teaching by transforming a well-known street activity in Brazil
called carrinhos de rolimã (roller carts) into a sport practice.
D. C. Orey (*) · M. Rosa · R. B. Soares

Departamento de Educação Matemática – DEEMA, Universidade Federal de Ouro Preto –
UFOP, Ouro Preto, Minas Gerais, Brazil

200 D. C. Orey et al.
For the purpose of this study, traditional roller carts are constituted of a wooden
body with three or four ball bearing wheels with a movable axis for steering and
control. The construction of these carts is simple and diversified, and they vary
somewhat in size and shape as well as in regard to the use of different materials and
terrain they are used on. The roller carts are mostly made from recycled materials
such as demolition wood and discarded ball bearings. A roller cart is known as a toy
used by young people in many countries made by using simple tools such as a ham-
mer and a saw. Figure 18.1 shows one of the most common types of Brazilian roller
carts built with a wooden frame and discarded steel bearings from automotive
repair shops.
It is important that teachers are supported to develop activities that arise from the
students’ own interests. In this regard, modelling and sport can be used to develop
students’ interest in diverse and sophisticated academic content that enables them to
grow in their ability to use mathematics. Sport and its distinct forms of competition
may challenge students and motivate them to increase their willingness to learn
mathematics. Hence, Rosa and Orey (2007) state that a modelling approach is use-
ful in coming to understand and comprehend problems, situations, and phenomena
that emerge in the everyday life of students. Therefore, the following research ques-
tion was proposed: What are the contributions that a mathematical learning envi-
ronment involving the construction of a roller cart and subsequent sports competition
in which students fairly compete among themselves can provide? According to this
query, it is important that students are given opportunities to recognize how math-
ematics can play a role in organizing fair, competitive, and efficient sports events
(Malkevitch 2017). In this context, mathematical modelling is understood as a
learning environment in which students, through the development of mathematical
activities, are invited to both inquire and investigate situations originating in other
non-academic/textbook areas of reality (Barbosa 2006).
In this study, modelling is considered part of a learning environment that helps
students to develop and exercise data-based opinions, creativity, and criticality
through the analysis, generation, and production of mathematical knowledge in
relation to the construction of roller carts. Through modelling, this context provides
Fig. 18.1 Brazilian roller

cart. (Source: Authors’
personal file)
18 Mathematical Modelling as a Learning Environment to Transform a Street Activity… 201
both real and concrete opportunities for students to discuss the role of mathematics
as well the nature of mathematical models (Rosa and Orey 2015).
This context enabled the authors to state that modelling provides a learning envi-
ronment that aims to facilitate the investigation of phenomena (street activities and
sports) through the elaboration of contextualized pedagogical activities (construc-
tion of roller carts) that help students to use mathematical knowledge to solve prob-
lems proposed in the classrooms. Thus, Noubary (2010) states that the widespread
interest in sports, in our culture, provides a great opportunity to catch student atten-
tion in mathematics classes.
In this chapter, we outline a teaching methodology with focus on the critical and
reflective dimensions of mathematical modelling that is based on the understanding
of students’ daily activities. This teaching strategy enables students to learn how to
analyse and take action in regard to problems they face in their own contexts.
18.2 Mathematical Modelling and Sports: Roller Carts
Most people think that mathematics is only applied to the sciences and engineering,
yet it is also used in other disciplines. Although it is not always perceived this way
by the population, mathematics plays an important role in sports. For example,
coaches use mathematical formulas to improve athletes’ performance (Malkevitch
2017) and discuss statistics related to the achievement of players and teams.
In this context, Lewis (2003) focused on the Oakland Athletics baseball team’s
analytical evidence-based and statistical approach to assembling a competitive
team. For example, in the United States, baseball statistics possess almost a cult
status for many followers of this sport. The same is true in countries in which soccer
is one of their main sports. Noubary (2010) also stated that sports provide an inex-
haustible source of fascinating and challenging problems for many disciplines. For
example, the use of mathematics includes the determination of the best batting order
for a team to maximize the number of runs a player can score or the scoring systems
for some of the complex and subjective aspects of sports events.
In addition, mathematical influences can be seen in the formalization of instru-
ments used by athletes and para-athletes. This includes balls, rackets, wheelchairs,
prostheses, court sizes, fields and demarcations, and pools and their dimensions. In
this regard, we have proposed a practical way to develop mathematical modelling in
classrooms through the elaboration of models in relation to the development of
sports practices related to roller carts. According to Noubary (2010), this approach
shows that through sports, many students can be exposed to the basics of mathemat-
ical modelling and statistical reasoning using a context that interests them.
Considering that the roller carts are powered from the initial impulse given by
the pilot, and by gravity, this study helped the students to connect mathematics and
science to their daily life. Here, many learners were motivated to perform calcula-
tions that examined the development of, and acceleration achieved by, the carts.
Calculations in the design and construction of the roller carts included those related
to the mass, circumference and diameter of the wheels, the number of wheels, geo-
metric shapes of the fixed and movable parts of its structure, the measurements
related to the axles and the wheels, the length of the main axles, as well as the over-
all symmetry of the cart and its aerodynamics. These calculations may guarantee a
certain standardization of this equipment and, consequently, a more judicious and
fair competition for competitors.
In this regard, Orey (2011) argues that mathematical modelling establishes a
bridge between sports practice and mathematics, which brings important contribu-
tions to pedagogical action developed in classrooms by providing an autonomous,
cooperative, meaningful learning environment that leads learners to the develop-
ment of creativity and criticality. This approach is related to the transformation of
academic mathematics into a living subject that is connected to real situations in
time and space that allows students to critically analyse and reflect upon mathemati-
cal phenomena that occur in their own communities (D’Ambrosio 1999). The origin
of many mathematical ideas is the result of a process that sought to solve problems
or explain and understand phenomena observed in reality. These ideas can be trans-
lated from elaborations of mathematical models whose acquisition, application, and
evaluation are linked to the process of mathematical modelling.
18.3 Methodological Procedures
The data collection and analysis as well the interpretation of the results of this study
were performed by applying a Mixed Methods Study design QUAL+quan. In this
design, a quantitative approach was embedded within a predominantly qualitative
study. The plus sign refers to the qualitative and quantitative data that were collected
and analysed simultaneously (Creswell and Plano Clark 2007).
This investigation was conducted with 34 students in a public night school con-
text. The participants’ ages ranged from 18- to 33-years old, and all were in the
second year of high school in the Youth and Adult Education Program in the Belo
Horizonte metropolitan region, the state capital of Minas Gerais, Brazil. The third
author was the teacher of these students.
In this study, educational sports engineering (Brothwell 2016) was applied in
association with the modelling tasks related to designing, building, and testing (vali-
dating) the roller carts by regulating and standardizing their dimensions, which were
used to analyse as well as interpret data to understand the construction and perfor-
mance of competitors and their carts. Accordingly, in this investigation, concerns
associated with unfair competitive advantages helped in the promotion of student
critical and reflective discussions related to planning and development of the equip-
ment (James 2010) through the exploration of sports engineering and modelling.
For data collection, students were organized into four groups, two with eight and
two with nine students each, in order to elaborate and develop modelling projects
related to the designing and construction of roller carts in which they could partici-
pate in a race competition under equal conditions for the competitors. The data col-
lection with regard to the design of the model of the roller cart and its construction
was conducted from 29 May 2017 to 26 June 2017. Student groups developed mod-
elling projects in 12 classes that were each 45 minutes long.
Quantitative and qualitative data were collected by using three activity modules
related to the learning tasks through the modelling process. Quantitative data were
organized and analysed by using descriptive statistics. Subsequently, qualitative
data were quantified in order to develop two conceptual categories: Modelling as a
Learning Environment and Modelling and Sports.
18.4 esults and Discussion Gleaned

R
from Conceptual Categories
Mathematical modelling as a learning environment enables students to analyse and

interpret data, to formulate and test hypotheses, and to verify the effectiveness of the
elaborated mathematical models taken from their own reality. For example, 24
(70.6%) students stated that they used geometric concepts, measurement and trans-
formation units, reduction scales, ratio, proportion, as well measuring instruments
such as rulers, squares, compasses, and scalers to elaborate the geometric models of
the roller carts.
Modelling, as we perceive it, is a process that is not a neutral or passive activity.
In this approach, students in their groups designed a model and constructed their
roller carts by learning how mathematical concepts were used in the preparation,
analysis, and resolution of their models. Then, students chose a standardized model
of the roller cart to be used in the race competition. Figure 18.2 shows the models
of the roller carts developed by the students in each group.
Since there were minor differences between the elaborated models, the research-
ers asked students to exchange information with other groups in order to analyse
and standardize their models. During the process of the standardization of the cho-
sen roller cart, students in their groups decided to determine the measures of the
bearings of their roller cart. Then, students developed and applied mathematical
procedures such as how to use measurement units to identify and propose solutions
Fig. 18.2 Models of the roller carts. (Source: Authors’ personal file)
Fig. 18.3 Models of the front and back bearings of the standardized cart. (Source: Authors’ per-
sonal file)
to the construction of the roller carts. Figure 18.3 shows the models of the front and
back bearings of the standardized cart, which were 60 mm and 70 mm, respectively.
During the mathematization process, one of the students said: “But 60 and 70 are
big measures”, and another student replied: “No, they are good sizes”. Thereafter,
the researchers realized that some students did not notice that the measurements
were given in millimetres and explained about scales and the mathematical relation
between millimetres and centimetres.
Later, in the development of the modelling project, during the validation of the
measurement of the parts by the woodworker, students decided to change the size of
diameter of the bearings to 70 mm in the front and 80 mm in the back of the roller cart.
This decision was made because the thickness of the wooden parts was changed from
3 cm to 2 cm and students decided to maintain the same height of the standardized
roller cart with respect to the ground as determined in the initial modelling project.
Students in their groups applied modelling to develop a standardized roller cart
for the competition by using mathematical content to accomplish the proposed
activities related to the design of the model and the construction of the roller carts.
For example, one of the concerns of the students was to determine the dimensions of
the roller carts that were suitable for the students, regardless of their height and size
so that all of them could participate in the race. One of the students was considered
as a reference for the development of this task because she was small. Then, this
student positioned herself on two desks as if she was sitting on a roller cart so that
students could determine measures of the length of the seat and the main axis. Next,
students used cardboard to perform this simulation to facilitate the development of
the measurements. Figure 18.4 shows the development of this simulation process.
Next, students drilled four holes in the main axis so that the seat could be adjusted
in two different positions to allow all students in their groups to fit in the standard-
ized roller cart for the race competition. According to the students’ decision, these
holes were drilled in the centre of the seat so that they were aligned with the back
Fig. 18.4 Simulation process of the standardization of the size of the roller cart. (Source: Authors’
personal file)
Fig. 18.5 Model and wooden seat of the standardized roller cart. (Source: Authors’ personal file)
Fig. 18.6 Standardized roller cart modelled by the students. (Source: Authors’ personal file)
extremity of the centre axis of the roller cart when it is in its highest position.
Figure 18.5 shows the model and the wooden seat of the standardized roller cart.
During the process of elaborating this model, students described, analysed, and
interpreted the data collected about the dimension of the parts of the cart. Then, they
sent their notes to the woodworker for the validation of their results. Figure 18.6
shows the standardized roller cart modelled by the students in all groups.
For example, 26 (76.6%) students affirmed that the standardization of procedures
makes roller cart competitions fairer. Thus, one of the students argued that “all the
roller carts are mathematically the same and what really matters is the preparation of
the drivers and their dedication”. According to Barbosa (2006), it is important that
results obtained in this process are linked to the students’ perceptions and reality.
In this study, mathematical modelling was considered a learning environment in

which students inquired and investigated a phenomenon (model of a standardized
roller cart). The students worked with real problems and used mathematics as a
language for understanding, simplifying, exploring, and solving situations in an
interdisciplinary fashion through mathematical modelling (Rosa and Orey 2015).
To sum up, the students applied this pedagogical action when they observed the
attributes of designing a model and constructing a standardized roller cart, verified
if specific outcomes could be produced and reproduced, and understood how to use
rules and norms to select different and efficient variables to manipulate and elabo-
rate this mathematical model.
18.5 Final Considerations
The emphasis on data analysis by students in relation to problems they face in their
own communities is one of the fundamental characteristics of mathematical model-
ling as a learning environment. For example, Rosa and Orey (2015) argue that the
ongoing and developing critical perspective of students in relation to social condi-
tions that influence their own lives and experiences helps them to identify common
problems and collectively develop strategies to solve them.
In this study, the focus of mathematical modelling was to use data in a specific
sport competition related to roller carts that have been initially created by the social,
cultural, climatic, and economic influences in Brazil in which popular and diverse
forms of competition arose and are still practiced. This presents a rational transfor-
mation because it involves critical analysis of sociocultural phenomena through the
elaboration of mathematical models. In short, they learn to move away from high
emotional arguments and focus on the data.
When learners explore problems taken from their own reality, they begin to study
symbolic, systematic, analytical, and critical contexts by using mathematical tools
developed and learned in a modelling-learning environment. In this context, model-
ling as a learning environment emphasized the role of mathematics in society as
well as the function of mathematical models in solving everyday challenges. This
approach helped students to move away from emotional arguments and to focus on,
and then apply, data-based tools to build a model of a standardized roller cart for a
race competition. Hence, this project incorporated a type of transformative learning
that created conditions that helped the learners to challenge the worldviews and
values they saw around them. Hopefully, this experience allowed them to reflect
critically on these experiences in order to develop a data-based rational discourse in
new opportunities and contexts.
Mathematical modelling is, as we use it, a teaching methodology that focuses on
the development of students’ criticality and reflexivity in order to engage them in
contextualized activities that enable the development of a deeper and increasingly
more active involvement in the constructions of social significance of the world of
the learners (Rosa and Orey 2015). Modelling as a learning environment is based on
the comprehension and understanding of reality and the generated data. When we
borrow systems from reality, students begin to study and analyse them symbolically,
systematically, analytically, and critically in order to make informed decisions
based less on emotion and more on data.
References
Barbosa, J. C. (2006). Mathematical modelling in classroom: A critical and discursive perspective.

Brothwell, A. (2016). What is sports engineering? Sheffield: ISEA.
Creswell, J. W., & Plano Clark, V. L. (2007). Designing and conducting mixed-methods research.
Thousand Oaks, CA: Sage.
D’Ambrosio, U. (1999). Literacy, matheracy, and technocracy: A trivium for today. Mathematical
Thinking and Learning, 1(2), 131–153.
James, D. M. (2010). The ethics of using engineering to enhance athletic performance. In
Proceedings of the 8th conference of the International Sports Engineering Association
(pp. 3405–3410). Vienna: ISEA.
Lewis, M. (2003). Moneyball: The art of winning an unfair game. New York: W. W. Norton &
Company.
Malkevitch, J. (2017). Mathematics and sports. Feature column: Monthly essays on mathematical
topics. Providence, RI: American Mathematical Society – AMS.
Noubary, R. (2010). Teaching mathematics and statistics using tennis. In J. A. Gallian (Ed.),
Mathematics and sports (pp. 227–239). Washington, DC: The Mathematical Association of
America MAA.
Orey, D. C. (2011). The math trail at California State University, Sacramento. Sacramento, CA:
CSUS. Retrieved July 30th, 2017, from https://sites.google.com/site/mathtrailcsus/.
Rosa, M., & Orey, D. C. (2007). Cultural assertions and challenges towards pedagogical action of
an ethnomathematics program. For the Learning of Mathematics, 27(1), 10–16.
Rosa, M., & Orey, D. C. (2015). Social-critical dimension of mathematical modelling. In
G. A. Stillman, W. Blum, & M. S. Biembengut (Eds.), Mathematical modelling in education
research and practice: Cultural, social, and cognitive influences (pp. 385–395). New York:
Springer.
Chapter 19
Concurrent Development of Engineering
Technician and Mathematical Modelling
Competencies
Lidamari de Villiers and Dirk Wessels
Abstract Competency gaps have been a contentious issue in education for many
years. This chapter explains how mathematical modelling and engineering techni-
cian competencies can develop concurrently over six modelling tasks. By investi-
gating the mathematical modelling and engineering technician competencies that
can enrich mathematical reasoning and understanding, particular competencies
were identified to be investigated and assessed. Twelve first year civil engineering
technician students studying at a South African University of Technology volun-
teered to partake in a design-based research study that lasted one semester. The final
results of the six modelling tasks showed slow but consistent progress in both engi-
neering technician and mathematical modelling competence development.
Keywords Mathematical modelling · Model-eliciting activities · Mathematical

modelling competencies · Engineering technician competencies · Design-based
research · Hypothetical learning trajectory · Sense-making
19.1 Background
The focus of this research study was to provide a local instructional theory that
explains how mathematical modelling and engineering technician competencies can
develop concurrently through mathematical modelling education, to enhance the
L. de Villiers (*) · D. Wessels

Department of Curriculum Studies, Stellenbosch University, Stellenbosch, South Africa
e-mail: lida@surgeon.co.za; dirkwessels@sun.ac.za

210 L. de Villiers and D. Wessels
students’ mathematical reasoning and understanding. When investigating the situa-

tion in both mathematics and engineering education, the following concerns
emerged:
• Lawless (2005), of the South African Institute of Civil Engineers (SAICE),
Professional Development and Projects, cited that undergraduate students reveal
a poor understanding of mathematics in the classroom and school learners have
not grasped basic mathematical principles. Chick and Stacey (2013) noted that
mathematical knowledge is often expressed in terms of content aspects. To come
to terms with the content, students tend to revert to rote learning, instead of
developing a deeper understanding of the subject matter.
• Students’ poor knowledge base restricts them from being able to solve real-world
mathematical problems – a primary task of any engineer (Moore et al. 2013).
Their goals are limited in finding algorithms to produce instant answers, without
necessarily knowing what their mathematical solutions represent (Skemp 1976).
As poor problem solvers, they lack critical and reflective thinking abilities to
construct viable mathematical solutions from real-world problems, and this gap
continues to widen (Lawless 2005).
• Furthermore, Woollacott (2003) stresses a concern for the South African engi-
neering students that are under-prepared and not successful in achieving their
qualifications. Apart from an inadequate knowledge base, gaps also emerged in
terms of metacognitive competencies, such as team-working, decision-making
and effective communication, even though such skills are regarded as essential
building blocks to becoming successful students and professional engineering
technicians.
The goals of both engineering and mathematics education emphasise the impor-
tance of mathematical understanding, real-world problem-solving and adequate
metacognitive competencies (Schoenfeld 1992; Woollacott 2003). The above con-
cerns exposed the current gap in mathematics and engineering education on how to
allow students to develop the habits of interpretation and sense-making (Schoenfeld
1992). This study aimed to provide an explanation of how students can develop
competencies to enrich their mathematical reasoning and understanding.
19.2 Research Question
To address this complex situation, the following research question emerged:

To what extent can first year students’ mathematical modelling and engineering technician
competencies develop concurrently through mathematical modelling?
It is well documented that mathematical modelling has the potential to address the
gap between applying mathematics in the real world and addressing mathematical
concerns in the classroom (Kaiser 2007). Mathematical modelling authentically
connects mathematics to the real world and is used to explain real-world phenom-
ena and/or make predictions about future behaviour of a system in the real world.
19 Concurrent Development of Engineering Technician and Mathematical Modelling… 211
Mathematical modelling represents iterative cycles where students can come to

understand a real-world problem, simplify, mathematise, interpret and validate the
solution against the real-world situation, and opportunities for generalising obtained
solutions to similar situations are provided (Blum and Leiβ 2005). Modelling also
facilitates the development of competencies while students learn to formulate, solve
and make decisions concerning real-world problems (Maaß 2006). These processes
allow for multiple approaches and answers (Lingefjärd 2006). Cirillo et al. (2016)
regard the translation of a real-world problem into a mathematical form as the
“essence of mathematical modelling” – the problem needs to be simplified and
explained, variables must be identified, approximations are made and the solutions
need to be justified and reported. The active sense-making of meaningful situations
through invention, extending and iterative refining of the students’ own mathemati-
cal constructs are emphasised (Bahmaei 2013).
19.3 Identification of Competencies
To align with the requests from the workplace, it was necessary to examine the
competencies required from professional engineering technicians. The Engineering
Council of South Africa (ECSA 2016) is the only body in South Africa that is autho-
rised to register engineering professionals and bestow engineering titles on persons
who have met the requisite professional registration criteria. ECSA’s mission state-
ment focuses on establishing a South African engineering profession that can be
successful in creating socio-economic growth in the country. To meet these obliga-
tions, the competence of individuals must be ensured. In collaboration with other
international engineering bodies, ECSA has identified the following crucial cogni-
tive and metacognitive competencies for successful engineering technicians: defin-
ing and analysing problems, designing and developing solutions, comprehending
and applying knowledge, recognising and addressing factors, meeting health and
safety requirements, sound judgement, communicating, managing and
responsibility.
Blomhøj and Jensen (2007) define mathematical modelling competence as
“being able to autonomously and insightfully carry through all aspects of a mathe-
matical modelling process in a certain context” (p. 48). Through mathematical mod-
elling, students have opportunities to develop a multitude of mathematical modelling
competencies. Apart from the cognitive competencies relating to understanding the
modelling process, metacognitive competencies enhance critical reflection and
assist in managing and controlling one’s own thinking processes during the model-
ling activity (Kaiser and Brand 2015; Maaß 2006). To determine the particular com-
petencies to be investigated, engineering technician and mathematical modelling
competencies and sub-competencies were compared. Sub-competencies refer to the
skills and processes that need to be carried out to master the competencies. These
comparisons assisted towards the goal of understanding what the competencies
mean and determining a way to recognise and identify them in the students’ work.
Finally, the following criteria served as a framework to select specific competencies:
relevance, significance, personal judgement and the level of detail had to be man-
ageable and useful to categorise competencies (Woollacott 2003).
By comparing the competencies, the synergy between the goals of mathematics
education and engineering education was exposed, which allowed the selection of
particular competencies to investigate and assess. The resulting mathematical
modelling competencies which formed the focus of this study were internalising,
interpreting, structuring, symbolising, adjusting, organising, generalising, commu-
nicating, managing and responsibility. These cognitive and metacognitive compe-
tencies are all interrelated and do not develop in isolation. Students’ knowledge and
beliefs impact on how they develop cognitive intelligence and how they adjust their
ways of thinking to solve real-world problems (Brown 1992).
19.4 Context of the Teaching Experiment
Purposive homogenous sampling applied to this experiment, since all the students
were first-year civil engineering technician students studying at a University of
Technology. Twelve students volunteered to participate in this study. They had no
prior experience of mathematical modelling activities and were not exposed to
mathematics education during the semester. They were enrolled in a bridging pro-
gram, as they did not meet the entrance requirements to be allowed access to study
engineering.
The first session served to orientate the students to what mathematical modelling
entails. Socio-constructivist classroom norms were explained as well as the impor-
tance of developing mathematical modelling competencies. Over one semester, the
students worked in small groups of four students each, on six thought-revealing
model-eliciting activities (MEAs) which necessitated the construction of models to
describe, analyse and solve real-world problems. MEAs are real-world, client-
driven problems which students solve in small groups. The task descriptor of a
MEA elicits the need to construct a mathematical model that does not reside in the
problem statement (Baker and Galanti 2017). Students need to produce descrip-
tions, procedures and solution methods that reveal their ways of thinking. Rather
than processing information with a given procedure, they learn to make sense of
meaningful situations through inventing, extending and refining their own mathe-
matical constructs (Bahmaei 2013). Solutions are represented in a variety of ways,
including spoken language, written symbols, graphs, graphics and metaphors. Such
representations allow for students to develop a holistic set of competencies (Moore
et al. 2013).
Qualitative data were generated from the students by means of various assess-
ment instruments, including reflection questionnaires, observation guides, student
worksheets, field notes, informal discussions, and video and audio recordings.
These instruments assisted the students and the researcher to gain a better under-
standing about their ways of thinking and executing mathematical problems. During
informal discussions, the participants shared, compared, explained and justified
their solutions with the rest of the class and with the researcher/facilitator. Informal
discussions were continually employed to answer queries that arose while data col-
lection took place. Data generation occurred immediately after each session. Data
were analysed and coded to investigate competence development.
19.5 Methodology
In searching to develop the students’ competencies for enhanced mathematical rea-

soning and understanding, this study required a detailed investigation of compe-
tence development with an interactive, iterative and interventionist approach within
a socio-constructivist perspective. For this reason, design-based research (DBR)
was appropriate. Gravemeijer and Cobb (2006) define the purpose of DBR as to
“develop theories of both the processes of learning as well as the means designed to
support that learning” (p. 18). The study took place in the messy and complex class-
room situations that characterise real-life learning within a particular context rele-
vant to engineering technician students.
The retrospective analysis of a pilot study that took place prior to the main study
with different participants, who worked on similar modelling activities, formed the
basis of this design research. Two main arguments resulted from the pilot study:
procedural understanding and memorising seem to be of primary importance when
the students learn mathematics, and the students tend to do what they are told, rather
than trying to enhance their mathematical understanding and reasoning. The results
of the pilot study served as a link between the literature study and the teaching
experiment. DBR consists of three distinct phases: (a) the design and preparation
phase, (b) the implementation of such problems within a socio-constructivist per-
spective and (c) a reflection phase (Gravemeijer and Cobb 2006).
19.5.1 Design and Preparation Phase (Phase 1)
To answer the research question, a hypothetical learning trajectory (HLT) was con-
structed to serve as a guide to select and develop the instructional materials and to
select and design activities. Simon (2014) explains the HLT as threefold: it consists
of the learning goal that defines the direction, the learning activities and a prediction
of how the students’ thinking and understanding will evolve in the context of the
learning activities. The HLT thus includes information about the learning tasks, as
well as the variables that can affect outcomes such as teachers, students, classroom
ethos, contextual activities and reflection tools (Reeves 2006). Six modelling tasks
from literature that had proven successful in competence development were adjusted
by considering both the students’ current levels of understanding, as well as the
results of the pilot study. The last activity will be discussed in this chapter. It was
adapted from the Lost Cell Phone Problem of Anhalt and Cortez (2015). While
planning and designing the activity, anticipated questions were prepared to assist
students with difficulties and to guide them towards solution paths (Wake et al.
2016). Typical questions were:
• How can you improve/modify your model?
• Can you obtain a more specific answer by reconsidering the given data?
• Should you perhaps establish new assumptions?
• Do you know anything else that you have not considered yet?
19.5.2 Implementation Phase (Phase 2)
The second phase of DBR involved empirical data generation while the students
were engaged in the activities. During this phase, the HLT served as a guideline to
ensure that all scaffolding and observations were focused towards developing their
competencies (Simon 2014). In this activity, the students’ task was to locate a miss-
ing cell phone. Attached to a client’s letter was a topographic map, indicating the
co-ordinates of three cell phone towers that could pick up signals from the cell
phone, as well as the towers’ signal distances to the cell phone. While working in
groups, students externalised and revealed their ways of thinking. Two groups con-
sidered determining the midpoint between the three towers, while one group decided
to calculate the average value of the x- and y-coordinates.
The transition from initial to final interpretations followed the typical process as
described by Moore et al. (2013): Students initiated their activities from disorgan-
ised and inconsistent ways of thinking about their goals and possible solution steps.
They generally recognised the need to develop a model but ignored difficulties relat-
ing to surface-level details. The students attempted to find the point of intersection
by using the midpoint theorem of triangles rather than finding the distances between
the antennas and the cell phone. By comparing their interim results to the data, they
abandoned their ideas and started again. The researcher made use of the anticipated
questions that guided them to rethink the problem. The students revised and refined
their original understandings and eventually all the students engaged in constructing
circle equations, which directed them towards improved solutions.
Iterative “express-test-revise” cycles prompted the students to consider the use of
the topographic map. They adjusted their calculations and managed to obtain
improved results but were still unable to determine an exact location of the missing
phone. Group discussions and re-visiting their assumptions led them to consider the
position of the antennas with reference to the ground. As Lesh and Clarke (2000)
explained, their intermediate interpretations continued to evolve, and the students
repeated these “express-test-revise” cycles until they could produce results without
further adjustments. Throughout the activity, they reflected on their thinking strate-
gies to make appropriate revisions to their solutions. As they progressed through the
iterative sequences of interpretation-development cycles, their initial disorganised
Fig. 19.1 Students’ progress during a MEA
and inconsistent ways of thinking proceeded to more mature mathematical reason-

ing, as depicted in Fig. 19.1.
19.5.3 Retrospective Analysis (Phase 3)
The careful reporting of all the design processes allowed for the collecting and ana-
lysing of relevant data to develop a localised theory to explain the participants’
competence development. The consistent and logical methodology offered by DBR
enabled the researcher to link theoretical research with educational practice and to
build a theory of how the intervention worked (Van den Akker et al. 2006). The iter-
ative nature of MEAs offered opportunities to students to develop new cognitive
structures and deeper understanding than what would have been allowed by a single
iteration application of textbook formulas. This MEA allowed the students to remain
deeply engaged in the process to develop, explain and interpret specific situations
by themselves. Trails of documentation were generated that revealed important
aspects about the students’ ways of understanding and their ways of thinking. The
students’ progress reflected Piagetian-like stages of concept development, where
tension between accommodation and assimilation characterised the process of
learning, and they were able to modify their viewpoints to find a point of intersec-
tion. All episodes relating to competence development were transcribed, coded,
analysed and tested against the various data collection instruments as well as com-
pared to current literature to search for confirmation and counter-examples.
19.6 Results
The thorough investigation of engineering technician and mathematical modelling

competencies, combined with meticulous observation and recording of events as
they played out, allowed tracking the development of the students’ competencies.
Marks were allocated for each competency between 0 and 3, respectively, denoting
unsatisfactory, emergent/developing, proficient and exemplary performances.
Figure 19.2 indicates slow but consistent progress in all the competency
domains over the six modelling tasks. This gradual increase in competence develop-
ment complements Lesh and Harel’s (2003) views that students’ levels of develop-
ment not only change during the course of one modelling activity but also that
students progress from one activity to another which is structurally similar.
The long-term exposure to mathematical modelling instilled further growth in
the students’ abilities to internalise the problem situation. Rather than leaning
towards finding numbers and doing meaningless calculations, they proceeded
towards searching intentionally for a deeper understanding of the problem. The
competency of interpreting seemed to improve during the students’ further itera-
tions of the modelling process. Initially they created assumptions unintentionally
(e.g. antennas located at bottom of towers, surface area is flat) and only considered
and recognised factors that could influence the situation during subsequent itera-
tions. Even though the representations seemed to be problematic for some students,
they still recognised the need for creating (structuring and symbolising), revising
and refining their models (adjusting). Initially these constructs were primitive and
contained errors that would not lead to adequate solution paths. The development of
Fig. 19.2 Development of students’ mathematical modelling and engineering technician compe-
tencies over six modelling tasks
students’ capacities to evaluate and judge was revealed when they queried their
initial results as being too vague. This realisation indicated that they advanced to a
level where they also considered improved accuracy and more elegant solution pro-
cesses. This manifestation is indicative of improvement in their organising compe-
tencies. Generalising of their solution procedures was displayed during their
presentations, as the students explained their solution procedures independent of
specific coordinates, heights of antennas or locations of cell phones above sea level.
Management strategies enabled the students to manage and control the material and
resources at their disposal to reach their goals as well as to persevere and maintain
their intellectual engagement when grappling with difficult tasks. The students were
all actively engaged in the activity and revealed characteristics of responsible
behaviour as they all worked enthusiastically within their groups to obtain elegant
solutions. Even though this study anticipated competence development, none of the
competencies under investigation was defined in terms of what it would mean to
possess perfect competence. Competence development will remain a growing pro-
cess, directing the opinion that all students should always strive to continue to
develop themselves further, as no final stage of expertise exists.
Triangulation served to ascertain trustworthy, reliable and valid results for this
study and was accomplished through various methods, such as informal discussions
with the students, observations, analysing reflection questionnaires and evaluating
their final products. All these methods were employed to examine the same dimen-
sion of the research problem – engineering technician and mathematical modelling
competence to enhance reasoning and mathematical understanding. Figure 19.2
denotes the development of the engineering technician and mathematical modelling
competencies for the whole class over six activities on a bar chart. A consistent
improvement in competence development can be noticed.
19.7 Design Concerns and Limitations of the Study
This study was primarily concerned with the analysis and documenting to explain
how successive forms of reasoning emerged as a restructuring of prior forms of
cognition. Throughout the study, the researcher aimed to identify aspects of the
learning environment that could support students’ development of mathematical
modelling competencies. All potential risks and concerns were dealt with continu-
ously through careful and meticulous observation and recording of events (truthful
representation), by listening to the participants’ voices (teacher/researcher relation-
ship) and by utilising a vast range of data collection methods (trustworthiness).
Argumentative grammar guided the researcher to ensure that the learning pro-
cess justified the products of the research project (purposeful research). Kelly
(2004) explains argumentative grammar as “the logic that guides the use of a method
and that supports reasoning about its data. It supplies the logos in the methodology
and is the basis for the warrant for the claims that arise” (p. 118). Engeström (2011)
regards argumentative grammar as the golden thread that connects the theory, meth-
ods and empirical research and allows the researcher to establish a learning
trajectory that explains the shifts in students’ reasoning abilities while they are
engaged in the learning activities. The detailed data collection and analysis (DBR
methodology) allowed the researcher to ensure that the students’ shifts in reasoning
and the accompanied development of particular competencies resulted from the
support of the instructional design and that it did not occur due to a mere sequence
of events. The researcher was interested in identifying the mechanisms that can
provide credible explanations of causal relationships and processes under local con-
ditions (Maxwell 2004), which could pave the way to possible dissemination.
19.8 Conclusion
The results of this study indicated slow but consistent progress in all relevant math-
ematical modelling and engineering technician competencies. This slow and cum-
bersome development process complements Bereiter’s (2002) quest for a long-term
approach to the road of development in education. He pleaded for educational
visionaries rather than early adopters. Visionary quality inspires one to see ways
that an innovation can help to achieve particular long-term goals and not only to be
concerned with short-term gains as far as educational development is concerned.
This study showed that, over a period, continuous exposure to mathematical model-
ling can enrich our students with the needed competencies to enhance their mathe-
matical understanding and reasoning.
References
Anhalt, C. O., & Cortez, R. (2015). Mathematical modeling: A structured process. Mathematics
Teacher, 108(6), 446–452.
Bahmaei, F. (2013). Mathematical modelling in university, advantages and challenges. Journal of
Mathematical Modelling and Application, 1(7), 34–49.
Baker, C. K., & Galanti, T. M. (2017). Integrating STEM in elementary classrooms using
model- eliciting activities: Responsive professional development for mathematics
coaches and teachers. International Journal of STEM Education, 4(1), 10. https://doi.
org/10.1186/540594-017-0066-3.
Bereiter, C. (2002). Design research for sustained innovation. Cognitive Studies, 9(3), 321–327.
Blomhøj, M., & Jensen, T. H. (2007). What’s all the fuss about competencies? In W. Blum,
P. Galbraith, H.-W. Henn, & M. Niss (Eds.), Modelling and applications in mathematics edu-
cation: The 14th ICMI study (pp. 45–56). New York: Springer.
Blum, W., & Leiβ, D. (2005). “Filling up” – The problem of independence-preserving teacher
interventions in lessons with demanding modelling tasks. In M. Bosch (Ed.), Proceedings
CERME4 (pp. 1623–1633). https://www.researchgate.net/publication/46672512. Accessed on
20 May 2016.
Brown, A. L. (1992). Design experiments: Theoretical and methodological challenges in creat-
ing complex interventions in classroom settings. The Journal of the Learning Sciences, 2(2),
141–178.
Chick, H., & Stacey, K. (2013). Teachers of mathematics as problem-solving applied mathemati-
cians. Canadian Journal of Science, Mathematics and Technology Education, 13(2), 121–136.
Cirillo, M., Pelesko, J., Felton-Koestler, M., & Rubel, L. (2016). Perspectives on modeling in
school mathematics. In C. Hirsch & A. McDuffie (Eds.), Mathematical modeling and modeling
mathematics (pp. 3–16). Reston: The National Council of Teachers of Mathematics.
Engeström, Y. (2011). From design experiments to formative interventions. Theory & Psychology,
21(5), 598–628.
Engineering Council of South Africa (ECSA). (2016). Guide to the competency standards for
registration as a professional engineering technician. https://www.ecsa.co.za/register/
Professional%20Engineers/R-08-PN.pdf. Accessed 29 Apr 2016.
Gravemeijer, K., & Cobb, P. (2006). Design research from a learning design perspective. In J. Van
den Akker, K. Gravemeijer, S. McKenney, & N. Nieveen (Eds.), Educational design research
(1st ed., pp. 17–51). London: Routledge.
W. Blum, & S. Kahn (Eds.), Mathematical modelling (ICTMA 12): Education, engineering and
economics (pp. 110–119). Chichester: Horwood.
Kaiser, G., & Brand, S. (2015). Modelling competencies: Past development and further perspec-
tives. In G. A. Stillman, W. Blum, & G. Kaiser (Eds.), Mathematical modelling in educa-
tion research and practice: Cultural, social and cognitive influences (pp. 129–149). Cham:
Springer.
Kelly, A. (2004). Design research in education: Yes, but is it methodological? Journal of the
Learning Sciences, 13(1), 115–128.
Lawless, A. (2005). Numbers and needs: Addressing imbalances in the civil engineering profes-
sion. Halfway House: South African Institution of Civil Engineering (SAICE).
Lesh, R., & Clarke, D. (2000). Formulating operational definitions of desired outcomes of instruc-
tion. In A. Kelly & R. Lesh (Eds.), Handbook of research design in mathematics and science
education (pp. 113–149). Mahwah: Erlbaum.
Mathematical Thinking and Learning, 5(2–3), 157–189.
Lingefjärd, T. (2006). Faces of mathematical modeling. ZDM Mathematics Education,
38(2), 96–112.
113–142.
Maxwell, J. A. (2004). Causal explanation, qualitative research, and scientific inquiry in education.
Educational Researcher, 33(2), 3–11.
Moore, T. J., Miller, R. L., Lesh, R. A., Stohlmann, M. S., & Kim, Y. R. (2013). Modeling in engi-
neering: The role of representational fluency in students’ conceptual understanding. Journal of
Engineering Education, 102(1), 141–178.
Reeves, T. C. (2006). Design research from a technology perspective. In J. Van den Akker,
K. Gravemeijer, S. McKenney, & N. Nieveen (Eds.), Educational design research (1st ed.,
pp. 52–66). London: Routledge.
Schoenfeld, A. H. (1992). Learning to think mathematically: Problem solving, metacognition, and
sense making in mathematics. In D. Grouws (Ed.), Handbook for research on mathematics
teaching and learning (pp. 334–370). New York: Macmillan.
Simon, M. (2014). Hypothetical learning trajectories in mathematics education. In S. Lerman
(Ed.), Encyclopedia of mathematics education (pp. 272–275). Dordrecht: Springer.
Skemp, R. R. (1976). Relational understanding and instrumental understanding. Mathematics
Teaching, 77, 20–26.
Van den Akker, J., Gravemeijer, K., McKenney, S., & Nieveen, N. (2006). Introducing educational
design research. In Educational design research (1st ed., pp. 3–7). London: Routledge.
Wake, G., Swan, M., & Foster, C. (2016). Professional learning through the collaborative design
of problem-solving lessons. Journal of Mathematics Teacher Education, 19(2–3), 243–260.
Woollacott, L. C. (2003). Dealing with under-preparedness in engineering education. Part 1:
Defining the goal. A taxonomy of engineering competency. Proceedings of the 2003 WFEO/
ASEE e-Conference. American Society for Engineering Education. https://hdl.handle.
net/10539/21346
Part III
Pedagogical Issues for Teachers and
Teacher Educators Using Mathematical
Modelling and Applications
Chapter 20
Sense-making in Modelling Tasks: What
Can We Learn from Other Domains?
Piera Biccard
Abstract In this chapter, drawing on scholarly conceptions of sense-making from

domains outside of mathematics education, I will outline ways that mathematical
modelling tasks may meet some of these criteria. Myths around sense-making from
machine and human cognition as well as Weick’s seven elements of organizational
sense-making are presented to extend the discussion around sense-making within
mathematical modelling. An analogy is drawn between ideas around sense-making
from these domains and sense-making in mathematical modelling activities.
Although the discussions around sense-making from these diverse domains do not
match exactly the processes of mathematical modelling, the concepts may extend
our thinking about sense-making and provide us with vocabulary to discuss sense-
making in mathematical modelling.
Keywords Mathematical modelling · Modelling tasks · Sense-making · Sense-

making in mathematical modelling
20.1 Introduction
Sense-making in mathematics is often considered to mean deeper mathematical

understanding or flexible application of mathematical concepts (van Velzen 2016).
The National Council of Teachers of Mathematics (NCTM) has defined sense-
making as “developing understanding of a situation, or context, or concept by con-
necting it with existing knowledge” (NCTM 2009, 4). Indeed, the words
understanding and sense-making are often used interchangeably.
P. Biccard (*)
Department of Curriculum and Instructional Studies, University of South Africa,
Pretoria, South Africa
e-mail: biccap@unisa.ac.za

224 P. Biccard
The notion of sense-making in mathematical modelling specifically may need

elaborating since the competencies required for mathematical modelling (Maaß
2006; Kaiser 2007) are more demanding than solving procedural mathematical
problems. Modelling involves “quantifying, dimensionalizing, coordinatizing, cat-
egorizing, algebratizing and systematizing relevant objects, relationships, actions,
patterns, and regularities” (Lesh and Doerr 2003a, p. 5). In this chapter, the term
sense-making is considered from domains outside of mathematics. This may pro-
vide us with new lenses to view and understand sense-making in mathematical
modelling. Exploring sense-making in these domains may also suggest vocabulary
for discussing and understanding sense-making while modelling. Since “mathemat-
ical modeling authentically connects to the real world, starting with ill-defined,
often messy real-world problems with no unique correct answer” (Cirillo et al.
2016, p. 8), sense-making is a critical feature.
Two fields where sense-making is a prevalent discourse are human and machine
cognition and organizational science. These are set out in Sect. 20.2 while analogies
with mathematical modelling are integrated into the discussion. In Sect. 20.3, the
ideas are mapped onto a generic modelling process where the analogies are further
elaborated.
20.2 Sense-Making Defined by Other Domains
I now look more broadly at the idea of sense-making from other domains and at the
same time hold up the field of educational modelling to the scrutiny of specialized
fields to determine if modelling activities meet these sense-making criteria. The
idea that sense-making may have common ground across domains is presented by
Schoenfeld (2014, p. 405) when he stated that “resources, strategies, metacognition,
and beliefs are generally acknowledged to be essential parts of sense making in
every field”.
20.2.1 Sense-Making in Human and Machine Cognition
Klein, Moon and Hoffman (2006a, p. 71), scholars in the area of human and machine
cognition, provide the following psychological perspective of sense-making:
It is a motivated, continuous effort to understand connections (which can be among people,
places, and events) in order to anticipate their trajectories and act effectively.
Klein et al. (2006a) moved away from conventional conceptions of sense-making

within psychology such as sense-making being equivalent to, or a combination of,
creativity, curiosity, comprehension, mental models and situational awareness.
Sense-making involves all of these but cannot be defined by any one of these ideas.
These authors outline several sense-making functions: it satisfies a need to
20 Sense-making in Modelling Tasks: What Can We Learn from Other Domains? 225
comprehend; it is both retrospective and anticipates future action; it guides the

exploration of information and is often a social activity that facilitates the attain-
ment of common ground. Klein et al. (2006a) also discredit what they termed cer-
tain “myths” regarding sense-making.
Myth 1: Data Fusion Assists Sense-Making
Fused data refers to data that is synthesized or presented in a simpler or integrated
form, making it easily digestible. Klein et al. (2006a) cite research findings assert-
ing that people who are put in passive positions of receiving interpretations are less
able to notice emerging patterns and problems. Kim and Zhong (2017) also report
that structuring information negatively affects flexible thinking and creativity. In
terms of mathematical modelling, activities often require that students analyse and
systematize information before constructing or developing models. Often too much
data, messy data, or the need for generating your own data is required when model-
ling an open, real-world-based problem. It may be through developing models from
unfused data where sense-making emerges since students refine and revisit their
own interpretations of the data.
Myth 2: Sense-Making Is Not Simply About Connecting the Dots
According to Klein et al. (2006a), sense-making, in contrast, is about finding out
what a dot is (what is relevant) in the first place. In mathematical modelling litera-
ture, Bleiler-Baxter et al. (2016) found exactly this during a modelling task. Students
were unable to decide which information given in the problem was relevant. All the
information was treated as equally relevant. The need to “find the signal in the
noise” is a feature of open, reality-based modelling problems. Early assumptions,
narrowing down the information and decisions involved regarding the importance
(and relevance) of the information frame the solution process. Treilibs et al. (1980)
found that weak modellers avoid any decision-making whatsoever. This may be due
to students initially trying to use all the information available, whereas they should
make a judicious selection of important information before they “join the dots”
between the relevant pieces of information.
Myth 3: More Information Leads to Better Sense-Making
Too much information may result in people becoming increasingly overconfident
and not increasingly correct (Klein et al. 2006a). Too much information may result
in “finding” a preconceived answer in a small section of the information and not
considering all the information. When students are modelling, too much informa-
tion may lead to difficulties getting through the modelling cycle (e.g. Blum and
Leiss 2007), especially the initial phases of understanding, simplifying and structur-
ing. There is evidence within the modelling community that too much information
can be problematic with modelling activities (Bleiler-Baxter et al. 2016; Stender
and Kaiser 2016).
Bleiler-Baxter et al. (2016) found that a highly elaborate modelling context dis-
tracted students from simplifying the problem and they could not move forward in
solving the problem. They found students over-focused on the context. These
authors opted to de-contextualise the problem. In contrast, Stender and Kaiser
226 P. Biccard
(2016) opted for an intervention when students could not progress owing to an
undue focus on context. Their intervention was not to adjust the problem or context
but to explain the modelling cycle to the students, which enabled them to make
progress. Reusser and Stebler (1997) already identified this over-focusing on some
aspect of the context as a challenge in word-problem research. For Maaβ (2006),
real contexts should be easily understood, but not too much context is needed. The
challenge for the teacher is balance necessary and sufficient information to maintain
the real-life problem authenticity. Enough context should, however, allow for differ-
ent mathematical entry points into the problem.
Myth 4: Keeping an Open Mind Is Important
In fact, according to Klein et al. (2006a), people who stated and deliberately tested
an early initial hypothesis were the most successful problem solvers – not those who
kept an open mind to all possibilities. On the other hand, the worst performers are
those who jump to an early conclusion and stick to it. When groups work on model-
ling tasks, they may struggle with sense-making if they cannot formulate an early
hypothesis about the connections within the problem. They may spend too much
time elaborating and extending the context or try all possible solution paths. Just
how much of an “open mind” is necessary when modelling is an open question and
may need to be investigated further.
Myth 5: Sense-Making Follows the Form of
Information-Knowledge-Understanding
Sense-making does not simply flow effortlessly from enough information to under-
standing, nor does a certain level of information or knowledge necessarily lead to
sense-making. Lipton (2011) sets out that “any model of understanding that treats
what is necessary for mere knowledge as sufficient for understanding will be inad-
equate” (p. 50). Narrow definitions and conceptions of “understanding” tend to be
linear and unidimensional. Hills et al. (2007) explain that individualized linear
approaches to learning neglect the surrounding contextual situation. Purely behav-
iourist or cognitive approaches may not always provide comprehensive frameworks
for understanding sense-making when students tackle modelling activities. Lesh
and Doerr’s (2003b) publication attests to this as they formulated the term “beyond
constructivism”. They state that conceptual systems are dynamic and cannot be
reduced to “simpleminded checklists of condition-action rules” (p. 534).
If we were to dispel these myths in mathematical modelling, modelling activities
would enhance sense-making if they include unfused data, provide opportunities for
students to sort out relevant from irrelevant information, do not give too much infor-
mation, allow students to formulate an early hypothesis and endorse a non-linear
process in solving the problem (allowing enough time for several modelling cycles).
Klein et al. (2006b) present important information regarding the difference in
expert and novice sense-making strategies. Whereas novices tend to look for any
information that confirms their model, experts may deliberately look for informa-
tion that disconfirms their models. Klein et al. (2006b) propose that the progression
from novice to expert is a “process of learning whereby individual cases, or small,
contextually bound understandings with specific inference possibilities attached to
them, might develop into larger, more organized understandings” (p. 91). Similarly,
in the field of modelling, Lesh and Harel (2003) maintain that the process of devel-
oping powerful conceptual systems is more situated, unstable and multidimensional
than previously thought. By sorting out and making sense of local conceptual ideas
in modelling problems, the larger foundational ideas develop.
20.2.2 Sense-Making in Organization Science
Weick (1995) set out seven characteristics or properties of sense-making. Weick

considered his publication on sense-making in organizations to be a “developing set
of ideas with explanatory possibilities, rather than as a body of knowledge” (1995,
p. xi). I will now use these ideas to identify and extend the discussion on sense-
making in mathematical modelling. This section will try to determine if an analogy
exists and if it is feasible to conclude that modelling encompasses these character-
istics and that modelling can be seen as a sense-making activity.
20.2.2.1 Sense-Making Is Grounded in Identity Construction
According to Weick (1995), “sense-making begins with the sensemaker” (p. 18).
The sensemaker is an individual of many identities – what Weick (1995) called a
parliament of selves. Weick (2001) also stated that who the person is in a setting
provides a base from which he/she makes sense. Lesh and Kaput (2007, p. 299) also
state that (from a problem-solving perspective) successful problem solvers (sense
makers) are those individuals who are able to “adopt a variety of personae”. These
individuals embrace varying stances intuitively, subconsciously and not necessarily
formally. These stances are based on values, attitudes and beliefs. Savolainen (1993,
p.16), writing in the field of information processing, raises some key ideas about the
individual involved in sense-making:
Sense-making is a process; sense is the product of this process. According to Dervin “sense”
includes “knowledge”, but also a host of other subjective factors that reflect an individual’s
interpretation of a situation, including intuitions, opinions, hunches, effective responses,
evaluations, questions, etc”.
Niss (2010) presented four enablers for anticipated implementation of mathemati-

zation in modelling, two relating to personal beliefs: firstly, students need to believe
that modelling is a valid form of mathematics and secondly, they should be persis-
tent and confident. Daher (2015) found that four factors affected students’ position-
ing and emotions while working on modelling: the student’s individual
characteristics, the student’s history of learning experiences, the activity character-
istics and the modelling phases. Furthermore, Brown and Stillman (2017, p. 354), in
their study of year 6 primary school students, examined instances where students
displayed different modelling conceptions regarding problem solving in
228 P. Biccard
mathematics. Drawing on their findings the authors offered that modelling situa-
tions create ideal environments in which to develop a “sense of mathematics as a
way of thinking about life”.
20.2.2.2 Sense-Making Is Retrospective
Remembering and looking back are main sources of meaning-making (Weick 2001).
Reflection plays a critical role in the sense-making process. Traditional approaches
in mathematics classrooms often do not promote reflection as part of the learning
process. Skovsmose (2005) identified disposition, intention and reflection as essen-
tial to meaningful learning in mathematics classrooms generally. Modelling prob-
lems demand a great deal of reflective engagement from students. Due to the cyclical
nature of modelling, reflection is an embedded process and an important compe-
tency. Blum and Borromeo-Ferri (2016) state that student individual solutions,
rather than teachers favouring their own solution (Blum 2015), should be encour-
aged because they are the basis for reflection. Blum and Borromeo-Ferri add that it
is equally important to encourage multiple solutions so that individual thinking
styles are stimulated. Brown (2013) suggests that reflection in real-world modelling
tasks assists students in understanding that they play an important role in sense-
making in mathematics.
20.2.2.3 Sense-Making Involves Enactment
Weick (1995) explains that by acting in a sense-making situation, people create

objects that they project back into the sense-making environment. These objects can
constrain or enhance sense-making. However, making sense is not only about cog-
nitive interpretation; it resembles Bruner’s notion that we enact our way into mean-
ing (Colville et al. 2016). Mathematical modelling presents a highly enactive
learning environment. Students create objects through their words or actions while
solving a modelling task. By enacting, students place these objects back into the
modelling environment they are working in (usually with other students). Enactment
means that individuals shape their learning environment by projecting thoughts and
ideas into the learning space. Tasks that require projections of students’ own ideas
and thoughts support enactment.
Lesh and Doerr (2003a, p. 31) set out that modelling tasks require that students
create “auditable trails of documentation”. These documented objects or products
can take many forms. Smith (1994, p. 25) sets out three types of products that col-
laborating groups produce; these are tangible, intangible and ephemeral products.
The tangible products include the target products (that which students are expected
to produce at the end of the task, such as a model) and instrumental products such
as the data, tables or notes that were produced to support the group’s work. The
intangible products represent the individual thoughts of the group members – some
of which are shared with other members of the group. The ephemeral products are
things such as unrecorded conversations or temporary board sketches – these existed

only briefly but contributed to the other products and are part of the other products.
Lesh and Harel (2003, p. 159) explain that when modelling, students do not produce
simple answers to well-constructed questions – rather students have to produce
mathematical “conceptual tools” to construct, describe or explain a system. These
conceptual tools are enacted and shared with the rest of the group in order to advance
sense-making.
20.2.2.4 Sense-Making Is a Social Process
According to Weick (1995), “Those who forget that sense-making is a social pro-
cess miss a constant substrate that shapes interpretations and interpreting” (p. 39).
Sense-making does not only take place on an individual level but takes place in
social contexts. Learning as socio-constructivism has become a philosophy to guide
many learning theories (Ernest 1994). The view that meaning is a social product
(Yackel 2004) is widely acknowledged in mathematics education. Furthermore,
Vygotsky (1997, p. 49) states that the “social environment is the true level of the
educational process”. More specifically in modelling, Lesh and Doerr (2003b,
p. 525) clarify that “modelling is inherently a social enterprise”. Modelling tasks are
generally designed for groups of students to solve (Blum and Borromeo-Ferri 2016;
Lesh and Doerr 2003a) and by their very nature encourage and necessitate social
learning. Group member suggestions often provide cues for other group members to
act on (Brown and Stillman 2017), whether building on an existing group idea or
diverting group discussion in another direction. When mathematics is taught as an
isolated individual endeavour, students struggle to develop social learning skills
necessary for sense-making in modelling contexts. For Brown (2013, emphasis
added), novice younger modellers in her study “lack[ed] experience of shared math-
ematising and negotiation of understanding” (p. 304).
20.2.2.5 Sense-Making Is Ongoing
According to Weick (1995), sense-making does not have a specific starting point.
The “experience of sense-making is one of which people are thrown into the middle
of things and forced to act without the benefit of a stable sense of what is happen-
ing” (Weick 2001, p. 462). Perhaps similar to what Blomhoj and Jensen (2007,
p. 49) described as a key characteristic of the initial modelling phase as “learning to
cope with the feeling of perplexity due to too many roads to take and no compass
given”. Lesh and Harel (2003, p. 176) described learning as situated and described
the nature of local conceptual development while students are involved in solving
model-eliciting problems. They found that constructs are at some “intermediate
phase of development” and that the levels of development vary across tasks and
across time within a single task. Lesh and Harel (2003) re-iterate that students do
not progress along some ladder of conceptual development, while Lesh and Doerr
230 P. Biccard
(2003a, p. 4) state that understanding is not an “all or nothing” situation explaining

that ideas and constructs for solving real world problems are continually developing.
It may be more useful to consider conceptual development in terms of “net-
works” rather than “ladders”. The ladder metaphor is linear and limiting – students
are not immediately able to apply mathematical learning from one context to another
when taught concepts through atomistic ladder-like lists. A host of modelling litera-
ture attests to the various cycles in modelling (Cirillo et al. 2016) that describe or
predict that student progress is non-linear. Students often move through the model-
ling cycle several times before their ideas are crystallized and before they really
have a full understanding of the problem or the consequences of their choices. They
also learn about how their initial assumptions influence the outcome of the problem.
The same modelling problems can be given to students across a range of ages and
abilities, since student understandings and abilities as well as their mathematical
and modelling competencies are continually evolving.
20.2.2.6 Sense-Making Is Focused on, and by, Extracted Cues
Weick (1995) proposes that cues are “simple, familiar structures that are seeds from
which people develop a larger sense of what may be occurring” (p. 50). Cues help
us to notice things, and noticing is how we “perceive and interpret our way through
the day” and is a “natural part of sense making” (Ball 2011, p. xx). Cues, as sug-
gested by the stage acting industry, help us take action. Each individual notices cues
based on his/her own individual filter. This filter is comprised of his/her own iden-
tity, experiences and orientations.
Schatz and Bartlett (2013) describe intuition as “the unconscious awareness,
valuation, and integration of important cues”. They explain further that intuition “is
the ability to put together cues at a subconscious, nonverbal level, and recognize a
pattern worthy of notice before that pattern can be deliberately perceived” (p. 485).
They extended intuition to intuitive sense-making and describe it as “a conscious
process informed by subconscious intuitive mechanisms and moderated by deliber-
ate metacognitive effort, with the intention of understanding connections, interpret-
ing meaning, and anticipating trajectories, which support later decision making and
possible actions”.
20.2.2.7 Sense-Making Is More About Plausibility than Accuracy
According to Weick (1995), sense-making is driven by plausibility rather than accu-

racy. It is about “plausibility, coherence, and reasonableness” (p. 61). Weick (2001)
elaborated that sense-making is about being certain enough for the present condi-
tions and purposes. Plausible reasoning often “fits the facts, albeit imperfectly at
times”, and “the reasoning is based on incomplete information” (Isenberg 1986,
cited in Weick 1995, p. 56). These ideas around plausible thinking resonate with the
initial modelling phases.
Modelling (especially of more reality-based, open-ended tasks) is driven initially

by plausibility, while accuracy comes in during the latter phases of modelling. The
solutions presented at the end of a modelling task are often varied, based on plau-
sible assumptions that were made in the early phases of solving the task. Modelling
solutions are often a result of various plausible trade-offs between context and accu-
racy. These solutions often “fit well, but not perfectly, their constraints”, and the
mathematical systems students create are not “fully stable or comprehensive”
(Hamilton 2007, p. 5). Polya (1954) considered the differences between demonstra-
tive reasoning and plausible reasoning. The standards of plausible thinking are fluid,
while demonstrative reasoning follows strict rules and conventions. Polya’s plea is
still relevant today: “Certainly, let us learn proving, but also let us learn guessing”
(p. vi).
20.3 Sense-Making and the Modelling Perspective
This section presents a summary and integration of the sense-making ideas (pre-
sented in Sect. 20.2) with mathematical modelling. Although certain analogies were
pointed out in Sect. 20.2, this section synthesizes some of the ideas graphically and
extends the discussion. Figure 20.1 from Yoon and Thompson (2007) sets out the
Fig. 20.1 Modelling process (Yoon and Thompson 2007, p. 206)

232 P. Biccard
Fig. 20.2 Modelling is a sense-making activity. (Adapted from Yoon and Thompson 2007, p. 206)
modelling cycle or modelling process as a series of interpretation activities and

reflections on these interpretations to refine or extend them. Yoon and Thompson’s
exposition of what typically takes place when students are solving a modelling
problem is presented in blocks a to f.
In Fig. 20.2, modelling is shown as a sense-making activity where I show the
integration of Weick’s (1995, 2001) conceptions of sense-making depicted by the
labelled arrows, while Klein et al.’s (2006a) dispelled myths regarding sense-making
are included (in grey) underneath the blocks. The modelling process begins with
students (usually working in groups) interpreting the real situation presented to
them. Modelling tasks are usually complex and are initially driven by understanding
and simplifying (Blum and Leiss 2007). Students will need to bring their individual
prior experiences into the modelling space. Students spend time sorting out relevant
from irrelevant information. Contexts where a balance between complexity and vol-
ume of information are necessary. Weick’s (1995) elements of identity construction,
responding to cues and sense-making as a social process, correspond to this phase

of modelling, while the myths of data fusion and more information is necessary for
sense-making are dispelled.
In the next phase (block b), modellers will notice patterns and details in the real
situation. While students work through structuring the problem or extracting essen-
tials from non-essentials, they will be responding to cues from the task as well as
cues from the other members of the group. Students will also project their own
understanding and ideas into the working space. During these early stages, students
may work with plausible ideas, representations and methods before moving towards
accuracy later in the modelling problem. Weick’s elements of enactment and plau-
sible thinking may begin to feature during this phase.
As students move through the process or modelling cycle, their next step (block
c) is to express the patterns and details in a model of the real situation. In the model-
ling cycle mathematizing is essential at this point. This phase is further character-
ized by working within mathematics itself in order to present a mathematical model
of the real situation. Elements of enactment and social processes are necessary for
the development of the mathematical model of the situation. At this point, students
may be formulating an early hypothesis of the situation and working mathemati-
cally from that hypothesis.
In the next phase (block d), students interpret the mathematical results in terms
of the real problem. Here the group will check if their mathematical results make
sense in the real situation. Borromeo Ferri (2006, p. 93) describes interpretation as
the “transition from mathematical results to real results”. This interpreting phase of
modelling may require that the group comes to a shared understanding of their
results. The interpretation of their mathematical results involves reflective think-
ing – students will need to reflect on their earlier understanding of the problem and
refine their assumptions and the plausibility of their ideas before moving towards a
final model. Weick’s elements of retrospection and enactment are featured during
this phase of the modelling problem. However, the group may be working towards
even greater accuracy if they consider their plausible model to be suitable.
Moving through blocks e and f, the group may not find their model or ideas plau-
sible and may need to reconsider their assumptions and simplification of the prob-
lem before considering a new/revised model. The process of interpreting results in
the real situation will lead to validating their mathematical model. If the group finds
that the model does not solve the task requirements, they will start a new cycle of
modelling where they may find new patterns or details that conflict with their cur-
rent model. According to Yoon and Thompson (2007), “the model then becomes a
lens through which the modeler can re-interpret the situation” (p. 207). The iterative
nature of modelling will require that the students take further action through reflect-
ing on their current model until an equilibrium is reached between the model and
the real-world situation. In these cycles of refinement and re-interpretation, the
iterative nature of authentic problem solving dispels the myth that sense-making is
a linear process progressing from information to knowledge to sense-making (cf.
myth 5 in Sect. 20.2.1).
234 P. Biccard
It is possible to integrate the ideas from “outside” domains into modelling pro-
cesses and cycles to assist us in understanding and appreciating the sense-making
potential and sense-making processes in mathematical modelling. These outside
domains allow us to consider sense-making in modelling more holistically and pro-
vide us with extended vocabulary to discuss sense-making ideas.
20.4 Conclusion
In considering sense-making from other domains, some new perspectives emerge,

while existing perspectives within modelling literature are highlighted. From a pre-
liminary encounter with the sense-making literature, it is possible to attest that
sense-making is a diverse and complex notion. The dispelled myths that are set out
from the field of human and machine cognition present sense-making as an active
endeavour that may be supported by unfused data, opportunities to sort out relevant
from irrelevant information, formulating an early hypothesis and non-linear pro-
cesses. These ideas may allow us to understand sense-making when students are
involved in mathematical modelling, from a different perspective. Weick’s seven
elements of sense-making (although from the field of organizational science sense-
making) allow us a multi-faceted view of the individual involved in the sense-
making process. Sense-making is grounded in identity construction, retrospective,
enactive, social, ongoing, focused on and by extracted cues and driven by plausibil-
ity and coherence. Similar notions can be found in the modelling literature. Although
this cross domain analogy may not be a perfect fit or may not integrate fully with the
modelling processes, it may allow us an extended vocabulary with which to discuss
sense-making in teaching and researching mathematical modelling.
References
Ball, D. L. (2011). Foreword to mathematics teacher noticing (pp. xx–xxiv). London: Routledge.
Bleiler-Baxter, S. K., Barlow, A. T., & Stephens, C. D. (2016). Moving beyond context: Challenges
in modelling instruction. In C. R. Hirsch & A. M. McDuffe (Eds.), Mathematical modeling and
modeling mathematics (pp. 53–64). Reston: NCTM.
Blomhoj, M., & Jensen, T. H. (2007). What’s all the fuss about competencies? In W. Blum,
P. Galbraith, H. Henn, & M. Niss (Eds.), Modelling and applications in mathematics education
(14th ICMI study) (pp. 45–57). New York: Springer.
do? In S. J. Cho (Ed.), The proceedings of ICME 12 (pp. 73–96). Cham: Springer.
Blum, W., & Borromeo-Ferri, R. (2016). Advancing the teaching of mathematical modeling:
Research-based concepts and examples. In C. R. Hirsch & A. M. McDuffe (Eds.), Mathematical
modeling and modeling mathematics (pp. 65–75). Reston: NCTM.
engineering and economics, ICTMA 12 (pp. 222–231). Chichester: Horwood.
Borromeo Ferri, R. (2006). Theoretical and empirical differentiations of phases in the modeling
process. ZDM Mathematics Education, 38(2), 86–95.
Brown, J. P. (2013). Inducting year 6 students into “a culture of mathematising as a practice”. In
Connecting research and practice (pp. 295–306). Dordrecht: Springer.
Brown, J. P., & Stillman, G. A. (2017). Developing the roots of modelling conceptions:
‘Mathematical modelling is the life of the world’. International Journal of Mathematical
Cirillo, M., Pelesko, J. A., Felton-Koestler, M. D., & Rubel, L. (2016). Perspectives on modeling
in school mathematics. In C. R. Hirsch & A. M. McDuffe (Eds.), Mathematical modeling and
modeling mathematics (pp. 3–16). Reston: NCTM.
Colville, I., Pye, A., & Brown, A. D. (2016). Sense-making processes and Weickarious learning.
Management Learning, 47(1), 3–13. https://doi.org/10.1177/1350507615616542.
Daher, W. (2015). Discursive positioning’s and emotions in modelling activities. International
Journal of Mathematical Education in Science and Technology, 46(8), 1149–1164.
Ernest, P. (1994). The philosophy of mathematics and the didactics of mathematics. In R. Biehler,
R. W. Scholz, R. Strasser, & B. Winkelmann (Eds.), Didactics of mathematics as a scientific
discipline (pp. 335–349). Dordrecht: Kluwer Academic Publishers.
Hamilton, E. (2007). What changes are occurring in the king of problem-solving situations where
mathematical thinking is needed beyond school? In R. A. Lesh, E. Hamilton, & J. J. Kaput
(Eds.), Foundations for the future in mathematics education (pp. 1–6). Mahwah: Lawrence
Erlbaum.
Hills, T., Hurford, A. C., Stroup, W. M., & Lesh, R. (2007). Formalizing learning as a complex
system: Scale-invariant power law distributions in group and individual decision making. In
R. A. Lesh, E. Hamilton, & J. J. Kaput (Eds.), Foundations for the future in mathematics educa-
tion (pp. 225–244). Mahwah: Lawrence Erlbaum.
W. Blum, & S. Khan (Eds.), Mathematical modelling: Education, engineering and economics,
ICTMA 12 (pp. 110–119). London: Woodmead.
Kim, Y. J., & Zhong, C.-B. (2017). Ideas rise from chaos: Information structure and creativity.
Organizational Behavior and Human Decision Making Processes, 138, 15–27.
Klein, G., Moon, B., & Hoffer, R. R. (2006a). Making sense of sense-making 1: Alternative per-
spectives. IEEE Intelligent Systems, 21(4), 70–73.
Klein, G., Moon, B., & Hoffer, R. R. (2006b). Making sense of sense-making 2: A macrocognitive
model. IEEE Intelligent Systems, 21(5), 88–92.
Lesh, R., & Doerr, H. M. (2003a). Foundations of a models and modeling perspective on math-
ematics teaching, learning, and problem solving. In R. Lesh & H. M. Doerr (Eds.), Beyond
constructivism: Models and modeling perspectives on mathematics problem solving, learning
and teaching (pp. 3–33). Mahwah: Erlbaum.
Lesh, R., & Doerr, H. M. (2003b). In what ways does a models and modelling perspective move
beyond constructivism? In R. Lesh & H. M. Doerr (Eds.), Beyond constructivism: Models and
modeling perspectives on mathematics problem solving, learning and teaching (pp. 519–556).
Mahwah: Erlbaum.
Mathematical Thinking and Learning, 5(2&3), 157–189.
Lesh, R., & Kaput, J. (2007). What kind of instructional activities are needed to support the devel-
opment of new levels and types of understanding and ability? In R. A. Lesh, E. Hamilton,
& J. J. Kaput (Eds.), Foundations for the future in mathematics education (pp. 297–299).
Mahwah: Erlbaum.
Lipton, P. (2011). Mathematical understanding. In J. Polkinghorne (Ed.), Meaning in mathematics
(pp. 49–54). Oxford: Oxford University Press.
Maaß, K. (2006). What are modeling competencies? ZDM Mathematics Education., 38(2),
113–142.
236 P. Biccard
National Council of Teachers of Mathematics (NCTM). (2009). Focus in high school mathematics:
Reasoning and sense making. Reston: NCTM.
Niss, M. (2010). Modeling a crucial aspect of students’ mathematical modeling. In R. Lesh,
P. Galbraith, C. R. Haines, & A. Hurford (Eds.), Modeling students’ mathematical competen-
cies (pp. 43–59). New York: Springer.
Polya, G. (1954). Mathematics and plausible reasoning (Vol. 1). Princeton: Princeton
University Press.
Reusser, K., & Stebler, R. (1997). Every word problem has a solution – The social rationality of
mathematical modeling in schools. Learning and Instruction, 7(4), 309–329.
Savolainen, R. (1993). The sense-making theory: Reviewing the interests of a user-centered
approach to information seeking and use. Information Processing & Management, 29(1), 13–28.
Schatz, S., & Bartlett, K. (2013). Modeling cues for intuitive sense-making simulations.
In D. D. Schmorrow & C. M. Fidopiastis (Eds.), Foundations of augmented cognition
(pp. 484–491). Berlin: Springer-Verlag.
Schoenfeld, A. H. (2014). What makes for powerful classrooms, and how can we support teach-
ers in creating them? A story of research and practice, productively intertwined. Educational
Researcher, 43(8), 404–412.
Stender, P., & Kaiser, G. (2016). Fostering modelling competencies in complex situations.
In C. R. Hirsch (Ed.), Mathematical modeling and modeling mathematics (pp. 107–116).
Reston: NCTM.
Skovsmose, O. (2005). Meaning in mathematics education. In J. Kilpatrick, C. Hoyles,
O. Skovsmose, & P. Valero (Eds.), Meaning in mathematics education (pp. 83–104). New York:
Springer Science and Business.
Smith, J. B. (1994). Collective intelligence in computer-based collaboration. Hillside: Erlbaum.
Treilibs, V., Burkhardt, H., & Low, B. (1980). Formulation processes in mathematical modelling.
Nottingham: Shell Centre for Mathematical Education.
van Velzen, J. H. (2016). Evaluating the suitability of mathematical thinking problems for senior
high-school students by including mathematical sense making and global planning. The
Curriculum Journal, 27(3), 313–329.
Vygotsky, L. S. (1997). Educational Psychology. Florida: CRC Press.
Weick, K. E. (1995). Sense-making in organizations. London: Sage.
Weick, K. E. (2001). Making sense of the organization. Malden: Wiley-Blackwell.
Yackel, E. (2004). Theoretical perspectives for analyzing explanation, justification and argumen-
tation in mathematics classrooms. Journal of the Korea Society of Mathematical Education
Series D: Research in Mathematical Education, 8(1), 1–18.
Yoon, C., & Thompson, M. (2007). Cultivating modeling abilities. In R. A. Lesh, E. Hamilton,
& J. J. Kaput (Eds.), Foundations for the future in mathematics education (pp. 201–210).
Mahwah: Erlbaum.
Chapter 21
The Deployment of Mathematics in Areas
Other than Those Normally Associated
with Mathematical Modelling
and the Applications of Mathematics
Cyril Julie
Abstract The extra-mathematical domains that feature prominently for mathemat-

ical modelling and the applications of mathematics in ICTMA deliberations are the
physical, biological, economic, environmental, social and engineering sciences.
This chapter focuses on the deployment of mathematics in domains other than these
dominant ones. Sources such as novels, stories, monographs, journalistic-like com-
mentaries on mathematics, anecdotal accounts on mathematics by ordinary people
and so forth were mined to ascertain the uses mathematics is put to in such sources.
The uses are classified as backgrounding storylines through encryption and decryp-
tion; depictions of cleverness, dissent/avoidance and madness; rendering political
commentary; and entertainment through humour. It is suggested that students
should be afforded opportunities to put mathematics to use in a similar manner.
Keywords Applications · Cryptography · Cryptosystems · Fiction · Cleverness ·

Mathematical modelling · Mathematics “just for laughs”
21.1 Introduction
Aris (1978) captures the extra-mathematical domains that are dominant in mathe-
matical applications and modelling by articulating mathematical modelling as a
“complete and consistent set of mathematical equations which is thought to corre-
spond to some other entity, its prototype. The prototype may be a physical, biologi-
cal, social, psychological or a conceptual entity, perhaps even another mathematical
C. Julie (*)
School of Science and Mathematics Education, University of the Western Cape,
Bellville, South Africa
e-mail: cjulie@uwc.ac.za

238 C. Julie
model…” (p. 1). This raises the question of whether there are domains and areas of
human engagement other than those mentioned by Aris where mathematics is also
deployed. In this chapter, this issue is pursued by focusing on sources such as nov-
els, stories and oral accounts linked to aspects of mathematics. In order to distin-
guish between the domains at play in “pure” mathematical applications and
modelling and those of import in this chapter, the constructs of “uses of mathemat-
ics” and “mathematics put to use” are employed as organising devices. The con-
struct “uses of mathematics” pertains to extra-mathematical domains as normally
perceived in mathematical applications and modelling (see, e.g., Niss et al. 2007).
“Mathematics put to use” concerns certain other domains and areas in daily life.
This chapter proceeds in five sections. The next section discusses the making
visible mathematical dimensions in daily life activities in order to demonstrate the
“uses of mathematics” derived from ordinary activities. It includes a brief exposi-
tion of the purposes of mathematical applications and modelling and situates the
making visible examples in the section one of these purposes. The section closes
with making a distinction between making visible mathematical dimensions in
daily activities and the notion of “mathematics put to use” as utilised in this chapter.
The third section provides the data sources, the data analysis and the approach that
was followed. This is followed by a description of the themes forthcoming from the
analysis. The chapter concludes with a discussion of possible implications of allow-
ing students to engage with activities around “mathematics put to use”.
21.2 Making Visible Mathematics in Daily Activities
In reflecting on making mathematics accessible to as many school students as pos-

sible, Alsina (2002) suggests that the quest should be “to make visible mathematical
dimensions of our daily life on this planet” (p. 249). He exemplifies this making
visible with issues, such as staircases, from “our daily life” as follows:
…all humans need stairs which are easy to climb and almost all human beings use shoes.
Stairs are important objects. Measure them! (e.g. using electronic measurements). The
ideal steps have two important measurements: H (height) and D (depth), related by the
affine equation 2H+D = 63 cm. The inclination tan A = H:D is also interesting. What are
the upper and lower bounds for H, D and A? When is it convenient to have a ramp and not
a stair? In vertical ladders (like in submarines) you face the steps to go down, but in normal
stairs you come down the other way around: when is it better to face the steps?
(pp. 247–248)
Mathematics trails are also concerned with making visible “mathematical dimen-
sions”. A mathematics trail is defined as “a walk to discover mathematics [in] a
neighborhood, a business district or shopping mall, a park, a zoo, a library, even a
government building…where walkers formulate, discuss, and solve interesting
mathematical problems” (Shoaf et al. 2004, p. 6). In order to draw attention to
21 The Deployment of Mathematics in Areas Other than Those Normally Associated… 239
mathematics, mathematical activities which students have to do while on the walk

are provided. Shoaf et al. (2004) provide such an activity for students to do. They
are instructed to do the Slides activity (Shoaf et al. 2004, p. 20) when they reach a
play park on a certain trail route.
SLIDES
Slides are loads of fun for everyone! We never seem to outgrow the thrill of
coming down a slide – the steeper and faster, the more fun! One of the things
you learn in mathematics has to do with the slopes of lines. We can combine
the fun and thrill of a slide with the concept of the slope of the slide. Once the
walkers are comfortable finding the slope of the first slide, it is a good idea to
have them find other slides or ramps in the playground and compare
their slopes.
GUIDE TO STOP 6 – Straight Slides
Now measure and record how far it is along the ground from the tallest point.
Measure and record the height of the slide at its tallest point and at its lowest
point. Subtract the lowest from the highest and record your results in the
numerator of the fraction below.
Heightat tallest point − Heightat lowest point
Distance from tallest point toendof slide.
Although somewhat different, ethnomathematics can also be viewed as an

approach linked to the making visible “mathematical dimensions”. Gerdes’s ethno-
mathematical project (Gerdes 1988), for example, concentrated on the techniques
used by Mozambican peasants to construct artefacts such as baskets. He observed
peasants practising their craft and engaged in constructing the artefacts himself. A
diagram with mathematical annotations (Fig. 21.1), such as for the weaving of
“Litenga” baskets is shown. Mathematical questions related to the diagrammatic
Fig. 21.1 Diagrammatic

representation of part of
“Litenga” (cf. Gerdes
1988, p. 155)
240 C. Julie
Fig. 21.2 Graphical

representation of function
related to basket weaving
(cf. Gerdes 1988, p. 155)
representation are then pursued such as “What happens when the ‘horizontal’ and
‘standing’ strips are of different width, e.g. 1 (unit of measurement) and a?” (Gerdes
1988, p. 155).
Further mathematical issues are investigated and brought to the foreground to
make visible the mathematical dimensions. Figure 21.2 illustrates the graphical rep-
1
resentation of the function hex 1 a hex 1 a (Gerdes’s symbolism
2 cos
used for the function) for investigating the question: “How does the optimal angle α
depend on a?”.
These three examples of making “visible mathematical dimensions of our daily
life” demonstrate that the objective is to bring to the foreground mathematics and
mathematical objects in artefacts and activities in daily life. The artefacts and activi-
ties that are harvested for putting a mathematical gaze over them are not always
those which fall within the normal visible realm of domains which form the extra-
mathematical domains for mathematical modelling and applications of mathemat-
ics. However, they fall under the “uses of mathematics”. In terms of the purposes of
mathematical applications and modelling, they fall under desciptive purpose of the
three purposes (to describe, to predict or to presecribe) given for mathematical
applications and modelling. The three purposes are exemplified in the next three
paragraphs.
Davis and Hersh (1986) posit that mathematical applications and models “can
serve to describe, to predict, or to prescribe. These modes are interrelated, but they
are not identical” (p. 115). According to these authors, mathematical descriptions of
phenomena are arrived at by “focusing on a sufficiently limited, sufficiently
abstracted aspect of [a] phenomenon, we are able to summarize and replace that
aspect by an equivalent mathematical description” (p. 116). The vertical distance of
the trajectory of a projectile, for example, under certain considerations, can be
1
mathematically described by v Vy 0 t gt 2 with Vy0 the initial velocity along the
2
y-axis, t the time and g the gravitational constant.
Some mathematical models which have a descriptive purpose can also be used to
make predictions. In some sense, most descriptive models are constructed for pre-
dictive purposes; but it should be borne in mind that not all descriptive models could
be used for predictive purposes. However, the quadratic model given above for the
vertical distance of the trajectory of a projectile can be used to predict the vertical
distance of the projectile for a given time.
The prescriptive uses of mathematical models and applications are those that
lead to “human action or automatically to some sort of technological action” (Davis
and Hersh 1986, p. 120). The body mass index (BMI) and Gini coefficient of income
inequality are given as examples entailing prescriptive modelling by Niss (2015).
Davis and Hersh (1986, p. 121) refer to “…life insurance…proportional representa-
tion voting [systems]…school rating systems” (p. 121) as examples and assert that
prescriptive mathematisations “regulate and alter our lives and characterize our civi-
lization” (p. 121). Niss (2015), however, argues that although prescriptive model-
ling shares features with descriptive modelling, prescriptive models are, in principle,
not amenable to validation and falsification.
This chapter is also concerned with activities in daily life outside the realm of
these dominant domains. The spark for doing what is done in this chapter was
ignited by Lave’s (1991) method of researching and theorising about learning
through using empirical sites outside of normal school practice. For this chapter, the
activities are leisure-like ones such as recreational reading and telling and listening
to stories. It anchors around how “mathematics is put to use” in these activities by
considering texts and oral accounts, used in these leisure-like activities. The issue
being pursued is thus around the question: “How and for what purposes is mathe-
matics deployed in these sources?”
21.3 A Note on Texts Used, Selection of Sources and Method
In this section, a brief exposition is provided for the selection of sources and the
method used for analysis. This chapter is about texts, in the form of books and oral
stories, used for recreational and relaxation purposes. These texts are different from
the usual variety of mathematical texts. Table 21.1 below gives a typology of
Table 21.1 Typology of mathematical texts

Category of text Example Audience
Book: Expository Aris, R. (1978). Mathematical modelling techniques. Area
without exercises Marshfield: Pitman specialists
Book: Expository with Giordano, F.R., & Weir, M. D. (1985). A first course in Students
exercises (textbook) mathematical modelling. Monterey: Brooks/Cole
Journal article Herrero, M. A., & Velazquez, J.J.L. (1997). On the Area
melting of ice balls. SIAM Journal on Mathematical specialists
Analysis, 28(1), 1–32
Popular texts Paulos, J. A. (1996). A mathematician reads the General
newspaper. London: Penguin public
242 C. Julie
mathematical texts. Mathematical texts are those that are classified in the
Mathematics Subject Classification (MSC).
A distinguishing feature of a mathematical text is that when it is picked up and
paged through there is no uncertainty that it is about mathematics and primarily for
those studying mathematics. The objects in the text are undeniably mathematical
ones such as symbols, mathematics specific terminology and so forth. Further, in
terms of the bipartite demarcation of texts as fiction and non-fiction, mathematical
texts are non-fiction.
This specification of mathematical texts is to differentiate it from the kind of
texts forming the basis of this chapter. The written texts of import used in this chap-
ter are primarily fiction except for the monograph of an inaugral professorial address
referred to below.
The books were not purposefully selected. They are those that I have recently
read mostly for recreational and relaxation purposes. Also included are texts I have
encountered in my school-going days that could be recalled. In all of them reference
is made in one way or another to mathematics and mathematical workers or the
mathematical topics could be inferred. Books with mathematics-like titles but not
referencing mathematics further were not included. Furthermore, books with a
mathematical instructional slant in the sense of merely explaining and exemplifying
a mathematical idea without focusing on “mathematics put to use” such as the one
by Jacobs (2013) were excluded. The monograph is the inaugral professorial address
of a professor in Applied Mathematics which was delivered in 1974. In addition to
addressing issues in the discipline of the deliverer of an inaugral professorial lec-
ture, there is normally a section of the address to which the non-specialist members
of the audience can relate. My interest in this address was aroused by a curiosity to
find out whether applied mathematicians were also addressing political issues in
public presentations. The books used are all written in English due to my being
conversant only with English and my mother tongue, Kaaps, a langauge spoken
primarily in the environs of Cape Town which allows me to engage with other lan-
guages such as Dutch, Flemish and Afrikaans. The anecdotal stories are those with
some reference to mathematics I have encountered in conversations with colleagues,
students and “just plain folks”. They were communicated in Kaaps and English.
In essence the method I have followed is recalling what I have read, inspection-
ally re-reading it, recalling stories that I encountered and recounting them in my
mind and then creating categories around the question of interest – what use is
mathematics put to here? “Mathematics put to use” is analogous to the way authors
of fiction metaphorically use situations in their stories. For example, in his Wizard
of the crow, Ngũgĩ wa Thiong’o (2006) uses the biblical narrative of Jacob’s ladder,
albeit the building of a staircase to heaven, as background for dealing with corrup-
tion in some fictious country. In this chapter, thus, mathematics is not in the fore-
ground but in the background, or it can be inferred from the trend of the story.
Through the process discussed above, four broad overlapping categories of
“mathematics put to use” were identified: (1) backgrounding storylines through
encryption and decryption; (2) depictions of cleverness, dissent/avoidance and
madness; (3) rendering political commentary; and (4) entertainment through

humour. These are discussed in the next section. As with all category creations, they
are coloured by the lens of the categoriser. This is particularly the case when one
deals with fiction texts because one is, in the words of Paulo Freire, “reading the
word and reading the world” (see, e.g., Freire and Macedo 1987). Others, such as
those involved in literary critiques, might come up with totally different categories.
This reading is coloured by the life history and experiences of the reader of the texts.
Regarding quotations, fairly long ones are used in order for the reader to get the gist
of the “evidence”.
21.4 ategories of “Mathematics Put to Use” in Primarily

C
Non-mathematical Texts and Stories
Four categories of “mathematics put to use” were derived through the process
described above. These are discussed below.
21.4.1 ackgrounding Storylines Through Encryption

B
and Decryption
The mathematical constructs of encryption and decryption have been used in many
stories and novels. The first encounter I had with encryption and decryption was at
high school where we studied a poem from Hornblower and the Widow McCool
(Forester 1951) [I am using a date obtained from the Internet, which is obviously
not the publication we used at school many, many years ago] in our English lan-
gauge class. The short story deals with B. I. McCool who wanted to start a revolu-
tion, but he was executed at sea. His chest had his name on its lid. He wrote a poem
on the lid of the chest which had to be given to his wife after his execution. The
poem is:
Ye heavenly powers! Stand by me when I die!
The bee ascends before my rolling eye.
Life still goes on within the heartless town.
Dark forces claim my soul. So strike ‘em down.
The sea will rise, the sea will fall. So turn
Full circle. Turn again. And then will burn
The lambent flames while hell will lift its head.
So pray for me while I am numbered with the dead.
The teacher told us that there was a message contained in the poem and we had
to unravel it. After some initial struggling, she informed us that we should not read
it literally but try to find the instructions to open the chest by concentrating on some
letters indicated by some of the words. Eventually, the class figured out the hidden
244 C. Julie
instructions and the teacher gave the thrust of the poem. (See Hornblower.wikia.
com/wiki/Hornblower_and_the_Widow_McCool.) Mathematics is not explicitly
mentioned in the poem or the story. Encryption and decryption in this short story are
merely in the background.
More explicit “mathematics put to use” is found in Stieg Larsen’s Millenium tril-
ogy featuring Lisbeth Salander, a computer hacker, and Mikael Blomkvist, a jour-
nalist. David Lagercrantz (2015) continued the series with a fourth book, The girl in
the spider’s web. This crime fiction novel is about cyber crime and in the book fre-
quent reference is made to mathematics related to encryption and decryption.
Examples of this referencing of mathematics are:
They were staggeringly large numbers which formed a familiar pattern alongside the num-
bers next to them. As she looked through the papers and came across the simple sequence
641, 647, 653 and 659, there was no longer any doubt: they were sexy prime quadruplets,
sexy in the sense that they differ from each other by six.
There were also twin primes, and every other imaginable combinations of prime num-
bers. She could not help but smile. “Awesome”. (Lagercrantz 2015, p. 279)
…in which a pair of autistic and mentally disabled twins recite staggeringly high prime
numbers to each other, as if they could see them before their eyes in some sort of inner
mathematical landscape.
What these twins were able to do and what Salander now wanted to achieve were two
different things. But there was still a similarity, she throught, and decided to try, however
sceptical she might be. So she brought up the encrypted N.S.A. file and her program for
elliptic-curve factorization. (Lagercrantz 2015, p. 295)
In his acknowledgements, Lagercrantz thanks a security researcher at Kapersky

Lab in Sweden which deals with the development of anti-virus software and a pro-
fessor of Mathematics at Uppsala University. This is ostensibly done due to them
either informing or checking the validity of the mathematics. The mathematics
focuses on primes and elliptic-curve factorization, which are related topics in cryp-
tography and the quest to keep information secure through encryption. In fact the
Rivest, Shamir and Adleman (RSA) system for encryption and decryption of infor-
mation uses the product of large prime numbers. Elliptical curve factorization is
also associated with cryptosystems (Koblitz 1987). However, Lagercrantz uses
encryption as a background device in his storyline to, in some sense, create sus-
pense – the essence of crime novels – and not to foreground mathematics.
I cannot resist in mentioning Roald Dahl’s The BFG (Big Friendly Giant) (2013).
Dahl’s playing around with words and the creation of new words are well known.
The encryption part is that some words have a pig Latin – a secret language formed
from English by transferring the initial consonant or consonant cluster of each word
to the end of the word and adding a vocalic syllable (usually er) (online diction-
ary) – flavour. In speaking to Sophie, one of the main characters in the story, about
her parents, the BFG says “By now they must be jipping and skumping all over the
house…” (p. 30). Here Dahl interchanges the “sk” from “skipping” to replace the
“j” from “jumping” and vice versa to create words which the reader must decode. In
a similar way when the BFG was informing Sophie about a book (Nicholas Nickleby)
he is reading, he said “Dahl’s Chicken” (Dahl, p. 105) to refer to the author of the
book. Taking some poetic licence, the “Ch” of Charles replaced the “D” of Dickens
and vice versa. There is also the more elementary coding of shifting letters in the
same word such as in “porteedo” for “torpedo” (Dahl 2013, p. 170). Dahl is obvi-
ously not consciously backgrounding the BFG with encryption, but this can be
inferred from his creation of his language, Gobblefunk. I am still struggling to
decode “rotrasper”. As is common in expository mathematical texts, it might be
obvious and I leave it to the reader to decode.
21.4.2 Depictions of Cleverness, Avoidance and Madness
Cleverness is often associated with being adept at mathematics. When I was at

school, if you were “good” at mathematics in terms of scoring high marks in tests,
you were called a “slimmetjie” (clever one). As an aside, in the context of gang
membership, being a “slimmetjie” can land you in trouble. In this context, a “slim-
metjie” is deemed to be “too big for your boots” by being argumentative and want-
ing to rebel against gang rules and codes of behaviour. Coming back to a “slimmetjie”
in mathematics, it is widely reported that people in general are not ashamed to admit
that they were not good at mathematics at school. Similar statements about “not
being good at subject X at school” are rare for other school subjects in normal con-
versations. In this sense mathematics bestows a kind of status of cleverness on those
who excel at it.
Not only is mathematics a signifier of cleverness. Some novelists make use of
mathematical workers in the background to depict both madness and cleverness.
This is the case in The inside-out man (Strydom 2017). Strydom writes:
I did, however, once go into my neighbour’s apartment. He’d knock on my door to ask if I
could fix his radio. An old man with a balding head and long grey strands that fell to his
shoulders, he was fidgety and frustrated, barely acknowledging me even as he asked his
favour. I soon realised that his problems were graver than chronic restlessness.
The radio itself seemed to be working fine, and I turned the knob, from station to station,
to show him. He insisted it wasn’t. He said he needed the numbers from the voice on the
radio, and they weren’t just coming any more. He’d scrawled hundreds of these numbers on
his walls, refrigerator, and even on the grubby brown upholstery of his couch. But he
couldn’t find the pattern if the numbers stopped coming. Without the numbers there was
nothing, he said. No meaning behind any of it. Without the numbers there was only chaos
and madness…and he threw me out of his apartment…
It was only much later that I learnt that he’d once been the Head of Applied Mathematics
at the university—a genius in the field of graph theory. An agoraphobic schizophrenic, but
a genius nonetheless. (pp. 16–17)
The novel is not about the applied mathematician. He virtually disappears after this
referencing to him. The passage brings forth the dual nature of the depiction of
cleverness and madness of mathematical workers and how it is deployed in the
novel to provide a backdrop for the storyline.
School children sometimes invent uncanny methods to express their feelings
about mathematics. A teacher relayed the following story. For the mid-year
246 C. Julie
mathematics examination, one of her pupils submitted his examination script. On

the front page of the four-page examination book, the subject name (Mathematics,
of course), the pupil’s name, his school class code and nothing more were written.
On the second and third pages, nothing were written. On the last page, there was a
drawing of a rock and an arrow probably pointing to something underneath the rock.
In the call-out at the start of the arrow was written “Miss, all the answers are under
the rock”. My reading of this story is that the pupil was messaging his plea of want-
ing to avoid mathematics, albeit in a somewhat creative way.
21.4.3 olitical Commentary Related to Inequality

P
in Social Arrangements
The interesting book, Flatland (Square 1884), was my first encounter with a novel
in which mathematical objects are anthropomorphized. A. N. Square was a pseud-
onym, and in later editions of the book, the “real” name, Edwin A. Abbott, is given
as the author. An animated movie based on Flatland was released in 2007.
In part I of the book, plane figures are the inhabitants of the plane stated as:
Imagine a vast sheet of paper on which straight Lines, Triangles, Squares, Pentagons,
Hexagons and other figures instead of remaining fixed in the places, move freely about, on
or in the surface, without rising above or sinking below it… (p. 4)
Women in this society are lines. The structure of the rest of society in Flatland is as
given in Table 21.2.
There are rules for sons moving up the social ladder mostly through birth, but for
the soldiers and lowest class of workmen, it can only happen
after a long series of military successes, or dilligent and skilful labours…that the Artisan
and soldier classes manifest a slight increase of their third side or base, and a shrinkage of
the other two sides. Intermarriages (arranged by the Priests) between the sons and daughters
of these more intellectual members of the lower classes generally result in an offspring
approximating still more to the type of the Equal-sided Triangle. (Square 1884, p. 9)
Many later editions of Flatland were published, and some of these had explana-
tory notes as part of the edition. Abbott (2008) is such an edition with notes and
comments written by Ian Stewart. Stewart writes “Just below the surface…it
Table 21.2 Social structure of Flatlanders

Social position in society Two-dimensional shape
Soldiers and lowest class of Isosceles triangles
workmen
Middle class Equilateral triangles
Professional men and gentlemen Squares and pentagons
Nobility Hexagons rising in the number of their until they receive the
title polygonal
Priestly class Circles
[Flatland] is a biting satire on Victorian values—especially as regards women and

social status…” (Abbott, p. ix). This brings to the fore the deployment of mathemat-
ics to render commentary on social arrangements in nineteenth-century England.
In his inaugural lecture as professor of Applied Mathematics at the University of
Cape Town, Ellis (1974), in addition to rendering a panoramic view of areas applied
mathematicians deal with, including his own research in cosmology, refers to a
“statement of values…which should be taken seriously” (p. 7) by all academics. This
he insists should “reflect a concern for the conditions of life that surround us” (p. 9).
He engages with this issue in the final part of his lecture by dealing with a model
derived from data related to African people who were prosecuted for “transgressing”
the draconian pass laws – people having to carry permits to be in certain areas – of
the white minority rule era in South Africa. The graph of the model is given below
in Fig. 21.3. He calls this “the plight of the urban African [and] how he (the African)
lives his life being continually reminded that although his work is needed, his exis-
tence is resented and discouraged: in fact his life is made extremely difficult” (p. 15).
He concludes the inaugural address with “…the only true basis for freedom is a
realistic vision of alternative possibilities before us. Mathematical studies can
sometimes help us in understanding what these alternative possibilities are. But
Fig. 21.3 Exponential

growth of the number of
pass law prosecutions per
years (thousand of
individuals) (Ellis
1974, p. 14)
248 C. Julie
such an understanding is quite valueless unless it affects our actions. An understand-

ing of any social wrong, which does not lead to some corrective action to right the
wrong, is meaningless” (p. 17). In the closing sections of his address, Ellis thus not
only brings to the foreground political commentary on an unsavoury aspect of South
African life during a particular historical period but also calls for action to address
inequalities in societal arrangements. In this sense, mathematics is shifted to the
background.
21.4.4 Entertainment
Instances of “mathematics put to use” related to “just for laughs” abound. These
instances of the deployment of mathematics “just for laughs” are normally, in my
opinion, meant for “laughing with” and “laughing at” mathematics. A Google
search on “mathematical cartoons” resulted in more than three million hits.
An example I frequently use with students and teachers is a somewhat naughty
one which I picked up in a Dutch journal for teachers when I browsed through jour-
nals at the Freudenthal Institute some years ago. I was not very mindful of
Freudethal’s advice of writing down the source so it is absent from the references.
In any case the issue deals with the familiar parent/child’s age word problems
accompanying the solutions of simultaneous linear equations. The problem goes as
follows: Currently a father is 25 years older than his child. In 7 years he will be five
times as old as his child. The punchline lies in the problem statement: What is the
father doing now?
Coming nearer to home and to visiting vineyards, teachers were requested to
bring to a continuing development activity episodes of people using mathematics in
everyday activities. A teacher, whose brother is a wine farmer, brought the fol-
lowing one:
Farmworkers have their own language and ways of measuring when they must plant new
vines. Getting their tools and stuff ready the foreman would shout: Bring net die Engelse
grawe (Bring only the English spades.) At the site for digging the holes he would instruct
the others: Spit tot by Engeland (Dig until you come to England!)
Actually they must dig until they reach the “Made in England” mark on the handle
which is the measurement of the depth of the hole.
The “just for laughs” idea can also be taken to words. An endearing word in
South Africa is “gatvol” (ghut-foll). Although there are attempts to give English
equivalents such as “fed up”, “discouraged” and “dissatisfied” to it, it is generally
not translated in South Africa. “Gat” has the connotation of “hole” and is also used
for “arse”. Playing with these connotations in a presentation to teachers at a certifi-
cate award ceremony had them in stiches. The Sierpinski carpet was depicted as the
x2 2 x 3
smart “gat”; the discontinuity of as a “gat” in the graph and the Big
x 1
Hole, an open-pit and underground mine in Kimberley, South Africa, was linked to
an ellipse and finally a rugby ball.
Humorous video clips featuring mathematics are also around. The better known
one is 25 ÷ 5 = 14 (https://www.youtube.com/watch?v=t8XMeocLflc). The inten-
tion of “just for laughs” making use of mathematics is not to foreground the math-
ematics but to entertain.
21.5 Concluding Remarks
In this chapter, it was demonstrated how “mathematics is put to use” in areas not
normally used in mathematical applications and modelling. Fictional stories and
novels, a monograph of an inaugural professorial lecture and oral accounts linked to
the mathematical education enterprise were mined to ascertain how “mathematics is
put to use” in these sources. Four broad themes were identified. In these “mathemat-
ics is put to use” for backgrounding story lines; caricaturing mathematical workers
as clever, mad or both; rendering political commentary; and entertaining through
humour. It is clear that this deployment of mathematics is different from the pur-
poses of description, prediction and prescription in mathematical applications and
modelling posited by Davis and Hersh (1986).
Driven by Noddings’s (1993, p. 159) suggestion that “Teachers should attend to
generative themes…and explore how mathematics is and is not involved in them,”
albeit that her suggestion is about the politicisation of mathematics classrooms, my
contention is that the kind of “mathematics put to use” revealed in this chapter
should be considered for incorporation in the repertoire of engagements students
have with mathematics. It is well known that not all students are equally interested
in mathematics and by implication mathematical applications and modelling.
Students do mathematics for different reasons. Some of these reasons resonate with
Feynman’s (1998, p. xxv) observation regarding students studying physics that
“…many of the students in our course study physics as a background for their pri-
mary interest in other fields”. Similarly, school students are not equally interested or
have the same motives for doing mathematical applications and modelling. Drawing
on Miller’s (1983) stratified model for students’ interest in scientific literacy, I con-
structed a similar configuration for students’ interest doing school mathematics
(Julie 2012). The configuration is given in Fig 21.4.
For learners at the bottom of the configuration, it does not mean that they are not
interested in other domains or school subjects. My modest proposal is that students
should be afforded opportunities to engage in activities in “mathematics put to use”
of the kind of leisure activities mentioned above and others that might emerge. This
might allow them to creatively bring to the fore their talents with issues that interest
them in imaginative ways. For many school children, especially those from socially
and economically depressed environments, schools are most of the time the only
places where they can productively unleash their hidden talents. It might just be that
engaging them in schools in such activities as used in this chapter might engender a
different interest in mathematics.
250 C. Julie
Area represents number of learners

adhering to rationale . Learners’ Rationales
Disciplinary
Cognate disciplinary (e.g.
physics, accounting)
Utilitarian (access to low

or non-mathematically
driven ‘careers’)
Instrumental (compliance
with what is prescribed to
be done in school)
Disinterested
Fig. 21.4 A configuration of students’ rationales for doing mathematics
I experimented with some of the ideas above with preservice mathematics teach-
ers (Julie 1993). It was a time of turbulence and protests against apartheid in South
Africa at school and higher education institutions attended by the oppressed sectors
of our society. One of the rallying calls was for political relevance to be included in
the subjects students studied. Mathematics was not spared. Based on Flatland, the
preservice teachers were given the project of developing an essay in which mathe-
matical objects are anthropomorphized. In a similar way to Abott, they used the idea
of the existence or not of formulae for solution of polynomomial equations to render
commentary on the political situation in the country at that time. This is indicative
of the possibility of “mathematics put to use” in domains not normally associated
with modelling and applications.
References
Abbott, E. A. (2008). The annotated Flatland: A romance of many dimensions (Introduction and
notes by I. Stewart). New York: Basic Books.
Alsina, C. (2002). Too much is not enough teaching maths through useful applicationswith local
and global perspectives. Educational Studies in Mathematics, 50(2), 239–250.
Aris, R. (1978). Mathematical modelling techniques. Marshfield: Pitman.
Dahl, R. (2013). The BFG. London: Penguin Books.
Davis, P., & Hersh, R. (1986). Descartes’ dream: The world according to mathematics. San Diego:
Harcourt.
Ellis, G. F. R. (1974). On understanding the world and the universe. Cape Town: University of
Cape Town.
Feynman, R. P. (1998). Six easy pieces. London: Penguin.

Forester, C.S. (1951). Hornblower and the widow McCool. Avaliable from hornblower.wikia.com/
wiki/Hornblower_and_the_Widow_McCool. Accessed 22 May 2017.
Freire, P., & Macedo, D. (1987). Literacy: Reading the word and the world. South Hadley: Bergin
& Garvey.
Gerdes, P. (1988). On culture, geometrical thinking and mathematics education. Educational
Jacobs, J. (2013). Oor ‘n motorfiets, ‘n zombiefliek en lang getalle wat deur elf gedeel kan word.
Pretoria: LAPA Uitgewers.
Julie, C. (1993). People’s mathematics and the applications of mathematics. In J. de Lange,
C. Keitel, I. Huntley, & M. Niss (Eds.), Innovation in maths education by modelling and appli-
cations (pp. 31–39). Chichester: Ellis Horwood.
Julie, C. (2012). The primacy of teaching procedures in school mathematics. In S. Nieuwoudt,
D. Laubser, & H. Dreyer (Eds.), Proceedings of the 18th annual national congress of the
Association for Mathematics Association of South African (pp. 16–31). Potchefstroom: North-
West University.
Koblitz, N. (1987). Elliptic curve cryptosystems. Mathematics of Computation, 48(177), 203–209.
Lagercrantz, D. (2015). The girl in the spider’s web. London: Maclehouse Press.
Lave, J. (1991). Situating learning in communities of practice. In L. B. Resnick, J. M. Levine, &
S. D. Teasley (Eds.), Perspectives on socially shared cognition (pp. 63–82). Washington, DC:
American Psychological Association.
Miller, J. D. (1983). Scientific literacy: A conceptual and empirical review. Daedalus, 112(2), 29–48.
Niss, M. (2015). Prescriptive modelling—Challenges and opportunities. In G. Stillman, W. Blum,
& M. S. Biembengut (Eds.), Mathematical modelling in education research and practice:
Cultural, social and cognitive influences (pp. 67–79). New York: Springer.
& M. Niss (Eds.), Modelling and applications in mathematics education: The 14th ICMI study
Noddings, N. (1993). Politicizing the mathematics classroom. In S. Restivo, J. P. Van Bendegem,
& R. Bielher (Eds.), Math worlds: Philosophical and social studies of mathematics and math-
ematics education (pp. 150–161). Albany: State University of New York Press.
Shoaf, M. M., Pollak, H., & Schneider, J. (2004). Math trails. Bedford: COMAP.
Square, A. (1884). Flatland: A romance of many dimensions. London: Seely & Co.
Strydom, F. (2017). The inside-out man. Century City: Umuzi.
wa Thiong’o, N. (2006). Wizard of the crow. London: Harvill Secker.
Chapter 22
Rising to the Challenge: Promoting
Mathematical Modelling as Real-World
Problem Solving
Peter Galbraith, Derek Holton, and Ross Turner
Abstract The International Mathematical Modelling Challenge (IM2C) adds a new

incentive for applications of mathematics to problems beyond the classroom. Here
we describe the design of a single package of support material for multiple audi-
ences: on the one hand, for teachers intending to introduce modelling into curricula
independent of Challenge involvement, on the other hand, for those wishing to enter
teams in the Challenge. Some of the latter may also need modelling support pro-
vided for the former. It is necessary to identify and apply principles, some common
to both and others directed to serving differing priorities within the two audiences.
This involves addressing issues that continue to engage the international commu-
nity, such as the purpose of modelling cycles, competencies, and the mentoring of
student modellers. The chapter engages and illustrates such considerations.
Keywords Competency · Mentoring · Modelling challenge · Modelling processes

· Support materials
P. Galbraith (*)
School of Education, University of Queensland, Brisbane, QLD, Australia
e-mail: p.galbraith@uq.edu.au
D. Holton
MGSE, University of Melbourne, Melbourne, Australia
e-mail: dholton@unimelb.edu.au
R. Turner
Australian Council for Educational Research, Camberwell, Australia
e-mail: Ross.Turner@acer.org

254 P. Galbraith et al.
22.1 Introduction
This chapter has a theoretical focus. Hence, samples of text and data, where used,
are for purposes of illustrating principles, not for confirming or rejecting conjec-
tures or hypotheses. The first memorable contributions to the field of mathematical
modelling in education were not put forward by professional educators nor did they
emerge from conventional educational settings. One of the earliest articles noting
the lack of attention to real-world problem solving in mathematics teaching (Pollak
1969) made no mention of mathematical modelling as such. Simply titled ‘How can
we teach applications of mathematics?’, it focused on the unrealistic nature of word
problems posing as ‘application problems’ in school textbooks. In later contribu-
tions, Pollak included descriptions of mentoring learners to become modellers –
where the setting was not a classroom but the research laboratory of the Bell
Telephone Company. Similarly, the early years of ICTMA1 featured staff in British
Polytechnics who had previously been engaged in applying mathematics to solve
real problems in industrial settings. They were concerned that their (tertiary) stu-
dents did not know how to go about solving problems in real-world contexts, or
communicating the results of their actions, and sought to promote this ability. Since
those times we have seen the growth of mathematical modelling programs in educa-
tion at all levels, many of which use the label indiscriminately, with scant regard to
maintaining the integrity of modelling as real-world problem solving. The descrip-
tor ‘modelling-as-vehicle’ has been coined to describe the appropriation of the term
for curricular purposes other than the enhancement of student ability to solve real-
world problems (e.g., Julie and Mudaly 2007, p. 504).
22.1.1 National Priorities
Statements indicating the importance of students enabled to apply mathematics to

real-world situations, including everyday life, society, and the workplace, have been
found increasingly in national curriculum statements (ACARA 2017; CCSSI 2012),
for example:
Mathematics aims to ensure that students are confident, creative users and communicators
of mathematics, able to investigate, represent and interpret situations in their personal and
work lives and as active citizens. (ACARA 2017:1, emphasis added)
However, these intentions have rarely progressed into mandated components of

courses. This has been exacerbated in part by a ‘one-size-fits-all’ approach in educa-
tion, typified by assertions such as the following: ‘The notion of mathematical mod-
elling at school level is different from the way it is understood by practising
mathematical modellers’. Responding to opportunities provided by the IM2C, in
1
International Community for the Teaching of Mathematical Modelling and Applications.
22 Rising to the Challenge: Promoting Mathematical Modelling as Real-World… 255
focusing on modelling problems that involve real situations, can assist in countering
such beliefs.
22.1.2 The International Mathematical Modelling Challenge
Strengthening the connection between mathematical knowledge and skills and their
deployment to address real-world problems is a central objective of the International
Mathematical Modelling Challenge (IM2C 2017). Initiated by the COMAP group,
it aims to make skills and approaches used by professional modellers accessible to
school students. In doing so, an overarching objective is to stimulate and facilitate
change in some of the classroom practices of mathematics teachers and learners.
The use of ‘promoting’ in the title of this chapter refers to the introduction, or wider
implementation, of mathematical modelling as real-world problem solving in edu-
cational settings. As elaborated later, there are two ways that this focus is being
developed in Australia. They are not part of the IM2C.
The Challenge began in 2015 to promote the learning of mathematical model-
ling, at all levels of education. It is a team-based competition, undertaken by groups
of up to four students, with a team advisor (usually a mathematics teacher). Teams
choose up to five consecutive days (within a fixed time period) to work on a model-
ling task that is set internationally, to write a report, and to submit it for evaluation.
Teams work on the problem independently, using any resources (short of direct
assistance), and use their collaborative efforts to develop a solution and written
report. Participating countries can select up to two reports for inclusion in the inter-
national round of judging.
Australia first participated in 2016, with 25 registered teams and 15 team reports
finally submitted for evaluation. By 2017, interest and engagement in the Challenge
had expanded threefold: team registrations increased to 82 and the number of sub-
mitted projects to 45 from 27 schools across Australia. For the 2017 Challenge, the
number of schools that made direct contact and expressed interest via the purpose-
built website for the IM2C in Australia reached a total in excess of 320. Teams which
submitted projects comprised 73 girls and 93 boys, with the grade distributions
being Year 11/12 (72), Year 9/10 (65), and Year 7/8 (29).
The spread across year levels, and responses from individual schools, indicates
that several schools have not seen the IM2C primarily as an opportunity to enter a
competition. Rather they see its visibility as presenting an opportunity to inject
something new and valuable (i.e., mathematical modelling), into their mathematics
programs. These responses, which accord with the aims of the Australian advisory
committee, have motivated the development of support materials. The materials are
independent of the IM2C, are designed to sustain, and encourage the two different
participant audiences identified above, viz.:
(1) Those wanting support for the teaching of modelling in schools up to senior
secondary level, with no intention to enter teams in the Challenge. They may
have modelling experience from zero upwards. (Also included are participants
in teacher education courses, and those who want to learn modelling indepen-
dently of a teacher.)
(2) Mentors intending to enter teams in the IM2C. This group varies in experience
and includes some who also value support at the level provided for the forego-
ing group.
This chapter illustrates how these dual aspects are being addressed within Australia
through the production of supporting resources for teachers and students. These
may be freely accessed through a continuing link, currently https://www.immchal-
lenge.org.au/supporting-resources/, a website of the Australian Council for
Educational Research (ACER 2020). The materials are suitable for use, and are
being used, beyond Australia.
22.2 The 2017 IMMC Problem: Jet Lag
The 2017 IM2C problem sets the scene in two ways. Firstly, it indicates the type of
problem and by inference the nature of the modelling abilities that the IM2C aims to
foster. Secondly, in focusing attention on such abilities, it provides a basis for
designing materials to support the introduction of modelling principles and exam-
ples for real-world problem solving in general. Foundation components of the mate-
rials are consistent with but not linked to the specifics of IM2C problems (audience
1 above). Some components are written with the needs of audience 2 in mind. We
now proceed to discuss issues influencing design priorities.
Organizing international meetings is not easy, including the problem that

some of the participants may experience the effects of jet lag after recent
travel from their home country to the meeting location which may be in a dif-
ferent time zone, or in a different climate and time of year, and so on. All these
things may dramatically affect the productivity of the meeting.
The International Meeting Management Corporation (IMMC) has asked your

expert group to create an algorithm that identifies the best place(s) to hold a
meeting given the number of participants, their home cities, and approximate
dates of the meeting.
Participants are usually from all corners of the Earth, and the meeting implies
doing hard intellectual team work for three intensive days, with the partici-
pants contributing approximately equally to the end result. Assume that there
are no visa problems or political limitations, and so any country or city can be
a potential meeting location.
(continued)
The output of the algorithm should be a list of recommended places (regions,

zones, or specific cities) that maximize the overall productivity of the meet-
ing. The questions of costs are not of primary importance, but may be a sec-
ondary consideration, as the IMMC cannot afford to bring participants in a
week before the meeting to acclimatize them or provide rest time after an
exhausting journey. Test your algorithm at least on the two following
datasets:
Scenario (1) “Small

Meeting” Scenario (2) “Big Meeting”
Time: mid-June Time: January
6 individuals from: 11 individuals from:
Monterey CA, USA Boston MA, USA (2 people) Warsaw, Poland
Zutphen, Netherlands Beijing, China Denmark, Copenhagen
Melbourne, Australia Hong Kong (SAR), China (2 people) Melbourne, Australia
Shanghai, China Moscow, Russia Singapore
Hong Kong (SAR), China Utrecht, Netherlands
Moscow, Russia
2
Your submission should consist of a one-page Summary Sheet and your solu-
tion cannot exceed 20 pages for a maximum of 21 pages.
22.3 Modelling as Real-World Problem Solving
In addressing the scope of real-world modelling, we recognise the need to encom-

pass two significant families – descriptive modelling and prescriptive modelling.
22.3.1 Descriptive Modelling
Descriptive modelling is the most familiar form of modelling, both in its pervasive-
ness and in its appearance within the ICTMA tradition. In its most common form, a
real-world problem is specified, a model formulated, and the resulting mathematical
results interpreted and evaluated in relation to the original problem. Where neces-
sary, further (cyclic) modelling activity is conducted until a satisfactory solution is
achieved. Descriptive modelling does not necessarily involve the validation of an
initial solution in an existing context. For example, modelling phenomena, such as
global warming or population projections, is undertaken without access to future
data for purposes of evaluation. Other methods of evaluation are required, for exam-
2
Appendices and references should appear at the end of the chapter and do not count towards the
20-page limit.
ple, completeness of model structure to handle known influences, accuracy of data

values used, and the reasonableness of the resulting outcome.
22.3.2 Prescriptive Modelling
In prescriptive modelling (e.g., Niss 2015), the purpose is not to explain or make
predictions about real-world phenomena but to organize or structure a situation, for
example, where a new hospital should be located. (In asking for an algorithm to
determine the optimum location of an international meeting, the IM2C jet lag prob-
lem for 2017 will be recognised as involving prescriptive modelling.) Every pre-
scriptive modelling project also has an inbuilt cyclic expectation, for any ‘solution’
needs to be subjected to sensitivity testing, a procedure which tests how vulnerable
the ‘solution’ is to changes in input variables. As a minimum, this involves recalcu-
lations of the model using modified assumptions, slightly amended parameter val-
ues, interpretation, and an evaluative judgment of the outcome. If small changes in
parameter values create big changes in outcome, a total revisiting of the model
is needed.
22.3.3 Representation of the Modelling Process
The cyclic aspect of real-world modelling is triggered by the outcome of an evalua-

tion procedure – which differs in mechanism but not in purpose (quality assurance)
between descriptive and prescriptive modelling. There has been a proliferation of
‘cycles’ associated with modelling. See, for example, Perrenet and Zwanefeld
(2012) or Greefrath and Vorhölter (2016) for a discussion of several of these. Some
have little to do specifically with the conduct of modelling as a problem-solving
process, and without a justification in terms of such a process, the use of a term such
as ‘cyclic activity’ is rendered meaningless. Any mathematical activity that contains
metacognitive reflection and subsequent action will involve cyclic processing,
whether or not modelling is involved. In summary, cyclic features of modelling
activity are problem-solving based, being generated by progress through phases of
a modelling process. Such cyclic activity may require iterating in the search for a
solution to a problem. It does not represent the path of mental activity of
individuals.
Specifying the modelling task ➔ formulating a mathematical model ➔ solv-

ing the mathematics ➔ interpreting outcomes ➔ validating/evaluating the
outcomes in terms of the real context ➔ revisiting the solution process or
documenting and reporting outcomes.
The principal emphasis of the present project with respect to cyclic activity is on
the provision of a scaffolding aid, which students can learn, internalise, and apply
independently to new situations within both descriptive and prescriptive modelling
contexts. We refer to an essential modelling cycle (see box). Other representations
exist, for example, diagrams as in Berry and Davies (1996).
The components are necessary in the sense that every modelling project contains
them. They are analytical in the sense that they describe a problem-solving process,
consistent with the activities of professional modellers, including mathematicians
(e.g., Pedley 2005). They do not purport to describe cognitive activities that take
place inside a modeller’s head but act as a guide to checking and progressing
solution attempts. The essential modelling cycle may be augmented further for pur-
poses of pedagogy (Blum and Leiß 2007; Lesh and Doerr 2003) or research
(Stillman 2011). With audience 1 in mind, we use an essential cycle, minimising
information processing load. It is supported by illustration and discussion, in work-
ing through a selection of example problems of various levels of depth. Additionally,
its components are exhibited in the successful modelling projects of student teams
within the IM2C (audience 2).
22.3.4 Modelling Competency(ies)
Sadler (2007) writing in the context of assessment refers to extremes whereby ‘cri-
teria have been reduced to pea-sized bits to be swallowed one at a time’, resulting in
a ‘focus on a multitude of discrete competencies, rather than on competence’. He
writes of the danger of losing sight that overall competence involves an ability to
orchestrate a set of competencies. ‘Scaffolding’ is identified as a positive educa-
tional principle that has had its original intention (to provide supports during learn-
ing) hijacked to another purpose. He argues that scaffolding can become so
elaborate, and the level of assistance so detailed, that the learner cannot help but
‘succeed’. Remove the scaffolding, and failure ensues. Yet scaffolding is supposed
to be a temporary arrangement that supports the building process, following which
the building needs to stand on its own. This is a salutary warning that similar mis-
takes can be committed in searching to define and promote modelling competence.
There is no dispute about the importance of individual competencies, nor the
valuable work done in identifying separate abilities that together contribute to mod-
elling competence (e.g., Maaß 2006; Greefrath 2015). There is however reasonable
debate about the extent to which the modelling whole is more than the sum of sub-
competency parts, that is, what does the demonstration of individual competencies
tell us about the ability to select and employ them effectively in the context of an
extended problem? Our approach has been to regard the essential indicator (in the
sense of Sadler) of the possession of individual competencies as their successful
deployment in extended activity. Of course, not all sub-competencies will be equally
applicable across different problems and contexts. In our approach, competencies
needed are dealt with as they emerge naturally in the solution process, and the suc-
1. Problem definition
● Can identify a real-world problem to be solved, and specify precise mathematical
questions from the general problem statement
2. Model formulation
● Identify assumptions with justification ● Choose variables appropriately
● Identify and gather relevant (needed) data ● Choose and justify parameter values
● Develop mathematical representations
3. Mathematical processing
● Apply relevant mathematics ● Invoke and use appropriate technology
● Check mathematical outcomes for procedural accuracy
● Interpret mathematical outcomes in terms of the problem situation
4. Model evaluation
● Explore adequacy and relevance of findings in relation to problem situation
● Provide further elaboration or refinement of problem
● Assess relevance of revised solution(s) following revisiting and further work within
earlier criteria
● Conduct evaluation of sensitivity of solution to changed assumptions or conditions
● Reflect on quality of answers to specific questions posed in problem statement
Fig. 22.1 Competency-related assessment criteria
cession of competencies required within a given problem is merged therein. (Refer

to the weblink provided in the previous section.)
A list of related assessment criteria is given in Fig. 22.1. These have been devised
in concert by the Australian advisory group, several of whose members have experi-
ence in implementing and assessing modelling projects over a period of more than
20 years. For Challenge participants and their mentors (audience 2), these also form
criteria for the assessment of project reports. For those using the materials to intro-
duce and develop modelling at school (audience 1), the training problems ‘walk’
readers through a solution process, in which the sub-competencies are systemati-
cally invoked and integrated. (See website material.)
22.3.5 Mentoring
Mentoring has a special meaning within IM2C as the role of advisors of student
teams. It can also be taken more broadly to describe anyone who introduces model-
ling into a classroom and acts as a facilitator of student progress.
For audience 2, mentoring advice is provided through commentary referring
readers to specifics of team reports that demonstrate effective ways of expressing or
pursuing objectives within IM2C (see illustrations below). They are identified by
team number and can be sighted by accessing the corresponding numbered report
via the link https://www.immchallenge.org.au/supporting-resources/previous-
immc-problems/. These comments relate to Australian reports for the 2016 prob-
lem, which has been chosen as an exemplar. For those new to the field, mentoring
activities are embedded in the way that the training problems are developed, in the
modelling dialogue displayed and articulated within phases of the modelling pro-
cess, and in the transitions between them. (Fig. 22.2)
1. Problem definition: Examples of the specification of precise mathematical questions from a general
problem statement are displayed well in reports 2016026 and 2016021.
2. Model formulation: Good illustrations of the identification of assumptions with justification: are
provided in team reports 2016021, 2016025 (prose) and 2016026 (tabular format)
3. Choice of variables. Teams 2016021, and 2016033 each approached sub-questions in two different
ways in efforts to establish understanding and robustness.
4. Relevance of revised solution(s) following revisiting and further work. Team 2016026 followed through
two models produced by different approaches, which enhanced their approach to evaluation.
5. Model evaluation. Teams 2016021, 2016015, and 2016036, specifically included a discussion of
‘strengths and weaknesses’ in evaluating the output of their models. Team 2016026 discuss model
limitations, noting the large number of variables that they introduced - and add a concern about the
accessibility of their model in the absence of a computer.
Fig. 22.2 Mentoring feedback to teams (samples)
● What is mathematical modelling?

● Mathematical modelling framework
● Example (training) problems: Simple modelling activities
Self-generated modelling activities
Suggestions for student-generated modelling
Harder modelling activities
Previous IM2C problems (including team solutions)
● Teamwork and report writing
Fig. 22.3 Structure of support materials
22.4 Structure and Availability of Support Materials
The components shown in Fig. 22.3 have been developed from principles based on
the rationale and theoretical orientation described in Sects. 22.3.3 to 22.3.5. These
principles are elaborated in the materials. The structure is hierarchical, developing
progressively from the top entry. Given the two audiences, the materials are struc-
tured in modular form – so respective readers can target sections relevant to their
needs. As indicated, they are freely available on the website https://www.immchal-
lenge.org.au/supporting-resources/ ACER 2020).
For example, within ‘Example problems’, new modellers will begin at the start and
finish when they reach a level appropriate to their needs. Experienced users will typi-
cally not require introductory material and start somewhere in the middle of the hierar-
chy. The presentation is problem driven, in the sense that modelling requirements drive
the selection and ordering of material, not the production of lesson plans. All training
problem examples have previously been used with students and/or teachers in some form.
22.5 Feedback from Schools (Selected)
A survey following the 2017 Challenge registered a range of comments about the
benefits of participating, for example:
The students really enjoyed participating in the real-world problem solving challenge. They
enjoyed the pressure it created, and the creativity that came about from working together in
a team. The school will benefit as we take on board some the strategies required to solve this
International Mathematical Modelling Challenge.
The students learnt to be resourceful, to be self-reliant, to work cooperatively and inde-
pendently, as well as how to problem solve when the question is open ended. I am hoping
that the benefit to the school will increase interest in this type of activity and more participa-
tion in the future. This group of girls is looking forward to participating next year.
There was comment on difficulties fitting the modelling challenge into a crowded
school program. Commonly, team members used some mathematics class time
spread over 3 days and their own time, including a weekend. Often teams under-
rated the problem’s demands and spent less time than needed. A key to successful
teamwork was strong collaboration, sometimes initiated and maintained by the use
of web-based collaborative tools. Failure to submit reports was often due to school
arrangements that did not enable effective teamwork or to lack of experience with
modelling processes. Some reports were weak, as to be expected from people dip-
ping their toes into a new field. This underlines the dual challenge of enhancing the
role of modelling as real-world problem solving in education. It involves increasing
the numbers involved but also the quality of resulting student work.
22.6 Summary
This chapter has illustrated how opportunities presented by a new initiative are
being developed independently in the Australian context. The goal is to increase the
capacity of students, schools, and education systems, to embrace and enhance mod-
elling as real-world problem solving. In designing support material to cover the
needs of different audiences, it has been necessary to confront issues in theory and
practice that engage the modelling community. These include the nature of model-
ling tasks (descriptive and prescriptive), the structure and purpose of modelling
cycles, notions of competency, and challenges of mentoring. While an Australian
project, there is also international interest. Of more than 6300 website page views in
the 3 months to January 2018, 63% were from Australia. The next highest inci-
dences were from India (6.5%), the USA, South Africa, the UK, and Malaysia.
And while the quality of student work is variable, the following comment
received within teacher feedback is illuminating.
This is the second year the school has been involved and (I’ve) been blown away by the
efforts of the students and the complexity of the solutions they’ve provided.
It confirms observations of practitioners, within and outside education, that students

are capable of extraordinary initiative and quality when given appropriate incentive,
freedom, and mentoring. This adds another option for seeking support. Rather than
needing to argue up front in theoretical terms as to what modelling can achieve,
there is opportunity to demonstrate, through work samples, what students have actu-
ally done. This suggests another avenue of approach to both sponsors and education
authorities, to enhance both the provision and the outcomes achievable through
modelling as real-world problem solving, in particular the ability to use mathemat-
ics to address problems in the living environment.
References
ACARA. (2017). Australian Curriculum and Reporting Authority. Understand how Mathematics
works: Aims. Retrieved September 28, 2017, from https://australiancurriculum.edu.
au/f-10-curriculum/mathematics/
ACER. (2020). IM2C: International Mathematical Modelling Challenge. https://www.immchal-
lenge.org.au/about-the-immc
Berry, J., & Davies, A. (1996). Written reports. In C. R. Haines & S. Dunthorne (Eds.), Mathematics
learning and assessment: Sharing innovative practices (pp. 3.3–3.11). London: Arnold.
Blum, W., & Leiß, D. (2007). How do students and teachers deal with mathematical modelling
problems? The example “Filling up”. In C. R. Haines, P. Galbraith, W. Blum, & S. Khan (Eds.),
CCSSI. (2012). Common Core State Standards Initiative Standards for Mathematical Practice –
Model with Mathematics. Retrieved September 27, 2016, from: http://www.corestandards.org/
Math/Practice/MP4
Greefrath, G. (2015). Problem solving methods for mathematical modelling. In G. Stillman,
Greefrath, G., & Vorhölter, K. (2016). Teaching and learning mathematical modelling: Approaches
and developments from German speaking countries. In Teaching and learning mathematical
modelling. ICME-13 Topical Surveys (pp. 1–42). Cham: Springer.
IM2C. (2017). International Mathematical Modelling Challenge. http://immchallenge.org/
Julie, C., & Mudaly, V. (2007). Mathematical modelling of social issues in school mathematics in
South Africa. In W. Blum, P. Galbraith, H.-W. Henn, & M. Niss (Eds.), Modelling and appli-
cations in mathematics education: The 14th ICMI study (pp. 503–510). New York: Springer.
Lesh, R., & Doerr, H. M. (Eds.). (2003). Beyond constructivism: Models and modeling perspec-
tives on mathematics problem solving, learning, and teaching. Mahwah: Lawrence Erlbaum.
113–142.
Niss, M. (2015). Prescriptive modelling – Challenges and opportunities. In G. Stillman, W. Blum,
& M. S. Biembengut (Eds.), Mathematical modelling in education research and practice
Pedley, T. J. (2005). Applying mathematics. Mathematics Today, 41(3), 79–83.
Perrenet, J., & Zwanefeld, B. (2012). The many faces of the mathematical modelling cycle. Journal
of Mathematical Modelling and Application, 1(6), 3–21.
Pollak, H. (1969). How can we teach applications of mathematics? Educational Studies in
Mathematics, 2(2–3), 393–404.
Sadler, D. R. (2007). Perils in the meticulous specification of goals and assessment criteria.
Assessment in Education: Principles, Policy and Practice, 14, 387–392.
Stillman, G. (2011). Applying metacognitive knowledge and strategies to applications and mod-
(Eds.), Trends in teaching and learning of mathematical modelling (pp. 165–180). New York:
Springer.
Chapter 23
Mathematics Trails and Learning Barriers
Iwan Gurjanow and Matthias Ludwig
Abstract Mathematics trails are a popular way to implement outdoor education

outside mathematics classrooms. The MathCityMap project combines the idea of
maths trail with the possibilities of modern mobile devices. During a study, in which
ninth graders were asked to complete a maths trail with the MathCityMap applica-
tion, we observed them having trouble with measuring and handling quantities.
Analyses of the notes and calculations of the students revealed that over one third
had a poor measurement sense. This learning barrier is a big issue, since these com-
petencies are prerequisites to successfully complete a maths trail as well as for
modelling and should be addressed prior to carrying out these activities.
Keywords Maths trails · Technology · Learning barrier · Measurement sense ·

MathCityMap · MCM app · Grade 9 students
23.1 Introduction
The British Department of Education and Skills recommends carrying out more les-
sons outside the classroom. Learning outside can “nurture creativity, develop skills,
improve attitude to learning, stimulate and improve motivation” just to name a few
benefits (DfES 2006, p. 2). A meta-analysis on outdoor learning indicates that not
only students remember fieldwork and outdoor visits for many years but also the
experience of outdoor learning is considered “more effective for developing cogni-
tive skills than classroom-based learning” (Dillon et al. 2006). One suitable way to
implement learning outdoors in the mathematics classroom is the mathematics trail
concept. Blane and Clarke (1984) were among the first to present the mathematics
trail idea to a broad, scientific audience. According to Shoaf et al. (2004), maths
trails are accompanied by a trail guide booklet showing mathematically interesting
places and tasks, usually including a map, for the user to follow.
I. Gurjanow (*) · M. Ludwig

Institute of Mathematics Education, Goethe University Frankfurt, Frankfurt, Germany
e-mail: gurjanow@math.uni-frankfurt.de; ludwig@math.uni-frankfurt.de

266 I. Gurjanow and M. Ludwig
Due to a high availability of mobile devices, their important role in students’

daily lives (Kinshuk and Chen 2005) and freely accessible maps (e.g. OpenStreetMap
or Google Maps), it was a logical step to support the maths trail activity with a digi-
tal application. Jesberg and Ludwig published a first conceptualization for
technology-supported maths trails in 2012 (Jesberg and Ludwig 2012). In 2016, the
MathCityMap (MCM) project was launched. Since the launch, several studies have
been conducted using the MCM app; for instance, Cahyono and Ludwig (2019)
examined motivation of Indonesian students; and Zender and Ludwig (2018) stud-
ied students’ performance in geometry. One of the key features of the application is
a stepped hint system, which provides up to three hints. Usually, the third hint con-
tains already a large part of the solution to a task. Nevertheless, we observed many
groups having trouble finding the correct answer during the studies. In particular,
many times the use of measuring tools and working with units seemed to be a chal-
lenge for students. A literature review revealed that we as researchers and trailblaz-
ers have underestimated the role of measurement sense while doing mathematics
outdoors. This chapter aims to highlight the importance of measurement sense for
modelling in a real-world context (Hagena 2015) as well as to examine it empirically.
23.2.1 Mathematics Trails and Modelling
Blane and Clarke (1984) as well as Shoaf et al. (2004) considered mathematics
trails a good way to popularize mathematics. In this context, a maths trail should be
voluntary, self-directed and not only for students but for everyone (e.g. families).
Going on a maths trail means to experience and discuss cooperatively mathematics
in a “non-threatening environment” (p. 5). Doing mathematics in an authentic out-
door setting has also properties that are desirable in the school context. The connec-
tion of measurement (enactive action), calculating with the gathered data (cognitive
action) and translating the given problem into a mathematical model (abstraction) is
valuable, because one is more likely to remember what is learned later (Rösler
2011). Furthermore, this collaborative activity has the potential to promote many
process and content standards, as the National Council of Teachers of Mathematics
(NCTM 2000), for example, describes them. Students need to understand the prob-
lem, find an appropriate mathematical model, collect data, work mathematically
and validate or discuss their answer. These activities come very close to the model-
ling cycle as described by Blum and Leiß (2005) and help to promote modelling.
Additionally, students are naturally required to discuss and convince their team-
mates of their solution approach, which promotes the process competencies com-
munication and reasoning. The mathematical content depends on the tasks that a
23 Mathematics Trails and Learning Barriers 267
maths trail offers. Problems in the field of geometry and measurement are obvious,
since different forms shape our environment. Nevertheless, it is also possible to find
examples in the area of algebra and probability.
In contrast to the original trails, maths trails at school are not voluntary, are
mostly prepared by teachers for their students with a focus on a particular topic and
are subject to organizational obstacles. Since the available time is usually set by
lessons of a certain period (e.g. 45 or 90 min) and the learning group consists of
about 20–30 individuals, it is necessary for the teacher to structure the maths trail
activity well to be a success. Methodologically, a maths trail can be categorized as
outdoor learning at stations. At predefined stations, students must complete tasks
cooperatively without teacher interference. To run smoothly, each station should
demand roughly the same amount of time, offer stepped hints, and students should
be able to complete it “without having completed another station beforehand” (Eilks
2002). The number of stations should be high enough so that students will not hin-
der others from working. From our experience, at least one station per group of
students proved to be a good guideline. Due to these organizational obstacles, it is
challenging and takes effort to create a maths trail with complex modelling tasks,
since one has to create many tasks with a similar difficulty level that connect to dif-
ferent objects. It is therefore more useful to have simpler modelling tasks that focus
on particular steps of the modelling cycle that can be completed within a period of
about 15 min.
Yet, to be able to solve these tasks, students have to go through the following
steps of the modelling cycle as described by Blum and Leiß (2005). Firstly, they
need to simplify the problem by choosing an appropriate real-world model and later
on translate it to a mathematical model. Secondly, based on the chosen model, they
will take necessary measurements and, finally, calculate the answer. The
MathCityMap app provides an automatic validation of the answer, which is based
on a pre-defined interval for correct answers, formulated by the author of the task,
but may additionally take place during a discussion between students. The result of
the validation process determines if the students need to modify their model and go
through the modelling cycle once again. To focus on the appropriateness of the
chosen model, teachers should ask students to write down their arguments on the
corresponding trail guide and bring them to the following lessons. This way, the
maths trail integrates into traditional mathematics classroom practice and serves as
a source for discussion about different possible models.
The Camel Thorn Tree task is a typical example for a MCM task. One possible
approach is to model the tree as a cylinder, to measure the circumference, to divide
the circumference by Pi to obtain the diameter and, finally, to divide the diameter by
3 mm to find an answer.
Camel Thorn Tree (Displayed in Fig. 23.1)

The diameter of a camel thorn tree grows (to a height of 1.5 m) 3 mm every
year. Estimate the age of this camel thorn tree.
(continued)
Fig. 23.1 Screenshot of

the MathCityMap app
displaying the Camel
Thorn Tree task
23.2.2 Measurement Sense as a Learning Barrier
One main difference between modelling inside the classroom and outside the class-
room is the way a model is chosen. Inside the classroom, usually data, which are
available in the text or image, lead to a choice of model. Outside, learners have to
choose a model prior to collecting necessary data via measuring. In both cases han-
dling units and converting them are often necessary to complete the task.
According to Leuders and Barzel (2014, p. 48), the concept of measuring is a
fundamental invention of humankind to deal with many challenges of the real world.
Furthermore, in this content area, the arithmetical knowledge about numbers and
operations, geometry and knowledge about other subjects as well as daily life are
deeply interconnected. Working with units of measure and their notations (m2, cm,
etc.) has become so commonplace in everyday life that learning barriers are often
underestimated or even overlooked (Leuders and Barzel 2014; Thompson and
Preston 2004). Hagena (2015, p. 186) defines five central aspects of measurement
sense (MS) that help handling quantities:
MS 1 Being able to identify and distinguish different quantities of measurement
MS 2 Being able to measure, to estimate and to round
MS 3 Being able to decide whether to estimate, to measure and to round
MS 4 Having knowledge about units of measurement which includes calculating
and converting units
MS 5 Having a set of meaningful benchmarks for these units and being able to
use these benchmarks
In the context of a mathematics trail, we observed students having trouble with the
following aspects and declared them as learning barriers (LB):
LB 1 Using measuring tools to gather data and converting units (related to MS 2
and MS 4)
LB 2 Further calculations to complete the task
In mathematics classes, students often learn how to use rulers to measure centi-
metres and millimetres. Rarely, do they learn how to use a measuring tape or folding
rulers to measure longer distances. The lack of experience in handling measurement
tools leads to LB 1. Additionally, learners usually collect data in different units,
which makes it necessary to convert units before calculating. Furthermore, tasks
may demand answers in a pre-defined unit. LB 1 includes mistakes that happen in
those two cases as well. LB 2 describes challenges of determining, with the aid of
already-measured quantities, quantities that cannot be measured directly (e.g. vol-
ume or in some cases diameter) which usually results in arithmetical mistakes.
Over and above that, handling quantities plays also an important role in mathemati-
cal modelling inside the classroom. Hagena (2015) conducted a study on mathematical
modelling competencies with pre-service teachers. The participants of the experimen-
tal group who had trained their measurement sense showed significantly higher scores
in the modelling tests than did members of the control group who had not.
23.2.3 The MathCityMap Project
The MathCityMap project combines the idea of maths trails with the possibility of
modern mobile devices. The project is divided into two components. The MCM
web portal is built for trailblazers and aims to support creation of maths trails by
standardizing and simplifying the creation process. Learners use the MCM app for
iOS and Android to walk a prepared maths trail. The app gives users the possibility
to go on a math trail more independently by using mobile devices to find the task
locations, by giving feedback on the users’ answer and by providing hints in the
case that one becomes stuck at a particular task.
23.2.3.1 MCM Web Portal: www.mathcitymap.eu
The core of the MathCityMap web portal is a map of the world (OpenStreetMap)
that displays available trails all over the world. At the moment of writing this chap-
ter, over 1000 trails in more than 10 languages are available. The intention is to
establish a digital community of trailblazers that publish and share mathematics
trails and their ideas among each other. After a short registration, users can create
their own tasks and compile them to trails. For every maths trail, the system auto-
matically creates a PDF trail guide that contains all task information, a map over-
view and a title page. It is the teacher’s decision if she or he wants the learners to
walk the maths trail by using the digital trail guide of the MCM app, the traditional
paper trail guide or a combination of both.
23.2.3.2 MCM App for Mobile Devices
The MCM app (see Fig. 23.1) allows the user to access maths trails created with the
web portal. The app downloads trail data, such as images and map tiles, to the
mobile device so that it is possible to use a trail without any Internet connection.
This design decision minimizes technical issues when using the app offline or in an
area with low connectivity. Furthermore, the app offers an open street map overview
for orientation purposes, feedback on the entered answers and a stepped hint sys-
tem. The hint system has the purpose of enabling students to solve the tasks inde-
pendently and additionally has a positive impact on learning performance, learning
experience and communication (Franke-Braun et al. 2008). To help organize a
maths trail session, the MCM app allows users to start at a certain task. In this case,
all other tasks will be hidden. The next task becomes available, after a user com-
pletes the previously selected task. This feature allows teachers to assign a different
starting task for each group of students and to implement the learning at stations
teaching method.
23.3 Research Study
The literature review revealed that learning barriers in connection with measure-
ment and handling quantities already well describe and explain observations we
made during the maths trail interventions. Nevertheless, our research aim is to
determine if LB 1 is also empirically a major challenge for Grade 9 students while
completing a maths trails with MCM.
23.3.1 Methodology
The data that will be presented and analysed in the following sections were col-
lected in May and June 2017 during a study about learning performance (Zender
and Ludwig 2018) and motivational aspects (Gurjanow and Ludwig 2017). Grade 9
students from the region of Frankfurt/Main (Germany) were invited to complete a
maths trail about cylinders and cones with the MCM app. Their notes and sketches
are the main data source.
23.3.1.1 Study Design
Overall, 18 ninth grade school classes (aged 14–15 years) with 426 students (220
female, 202 male, 4 unknown gender) participated in the conducted study. In the
first 10 min, we welcomed and introduced the students to the MathCityMap app and
to the use of the provided measurement tools (folding ruler and measuring tape).
Subsequently, the teachers divided their students into groups of three. Each group
picked up their equipment, which consisted of a trail guide, measurement tools as
well as a smartphone, and walked a prepared maths trail for about 75 min. In the last
5 min, the subjects filled in a questionnaire on motivation. Since the results of the
questionnaire do not contribute directly to the research aim, they will not be part of
this chapter. Nevertheless, for the sake of completeness, we inform the reader of the
nature of the intervention. Altogether, the duration of the intervention was 90 min,
and it was designed in such a way that teachers could easily replicate it in a school
setting. The trail consisted of 20 tasks so that 10 groups of 3 could easily work on
tasks without hindering others. The Memorial Plate is one example task (including
stepped hints) to illustrate the spirit of the trail.
Memorial plate
What is the mass of the stainless steel plate? 1 m3
of stainless steel weights 7900 kg. Give your result
in kg.
Hint 1
Examine the plate more closely to determine the
height of it precisely.
Hint 2
The plate can be approximated by a compressed
Image of the memorial plate cylinder. Calculate the volume in a suitable unit.
Hint 3
Calculate the mass by multiplying the volume with
the given density of 7900 kg per m3.
23.3.1.2 Data Analysis
Overall, the 426 students formed N = 149 small groups of 2 or 3. The groups of
students made notes, sketches and calculations on their trail guides while working
on the trail. The guides were collected afterwards and scanned. Each task of the trail
was examined and categorized. The following categories were used:
Category 0 No attempt to complete the task (e.g. empty page)
Category 1 Mistakes connected with LB 1 (e.g. strongly deviating measured
values and mixing up different units in one term, see Fig. 23.2a;
wrong unit conversion, see Fig. 23.2b)
Category 2 Mistakes connected with LB 2 (e.g. wrong term conversion)
Category 3 Correct answer
23.4 Results
Table 23.1 shows a summary of all analysed trail guides. On average, students tried
to complete four tasks during the intervention and managed to complete two of
them correctly. In 19% of the cases, they made a mistake connected to LB 1. LB 2
occurred in 21% of the cases. The students failed on almost every second maths trail
task due to the learning barriers described. The SD value (shown in brackets in
Table 23.1) of correct completed tasks is remarkably high at 2.05. When a group
worked on a task (made notes or sketches on the trail guide), it was counted as “task
processed” even if they did not manage to find a correct answer. The number of
processed tasks (see Fig. 23.3a) and correctly completed tasks (see Fig. 23.3b) for
groups ranges from 1 to 13 and 0 to 10, respectively. The large ranges indicate dif-
ferences among the students concerning their engagement and mathematical perfor-
mance level. We conducted a second analysis with respect to three levels of
performance (see Table 23.2). The low performance level (Low) consists of groups
that completed zero or one task. The students of the average performance level
Fig. 23.2 Category 1 examples: (a) students mixing up metres and centimetres and (b) students
dividing by 100 to convert cm2 to m2
Table 23.1 Summary of mean number of completed and categorized tasks in all analysed
trail guides
Number of processed Category 1 (LB Category 2 (LB Category 3 (correct
N tasks 1) 2) answer)
149 3.8 (±2.5) 0.74 (±0.84) 0.78 (±1.00) 2.30 (±2.05)
19% 21% 60%
Fig. 23.3 Boxplots of

number of processed tasks
for (left) and correct
answers (right) for
the groups
(Avg) completed two to four tasks. The high performance level (High) completed
five or more tasks. The division is based on the mean and the standard deviation of
tasks completed.
The performance levels differ mainly in the count of processed tasks and the
quota of mistakes. The number of tasks that the groups in each level tried to com-
plete increases from 2.1 (low) to 4.3 (avg) up to 8 (high) (see Table 23.2).
Simultaneously, the quota of mistakes decreases with the performance level. In
about 70% of the processed tasks, the low performers encountered a learning bar-
rier, whereas the average level groups encountered learning barriers in 36% of cases
and the high performers only in 22% of the cases. In each performance level, the
sources of mistakes were identified as being split almost equally between category
1 and category 2.
Table 23.2 Analyses divided into performance levels

Performance Number of Category 1 Category 2 Category 3 (correct
N level processed tasks (LB 1) (LB 2) answer)
66 Low 2.1 (± 1.1) 0.73 (±0.73) 0.71 (±0.88) 0.62 (±0.49)
35% 35% 30%
63 Avg 4.3 (± 1.8) 0.76 (±0.97) 0.78 (±1.10) 2.81 (±0.81)
18% 18% 64%
20 High 8.0 (± 2.1) 0.70 (±0.71) 1.00 (±0.95) 6.25 (±1.65)
9% 13% 78%
23.5 Discussion
In Germany, measuring and working with units are addressed in Grade 5. One could
expect students to have internalized these topics by Grade 9. Nevertheless, over one
third of the students failed to solve tasks due to these aspects. Instead, a poor under-
standing and knowledge of measurement and handling units were identified as
major learning barriers. These prevented the participants from completing the math-
ematics trail successfully and reaching the actual goal of becoming aware of the
mathematics that surrounds them in their daily lives. The results are in line with
Leuders and Barzel (2014) as well as with Thompson and Preston (2004) and con-
firm empirically their thesis that handling quantities is a weakness among students.
Furthermore, the results indicate and confirm the interconnection between measure-
ment sense, arithmetic and geometry. Students who made many mistakes in han-
dling units (LB 1) also had a high percentage of mistakes in arithmetic (LB 2).
Improving students’ measurement sense prior to carrying out a mathematics trail is
crucial and promises good results (cf. Hagena 2015). Finally, we need to explore
more deeply the potentials of mathematics trails to promote measurement sense. It
seems reasonable that measuring real objects and calculating their quantities can
help to gather experience in this field as well as to establish meaningful benchmarks
of units.
Regarding the stepped hint system of the MathCityMap app, we need to recon-
sider and improve our current practices. Learning barriers in the field of measure-
ment need to be addressed as well as modelling processes.
References
Blane, D. C., & Clark, D. (1984). A mathematics trail around the city of Melbourne. Melbourne:
Monash Mathematics Education Centre, Monash University.
Blum, W., & Leiss, D. (2005). “Filling Up” – the problem of independence-preserving teacher
interventions in lessons with demanding modelling tasks. In M. Bosch (Ed.), Proceedings of
the 4th Congress of the European Society for Research in Mathematics Education. Barcelona:
FUNDEMI IQS.
Cahyono, A. N., & Ludwig, M. (2019). Teaching and learning mathematics around the city
supported by the use of digital technology. Eurasia Journal of Mathematics, Science and
Technology Education, 15(1), em1654. https://doi.org/10.29333/ejmste/99514.
Department for Education and Skills (DfES). (2006). Learning outside the classroom manifesto.
Nottingham: DfES Publications.
Dillon, J., Rickinson, M., Teamey, K., Morris, M., Choi, M. Y., Sanders, D., & Benefield, P. (2006).
The value of outdoor learning: Evidence from research in the UK and elsewhere. School
Science Review, 87(320), 107.
Eilks, I. (2002). Learning at stations in secondary level chemistry. Science Education International,
13, 11–18.
Franke-Braun, G., Schmidt-Weigand, F., Stäudel, L., & Wodzinski, R. (2008). Aufgaben mit
gestuften Lernhilfen - ein besonderes Aufgabenformat zur kognitiven Aktivierung der
Schülerinnen und Schüler und zur Intensivierung der sachbezogenen Kommunikation. In
R. Messer & W. Blum (Eds.), Lernumgebungen auf dem Prüfstand. Zwischenergebnisse aus
den Forschungsprojekten (pp. 27–42). Kassel: Kassel University Press.
Gurjanow, I., & Ludwig, M. (2017). Gamifiying math trails with the MathCityMap app: Impact of
points and leaderboard on intrinsic motivation. In G. Aldon, & J. Trgalova (Eds.), Proceedings
of the 13th International Conference on Technology in Mathematics Teaching (ICTMT 13)
(pp. 105–112), Lyon, France.
Hagena, M. (2015). Improving mathematical modelling by fostering measurement sense: An
intervention study with pre-service mathematics teachers. In G. A. Stillman, W. Blum,
& M. S. Biembengut (Eds.), Mathematical modelling in education research and practice
Jesberg, J., & Ludwig, M. (2012). MathCityMap – Make mathematical experiences in out-of-
school activities using mobile technology. In Proceedings of the International Conference on
Mathematics Education 12 (pp. 1024–1031). Seoul.
Kinshuk, D., & Chen, J. (2005). Mobile technology in educational services. Journal of Educational
Multimedia and Hypermedia, 14(1), 89–107.
Leuders, T., & Barzel, B. (2014). Größen, Maße und Messen. In H. Linnenweber-Lammerskitten
(Ed.), Fachdidaktik Mathematik. Grundbildung und Kompetenzaufbau im Unterricht der Sek. I
und II (pp. 48–68). Bad Langensalza: Klett Verlag.
National Council of Teachers of Mathematics (NCTM). (2000). Principles and standards for
school mathematics. Reston: NCTM.
Rösler, F. (2011). Psychophysologie der Kognition. Heidelberg: Spektrum Akademischer Verlag.
Shoaf, M., Pollak, H., & Schneider, J. (2004). Math trails. Lexington: The Consortium for
Mathematics and its Applications (COMAP).
Thompson, T., & Preston, R. V. (2004). Measurement in the middle grades. Insights from NAEP
and TIMSS. Mathematics Teaching in the Middle School, 9, 514–519.
Zender, J., & Ludwig, M. (2018). Einsatz von MathCityMap in der Sekundarstufe I – Eine Studie
über den Leistungszuwachs bei Schülern aus Klasse 9. In Beiträge zum Mathematikunterricht
2018 (pp. 2051–2054). Gesellschaft für Didaktik der Mathematik. WTM-Verlag.
Chapter 24
Validating with the Use of Dynamic
Geometry Software
Corinna Hankeln
Abstract A difficult step in the modelling process is to validate results, models or

assumptions. Using the Dynamic Geometry Software GeoGebra to do modelling
appears to be a promising means to foster students’ validating competencies. This
chapter presents results from a quantitative study in which the development of the
validating competency in 15 classes that worked with GeoGebra throughout a mod-
elling project is compared to the development in 15 classes that worked with paper
and pencil. Results show that both groups significantly and lastingly improved their
validating competencies, but that no significant difference between the groups could
be found using tasks without tools in testing.
Keywords Dynamic Geometry Software · Competency · DGS GeoGebra ·

Mathematical modelling · Technology · Validating · Validating competency
24.1 Introduction
The introduction of technology and digital tools in the modelling process has been
a topic of interest in the last decades. Already in the second ICTMA book in 1986,
there was a section concerning the use of microcomputers and simulations (Berry
et al. 1986). Nevertheless, still today numerous questions remain unanswered. For
example, the question in the ICMI topic study of how technology should be used at
different educational levels to effectively develop students’ modelling abilities and
to enrich the students’ experience of open-ended mathematical situations in applica-
tions and modelling (Blum et al. 2007) has not been given a satisfying answer yet
(English et al. 2016). This chapter can be seen as a contribution to this discussion
with a limited focus on Dynamic Geometry Software (DGS) and the competency of
validating.
C. Hankeln (*)
University of Münster, Münster, Germany
e-mail: c.hankeln@uni-muenster.de

278 C. Hankeln
In research, special attention has been given to the different phases of the model-
ling process and the competencies necessary to carry out those modelling steps
(Kaiser and Brand 2015). Within this discussion, especially the phase of validating
a found solution has been identified to present a challenge for students (Blum 2007).
This is also valid when modelling is done with digital tools. Lingefjärd and
Holmquist (2003) even considered it “obvious that a clear focus on the validation
part of mathematical modelling is more essential in the presence of technology than
maybe ever before” (p. 214). Doerr and Pratt (2008, p. 28) were convinced that there
is a need to appreciate more deeply the relationship between the design of tasks and
tools and the modelling activity that ensues and to understand how the (tool, task)
pair might encourage validation activity. This chapter approaches these issues by
analysing what influence the use of a Dynamic Geometry Software might have on
the development of the competency of validating.
24.2.1 Validating
First, it must be clarified what is understood by validating: Validating connects the

real results back to the situation model in terms of the modelling cycle by Blum and
Leiß (2006). Validating competency also means, following Blum and Kaiser (as
cited in Maaß 2006), to be able to critically check and reflect on found solutions
found, to review some parts of the model or go through the modelling process again
if solutions do not fit the situation, to reflect on other ways of solving the problem
or if solutions can be developed differently and to generally question the model.
This description, as well as explanations by Blum and Leiß (2006), shows the spe-
cial role of this competency in the modelling process. More than other steps of the
modelling cycle, the step of validating requires substantial reflections on the solu-
tion processes and sometimes even the willingness to accept the need for a second
run through the complete modelling cycle in order to consider additional factors.
Hence, this step needs more reflection from a meta-perspective than any other step
of the modelling cycle. Therefore, the validating competency is also intertwined
with students’ abilities in self-regulation and self-evaluation as well as with their
abilities to monitor their work. Nevertheless, to be able to assess validating compe-
tencies, the presented study only considered those abilities formulated by Maaß
(2006). Especially the abilities to reflect on found solutions and on models and their
fit to the real situation were seen as central for the validating competency.
Even though this step is an essential part of the modelling process, empirical
investigations have shown that independent processes of validation are rarely to be
found in students’ modelling activities (Blum 2007). Using digital tools, however,
might promote and support these important mathematical activities like monitoring
and checking resulting solutions (Hegedus et al. 2016). Mousoulides (2011), for
24 Validating with the Use of Dynamic Geometry Software 279
example, found an increase in students’ engagement in self-evaluating through the

use of software features and tools of a DGS in his empirical study: “groups were
constantly questioning the validity of their solutions, and wondering about the rep-
resentativeness of their models” (p.117).
24.2.2 ffordances and Constraints of a Dynamic Geometry

A
Software
To better understand the role of a digital tool like a DGS when validating, it is
worthwhile to look at the affordances (Gibson 1986) a DGS offers and how they can
be used to validate. Affordances can be seen as “the perceived and actual properties
of the thing, primarily those fundamental properties that determine just how the
thing could possibly be used” (Norman 1988, p. 9). A DGS, for example, bears the
affordance to change a visual mathematical representation by relocating certain
points in the drag mode. But the affordances also often imply certain constraints.
Hammond (2010) gives the example of a sledge-hammer that can be used for break-
ing rocks, but at the same time constraints its user by its weight. But it is just the
weight and thus the constraint that makes the functionality of breaking rocks pos-
sible. This principle can easily be transferred to a DGS, for example, to the drag
mode. A DGS, for example, GeoGebra, makes it possible to construct the intersec-
tion of two lines as a point that dynamically changes its location when the lines are
moved so that it always remains an intersection. However, this can just be done if
the right tool is chosen, which is a point on an object and not a free point. However,
this constraint of having to differentiate between different sorts of points is just what
makes interactive constructions possible. Therefore, when using a DGS in a model-
ling process, it has to be kept in mind that the benefits of using the software can only
be fully exploited when students are able to cope with the respective constraints.
Speaking in terms of the cognitive load theory (Sweller et al. 2011), students might
thus be burdened with a higher intrinsic load due to the numerous interactive ele-
ments of the software.
Considering the explanations above, the research questions for the study presented
in this chapter are as follows:
1. Do students improve their competencies in validating throughout a modelling
intervention with or without the use of a DGS?
2. Is there a difference between a DGS and a control group (working with or with-
out a DGS) concerning the development of their competencies in validating
throughout a modelling intervention?
280 C. Hankeln
24.4 Methodological Background
To investigate the influence of working with a DGS while learning how to vali-
date, an intervention study with a pre-, post- and delayed post-test was conducted.
A total of 821 grade 9 students from 30 different classes from North Rhine-
Westphalia worked for 4 consecutive mathematics lessons on 4 open modelling
problems (e.g. on the Playground Task). Fifteen of these classes (413 students) used
the DGS GeoGebra throughout the whole intervention, while the remaining 15
classes (408 students) worked on the same tasks but with the use of pen and paper.
Playground Task
Dear city councillors,
This year, a new playground will be built in the
Wienburg park in the north of Münster. On the
right, you see a map of the park. In this map, the
four entrances of the park are marked. Only by
these four entrances you can enter and leave the
park. For us as city council it is important that
no family is disadvantaged, no matter at which
entrance it enters the park. Today we are going
to discuss the question:
Where should the playground be built?
All classes were taught at regular German high schools (gymnasiums). Students
were not previously trained in using the DGS GeoGebra, but it was ensured that
they had already worked with the programme before, as it is also required by the
North Rhine-Westphalian syllabus to do so. Therefore, the students were not experts
in using GeoGebra but not complete novices either.
In both groups, the lessons started with a short presentation of the problem by the
teacher, which was followed by approximately 25 min of students’ independent
work. In the last 15 min, some of the students presented their suggestions for a solu-
tion to the problem, and the whole class discussed the benefits and disadvantages of
the proposed model as well as differences between students’ work with regard to the
different steps of the modelling process. The four modelling lessons were taught by
the regular mathematics teachers who before had participated in a preparatory train-
ing for the study. All lessons took place following a detailed lesson plan, the imple-
mentation of which was spot-checked in all classes. After each lesson, all teachers
additionally filled out questionnaires about how the implementation succeeded.
Students’ answers to all tasks were collected, and in the DGS group, some computer
screens were filmed to analyse how the students were using the DGS.
The tasks used in the intervention could all be solved using non-digital tools like
ruler and compass or using features of the DGS GeoGebra. In the Playground Task,
for example, which was the third exercise in the intervention, students in both groups
were able to construct mid-perpendiculars to find a mid-point between four entrance
points of the park. In both groups, students were asked to take on the role of the
decision-makers (the city council). Therefore, it was an essential part of the exercise
to discuss the consequences of the found mathematical result as the (nearly) exact
mid-point of the four entrances lies in the middle of a lake. The students were thus
encouraged to consider other conditions than just the mathematical one of finding a
mid-point. Therefore, this exercise differs from a wrapped-up word problem, even
though students often started by ignoring the given situation and by searching for a
mathematical solution to the problem. The aim of this exercise was thus to show
students the importance of considering the real situation. Once students saw the
need for a reflection on the realistic context, they came up with assumptions con-
cerning, for example, the costs of construction and safety regulations. For instance,
they proposed using existing pathways instead of measuring as the crow flies. This
assumption inspired them to use polygonal chains as mathematical models.
Item Rock
Emilia calculated that the rock she has seen in her vacation is 8 m
tall. With the help of the photo on the right, judge if her result is
plausible.
To assess the progress in modelling competencies, students filled out an achieve-

ment test in modelling both directly before and after the intervention and with a
3-month delay. This test instrument was newly constructed using a multi-matrix
design and, with the help of short answer and multiple-choice questions, focused on
separate modelling competencies (cf. Hankeln et al. 2019). Item Rock and Item Way
to School give examples of how the validating competency has been measured: The
Item Rock is designed to assess students’ abilities to critically check a solution
found. Even though students are not asked to reflect on a solution they found them-
selves, the task assesses essential components of the validating competency.
Students must find a suitable measure to compare the rock to and argue why the
result cannot be plausible. They must set up a maximum height for the girl on the
picture, for example, 2 m, and compare her size to the rock. As the rock is three
times as tall as the girl, who measures at most 2 m, the given result does not seem
to be plausible.
Item Way to School, in contrast, focuses more on the abilities to generally ques-
tion a given model. In this task, students were asked to reflect on if the simple addi-
tion of the given two numbers is an adequate way to model the real situation. In
contrast to Item Rock, which aims at the critique of the plausibility of a given result,
282 C. Hankeln
this item focuses more on the reflection if constraints introduced by the real situa-
tion have been adequately integrated into the model used. A correct answer to Item
Way to School is that Melissa is not always right as there is no information given
where both students are living in relation to the school. She is just calculating the
maximal distance possible.
Item Way to School

Melissa lives 3 km away from school, Piotr has a 4.5-km-long way to school.
Melissa says: “So I have to drive 7.5 km to visit you!”
Is Melissa always right? Give a short explanation.
By using a variety of tasks that all assessed different facets of the validating
competency, the test tried to cover as many aspects as possible from the definition
formulated by Maaß (2006). In addition to validating, the test instrument also mea-
sured students’ abilities to simplify, mathematize and interpret. With the help of
item response theory, different models were calculated and compared to each other.
The model with the best model fit (e.g. with the smallest information criteria AIC1
and BIC) was a four-dimensional between-item model in which each of the four
competencies was represented by a separate dimension. Therefore, the data were
analysed for each competency separately. This chapter only draws on data concern-
ing the validating competency, which could be measured with an EAP/PV reliabil-
ity of 0.66 in the pre- and delayed post-test and 0.75 in the post-test. The EAP/PV
reliability is calculated as the ratio of the modelled variance to the observed vari-
ance and therefore comparable to Cronbach’s alpha. This relatively low reliability
can be explained by the variety of items that have been used as the two examples
presented above have shown. Nevertheless, the reliability is sufficient to conduct a
comparison of groups, which was the aim of the study. Two independent raters rated
40% of the open-ended tasks. The interrater reliability lay in a range of 0.81 and
0.95 and can thus be judged as very good.
As classes, not individual students, were assigned to the DGS and the control
group, the data is hierarchically clustered. Students belonging to the same class are
likely to be more similar to each other than to students belonging to different classes.
Ignoring this circumstance might lead to a distorted calculation of standard errors
and thus to incorrect tests of significance. To avoid this distortion, a type = complex
analysis in Mplus was used. Therefore, the data were analysed using regression
models in which students’ achievements at the second and third points of measure-
ment were controlled for their achievement at the first point of measurement. By
this means, it was possible to analyse differences between the DGS and the control
group in the post-test and follow-up test independently from differences that already
existed at the beginning of the intervention.
1
AIC (Akaike Information Criterion) defined by AIC = − 2 ln (L) + 2np and BIC (Bayes Information
Criterion) by BIC = − 2 ln (L) + ln (N) ∙ np, when np is the number of estimated parameters and
N the sample size. A smaller value indicates a better fit.
24.5 Findings
24.5.1 Do Students Improve Their Validating Competencies?
To evaluate the intervention, the mean achievements in the different points of mea-
surement were compared using Wald’s t-test. This test was used since it is imple-
mented in Mplus and thus allowed to control for the clustered data as described
above. The intraclass correlation, which was up to 0.08, confirmed the necessity of
using this more complex method than a simple ANOVA.
Figure 24.1 shows that both groups were able to improve their validating competen-
cies throughout the intervention. The ability estimates are fixed at the mean of all three
points of measurements and can thus not be interpreted in themselves. They increase in
both groups significantly from pre- to post-test, as Table 24.1 shows. Even though in
both groups the mean achievement in the delayed post-test is lower than in the post-test,
this difference is not significant, so that the improvement seems to be sustainable within
a 3-month delay. In comparison to the mean achievement in the pre-test, both groups
show a significant higher achievement in the delayed post-test. But all improvements
are just of a small effect size, which is not unexpected in a relatively short intervention.
Fig. 24.1 Development of ability estimates of DGS and control group (mean and standard devia-
tion (SD) in the pre-, post- and follow-up test)
Table 24.1 Development of the sub-competency validating in both groups

Criterion Group diff. t p d
Pre to post Control +0.23 13.08 <0.001 0.31
DGS +0.21 10.22 <0.01 0.26
Post to delayed post Control −0.08 1.08 0.30 0.09
DGS −0.05 0.59 0.44 0.06
Pre to delayed post Control +0.16 6.79 <0.01 0.20
DGS +0.16 16.09 <0.001 0.21
284 C. Hankeln
Table 24.2 Regression analysis for the post- and delayed post-test with pre-test and affiliation to
the DGS group as predictors
Criterion Coefficient b SE β p
Post-test Intercept −0.02 0.07 −0.02 0.79
Pre-test 0.37 0.04 0.32 <0.001
DGS group 0.05 0.09 0.06 0.59
Delayed post-test Intercept 0.04 0.07 0.05 0.59
Pre-test 0.44 0.04 0.37 <0.001
DGS group 0.07 0.07 0.08 0.31
24.5.2 I s There a Difference Between the DGS and the Control

Group?
As Table 24.2 shows, students’ achievement in validating in the pre-test is a signifi-

cant predictor of students’ achievement at both the second and the third points of
measurement. This is not surprising as students who achieve a high ability estimate
in the pre-test also achieve a high ability estimate at the other two points of mea-
surement. More interestingly, the affiliation to the DGS group has no significant
influence on students’ achievement. Apparently, the use of a DGS did not lead to a
different development of the validating competency than using traditional tools like
compass and ruler. This implies on the one hand that the use of a DGS did not favour
the development of the validating competency more than traditional tools. On the
other hand, it did not hinder it either. Regarding the constraints that the use of a
DGS implies, it could have been possible that students’ learning processes concern-
ing the use of the software interfered with their progress in modelling.
24.6 Summary and Discussion
The analysis has shown that the development of the validating competency in the
DGS group did not differ from the control group, while both groups lastingly and
significantly improved their competencies. This is in accordance with existing
research concerning the influence of DGS on achievements. As Drijvers et al. (2016)
state, the benefit of using technology in mathematics education does not appear to
be very strong, at least from the view of quasi-experimental studies. But one has to
keep in mind that many studies investigating the use of technology in mathematics
education often used qualitative methods and just a few participants. Additionally,
most studies that investigated the relationship between the use of a DGS and
achievement focused on geometrical knowledge (Chan and Leung 2014). There are
almost no quantitative studies focusing on modelling competencies. Some studies
used application tasks in their achievement tests (e.g. Dimakos and Zaranis 2010;
Gawlick 2002) but did not evaluate those tasks separately. The quantitative study
presented in this chapter is a first attempt to gain an impression of how using a DGS
might influence “traditional” modelling competencies necessary to solve modelling

tasks without a digital tool. Here, the digital tool was only used during the interven-
tion as a tool to learn but could not be used for the modelling test. On the one hand,
this method is more representative for the way a DGS is used in class, at least in
Germany. On the other hand, it is possible that the DGS group has developed differ-
ent aspects of the validating competency that are not covered by the definition by
Maaß (2006) on which the test items were based. As students in the DGS group
often used DGS features to control or adapt their models, they might have, for
example, developed higher abilities to use the software’s interactive feedback to
control their results. The comparison of the two groups should thus always be inter-
preted against the background of the used test items.
Another aspect concerns the role of tool competencies when working with a
DGS on modelling problems. Since learning to use a digital tool can itself be a
complex process, using a not so familiar tool might be a disadvantage to students
who are or feel not so secure in operating this software. Such indications could even
be found in the presented study as in the DGS group but could not be explained in
detail in this chapter due to space limitations: the inclusion of students’ self-efficacy
in relation to the used software was also a significant predictor of students’ achieve-
ment in validating (see Hertleif 2017). A careful look is thus necessary to better
understand to which extent students already must have learnt how to operate a soft-
ware to fully exploit its benefits for modelling.
A third aspect lies in the choice of tasks used in the intervention. As it was the
aim to attribute possible differences between the DGS and the control group to the
use of the digital tool, the tasks in the intervention did not exploit the full potential
a DGS might offer. It remains unclear how modelling competencies might be devel-
oped when students are using a digital tool for tasks that they cannot solve without
the tool. The study presented in this chapter has shown that the (task, tool) pair
(Doerr and Pratt 2008), whereby the tasks remain comparable to tasks without tools,
does not foster students’ validating competencies more than tasks and non-digital
tool pairs, at least for geometric modelling tasks and a DGS as tool.
References
Berry, J. S., Burghes, D. N., Huntley, I. D., James, J. J. G., & Moscardini, A. O. (Eds.). (1986).
Mathematical modelling methodology, models and micros. Chichester: Ellis Horwood.
Blum, W. (2007). Mathematisches Modellieren – zu schwer für Schüler und Lehrer? Beiträge zum
Mathematikunterricht, 3–12.
Blum, W., & Leiss, D. (2006). “Filling up” – The problem of independence-preserving teacher
interventions in lessons with demanding modelling tasks. In M. Bosch (Ed.), CERME-4 –
Proceedings of the fourth Conference of the European Society for Research in Mathematics
Education. Sant Feliu de Guíxols: Universitat Ramon Llull.
Blum, W., Galbraith, P. L., Henn, H.-W., & Niss, M. (2007). Modelling and applications in math-
ematics education. The 14th ICMI study. New York: Springer.
286 C. Hankeln
Chan, K. K., & Leung, S. W. (2014). Dynamic Geometry Software improves mathematical achieve-
ment: Systematic review and meta-analysis. Journal of Educational Computing Research,
51(3), 311–325.
Dimakos, G., & Zaranis, N. (2010). The influence of the Geometer’s sketchpad on the geometry
achievement of greek school students. The Teaching of Mathematics, 8(2), 113–124.
Doerr, H. M., & Pratt, D. (2008). The learning of mathematics and mathematical modeling. In
mathematics: Syntheses, cases, and perspectives. Research syntheses (pp. 259–285). NCTM:
Charlotte.
Drijvers, P., Ball, L., Barzel, B., Heid, M. K., Cao, Y., & Maschietto, M. (2016). Uses of technology
in lower secondary mathematics education. A concise topical survey. Cham: Springer Open.
English, L. D., Ärlebäck, J. B., & Mousoulides, N. (2016). Reflections on progress in mathematical
modelling research. In Á. Gutiérrez, G. C. Leder, & P. Boero (Eds.), The second handbook of
research on the psychology of mathematics education (pp. 383–413). Rotterdam: Sense.
Gawlick, T. (2002). On dynamic geometry software in the regular classroom. ZDM Mathematics
Education, 34(3), 85–92.
Gibson, J. J. (1986). The ecological approach to visual perception. New York: Psychology Press.
Hammond, M. (2010). What is an affordance and can it help us understand the use of ICT in educa-
tion? Education and Information Technologies, 15(3), 205–217.
Hankeln, C., Adamek, C., & Greefrath, G. (2019). Assessing sub-competencies of mathematical
modelling with the help of Item Response Theory – Presentation of a new test instrument. In
G. A. Stillman & J. P. Brown (Eds.), Lines of inquiry in mathematical modelling research in
education. Cham: Springer.
Hegedus, S., Laborde, C., Brady, C., Dalton, S., Siller, H.-S., Tabach, M., & Moreno-Armella, L.
(2016). Uses of technology in upper secondary mathematics education. Cham: Springer Open.
Hertleif, C. (2017). Dynamic geometry software in mathematical modelling: About the role of
programme-related self-efficacy and attitudes towards learning with the software. In A. Gilles
& J. Trgalova (Eds.), Proceedings of the 13th International Conference on Technology in
Mathematics Teaching (pp. 124–133). Lyon, France. (hal-01632970).
tives. In G. A. Stillman, W. Blum, & M. S. Biembengut (Eds.), Mathematical modelling in
education research and practice (pp. 129–149). Cham: Springer.
Lingefjärd, T., & Holmquist, M. (2003). Learning mathematics using dynamic geometry tools.
In S. J. Lamon, W. A. Parker, & K. Houston (Eds.), Mathematical modelling: A way of life
113–142.
Mousoulides, N. G. (2011). GeoGebra as a conceptual tool for modelling real world problems.
In L. Bu & R. Schoen (Eds.), Model-centered learning: Pathways to understanding using
GeoGebra (pp. 105–118). Rotterdam: Sense.
Norman, D. A. (1988). The psychology of everyday things. New York: Basic Books.
Sweller, J., Ayres, P., & Kalyuga, S. (2011). Cognitive load theory. New York: Springer.
Chapter 25
Using a Mathematical Modelling Activity
to Assist Students to Make Sense of a Limit
Theorem in Trigonometry
Toshikazu Ikeda and Max Stephens
Abstract The aim of this paper is to investigate whether mathematical modelling,

interpreted as a series of interactive translations, can also be effective to assist stu-
dents to construct mathematical knowledge. We describe a series of possible teach-
ing activities in which the theorem of a limit of a trigonometric function can be
approached through solving a real-world problem how to determine the shortest air
route between two cities. In this way, we show that these transformations have the
potential to build up mathematical concepts. These findings are intended to form a
framework for subsequent experimental teaching.
Keywords Limit · Mathematical knowledge · Mathematical modelling ·

Trigonometric function
25.1 Background
In our previous chapters (Ikeda and Stephens 2015, 2017), we introduced a perspec-
tive where mathematical modelling is interpreted as interactive translations, not
between two fixed worlds – between a real world and a mathematical world – but
actually among plural worlds. In this perspective, we have argued that plural worlds
are not simply the result of arbitrarily changing mathematical representations, but
arise fundamentally because of comparisons and contradictions between competing
T. Ikeda (*)
College of Education and Human Sciences, Yokohama National University, Yokohama, Japan
e-mail: toshi@ynu.ac.jp
M. Stephens
Melbourne Graduate School of Education, The University of Melbourne,
Melbourne, Australia
e-mail: m.stephens@unimelb.edu.au

288 T. Ikeda and M. Stephens
perspectives. These comparisons or contradictions can deepen students’ modelling

activities once we see the modelling process as operating in a plurality of worlds.
These ideas provide indispensable clues to further promote modelling activities.
In this chapter, we discuss one further possibility, namely, where mathematical
modelling may also be effective to construct or at least point to new mathematical
knowledge. Constructing mathematical knowledge starting from modelling has
been discussed as one of the perspectives of teaching modelling. In the international
classification of modelling perspectives by Blum and Niss (1991), “Prompting a
mathematics learning argument” was identified. Conceptual modelling was also
identified by Kaiser and Sriraman (2006). More recently, modelling as a vehicle for
something else has been discussed by Niss (2018), identifying “motivation, sense-
making, underpinning and consolidation of concepts, methods and results” as one
of the representative examples of this purpose. In these different perspectives,
introduction of new mathematical concepts and methods starting from modelling is
focused on. Our research question is: What is the effect of applying the framework
of interactive translations among plural worlds to introducing and developing
mathematical concepts starting from a modelling problem to determine the shortest
air route between two cities? Specifically, we develop a series of activities in which
mathematical knowledge about a fundamental theorem of the limit of a trigonomet-
ric function can be approached through solving a real-world problem to determine
the shortest air route between two cities.
To investigate the possibilities of this new perspective, we adopt the following
procedures:
1. We review general ideas about constructing a mathematical knowledge and anal-
yse the process of constructing mathematical knowledge starting from model-
ling. Then we focus on a specific problem of finding the shortest route between
two points to a fundamental theorem of the limit of a trigonometric function and
set up how many plural worlds are effective to analyse.
2. We illustrate the process from solving this real-world problem into constructing
new mathematical knowledge.
3. We interpret this construction process as an interactive translation among plural
worlds and clarify the effect to apply the framework of interactive translations
among plural worlds.
25.2 onstructing Mathematical Knowledge Starting

C
from Modelling
25.2.1 Theoretical Perspectives
Constructing mathematical knowledge starting from modelling has been discussed

as one of the perspectives of teaching modelling. For example, Kaiser and Sriraman
(2006) identified “conceptual modelling” as one form of the category “educational
modelling” which is concerned with concept introduction and development.
25 Using a Mathematical Modelling Activity to Assist Students to Make Sense… 289
Regarding this perspective, Treffers (1987) distinguished the two aspects of math-
ematization: horizontal and vertical. Further, based on the idea of “model-of/model-
for shift” by Streefland (1991), Gravemeijer (1999) identified four types of activity:
(1) activity in the task setting, (2) referential activity, (3) general activity and (4)
formal mathematical reasoning.
In these frameworks, mathematization is restricted to two types: horizontal and
vertical. Our framework treats multiple mathematizations (in our terminology,
translations) from one world into another world and focuses on the comparisons and
contradictions between competing perspectives among plural worlds. We identify
the effect of our framework in constructing mathematical knowledge starting from
modelling.
25.2.2 rom Solving a Problem to Constructing Mathematics

F
Knowledge Objective
At first, we consider mathematical activity as two mutually related basic components:

solving a problem and constructing mathematics knowledge. Solving a problem starts
by clarifying the problem arising in a particular world, possibly in the real world.
Mathematical modelling is a prime example of solving a problem. The initial problem
arises from the real world. After solving a problem, it is important for students to
reflect on the process and consider its relevance to what they have experienced or to
the mathematics they may have learned at elementary or secondary school. In this
respect, we refer to the idea of Hesse (1966), who threw light on the role of models in
science. Here models are treated as physical/mental entities for comparing and con-
trasting and are also used as places from which to draw out analogies (Ikeda and
Stephens 2015). We argue that constructing mathematical knowledge starts with this
reflection, comparing recently established knowledge with prior knowledge and clari-
fying their relevancies. Or, it might be finding previously unknown elements of
recently established knowledge. These unknown elements may help to clarify whether
the applicable range of established knowledge can be widened or not. If it is not
possible to widen the range of established knowledge, then it is important to delin-
eate the borders between applicable ranges and non-applicable ranges.
25.2.3 he Limit in a Trigonometric Function

T
as the Mathematical Knowledge Objective
Trigonometry is used throughout mathematics, especially in calculus. The key to

trigonometry in calculus is finding the derivatives of the sine, cosine and related
functions. Almost everything else follows from those. Derivatives are defined in
terms of limits, so that means students need to know something about limits and
trigonometric functions.
sin θ
The fundamental theorem of the limit in trigonometric function limθ →0 =1
θ
is a foundation to study differentiation of trigonometric function. In Japanese text-
books, this theorem is usually introduced without explanation, such as occurred in a
textbook by Iidaka and Matsumoto (2004) where they say: “The basic formula to
sin θ
consider the limit of trigonometric functions is as follows: limθ →0 = 1. At first,
θ
let’s investigate this theorem of a limit”. A consequence of this abstract approach is
that few students can understand the reasoning behind this formula. Further, the
proof of this formula is frequently done by geometric analysis involving areas of
triangles and sectors of circles. However, some argue that this proof involves circular
reasoning; therefore, it may not be mathematically consistent (Hitotsumatsu 1962).
25.3 onstructing the Knowledge Objective

C
from the Modelling Activity
25.3.1 Overview of Developmental Material
The specific focus of this study is to show how a mathematical modelling activity can
be used to assist students to make sense of a limit theorem in trigonometry which, as
mentioned, is normally treated abstractly. Four different “worlds” will be accessed
starting from modelling. The first is a real world; the second is a concrete operational
world where students manipulate objects; the third is a geometric operational world
which allows them to draw/measure a model; and the fourth is a symbolic operational
world which represents the situation mathematically. A series of activities focused on
in this study are composed of the following four phases. The problem in phases 1 and
2 arises from a real world, and the problem in phases 3 and 4 arises in students’ inner
minds. In phase 3, it is expected students become willing to re-organize and draw upon
their own mathematical knowledge through comparing such recently established
knowledge with prior knowledge and clarifying their relevancies.
Phase 1: introduce a real-world problem such as: “Why is our airplane over the
Baltic Sea?” which is derived from a possible flight path from Tokyo to London,
knowing that the shortest (straight line) route between Tokyo and London on a
world map does not pass through the Baltic Sea. This raises the question: How
do we find the shortest route between two points on the surface of the Earth?
Phase 2: manipulate the globe to investigate possible circular routes between Japan
and London, and consider why the shortest route takes the Great Circle route by
drawing several circles passing through two known points. As a result, it is pro-
posed that as the radius gets bigger, the arc between two points is getting closer to
the straight line passing through two points. Thus initial knowledge is developed.
Phase 3: asking the question such as: “Are there any similar cases to this fact in our
prior experiences?” Two of these are the equivalent area transformation from a
circle to a parallelogram, and increasing the number of congruent semicircles on

a diameter to examine their relationship with the circumference of a given semi-
circle. These will be considered specifically below.
Phase 4: illustrate how the underlying structure among these plural worlds can be
sin θ
related to the fundamental theorem of the limit, limθ →0 =1.
θ
25.3.2 A Prospective Scenario for Teaching and Learning
25.3.2.1 The Opening Problem or Situation
The following problem situation is based on a hypothetical air journey.

This raises the question: How do we find the shortest route between two points on
the surface of the Earth? Students can be led to see that the problem arises from
comparisons and contradictions between an experience in a real world and prior
knowledge derived from a two-dimensional world map.
I went to London by airplane from Tokyo. During the flight, when I looked
through the window, I could see the Baltic Sea. It was wonderful and beautiful,
but I wondered why our airplane is over the Baltic Sea. Then, I checked a world
map and noticed that the shortest (straight line) route between Tokyo and
London on the world map doesn’t pass through the Baltic Sea (Fig. 25.1). Why
does our airplane take this circuitous route?
Fig. 25.1 Shortest route on a map

25.3.2.2 Solving a Real-World Problem
We illustrate the prospective scenario. By manipulating another model, namely, a

globe, we want students to notice that the shortest route in a two-dimensional
world map is different from the shortest route on a globe. When manipulating the
globe, it is important for the teacher to let students note carefully the assumptions
that the earth is a sphere, also ignoring the fact that the earth is revolving on its axis
when the plane is in the air. After manipulating a globe several times, it is expected
for students to understand the shortest route between Japan and London is not a
straight line but a curved line (arc). But at first sight it is not easy to see which
curved line will be shortest because there are many possible curved lines joining
two points, namely, Japan and London. The teacher can ask students: “Which of
these is the shortest?”
To consider the shortest arc of cutting surface within a two-dimensional plane,
students can be asked to draw many circles which pass through two points and to
guess which arc is shortest. We want students to connect what is being depicted on
a two-dimensional page with circles of increasing radius to circles of different
radius on the surface of the Earth (or in the case of an aircraft at a fixed and rela-
tively small distance above the surface of the Earth), as shown in Fig. 25.2.
From Fig. 25.2, students are expected to ascertain that as the radius gets bigger,
the arc between two points (A, B) gets closer to the straight-line segment passing
through two points (A, B). In this stage, it should be noticed by students or the
teacher that students can ignore the fact the arbitrary “great arc” cannot be made
smaller than a fixed amount when two points on the earth are fixed and consider the
problem in the ideal situation that the earth can become bigger without end. By ask-
ing students what is the biggest circle derived from cutting the sphere, students will
notice that the biggest circle has its centre at the centre of the sphere. Then it is
Fig. 25.2 Drawing circles

of increasing radius
expected for students to conclude that the shortest air route is the arc of a Great
Circle passing through the centre of the sphere. Using the globe, it is confirmed by
students that the shortest route can be expected to pass close to the Baltic Sea.
Regarding the initial question why the trajectory of the plane should veer off the
straight-line path between two points on the earth, it is expected that students will
explain that the earth is not flat and make the point that two-dimensional maps nec-
essarily distort the shortest route.
25.3.2.3 Reflecting Similar Cases and Abstracting Substantial Properties
To re-organize and draw upon students’ own mathematical knowledge, it is impor-

tant to establish the connections by reflecting upon similar cases to abstract substan-
tial properties. The teacher can focus on the fact derived from previous activity that
as the radius is getting bigger, the arc between two points (A, B) is getting closer to
the straight line passing through two points. Asking the question for students such
as “Are there any similar cases to this fact in our prior experiences?” is the starting
point to re-organize and draw upon students’ own mathematical knowledge.
The following two examples can be presented by the teacher. The first is the
equivalent transformation from circle to a “parallelogram” as shown in Fig. 25.3.
Students have met this example before in finding the area of a circle. As the central
angle is made smaller, the number of sectoral arcs increases and eventually approxi-
mates two straight horizontal lines, making a parallelogram whose area is π r × r.
The second example as shown in Fig. 25.4 allows the teacher to ask students: “If we
increase the number of congruent semicircles resting on the diameter, what happens
about the sum of the lengths of the semicircular arcs?” Some students who have not
met this problem before might answer that it is getting closer to the length of the
diameter of the original half circle as shown in Fig. 25.4.
However, students who have met this problem before might answer: “Even
as the number of semicircles increases, the total length of these half circles
must remain constant, as shown by manipulating the formula such as
 πr   πr   πr 
“ π r = 2   = 4   =  = n   ”. These students are in a position to notice
2
   4  n 
that this example is different in one key respect from the original air route problem
Fig. 25.3 Transformation of area of circle

Fig. 25.4 Sequence of semicircles
Fig. 25.5 Substantial

structure underlying two
phenomena
and the transformation of circle as shown in Fig. 25.3. The example shown in
Fig. 25.4 is intended to remind students that decreasing size of each semicircular arc
length does not necessarily change the total length of these arc lengths. However,
the transformation of circle in Fig. 25.3 appears to have a similar structure to the
original problem. We want students to be able to explain that in the original prob-
lem, phenomenon A, as the radius r is getting bigger, the arc between two points is
getting closer to the straight line passing through two points (A, B), whereas, in the
transformation of a circle into a “parallelogram”, phenomenon B, as the central
angle θ gets smaller, the arc between any two points is getting closer to the straight
line. At this stage, it is expected for students to find out the differences between
them, namely, phenomenon A where we increase the radius r and phenomenon B
where we decrease the angle θ. Then, we have a key teaching point where the teacher
asks students: “How can we see two phenomena as the same? Can we integrate the
two phenomena?” In the original problem and in the sequence of semicircles, the
radius is changing, whereas in the original problem and in the transformation of
circle, the central angle θ is decreasing. By identifying the decrease of central angle
θ as essential, students will be able to see a similar structure between the original
problem and the transformation of circle as shown in Fig. 25.5.
From Fig. 25.5, by asking students to represent the length of segment AB and arc
(AB) using radius “r” and angle “θ”, the following two formulas can be derived
from students:
AB = 2 ⋅ r ⋅ sin θ
arc ( AB) = 2π r ⋅ 2θ / 2π = 2rθ
As angle θ becomes smaller and the arc AB is getting closer to the line segment
AB, students can get a sense of the genesis of the following fundamental theorem of
the limit of a trigonometric function:
2r ⋅ sin θ sin θ
limθ →0 = limθ →0 =1
2rθ θ
25.4 I nteractive Translations in Solving a Real-World

Problem
The initial question “Why is our airplane flying over the Baltic Sea?” was derived
from a paradox or contradiction between the real world and a concrete operational
world. In order to analyse the shortest route between Tokyo and London, it is neces-
sary to perceive that the route can be interpreted as the arc of circle above the Earth’s
surface, but with its critical features identical to the corresponding arc on the Earth’s
surface. Starting with the longest arc located on or above the circle of latitude, other
arcs of decreasing length belong to circles of increasing radius, ending with the
shortest arc belonging to a circle with the largest available radius, that is, belonging
to a Great Circle. From this result, a translation from a concrete operational world
into a geometrical operational world is promoted. By constructing multiple circles
passing through two points, we intend for students to see that as the radius gets big-
ger, the length of the arc becomes smaller and smaller as shown in Fig. 25.2. This
allows them to see why the shortest route is located on a Great Circle of the sphere.
This requires a translation from a geometrical operational world into a concrete
operational world. Using the globe, it makes sense that the shortest route passes
near the Baltic Sea. Analysing the globe in a concrete operational world acts as an
intermediate role between perception in a real world and geometrical construction
in a geometrical operational world. In other words, analysing the globe in a con-
crete operational world can build up the further abstract model in a geometrical
operational world and verify the result in a real world. This point supports the find-
ing of our previous chapter (Ikeda and Stephens 2017). Regarding the modelling
task of the Great Circle route, it is usually explained intuitively and the distance
calculated by using three-dimensional coordinate systems (Hunt 1999). However, in
this chapter, we focus on reasoning deductively why a Great Circle route is the
shortest through interactive translations among three worlds based on the fact the
straight line between two points is the shortest and connecting this to the limit of a
trigonometric function.
25.4.1 I nteractive Translations in Constructing Mathematical

Knowledge
The second process is constructing mathematical knowledge. Based on the recently

established knowledge “when the radius is getting bigger, the arc is getting closer to
the straight line”, our aim has been to find the similar models to re-organize and
draw upon a person’s own mathematical knowledge. The first was the equivalent
transformation from circle to parallelogram in a geometrical operational world as
shown in Fig. 25.3. Second is the transformation of a semicircle in a geometrical
operational world as shown in Fig. 25.4. By transforming the semicircle in a geo-
metrical operational world into a symbolic operational world, we can find that the
repeated transformations of a semicircle are different from the shortest route
between two points. Then, the substantial property is analysed between the shortest
route between two points and the area transformation of a circle. By identifying the
decrease of central angle θ as essential, students can be assisted to see that “as the
angle θ is getting smaller, the arc AB is getting closer to the segment AB”.
However, we note that it is not easy to represent our two phenomena (the shortest
route between two points and the area transformation of a circle) in a symbolic
operational world, while the progressive transformation of a semicircle can be
expressed in a symbolic operational world. It is, however, possible to ask: Is it pos-
sible to represent a mathematical formula about two phenomena just like the case of
progressive transformation of a semicircle? We can see that the example of progres-
sive transformation of a semicircle plays a crucial role to promote the operational
abstraction about the two phenomena from a geometrical operational world into a
symbolic operational world. As a result, the limit of the trigonometric function can
be abstracted and made sense of by students. Comparing relations between a geo-
metrical operational world and a symbolic operational world of the three examples
is effective to make a connection to the limit of the related trigonometric function.
However, when a teacher looks at these interactive translations among plural worlds,
it is necessary from a pedagogical point of view to identify those which are likely to
be meaningful and effective for students.
25.5 Conclusion
The aim of this chapter was to investigate whether our theoretical perspective in
which mathematical modelling is interpreted as a series of interactive translations
can be useful and sense-making for students in constructing and integrating math-
ematical knowledge. As one means to that end, we used a modelling situation help-
ing students to make connections between the requirements of the modelling task
and integrating distinct elements of mathematical knowledge that are rarely inter-
related. Concretely, we developed a series of interrelated activities in which math-
ematical knowledge about a fundamental theorem of the limit in trigonometric
functions could be approached through solving a real-world problem, how to deter-

mine the shortest route between two cities in a world.
By analysing prospective students’ learning activity from the perspective of
interactive translations among plural worlds, two points can be made evident. First,
we clarified that analysing the globe in a concrete operational world can be effec-
tive to build up the further abstract model in a geometrical operational world and
check the answer of a real world. Second, in constructing mathematics knowledge,
we compare knowledge established in this modelling situation with students’ prior
knowledge and draw attention to their connections. Moving now within a mathe-
matical world, our goal is to re-organize and draw upon a person’s own mathemati-
cal knowledge consisting of newly appreciated mathematical connections and
relationships to provide a context and methods for investigating a fundamental theo-
rem of the limit of a trigonometric function.
Our framework comes into line with Gravemeijer’s idea (1999) presenting a
“model-of/model-for shift” where both the model and the modelling facilitate
reflection. Multiple mathematizations (in our terminology, translations) are expected
among plural worlds. The principle underpinning this framework is that mathemat-
ics can be abstracted repeatedly from one world to another world. Further, our focus
is on the comparisons and contradictions between competing perspectives among
plural worlds which will promote constructing mathematical knowledge as well as
modelling. The intention is not only to promote the developmental progression but
also to deepen students’ mathematical knowledge by connecting and integrating the
outcomes constructed in each world. This is expected to require the development of
different kinds of material appropriate to secondary school level to investigate in
real classrooms the possibilities and challenges of this new perspective where math-
ematical modelling can be enacted with, and by, students as a series of interactive
translations.
References
Blum, W., & Niss, M. (1991). Applied mathematical problem solving: Modelling, applications,
and links to other subjects. Educational Studies in Mathematics, 22, 37–68.
Gravemeijer, K. (1999). How emergent models may foster the constitution of formal mathematics.
Mathematical Thinking and Learning, 1, 155–177.
Hesse, M. B. (1966). Models and analogies in science. Notre Dame: University of Notre Dame
Press.
Hitotsumatsu, S. (1962). Introduction to analysis. Tokyo: Shokabo Press (In Japanese).
Hunt, R. (1999). Time and motion. https://plus.maths.org/content/os/issue7/features/greatcircles/
index. Accessed 30 Apr 2018.
Iidaka, S., & Matsumoto, S. (Eds.) (2004). New mathematics III. Tokyo: Tokyo Syoseki (In
Japanese).
Ikeda, T., & Stephens, M. (2015). Reconsidering the roles and characteristics of models in math-
ematics education. In G. A. Stillman, W. Blum, & M. S. Biembengut (Eds.), Mathematical
modelling in education research and practice: Cultural, social and cognitive influences
Ikeda, T., & Stephens, M. (2017). Modelling as interactive translations among plural worlds –
Experimental teaching using the night time problem. In G. A. Stillman, W. Blum, & G. Kaiser
(Eds.), Mathematical modelling and applications: Crossing and researching boundaries in
mathematics education (pp. 349–358). Cham: Springer.
Niss, M. (2018). Advances in research and development concerning mathematical modelling in
mathematics education. In F.-J. Hsieh (Ed.), Proceedings of the 8th ICMI-East Asia Regional
Conference on Mathematical Education (Vol. 1, pp. 26–36). Taipei: EARCOME.
Streefland, L. (1991). Fractions in realistic mathematics education. A paradigm of developmental
research. Dordrecht: Kluwer Academic.
Treffers, A. (1987). Three dimensions: A model of goal and theory description in mathematics
instruction—The Wiskobas project. Dordrecht: D. Reidel.
Chapter 26
Exploring Habits and Habitus
of Biomedical Students with Modelling
Tasks
Hanti Kotze
Abstract This chapter aims to explore the potentialities of mathematical modelling

tasks that underscore first year biomedical technology students’ habits and habitus.
When students are engaged in modelling tasks for the first time, ingrained habits –
imposed by traditional teaching and learning approaches – are challenged and, in
turn, affect students’ future expectations and aspirations, a notion that Bourdieu
calls habitus. Three questionnaires were employed to analyse students’ perceptions
of four modelling tasks, spread over a semester. The novel modelling environment
was an impetus for change; however, modification of habits and a refinement of
habitus were marginally demonstrated.
Keywords Characteristics · Habits · Habitus · Semantic differential questionnaire

· Sense-making
26.1 Introduction
Learning experiences influence students’ habits (Dewey 1938), and in turn, habits
influence students’ future expectations, aspirations and career outlooks or habitus
(Bourdieu 1990). Habits and habitus are developed through socialisation and shape
the way students act, think and perceive the world. Students’ long-term behaviours
develop into a framework of thoughts which structure the way they think about
mathematics and its potential value in the workplace (Resnick 1989).
In many South African schools, traditional teaching and learning approaches pre-
vail, and therefore, model-eliciting activities are rarely implemented in classrooms
(Biccard 2010). Subsequently, first year students enter university with habits shaped
by traditional classroom activities that focus mainly on procedural knowledge.
H. Kotze (*)
Applied Physics and Engineering Mathematics, University of Johannesburg,
Johannesburg, South Africa

300 H. Kotze
Paradoxically, the biomedical industry requires technologists who can collect, anal-
yse and interpret real-world data where different knowledge bases are integrated
through interdisciplinary and team-based research (Humphrey et al. 2005).
According to these authors, biomedical contexts integrate aspects of biology, bio-
mechanics, biomedical optics, mathematical modelling and physiology. The inter-
disciplinary nature and cognitive challenges that biomedical students will encounter
in their careers raise concerns when classroom experiences do not impose habits
reflective of workplace demands. It could also be argued that traditional classroom
experiences may instil a school habitus that is fundamentally different to the percep-
tions deemed applicable to a career habitus. This chapter aims to explore the poten-
tialities of modelling tasks in cultivating habits and inspiring habitus.
26.2 Theoretical Perspectives
In the South African context, there is a need for more classroom experiences that
promote enriched thinking, knowledge that is contextualised and solutions that are
meaningful. Inasmuch as specialised competencies, strategies and knowledge are
vital, a good problem-solver also requires “the habits and dispositions of interpreta-
tion and sense-making” (Resnick 1989, p. 39). For Dewey (1944, p. 27), a habit is
the “more or less fixed way of doing things” and includes desires and attitudes
ingrained in traditions. A new learning environment and/or conditions may spark
the formation of different habits that are aligned to the new circumstances. However,
habit formation is initially in a phase of inertia as deep-rooted traditions may persis-
tently guide or lead actions and behaviours.
Complementary to the notion of habits, Bourdieu (1990) proposes habitus as the
framework embedded in an institutional field that influences the way students
behave and think. Habitus is embodied in sociocultural experiences that shape stu-
dents’ expectations and aspirations. Bourdieu refers to capital as the resources
available in the educational field which help students navigate through their studies
(Dumais 2002). Capital also refers to educational credentials and the appreciation
thereof. Students orientate themselves according to the available capital in a field,
and therefore, habitus becomes a reflection of attainable opportunities. As such, the
characteristics of a field can be traced back to its capital which includes cultural
capital (including academic success) and social capital. Habitus is also porous;
when certain restrictions in the field are lifted and new capital becomes available,
new experiences can change students’ views of the world and set an adapted habitus
in motion.
The social organisation of a modelling environment can contribute to students’
learning, expertise, motivation and confidence which affects their orientation and
outlooks, not only while they learn but also in future practices, thus developing
habitus. This chapter explores habits and habitus as embodied in biomedical tech-
nology students’ behaviours in, and expectations of, a mathematical modelling
environment that represents Bourdieu’s field. The notion of capital is represented by
26 Exploring Habits and Habitus of Biomedical Students with Modelling Tasks 301
the richness of the modelling environment, in particular the social structure embed-
ded in group work, the real-world contexts of tasks and potential modelling compe-
tencies that can be accumulated.
26.3 Empirical Design and Method
26.3.1 The Setting and Tasks
The participants (n = 51) were a cohort of first year students who enrolled in a
National Diploma in Biomedical Technology at the University of Johannesburg,
South Africa. Although students normally work by themselves during classroom
activities, ten groups were purposively structured for the sake of the modelling
tasks. Using scores from a previous semester test, each group included low, average
and high achievers. The social organisation was preserved throughout the first
semester of 2016 when this study took place. None of the participants had previous
experience of mathematical modelling.
Students completed four modelling tasks over one semester during tutorial peri-
ods, and each task lasted for 90 min. Tasks were inspired by newspaper articles that
related to real-world phenomena. Task 1 related to a community project wherein
children could exchange recyclables for “Moolas” (a made-up currency); in turn,
Moolas could be exchanged for “treasures”.
Moolas for Trash Task

A recycling project – aptly called “turning trash into treasure” – assists school-
aged children to provide for their basic needs by exchanging recyclable mate-
rials like glass and plastic for items like toiletries and stationery in a “swop
shop”. Use matrices to create a poster that can help children understand the
trading process. Think about the following: (1) Moolas are awarded for recy-
clables and can then be used as “buying power” to acquire “treasures”. (2)
Choose your own categories of recyclables (e.g., cans, newspapers, glass,
etc.), and choose categories of treasures (e.g., toiletries, clothes, toys, etc.).
You should use at most three categories of recyclables and treasures. (3) You
can use your own values/quantities to represent the trading power of Moolas.
These values must be realistic and reflect the real-world value of items. For
example, clothes are more expensive than toiletries. Use the following
strategy:
(a) Relate recyclables to Moolas.
(b) Relate Moolas to treasures.
(c) Find a relationship between recyclables and treasures.
(d) Check that your poster makes sense, it should be understandable to young
children!
302 H. Kotze
Three of the four tasks (Tasks 1, 2 and 4) involved a complete modelling cycle
(Blum and Leiß 2007) where data had to be collected from real-world narratives. In
Task 3 (Kotze 2018), a mathematical model was given, but students had to search for
relevant data using technology. All four tasks were delineated into subtasks to assess
modelling competencies as proposed by Kaiser and Brand (2015); these competen-
cies involved making sense of the narratives, setting realistic assumptions, mathe-
matising the real-world problem with symbol sense, working mathematically with
algebraic sense and interpreting solutions to give real-world meaning to the task.
The study adopts an exploratory sequential mixed methods design (Creswell 2014).
Qualitative data were collected with two open-ended questionnaires conducted after
Task 1 and Task 4, respectively. The aim of these questionnaires was to compare
modelling outlooks over time. In addition, quantitative data were also collected by
means of a semantic differential questionnaire which was completed after Task 4.
According to Heise (1970), semantic differentials are contrasting adjectives that
describe a concept on a bipolar scale. For example, students had to describe model-
ling tasks as “hurried” or “time-consuming”. Adjectives were selected to contrast
the modelling environment with a traditional learning environment. The semantic
differential questionnaire incorporated four sections that could potentially reveal
characteristics of the modelling tasks that underscore habits and habitus: (1) feel-
ings about group work, (2) the nature of the tasks, (3) the value of the tasks and (4)
the procedural demands. The questionnaire measured students’ impressions of the
modelling tasks on a five-point Likert scale from a negative outlook (score = 1) to a
positive outlook (score = 5). Contrasting pairs were randomly ordered from a nega-
tive to positive outlook or vice versa. For example, by contrasting “hurried” with
“time-consuming”, “hurried” would be regarded as a negative outlook (Likert
score = 1) in modelling terms and “time-consuming” as a positive outlook (Likert
score = 5). Where contrasting pairs were stated in the reverse direction on the ques-
tionnaire, responses were reverse coded. This way, all 31 questionnaire items were
rated on a low-to-high outlook.
The analysis of data was performed in three phases. Firstly, the open-ended ques-
tionnaire completed after Task 1 was analysed with a directed approach to content
analysis by using existing theoretical concepts as coding themes (Hsieh and Shannon
2005). When supported by exemplary evidence, findings from a directed content
analysis can either reinforce or contest a theory.
Secondly, the semantic differential questionnaire was analysed quantitatively
with principal component analysis (PCA) using IBM SPSS Statistics (version 24).
Three tests confirmed that data were suitable to perform PCA (Field 2009): First,
the correlation matrix was inspected for negative or small Pearson coefficients
(r < 0.3) which would indicate to weak correlations; these items were iteratively
removed. Second, the Kaiser-Meyer-Olkin value of sampling adequacy measured
0.591 and exceeded the acceptable lower bound of 0.5. Third, data passed Bartlett’s
Test of Sphericity with χ2 (171) = 443.539, p < 0.001. To control the process of
extracting the principal components, data were subjected to three criteria, namely,
the Kaiser criterion that specifies a minimum eigenvalue of one, a scree plot which
suggested a stable four-component structure and the cumulative variance explained
by the data (60.5%). When these three criteria were applied, four principal compo-
nents were extracted. In total, 19 of the semantic differential items loaded onto the
four extracted components which explained, respectively, 24.7%, 16.6%, 11.6%
and 7.6% of the total variance. The quality of PCA does not only depend on the
sample size. In fact, the effect of the sample size decreases when communalities are
high (average > 0.6) and when the number of variables (questionnaire items) are
five times the number of extracted components (MacCallum et al. 1999). Since the
average of communalities was 0.605 and the variable to component ratio was 4.75
to 1, the small sample size (n = 51) could be justified. The reliability of the four
principal components was calculated with Cronbach’s alpha and measured as 0.876,
0.671, 0.682 and 0.733, respectively. The PCA empirical design provided evidence
of construct validity of the semantic differential questionnaire (Field 2009). To
ensure content validity of the semantic differential questionnaire, modelling prin-
ciples and characteristics were considered in order to create items as bipolar pairs.
Thirdly, the open-ended questionnaire completed after Task 4 was analysed with
a directed approach to content analysis, using the four extracted principal compo-
nents deductively as themes (Hsieh and Shannon 2005). This analysis provided
insights into habits and habitus trajectories over the three-month period between
Task 1 and Task 4. The credibility of results in the study was enhanced by using
three questionnaires.
26.4 Findings
26.4.1 pen-Ended Questionnaire 1: Prospects of Field,

O
Capital and Habitus
New Learning Environment Task 1 presented students with the opportunity to

work together as a team for the first time in a mathematics classroom at university.
According to Dewey (1938), a new learning environment (or field) can act as a stim-
ulus for the modification of habits. Students could witness how their peers reasoned
and acted. Perhaps due to the complexity and the unusual demands of the modelling
task, one student was convinced that “the task made more sense working as a team”.
The new modelling environment was laden by social capital, a valuable resource
when faced by higher challenges. Alas, not all students were enthusiastic about the
first modelling experience: “I didn’t like the task and the group just confused me”.
Habits can only be modified when a “fixation” on new ways of doing is consolidated
with a “weakening” of old behavioural activities (Dewey 1938, pp. 31–32).
304 H. Kotze
Adaptive Initiatives In the novel modelling environment, students were challenged

to adjust their customary ways of thinking, doing and acting. Some students seemed
to be stimulated as different behaviours and patterns of thought had to be activated.
The task “inspired me to be inquisitive and expanded my horizons to become smart”.
Also, the task required students to operate “out of [their] comfort zone”, to argue
logically and to reason sensibly. Students had to adapt to the unfamiliarity of the
modelling task, signalling desires to be creative and productive. “We came [up] with
different approaches”, and we had to “combine ideas and produced something
strong”. Some students felt motivated to adapt their ways of doing; “[the task]
allows you to see things in a different way and do things in different ways”. An
innovative spirit was embraced by one student who felt that “[Task 1] made me open
up and made me really inquisitive about things I’ve never met”. However, percep-
tions were undoubtedly polarised as other students were reluctant to change: “I
would not want to do more of these tasks because we are used to the normal way of
doing things, and this method will take time to get used to”.
Capital Capital refers to the gains or competencies resulting from experiences and
learning opportunities. When the field is adjusted, students may recognise a short-
age/lack of desirable competencies which may even inspire an “avowal of a self-
seeking aspiration” towards such capital (Bourdieu 1990, p. 11). Accordingly, one
student sensed that in order to meet future workplace demands, “a biomedical tech-
nologist needs to analyse on a daily basis”. This awareness is a clear reflection on
the need for adaptation. Another student discovered that in Task 1, concepts could
be practically applied to a real-life situation: “Matrices are a good mathematical
tool since most of them are about arrangement of general things in life”. The atypi-
cal modelling demands also challenged conventional behaviours: “Such approach
should be done frequently in order to think open-minded”, a call for adjustment in
favour of more modelling capital. Such views differ from established classroom
practices that often require stereotyped approaches embodied in a definition-
examples-exercise routine (Resnick 1989). Scepticism was also evident: “I did not
enjoy [the task] at all, I’d prefer the traditional theoretical approach”. The state of
the field and its potential capital will impact on students’ habitus.
Habitus Students arrive at university with predisposed and ingrained school expe-
riences which impact on the way they think, reason and strategise. When students
behave in a certain way under certain circumstances, habits are born. In turn, past
schooling habits structure students’ habitus. The new modelling field demanded dif-
ferent strategies and approaches. Students’ “habitus, their socially constituted
nature, is immediately adjusted to the immanent demands” (Bourdieu 1990, p. 11).
Habitus also paves the way for future behaviour. The first modelling task was expe-
rienced by some students as more hands-on: “I now know better how to arrange data
available in daily life than [when doing] tasks given in textbooks”. The arrangement
and analysis of data are an expertise encultured not only in the biomedical industry
but also in ordinary daily life; as such, this aspiration reasserts habitus. The complex
modelling task was a benchmark for learning: “this approach may assist me in the
near future”. Task 1 presented students with the opportunity to self-create meaning
from the narrative in the newspaper article: “we were able to generate our own
matrices to make sense from the story”. This student seems to take responsibility for
mathematical meaning-making while reflecting a conscientious orientation to real
life, impacting on habitus. For some students, the competencies they gained at
school were virtuous: “[the task] was too difficult, I didn’t know what was going
on”, and “it would be impossible to apply matrices to other contexts”.
26.4.2 Semantic Differential Questionnaire
Table 26.1 shows the semantic differentials that loaded onto the four principal com-
ponents. Labels were created to describe the meaning of the underlying semantic
differentials associated to each component.
The first component related to students’ feelings about the task and was labelled
modelling affinity. The semantic differentials that loaded onto modelling affinity
supported students’ views on motivation, stimulation, enthusiasm, receptiveness to
other opinions/approaches and scientific curiosity. These characteristics afforded by
peer-interaction during the modelling tasks could also reveal opposing patterns of
long-standing mathematical outlooks and attitudes. The second component was
labelled sense-making and was related to the nature of the modelling tasks. Semantic
differentials that loaded onto sense-making described the usefulness, meaningful-
ness and purposefulness of modelling tasks which also elicited critical thinking.
Schoenfeld (1992, p. 18) argues that through meaningful types of classroom experi-
ences, students come to realise that “mathematics [is] an act of sense-making”.
When sense-making is stimulated by a certain experience-based context, students
Table 26.1 Principal component analysis of semantic differential questionnaire

Modelling affinity Sense-making Career aspiration Modification
Disappointed – Useful – useless Develop own solution – Follow routines –
stimulated target specific solution adjust routines
Receptive – Meaningful – not Superficial – realistic State answers –
unreceptive meaningful interpret answers
Productive – Purposeful – aimless Textbook like – unlike Understand
unproductive textbook facts – memorise
facts
Understanding – Stereotyped Theoretical – practical Tradition –
confused thinking – critical discovery
thinking
Motivated –
unmotivated
Enthusiastic –
unenthusiastic
Scientifically
curious – uninterested
306 H. Kotze
can develop a certain frame of mind that becomes part of their habitus (Bourdieu
1990). The third component was labelled career aspiration since it related to the
perceived future value of the modelling tasks that students could aspire to. The
seman tic differentials that loaded onto career aspiration referred to the culture of
developing justifiable solutions in a post-textbook-like climate where problems
have realistic contexts of practical value, thus simulating workplace demands. The
fourth component was labelled modification since it related to the procedural
demands of learning in a modelling environment. Semantic differentials that loaded
onto modification referred to the need to adjust routines, to interpret answers in real-
world terms, to understand concepts and the discovery of new ideas and processes.
Correlations between components were also investigated. To investigate possible
patterns, component scores were calculated by using the Anderson-Rubin procedure
in SPSS (Field 2009). A component score was generated for each student by com-
bining responses of all individual scores that loaded onto a particular component.
By default, component scores have mean 0 and standard deviation 1. To illustrate
correlations between components, scatterplots were generated. Figure 26.1 repre-
sents the overarching views of students after completing all four modelling tasks
and assists with the interpretation of typical student profiles: (a) depicts modelling
affinity versus sense-making and (b) contrasts modification versus career aspiration.
To help identify patterns, higher densities of scores were emphasised by superim-
posing ellipses onto the scatterplots. In Fig. 26.1a, the majority of scores are located
in the first quadrant which represents students with a positive attraction towards
modelling and sense-making. Figure 26.1b suggests a correlation between modifi-
cation and career aspiration that is more centralised. The elongated ellipse gives the
impression that the majority of students were positively inclined towards career
aspiration.
After completing Task 4, students’ responses to the second open-ended question-
naire were coded according to the four extracted components, namely, modelling
affinity, sense-making, career aspiration and modification.
Fig. 26.1 Scatterplots of (a) modelling affinity versus sense-making and (b) modification versus
career aspiration
26.4.3 Second Open-Ended Questionnaire
The modification of the field revealed an affinity towards modelling which mobil-
ised new habits. The four modelling tasks encouraged the exploration of new ideas
that had to be reconciled with known ideas and stirred a sense of scientific curiosity.
“I could sense the real-life connection with mathematics since the stories helped me
a lot to understand”, reflecting the view of a Fig. 26.1a quadrant one candidate. The
tasks imposed different habits on students, and in turn, they could question or recon-
sider their future outlooks, study goals and career aspirations: “I can use my knowl-
edge of the world out there to develop other maths skills, something I will need
later”, a Fig. 26.1b quadrant one response that signals a changing habitus. “I learned
to make better decisions” is testament of a student being receptive to developing
habits and competencies as afforded by the modelling tasks.
According to Schoenfeld (1992), classroom experiences impact largely on stu-
dents’ beliefs and sense-making of formal mathematics. Activities should emulate
mathematics as a sense-making activity. Indeed, one student discovered the purpose
and usefulness of learned mathematical concepts: “these tasks helped me make
sense of all my maths”. Students’ consistent pursuit to make sense of, and determi-
nation to assemble, ideas as a whole is a sign of a developing habitus. After com-
pleting four tasks, Fig. 26.1a suggests a slight positive inclination by students
towards sense-making due to their modelling experiences. This trend, albeit mar-
ginal, is evidence that a mathematical modelling approach can offer the due scope,
power and utility of mathematics in the classroom (Schoenfeld 1992). The model-
ling tasks alerted students to the rightful value of mathematics: “it is necessary to
relate math problems into real-life applications”, a perception not only value-laden
but also reflecting the desire to experience mathematics in action.
Almost half of the cohort were enthusiastic to embrace career aspirations as
operationalised in the modelling tasks. The concentration of scores in the first and
second quadrants of Fig. 26.1b reveals that more students were positively inclined
to modelling potentialities which advance students’ own solutions within realistic
contexts – unlike traditional textbook problems that often target particular solutions.
Different capital could be untapped from the realistic contexts of tasks: “[the tasks]
are sort of practical which helps us to apply mathematics and learn new things”. The
practical nature of the modelling tasks presented students with expectations and
learning goals that helped to ground learning. Undeniably, some students regarded
the modelling tasks with resistance: “I prefer routine problems, [we] won’t need this
[approach] in other subjects”, a Fig. 26.1b quadrant 3 profile. The modelling envi-
ronment necessitated a modification in established ways. Without appropriating
resources available from social capital and competencies gained from the modelling
field, students would be inclined to keep learning according to past perspectives,
thereby forfeiting prospects to an evolving habitus.
Deeply rooted actions, thought patterns and behaviours were challenged by the
modelling tasks that elicited modification of traditions to acquire “the habits and
dispositions of interpretation and sense-making” (Resnick 1989, p. 58). With
308 H. Kotze
dispositions still rooted in old-school habits, some students were perturbed with the
shift in task boundaries, namely, that modelling tasks do not necessarily have only
one correct answer. The scores of students who were resistant or hesitant to adapt
routines after completing four tasks can be observed in the second and third quad-
rants of Fig. 26.1b. The tendency to oppose change concurs with Dewey’s (1938)
belief that the formation of new habits is initially in a phase of inertia, as can be
expected when students are first exposed to modelling tasks. Indeed, as the new
learning demands actualised, limitations were acknowledged: “I cannot do this
[task] if I still struggle to understand the math concepts”. After all, there is an ines-
capable interplay between habits and habitus on the continuum from past to future.
26.5 Conclusion
Using four modelling tasks to explore habits and habitus was undeniably a limita-
tion of this study. There is scope for inquiries into a prolonged exposure to mathe-
matical modelling. However, while students are being short-changed at school
where the potentialities of mathematical modelling are overlooked, it was encourag-
ing to find evidence of evolving dispositions towards modelling that sparked the
importance of different habits and habitus.
References
Biccard, P. (2010). An investigation into the development of mathematical modelling competen-

cies of Grade 7 learners. Masters thesis. University of Stellenbosch, Cape Town, South Africa.
C. Haines, P. Galbraith, W. Blum, & S. Kahn (Eds.), Mathematical modelling: Education,
Bourdieu, P. (1990). In other words, essays towards a reflexive sociology. Stanford: Stanford
University Press.
Creswell, J. (2014). Research design – Qualitative, quantitative and mixed methods approaches
(4th ed.). Washington DC: Sage.
Dewey, J. (1938). Logic – the theory of inquiry. New York: Henry Holt.
Dewey, J. (1944). Experience and education. New York: Macmillan.
Dumais, S. A. (2002). Cultural capital, gender, and school success: The role of habitus. Sociology
of Education, 75(1), 44–68.
Field, A. (2009). Discovering statistics using SPSS (3rd ed.). London: Sage.
Heise, D. R. (1970). Semantic differential and attitude research. In G. F. Summers (Ed.), Attitude
measurement (pp. 235–253). Chicago: Rand McNally.
Hsieh, H. F., & Shannon, S. E. (2005). Three approaches to qualitative content analysis. Qualitative
Health Research, 15(9), 1277–1288.
Humphrey, J. D., Coté, G. L., Walton, J. R., Meininger, G. A., & Laine, G. A. (2005). A new para-
digm for graduate research and training in the biomedical sciences and engineering. Advances
in Physiology Education, 29(2), 98–102.
Kaiser, G., & Brand, S. (2015). Modelling competencies: Past development and future perspec-
tives. In G. A. Stillman, W. Blum, & M. S. Biembengut (Eds.), Mathematical modelling in edu-
cation research and practice: Cultural, social and cognitive influences (pp. 129–149). Cham:
Springer.
Kotze, H. (2018). Competencies in mathematical modelling tasks: An error analysis. Eurasia
Journal of Mathematics, Science and Technology Education, 14(8), em1567. https://doi.
org/10.29333/ejmste/91922.
MacCallum, R. C., Widaman, K. F., Zhang, S., & Hong, S. (1999). Sample size in factor analysis.
Psychological Methods, 4(1), 84–99.
Resnick, L. B. (1989). Treating mathematics as an ill-structured discipline. In R. Charles &
E. Silver (Eds.), The teaching and assessing of mathematical problem solving (pp. 32–60).
Hillsdale: Erlbaum.
Schoenfeld, A. (1992). Learning to think mathematically: Problem solving, metacognition and
sense-making in mathematics. In D. Grouws (Ed.), Handbook for research on mathematics
Chapter 27
Metacognitive Strategies in Group Work
in Mathematical Modelling Activities –
The Students’ Perspective
Alexandra Krüger, Katrin Vorhölter, and Gabriele Kaiser
Abstract Metacognitive competency can be described as a sub-competence of

mathematical modelling competence. Although there is consensus about its rele-
vance, until now research on metacognition in modelling processes is very limited.
In particular, students’ perspective on metacognitive competencies has not been
examined yet. The present study aims at investigating students’ perception of meta-
cognition in modelling processes. Fourteen groups of students were interviewed at
the beginning and at the end of a teaching unit. The results presented in the paper
focus on one group, who used extensively, at the end of the study, cognitive strate-
gies and skills showing a sense of direction in the formulation phase of the model-
ling process as an indicator of metacognitive strategies.
Keywords Cognitive strategies · Metacognition · Metacognitive strategies ·

Metacognitive knowledge · Metacognitive skills · Modelling problems ·
Monitoring · Planning · Regulating
27.1 Introduction
Based on the discussion of how to conceptualise modelling competencies (cf. Kaiser

and Brand 2015), one of the main research aims in mathematical modelling in
recent years was to investigate how to foster students’ modelling competencies (for
an overview, see Greefrath and Vorhölter (2016)). The overall modelling compe-
tence includes a metacognitive facet, which has been often neglected in past
A. Krüger (*) · K. Vorhölter · G. Kaiser

Faculty of Education, Universität Hamburg, Hamburg, Germany
e-mail: alexandra.krueger@uni-hamburg.de; Katrin.vorhoelter@uni-hamburg.de;
Gabriele.kaiser@uni-hamburg.de

312 A. Krüger et al.
research, although there is a strong consensus about its relevance (Blum 2015). For
example, students use metacognitive strategies in order to transit between stages in
the modelling process and to identify and overcome cognitive barriers (Stillman
2011). However, until now the promotion of students’ metacognitive modelling
competencies has not been investigated in depth. As the usage of metacognition
only takes place, if students are convinced of the benefits of it, it is necessary to
survey students’ perception of metacognition before fostering metacognition.
In this paper, we present first results of a research study, which examines the
students’ perspective on their metacognitive competencies within a cooperative
working environment during modelling activities. Due to space restrictions we
focus on one group of students and their metacognitive development.
27.2.1 The Concept of Metacognition
The concept of metacognition was already introduced by Flavell (1979) and Brown
(1978) in the 1970s (for details, see overview by Veenman (2011)). Despite this
early work, no consistent definition of metacognition in the current discussion can
be identified. However, the distinction between metacognitive knowledge (declara-
tive and conditional knowledge) and metacognitive skills (procedural metacogni-
tion, metacognitive strategies and skills) is widely accepted (Veenman 2011).
Besides, a third component is often included, which comprises metacognitive expe-
riences and an appropriate attitude (Efklides 2008). Therefore, in this study, the
following definition of metacognition is used, since this definition focuses on per-
formance rather than knowledge:
Metacognition is generally understood as knowledge, skills, and attitudes that are available,
necessary, or helpful in order to initiate, organise, and control (implicit as well as explicit)
processes of strategic decisions for learning or thinking. (Weinert 1994, p. 193, translated
by the authors)
We further assume in accordance with Efklides (2008) that parts of metacognition

not only take place on a conscious level, but using metacognitive strategies is some-
times an unconscious process. Thus, like Hartman (2001) in the following, we will
distinguish linguistically between metacognitive strategies, which need to be
applied consciously, and metacognitive skills, which often run automatically in the
background and therefore are not necessarily conscious. Correspondingly, we will
talk about metacognitive strategies, as the students themselves comment on the use
of this strategies, which means that they are aware of the strategies or have at least
become aware of them afterwards.
In addition, in modelling processes, not only metacognitive competencies of
single students but also social metacognition relating to the whole group has to be
27 Metacognitive Strategies in Group Work in Mathematical Modelling Activities … 313
considered (Goos and Galbraith 19961). Besides, there are general metacognitive
skills that can be transferred to different domains, and a personal repertoire of meta-
cognitive strategies can be distinguished (Veenman 2011). However, the benefit of
particular strategies may vary in different contexts. Furthermore, task-specific meta-
cognitive knowledge depends on the nature of the task. In the following, we will
elaborate on metacognition in modelling processes.
27.2.2 Metacognition in Modelling Processes
In the last decades, it became obvious that there are different perspectives on math-
ematical modelling that have an influence on the modelling problems chosen for
reaching the connected aims. For working on the modelling problems we use in our
studies (see, for example, Vorhölter (2019)), students have to develop their own
approach autonomously. The modelling problems are solvable with different math-
ematical procedures, which had not necessarily been content of the lessons imme-
diately before. The data required for solving the problem with the help of the
developed model had to be researched (e.g., on the internet), judged or calculated.
This description reveals that the handling of these tasks is demanding. Thus, meta-
cognitive strategies can be helpful.
Referring to the general metacognitive competence, metacognitive modelling
competencies can be divided into metacognitive knowledge and metacognitive
strategies. Since the focus of this paper is on metacognitive strategies, metacogni-
tive knowledge will not be considered in more detail in the following. Metacognitive
strategies are composed of the following:
• Strategies for planning the solution process considering the task that has to be
worked on, the group members involved and specific circumstances. Planning
strategies can be identified, when cognitive strategies are used purposefully, for
example, rereading the text, brainstorming, identifying missing information,
separating relevant from irrelevant information as well as agreeing on a common
understanding.
• Strategies for monitoring and, if necessary, regulating the working process,
which can be realised, for example, by using the modelling cycle as a tool.
Identifying cognitive barriers and seeking help by, for example, asking the
teacher and classmates or researching on the internet are indicators for monitor-
ing strategies and regulation strategies.
• Strategies for evaluating the work in order to improve the modelling process.
Indicators for using strategies to evaluate the work are questions and statements
about unsuccessful group work or personal involvement in the group work.
1
For a short overview about the usage of metacognitive strategies in groups while working on
modelling problems, see Vorhölter (2019).
Referring to Goos (1998), Stillman (2011) describes the usage of students’ meta-
cognitive strategies during modelling activities while experiencing difficulties. She
also identifies different types of productive metacognitive acts and non-productive
metacognitive acts during mathematical modelling. Therefore, the aim of the usage
of metacognition is gaining a sense of direction, especially in the model formulation
phase (Treilibs 1979) and not losing track of one’s own approach.
27.2.3 mpirical Research Results of Metacognition

E
in Modelling
Research on metacognition in modelling is at its beginning. However, there exists

comprehensive research on metacognition in problem-solving processes and in
mathematics education in general (for an overview, see Desoete and Veenman
(2006), Schneider and Artelt (2010)). In the following, we will limit the presenta-
tion of research results to the field of modelling.
In her study, Maaß (2006) analysed the metacognitive modelling knowledge of
students. She demonstrated that their knowledge of the modelling process as part of
the metacognitive knowledge increased in the course of the study. In addition, she
could reconstruct misconceptions about the students’ modelling processes and a
connection between metacognitive knowledge and modelling competence; that is,
due to missing or low meta-knowledge about the modelling process, problems could
arise while working on modelling problems.
In her studies, Stillman (2004, 2011) focused on emerging problems in the form
of cognitive blockages and how students can overcome these problems by using
metacognitive strategies. Thereby, they were able to progress and to transit between
different stages in the modelling process. She reconstructed different kinds of cog-
nitive and metacognitive strategies that can help to avoid difficulties or facilitate the
working process, for example, strategies for overcoming memory-related problems,
strategies for overcoming representational problems and strategies for benefiting
from facilitating perceptual conditions (Stillman 2004). The importance of a well-
developed repertoire of cognitive and metacognitive strategies, of an extensive
knowledge of mathematical concepts, facts and procedures as well as the relevance
of experiences in the use of it became apparent.
In the last years, various studies focusing on the usage of a modelling cycle as
metacognitive tool were carried out (Blum 2015; Maaß 2006; Schukajlow et al.
2015). Especially in the case of difficulties, the modelling cycle can serve as a meta-
cognitive aid as students can orient themselves on it (Blum 2015). Schukajlow and
Leiss (2011) investigated the self-reported use of cognitive and metacognitive strat-
egies during modelling activities. The results of this study show no significant cor-
relation between self-reported strategies (neither in general nor task-oriented) and
mathematical modelling competency (Schukajlow and Leiss 2011). This result can
be explained in two ways: On the one hand, it is possible that the use of a certain
strategy does not have a high influence on the working process; on the other hand,
the method for measuring those metacognitive strategies via self-reports may not be
valid or precise enough. However, “[t]here are many indications that metacognitive
activities are not only helpful but even necessary for the development of modelling
competency” (Blum 2011, p. 22).
27.2.4 Research Questions
As the application of metacognition is demanding, students should be introduced to

the use of it over a long period of time. Furthermore, fostering students’ metacogni-
tion will only be successful if students become aware of the benefits of using meta-
cognition. Therefore, students’ perception of the usage of metacognitive strategies
is important in order to develop appropriate teaching units and teacher interven-
tions. Therefore, the following research questions are examined in this paper:
1. Which metacognitive strategies do students express at the beginning and at the
end of an intervention study?
2. How do students’ metacognitive strategies develop from the beginning to the end
of the study?
3. How do the students evaluate this development with regard to group work and
the work on the modelling problem?
Thus, the aim of the study is to investigate students’ perspective on metacognitive
strategies in group work during mathematical modelling.
Fig. 27.1 Design of the study (TT teacher training)

27.3 Design and Methods of the Study
This study is embedded in the research project MeMo (Metacognition in Modelling)

at the University of Hamburg (Vorhölter et al. 2019). This research project aims at
evaluating a teaching unit fostering students’ usage of metacognitive strategies dur-
ing modelling processes.
The eighteen participating classes of grade nine and ten worked on six different
modelling problems for a 90-min lesson (see Figs. 27.1, 1–6). The teachers were
prepared for the study in three teacher trainings of three hours (see Fig. 27.1, TT1-3),
where theoretical aspects of mathematical modelling as well as useful teacher inter-
ventions within modelling activities were presented to them. The modelling prob-
lems they had to use in class were introduced to them, and possible cognitive
obstacles as well as suitable interventions were discussed. The modelling problems
were complex, reality-based tasks which can be solved in different ways (for an
example, see Vorhölter (2019)). All teachers were introduced to the modelling cycle
in the teacher training and were asked to refer to it when students asked for help as
well as during students’ presentation of possible solutions.
The presented study is a qualitatively oriented study focusing on students (for the
parallel study on teachers, see Wendt et al. this volume). For data collection, the
three-step design of Busse and Borromeo Ferri (2003) was used, which includes a
group observation with videotaping, a stimulated recall and a guided interview.
While working on the first and last modelling problem, 57 groups of students were
videotaped. The videos of 14 of these groups were analysed focusing on situations,
in which cognitive and metacognitive strategies as well as metacognitive knowledge
were demonstrated. On the following day, the selected situations formed the basis
for the stimulated recall. After having seen the scenes, students were asked to
describe the situations, their thoughts and their feelings during the situations. In
addition, a few focused questions adapted to the previous statements of the students
about the use of cognitive and metacognitive strategies were asked afterwards. The
interviews were carried out on an individual basis with all students of a group.
For data analysis, a qualitative content analysis (Kuckartz 2016) is used. The
codes used were developed inductively as well as deductively, based on literature
reviewed and the authors’ experiences with, and observations of, modelling pro-
cesses (Vorhölter et al. 2019). Students’ statements were coded as metacognitive
(and not only cognitive) if the broader context of the statements indicates that the
usage of the strategy was aiming at initiating, organising and controlling (Weinert
1994, p. 193) of the modelling process. The data has been coded several times with
the content-structuring analysis by Kuckartz (2016) and presented in case summa-
ries. Finally, the results were discussed and revised in expert groups.
27.4 First Results
In the following, we focus on one group of three boys from a higher track school in
Hamburg, who were taught by Ms. Schmidt, a young female teacher (for further
information about Ms. Schmidt, see Wendt et al. this volume). The group was cho-
sen because they changed their working behaviour from working alone to working
as a group during the study, which has positively influenced the use of metacogni-
tive strategies.
27.4.1 esults Concerning Metacognitive Strategies

R
of Planning
At the beginning of the study, the group used several cognitive strategies to under-
stand and simplify the real problem; for example, they distinguished relevant from
irrelevant information, and they identified missing and given information as well as
assumptions made. However, a metacognitive usage of these strategies could not be
reconstructed. In addition, the students mentioned at the beginning of the study
advantages of, and necessity for, a planning phase of the modelling process. After
working on the first modelling problem, they recognised the importance of cooper-
ating in the group as a requirement for working on the same approach, which is a
part of the planning strategies, and intended afterwards to use more planning activi-
ties during the modelling activities. During the next lessons, they developed their
own procedure for planning. Apart from the cognitive strategies they had already
used at the beginning, they did brainstorming and task sharing. Thereby, they
extended their use of cognitive strategies and used them on a metacognitive level.
Dustin: # Well … # And then ar- eh at first, we wanted to have a look and discuss the way
of solving the problem, because this has always helped us and every member of the group
can do something. And you can monitor mutually. […] It actually just keeps getting ahead.
All of the group members can participate, nobody has open questions and you can then, as
already said, also monitor mutually.
In this quotation from the end of the study, the student clarified that a planning
phase is important because in this way, all group members use the same approach
and can discuss their questions. Therefore, no open questions are left and the stu-
dents can monitor their processes mutually, which shows the connection between
the different procedural strategies of metacognition.

R
of Monitoring
At the beginning of the study, the three boys did not work together as a group: Most
of the time, they worked on their own or split up into two subgroups and worked on
different approaches. However, they were not satisfied with their group work and
mentioned that working together cooperatively and monitoring mutually would
have improved their solution.
Dustin: If we had exchanged our ideas, we would have realised that Furkan’s approach was
the right, ugh, better one. Then, we could have worked on Furkan’s approach together and
perhaps, we had found a solution faster.
At the end of the study, the boys monitored, according to their own description, their
processes in the following phases of the modelling process:
• The process of making assumptions
• The selection of the mathematical model
• Precision of the mathematical work
• Interpretation of the solution
• Validation of the solution
• Usage of the modelling cycle and phases of the modelling cycle
Overall, at the end of the study, the boys monitored not only their own behaviour but
also the behaviour of their group mates. They experienced the benefits of working
together in a group and monitored their processes mutually. Thus, the importance of
the group process for the development of social metacognition becomes apparent
(see Sect. 27.2.1).

R
of Regulation
Most strategies for regulating the modelling process are based on monitoring strate-
gies. Thus, it is not astonishing that the boys did not mention many strategies for
regulating in their first interviews. However, at the end of the study, they mentioned
more strategies for regulating their working process. The strategies named can be
divided into those used within the group and those aiming at getting external help
(see Table 27.1). At first glance, the strategies presented seem to be of a cognitive
nature. However, the broader context of the statements made clear that they were
used purposefully for improving the modelling process. Thus, these remarks were
coded as metacognitive.
At the beginning of the study, the only group-related strategy for regulating that
could be reconstructed was choosing a different model when recognising that theirs
Table 27.1 Strategies mentioned for regulating the modelling process

Focus Timing
Beginning of the study End of the study
Group Choosing a different Rereading the task for developing a real model
related mathematical model Choosing a different mathematical model
External Doing research on the internet Comparing with notes about the other
help modelling problems
Using a formulae book
Doing research on the internet
Asking the teacher
Asking classmates
is not useful. At the end of the study, they also reread the task when developing the
real model.
In addition, they made use of external help. At the beginning, they used informa-
tion from the internet to solve the problem. Besides, at the end of the study, they
looked at their notes from the other modelling problems to find a hint on how to
proceed and used a formulae book to find a formula. Furthermore, they asked the
teacher and classmates, when they needed help. Overall, they were able to extend
their usage of regulation strategies. As these strategies are more basic and self-
evident, it can be assumed that the boys did not learn these strategies explicitly but
became aware of their value for working on a modelling problem.
27.5 Conclusions
The group presented has extended their usage of metacognitive strategies. At the
beginning of the study, cognitive strategies were used more seldom and not always
aiming at improving the modelling process in contrast to an extensive and purpose-
ful usage at the end of the study. Furthermore, the students realised the necessity of
using metacognitive strategies.
They mentioned that they had learnt to transfer a real problem into a mathemati-
cal model, that is to say, they have achieved metacognitive knowledge about task
requirements and strategies. For example, they had the experience that it is neces-
sary to simplify the real problem by making adequate assumptions. Additionally,
they developed a critical attitude regarding their modelling process. This became
clear by monitoring their modelling process, where a sense of direction developed.
Furthermore, they improved the communication within the group by explaining all
steps of their working process to the other group members, that is to say that they
shared their metacognitive competencies.
To summarise, these first results show that the promotion of metacognitive strate-
gies or overall metacognitive competencies is possible, if it is explicitly promoted
by the teachers and the learning environment. The first impression from the analysis
of further groups suggests that the usage and the development of the usage of meta-
cognitive strategies are strongly dependent as well on the individual as on the group
they were in. However, the results of this study depend on the teachers, the school
context, the selected modelling problems and the selected group. Thus, it will be
necessary to extend these results to other groups of this study in order to create a
typology of different types of students’ perspectives on metacognitive strategies
during modelling processes.
References
G. Kaiser, W. Blum, R. Borromeo Ferri, & G. A. Stillman (Eds.), Trends in teaching and learn-
ing of mathematical modelling (pp. 15–30). Dordrecht: Springer.
education (pp. 73–96). Cham: Springer.
Brown, A. L. (1978). Knowing when, where, and how to remember: A problem of metacognition.
In R. Glaser (Ed.), Advances in instructional psychology (pp. 77–165). Hillsdale: Erlbaum.
Busse, A., & Borromeo Ferri, R. (2003). Methodological reflections on a three-step-design com-
bining observation, stimulated recall and interview. ZDM Mathematics Education, 35(6),
257–264.
Desoete, A., & Veenman, M. V. J. (2006). Metacognition in mathematics: Critical issues on nature,
theory, assessment and treatment. In A. Desoete & M. Veenman (Eds.), Metacognition in math-
ematics education (pp. 1–10). New York: Nova Science.
Efklides, A. (2008). Metacognition. Defining its facets and levels of functioning in relation to self-
regulation and co-regulation. European Psychologist, 13(4), 277–287.
Flavell, J. H. (1979). Metacognition and cognitive monitoring: A new area of cognitive-
developmental inquiry. American Psychologist, 34(10), 906–911.
Goos, M. (1998). I don’t know if I’m doing it right or I’m doing it wrong! Unresolved uncer-
tainty in the collaborative learning of mathematics. In C. Kanes, M. Goos, & E. Warren (Eds.),
Teaching mathematics in new times (pp. 225–232). Gold Coast: Mathematics Education
Research Group of Australasia.
Goos, M., & Galbraith, P. (1996). Do it this way! Metacognitive strategies in collaborative math-
ematical problem solving. Educational Studies in Mathematics, 30(3), 229–260.
Greefrath, G., & Vorhölter, K. (2016). Teaching and learning mathematical modelling. Approaches
and developments from German speaking countries. Cham: Springer.
Hartman, H. J. (2001). Developing students’ metacognitive knowledge and skills. In H. J. Hartman
(Ed.), Metacognition in learning and instruction. Dordrecht: Springer Neuropsychology and
Cognition.
tives. In G. A. Stillman, W. Blum, & M. S. Biembengut (Eds.), Mathematical modelling in
education research and practice (pp. 129–149). Cham: Springer.
Kuckartz, U. (2016). Qualitative Inhaltsanalyse, Methoden, praxis, Computerunterstützung.
Weinheim: Beltz Juventa.
113–142.
Schneider, W., & Artelt, C. (2010). Metacognition and mathematics education. ZDM Mathematics
Education, 42(2), 149–161.
Schukajlow, S., & Leiss, D. (2011). Selbstberichtete Strategienutzung und mathematische
Modellierungskompetenz. Journal für Mathematikdidaktik, 32, 53–77.
Stillman, G. (2004). Strategies employed by upper secondary students for overcoming or exploit-
ing conditions affecting accessibility of applications tasks. Mathematics Education Research
Journal, 16(1), 41–76.
Stillman, G. (2011). Applying metacognitive knowledge and strategies in applications and model-
ling problems at secondary school. In G. Kaiser, W. Blum, R. Borromeo Ferri, & G. Stillman
Springer.
Treilibs, V. (1979). Formulation processes in mathematical modelling. Thesis submitted to the
University of Nottingham for the degree of Master of Philosophy.
Veenman, M. V. J. (2011). Learning to self-monitor and self-regulate. In R. Mayer & P. Alexander
(Eds.), Handbook of research on learning and instruction (pp. 197–218). New York: Routledge.
Vorhölter, K. (2019). Enhancing metacognitive group strategies for modelling problems. ZDM
Mathematics Education, 51(4), 703–716.
Vorhölter, K., Krüger, A., & Wendt, L. (2019). Metacognition in mathematical modeling – an over-
view. In S. Chamberlain & B. Sriraman (Eds.), Affect and mathematical modeling (pp. 29-51).
Cham: Springer.
Weinert, F. E. (1994). Lernen lernen und das eigene Lernen verstehen. In K. Reusser & M. Reusser-
Weyeneth (Eds.), Verstehen. Psychologischer Prozess und didaktische Aufgabe (pp. 183–205).
Bern: Huber.
Wendt, L., Vorhölter, K., & Kaiser, G. (this volume). Teachers’ perspectives on students’ meta-
cognitive strategies during mathematical modelling processes - a case study. In G. A. Stillman,
G. Kaiser, & E. Lampen (Eds.), Mathematical Modelling education and sense making. Cham:
Springer.
Chapter 28
Fostering Students’ Construction
of Meaningfulness of Mathematics
with Mathematical Modelling Problems
Katrin Vorhölter and Björn Schwarz
Abstract Working on modelling problems in school can show students how fre-
quently mathematics is used in society and in everyday life. As a result, their moti-
vation to learn mathematics can increase. However, there has not been many
systematic investigations into whether, and how, these aims are achieved, nor is it
known which aspects of modelling problems are meaningful to the students. In this
chapter, we present a theoretical framework for analysing students’ personal mean-
ingfulness of mathematics with regard to influencing factors and consequences.
Furthermore, we present the results of a qualitative study with 15 students of grade
10 that show which characteristics of modelling problems and of working with
modelling problems in class have a direct or an indirect influence on the students’
construction of personal meaningfulness.
Keywords Basic needs · Beliefs · Constructivism · Mathematics classroom ·

Meaningfulness · Perspectives of modelling · Modelling problems
28.1 Introduction
A central aim related to the teaching and learning of modelling in mathematics

classrooms is to enhance students’ perception of mathematics itself as well as mak-
ing learning mathematics meaningful (Kaiser 2017). However, this aim has hardly
been verified, although the importance of meaningfulness for learning is well known
(Mitchell 1993). Important questions like what meaning students see in learning
K. Vorhölter (*)
e-mail: Katrin.vorhoelter@uni-hamburg.de
B. Schwarz
Faculty II, University of Vechta, Vechta, Germany
e-mail: bjoern.schwarz@uni-vechta.de

324 K. Vorhölter and B. Schwarz
mathematics or which aspects influence their construction of meaningfulness con-

cerning a special mathematical topic have not been tackled until now. Research in
this area often lacks clarification of the different denotations of meaningfulness as
well as the assumptions concerning the way in which the students ascribe this mean-
ingfulness. It is also relevant to look at the aspects influencing the process of ascrib-
ing and also at the consequences.
28.2.1 onstruction of Personal Meaningfulness

C
of Mathematics
A common observation in mathematics teaching is that students are not necessarily

convinced that the content they are taught in mathematics classrooms is of great
importance to them. In consequence, they regularly ask why they have to learn
mathematics, how learning mathematics might support their lives, or what the value
of engaging in mathematics is. In the terms of the study presented here, this implies
that the personal meaningfulness of mathematics is often not obvious to students.
We therewith closely follow Mitchell’s use of the term: “Meaningfulness refers to
students’ perception of the topics under study in their mathematics class as mean-
ingful to them in their present lives” (1993, p. 427). More precisely, in the follow-
ing, we understand “personal meaningfulness of mathematics” as the relevance and
significance an individual ascribes to mathematics. The term therefore covers the
relevance students see in learning and doing mathematics for themselves and for
their lives, and it is always related to an individuals’ perspective in the sense of
personal relevance. For a detailed description of this term and related terms see
Vorhölter (2009).
In the following, it is assumed that each learner has to both actively and individu-
ally construct his or her personal meaningfulness of mathematics. We describe this
process with the term “construction of personal meaningfulness”. Even though this
act of construction is a personal one within the respective individual student, social
transmissions—that means external influences—can possibly affect the construc-
tion of personal meaningfulness. With regard to school mathematics, these social
transmissions are especially, but not exclusively, represented by the learners’ class-
room experiences with mathematics (Schoenfeld 1992). This in particular covers
adequate learning environments or more general teachers’ activities which influ-
ence the construction of personal meaningfulness.
As every student has to construct his or her own personal meaningfulness, it is
possible that students construct different kinds of personal meaningfulness concern-
ing the same mathematical task or problem. In turn, although everyone has to con-
struct his or her own personal meaningfulness, the kind of meaning students
28 Fostering Students’ Construction of Meaningfulness of Mathematics… 325
Fig. 28.1 Framework of the process of construction of personal meaningfulness (Vorhölter 2009,
translated by authors)
construct is not arbitrary. It depends on the one hand on what the teachers, the par-
ents and also the society offer in terms of meaning. On the other hand, the students’
personal experiences, abilities, dispositions, wishes and intentions play an impor-
tant role. Combining these assumptions leads to the following model of construc-
tion of personal meaningfulness (Vorhölter 2009), which forms the basis of the
work presented here. Figure 28.1 depicts the central aspects of the model. According
to the model, construction of personal meaningfulness takes place within an indi-
vidual, which is influenced by the situation he or she is in. The process is affected
by the individual’s prerequisites, which are distinguished into background charac-
teristics and personal characteristics. Here, background characteristics cover invari-
able characteristics such as gender and age, while personal characteristics cover
those characteristics of a person, which can be influenced and changed, such as the
learner’s ambitions, his or her beliefs and the individual’s perception of a situation.
The individual’s construction of personal meaningfulness leads to consequences,
which cover both the individual’s judgements and activities.1
Following this model, the process of construction of personal meaningfulness is
influenced by the situation in which it takes place. In order to describe this influence
of the surrounding situation more precisely, we distinguish between direct and indi-
rect influences of the situation. Direct influence covers all aspects of the situation,
which are completely, or in part, designed to foster students’ construction of
personal meaningfulness. This especially covers all respective teaching activities,
1
A different kind of model for the process of constructing personal meaningfulness based on
Cultural-Historical Activity Theory is given in Hernandez-Martinez and Vos (2018).
for example, the choice of appropriate topics. These teaching activities of the teacher
can be seen as sense offered to the learners, which the learners can accept, change
or reject. Indirect influence, in contrast, means an influence of the situation, for
example, given by teaching activities, on personal characteristics. As these personal
characteristics also influence the process of construction of personal meaningful-
ness, this influence by the situation on the personal characteristics is called indirect.
Direct as well as indirect influences provided by modelling problems are described
in the next paragraph in more detail.
28.2.2 Modelling as an Offer for the Construction of Meaning
The idea that teachers provide offers of meaning of mathematics to learners through
their teaching activities leads to the question which learning environments in math-
ematics teaching are especially helpful to facilitate the learners’ construction of
personal meaningfulness. One prominent answer to this question is to strongly
incorporate references to reality or modelling into mathematics teaching. For exam-
ple, Kaiser (2017) states that one aim pursued by integrating modelling problems
into the mathematics classroom is to promote the learners’ insight into the meaning-
fulness of mathematical content. According to Blum (2015), modelling problems as
part of mathematics teaching can give more meaningfulness to lessons. A central
reason for implementing modelling into mathematics teaching is to show learners
the relevance of mathematics and thereby its value. For example, Lingefjärd (2006,
p. 111) indicates that “for most people, the value of mathematics most likely lies in
applications, and modeling is one of the main and most useful applications of math-
ematics”. In the same context, Burkhardt (2006, p. 191) states that “most students
find modelling courses more relevant to their lives than the mathematics they are
used to”. Thus, the use of authentic problems in mathematics teaching has a direct
influence on the construction of personal meaningfulness.
In contrast to this direct influence, modelling problems can also have an indirect
influence on the construction of personal meaningfulness. For the work presented
here, it is expected that especially the following three groups of personal character-
istics can be influenced by modelling problems:
–– Interest: It can be expected that teaching and learning mathematics with model-
ling tasks enhances the learners’ interest in mathematics. For example, the
authenticity of modelling problems fosters the learners’ interest in mathematics
(Kaiser et al. 2011).
–– Beliefs: It can be assumed that the characteristics of modelling tasks as well as
the way the students work with these tasks in class can have an influence on the
learners’ beliefs towards mathematics. For instance, modelling problems can
help to change learners’ beliefs towards a perspective emphasising the use of
mathematics (Maaß 2007).
–– Basic needs: The third group is formed by the basic needs as stated by Ryan and
Deci (2000). Ryan and Deci name autonomy, competence and relatedness in this
context, and it can be expected that modelling problems also influence these
fundamental psychological personal characteristics.
Based on the theoretical considerations of the preceding paragraphs, the work pre-
sented here focuses on two different research questions. In the following, both ques-
tions are described separately. The first question is:
(1) Which constructions of personal meaningfulness do students develop when
working with modelling problems in mathematics teaching?
The first question thus focuses on situations, in which students ascribe personal
meaningfulness to mathematics when working with modelling problems, and the
aim of the research question is to describe more precisely the kind of personal
meaningfulness the students construct. Against the background of the theoretical
model of construction of personal meaningfulness, this question thereby focuses on
the consequences of construction of personal meaningfulness, and the aim is to
describe the individual’s attitude towards the meaningfulness of mathematics.
The second question is:
(2) Which characteristics of modelling problems and which characteristics of
working with modelling problems in class have a direct or an indirect influence
on the construction of personal meaningfulness?
Thus, the second question again focuses on situations, in which students ascribe
personal meaningfulness to mathematics when working with modelling problems.
In this case, modelling problems obviously successfully offer personal meaningful-
ness, and the second research question aims at a more detailed description of those
characteristics of modelling problems, which make them successful offers of per-
sonal meaningfulness. Taking into consideration the distinction between factors
influencing the process of ascribing personal meaningfulness directly as well as
indirectly, it is also an aim of the research question to reconstruct those influencing
factors.
28.4 Method
For data collection, five classes of grade 10 (age 16–17) of two different German
higher-achieving schools (so-called Gymnasiums) were chosen. One reason for
choosing students of this age is that these students are capable of reflecting on their
experience during mathematics lessons. Further reasons, not less important, are that
students at this age often cannot apply the mathematical methods they have learnt in
their mathematics lessons in everyday life, and that students of this age often think
that there are much more important things than school.
For analysing the direct and indirect influence of modelling problems on stu-
dents’ construction of personal meaningfulness according to the research question,
the students of each class were videotaped twice: First, in each class, an ordi-
nary mathematics lesson was videotaped with two cameras. The only advice given
to the teachers for this lesson was that the lesson should be as usual for the students.
In the afternoon (or in the afternoon of the next day), a stimulated recall (see Gass
and Mackey (2000)) and an interview were done with all students of these classes
volunteering to be videotaped and interviewed. As the interviews were done in the
students’ free time, the number of students varies from two to six students per class.
However, students differed in mathematics performance, interest and gender, but
were similar in their willingness to provide information. During the stimulated
recall, they were shown the whole lesson and asked to comment on it. They were
mainly asked to say what they had been thinking during the lesson. Following this,
they were asked further questions about what they had said during the stimulated
recall. They were also interviewed about their attitude towards mathematics and
mathematics lessons, their feelings during mathematics lessons, and whether they
thought they needed or would need mathematics outside of school.
After three months, the teachers of the participating classes were given four dif-
ferent modelling problems and requested to act in the following way: At first every
student should read all four tasks and decide for herself or himself, which one she
or he would like to work on. Then the students, who had chosen the same task, were
asked to work on this task in a group. The teachers were asked to only assist if the
students asked for help during group work. At the end of the lesson, students got the
chance to present and discuss their results in class. The group work of the students
interviewed the previous time was videotaped, and another stimulated recall and
interview was done with them. In total, there are stimulated recalls and interviews
of 17 students relating to two different mathematics lessons.
In order to find out different types of personal meaningfulness and to analyse the
influence of modelling problems from the data, these were analysed by coding
according to the rules of Grounded Theory (see Strauss and Corbin (2003)). Firstly,
the interviews were analysed using open coding for conceptualisation on the first
level. By doing so, all codes were subordinated into nine categories (comprehen-
sion, learning process, mathematics, mathematics lessons, modelling, students,
task, teacher, type of work). Secondly, axial coding was done by relating the codes
of the categories mentioned above to each other, taking into account the framework
of personal meaningfulness outlined above. In a third step, the outcome of the sec-
ond step was used to compare and contrast different types of personal meaningful-
ness and the factors influencing them.
28.5 Results
28.5.1 ypes of Personal Meaningfulness Constructed

T
by the Students
Nine different types of personal meaningfulness with regard to modelling lessons

were reconstructed. They can be differentiated into four areas (see Fig. 28.2): The
first area is the construction of personal meaningfulness in mathematics as an aid
for life. The learners would like to use mathematics in their current or later life, in
their future occupation, and they want to be prepared for this at school. Some stu-
dents, though comparatively few, want to achieve self-fulfilment through mathemat-
ics: they want to experience joy and feel challenged. These constructions of personal
meaningfulness do not refer to mathematics itself but rather to doing mathematics,
which is of great importance to these students. The third category is evidence of the
students’ engagement with mathematics lessons. It becomes clear that these con-
structions of personal meaningfulness are purely related to mathematical lessons,
and the content of the lessons is indifferent to them. Queries of the last category,
mathematical competencies, lie transversely to these categories. The reason for this
is that students often consider this construction of personal meaningfulness as a
prerequisite for the attainment of further constructions. Examples for types of per-
sonal meaningfulness of two students are the following, which were assigned to the
categories of aid for life, whereby Kim was not able to realise her personal mean-
ingfulness, but Robin was:
Kim: Well, I don’t need this myself, [….]. So as a normal person you don’t actu-
ally need so much.
Robin: And yes, there you can see that mathematics is not really so much about
the math you learn in school, but that it is actually everywhere. And that it would
aid for life self-fulfillment engagement with

mathematics
lessons
preparation for life desire for challenge solving of tasks
tool fun marks
mathematical competencies
cognitive skills mathematical knowledge ability to reason
Fig. 28.2 Reconstructed types of personal meaningfulness of the modelling lessons

hardly work without mathematics, and. So, uh, in the real sense and yes, that some-
times you don’t even notice where mathematics is everywhere.
In addition to the types of personal meaningfulness mentioned above, with regard
to regular lessons without reality-based tasks, three further types could be recon-
structed: As a third type of the area engagement with mathematical lessons, fulfil
external expectations was reconstructed. In addition, a fifth category, social recog-
nition containing the two types general education and access requirements, was
found. Students wanted to gain social recognition through mathematics, for exam-
ple, by having a good general education or by obtaining a training place in a presti-
gious profession by means of good mathematics knowledge (Vorhölter 2009).
By contrasting the types of personal meaningfulness, the same students con-
structed with regard to the different lessons, it was analysed that the constructions
of the category aids for life as well as self-fulfilment furthermore have a special
position: Only if learners were able to make sustainable constructions in these areas,
they were not looking for further types of personal meaningfulness. In addition,
both areas reconstructed as most important for students, aid for life and self-
fulfilment, could be realised by the students more often in the modelling lesson.
Working on the modelling problem led to a modification of one student’s construc-
tion of personal meaningfulness; two others concretised their constructions of per-
sonal meaningfulness. The areas of personal meaningfulness as described above can
be assigned to the aims and perspectives of modelling (e.g., Blum 2015; Kaiser
2017) with the exception of the social-critical perspective. One can assume that the
latter could also be found in general but could not be reconstructed within the sam-
ple of the qualitative study presented here.
28.5.2 I nfluence of Modelling and Modelling Problems

on the Construction of Personal Meaningfulness
Getting back to the distinction between direct and indirect influences on the con-
struction of personal meaningfulness, in the following, the second research question
is to be answered. A central result is that the reality-based context of a modelling
problems has a direct influence on the students’ constructions of personal meaning-
fulness. In other words, the respective context is an offer of meaning in itself, which
is picked up by the students within their individual process of construction of per-
sonal meaningfulness. Moreover, it interestingly turned out that the context was the
only direct influence, which could be observed. In turn, this means that all other
characteristics of modelling in school and of modelling problems only served as
indirect influences on the construction of personal meaningfulness. Alongside, it
could be observed that the reality-based context also served as an indirect influenc-
ing factor, that is, that the context has both direct and indirect influence. Other
examples of characteristics having an indirect influence on the construction of per-
sonal meaningfulness are the lack of concrete operating instructions and the
reality-related context construction of meaning
result interest
group work
basic need: relatedness
teacher’s acting
self-activeness
basic need: autonomy
possibility to choose
basic need: competence
own solution
project-oriented work
beliefs
lack of information
missing instruction
Fig. 28.3 Indirect and direct influencing factors of working with modelling problems on the stu-
dents’ personal characteristics
incomplete information on the problems as well as the group work and the possibil-
ity to develop an individual solution. For example, the following answers were
given to the question, what the students thought to be different in the modelling
lesson compared to the lessons they are used to:
Thomas: Well, I actually thought it was pretty good. That you first have to think
about what information is missing, that is given more than you need, that
you have to think a little bit harder than if all is given.
Katja: […] that you can get out a task without knowing which task type you
have to use and which arithmetic operations and so on. I enjoyed that.
A complete list of all the observed indirect influencing factors can be found in the
left column of Fig. 28.3. It is important to emphasise that the order of the factors
does not imply any hierarchy. In contrary, the priority of the factors varies from
person to person and therefore can only be ranked on an individual level but not on
a summarised level.
As expected it could indeed be observed that the indirect influencing factors have
an effect on the interest, the beliefs and the basic needs (that is autonomy, compe-
tence and relatedness) of the students. The full result can also be found in Fig. 28.3.
It is to be especially emphasised that some of the indirect influencing factors, such
as group work, only have an influence on one area of the personal characteristics,
while other factors, such as own solution, influence several areas of personal
characteristics.
28.6 Conclusions
Acting on the assumption that construction of personal meaningfulness is an impor-

tant aim of teaching in general and mathematics teaching in particular, the study
confirms that the useage of modelling problems in mathematics classrooms can
foster these processes of students’ construction of personal meaningfulness. More
precisely, several ways of construction of personal meaningfulness could be
observed which were in this case initiated by the work of students with modelling
problems. This result therefore strengthens the plea for a stronger consideration of
modelling problems in mathematics teaching.
On the other hand, the results do not imply that there is any exclusive contribu-
tion to processes of construction of personal meaningfulness which only modelling
problems can offer. Indeed, the study identified several characteristics of modelling
problems and of working with them in the mathematics classroom which directly or
indirectly influence construction of personal meaningfulness. But none of these fac-
tors is solitarily bound to modelling. Instead, most of the factors, such as group
work, individual solutions and students’ self-activeness, can be regarded as consti-
tutional factors of constructivism-orientated teaching in general. In turn, this means
that the results presented from a more general perspective support the demand of
orienting mathematics teaching towards the principles of constructivism. Doing so
obviously helps students to individually construct personal meaningfulness of
mathematics. On a more detailed level, the results then offer empirical confirmation
for the assumption that one way of fostering these processes of students’ construc-
tion of personal meaningfulness is the use of modelling problems.
References
education (pp. 73–96). Cham: Springer International.
Burkhardt, H. (2006). Modelling in mathematics classrooms: Reflections on past development and
the future. ZDM Mathematics Education, 38(2), 178–195.
Gass, S. M., & Mackey, A. (2000). Stimulated recall methodology in second language research.
Mahwah: Erlbaum.
Hernandez-Martinez, P., & Vos, P. (2018). “Why do I have to learn this?” a case study on stu-
dents’ experiences of the relevance of mathematical modelling activities. ZDM Mathematics
Education, 50(1–2), 245–257.
Kaiser, G. (2017). The teaching and learning of mathematical modelling. In J. Cai (Ed.),
Compendium for research in mathematics education (pp. 267–291). Reston: NCTM.
Kaiser, G., Schwarz, B., & Buchholtz, N. (2011). Authentic modelling problems in mathematics
education. In G. Kaiser, W. Blum, R. Borromeo Ferri, & G. Stillman (Eds.), Trends in teach-
ing and learning of mathematical modelling (pp. 591–601). Dordrecht: Springer Netherlands.
Lingefjärd, T. (2006). Faces of mathematical modeling. ZDM Mathematics Education,
38(2), 96–112.
Maaß, K. (2007). Modelling in class: What do we students want to learn? In C. Haines,

P. L. Galbraith, W. Blum, & S. Khan (Eds.), Mathematical modelling (ICTMA 12): Education,
Mitchell, M. (1993). Situational interest: Its multifaceted structure in the secondary school math-
ematics classroom. Journal of Educational Psychology, 85(3), 424–436.
Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic moti-
vation, social development, and Well-being. American Psychologist, 55(1), 68–78.
Schoenfeld, A. H. (1992). Learning to think mathematically: Problem solving, metacognition,
and sense-making in mathematics. In D. A. Grouws (Ed.), Handbook of research on math-
ematics teaching and learning: A project of the National Council of Teachers of Mathematics
(pp. 334–370). New York: Macmillan.
Strauss, A. L., & Corbin, J. M. (2003). Basics of qualitative research: Techniques and procedures
for developing grounded theory. Thousand Oaks: Sage.
Vorhölter, K. (2009). Sinn im Mathematikunterricht: Zur Rolle von Modellierungsaufgaben bei
der Sinnkonstruktion von Schülerinnen und Schülern. Opladen: Budrich.
Chapter 29
Teachers’ Perspectives on Students’
Metacognitive Strategies during
Mathematical Modelling Processes –
A Case Study
Lisa Wendt, Katrin Vorhölter, and Gabriele Kaiser
Abstract The usage of metacognitive strategies can help students to avoid or to

overcome cognitive barriers during modelling processes. For imparting relevant
metacognitive strategies and supporting students during the modelling process, it is
necessary that teachers act on a meta-meta level and monitor their students’ usage
of metacognitive strategies. Therefore, the teachers’ perception of students’ meta-
cognitive activities during modelling processes is of great importance. However,
until now, there does not exist such a study. The study described in this chapter
examines this topic in a qualitatively oriented intervention study. The results focus
on one teacher, who showed a broadening of her perception of the students’ usage
of metacognition in mathematical modelling processes during the intervention.
Keywords Conditional knowledge · Evaluating · Metacognition · Metacognitive

knowledge · Metacognitive strategies · Modelling problems · Monitoring ·
Planning · Regulating · Teachers’ perception
L. Wendt (*) · K. Vorhölter

e-mail: Lisa.wendt@uni-hamburg.de; Katrin.vorhoelter@uni-hamburg.de
G. Kaiser
Australian Catholic University, Brisbane, QLD, Australia
e-mail: Gabriele.kaiser@uni-hamburg.de

336 L. Wendt et al.
29.1 Introduction
Mathematical modelling processes are cognitively demanding and complex for stu-
dents as every step of the modelling cycle can contain cognitive barriers. Modelling
activities are also challenging for teachers because they have to act on a meta-meta
level in order to support their students appropriately (Stillman 2011). When acting
on a meta-meta level, teachers’ interventions are predicated on their reflecting on
and taking into account the metacognitive activity of their students. Thus, teachers’
perceptions of those situations are crucial.
In the following, the reconstructed perception of one teacher during an interven-
tion study lasting eight months is presented with a special focus on the change of her
perception of the students’ usage of metacognition and necessary or implemented
interventions by her.
As there are different perspectives on mathematical modelling pursuing different

aims (see Kaiser and Sriraman [2006]), we would like to stress those points which
are important for our understanding of modelling, because they had an influence on
the modelling problems chosen for the study as well as the way of implementing
them in classrooms. The modelling problems used in the study were authentic tasks,
that is, tasks which were relevant at least for a special group of people, at best for
the students themselves. While working on the modelling problems, students had to
develop their own approach autonomously. The problems were solvable with differ-
ent mathematical means, which had not necessarily been content of the lessons
immediately before implementation. The data required for solving the problem with
the help of the developed model had to be researched (e.g., on the internet), judged
or calculated. Some of the modelling problems used are presented in Vorhölter and
Kaiser (2016) and Vorhölter (2018).
Working on such complex modelling problems is quite challenging for students.
Because students often lose a sense of direction (Treilibs 1979), planning and moni-
toring the solution process can help to keep an overview. As these processes involve
metacognitive strategies, metacognition can support students to work on the model-
ling problem in a goal-oriented way.
29.2.1 Metacognition in Modelling Processes
In the last decades, different conceptions of metacognition were carved out due to
differences in the perspectives on metacognition and roots of metacognitive
research. However, there exists a consensus about the subdivision of metacognition
29 Teachers’ Perspectives on Students’ Metacognitive Strategies during Mathematical… 337
into different components, namely, metacognitive knowledge, metacognitive experi-

ences, metacognitive strategies as well as a motivational component (Brown 1987;
Efklides 2008; Flavell 1979; Veenman 2011).
Metacognitive knowledge is characterised by knowledge about person, task and
strategy variables (Flavell 1979). Firstly, knowledge about personal characteristics
includes information about the person her/himself as well as about others as cogni-
tive subjects. In the modelling process, among others, teachers should have knowl-
edge about the mathematical competencies of the students involved in the modelling
process as well as knowledge about their working preferences. The students work-
ing on the modelling problem should be aware of their own preferences and skills
as well as of those working together with them. Secondly, knowledge about task
characteristics comprises the task and relations between the tasks and the ways to
solve them. This includes, for example, knowledge about the requirements for
developing an individual approach as well as for investigating the data as input or
knowledge required to make adequate assumptions. Lastly, knowledge about strat-
egy characteristics contains knowledge of strategies as well as the conditions for
their usage, which means knowledge about when and why a strategy should be
implemented and to what extent. The last aspect about the “when” and “why” of the
usage of strategies is also called conditional knowledge (cf. Veenman 2011). A goal-
oriented usage of cognitive strategies, such as rereading the task or underlining
important information given in the task, and the usage of heuristic strategies, such
as usage of analogies and working backwards, are helpful for productive work on
modelling problems.
Metacognitive skills “refer to the deliberate use of strategies (i.e., procedural
knowledge) in order to control cognition” (Efklides 2008, p. 280). First of all, this
includes strategies for planning. According to Hasselhorn and Gold (2006), the
planning process is about realising which goal should be and how it can be reached,
including choosing the strategy as well as agreeing on the order of the strategic
action. With regard to modelling processes, planning skills include (among others)
considering and understanding the task that has to be worked on by all involved
individuals. Students should discuss the aim of the modelling task and their solution
approach together. This should contain a discussion about different ideas for solving
the problem in order to identify the most appropriate ones. By doing so, the students
can guarantee that all students involved can follow the planning idea to work on the
problem, and further, they can make sure that they discuss about a goal-oriented
work (Chalmers 2009).
Furthermore, strategies for monitoring are useful not only for identifying the
discrepancy between the status quo and the aspired goals but also to critically
accompany an individual’s working progress. It is also important that students think
about impacts, namely, what their chosen strategies result in. If necessary, stimu-
lated by monitoring processes, regulating strategies should be used (Hasselhorn and
Gold 2006). During mathematical modelling processes, students can monitor their
process, for example, by using the modelling cycle as a tool for orientation. Skills
for monitoring require that the strategies applied are used in a systematic and
338 L. Wendt et al.
goal-oriented way. Students should be aware of cognitive barriers and should also
monitor their working behaviour.
Finally, strategies for evaluation are used after solving the modelling problem
and include an evaluation of the results, that is, whether they fit the chosen goals or
not, and an examination of the whole process, of the chosen strategies and of the
time management (Hasselhorn and Gold 2006). Strategies for evaluating the model-
ling process comprise, among others, a reflection on the processing of the whole
task concerning methods of operations as well as implemented strategies and its
benefits. Strengths and deficits of the working process need to be recognised in
order to improve the whole modelling process as well as to remember and apply it
in future modelling activities. For a detailed and task-specific discussion of meta-
cognitive processes in modelling, see Vorhölter and Kaiser (2016) and
Vorhölter (2018).
The respective discussion about problem-solving emphasises the importance of
using metacognitive strategies while processing a task. At first, a careful planning
process is important when working on complex problems (Schoenfeld 1992;
Verschaffel 1999). Furthermore, there is a positive impact of mutual support and
control in solving nonroutine tasks on the successful solution process (Goos 1998).
Regarding metacognitive knowledge, Schneider and Artelt (2010) point out that
high metacognitive knowledge leads to high mathematical performance. According
to Stillman (2011), students often have a low (or a lack of) meta-knowledge regard-
ing the transitions between single phases of the modelling cycle, which may lead to
cognitive barriers. At this point, the usage of metacognitive strategies is crucial
(Blum 2015). In this chapter, the focus is on metacognitive strategies.
29.2.2 eachers’ Perception of Students’ Metacognition

T
in Modelling Processes
Although metacognition can help strongly when working on a modelling problem,

students often do not use metacognition on their own, as using metacognition
requires not only knowledge and strategies but also an insight into the usefulness of
metacognition. Furthermore, students have to be taught how and when to use meta-
cognition (Veenman 2011). Therefore, teachers need to stimulate the students’
usage of metacognitive strategies.
When doing so, teachers are acting on a meta-meta level: They have to monitor
their students’ metacognition, which is an ambitious task for teachers. The success
of teacher interventions depends on their knowledge about modelling (Stillman
2011). Thus, teachers need to be aware which metacognitive strategies their stu-
dents apply. Referring to the concept of teachers’ noticing (Sherin et al. 2011), the
perception of students’ metacognitive modelling activities is considered. Secondly,
taking their perception as a basis, the teachers need to judge whether the
metacognitive strategies used are appropriate or not. If not, they have to redirect
their students in using metacognitive knowledge during the modelling process.
For diagnosing, teachers can perceive, according to Stillman (2011), whether
students use productive metacognition on one of the following levels: recognising
the relevance of strategies, choosing an appropriate strategy for implementation and
implementing this strategy successfully. That means among others, teachers have to
identify in which step students need support. Furthermore, they should identify the
so-called red flag situations, which are particularly significant and appear “when
students become aware of specific difficulties” (Goos 1998, p. 226) during the mod-
elling process. Red flag situations can be seen as metacognitive triggers (Goos
1998). When students are not able to implement appropriate metacognitive strate-
gies in those situations themselves, teachers need to guide them.
The theoretical discussion shows the importance of the teachers’ perception of the
students’ usage of metacognition. However, there are still no empirical findings
concerning the teachers’ perceptions of the students’ usage of metacognitive strate-
gies during mathematical modelling activities and how this teacher perception can
be fostered. The study aims at the reconstruction of teachers’ metacognitive percep-
tion relevant for teaching modelling, especially focusing on the perception of meta-
cognitive strategies used by students while working on a modelling problem.
Furthermore, the development of their perception during a ten-month intervention
is of special interest. In this chapter, the following research questions are tackled,
relating to the evaluation of the perception of one teacher:
• Which aspects influencing the teacher’s perception of metacognitive strategies
can be reconstructed at the beginning and the end of the intervention study?
• Which metacognitive strategies does the teacher explicitly notice while analys-
ing videotaped modelling processes? Is there a change from the beginning to the
end of the study?
29.4 Research Design and Methods of the Study
The qualitative study described in this chapter is embedded into a larger project car-
ried out at the University of Hamburg, called MeMo (Metacognitive Modelling
Competencies) (Vorhölter et al. 2019). It aims at the evaluation of a learning envi-
ronment fostering students’ metacognitive skills during modelling processes. In this
setting, this study focuses on the participating teachers’ perspective on students’
metacognition, whereas a parallel study focuses on the students’ perspective within
the same classes, which allows interesting comparisons (Krüger et al. this volume).
340 L. Wendt et al.
In this study, 14 teachers participated with their mathematics classes. They

implemented six modelling activities over a period of eight months. The students
worked on the problems in small groups of three to four students. The teachers
received the modelling tasks from the project group together with possible solu-
tions. Before working on the modelling problems, the teachers introduced the mod-
elling cycle of Kaiser and Stender (2013) to the students. Most students had not
worked on a modelling problem before and therefore needed to be introduced into
the modelling process. Some teachers visualised single steps of the modelling cycle
with the help of a specific modelling problem. This encouraged the students to
develop metacognitive knowledge about task characteristics (Vorhölter and
Kaiser 2016).
A detailed description of the whole study including information about the sam-
ple and an example of one of the modelling tasks used in the study can be found in
Vorhölter (2018). In this chapter, the presented case of Ms. Schmidt is an example
of a young teacher, who had been working for five years at the time of the study at
a higher academic track school within an urban area of higher socio-economic sta-
tus. Overall, the school has a learning supportive climate and performs above aver-
age in the annual federal testing.
Each of the modelling problems aimed at deepening one of the metacognitive
skills. Teachers were asked in accompanying training courses to focus on this par-
ticular metacognitive skill in these modelling activities by stimulating it during the
work on the modelling problems as well as discussing the relevance of these kinds
of metacognitive strategies and their implementation in the modelling process at the
end of each modelling activity. During the first two modelling activities, the usage
of evaluation strategies was focused on, whereas during the next two activities, the
planning strategies were highlighted. Finally, monitoring strategies were empha-
sised during the last two modelling activities.
During the school year, three teacher trainings were offered, one immediately
before the first two modelling activities and each lasting two to three hours. The
teacher trainings included theoretical discussions, a didactical preparation focusing
on the six modelling problems as well as a session based on own activities, where
the teachers worked on the modelling problems themselves. The didactical prepara-
tion of the modelling problems comprised (among others) different ideas to solve
each of these modelling problems in order to make the teachers aware of different
solution approaches, which are always possible in modelling processes. Furthermore,
the teachers were supported in anticipating students’ difficulties and were acquainted
with possible adaptive teacher interventions related to each modelling problem.
For evaluating the teachers’ perceptions of metacognition, different methods can
be used. Firstly, teachers’ perception can be reconstructed from their classroom
activities when working with the students as perception is a basis for acting.
However, in classrooms, other factors than teachers’ perceptions of students’ cur-
rent problems may influence their actions resulting in activities not consistent with
their perception. Secondly, teachers can be asked to describe their perceptions. As
this is not possible during a lesson, teachers’ perceptions have to be collected after-
wards. Actually, in this case, one does not gather teachers’ perceptions but only
those perceptions expressed. Using interviews especially helps understanding the
teachers’ motivation and purposes for their activities and allows further requests in
the interview if needed.
Weighing pros and cons of the different approaches, in this study, data was col-
lected by following the three-step design of Busse and Borromeo Ferri (2003),
which includes a process observation, a stimulated recall and a focused interview.
Therefore, the first modelling activity within the intervention study and the last one
were videotaped by using one camera for each student group comprising three to
four students, that means in total about six to seven groups. The videos of the stu-
dent groups were analysed directly after the lesson concerning interesting video
scenes relating to either students’ usage of metacognitive strategies or situations, in
which students failed to use metacognitive strategies or video scenes dealing with
the teacher’s support of the students in using metacognitive strategies. The selected
video scenes were cut to make video vignettes of approximately two min length. In
the interviews, seven scenes were shown on average as part of the stimulated recall.
In every interview, at least one intervention scene was presented; in the other scenes,
the teachers were not involved in the working processes and thus most times had not
noticed the process shown in the scene before. The interviews took place on the next
day of schooling, and the teachers were shown scenes of their own classes. The
teachers were asked to describe and evaluate the shown video scenes and answer
further questions. All in all, the interviews lasted from approximately 30 to 60 min
at the beginning of the study and from 45 to one and a half hours at the end of the
study. The differences in the lengths can be explained by the elaborateness of the
teachers’ answers on the one hand but also by the quantity of video scenes shown to
the teachers on the other hand. Because no quantitative evaluation of the data was
planned, we made an effort to show video scenes of broad variety in foci and issues
tackled.
Data collection is already finished, and the collected data are currently being
analysed using qualitative text analysis according to Kuckartz (2016). The inter-
views were transcribed and coded with the computer software Maxqda. After exe-
cuting a qualitative content analysis, an evaluative content analysis is currently
being carried out to develop a typology of teachers’ perceptions on metacognition.
Therefore, the category system was developed deductively and inductively by dif-
ferentiating theory-based main categories mentioned above, namely, metacognitive
knowledge, metacognitive experiences and metacognitive strategies during teacher
interventions, that is, orientation, planning, monitoring and regulation and evalua-
tion. The category system as well as the whole coding manual was discussed in
groups of experts several times. All results are based on the analysis of the inter-
views (not of the videos) as well as a short questionnaire on personal data.
342 L. Wendt et al.
29.5 First Results: The Case of Ms. Schmidt
The results presented in this chapter describe the development of a selected teacher,
who showed a remarkable change in her perception on students’ metacognitive
activities. At first, background information about the school and the teacher are
given. Then, changes in reconstructed aspects influencing the perception of meta-
cognitive strategies are presented. Furthermore, the development of the teacher’s
perception of the students’ usage of metacognitive strategies for planning, monitor-
ing and evaluating is described.
The teacher—Ms. Schmidt—was highly motivated from the beginning of the
study onwards. She even convinced her colleagues to participate as well. After the
last modelling activity, she decided to support all the other mathematics teachers of
her school when introducing mathematical modelling into their lessons. Due to the
first teacher training, which took place before the first modelling activity, she had
basic knowledge about the modelling process and requirements of modelling prob-
lems. She reported before the study that she had low experience in teaching model-
ling problems and that her students had also not been familiar with modelling
problems.
29.5.1 Aspects Influencing the Perception of Metacognition
At the beginning of the study, although being introduced to the concept of metacog-
nition, Ms. Schmidt was not able to differentiate reliably between different meta-
cognitive skills of the students, at least on a conceptual level: When she was asked
to highlight her impression of the students’ usage of metacognitive strategies, she
was uncertain about the metacognitive strategies and, thus, demanded the defini-
tions of the individual metacognitive strategies:
Ms. Schmidt: Huh, so that they think about their results or what do you mean
with those?
Interviewer: For example.
Ms. Schmidt: That they check themselves or each other?
This excerpt shows that she had an idea of individual constituents of modelling
processes and metacognition, but she had no deeper understanding of all parts. In
addition, at the beginning of the study, she had a limited understanding of evaluation
that means she reduced this strategy to validating the mathematical solution instead
of evaluating the whole modelling process.
However, she noticed that her students had very little metacognitive knowledge
about the modelling process, especially about what is required with respect to the
task characteristics. Due to low or even lacking experience in mathematical model-
ling, the students were not aware that it is essential to make assumptions and that
working together during modelling processes is beneficial. The teacher’s descriptions
of her perception of the students’ activities were indicators for metacognitive knowl-
edge by Ms. Schmidt about the task, personal knowledge of the modellers involved
and strategy characteristics, which were presented to her during the first teacher
training. The teacher was able to apply these aspects in order to analyse the scenes
shown to her during the interview.
In comparison, at the end of the study, she noticed a development of the students’
metacognitive knowledge regarding the modelling process. According to Ms.
Schmidt, the students knew that it was crucial to make assumptions and consider the
whole modelling process (at the beginning of the study, they did not validate their
mathematical model and mathematical solutions):
Ms. Schmidt: At some point, they just used the modelling cycle without any assis-
tance or support and without even thinking about it too much,
because they realised that they need to make assumptions. Without
making assumptions, it’s not possible. Therefore, they definitely
took a step forward.
Furthermore, the teacher noticed that at the end of the study, her students were
aware of the necessity to compare different solution ideas and based on this obser-
vation described an increase in the students’ metacognitive knowledge with regard
to task characteristics. This observation can serve as an indicator for an increase in
the teacher’s metacognitive knowledge about strategy characteristics.
29.5.2 Metacognitive Skills
In the following, changes in the teacher’s perception of planning processes as part

of metacognitive skills are presented followed by elaborations on skills on monitor-
ing and regulating and, lastly, evaluation skills.
Planning Skills Concerning metacognitive strategies for planning at the beginning
of the study, the teacher perceived that joint planning processes were missing. Ms.
Schmidt noticed individual approaches to plan the solution process as only one
person, the leader of the group, tried to plan the process by comparing different
solution ideas. Nevertheless, the group did not discuss these ideas together:
Ms. Schmidt: Sarah has an idea and presents her idea to the other students, and
says “let’s do this”, she does not ask whether the other group mem-
bers have an alternative idea.
When perceiving this, the teacher stated that the students needed to become more
active as a group and needed to consider possible alternatives suggested by all group
members. Consequently, she emphasised this aspect by trying to stimulate the com-
munication and participation of all group members in the following modelling
activities. At the end of the study, the teacher showed an increased awareness and
intensified perception of metacognition. She described weaknesses and strengths of
the different groups in depth; for example, planning activities of the whole
344 L. Wendt et al.
modelling process were still missing in a few groups; in some cases, the students
still planned their modelling process step by step. Moreover, the teacher recognised
a sense of direction of her students (Treilibs 1979) by perceiving that they followed
the structure of making assumptions, mathematising and reflecting.
Overall, Ms. Schmidt’s efforts of fostering her students’ planning skills showed
her own belief about the importance of planning and working together coopera-
tively. These efforts went along with an improvement of the students’ planning
skills, as reported in the connected study within this project focusing on students. A
few student groups even developed their own procedures for planning during the
intervention (Krüger et al. this volume).
Monitoring and Regulating Skills At the beginning of the study, Ms. Schmidt
perceived strategies for monitoring and regulating only partially. Again, she identi-
fied an individual student who monitored the process and at the same time led the
group work. For example, this student monitored every group member’s under-
standing of the modelling problem in order to guarantee that everybody is able to
work on the problem by questioning other students. However, the other group mem-
bers did not monitor each other mutually. The teacher did not perceive (or at least
did not express) this lack of mutual monitoring but described that there was a joint
discussion of the mathematical model. The whole group discussed whether their
chosen mathematical model was appropriate or not. Furthermore, the teacher
became aware of the fact that the students failed to monitor with respect to the
whole modelling process, especially because they failed to validate the mathemati-
cal model and solutions.
In comparison, at the end of the study, Ms. Schmidt showed a more holistic per-
ception of the increased student usage of monitoring strategies. She recognised an
automatisation of monitoring processes and an independent application of monitor-
ing strategies without further stimulation by the teacher:
Ms. Schmidt: Well, I would say it is almost ideal that they act independently with-
out me involved (…) but they’re thinking of ways to improve their
work. Or where are open questions? Are there assumptions that are
not totally correct? All in all, the whole time quite well monitored
and regulated.
Moreover, the teacher perceived a successful stimulation of monitoring pro-
cesses by using the modelling cycle as well as the successful implementation of
systematic reflection of strategies. She realised that by setting a goal for each indi-
vidual group for every modelling activity, the students remembered their goals dur-
ing task processing and monitored these.
Evaluating Skills As already mentioned, at the beginning of the study, the teacher
had a limited understanding of evaluation. Her perception of validation was reduced
to validation of a mathematical solution instead of evaluating the whole modelling
process. With this incorrect understanding, she perceived only that a student was
lacking validation activities towards the mathematical solutions.
At the end of the study, her understanding of evaluation was extended to an

assessment of the whole modelling process. The teacher perceived the students’
development in a positive way: At the end of the intervention study, the students
evaluated their modelling process. After going through the modelling cycle, they
reflected on what they had done, and in the case that they were not satisfied with
their evaluation of the first round through the modelling cycle, they carried out
another round until they were satisfied with their evaluation.
29.6 Summary and Limitations
These first results show that teachers’ perception about metacognitive strategies and
the modelling process is developing by own modelling activities with their class and
by professional development in mathematical modelling. Thereby, the teachers’
perceptions of students’ usage of metacognitive strategies can be strengthened.
Based on their own perception, teachers can support their students in using meta-
cognitive strategies while working on modelling problems.
The results were reconstructed by analysing the expressed teacher’s perception
of students’ metacognitive processes by interviews. Surveying teachers’ percep-
tions could also be done by analysing video material and analysing the observable
behaviour in the videos. The study offers a wide range of video material so that
results can be validated. Based on the analyses presented in this paper on a teacher’s
perception of metacognition in modelling, the focus will be extended in the further
analyses to teacher interventions, which build on the teachers’ perceptions.
Furthermore, the results will be related to the results of further studies within the
project; in particular, students’ perspectives on metacognitive strategies will be
compared with the corresponding teachers’ perspectives (Krüger et al. this volume)
as well as students’ development of metacognitive modelling competencies (first
results can be found in Vorhölter 2018) with teachers’ perceptions.
References
Blum, W. (2015). Quality teaching of mathematical modeling: What do we know, what can we do?
In S. J. Cho (Ed.), The proceedings of the 12th international congress on mathematical educa-
tion (pp. 73–96). Cham: Springer.
Brown, A. L. (1987). Metacognition, executive control, self-regulation, and other more mysterious
mechanisms. In F. E. Weinert & R. H. Kluwe (Eds.), Metacognition, motivation, and under-
standing (pp. 65–116). Hillsdale: Lawrence Erlbaum.
Busse, A., & Borromeo Ferri, R. (2003). Methodological reflections on a three-step-design com-
bining observation, stimulated recall and interview. ZDM Mathematics Education, 35(6),
257–264.
Chalmers, C. (2009). Group metacognition during mathematical problem solving. In R. K. Hunter,
B. A. Bicknell, & T. A. Burgess (Eds.), Crossing divides (pp. 105–111). Palmerston
North: MERGA.
346 L. Wendt et al.
Efklides, A. (2008). Metacognition. Defining its facets and levels of functioning in relation to self-
regulation and co-regulation. European Psychologist, 13(4), 277–297.
Flavell, J. H. (1979). Metacognition and cognitive monitoring: A new area of cognitive-
developmental inquiry. American Psychologist, 34(10), 906–911.
Goos, M. (1998). I don’t know if I’m doing it right or I’m doing it wrong! Unresolved uncer-
tainty in the collaborative learning of mathematics. In C. Kanes, M. Goos, & E. Warren (Eds.),
Teaching mathematics in new times (pp. 225–232). Gold Coast: MERGA.
Hasselhorn, M., & Gold, A. (2006). Pädagogische Psychologie. Erfolgreiches Lernen und Lehren.
Stuttgart: Kohlhammer.
Kaiser, G., & Stender, P. (2013). Complex modelling problems in co-operative, self-directed learn-
ing environments. In G. Stillman, G. Kaiser, W. Blum, & J. P. Brown (Eds.), Teaching math-
Krüger, A., Vorhölter, K., & Kaiser, G. (this volume). Metacognitive strategies in group work
in mathematical modelling activities. In G. A. Stillman, G. Kaiser, & E. Lampen (Eds.),
Mathematical modelling education and sense making. Cham: Springer.
Kuckartz, U. (2016). Qualitative Inhaltsanalyse, Methoden, Praxis, Computerunterstützung.
Beltz: Weinheim.
Schneider, W., & Artelt, C. (2010). Metacognition and mathematics education. ZDM Mathematics
Education, 42, 149–161.
Schoenfeld, A. H. (1992). Learning to think mathematically: Problem solving, metacognition, and
sense-making in mathematics. In D. A. Grouws (Ed.), Handbook of research on mathematics
Sherin, M. G., Jacobs, V. R., & Philipp, R. A. (2011). Situating the study of teacher noticing: Seeing
through teachers’ eyes. In M. G. Sherin, V. R. Jacobs, & R. A. Philipp (Eds.), Mathematics
teacher noticing: Seeing through teachers’ eyes (pp. 4–13). New York: Taylor and Francis.
Stillman, G. A. (2011). Applying metacognitive knowledge and strategies in applications and mod-
Springer.
Treilibs, V. (1979). Formulation processes in mathematical modelling. Thesis submitted to the
University of Nottingham for the degree of Master of Philosophy.
Veenman, M. V. J. (2011). Learning to self-monitor and self-regulate. In R. E. Mayer &
P. A. Alexander (Eds.), Handbook of research on learning and instruction (pp. 197–218).
New York: Routledge.
Verschaffel, L. (1999). Realistic mathematical modelling and problem solving in the upper ele-
mentary school: Analysis and improvement. In J. H. M. Hamers, J. E. H. van Luit, & B. Csapó
(Eds.), Contexts of learning: Teaching and learning thinking skills (pp. 215–240). Lisse: Swets
& Zeitlinger.
Vorhölter, K. (2018). Conceptualization and measuring of metacognitive modelling competen-
cies – empirical verification of theoretical assumptions. ZDM Mathematics Education, 50(1),
343–354.
Vorhölter, K., & Kaiser, G. (2016). Theoretical and pedagogical considerations in promoting stu-
dents’ metacognitive modeling competencies. In C. Hirsch (Ed.), Annual perspectives in math-
ematics education 2016: Mathematical modeling and modeling mathematics (pp. 273–280).
Reston: NCTM.
Vorhölter, K., Krüger, A., & Wendt, L. (2019). Metacognition in mathematical modeling – An over-
view. In S. Chamberlain & B. Sriraman (Eds.), Affect and mathematical modeling (pp. 29-51).
Cham: Springer.
Part IV
Assessment of Mathematical Modelling
in Schools
Chapter 30
Cognitive Diagnostic Modelling
for Mathematical Modelling Assessment
Cigdem Alagoz and Celil Ekici
Abstract In a professional development workshop, we measured mathematical

modelling as a performance assessment representing a multi-dimensional construct
where dimensions are established with the performance indicators provided in a
rubric. The study showed that a multi-dimensional non-compensatory psychometric
model for scoring mathematical modelling performance provided a statistical model
that fit to the assumptions of the mathematical modelling assessment. Empirical
results are presented from a mathematical modelling task. We demonstrated that we
can identify distinct attribute mastery profiles and provide fine-grained individual-
ized feedback. Performing such analysis for teachers helped us identify professional
development needs for teaching mathematical modelling and applications.
Keywords Cognitive diagnostic models · Mathematical modelling (assessment) ·

Multidimensionality (of mathematical modelling construct) · Professional
development · STEM · Teachers
30.1 Introduction
Mathematical modelling has been gaining emphasis in education. Adopting an edu-

cational perspective, mathematical modelling in the teaching of mathematics, both
as a means for learning mathematics and a competence in its own right, is of crucial
importance in not only mathematics but also in the sciences (Blum and Niss 1991).
In the USA, for example, it is one of the eight standards of mathematical practice as
recommended by the Common Core State Standards for Mathematics (CCSSM)
C. Alagoz (*)
Division of Research and Analysis, Texas Education Agency, Austin, TX, USA
C. Ekici
Department of Mathematics and Statistics, Texas A&M University-Corpus Christi,
Corpus Christi, TX, USA
e-mail: celil.ekici@tamucc.edu

350 C. Alagoz and C. Ekici
and as a conceptual category for high school (NGACBP and CCSSO 2010, p. 7).
The CCSSM classify modelling as “the process of choosing and using appropriate
mathematics and statistics to analyze empirical situations, to understand them bet-
ter, and to improve decisions” (p. 72). Moreover, students are expected to practice
their mathematical modelling skills in science classes as well (NGSS Lead
States 2013).
Researchers have tried to answer questions such as “What do we mean by model-
ling competencies?” (Maaß 2006). Mathematical modelling competency refers to
the ability to construct, or use, or apply mathematical models to propose solutions
to a real-life problem (Blum and Borromeo Ferri 2009). The modelling process
begins the process of simplifying a real-world problem with a real model, idealizing
it with an explicit set of assumptions, and mathematizing it with a mathematical
model. The mathematical problem-solving process then produces a solution to be
interpreted and validated (Blum 1996, p.18). If the process does not yield a realistic
solution, the entire modelling process or its parts need to be worked through again.
The Guidelines for Assessment and Instruction in Mathematical Modelling
Education were recently offered as a resource for teachers to incorporate mathemat-
ical modelling into their teaching practices (Bliss et al. 2016). These state the main
idea for assessment is that it “should focus on the process and not on the product or
pieces only” (Bliss et al. 2016, p. 47). The focus on the process has been a consistent
pattern in the NCTM’s curricular recommendations since the standard’s movement
with five sub-competencies began (Bleiler-Baxter et al. 2017; NCTM 2000).
In assessing mathematical modellers’ performance, Jensen remarked that “multi-
dimensionality is necessary yet challenging” (Jensen 2007, p.141). Ekici and Alagoz
(2018) demonstrated the multidimensionality of the mathematical modelling con-
struct with Confirmatory Factor Analysis for the Queen Conch Modelling Task
which is the same task used in the current study. Towards incorporating multidimen-
sionality in modelling assessments, Zöttl et al. (2011) presented a multidimensional
IRT approach for scoring a mathematical modelling test and compared it to scoring
with a unidimensional IRT model. They showed empirically the superiority of a sub-
dimensional model. These authors noted that the compensatory assumption which is
made for multidimensional IRT analysis is problematic for scoring mathematical
modelling. A compensatory model assumes that a low proficiency in one sub-dimen-
sion can be compensated by high proficiency in another sub-dimension. For mathe-
matical modelling assessment, this assumption is unrealistic as competency in
problem-solving may not be compensated by competency in communication, for
example. Zöttl et al. (2011) further suggested that it would be more appropriate to
use a non-compensatory approach and noted the unavailability of such implementa-
tion in psychometrics and in the common software systems at the time.
Currently, there are non-compensatory models and software available to answer
this mathematical modelling assessment problem. Building on these recent devel-
opments in psychometrics (Rupp et al. 2010), we are analysing mathematical mod-
elling with a non-compensatory model. Furthermore, we demonstrate how to
provide rich information about students’ knowledge states in the measurement and
assessment of mathematical modelling from a cognitive diagnostic assessment
perspective.
30 Cognitive Diagnostic Modelling for Mathematical Modelling Assessment 351
30.1.1 Purpose of the Study
Here, we suggest a mathematical modelling assessment approach construed as a

process which can lend itself to demonstrate the multidimensionality of a modelling
task performance and provide feedback regarding the fine-grained performance of
each learner and the latent features of the task itself. This process is to inform teach-
ers and modelling assessment task designers about inherent features of mathemati-
cal modelling performance as enacted by the intended assessments. In finding a
best-fitting statistical model to represent mathematical modelling which is both
multidimensional and non-compensatory, our research question is: Based on empir-
ical evidence, is it statistically appropriate to measure the mathematical modelling
construct by use of a non-compensatory cognitive diagnostic model?
30.2 Method
Cognitive Diagnostic Models (CDM) are used to provide diagnostic feedback to

examinees about their strengths and weaknesses in responding to test items. Scores
obtained using these models are different from score-based information that is pro-
vided by the usual classical test and item response theory models. These CDMs
have the potential to be a new way of providing feedback to teachers and students.
The non-compensatory model used in this study is the Deterministic Inputs, Noisy
“And” gate (DINA) model (Haertel 1989).
30.2.1 A Non-compensatory CDM: DINA Model
Item response analysis starts with the construction of a Q-matrix. A Q-matrix is

defined as a matrix in which attributes are indicated that are required to solve each
item on the test. Q is a J × K matrix of zeros and ones, its element on the jth row,
and kth column is qjk, which is the indicator of whether skill k is required to produce
a correct response for the jth item. A correct response is produced when an exam-
inee masters all the attributes indicated in the Q-matrix. The probability of a correct
response for an examinee is the same if he/she has only some of the required attri-
butes or none of the required attributes. In this sense, the DINA model is a conjunc-
tive stochastic model where mastering all the attributes indicated in the Q-matrix
does not guarantee a correct response and missing one or more of these attributes
does not result in an incorrect response. For each examinee and for each item, an
indicator, η, is calculated for mastering required skills for that item. η equals 1 when
an examinee possesses all the required skills for a correct response and equals 0
when an examinee does not possess one or more of the required skills. s refers to the
probability of slipping and missing the item when an examinee possesses all the
required skills. g refers to the probability of guessing and giving a correct response
when an examinee does not possess all the required skills. When an examinee has
all the required skills to solve an item, the probability of correct response equals
1 − s. When an examinee misses any one of the skills necessary to solve the item,
probability of correct response is equal to g. The probability of giving a correct
response for a given skill set is defined as P(X = 1) = (1 – s)η g 1-η. The DINA model
is appropriate for tests for which the conjunction of several attributes is required to
achieve a correct response and missing one required attribute produces an incorrect
response. It is easy to interpret, and the model and the algorithms are available as
well as a number of applications.
30.3 Setting and Participants
As a part of an NSF-funded STEM project, in-service secondary mathematics and

science teachers attended a professional development program. In summer of 2017,
one of the workshop themes was integrated project-based learning with mathemati-
cal modelling as a common interdisciplinary practice, and another was educational
measurement of mathematical modelling and integrated STEM learning. This is a
similar in-service project-based mathematical modelling training to that of Blomhøj
and Kjeldsen (2006) with more interdisciplinary collaboration among participants.
Mathematical modelling as an interdisciplinary process was discussed in workshop
sessions from both an instructional perspective in one session and from an assess-
ment perspective in another session focusing on educational measurement of learn-
ing outcomes with mathematical modelling. Based on their needs and backgrounds,
teachers were placed in two groups as a modelling group and an assessment group
with approximately 20 teachers in each. In-service teachers engaged in the model-
ling process, using their own ideas in creating mathematical models for locally rel-
evant problems such as the queen conch population. In total, 28 teachers completed
mathematical modelling tasks. The assessment rubrics and analysis of the results
provided a hands-on investigation of content and processes involved in mathemati-
cal modelling practice and its assessment.
30.3.1 Mathematical Modelling Tasks
The mathematical modelling tasks used in this study were originally produced by
Dance and Sandefur (2004) as a part of an earlier NSF grant producing nontradi-
tional locally relevant curriculum materials to increase the participation and success
of underrepresented minority students in mathematics and the sciences with accom-
panying assessment materials for introductory college mathematics courses. The
modelling activity that we used for this study had the following driving question:
“Will your grandchildren have queen conch?” Fig. 30.1 is an example page of this
mathematical task used with a teacher’s work. The formative assessment tool asks
Fig. 30.1 Sample page from the Queen Conch Population Modelling Task
teachers to model the changes in the queen conch population (Dance and Sandefur
2004). Tasks include a) making a graph and evaluating the graph at different values
of x, b) evaluating the function where x = 100 and finding x-intercepts and the point
x = 100, and c) providing a conch fishing scenario, revising the growth function, and
finding out what happens when the fishing continues to be at the rate of 10%. Based
on their models, participants are expected to answer the question: “Would you be
worried about catching 10% of the conchs if you were interested in having conch
when your children grow up?” In addition to this model which represented a con-
tinuous model to study logistic behaviour, we introduced a discrete method as well
to model the data for the local conch population.
30.3.2 Rubric as a Developmental Tool
During the workshop, teachers were presented with a rubric and asked to critically
evaluate the modelling task performances. This rubric was developed by the
New York Performance Standards Consortium (NYPSC 2016) to assess mathemati-
cal modelling task performance. Consistent with our approach to mathematical
modelling, mainly as a process in learning and applying mathematics rather than as
separate content, we studied initially the competencies in the mathematical model-
ling process aligning with the process standards described by the NCTM (2000).
Teachers were first asked to work on the modelling task in pairs and prepare their
responses. Then, the first group evaluated the responses of the other group and vice
versa. Each answer was evaluated by at least two teachers allowing us to observe
rater consistency.
The NYPSC rubric accommodates scoring of a student’s performance with the
mathematical modelling by attending to five process standards of school mathemat-
ics: problem-solving (PS), reasoning and proof (R&P), communications (Com),
connections (Conx), and representations (Rep) (NCTM 2000). Anhalt and Cortez
(2015) observe that within the process of modelling, students are naturally engaged
with other standards of mathematical practice and process standards (NCTM 2000,
2010). As seen in Table 30.1, a Q-matrix is formed from the expected learning out-
comes driven by the descriptions provided in the NYPSC rubric. Q-matrix columns
are created with the five processes. Q-matrix entries indicate the relationship
between these five processes and the performance indicators measured by the rubric.
These performance indicators have commonalities with the modelling competencies
described by Blum and Leiβ (2007) as (1) understanding the task, (2) simplifying/
Table 30.1 Q-matrix identification for modelling performance indicators and process skills
Modelling process skills
Modelling performance indicators PS R&P Com Conx Rep
Selects appropriate strategies 1 0 0 0 0
Executes conceptually sound mathematical procedures accurately 1 0 0 0 0
Makes major conceptual or procedural errors 1 0 0 0 0
Justifies mathematical statements accurately 0 1 0 0 0
Draws valid conclusions 0 1 0 0 0
Constructs a generalization and uses to make predictions 0 0 1 0 0
Uses mathematical terminology and notation appropriately 0 0 1 0 0
Clearly communicates process and solution 0 0 1 0 0
Discusses how mathematics interconnect and build on each other 0 0 0 1 0
Applies concepts to real-world situations 0 0 0 1 0
Creates appropriate models inherent to the tasks that represent the 0 0 0 0 1
problem accurately
Note: Mathematical modelling skills are abbreviated as follows: PS problem-solving, R&P reason-
ing and proof, Com communication, Conx connection; and Rep, representations
structuring the task, (3) mathematizing, (4) working mathematically, (5) interpret-
ing, and (6) validating the solution.
The modelling of the groups with similar mathematical modelling competency
levels starts with the creation of the Q-matrix by content experts. Each mathemati-
cal modelling performance task measures one or more competencies as depicted by
the Q-matrix. Table 30.1 presents performance indicators along with the Q-matrix
which corresponds to the required performance indicators for each attribute mea-
sured by the Queen Conch Modelling Task. Utilizing this Q-matrix and DINA model
(de la Torre and Douglas 2004), item response data are analysed. The model is
estimated using an Ox program written by de la Torre (2009).
30.4 Results
First, we evaluated the inter-rater consistency by looking at the correlations between

two ratings. As seen in Table 30.2, the correlations are found to be strong and sig-
nificant for problem-solving and reasoning and proof (>.86). The ratings are not
consistent for communication skill according to the correlation of .355 which is not
significant. Teachers do not seem to agree on Com skill when they used the rubric,
but their ratings were somewhat consistent for Conx skill. These results indicate that
Com skill requires further calibration among teachers as it applies to this modelling
task. This is also informative from the task design perspective as the modelling task
can be revised to provide clearer prompts regarding communication skill for a mod-
eller to demonstrate competence.
The item parameter estimates are provided in Table 30.3 with corresponding
standard errors. Most items had very small guessing and slipping estimates, mean-
ing teachers who had not mastered all the required skills for each item had a very
small chance of answering that item. Similarly, teachers who had mastered each
skill had very small chance of not answering the item correctly. Only item 6 had 0.4
slipping estimate which means that 40% of teachers who mastered all the required
skills to answer item 6 correctly could not/did not do so.
Our assessment measured five attributes, which creates 25 possible attribute mas-
tery patterns. These mastery patterns establish latent groups based on the individu-
als’ attribute probabilities estimated from the CDM analysis. They are called latent
because they are estimated from the data and not observed directly. The latent
classes and their estimated probabilities are provided in Table 30.4. The most likely
Table 30.2 Correlations between two sets of ratings for modelling skills
Modelling process skills
Problem- Reasoning
solving and proof Communication Connections Representations
Rater accuracy .86∗∗ .89∗∗ .355 .74∗ .66∗
correlations
Table 30.3 Item parameter estimates

Item Guess SE(Guess) Slip SE(Slip)
1 .000001 0.7 0.000001 0.2
2 .000001 0.7 0.000001 0.2
3 .000001 0.7 0.000001 0.2
4 .000001 0.4 0.000001 0.2
5 .000001 0.4 .000001 0.2
6 .000001 0.4 .4 0.1
7 .000001 0.3 .000001 0.3
8 .000001 0.3 .000001 0.3
9 .000001 0.2 .000001 0.4
10 .000001 0.2 .000001 0.4
11 .000001 0.2 .0005 0.3
Table 30.4 Latent classes and their posterior probabilities

Latent Latent Latent Latent
class# Probability class# Probability class# Probability class# Probability
00000 .07 01000 <.001 10,000 .11 11,000 0.25
00001 <.001 01001 <.001 10,001 <.001 11,001 0.11
00010 <.001 01010 <.001 10,010 <.001 11,010 <.001
00011 <.001 01011 <.001 10,011 <.001 11,011 <.001
00100 <.001 01100 <.001 10,100 <.001 11,100 .07
00101 <.001 01101 <.001 10,101 0.07 11,101 .07
00110 <.001 01110 <.001 10,110 <.001 11,110 .07
00111 <.001 01111 <.001 10,111 <.001 11,111 0.18
profile was 11,000 where the first two attributes were mastered and the last three
attributes were not mastered yet. This latent group comprises 25% of the sample.
The second largest group was 18% of the sample where all the attributes have been
mastered. The latent group mastering only the first attribute was 11% of the sample.
Attribute mastery estimates are provided in Table 30.5. The first attribute which
was problem-solving has been mastered by 93% of our teacher sample. The fourth
attribute, connections, has been mastered by the least percentage (25%) of the sam-
ple. Both communication and representation attributes have been mastered by less
than 50% of the teachers in our sample.
An example of individual feedback is depicted in Fig. 30.2 where mastery levels
in the process skills are compared for two individuals. From Fig. 30.2, it can be seen
that Teacher A mastered problem-solving and reasoning and proof and did not mas-
ter connections and representations. This teacher’s mastery in communication as
proficiency was 50% and therefore needs improvement to reach mastery. Teacher B
mastered problem-solving, reasoning, and proof, and representations, communica-
tion, and connections were not mastered.
Table 30.5 Estimates of Attribute Mastery

attribute mastery prevalence
1 Problem-solving .93
2 Reasoning and proof .75
3 Communication .46
4 Connections .25
5 Representations .43
Fig. 30.2 Example feedback presenting profiles for two individuals’ attribute mastery probabili-
ties based on their mathematical modelling performance
This study aimed to evaluate the assessment of mathematical modelling with a cog-
nitive diagnostic model. The current statistical model answered the need for a mul-
tidimensional and non-conjunctive modelling of mathematical modelling
assessment. The results showed the feasibility and interpretation of a multidimen-
sional non-compensatory psychometric model for scoring mathematical modelling
performance providing individualized and group feedback to learners, in this case
teachers in a professional development workshop.
We have demonstrated that we can identify distinct attribute mastery profiles
among learners and provide fine-grained individualized feedback. Performing such
analysis for teachers helped us to identify professional development needs on teach-
ing mathematical modelling and applications. To illustrate, the data analysis indicated
that teachers had difficulty with connections with fewer teachers demonstrating mas-
tery than other attributes. On the other hand, problem-solving is the dimension where
teachers performed best. As making connections was the weakest aspect of mathe-
matical modelling, it can be concluded that more training and support are needed for
teachers. To support interdisciplinary connections, we recommend interdisciplinary
professional development programs where mathematics and science teachers can
support each other to develop richer and more meaningful connections and interpreta-
tions with modelling and applications. Communications and representations as per-
formance attributes are areas where these teachers need more improvement as well. It
should also be noted that the inter-rater reliability among the teacher raters for
communications was not within the acceptable range which in turn increased the
measurement error for this attribute. This indicates that a consensus on a clear defini-
tion and expected mastery from the communications attribute is lacking among teach-
ers. Training and further professional discussion on this topic would increase our
understanding of the outcome expectations from mathematical modelling and
increase the quality of measurement and the diagnostic score for this attribute.
The results of this study provided insight into the mathematical modelling con-
struct as elicited by the modelling task used while analysing the scoring of this
multidimensional construct. As a follow-up, impact of teachers’ alternative specifi-
cations and mis-specifications of Q matrices in their teaching practices with math-
ematical modelling and applications could be examined. Operationalization of the
mathematical modelling ability within the assessment material, as well as the rubric,
played a crucial role in the outcome of the psychometric analysis. This process also
informs modelling task designers in how to enhance their assessments to better
elicit skills such as representations based on the modelling task performance data.
Acknowledgements Part of this work was supported by National Science Foundation Grant No.
1355437. Any opinions, findings, and conclusions or recommendations expressed in this material
are those of the author(s) and do not necessarily reflect the views of the National Science
Foundation.
References
Anhalt, C. O., & Cortez, R. (2015). Mathematical modeling. Mathematics Teacher, 108(6),
447–452.
Bleiler-Baxter, S. K., Stephens, D. C., Baxter, W. A., & Barlow, A. T. (2017). Modeling as a
decision-making process. Teaching Children Mathematics, 24(1), 20–28.
Bliss, K., Fowler, K., Galluzzo, B., Garfunkel, S., Montgomery, M., et al. (2016). GAIMME:
Guidelines for assessment and instruction in mathematical modeling education. Philadelphia:
Consortium for Mathematics and Its Applications (COMAP) & Society for Industrial and
Applied Mathematics (SIAM).
Blum, W. (1996). Anwendungsbezüge im Mathematikunterricht—Trends und Perspektiven.
Schriftenreihe Didaktik der Mathematik, 23, 15–38.
Blum, W., & Borromeo Ferri, R. (2009). Mathematical modelling: Can it be taught and learnt?
Blum, W., & Leiβ, D. (2007). How do students and teachers deal with modelling problems? In
C. Haines, P. Galbraith, W. Blum, & S. Khan (Eds.), Mathematical modelling- education, engi-
neering and economics (pp. 222–231). Chichester: Horwood.
Blum, W., & Niss, M. (1991). Applied mathematical problem solving, modelling, applications, and
links to other subjects. Educational Studies in Mathematics, 22, 37–68.
Dance, R. A., & Sandefur, J. T. (2004). Will your grandchildren have queen conch. Available http://
www.uvi.edu/files/documents/College_of_Science_and_Mathematics/math_skills/19.pdf
de la Torre, J. (2009). DINA model and parameter estimation: A didactic. Journal of Educational
and Behavioral Statistics, 34, 115–130.
de la Torre, J., & Douglas, J. A. (2004). Higher-order latent trait models for cognitive diagnosis.
Psychometrika, 69, 333–353.
Ekici, C., & Alagoz, C. (2018, Feb). Teachers’ validity arguments for mathematical modeling com-
petency measures and connections. Presented at the 2018 Annual Association of Mathematics
Teacher Educators Conference, February 8–10. Houston, TX.
Haertel, E. H. (1989). Using restricted latent class models to map the skill structure of achievement
items. Journal of Educational Measurement, 26, 333–352.
Jensen, T. H. (2007). Assessing mathematical modelling competency. In C. Haines, P. Galbraith,
W. Blum, & S. Khan (Eds.), Mathematical modelling (ICTMA 12): Education, engineering and
113–142.
National Council of Teachers of Mathematics. (2000). Principles and standards for school math-
ematics. Reston: Author.
National Governors Association Center for Best Practices & Council of Chief State School Officers
(NGACBP&CCSSO). (2010). Common core state standards for mathematics(CCSSM).
Retrieved from http://www.corestandards.org/Math/Practice/
New York Performance Standards Consortium. (2016). Math rubric for performance based assess-
ment tasks. Available at performanceassessment.org/performance/prubrics.html.
NGSS Lead States. (2013). Next generation science standards: For states, by states. Washington,
DC: The National Academies Press.
Rupp, A. A., Templin, J. L., & Henson, R. A. (2010). Diagnostic assessment: Theory, methods, and
applications. New York: Guilford Press.
Zöttl, L., Ufer, S., & Reiss, K. (2011). Assessing modelling competencies using a multidimen-
sional IRT approach. In G. Kaiser, W. Blum, R. Borromeo Ferri, & G. Stillman (Eds.), Trends
in teaching and learning of mathematical modelling (pp. 427–437). Dordrecht: Springer.
Chapter 31
Occurrences of Mathematical Modelling
Competencies in the Nationally Set
Examination for Mathematical Literacy
in South Africa
Mbulelo Bali, Cyril Julie, and Monde Mbekwa
Abstract Mathematical literacy is generally viewed as embedded in mathematical

modelling and the applications of mathematics. It is therefore not unreasonable to
expect that the development of mathematical modelling competencies should be
accorded priority in mathematical literacy. In South Africa, mathematical literacy is
examined through a nationally set examination at the end of 12 years of schooling.
Schemes of marking of five consecutive examinations for mathematical literacy
were analysed to ascertain the prioritisation of mathematical modelling competen-
cies in these examinations. It was found that more than 80% of marks were awarded
to mathematical calculations and that the formulation of a problem statement and
the validation of a model did not appear in any of the schemes of marking. The
implications of these results are discussed.
Keywords Content analysis · Critical case sampling · Mathematical literacy ·

Mathematical modelling · Assessment · Mathematical modelling competency
31.1 Background and Introduction
The South African public schooling system consists of two phases, namely, General
Education and Training (GET) and Further Education and Training (FET). The GET
phase is compulsory and constitutes the first ten years of schooling. In GET, learn-
ers follow a general curriculum consisting of languages and other school subjects
M. Bali (*) · C. Julie · M. Mbekwa

Cape Town, South Africa
e-mail: cjulie@uwc.ac.za; mmbekwa@uwc.ac.za

362 M. Bali et al.
including mathematics. On completion of the GET phase, learners go to the non-

compulsory FET phase. It is offered in schools and FET colleges. Students in the
FET phase must choose either mathematics or mathematical literacy, not both.
Mathematical literacy is intended to develop in learners’ competencies that will
allow them to make sense of, participate in and contribute to the twenty-first-century
world characterised by numbers, numerically based arguments and data represented
and misrepresented in a number of ways. Furthermore, it is hoped that by doing
Mathematical literacy learners will develop a critical competence that will enable
them to become critical citizens who participate meaningfully in social, economic
and political issues (Department of Basic Education 2011). Such competencies
should be developed through teaching and learning characterised by the provision
of opportunities to analyse real-life authentic problems and devise ways to work
mathematically to solve such problems. Similar arguments for the inclusion of a
subject that will enable learners to use mathematical knowledge to deal with real-
life issues have been made worldwide (e.g., Blum and Niss 1991; Blum and
Borromeo Ferri 2009; Jablonka 2003; Julie 2007; OECD 2003; De Corte et al. 2000
and many others). Although there are differences in different mathematics curricula
for different countries and amongst mathematics educationists regarding the name
of this subject, they all agree that mathematical modelling is central to the develop-
ment of mathematical literacy. In essence, as Cai et al. (2016) point out, the
Programme for International Student Assessment (PISA) process for the assessment
of mathematical literacy is similar to the mathematical modelling process. It is
against this background that the following research question has been investigated:
What mathematical modelling competencies are afforded priority in the national
examinations for grade 12 in Mathematical literacy in South Africa?
National examinations were chosen because they are nationally set, moderated
and administered and therefore are a reliable source of data about the intended cur-
riculum. This chapter overviews the processes undertaken to answer the above
research question.
31.2.1 athematical Literacy, Mathematical Modelling

M
and Mathematical Modelling Competency
The theoretical framework used in this research is drawn from both the research and
theoretical literature on mathematical literacy, mathematical modelling and mathe-
matical modelling competency. The Organization for Economic Cooperation and
Development (OECD) (2003) defines mathematical literacy as an individual’s capac-
ity to identify and understand the role mathematics plays in the world, to make well-
informed judgements and to use mathematics in ways that meet the needs of that
individual as a constructive, concerned and reflective citizen. Member countries of
31 Occurrences of Mathematical Modelling Competencies in the Nationally Set… 363
the OECD participate in PISA which, amongst other competencies, assesses stu-
dents’ mathematical literacy levels. Because of the substantial number of countries
taking part in the PISA, it is reasonable to expect that the OECD/PISA view of math-
ematical literacy is the view of the member countries. Different countries have differ-
ent names for mathematical literacy; for example, the name Numeracy is used in
many countries, Quantitative Literacy in the United States of America (USA) and
Functional Mathematics in England. In all these, the purpose of the subject is the
same as that encapsulated in the OECD/PISA definition of mathematical literacy. The
South African definition is no different to PISA’s except in wording. According to the
South African Department of Basic Education (DBE) (2011), mathematical literacy
is defined as the subject intended to allow individuals to use elementary mathematical
concepts and skills to make sense of numerically and statistically based scenarios in
their everyday lives, the workplace, and to participate as critical citizens in social and
political discussions (DBE 2011, p. 8). Although there are minor differences amongst
protagonists of mathematical literacy in terms of the subject name and purpose, they
all agree that mathematical modelling competency has a direct relationship with
mathematical literacy. As such, PISA 2012/2015 sees the process of mathematical
literacy development as the same as that of mathematical modelling competency
development (Cai et al. 2016). But then what is mathematical modelling? It is vari-
ously defined (see, e.g. Blum and Niss (1991), Blum and Borromeo Ferri (2009),
Kang and Noh (2012)). The definition adopted in this study for mathematical model-
ling is that given by Blum and Borromeo Ferri (2009). They define mathematical
modelling as a cyclical process of translating between the real world and mathemat-
ics in both directions. In essence, the process involves creating and modifying models
of empirical situations to understand them better and make informed decisions.
Figure 31.1 is a diagrammatic representation of the mathematical modelling process.
The modelling cycle of Blum and Borromeo Ferri (2009) specifies knowledge
and skills required to build a mathematical modelling competency. The verbs that
describe the steps (1–6) spell out clearly what actions are necessary to execute each
step, while the nouns (represented by circles and squares) give the end product for
each step. Mathematical modelling competency may be regarded as a compound
Fig. 31.1 Mathematical modelling cycle from Blum and Borromeo Ferri (2009, p. 46)
364 M. Bali et al.
Table 31.1 Arguments for inclusion of mathematical modelling in classroom mathematics

teaching and Mathematical literacy in South African school curriculum
Arguments for mathematical modelling in Arguments for Mathematical literacy in the
mathematics classrooms (Blum and Niss 1991) South African school curriculum (DBE
2011)
Develop general personal competencies Use mathematics to develop self-managing
orientated towards fostering explorative, creative competencies (self-managing person)
and problem-solving capacities
Develop critical competence for acting with Develop individuals that can participate as
integrity as private and social citizens critical citizens in social and political
discussions
Utilise mathematics for solving problems or Use mathematics to make sense of
describing aspects of specific extra-mathematical numerically and statistically based scenarios
areas or situations in other subjects, occupational, in the workplace (contributing worker)
actual or future everyday life contexts
Present to students a good picture of mathematics
as a science and an activity in society and culture
competency made up of various modelling sub-competencies. These sub-

competencies are precisely the activities involved in steps 1–6 of the modelling
cycle in Fig. 31.1 (see Maaβ (2006)). This modelling cycle was used to develop a
framework for the identification of competencies assessed in the mathematical lit-
eracy nationally set examinations in South Africa. The relationship between argu-
ments for incorporating mathematical modelling and applications of mathematics in
the everyday teaching of classroom mathematics provided by Blum and Niss (1991)
and the reasons for introducing the subject Mathematical literacy in the South
African school’s curriculum are compared in Table 31.1. This also provides a strong
case for the research question stated above.
31.2.2 Assessment of Mathematical Modelling Competency
Niss et al. (2007) define mathematical modelling competency as the ability to per-
form successfully all the steps in the modelling cycle and critique existing models.
Similarly, Maaβ (2006) defines mathematical modelling competencies as follows:
“Modelling competencies include skills and abilities to perform modelling pro-
cesses appropriately and goal-oriented as well as the willingness to put these into
action” (p. 117). In this chapter, mathematical modelling competency is viewed in
the sense of Niss et al. (2007).
Assessment of mathematical modelling competency is one major challenge fac-
ing educationists and researchers today. The main reason for this is the existence of
different perspectives on mathematical modelling in school mathematics. These
perspectives result in some of the mathematical modelling processes being given
more priority than others (see Borromeo Ferri (2013), Fredj (2013) and others).
In their analysis of modelling problems useful for the classroom, Kang and Noh
(2012) identify three types of modelling problems. Their classification of modelling
problems is based on ambiguity and completeness of information provided in the
problem statement. Level 1 problems are clearly stated with all the information
required to formulate a model provided. Kang and Noh (2012) explain the expected
solution process for such problems as follows: “Students are expected to search for the
needed information that is hidden in the problem, recall the (implicitly or explicitly)
called for procedure, and carry it out correctly” (p. 7). Level 2 problems have some
ambiguity about what needs to be done to solve them and often do not contain all the
information needed to complete the task. Students need to devise meaningful ways to
collect relevant data and produce reasonable answers. For level 3, the questions con-
tain open-ended information which is often incomplete and/or redundant. Students
must first analyse the task to determine what needs to be done and suggest possible
solution strategies and carry out these strategies. It is important to note that all these
categories of modelling problems require execution of the entire modelling process.
Regarding assessment tools, rubrics have been favoured over marking schema
during assessment of mathematical modelling competencies (see, e.g. Anhalt and
Cortez (2015), Chan et al. (2012), Leong (2012)). Such rubrics include all the steps
of mathematical modelling as criteria for scoring. Descriptors for mark allocation
are also included to indicate expected levels of performance.
31.3 Methodology
31.3.1 Research Type, Sampling and Data Collection
The study uses descriptive research with its purpose described by Boudah (2011) as
“to understand and report the characteristics of a current or past situation” (p. 12).
The sampling used is critical case sampling viewed as “the researcher [choosing]
the situations or participants because of their uniqueness or how important they are
to the issue” (Boudah 2011, p. 141). The complete data set on which critical case
sampling is applied consists of grade 12 final national examinations in mathemati-
cal literacy marking schema for five consecutive examinations held from November
2014 to June 2016. The examination consists of two papers, and these differ accord-
ing to purpose and taxonomy levels of questions from level 1 (knowledge) to level
4 (reasoning and reflection) (see Table 31.2). Table 31.2 shows the differences
between the two examination papers as stated in the national curriculum.
During critical case sampling, paper 1 examinations were excluded as these did
not have questions involving reasoning and reflection, and therefore, one cannot
expect assessment of a reasonable number of modelling sub-competencies in this
paper. Furthermore, not all topics prescribed in the syllabus were analysed. Prescribed
topics for the subject are finance, measurement, maps, plans and other representa-
tions, data handling and probability. For manageability reasons, only level 4 ques-
tions on the topic of measurement were analysed. One reason for this choice was that
366 M. Bali et al.
Table 31.2 A comparison of purpose and composition between papers 1 and 2 for Mathematical
literacy examinations
Paper Purpose Composition
1 Assess basic mathematical skills and 60% knowledge (level 1)
competency 35% routine procedures (level 2)
5% multistep procedures (level 3)
2 Use mathematical and non-mathematical 25% routine procedures (level 2)
techniques to solve context-based problems 35% multistep procedures (level 3)
40% reasoning and reflection (level 4)
Table 31.3 Content analysis for mathematical modelling sub-competencies assessed in the
national examinations for Mathematical literacy
Modelling sub-competencies Categories Coding rules
Production of a problem 1. Write down the ‘givens’ and the ‘required Satisfy both 1
statement from a messy real- to find’ and 2
world situation 2. Make assumptions OR
3. Simplify the problem by writing Satisfies 3
appropriate relationships. only
Mathematical model from 1. Identification of variables for inclusion in Satisfy 1, 2
real-world problem statement the model and 3
2. Uniquely defining variables 3 only
3. Represent relationships between variables acceptable
Mathematical solution from the 1. Applying formulae (model) Satisfy 1, 2
mathematical model 2. Algebraic simplifications and 3
3. Obtain mathematical results
Interpret mathematical solution in 1. Match mathematical results with situation Satisfy both 1
terms of the real world 2. Interpret mathematical results in terms of and 2
real situation
Accept solution, validate or 1. Accept final solution Satisfies 1 or
critique and revise model 2. Reconcile unexpected mismatches 2 or 3
between mathematical results and real-world
expectations
3. Revise the model
the topic sits in authentic contexts and contains many established mathematical mod-
els in the form of formulae for perimeter, area, volume, etcetera. Hence, there is a
high possibility to assess modelling competencies in level 4 questions in this topic.
31.3.2 Data Analysis Methods
The data analysis strategy used was qualitative content analysis. Hsieh and Shannon
(2005) define qualitative content analysis as a method of analysis for the subjective
interpretation of the content of text data through the systematic classification pro-
cess of coding and identifying themes or patterns. This implies the creation of cat-
egories derived from the underlying theoretical framework. The content analysis
framework for this research is given in Table 31.3. It is based on the modelling cycle
of Blum and Borromeo Ferri (2009) and the framework used by Galbraith and
Stillman (2006) to identify learner blockages in carrying out the modelling process.
Table 31.3 shows the coding rules used for the categories of modelling
sub-competencies.
Brink (1993) defines reliability as the extent to which a research method can
yield the same results consistently over different periods of testing. Since data ana-
lysed in this research is qualitative, reliability of the data analysis (i.e. content anal-
ysis framework) was addressed through inter-rater reliability or agreement. The
framework was given to experts in the subject (Subject Advisors), and the agree-
ment on coding rules application on the same data item between any two experts
was 97.76%. This also addressed trustworthiness and hence the credibility of the
data analysis instrument. The question paper and marking schema production fol-
low a strictly monitored process with internal and external moderators quality assur-
ing the final product. Hence, the national examination is a credible source of data.
31.4 Results
Table 31.4 shows the classification of knowledge and skills awarded marks in the
national marking schemas. Such knowledge and skills were classified into mathe-
matical modelling sub-competencies using content analysis.
The results of the analysis are summarised in Table 31.5. The first column con-
tains the five main modelling sub-competencies. In the second column are the
Table 31.4 Classification of knowledge and skills awarded marks in Mathematical literacy
examinations
Specific knowledge/skills (according to the
Modelling sub-competency marking schema)
Production of a problem statement from a Problem simplification or rephrasing
messy real-world situation
Mathematical model from real-world problem Recall a known mathematical formula (not
statement given)
Express a worded relation mathematically
Determine the solution path or method
Mathematical solution from the mathematical Use correct formula (given)
model Substitution of values into formulae
Simplification (BODMAS)
Unit conversion
Use ratio and proportion to determine unknown
quantities
Interpret mathematical solution in terms of the Round numeric answers off or up or down
real world according to contexts
Express the calculated answer in terms of the
given context
Accept solution or validate or critique and Make conclusions based on calculated answers
revise model
368 M. Bali et al.
Table 31.5 Summary of the results

Examinationa
Nov Mar Nov Mar Jun
2014 2015 2015 2016 2016 Percentage of
Modelling sub-competency [11] [13] [26] [15] [15] total mark (%)
Production of a problem statement 0 0 0 0 0 0
from a messy real-world situation
Mathematical model from 0 1 2 2 2 8.75
real-world problem statement
Mathematical solution from the 10 12 22 12 13 86.25
mathematical model
Interpret mathematical solution in 1 0 2 1 0 5
terms of the real world
Accept solution or validate or 0 0 0 0 0 0
critique and revise model
a
Note: https://www.education.gov.za/Curriculum/NationalSeniorCertificate(NSC)Examinations.
aspx
examination papers (paper 2) analysed, and the number in square brackets indicates
the total mark for level 4 questions in the paper on the topic of measurement. Values
in the columns for each paper analysed indicate the total marks awarded for each
sub-competency. The last column gives a percentage of marks allocated to each sub-
competency for all the papers analysed. For example, for November 2014 examina-
tion, 11 marks were awarded for level 4 questions on the topic of measurement.
Zero out of eleven marks was awarded for sub-competencies: Production of a prob-
lem statement, Mathematical model from real-world problem statement and Critique
and revise model. Ten marks were awarded to sub-competency: Mathematical solu-
tion from the mathematical model and one mark to Interpret mathematical solution
in terms of the real world.
The analysis of the results reveals firstly that for the five examination papers anal-
ysed, 86.25% of the total marks on the topic of measurement were awarded to math-
ematical calculations, techniques and skills. Thus, the majority of marks were
awarded to competence in producing a Mathematical solution from the mathemati-
cal model. According to the schema of marking, these mathematical solutions are
easily performed with calculators, which examinees are allowed to use in examina-
tion, but there might be some conversion of measurement units to be done. Secondly,
the competence associated with the formulation of a mathematical model only com-
prised 8.75% of the total marks. In most cases, the mathematical model was pro-
vided in the form of formulae which candidates were instructed to use. There were,
however, some, albeit elementary, adaptations to be made to the supplied model like
changing the subject of the formula or, simply put, the quantity to be calculated was
not the subject of the formula. The skill critiquing existing models was also not
assessed. This is antithetical to the intended curriculum’s explicit outcome of math-
ematical literacy as the development of critical citizenship. This can be ascribed to
the topic, measurement, focusing more on the utilitarian ideal (Blum and Niss 1991)
of work-related applications. It might be the case this competency is deemed more
appropriate for other topics, such as financial matters, dealt with in the curriculum
and thus needs further investigation.
Overall, the results reveal that there is some misalignment between competen-
cies the subject mathematical literacy aims for and the actual competencies priori-
tised in examinations. All analysed questions do not meet even the level 1 criteria
for modelling problems according to the Kang and Noh (2012) classification.
National examinations provide a reliable form of guidance to both teachers and
textbook writers regarding key knowledge and skills of the intended curriculum.
Therefore, there should be a close relationship between competencies envisaged for
learners in the national curriculum and the actual competencies assessed in the
national tests.
Acknowledgement This research is supported by the National Research Foundation under grant
number 77941. Any opinions, findings and conclusions or recommendations expressed in this
material are those of the authors and do not necessarily reflect the views of the National Research
Foundation (NRF) of South Africa.
References
Anhalt, C. O., & Cortez, R. (2015). Modelling: A structured process. Mathematics Teacher, 108(6),
447–452.
Journal of Mathematical Modelling and Applications, 1(1), 45–58.
Blum, W., & Niss, M. (1991). Applied mathematical problem solving, modelling, applications,
and links to other subjects – State, trends and issues in mathematics instruction. Educational
Studies in Mathematics, 22, 37–68.
Boudah, D. (2011). Conducting educational research: Guide to completing a major project.
Brink, H. I. L. (1993). Validity and reliability in qualitative research. Curationis, 16(2), 35–38.
Cai, J., Mok, I. A. C., Reddy, V., & Stacey, K. (2016). International comparative studies in math-
ematics: Lessons for improving student learning. In J. Cai, I. A. C. Mok, V. Reddy, & K. Stacey
(Eds.), International comparative studies in mathematics. ICME-13 topical surveys (pp. 1–35).
Cham: Springer Open. https://doi.org/10.1007/978–3–319–42414–9_1.
Chan, E. C. M., Ng, D., Widjaja, W., & Seto, C. (2012). Assessment of primary 5 student’s math-
ematical modelling competencies. Journal of Science and Mathematics Education in Southeast
Asia, 35(2), 146–178.
De Corte, E., Verschaffel, L., & Greer, B. (2000). Connecting mathematics problem solving to
the real world. Proceedings of the International conference on mathematics education into
the 21st century: Mathematics for living (pp. 66–73). Amman: The National Center for Human
Resource Development.
Department of Basic Education. (2011). Curriculum and Assessment Policy Statement (CAPS)
Grades 10–12: Mathematical Literacy. Pretoria: Department of Basic Education.
370 M. Bali et al.
Fredj, P. (2013). Modes of modelling assessment – a literature review. Educational Studies in

Mathematics, 84, 413–438.
Hsieh, H., & Shannon, S. E. (2005). Three approaches to qualitative content analysis. Qualitative
Jablonka, E. (2003). Mathematical literacy. In A. J. Bishop, M. A. Clements, C. Keitel,
J. Kilpatrick, & F. K. S. Leung (Eds.), Second international handbook of mathematics educa-
tion (pp. 75–102). Dordrecht: Kluwer Academic.
Julie, C. (2007). Critique of mathematical models and applications: A necessary component of
mathematical literacy. (Unpublished class notes). Bellville: University of the Western Cape.
Kang, O., & Noh, J. (2012, 8–15 July). Teaching mathematical modelling in school mathematics.
Regular lecture presented at 12th International Congress on Mathematics Education. Seoul,
Korea. Available at: docplayer.net/18511636-Teaching-mathematical-modeling-in-school-
mathematics.html
Leong, R. K. E. (2012). Assessment of mathematical modelling. Journal of Mathematics Education
at Teacher College, 3(Spring-Summer), 61–65.
Maaβ, K. (2006). What are modelling competencies? ZDM Mathematics Education, 38(2),
113–142.
Niss, M., Blum, W., & Galbraith, P. (2007). Introduction. In W. Blum, P. L. Galbraith, H. Henn,
& M. Niss (Eds.), Modelling and applications in mathematics education: 14th ICMI Study
Organisation for Economic Co-operation and Development (OECD). (2003). The PISA 2003
assessment framework – Mathematics, reading, science and problem solving knowledge and
skills. Available at: http://www.oecd.org/education/school/programmeforinternationalstuden-
tassessmentpisa/pisa2003assessmentframeworkmathematicsreadingscienceandproblemsolv-
ingknowledgeandskills-publications2003
Chapter 32
Expectations for Challenge in Modelling
and Its Assessment
Jill P. Brown
Abstract This chapter reports a document analysis of key curriculum documents in

two states of Australia seeking to answer the research question: What level of chal-
lenge is expected in modelling tasks in the intended curriculum? Both states,
Queensland and Victoria, include mathematical modelling tasks as part of the
assessment in the final year of schooling. Analysis found that information provided
to teachers in Queensland was broader and more substantive including with respect
to sample assessment instruments and how these would be implemented and
assessed. A framework of complexity used to analyse two specific sample tasks
found the indicative levels of expectations, in this case with respect to degree of
challenge, were lower in Victoria than in Queensland.
Keywords Assessment · Challenge · Challenging tasks · Complexity · Degree of

contextualisation · Degree of openness · Productive struggle
32.1 hallenge: A Critical Component for Learning

C
and Assessment
In the volume arising from the 16th ICMI Study, Challenging Mathematics In and
Beyond the Classroom, Barbeau (2009) describes challenge as a human activity
(p. 1). Furthermore, in arguing for the essential nature of challenge in learning
mathematics, he claims that facing challenging mathematical tasks allows learners
to be patient, open-minded, resilient and draw on internal resources. More
J. P. Brown (*)
Mathematics Education|Faculty of Arts and Education, Deakin University,
Melbourne, VIC, Australia
e-mail: jill.brown@deakin.edu.au

372 J. P. Brown
specifically, learning “growth is promoted through the presence of an appropriate

challenge” (p. 3). Modelling tasks provide opportunities for all students to face such
challenges according to Shimizu and Williams (2013).
Not surprisingly, challenge is frequently used in the mathematics education lit-
erature without definition, as if it is well understood and has no variation in the way
it is used. Barbeau defines “challenge as a question posed deliberately to entice its
recipient to attempt a resolution, while at the same time stretching their understand-
ing and knowledge of some topic” (2009, p. 5). He describes a good challenge as
when the task solver has the.
necessary mathematical apparatus or logical skill, but needs to use them in a nonstandard or
innovative way [and]…. involve[s] explanation, questioning and conjecturing, multiple
approaches, evaluation of solutions for effectiveness and elegance, and construction and
evaluation of examples. (p. 5)
Engaging with challenging or cognitively demanding tasks is an important part of

mathematical learning, involves student struggle and is a natural component of
mathematical modelling, problem-solving and applications. Both challenge and the
resulting struggle are essential elements of learning to model and assessment of
modelling. Questions that arise for a mathematics education researcher and educa-
tor preparing pre-service teachers for secondary mathematics include the following:
Are upper secondary mathematics students engaging with challenging tasks? Is this
intended? The specific research question addressed in this chapter is: What level of
challenge is expected in modelling tasks in the intended curriculum?
32.2 Background
The importance of challenge, or challenging problems, in the learning of mathemat-

ics seems well acknowledged in the educational research community. Within
Australasia, for example, English (2016) notes that mathematics, and the capability
to engage with challenging problems that require mathematics to be solved, under-
pins much of the science, technology and engineering disciplines in STEM educa-
tion in schooling and beyond (English 2016). By challenging problems, English is
particularly referring to those involving problem-solving and mathematical
modelling.
Internationally, the work of Smith and Stein (2011) evaluating the level of cogni-
tive demand of tasks as designed and implemented is well known. Along with
English (2016), they describe challenging tasks similarly and expand on the rela-
tionship to the cognitive demand of the task. In this chapter, challenging tasks
include problem-solving, mathematical modelling where the focus is on solving a
real-world problem and investigative tasks and involve what is described by Smith
and Stein as “doing mathematics” (as distinct from memorisation, procedures with-
out connections and procedures with connections) and are of high cognitive demand.
32 Expectations for Challenge in Modelling and Its Assessment 373
Of critical concern, however, are findings from substantive research by Smith and
Stein (2011), and more recently Russo and Hopkins (2017), that teachers are reluc-
tant to pose challenging tasks or, when they do, the implementation reduces the
intended level of challenge.
Of course, engaging with the challenge of a mathematical task and struggle
coexist. One necessarily presupposes the other. Hiebert and Grouws (2007) describe
productive struggle as occurring when “students expend effort in order to make
sense of mathematics, to figure something out that is not immediately apparent”
(p. 387) and argue that struggling to make sense of mathematics is a necessary com-
ponent of learning mathematics with understanding. In the same way that persis-
tence and perseverance are distinguished, with the latter occurring when one’s
efforts can result in progress in problem-solving (versus continuing to try but on an
unproductive path), the construct productive struggle is used to describe actions,
cognitive and metacognitive, during a task solver’s struggle to make sense of a prob-
lem. Struggle is seen as an opportunity for learning or progress toward task solution.
It is recognised as one of eight important teaching practices, that is, to “support
productive struggle in learning mathematics” (NCTM 2014, p. 10).
A recent study by Warshauer (2015) investigated the role of productive struggle
in middle school mathematics in a study with six teachers and 317 grade 6 and 7
students. She researched how teachers can maintain the (intended) cognitive demand
by addressing the struggle and building on student thinking (p. 381). Reinforcing
the notion that challenge and struggle coexist, Warshauer found that “posing prob-
lems of high cognitive demand gave the students opportunities to think, reason and
problem-solve in ways that meant the students often had to struggle about the math-
ematics” (p. 396).
However, there is still limited advice in many curricula, including within the
Australian Curriculum Mathematics F-10, regarding challenge and productive
struggle. “Teachers [are encouraged] to help students become self-motivated, confi-
dent learners through inquiry and active participation in challenging and engaging
experiences” (ACARA 2015, online), but little advice about the types of activities
or pedagogies that support students learning how to solve challenging mathematical
problems is provided. At the upper secondary level, different states still have their
own curriculum documents, and the statutory assessment and reporting bodies in at
least two Australian states, Queensland and Victoria, provide sample tasks and cur-
riculum documents for senior secondary curricula related to the assessment of
mathematical modelling, problem-solving and applications. In these states, in the
first, modelling has been part of the learning and assessment regime for several
decades, and in the second, modelling has been optional for some time, but from
2016, all students must do at least two ‘modelling or problem-solving tasks’ as part
of their formal assessment.
374 J. P. Brown
32.3 Methodology
This study is attempting to determine the level of challenge in tasks expected as

evident in curriculum documents – in the final year of schooling – by the statutory
authorities in two Australian states, Queensland and Victoria. In both states, model-
ling tasks form part of the assessment regime. The methodological approach of
document analysis, that is, a systematic approach for evaluating documents, involves
examining and interpreting data to elicit meaning and gain understanding (Corbin
and Strauss 2008). The intention is to provide rich descriptions of a single phenom-
enon, namely, the level of challenge. Further, McNaught and Lam (2010) argue
word clouds provide a rich visualisation of text providing a useful initial analysis.
Watson (2017) undertook a document analysis of the mathematics and science
curricula in three countries (Australia, NZ, and USA), searching to ascertain where
science could “provide meaningful contexts within which to set statistical investiga-
tions” (p. 1057). She used several frameworks, including a model for statistical
investigation to analyse the curricula documents, and argued these documents are
“responsible for covering the required detail for each stage of a statistical investiga-
tion as well as its overall structure” (p. 1059). By inference, curriculum documents
that require mathematical modelling and/or problem-solving should ensure these
types of mathematical activity are clearly defined as to the process such activity
involves and the depth of treatment. The documents analysed, for this study, were
those related to the Queensland Senior Mathematics syllabuses for Year 11 and 12,
and in Victoria, the Victorian Certificate of Education (VCE) mathematics
study design.
Following Stillman (2002), tasks were classified using two dimensions from
Blum and Kaiser’s (1984) “categorising types of applications and conceptions for
an application – oriented mathematics instruction” (p. 399), namely, (i) nature of
connection to real-world and extra mathematics fields and (ii) function for the appli-
cation. An additional two dimensions from Stillman (2002), (iii) degree of contex-
tualisation and (iv) degree of open-endedness, completed the initial categorisation.
Stillman’s (2002, p. 56) synthesis of the Blum and Kaiser’s dimensions and catego-
ries describe categories for dimension (ii) as mathematics contributes to solution of
real problem (analytical model); mathematics contributes to comprehension of real
situation (descriptive model); develops translation skill; develops meta-knowledge;
as an illustration of mathematical concepts; motivating mathematical concepts;
organising mathematical areas; or promoting general goals. For dimension (i) the
categories are close (real assumptions and real problem-setting); remote (real
assumptions and unreal problem-setting or unreal assumptions and real problem-
setting); or unreal (unreal assumptions and unreal problem-setting).
Stillman (2002) proposed the dimension, degree of contextualisation, “to convey
the range of embedding that occurs between the mathematics that can be used to
model a situation and the description of the situation” (p. 70). She described three
metaphors to denote the level of embeddedness. These are Border problems where
the context and mathematics are completely separate and do not interact in any way;
Wrapper problems with the mathematics hidden inside the context, but once found,
the task context can be thrown away, although “more diligent problem-solvers make
pick up the wrapper…to check [their solution] makes sense” (p. 77); and Tapestry
problems where the task context and mathematics are entwined, “the solution pro-
cess proceeds by the problem solvers continually returning to the task context to
check they are on track” (p. 77). In addition, she presents degree of openness as a
continuum from closed to open-ended. In a closed task, the direction of task solu-
tion is specified by the task setter, whereas a task is more open-ended where some
or all direction required to solve the problem comes from the task solver.
Stillman used these four categories to classify tasks used by teachers in two
Queensland schools when the change from abstract mathematics syllabuses to mod-
elling- and applications-focused syllabuses became statewide in the 1990s. These
tasks were designed or chosen by the teachers for implementation as part of the
extended time unsupervised assessment tasks. Stillman found “the degree of con-
textualisation [to be] of significant distinction” (2002, p. 51) for extended time tasks
designed or selected by the teacher.
32.4 Analysis of Curriculum and Assessment Documents
In Australia, the states have jurisdiction over education. Hence, different states have
different systems. The Queensland Studies Authority [QSA] syllabi (QSA 2014a, b)
provide detailed advice for each mathematics subject including Mathematics A, B
and C. In contrast, in Victoria, the Victorian Curriculum and Assessment Authority
[VCAA] study designs specify “the content for the study and how the students’
work is to be assessed. Schools … must adhere to the requirements in the study
designs” (online). The VCE mathematics study design explicitly states it aims to
“provide access to worthwhile and challenging mathematical learning” (VCAA
2015, p. 6).
The Queensland syllabuses (one per subject, e.g. Mathematics A, B and C)
include topics (e.g. focus, subject matter, suggested learning experiences), assess-
ment techniques, criteria and standards and learning experiences. The VCE mathe-
matics study design includes areas of study, content dot points and outcomes
mapped to key knowledge and key skills. Sample assessment tasks are available on
the respective websites. In Queensland, those with a focus on extended modelling
and problem-solving include the suggested task, student samples and associated
task-specific criteria and standards. The sample assessment instrument, and student
responses, for the cooling of coffee, the Coffee Task, is 32 pages. In Victoria, assess-
ment advice (online) includes teaching and learning activities, sample tasks includ-
ing for mathematical modelling or problem-solving (approx. 3 pp.) and performance
criteria. At the end of 2016 after the new assessment practices including mathemati-
cal modelling were introduced, implementation reports were published.
As an initial analysis, a word cloud (http://www.wordle.net) mapping two key
documents, the Queensland Mathematics C syllabus (2014) and all parts of the VCE
376 J. P. Brown
Fig. 32.1 Word clouds comparing sample key curriculum documents: (a) Queensland and (b)
Victoria
mathematics study design (2015) related to Specialist Mathematics, was produced

(Fig. 32.1). Word clouds provide an indication of the frequency of words and hence
some indication of the emphasis of the documents. These graphics appear to indi-
cate a differing emphasis with most frequent words in Queensland being students,
mathematical, assessment, learning and investigate, whereas in Victoria, the most
frequent words included study, unit, functions, mathematical and technology. The
emphasis in Queensland appears to be on learning mathematics at a global level,
whereas in Victoria, it is more focused on ‘pure content’.
In Queensland, teachers are advised that assessment instruments are developed
by the school to provide “information on which teachers may make judgments
about student achievement of the general objectives; [and] a level of challenge suit-
able for the whole range of students” (e.g. QSA 2014b, p. 39). There are ten global
aims and general objectives organised into four categories: knowledge and proce-
dures, modelling and problem-solving, communication and justification and affec-
tive. The first three categories are directly linked to three exit criteria, respectively.
Assessment is 100% school-based assessment (SBA) with moderation (although
this is to change from 2019).
In Victoria, the study design includes Further Mathematics, Mathematical
Methods and Specialist Mathematics at the Year 12 level. The aims of this study are
to enable students to (1) develop mathematical concepts, knowledge and skills; (2)
apply mathematics to analyse, investigate and model a variety of contexts and solve
practical and theoretical problems in situations that range from well-defined and
familiar to open-ended and unfamiliar; and (3) use technology effectively as a tool
for working mathematically. These aims link directly to Outcomes 1, 2 and 3 that
students are required to demonstrate. Assessment is 34% SBA, and 66% externally
assessed (statewide examinations). Aim 3 is interlinked with aims 1 and 2.
In both jurisdictions, modelling tasks are part of the assessment regime. In
Queensland, modelling and problem-solving are to be used at least twice a year in
all subjects. In Victoria, modelling or problem-solving tasks are included, three per
year in Further Mathematics and two per year in Mathematical Methods and
Specialist Mathematics. Technology use is an integral part of teaching, learning and
assessment in both states.
In Queensland, one exit criterion relates specifically to modelling and problem-

solving. This “refers to the student’s ability to apply problem-solving strategies to
investigate and model situations, generate and use data and analyse and interpret
results in the context of problems to investigate the validity of mathematical argu-
ments and models” (QSA 2014a, p. 33). There are published (descriptive) Standards
A–E associated with the exit criteria.
In Victoria, Outcome 2 (see aim 2, this section) varies slightly for each specific
‘subject’. For Mathematical Methods, it is described as follows: “apply mathemati-
cal processes in non-routine contexts, including situations requiring problem-
solving, modelling or investigative techniques or approaches, and analyse and
discuss these applications of mathematics” (VCAA 2015, p. 76). For Specialist
Mathematics, the outcome is described as “Apply mathematical processes, with an
emphasis on general cases, in non-routine contexts, and analyse and discuss these
applications of mathematics” (p. 88).
32.5 A More Detailed Analysis of Illustrative Sample Tasks
The categories (see Sect. 32.3) were used to classify tasks offered by the QCA and
VCAA as modelling and/or problem-solving tasks. Analysis of the tasks showed
clear differences between the sample tasks with respect to the level of challenge
expected in the two states. For space reasons, two representative illustrative tasks
are used, from Queensland The Coffee Task and from Victoria The Geese Task.
The Coffee Task presents Newton’s Law of Cooling, and the task is to model the
relationship between the differences in the time of cooling during the cooling of a
liquid (coffee) in an open cup composed of a variety of different materials. The
instrument-specific criteria and standards show that for the criteria, modelling and
problem-solving, at Standard A, students are required to demonstrate the following:
use of problem-solving strategies to interpret, clarify and analyse problems to
develop responses from routine simple tasks to nonroutine complex tasks in life-
related and abstract situations; identification of assumptions and their associated
effects, parameters and/or variables; and use of data to synthesise mathematical
models and generation of data from mathematical models in simple through com-
plex situations. In addition, two sample solutions are provided: one meeting the
criteria at Standard A and one at Standard C.
The Geese Task is a ‘modelling or problem-solving task’ according to the VCAA
that involves modelling a population of geese using differential equations. There are
three parts each with subparts (3 pp.). The assessment criteria related to Outcome 2
are identification of important information, variables and constraints; application of
mathematical ideas and content from the specified areas of study; and analysis and
interpretation of results.
In category (i) nature of connection to real world and extra-mathematical fields,
the Coffee Task was classified as immediate use of mathematics in everyday adoles-
cent life, whereas the Geese Task was categorised as use of mathematics in
378 J. P. Brown
situations in the wider environment; however, the context transcended reality. For
category (ii) function for the application, the Coffee Task was classified as mathe-
matics that contributes to comprehension of a real situation (descriptive model),
whereas the Geese Task was classified as illustration of mathematical concepts. For
(iii) degree of contextualisation, the Coffee Task was classified as a wrapper task as
students must engage with the context initially. In addition, they could use the real-
world context in interpreting the task. The Geese Task was classified as a border task
with no need to consider the context in solving the problem except possibly to con-
sider a suitable domain for any function used in task solution. For (iv) degree of
open-endedness, the Coffee Task tended toward the open-ended end of the contin-
uum, whereas the Geese Task was closed.
Although the analysis presented so far shows clear differences in the tasks and
expectations of challenge, it was felt the analysis was not sufficient to clearly dif-
ferentiate between the levels of challenge expected. Using grounded theory, Stillman
(2002) developed a Framework of Categories of Task Complexity based on empiri-
cal data (see also Stillman and Galbraith (2003)). The major categories are concep-
tual complexity, mathematical complexity, linguistic complexity, intellectual
complexity, representational complexity and contextual complexity. The attributes
that may contribute to each category are many, and each is accompanied by a dimen-
sional range. In 2004, Stillman, Edwards and Brown extended the framework to
include the category complexity of technology use as illustrated in Table 32.1, giving
an indication of the depth of analysis required in each category.
Using the framework to analyse the tasks, it was found that for almost every
attribute within each category of complexity, the Coffee Task [C] was at the same
Table 32.1 The category complexity of technology use

Dimensional ranges (simple ….
Attribute (level of complexity) complex)
How many electronic technologies are involved? 1…many
How are these technologies used? Analysis tool, real-world interface
How much technological knowledge is required? Little…a lot
How easy is the technology to use? Easy…very difficult
How obscure is the choice of techniques? Fairly obvious…quite obscure
How complex is each technique? Quite simple…quite complex
How complex is the combination of techniques? All quite simple…most quite complex
How visible are the links between techniques? Fairly apparent…quite obscure
How many steps are involved? 1…many
How many features of the technology are involved? 1…many
What amount of guidance is given? High …. none
How much decision-making is necessary? None…a lot
How many representations can the technology 1… many
provide?
Table 32.2 Mapping the Coffee Task and the Geese Task to complexity of technology use
Attribute (level of complexity) Dimensional ranges (simple …. complex)
How many electronic technologies are 1[G] … [C] … many
involved?
How are these technologies used? Analysis tool [G, C], real-world interface [C]
How much technological knowledge is Little…[G] [C] … a lot
required?
How easy is the technology to use? Easy [G]… [C] …very difficult
How obscure is the choice of techniques? Fairly obvious[G]… [C] quite obscure
How complex is each technique? Quite simple… [G, C] … quite complex
How complex is the combination of techniques? All quite simple [G, C] …most quite complex
How visible are the links between techniques? Fairly apparent [G, C] …quite obscure
How many steps are involved? 1… [G, C] many
How many features of the technology are 1… [G, C] …many
involved?
What amount of guidance is given? High [G], …. None [C]
How much decision-making is necessary? None… [G] … a lot [C]
How many representations can the technology 1… [G, C] many
provide?
[G, C] or a higher level of complexity than the Geese Task [G]. Table 32.2 illustrates
this for complexity of technology use. This more fine-grained framework was found
to be particularly useful.
In Queensland, although the Coffee Task was fairly open, the expectations, previ-
ous experiences and familiarity with the criteria of the teacher and students seemed
to result in an expectation of the cognitive demand remaining high and the task
being challenging for the students. Given that sample solutions at Standards A and
C were provided, it is possible to compare the expected level of challenge of these
and to the Geese Task. Based on the extended framework from Stillman et al. (2004),
from least to most challenging, we have the Geese Task→ the Coffee Task C → the
Coffee Task A. It appears the difference may, in part, be related to student under-
standing of expectations of the criteria and standards which in some way ‘extends’
the task to a higher level. The Geese Task, perhaps at a higher level mathematically,
was explicit in the mathematical requirements (e.g. Set up a differential equation to
model the situation and find the general solution…. Use exponential regression
to…). In addition, the link to the relevant assessment criteria for this task is much
less explicit with very broad criteria (i.e. analysis and interpretation of results).
In both jurisdictions, the sample tasks could be redesigned by the teachers and
implemented in a variety of ways, with certain restrictions. In Queensland, the
teacher determines if the task is completed in/out of class. In Victoria, the expecta-
tion is the task be undertaken “mainly in class time” (VCAA 2015, p. 88), but much
anecdotal evidence suggests that not only are they implemented in most cases ‘all’
in class time but also that class time is under examination conditions. However,
analysis of the available documents provides clear evidence of mathematical
380 J. P. Brown
challenge embedded in the Geese Task, but the challenge was mitigated by the task
setter’s specific directions to the solver. In contrast, in Queensland, the open task in
conjunction with familiar criteria appeared to provide challenge for student perfor-
mance at the C standard still allowing other students to be extended further and to
engage in analysis and synthesis at the A standard of performance.
32.6 Findings and Implications
On the basis of the analysis undertaken and presented in this chapter, it appears the
level of challenge in Victoria as pertaining to mathematical modelling is low. Whilst
the sample tasks may have enticed students to solve them (Barbeau 2009), there is
little evidence in the sample tasks, particularly for Mathematical Methods and
Specialist Mathematics, of opportunities for students to demonstrate their mathe-
matical capabilities in non-standard or innovative ways (Barbeau 2009), or as Smith
and Stein (2011) describe, doing mathematics. All students should experience chal-
lenging tasks where a solution pathway is not well-rehearsed or explicit, metacogni-
tive activity is necessary and “the unpredictable nature of the solution process
required” (p. 16) and necessitates challenge and associated productive struggle. In
contrast, the significantly more open-ended nature of the tasks from Queensland did
provide such opportunities.
Curriculum writers have potential to influence the level of challenge faced by
students during mathematics tasks – especially mathematical modelling and
problem-solving tasks. This includes the intended and implemented level of chal-
lenge. These are influenced by task design, assessment processes, teacher and stu-
dent understanding of mathematical modelling and problem-solving, expectations
of learners, sample tasks, past practices, value given to ‘real world’ and ongoing
advice and feedback from curriculum authorities.
The level of challenge expected by curriculum writers, as evidenced by the avail-
able curriculum documents, supporting materials online and illustrated by the
Coffee Task and the Geese Task, has been shown to be quite different in the two
educational jurisdictions. As it is the teachers who design and implement the tasks,
a selection of actual tasks used needs to be collected and analysed. Anecdotal evi-
dence, to date, suggests in Queensland, open implementations such as the sample
Coffee Task are not uncommon. Whereas in Victoria, mathematical modelling or
problem-solving tasks appear to be ‘very structured’ and in addition implemented
‘all rather than mainly’ in class and under test conditions. These practices may con-
tribute to the use of less challenging tasks in Victoria where opportunities for
authentic modelling are very much constrained. Not only the presence of so-called
modelling tasks in high stakes assessment but also indicative levels of expectations,
in this case with respect to degree of challenge, in supporting documents such as
sample tasks, are necessary to ensure the depth of challenge required for twenty-
first-
century capabilities, as espoused by English (2016) and Shimizu and
Williams (2013).
References
ACARA. (2015). Australian curriculum: Mathematics. Retrieved http://www.australiancurricu-

lum.edu.au/
Barbeau, E. (2009). Introduction. In E. Barbeau & P. Taylor (Eds.), Challenging mathematics in
and beyond the classroom. The 16th ICMI study (pp. 1–10). New York: Springer.
Blum, W., & Kaiser, G. (1984). Analysis of applications and of conceptions for an application-
oriented mathematics instruction. In J. Berry, D. N. Burghes, I. D. Huntley, D. J. G. James,
& A. O. Moscardini (Eds.), Teaching and applying mathematical modelling (pp. 201–214).
Chichester: Ellis Horwood.
Corbin, J., & Strauss, A. (2008). Basics of qualitative research (3rd ed.). Thousand Oaks: Sage.
English, L. D. (2016). Advancing mathematics education research within a STEM environment.
In K. Makar, S. Dole, J. Visnovska, M. Goos, A. Bennison, & K. Fry (Eds.), Research in math-
ematics education in Australasia 2012–15 (pp. 353–372). Singapore: Springer.
Heibert, J., & Grouws, D. A. (2007). The effects of classroom mathematics teaching on students’
learning. In J. F. Lester Jr. (Ed.), Second handbook of research on mathematics teaching and
learning (pp. 371–404). Charlotte: Information Age.
McNaught, C., & Lam, P. (2010). Using Wordle as a supplementary research tool. The Qualitative
Report, 15(3), 630–643.
NCTM. (2014). Principles to actions: Ensuring Mathematical success for all. Reston: NCTM.
Queensland Studies Authority. (2014a). Senior syllabus: Mathematics B 2008. Brisbane: QSA.
Queensland Studies Authority. (2014b). Senior syllabus: Mathematics C 2008. Brisbane: QSA.
Russo, J., & Hopkins, S. (2017). How does lesson structure shape teacher perceptions of teaching
with challenging tasks? MTED, 19(1), 30–46.
Shimizu, Y., & Williams, W. (2013). Studying learners in mathematical contexts. In K. Clements,
A. Bishop, J. Kilpatrick, F. Leung, & C. Keitel (Eds.), Third international handbook in math-
ematics education (pp. 145–168). New York: Springer.
Smith, M., & Stein, M. K. (2011). 5 practices for orchestrating productive mathematics discus-
sions. Reston: NCTM.
Stillman, G. A. (2002). Assessing higher order mathematical thinking through applications.
Unpublished doctor of philosophy thesis. Brisbane: University of Queensland.
Stillman, G. A., & Galbraith, P. (2003). Towards constructing a measure of the complexity of appli-
cation tasks. In S. Lamon et al. (Eds.), Mathematical modelling: A way of life (pp. 179–188).
Stillman, G., Edwards, I., & Brown, J. (2004). Mediating the cognitive demand of lessons in real
world settings. In B. Tadich, S. Tobias, C. Brew, B. Beatty, & P. Sullivan (Eds.), Proceedings
of MAV41 (pp. 489–500). Melbourne: MAV.
VCAA. (2015). Victorian certificate of education Mathematics study design. Melbourne: VCAA.
Warshauer, H. K. (2015). Productive struggle in middle school mathematics classrooms. Journal
of Mathematics Teacher Education, 18(4), 375–400.
Watson, J. M. (2017). Linking science and statistics: Curriculum expectations in three countries.
International Journal of Science and Mathematics Education, 15(6), 1057–1073.
Chapter 33
Modelling Tasks in Central Examinations
Based on the Example of Austria
Hans-Stefan Siller and Gilbert Greefrath
Abstract In Austria, the written school leaving examination was centralised in

2015. This resulted from the government programme of the 24th legislation period
of November 2008. These newly designed tasks will be analysed as an example of
written central examination tasks in terms of modelling content. For modelling
tasks, a range of significant criteria are known. On this basis, evaluation criteria for
the modelling content of final examination tasks have been developed and applied
to Austrian examples. With the help of the available empirical data, connections
between the modelling content of a task and the success rate can be investigated.
Keywords Assessment · Competence · Authenticity · Modelling · Modelling

character · Openness · Relation to reality · Sub-competencies · Relevance
33.1 Introduction
General mathematical competencies are being focused on more and more in relation
to examination tasks in mathematics. Over the past years, educational standards for
the higher-education entrance qualification, as well as for designing examinations,
have been implemented in Germany (KMK 2012) as well as in Austria (BIFIE 2012).
Modelling tasks can be assigned a specific role in relation to examinations. There
have been many empirically meaningful results concerning the use of modelling
tasks in examinations in lower secondary education (Greefrath et al. 2017). Within
H.-S. Siller (*)

Institute of Mathematics, University of Würzburg, Würzburg, Germany
e-mail: siller@dmuw.de
G. Greefrath
Institute of Education in Mathematics and Computer Science, University of Münster,
Münster, Germany
e-mail: greefrath@uni-muenster.de

384 H.-S. Siller and G. Greefrath
the scope of upper secondary education, initial empirical investigations regarding

critical task orientation are available in the context of central school leaving exami-
nations (Greefrath and Vorhölter 2016).
For modelling tasks, several important criteria are known (Maaß 2010). On the
basis of these criteria, we have developed an assessment of the modelling content of
school leaving examinations (Greefrath et al. 2017) and applied it also to Austrian
final examination tasks. The reasons for applying this tool to the tasks are diverse.
The most obvious reason is exploring the instrument in addition to Greefrath et al.
(2017). Another reason for looking at these tasks is strengthening the competency-
level model, developed by Siller et al. (2016), to create and analyse examination
tasks regarding their difficulty. Using the available data, it is possible to investigate
the relationships between the modelling content and the success rate of the tasks.
The Austrian school leaving examination consists of two parts – details can be
found in Siller et al. (2016). We are focusing on the second part of this examination
and investigating the tasks of the years 2015, 2016 and 2017. These tasks are explic-
itly characterised in the examination conception of the Federal Institute for
Educational Research, Innovation and Development of the Austrian School
System – called BIFIE – based on a conception from the year 2012 (BIFIE 2012).
A crucial target of such tasks is the link of “basic skills to clearly defined contexts
and application areas”. As a result, these tasks are more extensive and also have a
clearer relation to reality.
33.2.1 Modelling Competency
According to Blum et al. (2007), we describe modelling competence as the ability

to perform the necessary process steps when switching between reality and mathe-
matics in problems adequately and to analyse or comparatively assess given models
(see Fig. 33.1). As a consequence, the diverse processes during a modelling activity
can be analysed and accentuated in different ways. The ability to perform a sub-
process such as interpreting is accepted as a sub-competency of modelling (Kaiser
2007). Our evaluation of an examination task is limited to sub-competencies of
modelling presented in Fig. 33.1 and characterised in Table 33.1. Whether under-
standing and exposing take place can often not be attributed to the tasks but rather
to the participating students.
33 Modelling Tasks in Central Examinations Based on the Example of Austria 385
Fig. 33.1 Modelling cycle. (Blum and Leiss 2007, p. 225)
Table 33.1 Selected sub-competencies of modelling (Greefrath and Vorhölter 2016, p. 19)
Sub-
competency Description
Simplifying Students identify relevant and irrelevant information from a real problem
Mathematising Students translate specific, simplified real situations into mathematical models
(e.g. terms, equations, figures, diagrams and functions)
Interpreting Students relate results obtained from manipulation within the model to the real
situation and thus obtain real results
Validating Students judge the real results obtained in terms of plausibility
33.2.2 Criteria for the Selection of Modelling Tasks
The selection of modelling tasks, especially in consideration of examination situa-

tions, is not easy. The criteria that such tasks should ideally fulfil can be formulated
as a reference point. In her broad classification scheme for modelling tasks, Maaß
(2010) especially focuses on the relation of the task to reality, the openness of a task
related to the question and the modelling activity as criteria for modelling tasks.
We will discuss these aspects in more detail below (based on Greefrath et al.
2017). Concerning the relation of the task to reality, we shall look at the authenticity
and relevance of the context. Taking the openness of a task into account, multiple
ways of solution will be the criteria we are looking at. In respect to the modelling
activity, we will take a closer look at the sub-competencies of modelling (see
Table 33.1).
33.2.2.1 Authenticity
Authenticity is a key feature of modelling tasks. The term authenticity is not used
consistently in mathematics education (Galbraith 2007). An overview of different
definitions of authenticity can be found in Vos (2011). Authenticity refers to an
extra-mathematical context which must be discussed by mathematical means in the
respective situation. The extra-mathematical context should be authentic instead of
just being constructed for this special mathematical task. Concerning the extra-
mathematical context, we follow the definition of the OECD (2001), which consid-
ers authentic tasks as “based on situations which … represent the kinds of problem
encountered in real life” (p. 23). The use of mathematics in this situation should not
be limited to mathematics education. It should be meaningful and realistic. Authentic
modelling tasks are problems that belong to an existing area and that are accepted
as such by the people who work in the respective field (Niss 1992). As a result, an
authentic task is credible for students and at the same time realistic in relation to the
environment. While working on authentic tasks, students can be sure that they are
working on a real problem which also occurs outside of mathematics education.
However, authenticity does not yet mean that the tasks are necessarily significant for
the present or future life of the students.
33.2.2.2 Relevance
A task is relevant if it can be considered meaningful for the present or future life of
the students. In the case of a task regarded as meaningful to the current life of the
students, we talk about relevance for the students. In contrast, we talk about life
relevance if a task becomes relevant for future situations. One can also talk about
relevance if the respective tasks can be linked to the present life of the students
(Leuders 2001, p. 100). Thus, relevance is more connected to students than authen-
ticity (Maaß 2010). This also fits the notion of Browder (1976, p. 251) that “the
criterion of relevance … is social utility and (…) we tend to assign such value in
terms of the effectiveness or efficiency of the mathematical techniques or practices
involved towards specific goals of social effort”.
33.2.2.3 Openness
Tasks that allow various approaches and solutions are called open tasks. The open-
ness of a task enables students to adopt an individual approach towards it. There
have been diverse classification schemes for open tasks. We limit ourselves to the
investigation of a task concerning its initial state, the transformation and its final
state based on Bruder (2000, 2005) and Wiegand and Blum (1999). These classifica-
tion schemes make use of the description of a problem through an initial state, a
final state and a transformation, as known in problem-solving psychology, and are

not limited to modelling tasks (Klix 1971). Open tasks are always distinguished by
the clarity of the initial and the final state, as well as the transformation. Another
classification of open tasks based on Bruder (2003) entails seven different types of
tasks and can be found in Maaß (2010). Here, there is a differentiation between
over- and under-determined tasks.
33.2.3 Research Questions
In the international discourse, mathematical modelling has been generally accepted

as a process-related competence. This fact, along with the overall importance of
competence orientation in the international didactic discussion, was the reason for
our investigating examination tasks in terms of this competence in more detail.
Within this scope, we have considered examination tasks taken from the stan-
dardised written school leaving examination of Austria, which was established
nationwide in 2015. The starting point of our consideration is the theory-driven
criteria for modelling tasks, which we are adopting to these examination tasks. Our
approach is supported by the introduction of a model of competence levels within
the scope of the written school leaving examination, because mathematical model-
ling plays a central role in Austria (Siller et al. 2016).
In relation to our approach, the question arose of how the quality of Austrian
school leaving examination tasks that contain modelling parts can be evaluated
from a didactic point of view relating to modelling tasks. Within this framework, we
limit ourselves to the so-called part 2 tasks (see BIFIE 2012). As a result, a first
research question arises: How can Austrian school leaving examination tasks be
evaluated concerning the sub-competencies of modelling, authenticity, relevance
and openness? In the following, we subsume all those characteristics of a task under
the term modelling character. In this context, we are looking at characteristics of
modelling tasks that often occur and ones that do not. Furthermore, we want to find
out if typical characteristics of modelling tasks can be identified in Austrian school
leaving examinations. In summary, the aforementioned question can be formulated
as such:
What kind of modelling character do Austrian school leaving examination
tasks have?
It is also interesting for the design of modelling tasks in examinations whether
these can be constructed at different levels of difficulty or whether the frequency of
solution is related to the modelling character of a respective task. In Austria, the data
concerning the frequency of solution is available, so a second question is also of
great interest:
Is there a relation between the modelling character of final examination tasks and
the frequency of solution?
33.3 Research Design
Tasks of the second part, so-called part 2 tasks, from the Austrian standardised
school leaving examinations of 2015, 2016 and 2017 are the basis for our investiga-
tion. These tasks are explicitly characterised in the examination conception (BIFIE
2012) as tasks that target the linking of “basic skills to clearly defined contexts and
application areas”. As a result, these tasks are more extensive and also have a clearer
relation to reality.
In the examination, four to six “part 2 tasks” that are divided into two to six sub-
tasks can be expected. In total, we investigated 72 items, 24 per implementation
year. Each item was evaluated according to the criteria concerning the modelling
character presented above (authenticity, relevance and openness). Moreover, the
items were evaluated according to the sub-competencies of modelling that are
required in each case.
Using the criteria and descriptions presented in Table 33.2, the items were rated
by two coders – the authors – (Cohens k = .92) within the scope of a qualitative
research process. Finally, all items were clearly assigned the corresponding charac-
teristics. In the few cases in which the coders differed, the coders discussed the
problem until agreement and set a common code which was chosen as the final
code. This resulted in an assignment of all individual items to the mentioned criteria.
33.4 Results
In the following, the results are presented in the order of the chosen criteria concern-
ing the modelling character, and then this is followed by the frequency of solution.
Table 33.2 Criteria to evaluate the modelling character of the items

Criterion Description
Relation to reality Can the task be related to an extra-mathematical context?
Authenticity (context) Is the context authentic in relation to the present situation? Is it a
realistic situation?
Authenticity (use of Is the context authentic in relation to the use of mathematics? Is the
mathematics) use of mathematics realistic?
Relevance Is the context relevant to the students? Is it a real problem for the
students?
Openness Is it possible to solve the tasks by different approaches?
Sub-competencies What kinds of sub-competencies of modelling are necessary for
solving (simplifying, mathematising, interpreting, validating)?
33.4.1 Relation to Reality
By evaluating the relation to reality of the considered items, one can see that many
of them certainly show a relation to reality. Only a few items were categorised as
inner-mathematical (such as “Determine the inflection point of the functions!”,
BIFIE 2015, p. 5). Almost all of the investigated items were embedded into a factual
context (86%) and were thus categorised as being close to reality. An example for
such an item is “Show that the detection limit of 14C is reached after 10 half-life
periods” (BIFIE 2015, p. 6). One could discuss if this item is relevant to students.
Referring to BIFIE (2012) and to the Austrian curriculum, we note that students
should be familiar with such tasks, although – in our opinion – they are not so rel-
evant to students. All in all, the number of examination items related to reality can
be classified as very high.
33.4.2 Authenticity
The evaluation concerning the authenticity of the tasks shows a clear difference to
the one looking at the relation to reality. The number of items that contain an authen-
tic factual context is approximately one sixth (11 of 72 items). Within this scope,
radiocarbon dating, the inheritance of lifeblood, the acceleration of the German
high-speed train ICE and income tax in Austria could be regarded as authentic fac-
tual contexts. However, not all of the sub-tasks concerning the respective contexts
can be evaluated as authentic because sometimes they do not represent a real prob-
lem. Only a small number of the investigated items represent an authentic use of
mathematics (4 of 72 items) because, in reality, the problems are often solved by
different mathematical methods than represented in the task. To give an example,
the average rate of change of the velocity of a high-speed train in a given time inter-
val is in reality not determined on the basis of diagram data (as required in a task
from 2016) but rather on the basis of identified numerical data. This example shows
that an authentic factual context does not go hand in hand with an authentic use of
mathematics.
33.4.3 Relevance and Openness
Tasks that are relevant to learners could hardly be identified (1 of 72 items). The
single context that can be considered as adequate is the investigation of the velocity
of a runner over a 200-metre-long distance with the aim of optimising the training
performance. All the other contexts do not directly concern the students or only
become relevant in later life. In addition, open items were rare as well. Four items
could be classified as open here because the calculation of the inflection points is
possible in different ways (different sufficient criteria). However, in most items, the
initial and the final state are clearly described, and the respective transition does not
provide much flexibility.
33.4.4 Sub-competencies
Whereas almost all of the investigated items show a relation to reality, there are only
a few items addressing certain sub-competencies of modelling. Mostly, one can find
items concerning interpreting (13 of 72 items) and some concerning mathematising
(6 of 72 items).
33.4.5 Frequency of Solution
Within the scope of the Austrian examinations from 2015 to 2017, the frequency of
solutions of all the chosen items has been determined. The average number of stu-
dents in each of the three cohorts is about 17,000, and each cohort sat its own
examination.
The success rate for the items of the years 2015 until 2017 is 44% on average.
The success rate (Fig. 33.2) shows that 25 of the 72 items have a frequency of solu-
tion between 0.2 and 0.4. Items that show an authentic factual context and require a
sub-competency of modelling (interpreting) can be found among the items with the
lowest (≤0.2) as well as the items with the highest success rate (>0.8).
Fig. 33.2 Distribution of the success rate of the items

Our investigation of tasks in the last central examination in Austria found that many
of the investigated items show a relation to reality. However, this does not suggest
that the items really contain modelling tasks. The application of typical criteria for
modelling tasks, such as authenticity, openness and relevance, to the investigated
items, shows that we can only talk about modelling tasks to some extent.
Approximately one sixth of the items contain an authentic factual context, but an
authentic use of mathematics can only be found in a few cases. Moreover, hardly
any task is open or relevant.
The final examination tasks in Austria are typically related to reality, not very
authentic, not open and not relevant from the perspective of the students. The most
preferred sub-competencies of modelling are – apart from working mathemati-
cally – interpreting and, to a small extent, also mathematising. The situation in
Austria is not automatically transferable to written central examination tasks in
other countries, but there is previous work such as that by Vos (2013) on the Dutch
mathematics examination papers selectively in the period 2006–2010 and by Julie
(2015) in the South African context. The example of Austria was chosen because of
a new conception of examinations and the accessibility of empirical data on the
frequency of solutions. However, the criteria used can also be applied to other
examination tasks and thus allow a comparison with respect to the modelling
content.
References
BIFIE. (Ed.). (2012). Die standardisierte schriftliche Reifeprüfung in Mathematik. Inhaltliche und
organisatorische Grundlagen zur Sicherung mathematischer Grundkompetenzen. Available
at: https://www.srdp.at
BIFIE. (Ed.). (2015). Mathematik Teil-2-Aufgaben. Test booklet. Available at: https://
w w w. s r d p . a t / fi l e a d m i n / u s e r _ u p l o a d / d ow n l o a d s / M a t u r a _ 2 0 1 4 - 1 5 / 1 2 _ M AT /
KL15_PT1_AHS_MAT_T2_CC_AU.pdf
C. Haines, P. Galbraith, W. Blum, & S. Khan (Eds.), Mathematical modelling (ICTMA 12):
Blum, W., Galbraith, P. L., Henn, H.-W., & Niss, M. (Eds.). (2007). Modelling and applications in
mathematics education. The 14th ICMI study. New York: Springer.
Browder, F. E. (1976). The relevance of mathematics. The American Mathematical Monthly, 83(4),
249–254.
Bruder, R. (2000). Akzentuierte Aufgaben und heuristische Erfahrungen. Wege zu einem
anspruchsvollen Mathematikunterricht für alle. In Lothar Flade, Wilfried Herget (2000) (Eds.),
Mathematik lehren und lernen nach TIMSS: Anregungen für die Sekundarstufen (pp. 69–78).
Berlin: Volk und Wissen.
Bruder, R. (2003). Konstruieren – auswählen – begleiten. Über den Umgang mit Aufgaben.
Friedrich-Jahresheft „Aufgaben. Lernen fördern – Selbstständigkeit entwickeln“ (pp. 12–15).
Seelze: Friedrich Verlag.
Bruder, R. (2005). Working with tasks for the learning of problem solving in maths teaching as an
issue of the first teacher training phase. ZDM Mathematics Education, 37(5), 351–353.
Galbraith, P. (2007). Authenticity and goals – Overview. In W. Blum, P. L. Galbraith, H.-W. Henn,
& M. Niss (Eds.), Modelling and applications in mathematics education (pp. 181–184).
New York: Springer. https://doi.org/10.1007/978-0-387-29822-1_17.
Greefrath, G., & Vorhölter, K. (2016). Teaching and learning mathematical modeling. Approaches
and developments from German speaking countries. Cham: Springer.
Greefrath, G., Siller, H.-St., & Ludwig, M. (2017). Modelling problems in German grammar
school leaving examinations (Abitur) – Theory and practice. Available at: https://keynote.
conference-services.net/resources/444/5118/pdf/CERME10_0165.pdf
Julie, C. (2015). Learners’ dealing with a financial applications-like problem in a high-stakes
school-leaving mathematics examination. In G. A. Stillman, W. Blum, & M. S. Biembengut
(Eds.), Mathematical modelling in education research and practice: Cultural, social and cog-
W. Blum, & S. Khan (Eds.), Mathematical modelling (ICTMA 12). Education, engineering and
economics (pp. 110–119). Horwood: Chichester.
Klix, F. (1971). Information und Verhalten. Bern: Huber.
KMK. (2012). Bildungsstandards im Fach Mathematik für die Allgemeine Hochschulreife
(Beschluss der Kultusministerkonferenz vom 18.10.2012). Köln: Wolters Kluwer.
Leuders, T. (2001). Qualität im Mathematikunterricht der Sekundarstufe I und II. Berlin: Cornelsen
Scriptor.
Maaß, K. (2010). Classification scheme for modelling tasks. ZDM Mathematics Education, 31(2),
285–311.
Niss, M. (1992). Applications and modelling in school mathematics – Directions for future devel-
opment. Roskilde: IMFUFA Roskilde Universitetscenter.
OECD. (2001). Knowledge and skills for life. First results from the OECD programme for
Internationale Student Assessment (PISA) 2000. Paris: OECD. Available at: http://www.oecd.
org/education/school/programmeforinternationalstudentassessmentpisa/33691596.pdf
Siller, H.-S., Bruder, R., Hascher, T., Linnemann, T., Steinfeld, J., & Sattlberger, E. (2016).
Competency level modelling for school leaving examination. In K. Krainer & N. Vondrová
(Eds.), Proceedings of CERME 9 (pp. 2716–2723). Prague, Czech Republic. <hal-01289473>.
Vos, P. (2011). What is ‘authentic’ in the teaching and learning of mathematical modelling? In
of mathematical modelling (pp. 713–722). Dordrecht: Springer.
Vos, P. (2013). Assessment of modelling in mathematics examination papers: Ready-made mod-
els and reproductive mathematising. In G. A. Stillman, G. Kaiser, W. Blum, & J. P. Brown
Wiegand, B., & Blum, W. (1999). Offene Probleme für den Mathematikunterricht – Kann man
Schulbücher dafür nutzen? Beiträge zum Mathematikunterricht (pp. 590–593), Hildesheim:
Franzbecker.
Part V
Applicability at Different Levels
of Schooling and University
Chapter 34
Characterising Modelling Competency
in Students’ Projects: Experiences
from a Natural Science Bachelor Program
Morten Blomhøj
Abstract The chapter investigates modelling competency in problem-oriented

project work in the interdisciplinary natural science bachelor program (Nat Bach) at
Roskilde University, Denmark. The focus is on analyses of students’ project reports
within the area of infectious diseases modelled by the standard SIR compartment
model and extensions thereof. Within this domain, the students’ modelling compe-
tency is analysed with respect to two dimensions in modelling competency. These
are the degree of coverage with respect to the modelling process and the technical
level of the modelling work and of the use of mathematics (Blomhøj M, Højgaard
T, Modelling and applications in mathematics education: The 14th ICMI study.
Springer, New York, pp 45–56, 2007). In addition, two types of students’ related
reflections are pinpointed and analysed, namely, reflections related to the particular
modelling sub-processes (internal reflections) and reflections concerning the role
and function of the model and its results in a context of application (external reflec-
tions) (Blomhøj M, Kjeldsen TH, Trends in teaching and learning of mathematical
modelling. Springer, Dordrecht, 2011). Based on the analyses and experiences with
supervising and assessing students’ modelling projects in the Nat Bach program, a
framework including the two dimensions and the two types of reflections is pro-
moted as a means of characterising progress in students’ mathematical modelling
competency in problem-oriented project work.
Keywords Compartment model(s) · Degree of coverage · External reflections ·

Internal reflections · Modelling competency · Project work · Technical level
M. Blomhøj (*)
IMFUFA, Department of Science and Environment, Roskilde University, Roskilde, Denmark
e-mail: Blomhoej@ruc.dk

396 M. Blomhøj
34.1 Context and Relevance of Research Question
Some background information is needed for understanding the type of problem-

oriented project work, which forms the context for the chapter. The Natural Science
Bachelor program (Nat Bach) is a three-year interdisciplinary program leading to
bachelor degrees with specialisations in two subjects, which give access to subse-
quent two-year masters within the chosen specialisation at Roskilde University or at
other universities. The program runs in Danish and in English with around, respec-
tively, 120 and 40 new students per year. As in all programs at Roskilde University,
half of the students’ study time is devoted to problem-oriented group-based project
work. In each semester (half a year), the students conduct one project of 15 points
(half of a semester) and three subject organised courses of 5 points each. Two of the
three courses can be chosen according to interests and desired specialisations. The
projects are conducted in groups of 2–6 students under supervision of a researcher.
The students write project reports of typically 40–80 pages possibly supplemented
with appendices (e.g. a poster, a draft for a paper, IT products or other designs). The
common pedagogical foundation for, and various practices of, the project work at
Roskilde University is presented and discussed by Andersen and Hielesen (2014).
The students’ learning in the project work is assessed individually at an oral
group examination with two examiners, the project supervisor and an internal or
external examiner. The examination takes as its point of departure the project report.
This may sound like a contradiction in terms. However, each group member is
responsible for the outcome of the project work and should therefore be able to give
account for, defend and reflect upon all aspects of the project report. During the
examination, the students are asked individual questions on different parts of the
project. Moreover, in the first three semesters and in the bachelor project in semester
6, the students draw lots for an individual question in relation to their project 3 days
before the project exam. In the exam, to begin with, these questions are addressed
in five-minute oral presentations with follow-up questions from the examiners. In
addition, the exam includes an academic discussion of the project to which all the
students are expected to contribute. The students are graded individually based on
their performance at the exam. The duration is 30 min per student, so for groups of
4–6 students, there is ample time for both examining individually and discussing the
project in detail.
Mathematical modelling plays a natural and important role in projects at Nat
Bach. Often, setting up and analysing mathematical models is part of the projects
for semester 1, where the projects have to address an exemplary case of applications
of science in technology and society. In semester 2, modelling is essential since the
theme for the projects is interplay between model, theory, experiment and simula-
tion in the natural sciences. In semester 3, the theme is theory of science in the natu-
ral sciences. Here, mathematical modelling and models can play a role but are
typically not in focus in the projects. In the subject specialisations in semesters 4
and 5, there are different constraints on the project work. In the mathematics and
physics modules, the students should construct or analyse models eventually in
34 Characterising Modelling Competency in Students’ Projects: Experiences… 397
connection with experimental work. In chemistry and in environmental, medical

and molecular biology, mathematical modelling can play a role in the projects, but
typically only groups with students who have mathematics as one of their subjects,
include mathematical modelling in these projects. For the final project in semester
6 (the bachelor project), there is no thematic or subject matter constraints. So here,
it is possible to work with mathematical modelling and models in interdisciplinary
projects, and this is the case for 10–15% of the bachelor projects.
For more than 20 years, together with colleagues, I have been engaged in
researching how to support the students’ development of mathematical modelling
competency through problem-oriented project work in this context. This involve-
ment has led to the research question: How can progress in students’ mathematical
modelling competency in problem-oriented project work be characterised? The
research question is of relevance for the continuing development of our practice of
supervising modelling projects. The problem-oriented project work at Nat Bach
represents a rather extreme realisation of a holistic approach to modelling in a bach-
elor program (Blomhøj and Kjeldsen 2018). The investigation of the research ques-
tion in this context may contribute to the characterisation of students’ progress in
mathematical modelling competency in interdisciplinary science studies and maybe
in upper secondary mathematics education.
34.2 Theory and Methodological Reflections
The research question is posed and studied within the framework of modelling com-
petency as it is defined in the Danish KOM-project (Niss and Højgaard 2011,
pp. 50–52). Here, modelling competency is pinpointed as one of eight main compe-
tencies spanning – together with three different types of overviews and judge-
ments – what it means to be competent in mathematics. Progress in the development
of each of the competencies is conceptualised further with respect to three dimen-
sions (pp. 72–73). In relation to modelling, these dimensions are (1) degree of cov-
erage, according to which part of the modelling cycle the students can work with,
(2) technical level, according to modelling techniques and mathematics used, and
(3) radius of action, according to the domains in which the students can perform the
competence (Blomhøj and Højgaard 2007).
In addition, we have found it relevant to distinguish between two types of stu-
dents’ reflections in modelling projects at Nat Bach, namely, internal reflections,
which are connected to the different sub-processes involved in the modelling pro-
cess (e.g. assumptions, estimation of parameters, analyses of the model, interpreta-
tions of the model and model results and the validity thereof), and external
reflections, which are connected to the possible, and/or actual, functions and conse-
quences of a model and its results in an investigation or decision process in a soci-
etal or scientific context (Blomhøj and Kjeldsen 2011).
Within this framework and in the context of project work at Nat Bach, the
research question is investigated through analyses of seven project reports involving
398 M. Blomhøj
mathematical modelling of infectious diseases. For pragmatic reasons, the projects

are selected among projects supervised by the author or where the author has been
involved in the projects in other ways (e.g. as opponent supervisor). The study is
explorative in character and aims to illustrate and discuss the students’ development
of modelling competency in the area of modelling infectious diseases. Three of the
projects address problems related to so-called child diseases, in particular measles,
mumps and rubella (MMR), against which vaccinations were included in the Danish
vaccination program in 1986. In Sect. 34.3.2, these projects are used to illustrate the
developments in the students’ modelling competency and their related reflections,
in particular their external reflections. The analyses include in addition two projects
on influenza epidemics and pandemics, and strategies for vaccination against influ-
enza, as well as two projects on the modelling of health strategies against the sexu-
ally transmitted diseases gonorrhoea and chlamydia. These four projects have
contributed to the analyses of the students’ development of modelling competency
in relation to the SIR compartment model (Susceptibles, Infectious, Recovered) for
a single epidemic course as described in Sect. 34.3.1. In particular, the analyses
have pinpointed how the students developed their modelling competency in dimen-
sion (1) and (2) during their work with uncovering and reflecting on the basic
assumption behind the SIR model or the SIS model (Susceptibles, Infectious,
Susceptibles) for bacterial infections such as gonorrhoea and chlamydia. In the
work in the project groups with modifying and extending these models for address-
ing the particular problem of the project, the students further developed their model-
ling competency and their related internal reflections. This is illustrated in Sect.
34.3.2 in relation to the projects on the MMR diseases.
34.3 Mathematical Modelling Projects in Epidemiology
Models and modelling are essential in epidemiology in general. Models are used for
defining and communicating basic concepts and notions in the field, for describing
and explaining data and phenomena related to infectious diseases, for predicting
courses of epidemics, for designing and testing vaccination programs or other types
of regulations and for providing a basis for design of healthcare policies (Bailey
1986). The field of infectious diseases is thereby exemplary for the role that model-
ling can play in interdisciplinary research fields in natural science – in particular for
interplay with biology. Moreover, the field is exemplary for the role models can play
in political decision-making related to risk phenomena, which according to Beck
(1992) characterize post-modern societies.
34.3.1 he Simple SIR Model as a Basis

T
for Modelling Projects
All seven projects analysed have taken their point of departure in the SIR model or
modifications thereof. The SIR model is a compartment model for the course of one
single epidemic caused by a virus infection and leading to immunity in a closed
population (Fig. 34.1). The projects modify and extend this model and develop and
apply models for more or less specific infectious diseases involving epidemiologi-
cal data.
Generally, in these projects, students’ understanding and interpretation of the
basic assumptions behind the compartment approach for modelling the spread of
the disease in focus play an important role. First, the students need to understand the
nature and strength of the basic assumption behind the SIR model, namely, that the
compartment variables describe the number of individuals with a certain infection
status and that this implies an assumption of homogeneity within each compart-
ment. Next, the students should understand that the flows between the compart-
ments are determined by (some of) the compartment variables, some parameters
and possibly some external in- and out-flows. This paves the way for mathematisa-
tion. The net rate of change in time for each compartment variable can be expressed
as the sum of inflows minus the sum of outflows. Thereby, a compartment model
(represented by a diagram as in Fig. 34.1) leads to a system of ordinary differential
equations.
Learning and reflecting on these assumptions and their consequences for the
models belong to the dimension of the technical level of the students’ modelling
competency both mathematically and in terms of compartment modelling tech-
niques. However, the assumptions are anchored to the sub-processes of problem
formulation, systematisation and mathematisation in the modelling process.
Working with uncovering of, and reflecting on, the underlying assumptions, the
students have to engage with these processes in reverse. Thereby, the students
develop their modelling competency in the dimension of the degree of coverage
with respect to the modelling cycle.
Fig. 34.1 The SIR compartment model with three solution curves in the phase plan (S, I)
400 M. Blomhøj
In the projects, to begin with, the groups typically implemented a simple model
in MatLab, a professional IT tool for modelling introduced in a compulsory course
at the very beginning of Nat Bach. The students analysed their model numerically
in order to learn about the meaning and importance of the parameters and the initial
values of the variables. Thereby, the students used and developed further the techni-
cal level of their modelling competence.
The third dimension – radius of action – is not in play within the projects.
However, the generality of compartment modelling facilitates a subsequent and
gradual development of the students’ radius of action in future modelling projects in
other areas.
The students’ understanding of the assumptions in terms of the dynamic of the
spread of an infectious disease is very important as a basis for their modelling work
with modifying and extending the model for addressing their problem. The biologi-
cal and epidemiological meaning of the model parameters, that is, the effective con-
tact rate (c) and the rate of recovery (v), is essential in this respect (see Fig. 34.1). In
particular, the parameter c is complex. It denotes the number of effective contacts
per time that each individual has with other individuals in the modelled population.
“Effective contacts” means contacts that would lead to an infection if occurring
between a susceptible and an infectious individual. Therefore, this parameter sum-
marises information about the population in terms of density and of normal social
behaviour, as well as medical information about the infectiousness of the disease.
Understanding that a constant recovery rate v means that the average time of
retention in the compartment I is given by 1/v was able to be established as a math-
ematical result by some project groups. In any case, it took some reflection for stu-
dents to understand that calculating the average time of retention actually makes the
homogeneity assumption meaningful for the modelling of the overall dynamic of an
infectious epidemic. The parameter v is specific for the disease, and its value can be
estimated from clinical data. Although depending on the disease, it might vary
with age.
In the first phase of the projects, the groups analysed the qualitative mathemati-
cal aspects of the simple models. For the SIR model, this typically included the
following results:
An epidemic occurs if and only if I′(t) > 0, which yields cS t 1 as a necessary
N
condition for an epidemic. If the entire population is susceptible (worst-case sce-
c
nario) before an infection, (S(0) ≈ N), we have R0 1 as a necessary condition
for an epidemic to occur.
R0 is called the reproduction number and is specific for the disease in the particu-
lar situation where the parameters are estimated. It denotes the number of new
infections that one infected individual can produce in the situation where the entire
population is susceptible. Hence, R0 denotes the infectivity of the disease in the
population in that situation.
Through this type of analysis, the students developed their understanding of the
model so they could interpret their mathematical results in the context of an epi-
demic. The epidemic criterion R0 > 1 became meaningful for the students: if an
infected person on average has less than one effective contact during the average
period of infection, no epidemic can occur.
Herd immunity is another key notion, which becomes meaningful for the stu-
dents through their mathematical analyses of the simple SIR model. For a disease in
a population, herd immunity occurs when the number of susceptible individuals is
too small for the disease to spread (i.e. S is too small for causing I′(t) to be positive).
The level of immunisation, which causes herd immunity in a population, is called
the critical immunisation threshold (pc). In the model, this can be controlled by
introducing vaccination before the infection. So the initial number of susceptible
individuals, S0, is reduced to (1 – pc) S0, where pc can be interpreted as the vaccina-
tion degree, which with S0 ≈ N yields a condition for herd immunity:
c 1 pc S0 v 1
1 pc 1 1
vN c R0
All these concepts are key in epidemiology and defined in close connection to math-
ematical models. In the projects, the students establish these results and see the
connections to the mathematical behaviour of the solutions. By analysing the solu-
tions in the phase plan (Fig. 34.1), the students can establish the meaning of the
pc value.
The fact that the system of differential equations in the SIR model cannot be
solved analytically is surprising for most students. Depending on the students’
mathematical pre-requisites and interests, this can lead to further mathematical
analysis of the model yielding an analytical expression for the solution curves in the
phase diagram by solving the differential equation for I’/S’. Through these activities
and related reflections, the students develop their modelling competency in the
domain of degree of coverage to the extent that they recapitulate the steps in the
modelling process behind the SIR model and reflect on the different types of
assumptions and their implementation in the mathematisation of the model. In the
dimension technical mathematical level, the students’ competency develops through
their use of numerical and analytical methods for analysing the model and its results.
The SIR model can easily be modified into a SIS model for bacterial infections
where immunity is not developed and the infectious flow back into the compartment
of the susceptible after recovering. The projects on gonorrhoea and chlamydia did
this model modification, and the students had to extend the model by dividing the
population into two subpopulations – one large group with average rate of unpro-
tected sexual contacts with new partners and a much smaller group with a much
higher contact rate. Thereby, the students developed a model that could be repro-
duced and that explained the relatively constant prevalence of both these diseases in
the Danish population. That is an interesting result, noticing that the reproduction
rates based on average contact rates for both diseases are far below one. Therefore,
a simple SIS model will predict extinction of both diseases. In addition, the models
made it possible for the students to test different interventions. For chlamydia, the
group tested a current – at that time – initiative with home testing as a way of reduc-
ing the average duration of the period of infection. For gonorrhoea, the group tested
402 M. Blomhøj
the effect of reducing the number of people with a high contact rate, for instance, by
means of offering prophylactic antibiotic treatment or via targeted information.
In all projects, the students developed their modelling competencies in dimen-
sions 1 and 2 and with respect to both types of reflections. Of course, the students’
progress varied among groups and among the students in the group. In the initial
phase of the projects, the reflections were typically internal reflections in the form
of critique of the model assumptions from an epidemiological point of view. For
example, some of the groups discussed whether the values of the parameters c and
v really are the same for different age groups in the population.
One of the groups working with influenza implemented age group-dependent
contact rates in an SIR model with five age groups. The model was used to evaluate
vaccination strategies. The Danish policy of offering free vaccination to the oldest
group was critiqued based on model results indicating that people in this age group
could escape influenza nearly as well if the same number of vaccinations were used
in the younger age groups instead. Such a strategy results in a much lower overall
prevalence of influenza in the population. In groups with stronger interest in math-
ematics, the students investigated and reflected on the sensitivity of the model
results towards variation in the parameters and their influence on the solutions in
terms of the maximum prevalence of the disease or the duration of an epidemic in
the model.
34.3.2 Modelling the MMR Child Diseases
The vaccination program against the MMR diseases and the epidemic of these so-
called child diseases have been analysed in several modelling projects during the
latest decades. The introduction of the MMR program was actually based on a
mathematical model, and this model was analysed in a project at Nat Bach. The
model produced by the health authorities was basically an economical cost-benefit
model, balancing the cost of the program with the value of the expected reduced
loss of parents’ workdays due to the three child diseases. However, this model did
not include the dynamics of the epidemics. Instead, in the model, the prevalence of
the diseases was reduced according to the immunisation degree caused by vaccina-
tion. Thereby, the model underestimated the effect of the program on the one hand,
but on the other hand, the model could not take into account the medical side effects
and related costs of infections among adults, in the case when the critical vaccina-
tion degree was not reached. The project group uncovered the assumptions behind
the model and analysed and critiqued the limitations of the model by comparing it
with an SIR model. Thereby, the students developed their modelling competency in
dimension 1 and their critical reflections on the role of a model in a societal decision
process (i.e. their external reflections).
In the current situation in Denmark, the critical immunisation degree for mea-
sles, which is estimated to be around 95% (corresponding to a R0 number of 20), is
not reached. The vaccination degree has varied between 80% and 88% during the
last two decades. Accordingly, Denmark has had small outbreaks of measles with
around 100 reported cases in this period. Among the MMR diseases, measles is the
most infectious and serious disease in terms of complications among infected adults.
From time to time, there has been quite some public debate about the MMR vacci-
nation program offered to all children aged 15 months and again at age 4 years.
Some 10 years ago, autism was under suspicion for being a (rare) side effect of the
MMR vaccination, and at the same time, there was a group of parents and future
parents who argued that it was healthier for children to go through the “natural”
infections instead of taking the vaccine. Although the risk of complications is much
higher for infected than for vaccinated children, the debate caused a decrease in the
number of child vaccinations.
Two projects have analysed the situation by means of setting up age-structured
compartment models with birth and death, a cohort vaccination degree (immunisation
degree, p) and transitions from one age group to the next according to numbers of
years covered by the age group (Li). The compartment diagram is shown in Fig. 34.2.
One of these projects – a bachelor project – focused on the public debate about
the MMR program and the possible role that mathematical models can play in such
a debate. Scenarios for the next 20 years were simulated in the model with yearly
vaccination degrees in the youngest cohort around 85% and with a declining num-
ber of immune people in the older cohorts with 100% immunity from having the
infection before the MMR vaccination program. Although the model probably over-
estimates the frequency and magnitude of measles epidemics, the qualitative pattern
in the model results, with a radical shift towards an older age profile among the
infected individuals, was robust with respect to variation in the parameter values.
With a subcritical child vaccination level, measles will be a disease for middle-age
and old people, for whom complications are more frequent and more serious.
As part of the project, the student interviewed three persons who had expressed
scepticism about the MMR program in the public debate and on the internet. The
interviewees were presented with the model results and asked if the results effected
their opinion. Although people do not easily change opinion about such matters, it
became clear in the interviews that the phenomena of herd immunity and age shift
Fig. 34.2 Aged-structured SIR model for MMR diseases

404 M. Blomhøj
in the group of infected were not known by the interviewees beforehand. In that
sense, the model provided new aspects to the problematics for the interviewees.
The third project addressing the MMR problem field was a project in the math-
ematics specialisation module. Here, the group included an extra compartment for
exposed individuals. Thus, for each age group, the dynamics during an epidemic
were modelled with an SEIR (Susceptible, Exposed, Infectious, Recovered) com-
partment model. The main emphasis was on the mathematical analysis of the model.
In addition, the group also developed a simulation program, in which each individ-
ual is represented by a small circle moving randomly on the computer screen and
shifting colour depending on its infection status. If a susceptible green circle meets
an infectious red circle, it becomes grey for being exposed, and after the time of
latency, it becomes red and infectious until it turns blue and immune after the time
of infection. The simulation program was used to illustrate some aspects of the
mathematical behaviour of the model, for example, the importance of taking into
account the state of being exposed and the element of stochasticity in the outbreak
of an epidemic in situations where the herd immunity threshold is not reached. The
students discussed these matters based on their model results and simulations.
Thereby, they demonstrated advanced internal reflections. Moreover, the students
developed their external reflections in the project by discussing which role these
models can play in public discussions of health policy in relation to the MMR dis-
eases. For example, the group discussed if the model results could be used as an
argument for compulsory vaccination.
In both projects, the students developed their modelling competency in the
dimension of degree of coverage by uncovering and analysing the modelling pro-
cess behind the standard models. By modifying and extending the standard model,
the students gradually gained experiences with the full modelling cycle. In terms of
the technical level of the competency, in the last mentioned project in particular, the
students developed their mathematical and computer-based skills and competencies
supporting their modelling competency. However, in both projects, the students
developed the technical level of their modelling competency by analysing compart-
ment models both numerically, using MatLab, and analytically. Also in both proj-
ects, the students developed their internal as well as their external reflections related
to the role and function of models in societal contexts of application.
34.4 onclusion – Characterising Students’ Modelling

C
Competency in Projects in Epidemiology
From supervising and analysing projects at Nat Bach, it is evident that the area of
epidemiology constitutes a very fruitful problem field for developing students’
mathematical modelling competency through project work at bachelor level. The
area is accessible mathematically for first-year students, and the students can enter
the field of modelling infectious diseases based on their experiences, previous
knowledge and readings in epidemiology. In an interdisciplinary program such as

Nat Bach, the students can develop their knowledge about the biological and medi-
cal aspects of the specific infectious diseases during their modelling project.
The framework with the three dimensions and two types of reflections has proven
to be a possible answer to the research question with respect to the students’ devel-
opment of modelling competency in projects within the field of epidemiology.
Progress in the third dimension – radius of action – is not addressed in this study.
That will require a design where the same students can be followed in modelling
projects in different domains. However, the framework has been instrumental for
qualifying the assessment of the students’ modelling competency, for developing
the practice of supervising modelling projects and for tailoring the feedback to the
students on their progress in modelling competency in their projects.
References
Andersen, A. S., & Heilesen, S. B. (Eds.). (2014). The Roskilde model: Problem-oriented learning
and project work. Heidelberg: Springer.
Bailey, N. T. (1986). Macro-modelling and prediction of epidemic spread at community level.
Mathematical Modelling, 7(5–8), 689–717.
Beck, U. (1992). Risk society – Towards a new modernity. London: Sage (first published in
German, 1986).
Blomhøj, M., & Højgaard, T. (2007). What’s all the fuss about competences? Experiences with
using a competence perspective on mathematics education to develop the teaching of math-
ematical modelling. In W. Blum, P. L. Galbraith, H.-W. Henn, & M. Niss (Eds.), Modelling and
applications in mathematics education: The 14th ICMI study (pp. 45–56). New York: Springer.
Blomhøj, M., & Kjeldsen, T. H. (2018). Interdisciplinary problem oriented project work – A learn-
ing environment for mathematical modelling. In S. Schukajlow & W. Blum (Eds.), Evaluierte
Lernumgebungen zum Modellieren (pp. 11–29). Wiesbaden: Springer Spektrum.
Blomhøj, M., & Kjeldsen, T. H. (2011). Students’ reflections in mathematical modelling projects.
In G. Kaiser, W. Blum, R. Borromeo Ferri, & G. Stillman (Eds.), Trends in teaching and learn-
ing of mathematical modelling (pp. 385–396). Dordrecht: Springer.
Niss, M., & Højgaard, T. (2011). Competencies and mathematical learning: Ideas and inspiration
for the development of mathematics teaching and learning in Denmark. Roskilde: Roskilde
University, IMFUFA.
Chapter 35
From Royaumont to Lyon: Applications
and Modelling During the 1960s
Dirk De Bock and Bert Zwaneveld
Abstract At the Royaumont Seminar (1959), the New Math reform was officially
launched. In the decade between Royaumont and the first ICME congress in Lyon
(1969), many mathematics educators were involved in actions to facilitate the
implementation of the New Math reform. The New Math advocates were convinced
that a deep knowledge and understanding of the structures of modern mathematics
were prerequisites to arrive at substantial applications, but in actual classroom prac-
tices, the applied side of mathematics was often completely neglected. But already
in Royaumont, there were alternative voices who pleaded for taking the role of
applications seriously. We investigate the arguments for integrating applications in
mathematics education, as well as the kind of (new) applications that were envis-
aged, at the Royaumont Seminar and in the decade thereafter.
Keywords Hans Freudenthal · ICME-1 · Modern mathematics · New Math ·

Realistic Mathematics Education · Royaumont Seminar
35.1 Introduction
The Organisation for European Economic Cooperation (OEEC) Seminar held from
November 23 to December 5, 1959, at the Cercle Culturel de Royaumont in
Asnières-sur-Oise (France) is considered as a turning point in the history of math-
ematics education in Europe and in the USA (De Bock and Vanpaemel 2015). As
Bjarnadόttir (2008, p. 145) stated, “The Royaumont Seminar can be seen as the
beginning of a common reform movement to modernize school mathematics in the
world”. The famous slogan “Euclid must go!”, launched at Royaumont by the
D. De Bock (*)
Faculty of Economics and Business, KU Leuven, Leuven, Belgium
e-mail: dirk.debock@kuleuven.be
B. Zwaneveld
Welten Institute, Open Universiteit, Heerlen, The Netherlands

408 D. De Bock and B. Zwaneveld
Bourbakist Jean Dieudonné, became a symbol of the radical modernisation of

school mathematics. Most of the Royaumont proposals were strongly influenced by
Bourbaki, the French structuralist school whose members or adherents, such as
Gustave Choquet, Jean Dieudonné, Lucienne Félix and Willy Servais, were well
represented at the seminar. According to these scholars, the basic model for mod-
ernising school mathematics should be the academic discipline of mathematics, as
reconstructed and formalised from the late 1930s on by Bourbaki.
Less well known is that alternative reform proposals, emphasising the role of
applications, were also voiced at Royaumont. These application-oriented proposals
were however less decisive for developments during the 1960s than the dominant
structuralist ideas. In this chapter, we first discuss the kind of (new) applications that
were envisaged by some Royaumont lecturers, as well as their pleas for integrating
the applied side of mathematics in secondary school curricula. This discussion will
be based on New Thinking in School Mathematics (OEEC 1961a), the official report
of the Royaumont Seminar. Second, we examine the views of the more radical New
Math reformers on applications or, more generally, on the usefulness of mathemat-
ics. Third, we follow the debate on applications and modelling in the mathematics
education community between Royaumont and the first ICME congress (ICME-1)
in Lyon (August 24–31, 1969). For several reasons, this decade is less well docu-
mented in the history of our field. At that time, L’Enseignement Mathématique, the
only international journal on mathematics education until the late 1960s, had
become a purely mathematical journal (see Furinghetti [2009]), and the interna-
tional conference series, which are now strongly established in our field (ICME,
PME, ICTMA, HPM, …), had not yet started. An exception might be the annual
meetings of the International Commission for the Study and Improvement of
Mathematics Teaching (CIEAEM) founded in the early 1950s, but CIEAEM only
started publishing Proceedings of their meetings in 1974 (Bernet and Jaquet 1998).
Although there were no strong international communication channels in the
mathematics education community during the early and mid-1960s, it was a very
rich period of noteworthy international activity, including several seminars and
symposia organised under the auspices of OEEC/OECD, UNESCO or International
Commission on Mathematical Instruction (ICMI) (see, e.g., Furinghetti et al.
[2008]). These meetings mainly focused on issues related to the forthcoming intro-
duction of New Math (program development, renewal of geometry teaching, teacher
(re-)education and new didactical methods), but occasionally, concerns and propos-
als about the integration of applications in the curricula were expressed too. On the
basis of the Proceedings and other edited documents from these meetings, we more
generally review the visions on the role of applications in the international mathe-
matics education community of that time.
By the end of the 1960s, the debate on applications and modelling gained
momentum. Hans Freudenthal, at that time president-elect of ICMI, organised the
international colloquium “How to Teach Mathematics so as to Be Useful” (Utrecht,
August 21–25, 1967). The contributions to that colloquium were published in the
first issue of Educational Studies in Mathematics (May, 1968). In his introductory
address, Freudenthal took the opportunity to explain his views on the colloquium’s
35 From Royaumont to Lyon: Applications and Modelling During the 1960s 409
theme. He argued that students could not be expected to (be able to) apply the math-
ematics they had been taught in a purely theoretical way. Instead, to enable students
to apply the mathematics they have learned, mathematics education should start
from concrete contexts and patiently return to these contexts as often as needed
(Freudenthal 1968). It is the beginning of a new era in which applications and mod-
elling gradually became an essential part of mathematics education. In the
Netherlands, the theory and practice of Realistic Mathematics Education (RME)
were developed and inspired the teaching of mathematics in a large number of coun-
tries worldwide. We conclude this chapter with a more detailed discussion of
Freudenthal’s ideas on applications and modelling in the years preceding ICME-1.
35.2 Applied Mathematics at the Royaumont Seminar
The Royaumont Seminar was organised by the Office for Scientific and Technical
Personnel (OSTP) of the Organisation for European Economic Cooperation
(OEEC; later joined by nations outside Europe to form the Organisation for
Economic Co-operation and Development, OECD). The office was created for the
purpose of promoting international action to increase the supply and improve the
quality of scientists and engineers in OEEC countries (OEEC 1961a). The main
motive for OEEC/OSTP to organise a seminar aimed at upgrading mathematics
education was clearly economic: Industry and other branches of economic activity
were confronted with new applications of mathematics leading to a demand for
more mathematicians with new kinds of skills. Therefore, a reappraisal of the con-
tent and methods of school mathematics was needed. In his opening address,
Marshall H. Stone, at that time president-elect of ICMI, formulated the functional
argument as follows:
[…] the usefulness of mathematics in practical matters has been an added factor in its vital-
ity as a component of the school curriculum. In this period of history it is the rise of modern
science and the ensuing creation of a technological society which compels us to give
increasing weight to the utilitarian arguments for the more intensive teaching of mathemat-
ics. (OEEC 1961a, p. 17)
Stone also emphasised the need for a better coordination between mathematics and
science teaching: “It is not going to be sufficient to improve the mathematical cur-
riculum as an isolated part (…). It is of the first importance that instruction in math-
ematics and in the various sciences should be adequately co-ordinated” (OEEC
1961a, p. 21).
In view of the above, the Royaumont Seminar should thus have been a break-
through of an applied and interdisciplinary perspective in mathematics education,
but it turned out differently. Due to a dominance at the seminar of professional
mathematicians, most of them members or adherents of the French structuralist
school, pure academic mathematics was de facto adopted as a model for school
mathematics and most participants only paid lip service to the active application of
mathematics. For instance, Dieudonné admittedly referred to applications to
theoretical physics as a main argument for the inclusion of new topics in university
courses of analysis but left open the question whether any kind of “applied mathe-
matics” should already be integrated in the secondary school programs. Nevertheless,
he believed that a favourable consideration of his reform proposals, having a clear
Bourbaki orientation, would already provide the theoretical foundations for teach-
ing questions of applied mathematics (OEEC 1961a).
An alternative voice at Royaumont was that of Albert W. Tucker, a Canadian
mathematician at that time working at Princeton. Tucker discussed the aspect of
new uses of mathematics and their implication for mathematics education. Rather
than study problems which involve two variables – or at most three or four – as most
problems in classical physics, new branches of mathematics are developed to deal
with complex realities involving several variables, which often occur in the social
sciences, for instance in economics and psychology. Within these realities, Tucker
distinguishes problems of disorganised complexity and problems of organised com-
plexity. The first category refers to problems with numerous variables and asks for
techniques of probability theory and statistical interference, being effective for
describing “average behaviour”. Problems of organised complexity involve a siz-
able number of factors which are interrelated into an organic whole and require,
among other things, a knowledge and use of matrix algebra. Tucker exemplifies this
last category with a problem of linear programming, utilising inequalities, intersec-
tions, graphic methods and unique algebraic procedures for solving equations.
According to Tucker, integration in all secondary school programs of these newer
types of mathematics, in a suitable form, is feasible and desirable. He, however,
acknowledges that an effort is needed to enhance teachers’ knowledge about mod-
ern mathematics and its applications to teach the subject well (OEEC 1961a).
Tucker’s plea for the integration of probability theory and statistics in secondary
school curricula was supported by Luke N. H. Bunt from Utrecht University (the
Netherlands). Bunt presented at Royaumont the outline of a syllabus on this subject
matter taught in a Dutch experiment for the alpha streams of secondary schools (for
more details on this experiment, see, e.g. Bunt [1959]):
(a) Some elements of descriptive statistics, such as frequency distributions, histograms, mean,
median, and standard deviation.
(b) “Classical” probability theory, with proofs of some of the elementary theorems.
(c) Intuitive treatment of binomial probability distributions; application to physics.
(d) Testing of a hypothesis (Bernoulli type of distribution); null hypothesis; level of significance;
sample space; critical region; confidence limits; sign test; rank correlation. Only Type I errors
(accepting a false hypothesis) are considered. (OEEC 1961a, p. 91)
For Bunt, the problem of estimating some characteristics of a population on the
basis of the values of these characteristics in a sample should be the dominant objec-
tive of a course in statistics. Also, Bunt’s proposal went against the general trend of
the Royaumont Seminar because (a) he did not primarily focus on those mathemati-
cally gifted students that would become mathematicians or engineers but on future
students in economics, psychology and other social sciences and (b) his didactical
approach was pragmatic rather than mathematically rigorous.
35.3 ew Math Reformers’ View on Applications

N
and Modelling
Although at Royaumont and in the decade thereafter there were several calls for
mathematical instruction to take applications of mathematics seriously (Niss et al.
2007), New Math, strongly focusing on theoretical academic mathematics, was – at
least in continental Europe – the dominant reform paradigm. Originally, the ambi-
tions of the New Math reformers and practitioners’ call for a focus on useful math-
ematics were not in contradiction or as stated by Niss (2008):
It is worth noticing that despite the strong theoretical orientation of the New Math move-
ment, its founders insisted that one of the points of the reform was to provide an ideal
platform for dealing with the application of mathematics to matters extra-
mathematical. (p. 72)
Claims about the omnipresence and increased usability of (modern) mathematics

can be found in many contemporary sources. In the Charte de Chambéry, a main
French reform document prepared by the “Commission Lichnerowicz” and adopted
by the Association des Professeurs de Mathématiques de l’Enseignement Public
(APMEP), the broad usability of modern mathematics is emphasised (and used as a
main argument for the reform of mathematical teaching at all educational levels).
Contemporary mathematics is useful in many fields: theoretical physics of course, but also
computer science, operational research, stock management of companies, organization
charts of big administrations, planning for major projects, sociology, linguistics, medicine
(diagnosing), pharmacy … . (Charte de Chambéry 1968)
Georges Papy, the architect of the new mathematical curriculum in Belgium and
president of the CIEAEM during the mid-1960s, wrote in the Préface of
Mathématique Moderne 1, the first volume of his pioneering textbook series:
The scope of the material studied in the first 13 chapters [sets and relations] goes far beyond
the boundary of mathematics. The student is initiated into types of reasoning constantly
used in all spheres of thought, science and technology. (Papy 1963, p. vii)
A closer look at the Mathématique Moderne approach reveals that Papy indeed
occasionally leaves the pure mathematical path and presents “daily life” situations
to introduce new mathematical concepts and structures, but these situations do not
incorporate realistic or authentic problem situations to be solved with mathematical
tools. Their only purpose is to facilitate comprehension of an abstract formalised
definition of the mathematical concept or structure that is targeted. Moreover, the
newly learned mathematics is never (re)invested to analyse and to solve new chal-
lenging problems outside mathematics.
To better characterise the role of extra-mathematical situations in New Math
courses of the 1960s, Hilton’s (1973) distinction between illustration and applica-
tion might be helpful. The point Hilton made is essentially the following. A situa-
tion, within or outside mathematics, is an illustration of a mathematical theory if
and only if that situation clarifies the theory. A situation is an application of a math-
ematical theory if and only if that situation is clarified by the theory. For the
high-order mathematical structures of the New Math, such as groups, fields or vec-
tor spaces, no applications were available for the early-aged students to whom these
structures were taught, and thus, these structures only could be illustrated with con-
crete instantiations (e.g. concrete materials or games especially constructed for that
purpose). Although New Math advocates often referred to the universal applicability
of the powerful structures of the modern mathematics in today’s science and tech-
nology, they were unable to demonstrate this applicability to their students; for
them, it was just words, no tools for real problem-solving, application analysis or
modelling.
[Mathematical] structures are great and admirable machines, but they can produce, in early
mathematics education, only too small things and too small effects. These small things are
the naive examples of structures which embellish modern mathematics textbooks and
which have been designed especially for students. (Rouche 1984, p. 138)
35.4 Applications and Modelling in the Post-Royaumont Era
Based on a survey of 21 national reports, Kemeny (1964) observed that the main
interests of the international mathematics education community in late 1950s/early
1960s were one-sidedly directed towards pure mathematics. The debate was focused
on the type of new mathematical subjects that could find a place in secondary school
programs, on how the teaching of traditional topics could be improved by the adop-
tion of modern ideas, on the “right” way of teaching geometry, and so on. With the
exception of a widely supported plea for teaching some notions of statistics at the
secondary level, relatively little attention was paid to the applied side of mathemat-
ics. Also at the international meetings on mathematics education of the 1960s,
organised by OEEC/OECD, UNESCO or ICMI, only occasionally ideas for inte-
grating applications of mathematics in secondary school curricula were voiced. In
the next paragraphs, we briefly discuss three main sources of applications that, aside
from statistics, were mentioned at these forums.
First, reference was still made to applications of mathematics to classical phys-
ics. The Group of Experts, that met in Dubrovnik (1960) for the purpose of prepar-
ing a detailed synopsis for modern secondary school mathematics, as stipulated in
one of the Royaumont resolutions (OEEC 1961a), re-insisted on the need for a
better coordination between the teaching of mathematics and the teaching of sci-
ence (particularly of physics) but provided little or no concrete suggestions to put
that coordination into practice. An exception might be the early introduction of vec-
tors and the systematic development of their algebraic and geometrical properties in
a modern curriculum for school geometry, which they considered, at least poten-
tially, of the greatest use to the students and teachers of physics (OEEC 1961b).
From physical scientists, an increasing pressure was felt to teach a more or less
intuitive introduction to calculus in secondary schools – which was not the case in
many countries – but mathematics education reformers of that time did not have
clear ideas how such an introduction could be properly integrated in a modern math-
ematical curriculum (Kemeny 1964).
Second, there is the mathematics related to the upcoming computing machines
which began to fundamentally impact secondary school mathematics. Examples of
new computer-related applications and their curricular impact were thoroughly dis-
cussed at the OECD conference in Athens (1963). That conference provided a spe-
cial section on “applications in the modernisation of mathematics” (OECD 1964) in
which Henry O. Pollak (USA), one of the pioneers in the field of applications and
modelling in mathematics education, examined, among other things, new areas of
mathematics motivated by computer sciences. He stated that the basic notions of
programming, including the use of flow diagrams in the construction of algorithms,
should be essential parts of secondary school curricula. But as Hermann Athen, a
German contributor to that Athens conference, argued, computers may have a much
broader impact:
A factor not to be neglected is the technical and economic revolution which is taking place
as a consequence of the big automatic computers. This revolution in psychic and intellec-
tual functions of human thinking and computing is continuously leading to new investiga-
tions in the fields of logic and the analysis of thinking. There is practically no field of
mathematical investigation which is not dependent upon the use of computers, e.g. many
problems of the social, behavioural, managerial and economic sciences. (OECD
1964, p. 245)
Other topics, in some way or another related to computers or their use, were sug-
gested at that time, for instance, binary representations of numbers, coding, numeri-
cal analysis, discrete mathematics, electrical circuits, logic and Boolean algebra.
Third, as a genuine application to economics and other social sciences, linear
programming was repeatedly mentioned. The topic fitted well within a modern
course of linear algebra but also could strengthen students’ numerical skills related
to solving equations and inequalities. Moreover, it opened a window to operational
research, a recent field of applied mathematics that deals with the application of
advanced analytical methods to help make better decisions given certain constraints.
Probably more than other fields of application, optimisation involves mathematising
and modelling, that is, interpreting a real-world situation in terms of a precisely
formulated mathematical model (OECD 1964). Modelling and models were not yet
widespread notions during the 1960s, but they gained ground. In his Introduction to
the Proceedings of the UNESCO colloquium in Bucharest (1968), Nicolae
Teodorescu observed that the notion of model had acquired universal presence and
circulation and already acknowledged the cyclic nature of modelling processes.
Modelling the complex, heterogeneous reality is the deliberate aim of any modern research
method in sciences of nature, in social sciences and in humanities. The victorious penetra-
tion of mathematics in other scientific domains is accounted for by modelling which,
repeated successively, leads to mathematical models. (International UNESCO Colloquium
1968, p. 27)
35.5 Freudenthal and the Emerging RME Movement
The end of the 1960s was characterised by an increased interest for the didactics of
mathematics, particularly at the micro level. Not only the purely mathematical sub-
jects but also the way a child learns became a main guiding principle for developing
mathematics education. This new trend was reflected in a growing number of (inter-
national) congresses and meetings in the field. In the context of this chapter, the
ICMI colloquium initiated by Freudenthal around the theme “How to Teach
Mathematics so as to Be Useful” (1967) deserves our special attention. It was the
first meeting in which an international panel discussed the differences in opinion
about the role of the use of mathematics (La Bastide-Van Gemert 2015). In his
opening address, Freudenthal sketched, in a general way, his views on mathematics
education. He explained that teaching mathematics “so as to be useful” is not the
same as teaching useful mathematics:
Useful mathematics may prove useful as long as the context does not change, and not a bit
longer, and this is just the contrary of what true mathematics should be. Indeed it is the
marvelous power of mathematics to eliminate the context. […] In an objective sense the
most abstract mathematics is without doubt the most flexible. In an objective sense, but not
subjectively […]. (Freudenthal 1968, p. 5)
He further argued that we should teach neither “applied mathematics” nor “pure
mathematics” (and expect that the student will be able to apply it later). Mathematics
is rather learned by doing, as a human activity, as a process of mathematising reality
and, if possible, even of mathematising mathematics.
The problem is not what kind of mathematics, but how mathematics has to be taught. In its
first principles mathematics means mathematizing reality, and for most of its users this is
the final aspect of mathematics, too. For a few ones this activity extends to mathematizing
mathematics itself. (Freudenthal 1968, p. 7)
Contemporary mathematics educators would say that the teaching of mathematics

should aim at making sense of the mathematics that is taught, but in the decade we
discuss in this chapter, it seemed to be not a main objective, although its importance
from the child’s point of view was already pointed out by Piaget (1954).
Freudenthal’s colloquium is sometimes regarded as the symbolic beginning of a
new era in the history of applications and modelling in mathematics education, the
so-called advocacy phase (Niss et al. 2007) in which advocates of applications and
modelling provided arguments in favour of the serious inclusion of such compo-
nents in the teaching of mathematics. In several countries (such as, for example, the
UK and the USA), this phase was quickly followed by a second development phase,
mainly characterised by the development of new educational materials to put such
teaching into practice, sometimes by institutes especially created for that purpose.
In the Netherlands, for example, the implementation was driven by the Institute for
the Development of Mathematics Education (IOWO) – founded in 1971 by
Freudenthal, nowadays the Freudenthal Institute – which shaped the philosophy and
practice of RME. In the RME vision, mathematics teaching should start from rich
problems that embed the relevant mathematics. A more recent development is called
emergent modelling and refers to the modelling of problem situations that (in part)
have been invented by the students themselves and that evoke the need for a math-
ematics which is new to them (Gravemeijer 2007).
But these developments did not yet reflect a global trend in mathematics educa-
tion during the late 1960s. When in 1969 the first ICME congress was held, only
two of the twenty published papers were related to applications and modelling
(Editorial Board of Educational Studies in Mathematics 1969; Gispert 2003). The
neglect of the applied side of mathematics will however soon be undone: Already at
ICME-2 (Exeter, 1972), a working group, chaired by Pollak, was devoted to “appli-
cation of mathematics” (Howson 1973), an initiative that will pave the way for the
establishment of a conference series and a community of mathematics educators
developing and investigating applications and modelling at various educational lev-
els (ICTMA) in the early 1980s.
35.6 Conclusions and Discussion
The New Math movement, launched at the Royaumont Seminar, was dominated by
academic mathematicians who had a genuine interest in education, but most of them
were involved in pure more than in applied mathematical research. Applications
were not their first concern. But although most leading reformers of the late 1950s
and the early and mid-1960s focused on pure mathematics and only paid lip service
to the active application of mathematics, some mathematics educators highlighted
the role of (new) applications and regarded it as an essential element in the moderni-
sation of mathematics teaching. Calls for taking applications – and later also math-
ematical modelling – seriously grew louder by the end of the 1960s and culminated
in Freudenthal’s colloquium “How to Teach Mathematics so as to Be Useful”. This
meeting marked the beginning of a new and more favourable period in the history
of the teaching of mathematical modelling and applications.
Since then, the mathematics education community has travelled a long journey.
New Math no longer plays a role in decision-making about education, and in retro-
spect, the debates of the 1960s between pure and application-oriented mathemati-
cians were not that fruitful. Finding a good balance seems to be more important than
to polarise. Modelling and applications are now recognised as a key component of
mathematics education, both in the academic debate and in the educational practice.
Many mathematics education conferences nowadays have a modelling and applica-
tion study group, and ICTMA has been integrated into ICMI. In parallel, the volume
of research on modelling and applications has grown, and the field has been theo-
retically conceptualised. Instructional materials and experimental curricula for dif-
ferent educational levels were developed so that contemporary mathematical
curricula in most – if not all – countries include an introduction to statistics and pay
explicit attention to applications and to the notion of modelling, not only at the
secondary level.
References
Bernet, T., & Jaquet, F. (1998). La CIEAEM au travers de ses 50 premières rencontres [The
CIEAEM through its first 50 meetings]. Neuchâtel: CIEAEM.
Bjarnadóttir, K. (2008). Fundamental reasons for mathematics education in Iceland. In B. Sriraman
(Ed.), International perspectives on social justice in mathematical education (Monograph 1),
The Montana Mathematics Enthusiast (pp. 137–150). Missoula: The University of Montana.
Bunt, L. N. H. (1959). L’enseignement de la statistique dans les écoles secondaires des Pays-
Bas [The teaching of statistics in Dutch secondary schools]. Mathematica & Paedagogia,
17, 35–48.
Charte de Chambéry. (1968). Étapes et perspectives d’une réforme de l’enseignement des mathé-
matiques [Steps and perspectives of a reform of mathematical teaching]. Paris: APMEP.
De Bock, D., & Vanpaemel, G. (2015). Modern mathematics at the 1959 OEEC seminar at
Royaumont. In K. Bjarnadóttir, F. Furinghetti, J. Prytz, & G. Schubring (Eds.), Proceedings
of the third international conference on the history of mathematics education (pp. 151–168).
Uppsala University, Department of Education: Uppsala.
Editorial Board of Educational Studies in Mathematics. (1969). Addresses of the first international
congress on mathematical education. Educational Studies in Mathematics, 2(2–3), 134–418.
Freudenthal, H. (1968). Why to teach mathematics so as to be useful. Educational Studies in
Mathematics, 1(1–2), 3–8.
Furinghetti, F. (2009). The evolution of the journal L’Enseignement Mathématique from its initial
aims to new trends. In K. Bjarnadóttir, F. Furinghetti, & G. Schubring (Eds.), “Dig where you
stand”. Proceedings of a conference on on-going research in the history of mathematics educa-
tion (pp. 31–46). Reykjavik: University of Iceland, School of Education.
Furinghetti, F., Menghini, M., Arzarello, F., & Giacardi, L. (2008). ICMI renaissance: The emer-
gence of new issues in mathematics education. In M. Menghini, F. Furinghetti, L. Giacardi, &
F. Arzarello (Eds.), The first century of the international commission on mathematical instruc-
tion (1908–2008). Reflecting and shaping the world of mathematics education (pp. 131–147).
Istituto della Enciclopedia Italiana: Rome.
Gispert, H. (2003). Applications: Les mathématiques comme discipline de service dans les années
1950–1960 [Applications: Mathematics as a service discipline in the years 1950–1960]. In
D. Coray, F. Furinghetti, H. Gispert, B. R. Hodgson, & G. Schubring (Eds.), One hundred years
of L’Enseignement Mathématique: Moments of mathematics education in the twentieth century
(Monographie 39) (pp. 251–270). Genève: L’Enseignement Mathématique.
Gravemeijer, K. (2007). Emergent modelling as a precursor to mathematical modelling. In
W. Blum, P. Galbraith, H. W. Henn, & M. Niss (Eds.), Modelling and applications in math-
ematics education. The 14th ICMI study (pp. 137–144). New York: Springer.
Hilton, P. J. (1973). Le langage des catégories [the language of categories] (French translation by
J.-C. Matthys). Paris, Lyon: Cédic.
Howson, A. G. (Ed.). (1973). Developments in mathematical education. Proceedings of the second
international congress on mathematical education. Cambridge: Cambridge University Press.
International UNESCO Colloquium. (1968). Modernization of mathematics teaching in European
countries. Bucharest: Éditions Didactiques et Pédagogiques.
Kemeny, J. G. (1964). Which subjects in modern mathematics and which applications in modern
mathematics can find a place in programs of secondary school instruction? L’Enseignement
Mathématique, 10(1–2), 152–176.
La Bastide-van Gemert, S. (2015). All positive action starts with criticism. Hans Freudenthal and
the didactics of mathematics. New York: Springer.
Niss, M. (2008). Perspectives on the balance between applications and modelling and ‘pure’
mathematics in the teaching and learning of mathematics. In M. Menghini, F. Furinghetti,
L. Giacardi, & F. Arzarello (Eds.), The first century of the international commission on math-
ematical instruction (1908–2008). Reflecting and shaping the world of mathematics education
(pp. 69–84). Istituto della Enciclopedia Italiana: Rome.
Niss, M., Blum, W., & Galbraith, P. (2007). Introduction. In W. Blum, P. Galbraith, H. W. Henn,
& M. Niss (Eds.), Modelling and applications in mathematics education. The 14th ICMI study
OECD. (1964). Mathematics to-day. A guide for teachers. Paris: OECD.
OEEC. (1961a). New thinking in school mathematics. Paris: OEEC.
OEEC. (1961b). Synopses for modern secondary school mathematics. Paris: OEEC.
Papy, G. (1963). Mathématique moderne 1 [Modern mathematics 1]. Paris/Bruxelles: Didier.
Piaget, J. (1954). The construction of reality in the child. New York: Basic Books.
Rouche, N. (1984). Bilan de la réforme et idées pour l’avenir [Review of the reform and ideas for
the future]. Bulletin de la Société Mathématique de Belgique (Série A), 36(2), 133–145.
Chapter 36
Interactive Case Practice Teaching
on Mathematical Modelling Course
Xiaojun Duan, Dan Wang, and Mengda Wu
Abstract The design of one kind of interactive case practice teaching in a tertiary
mathematical modelling course is outlined in this chapter. The process of interactive
case creation and problem-solving between students and teacher is constructed with
the aim of promoting students’ creativity, initiative, and interest in learning, as well
as teaching efficiency. The interactive case practice teaching process has four stages:
interactive case creation, subject activity, open learning, and feedback and evalua-
tion. The students in these stages experience the whole teaching process including
problem creation, problem analysis, problem-solving, and knowledge application.
Finally, a case, the sensitive question survey, is used to demonstrate the interactive
case teaching process in practice.
Keywords Interactive case teaching · Mathematical modelling teaching ·

Teaching practice
36.1 Introduction
There is much literature (e.g., Stillman et al. 2017) that involves guiding students in
how to learn to solve practical problems in a mathematical modelling course.
Stillman (2015) notes that with mathematical modelling, the direction reality to
mathematics becomes the focus with the modeller asking: “Where can I find some
mathematics to help me with this problem?” Quite some time ago, Pollak (1997)
introduced a way of solving problems in the real world by mathematical modelling,
while more recently, Wedelin and Adawi (2015) outlined principles for designing
small realistic problems to allow engineering students to practice and apply problem-
solving involving problems of theoretical interest but nontrivial modelling. Hestenes
X. Duan · D. Wang (*) · M. Wu

College of Liberal Arts and Sciences, National University of Defense Technology,
Changsha, China
e-mail: mdwu@nudtwmd.com

420 X. Duan et al.
(2010) believes that the key to developing understanding of a phenomenon by math-

ematical activity is a problem design that facilitates the application of the main
pedagogical approach. In China, as in many other countries, the National Natural
Science Foundation has advocated problem-driven applied research, particularly
encouraging links between industry and mathematics. The intention is to inspire
innovations as well as all-round talents of young people to meet current demands of
the national economy and technological and scientific innovation (Li 2013).
Cumberbatch and Fitt (2001) demonstrate some cases from industry where mathe-
matical modelling could serve these purposes.
In the process of mathematical practice teaching, there are many ways to incor-
porate mathematical modelling into the problem solution to enhance cognitive abil-
ity (e.g. Wedelin and Adawi 2015) and strengthen innovation and improve the
practice ability of students (Wu et al. 2015). During mathematical modelling, mod-
ellers are simplifying and structuring a real-world problem, transferring it into the
world of mathematics, working mathematically, and interpreting a mathematical
result as well as validating a real result (Maaβ 2006). To meet this demand, it is very
important to create a multilevel practice environment (e.g. as described in Duan
et al. [2015]) to cultivate students’ self-learning ability, stimulating their learning
interests first, and then to improve their ability of acquiring new knowledge and
solving practical problems in the real world. Problem-based teaching is a main
method in the mathematical modelling teaching. The importance of cases is widely
supported in the literature on applied problem-solving (Almeida and da Silva 2015;
Jonassen 2011; Wedelin and Adawi 2014).
There are many different views taken on the objectives and the traditions of the
introduction of problems in mathematical classes (e.g. Zawojewski and Lesh 2003).
One kind of interactive case practice teaching and its corresponding teaching dem-
onstration in a mathematical modelling course is outlined and illustrated in this
chapter. The key to this teaching practice is to design a process of interactive case
creation and problem-solving, according to promoting both student initiative and
interest and stimulating their creativity gradually. The interactive case teaching has
four aspects: case creation, subject activity, open learning, and feedback and evalu-
ation. The four aspects are anticipated to attract students by finding a problem from
life via discussion between the teacher and students, exploring the basic model and
knowledge in subject activity, widening application of this knowledge and deepen-
ing the research by a corresponding motivation of self-learning, and ultimately
evaluating and providing feedback. The goal is to enhance student cognitive ability
and strengthen innovation ability (Wu et al. 2015). Finally, an example, the Sensitive
Question Survey, is demonstrated to show the efficiency of interactive case teaching.
36.2 Interactive Case Practice Teaching
A good mathematical modelling case could provide a positive environment for spe-
cial training for students as a small paradigm of scientific research. The case includes
understanding the background of a problem, exploring the essence of that problem,
36 Interactive Case Practice Teaching on Mathematical Modelling Course 421
and improving students’ mathematical application ability, innovation ability, knowl-

edge cognitive level, engineering accomplishment, and so on. In mathematical mod-
elling courses, the cases are usually proposed by teachers, and the students are
required to understand the background of the cases and discuss the problem solution
with their teachers. However, in some cases, students are likely to show a passive
response to the case that is provided and designed only by the teacher because of
lack of interest or experience. In contrast, in this section, we elaborate the design
goal for the case creation and problem-solving to indicate how student interest and
experience could be cultivated in a mathematical modelling course.
36.2.1 The Design Goal of the Interactive Case Creation
36.2.1.1 Demands and Interests-Oriented Case Creation
The teaching case should be attractive. Otherwise, it could not stimulate the enthu-
siasm of the students because the initiative of thinking of students is built on the
basis of rich interest. If we hope to create an environment of self-learning for stu-
dents, we should evoke their learning interests and attract them into the field of posi-
tive learning (Duan et al. 2015). On the other hand, the study is always related with
some special instance, according to Piaget’s constructivism theory (1983) which
looks at learning as two major principles guiding our learning: adaptation and orga-
nization. For individuals to survive in an environment, they must adapt to the stimuli
from the environment. Assimilation and accommodation are both parts of the adap-
tation process. Piaget believed that humans possess knowledge structures that
assimilate external events and convert them to fit their knowledge structures.
Moreover, their knowledge structures accommodate themselves to new, unusual,
and constantly changing aspects of the external environment. Organization refers to
the nature of these adaptive knowledge structures. Piaget (1983) suggests that
knowledge should be organized in complex and integrated ways. So it is very impor-
tant to create a very attractive circumstance related to the learning topic for students
to take advantage of these principles.
Cases that are elaborately designed by teachers will probably not achieve the
anticipated effect. Students might not become enthusiastically involved in the case
creation. Sometimes, the teacher complains that the students did not participate very
actively. If we analyse this situation carefully, one reason is an unimaginative and
uninteresting case could not stimulate student interest. If the problem is raised by
the students and comes from life and their concerns, this case would be more attrac-
tive to them (da Silva et al. 2015). Students will also have opportunities to perceive
the relevance and interrelatedness of what is learnt during problem finding and
knowledge application (Ng and Stillman 2015).
422 X. Duan et al.
36.2.1.2 Students-Involved Interactive Case Design
The case and problem design provided by the students in the first stage are raw in
most instances and need to be investigated and reprocessed. It is not appropriate that
these cases are directly applied in course teaching. Reasons include: (1) the raw
problem includes some ambiguous conception and unclear relationship which may
require enough time to understand and confirm, but course time is limited and the
time length available cannot satisfy it; (2) the complexity of the raw problem may
reduce student enthusiasm on the mathematical modelling; and (3) the case for
teaching purposes is not identical with problem-solving in the real world. The teach-
ing case should be carefully designed to focus on the mathematical modelling, inno-
vation, and knowledge cognition. Therefore, the interactive case design needs to
occur between student and teacher. The students provide the background of the case
and the interesting problem, and then the teachers and students work together to
complete the investigation and reprocessing of the problem. Participation in prob-
lem modification allows students to better focus on issues that really need to be
solved. They realize that the feasibility and computational complexity of the prob-
lem also become a concern. They also find that an appropriate and idealized assump-
tion to the problem does not only change the key model but also helps to better grasp
the main conflict and critical pathway in the problem. The interactive case design is
helpful for students to understand the problem deeply. On the other hand, varied
problems provided by students can broaden their horizons. Variation is essential for
being able to discern critical aspects of a problem (Marton and Trigwell 2000).
Through variation, students get a chance to experience differences and similarities
between problems (Wedelin and Adawi 2015).
36.2.1.3 Case Refinement
For the mathematical modelling course and practice, case teaching is a common
method. However, finding a good case, which is interesting and attractive for stu-
dents, is not an easy task. The problems raised by students can extend the source of
the teaching case. The teacher and students could work together to refine the prob-
lem. The main idea is to simplify real-world problems to make them smaller while
retaining important characteristics such that the solution to the problem is still of
practical or theoretical interest and the problem invokes nontrivial modelling and
problem-solving activities (Wedelin and Adawi 2015). A refined case should have
the following features.
• The real-world problems are often large and complex; if we want to teach stu-
dents to solve real-world problems, we should assume that the real-world prob-
lem will serve as a good exercise problem. An exercise problem is not necessarily
complicated and should retain important characteristics of real-world problems.
It is challenging work, but the simplified problem should keep the essence of the
real-world problem.
• The extra mathematical knowledge background of the problem should be simpli-

fied so that students focus more on the mathematical modelling. In fact, the real-
world problems may be often ill-defined (Jonassen 2011) and difficult for better
understanding by students. The teacher and students should find a relevant point
of view and make reasonable assumptions to reduce this professional background
knowledge. The problem should not require extensive new theory to be learned
before the problem is attempted (Wedelin and Adawi 2015).
• The method of solving the problem should not be unique. The problem should
leave appropriate room so that students can investigate and explore different
ways to solve it. This will help students to innovate and diverge their thinking.
The students may also need to learn and search more theory when they have dif-
ficulties in solving the problem, expecting that new given problems will be
solved with new methods (Wedelin and Adawi 2014). This will also help foster
self-learning ability.
• The problem cannot be too complex to describe in mathematical language, and
the concepts in the problem should be clear and can be expressed quantitatively.
Real-world problems are often qualitatively described by natural language. How
to create a quantitative measure of the problem is an important task in mathemat-
ical modelling. However, some concepts and relationships may be very abstract
and complex in the real-world problem so they cannot be easily described quan-
titatively. These problems may be good research projects in mathematical model-
ling application but not good teaching cases in mathematical modelling teaching.
• The solution should be of practical interest and useful, not just for practice of
application of some mathematical method (Wedelin and Adawi 2015). For the
result of a problem, this is natural not only for the solution of the problem itself
but also how it can be interpreted and contribute to its context. The useful and
interesting results can impress students and make them appreciate the role of
mathematical modelling.
• As a teaching case, the computational complexity is usually not too high. The
teacher could divide the problem into separate sub-problems to reduce complex-
ity, making it possible to produce more suitable tasks. Different tasks have differ-
ent difficulty of calculation so that students can make a choice according to their
own abilities. The computational complexity of the classroom case should be
low, and sub-problems with high computational complexity may be set as an
extension of the problem for students.
• The cases should be extensible, allowing students to discuss more complex mod-
els based on the teaching case. For students, a new problem may be a challenging
task. Extending the teaching case can transition between the teaching case and a
real-world problem. This may prevent new learners of mathematical modelling
from feeling helpless in facing new problems. It also helps build self-confidence
and enhance their interest in learning mathematical modelling.
424 X. Duan et al.
36.2.2 Case-Driven Practice Teaching Process
Case-driven practice teaching has four aspects: case creation, subject activity, open
learning, and feedback and evaluation. We now look at how each is cultivated in turn
to meet the goals of the case.
Case creation is the basis for case-driven practice teaching to some extent. The
created case should appeal to students by the selecting of an interesting subject from
their life via discussion among teachers and students. Personalized design is of ben-
efit to a student’s development. Different cases are designed to improve students’
ability to imagine, learn, and practise mathematical modelling knowledge. Two
ways to create the appealing case for students are from their life via discussion
among the students themselves as mentioned above and to design six to ten related
subjects by teachers. In reality, we found that most students prefer the former, which
is more intriguing for them. Teachers need to investigate the degree of appropriate-
ness of the case with the students, including the professional background, difficulty,
computational complexity, possible mathematical model, and extensibility, and then
give suggestions to them for better fulfilment of the task.
Subject activity is the key to the implementation of the problem-solving. It inte-
grates the modelling content into a series of subject cases by finding the problems,
attempting to collect data, analysing that data, and finding the strategy to solve the
problems. In this aspect, the students become the leading actor in acquiring knowl-
edge and experience, exploring new knowledge and self-learning independently. In
addition, they develop their modelling competency by participating actively in this
process. In this part, students start to finish the modelling task, but this process still
needs teacher guidance. If they encounter difficulties, teachers guide them to find
literature references or give suggestions. The main idea is to guide students to use
learned knowledge freely, experiencing new thinking and exploring new knowl-
edge. Finally, students are expected to improve their self-learning ability.
Open learning is a good way to investigate the study and research in case teach-
ing by combining students’ self-learning, cooperation, and teachers’ guidance. A
good case needs a good way of learning to realize it. The teacher provides the study
resources, and the students discuss among the group. They, thus, acquire new opin-
ions and know how to cooperate with each other (Duan et al. 2015). In the process
of open learning, students present to the class to show their research. All students
have opportunities to raise questions of others, including the teacher and each group
member. The teacher acts as a participant, not just an evaluator.
Feedback and evaluation create a spiral improvement loop in the case-driven
learning pattern. They are important parts of the case-driven practice-learning pat-
tern (Duan et al. 2015). Their purpose is ensuring a gain in learning activity and a
generalization of the practice teaching process. We recommend teachers give posi-
tive evaluation of the product of the students’ practice. This kind of appraisal stimu-
lates student enthusiasm and perseverance. In the feedback and evaluation process,
the teachers play an important role, evaluating students’ work and encouraging stu-
dents to improve their creative processes by using an iterative modelling process.
Thus, this process creates a spiral improvement loop for case-driven learning.
36.3 Example Case: The Sensitive Question Survey
In this section, a case is demonstrated to show how the case-driven practice teaching
unfolds. It involves a topic addressed previously in ICTMA (see Street and Street
[1998]).
36.3.1 Problem Creation
This case is a student activity that is carried out in the class. The goal of the problem
is to obtain the proportion of students answering “Yes” in a sensitive question sur-
vey. In this task, a sensitive question survey is made for college students, for
instance, “Are you in love?” The students are required to answer “Yes” or “No.”
Obviously, if the respondents would honestly answer the question, the investigator
would obtain the true result immediately. But it is actually impossible because the
questions are sensitive. Therefore, the design of the questions is very important
(Street and Street 1998).
A student raises this problem, but other students do not think it is a problem that
needs building a mathematical model to solve. They think that anonymous voting
can solve this problem of respondents not honestly answering the question.
Obviously, if this survey strategy was adopted, this problem will lose its value. The
teacher raises a question: How to design the sensitive question survey if anonymous
voting is not adopted? The students work together with the teacher to discuss this
question. Obviously, directly using a question “Have you ever been in love?” to
survey, the investigator cannot get the expected result. As it is a sensitive question,
it is not easy to get the true proportion of “Yes.” To solve this problem, questions
need to be designed in this survey that students will answer. How could the investi-
gator design questions to obtain a good estimate of the true proportion of “Yes”?
Firstly, the teacher and students discuss and design a strategy for this survey. Two
questions are designed:
Question A: Have you ever been in love? Question B: Have you never been in love?
These two questions are written on different cards called Question-A card and
Question-B card, respectively. Suppose the number of Question-A cards and
Question-B cards are a and b, which are determined by the investigator. All cards
are placed in a black box. The respondents take a card randomly and then answer
“Yes” or “No” for the question on the chosen card. In this process, the investigator
does not know which question the respondents answered; they only elicit the answer
“Yes” or “No.” So we assume the respondents honestly answer the question on their
chosen card. This method is called a stochastic survey.
426 X. Duan et al.
36.3.2 Subject Activities and Open Learning
For convenience, assume the number of respondents (students) is n, let M be the

number of students that have ever been in love, and n-M denotes the rest of the stu-
dents. Then the proportion of students who have ever been in love is M/n which is
denoted by x. The aim of the sensitive question survey is to estimate x.
Subject activity requires the students to find the answer to this problem via dis-
cussion among them. Let u = a/(a+b). The probabilities that the respondents took
Question-A card and Question-B card are u and 1-u, respectively. Furthermore, the
probability that the respondents answer “Yes” is denoted by
p = xu + (1 − x ) (1 − u ) .
where the students answering “Yes” include two kinds:

• The students who have ever been in love take a Question-A card.
• The students who have never been in love take a Question-B card.
Counting the number of “Yes” answers denoted by Y and using Y/n to approximate
the probability p, then
Y / n ≈ xu + (1 − x ) (1 − u ) .
Finally, the approximate probability x1 can be evaluated by
x1 = ( u − 1 + Y / n ) / ( 2u − 1) .
When u = 1 or 0, we could get the exact result about the proportion of students who
have ever been in love, but it is not realistic because the respondents may not hon-
estly answer the question in these sensitive cases. When u = 1/2, the respondents
would not feel anxious about sensitive information, but the investigators will not get
any information from this formula because the denominator is zero. This is a
dilemma, so we may take u = 1/3 or 1/4 in general. This modelling process leaves an
extensible space for students.
In order to overcome the problem that u cannot take 1/2, some students think the
second question (Question B) does not have to be related to the first question
(Question A). They discussed and designed another group of questions that is called
an unrelated survey.
Question A: Have you ever been in love?
Question B: Is your student identifier an even number?
The strategy of this survey is the same as the previous one. Let u = a/(a + b). The
probabilities that the respondents will take Question-A card and Question-B card
are u and 1 – u, respectively. Then the probability that the respondents answered
“Yes” can be expressed by
p = xu + (1 − u ) / 2.
Counting the number of “Yes” answers denoted by Y and using Y/n to approximate
the probability p, then
Y / n ≈ xu + (1 − u ) / 2,
Further,
x 2 = (Y / n − (1 − u ) / 2 ) / u.
This formula is better than that in the previous model. When u = 1/2, we can get
(
x 2 = 2 Y / n −1 / 4 . )
This method is much easier to acquire the trust of the respondents than the
first survey.
As an extension for research, students are required to discuss the accuracy of the
two methods and compare their advantages and disadvantages by a computer simu-
lation method.
36.3.3 Feedback and Evaluation
This case is used in the class of second year college students. In the problem cre-
ation phase, students have a keen interest in raising problems, but their cases lack an
elaborated design for such a problem. They expect this to be appreciated by the
teacher and are willing to refine the problem with the teacher. They ask various
problems: “What is the optimal ratio of a and b?” “Will the respondents be sure to
answer the question truthfully?” “Is their degree of trust 100%?” “Does this prob-
lem need to use a mathematical model?” “What model should be used?” The teacher
discusses these questions with students, which can deepen students’ understanding
of the problem. In the subject activity and open learning phase, the questions raised
by students become the best guide for mathematical modelling. However, students
do not give best advice on how to answer these questions, even though they have an
enthusiastic discussion. But when the teacher hints that the focus of this problem is
how to calculate the probability of answering “Yes”, students can build a mathemat-
ical model to solve it immediately and discuss the possible ratio of a and b. From
this activity, we find that students have good mathematical skills but lack of model-
ling competence, including holistic (requiring full-scale modelling) and atomistic
(concentrating on mathematisation and analysis of models). In the teaching prac-
tice, students experienced the entire process of problem creation and resolution.
Compared with solving the teachers’ problems, students have greater enthusiasm
for solving their own problems.
428 X. Duan et al.
36.4 Conclusions
One kind of interactive case practice teaching for a mathematical modelling course
is discussed here. The key to this practice teaching is integrating case-driven prac-
tice into the course to improve the comprehension and stimulate the creativity of
students progressively. The interactive case practice teaching has four sequential
aspects: case creation, subject activity, open learning, and feedback and evaluation.
The four aspects correspond to how it is intended to attract students and train them
by designing a case from their life via discussion between teacher and students,
integrating basic knowledge and topic training into subject activity, and synthesiz-
ing flexible application with motivated self-learning and feedback and improve-
ment. Through years of practice, a lot of achievements have been made in case-driven
practice activities. It is very effective, and it is a good form to improve the modelling
ability of students. We intend to pursue further research to improve the mechanism.
Acknowledgments This work is supported by the Course Construction Program and the
Innovation Practice and Internship Base for Mathematical Modelling of NUDT, China.
References
Almeida, L. M. W., & da Silva, K. A. P. (2015). The meaning of the problem in a mathematical
modelling activity. In G. A. Stillman, W. Blum, & M. S. Biembengut (Eds.), Mathematical
modelling in educational research and practice: Cultural, social and cognitive influences
Cumberbatch, E., & Fitt, A. (2001). Mathematical modelling: Case studies from industry.
Cambridge: Cambridge University Press.
da Silva, J. N. D., Santana, T., & Carneiro, C. H. (2015). Contributions of mathematical modelling
in education of youth and adults. In G. A. Stillman, W. Blum, & M. S. Biembengut (Eds.),
Mathematical modelling in educational research and practice: Cultural, social and cognitive
influences (pp. 557–555). Cham: Springer.
Duan, X., Wang, D., & Wu, M. (2015). Research-based modelling teaching activities: a case of
mathematical positioning with GNSS. In G. A. Stillman, W. Blum, & M. S. Biembengut (Eds.),
Hestenes, D. (2010). Modelling theory for math and science education. In R. Lesh, P. L. Galbraith,
C. R. Haines, & A. Hurford (Eds.), Modelling students’ mathematical modelling competencies
Jonassen, D. (2011). Learning to solve problems. New York: Routledge.
Li, T. (2013). Mathematical modelling education is the most important educational interface
between mathematics and industry. In A. Damlamian, J. F. Rodrigues, & R. Sträßer (Eds.),
Educational interfaces between mathematics and industry (pp. 51–58). Cham: Springer.
Maaβ, K. (2006). What are modelling competencies? ZDM Mathematics Education, 38(2),
113–142.
Marton, F., & Trigwell, K. (2000). Variatio est mater studiorum. Higher Education Research and
Development, 19(3), 381–395.
Ng, K. E. D., & Stillman, G. A. (2015). Exporing interconnections between real-world and appli-
cation tasks: case study from Singapore. In G. A. Stillman, W. Blum, & M. S. Biembengut
(Eds.), Mathematical modelling in educational research and practice: Cultural, social and
cognitive influences (pp. 207–217). Cham: Springer.
Piaget, J. (1983). Piaget’s theory. In P. Mussen (Ed.), Handbook of child psychology (Vol. 1, 4th
ed.). New York: Wiley.
Pollak, H. O. (1997). Solving problems in the real world. In L. A. Steen (Ed.), Why numbers count:
Quantitative literacy for tomorrow’s America (pp. 91–105). New York: College Entrance
Examination Board.
Stillman, G. A. (2015). Applications and modelling research in secondary classrooms: What have
we learnt. In S. J. Cho (Ed.), Selected regular lectures from the 12th International Congress on
Mathematics Education (pp. 791–806). Cham: Springer.
Stillman, G. A., Blum, W., & Kaiser, G. (Eds.). (2017). Mathematical modelling and applications:
Crossing and researching boundaries in mathematics education. Cham: Springer.
Street, A. P., & Street, D. J. (1998). Discrete approaches to mathematical modelling. In P. Galbraith,
W. Blum, G. Booker, & I. D. Huntley (Eds.), Mathematical modelling: Teaching and assess-
ment in a technology-rich world (pp. 207–219). Horwood: Chichester.
Wedelin, D., & Adawi, T. (2014). Teaching mathematical modelling and problem solving: A cogni-
tive apprenticeship approach to mathematics and engineering education. International Journal
of Engineering Pedagogy, 4(5), 49–55.
Wedelin, D., & Adawi, T. (2015). Applied mathematical problem solving: Principles for design-
ing small realistic problems. In G. A. Stillman, W. Blum, & M. S. Biembengut (Eds.),
Wu, M., Wang, D., & Duan, X. (2015). The teaching goal and oriented learning of mathematical
modelling courses. In G. A. Stillman, W. Blum, & M. S. Biembengut (Eds.), Mathematical
modelling in educational research and practice: Cultural, social and cognitive influences
Zawojewski, J. S., & Lesh, R. (2003). A models and modelling perspective on problem solving.
In R. Lesh & H. M. Doerr (Eds.), Beyond constructivism: Models and modelling perspectives
on mathematics problem solving, learning, and teaching (pp. 317–336). Mahwah: Erlbaum.
Chapter 37
Pre-service Teachers’ Sense-making
of Mathematical Modelling through
a Design-Based Research Strategy
Rina Durandt and Geoffrey V. Lautenbach
Abstract This chapter reports on the sense-making of pre-service teachers in and

about mathematical modelling by means of a well-planned design-based research
strategy over a period of two iterations. Modelling, and the teaching thereof, is
mostly difficult for teachers, but it can be learnt through the development of math-
ematical, modelling and facilitator competencies of pre-service teachers during
their formal education. This study explores how pre-service teachers translate the
modelling cycle and develop certain competencies over time through partaking in a
series of modelling activities as modellers themselves and as teachers of modelling.
Qualitative findings reveal how pre-service teachers improve in their sense-making
of the mathematical modelling process, but they disclose some shortcomings that
relate to preparing for the teaching of modelling.
Keywords Design-based research · Mathematical modelling · Modelling

competencies · Pre-service mathematics teachers · Sense-making · Teaching
modelling
R. Durandt (*)
Department of Applied Physics and Engineering Mathematics, Faculty of Science,
University of Johannesburg, Johannesburg, South Africa
e-mail: rdurandt@uj.ac.za
G. V. Lautenbach
Department of Science and Technology Education, Faculty of Education,
University of Johannesburg, Johannesburg, South Africa
e-mail: geoffl@uj.ac.za

432 R. Durandt and G. V. Lautenbach
37.1 Context and Purpose
Traditional problem-solving, applications and mathematical modelling are con-

nected, although not similar, and all contribute towards authentic learning experi-
ences. The Guidelines for Assessment and Instruction in Mathematical Modelling
Education report (COMAP-SIAM 2016) explains how a mathematics problem situ-
ation can be transformed into a word problem (requiring traditional problem-
solving) by adding labels, further transformed to an application problem (requiring
mathematical applications) by adding meaning in terms of context and finally the
application problem is transformed into a modelling problem (requiring the com-
plete modelling process) by adding interpretation. In South Africa, traditional
problem-solving has been an intrinsic focus of the school mathematics curriculum
for many years, and mathematical modelling only gained prominence later in 2011.
Subsequently, teachers are now expected to take part in the teaching and learning of
modelling activities. An overview of the literature confirms that teachers experience
modelling as difficult, they feel unprepared for teaching modelling, and they may
not be able to appreciate the value of mathematical modelling if they are not exposed
to such activities during their formal education (Jacobs and Durandt 2017; Ng 2013;
Stillman et al. 2007). On the other hand, Blum and Borromeo Ferri (2009) are con-
fident that mathematical modelling can be learnt and taught.
The decade of research on problem-solving in mathematics and decision-making
by teachers in mathematics classrooms by Schoenfeld (2014) that depends largely
on teachers’ resources (knowledge and competencies), orientations (mostly beliefs)
and goals (usually informed by resources and orientations) has led to the systematic
research process implemented in this inquiry. In this inquiry we purposefully con-
sidered a systematic design-based research (Anderson and Shattuck 2012) process
(over two iterations) containing iterative cycles of design, implementation and anal-
ysis to explore pre-service teachers’ sense-making of mathematical modelling as
they participated as modellers themselves and teachers of modelling. For this pur-
pose two research questions guided this research study: (1) how do pre-service
teachers develop in translating the modelling cycle after exposure to modelling
activities (as modellers and teachers of modelling), and (2) how do pre-service
teachers progress in mathematical, modelling and facilitator competencies through
partaking in a series of modelling activities?
This study forms part of a broader project that distils curricular guidelines aimed
at the adequate preparation of pre-service mathematics teachers for mathematical
modelling during their formal coursework. Some findings from this broader inquiry
have already been reported elsewhere (Jacobs and Durandt 2017), and together with
the findings reported in this chapter, we can start to make sense of how pre-service
teachers come to grips with mathematical modelling as content, practice and peda-
gogical content knowledge (PCK) as they develop competencies through the care-
fully designed interventions.
37 Pre-service Teachers’ Sense-making of Mathematical Modelling through… 433
37.2 Literature Perspectives
37.2.1 Suitable Framework for Pre-service Teacher

A
Preparation of Modelling
The chapter authors are of the opinion that the formal preparation of pre-service
teachers for teaching and learning of mathematical modelling is informed by the
notion of Teaching for Robust Understanding of Mathematics rubric by Schoenfeld
(2014), as well as Bruner’s 1973 Learning to be philosophy (as cited in Amory et al.
2008). Both of these theoretical perspectives require the pre-service teacher to
become a practitioner of knowledge within a professional domain. In this study,
they take on the role of both modeller and a teacher of modelling.
Schoenfeld (2014, p. 407) describes five dimensions of a powerful mathematics
classroom that links with the modelling context by focusing on the (1) “mathemat-
ics”, (2) “cognitive demand”, (3) “access to mathematical content”, (4) “agency,
authority, and identity”, and (5) “uses of assessment.” One of the aims of Teaching
for Robust Understanding of Mathematics rubric is to work productively with
teachers and not to evaluate their performance. In this study, these five dimensions
were considered in planning the series of modelling activities as teachers partici-
pated via different roles as learner, beginning modeller and teacher of modelling.
The dimensions were also meant to prepare student teachers for a complex world
that will enable them to act purposefully in difficult real-life situations.
Bruner (as cited in Amory et al. 2008) places emphasis on learning about (which
involves the learning of the different elements, including facts, concepts and proce-
dures, of the mathematical modelling process) and learning the practices of the
knowledge domain (which involves the learning of the entire modelling process
leading to an interpretation of results to arrive at a suitable solution). Learning,
according to Bruner, requires the application of the conceptual framework and theo-
ries of the domain, giving rise to practices of inquiry, interpretation and eventually
problem-solving. In this study, we purposefully exposed pre-service teachers,
through a series of modelling activities, to learning about mathematical modelling
as well as to learning the practices of modelling.
37.2.2 spects of the Teaching and Learning of Mathematical

A
Modelling
This study relates to the development of sense-making of mathematical modelling

as pre-service teachers engage in a series of interventions. Central to the discussion
are (1) a suitable modelling cycle, (2) the modelling task and (3) the development of
essential competencies. First, a cycle that was previously implemented in Singapore
with some success was used in this study (Balakrishnan et al. 2010). The modelling
process, cyclic in nature, includes four elementary elements. These four elements
include representing a real-world problem mathematically, using appropriate math-
ematics to solve the problem, making sense of the solution and reflecting on the
assumptions and limitations. Second, central to the modelling experience is the
modelling task to guide the design of a mathematical modelling learning experi-
ence. In this study, the framework by Ang that appeared in Tan and Ang (2012)
informed the design of the series of modelling activities by addressing the following
five questions for planning or designing mathematical modelling experiences:
WHICH level of learning experience?
WHAT is the skill or competency?
WHERE is the mathematics?
HOW to solve the problem or model?
WHY is this experience a success? (p. 715)
Third, Niss et al. (2007, p. 12) define modelling competency as “the ability to iden-
tify relevant questions, variables, relations or assumptions in a given real world situ-
ation, to translate these into mathematics and to interpret and validate the solution
of the resulting mathematical problem in relation to the given situation”. This defi-
nition aligns with the cycle from Balakrishnan et al. (2010) and highlights crucial
competencies needed to proceed through a modelling cycle. Stillman et al. (2007)
identify remarkable mathematical accomplishments by secondary school students
moving through the modelling cycle, but also numerous challenges in transitions
between the different elements of the cycle.
Research results published by the Centre for Development and Enterprise
(Bernstein 2013) mention the vast need for competency development in South African
mathematics teachers. In this study, we differentiate between pre-service teachers’
mathematical and modelling competencies based on research findings from Bernstein
(2013) and guidelines from Stillman et al. (2007). We also look at their abilities to
use, apply and validate mathematical models based on Meyer’s (2012) criteria, as
well as at facilitator competencies as participants took on the role of teachers of mod-
elling during reflective activities (based on criteria from Hess et al. 2009).
37.3 Research Design
37.3.1 Research Paradigm
The authors argue from a socially constructed knowledge viewpoint (Creswell

2013) relying as much as possible on the participants’ interpretations, mathematical
workings and reflections as they proceed through the modelling cycle. The overall
intention was to comprehend pre-service teachers’ sense-making of mathematical
modelling and to follow the development of their competencies as they participated
as modellers and teachers.
37.3.2 he Design-Based Research Strategy and Data

T
Collection
This study, with a focus to develop pre-service teachers’ sense-making of mathe-

matical modelling, forms part of a broader study conducted over three design-based
research phases (see Table 37.1). Anderson and Shattuck (2012, p. 2) view design-
based research as a “methodology designed by and for educators that seeks to
increase the impact, transfer, and translation of education research into improved
practice”.
Qualitative data were collected in iteration 1 as participants attempted the World
Cup Rugby 2015 task (designed by the authors) and in iteration 2 session 2 as par-
ticipants attempted the Traffic Flow task (adapted from Bester et al. 1998). Both
these modelling tasks are non-routine and open-ended, requiring a substantial mod-
elling demand (Tan and Ang 2012). Data were also collected in iteration 2 session 3
as participants took on the role of teachers of modelling, while they reflected on
their proposed solutions of the Traffic Flow task). Forty-nine participants (all Grade
10–12 mathematics student teachers with a mean age 22.4 years) took part in the
project in 2015, and they were expected to present their work on a predesigned
worksheet as they moved through the modelling cycle (Balakrishnan et al. 2010)
and on a poster as they reflected on and presented their proposed solution. The same
participants were involved in both iterations. They were divided into ten groups,
stratified by mathematical ability.
World Cup Rugby 2015 Task

The 2015 Rugby World Cup Tournament (RWT) will be hosted by England
from 18 September to 31 October 2015 and the final will be played at the
well-known Twickenham Stadium.
The official rugby ball is made of the following features: three-ply backing
material for good shape retention, standard grip, grippy rubber surface, hand
stitched and synthetic latex bladder for excellent air retention. The design also
features the official Tournament social media hashtag, and the ball comes
inflated for safeguarding a good smell and shape (approximate weight 460 g &
pressure 10PSI). Tournament rules stipulate each of the 20 teams should sup-
ply a sufficient number of the official rugby balls for team practices and prep-
aration. Tournament officials are responsible to supply the match ball only.
The South African Rugby Union (SARU) would like to receive a recommen-
dation on the maximum number of official rugby balls that can be transported
to England for the RWT 2015 on a South African Airways (SAA) flight.
Provide the SARU with a plan on how they can perform relevant calculations.
You need to explain the method you used as the SARU would like to apply
this method to other areas.
(continued)
Data collected from the SARU stipulates the official match ball is oval- or
egg-shaped and made from four different panels. For a size 5 rugby ball (rel-
evant for full-sized rugby), the length should be approximately 30 cm, the
length circumference approximately 77 cm and the width circumference
roughly 62 cm. In-flight safety regulations by SAA stipulates the weight per
new ball (460 g) and optimal ball pressure of approximately 10PSI can be
allowed (standard ball size). Luggage capacity is restricted to a maximum of
1 cubic metre per flight.
Traffic Flow Task

The Department of Town and Regional Planning would like to receive a rec-
ommendation on the best location for a Day Care Centre for toddlers. Provide
the Department with a plan on how they can select the best location. You need
to explain the method you used as the Department would like to apply this
method to other areas.
Data collected from the local Traffic Department gives information on a
section of the city’s street network (one-way streets and how many cars enter
or leave this section of the city via the indicated street in a certain 1-h period):
Both First and Second Street are one-way streets from north to south. Third
Avenue is a one-way street from west to east, but Fourth Avenue is a one-way
street from east to west. Street corners are identified at Third Avenue and First
Street, Third Avenue and Second Street, First Street and Fourth Avenue and
Second Street and Fourth Avenue. 180 cars enter First Street. 70 cars enter
Second Street. 200 cars leave First Street. 30 cars leave Second Street. 200
cars enter Third Avenue. 400 cars leave Fourth Avenue. 20 cars leave Third
Avenue. 200 cars enter Fourth Avenue.
Table 37.1 Sequence of modelling activities during the DBR phases of the overarching study
DBR phase 3
DBR phase 1 DBR phase 2 Focus on reflective
Focus on Focus on design development and analysis and
needs analysis implementation evaluation
Pilot study Iteration 1 Iteration 2
Participants’ Modellers Modellers Session 1: Modellers
role Session 2: Modellers
Session 3: Modellers
and teachers of
modelling
Session 4: Teachers of
modelling
37.3.3 ata Analysis Procedures and Trustworthiness

D
Measures
The qualitative data were analysed via the directed content analysis method. Hsieh
and Shannon (2005, p. 1278) define this method as “a research method for the sub-
jective interpretation of the content of text data through the systematic classification
process of coding and identifying themes or patterns”. The analysis was informed
by coding categories derived from the literature (Meyer 2012; Stillman et al. 2007;
Tan and Ang 2012) and grouped as modelling competence, mathematical compe-
tence, model ability and reflective ability. The authors coded the data and an exter-
nal subject specialist confirmed the analysis. Furthermore, we purposefully
addressed the trustworthiness of findings (Creswell, 2013) through a thick descrip-
tion of the methodology in the formal report of the overarching research project, the
triangulation of data collection and analysis methods (more than one modelling task
and other data collection instruments in the broader study) and the reflective and
iterative process of design-based research.
37.4 Findings
Tables 37.2 and 37.3 display the results of the analysis of mathematical and model-
ling competencies (based on guidelines from Stillman et al. 2007). The focus is on
how pre-service teachers’ modelling competencies and mathematical competencies
evolved as they took on the role of modellers. More challenges were evident in itera-
tion 1 than in iteration 2 where more groups are seen taking up the various compe-
tencies, perhaps showing the evolving understanding of the modelling cycle in this
study and improving competencies.
Table 37.2 Results displaying modelling competencies over two iterations

Number of groups
Iteration Iteration
1 2
Moving successfully through the modelling cycle (elements 1–4) 9 10
Identifying from the available information what is relevant and what is 9 10
irrelevant
Making simplified and relevant assumptions about the situation to enable 10 10
mathematics to be applied
Demonstrating the need to acquire new mathematical knowledge or data 5 8
or turn to other information sources (such as technology or textbooks)
Choosing appropriate methods to confirm calculations and for checking 1 1
and testing the model
Linking and/or critically checking mathematical results with the 8 10
real-world situation
Considering implications of decisions and results 5 9
Table 37.3 Results displaying mathematical competencies over two iterations

Number of groups
Iteration Iteration
1 2
Recognizing relevant variables 9 10
Accurately representing the real-world problem mathematically 7 10
Selecting appropriate mathematical formula consistent with the 6 9
representation
Applying mathematical content knowledge correctly 6 7
Investigating more than one mathematical procedure to solve the 3 1
problem
Table 37.4 Results displaying modelling ability over two iterations

Number of groups
Iteration Iteration
1 2
Accurate (model output expected to be near to correct) 2 8
Realistic (model based on original correct assumptions) 6 9
Precise (model output values should be definite mathematical quantities 8 7
or falling into a closed interval)
Practical (sensible solutions provided) 3 8
Robust (relatively safe to errors) 1 9
Sensible suggestions for improvement of the model 0 10
General applicability providing useful conclusions 6 10
Sensible suggestions on the applicability in other contexts 0 8
Table 37.4 displays the results of the analysis of group worksheet documents
with a focus on how pre-service teachers, as modellers, developed with relation to
their sense-making of mathematical models (based on guidelines by Meyer 2012
and the elements of the modelling cycle by Balakrishnan et al. 2010). More positive
results were displayed in iteration 2 than in iteration 1, indicating that participants
improved over time regarding sense-making of mathematical models.
Two criteria from Hess et al. (2009) were considered in the analysis of group
posters as pre-service teachers took on the role of teachers of modelling. The authors
view both these criteria, (1) overall appearance and (2) the display of coherent infor-
mation, as part of the professional development of teachers where they will one day
have to facilitate modelling in their classrooms. These criteria also align with both
the Learning to be and Teaching for Robust Understanding of Mathematics philoso-
phies. All groups created, to a greater or lesser extent, an appealing environment
through their particular display, which was intended to stimulate mathematical dia-
logue (Ng 2013). In most posters, key aspects of a high-quality presentation were
lacking. Three of the group posters were regarded as high quality, three of medium
quality, and three of lower quality (one group did not submit a poster). For example,
Fig. 37.1 Poster from Group 1 in iteration 2 suggesting a real-world solution for the Traffic Flow
task
Fig. 37.1 displays the poster from Group 1 (rated as of medium quality), clearly
indicating a solution for the Traffic Flow problem, but where information on their
deliberations is lacking. As future teachers this lack of documentation of thinking
and reasoning highlights a need for further growth as teachers of modelling becom-
ing role models demonstrating good communication skills.
37.5 Discussion
Findings related to the category modelling competence show how participants

mapped their way through the modelling cycle, gave meaning to the real-world
problem by enabling mathematics to be applied and provided a well-considered
real-world solution to the problem through interpretation of the mathematical
results. During the first iteration, some groups experienced blockages at the begin-
ning of the modelling cycle, similar to other studies (Ng 2013; Stillman et al. 2007),
but in iteration 2 all groups displayed improved modelling competencies by suc-
cessfully moving through the cycle. In both iterations, participants struggled to
choose appropriate methods to confirm calculations.
Findings related to the category mathematical competence explain the groups’
mathematical accuracy in their representations of the real-world problem, their
selection of applicable mathematical content knowledge and the efficiency of their
mathematical procedures. More groups demonstrated improved mathematical com-
petence in the later iteration in the context of the modelling examples, although in
both iterations some shortcomings could by identified especially in selecting more
than one mathematical procedure to perform calculations.
In relation to the category, model ability, participants showed progression in their
sense-making of mathematical models over the two iterations, but lacked the ability
to provide sensible suggestions for improving their mathematical models used in
these modelling tasks. They were also unable to argue for or against the applicability
of models in other contexts. Related to the category reflective ability, the participants
demonstrated an appealing surrounding (or atmosphere) important for a modelling
climate, although the posters from some groups showed shortcomings with respect
to role modelling good communication and documentation of thinking.
37.6 Conclusion
In this study, we purposefully considered a systematic design-based research pro-

cess (over two iterations) containing iterative cycles of design, implementation and
analysis to explore pre-service teachers’ sense-making of mathematical modelling
as they participated as both modellers themselves and teachers of modelling. With
the study embedded in the Teaching for Robust Understanding of Mathematics and
Learning to be philosophies (Schoenfeld 2014; Bruner as cited in Amory et al.
2008), the authors attempted to contribute towards the adequate professional devel-
opment of pre-service teachers in relation to mathematical modelling. The particu-
lar focuses of this inquiry were how pre-service teachers evolve in translating the
modelling cycle, that is, make sense of it, after exposure to modelling activities (as
modellers and teachers of modelling) and how pre-service teachers progress in
mathematical modelling and facilitator competencies, over two iterations of the
design-based research process, by engaging with a series of modelling activities.
Qualitative findings revealed that pre-service teachers’ mathematical and model-
ling competencies showed progression over the course of the two iterations in the
context of this study. Furthermore, participants developed an understanding of the
modelling cycle and how modelling is connected to everyday life. Some negative
trends were detected as participants examined assumptions made, searching for
alternative mathematics to verify properties or checking results and discussing the
applicability of mathematical models. Although these pre-service teachers managed
fairly well to create an inviting modelling surrounding with their poster presenta-
tions, they would require more guidance on the overall appearance and display to
improve coherence, documentation and communication of thinking. Overall, in the
context of this inquiry, pre-service teachers claimed to have improved their concep-
tualization of mathematical modelling. Teacher preparation for mathematical mod-
elling is multidimensional, and the results from this study lead towards further
endeavour in incorporating that multidimensionality explicitly into teacher prepara-
tion courses.
Acknowledgement This work is based on the research partially supported by the National
Research Foundation (NRF) of South Africa, Unique Grant No. 106978.
References
Amory, A., Gravett, S., & Van der Westhuizen, D. (2008). Teaching and learning at the University
of Johannesburg: A position paper. Document SEN, 252(2008), 5.
Anderson, T., & Shattuck, J. (2012). Design-based research a decade of progress in education
research? Educational Researcher, 41(1), 16–25.
Balakrishnan, G., Yen, Y. P., & Goh, L. E. E. (2010). Mathematical modelling in the Singapore sec-
ondary school mathematics curriculum. In B. Kaur & J. Dindyal (Eds.), Mathematical applica-
tions and modelling (pp. 237–257). Singapore: World Scientific.
Bernstein, A (Ed.) (2013). Mathematics outcomes in South African Schools. What are the facts?
What should be done? The Centre for Development and enterprise. South Africa Available from:
http://www.cde.org.za/wp-content/uploads/2013/10/MATHEMATICS%20OUTCOMES%20
IN%20SOUTH%20AFRICAN%20SCHOOLS.pdf
Bester, E. A., Ham, J., Loots, K., & Stark, A. (1998). Study and master mathematics grade 11 &
12 (HG). Goodwood: National Book Printers.
COMAP-SIAM. (2016). Guidelines for assessment & instruction in mathematical modeling edu-
cation (GAIMME). Bedford: COMAP & Philadelphia: SIAM. Available from: http://www.
siam.org/reports/gaimme.php
Creswell, J. W. (2013). Research design: Qualitative, quantitative, and mixed methods approaches.
Hess, G. R., Tosney, K. W., & Liegel, L. H. (2009). Creating effective poster presentations: AMEE
Guide no. 40. Medical Teacher, 31(4), 319–321.
Hsieh, H. F., & Shannon, S. E. (2005). Three approaches to qualitative content analysis. Qualitative
Jacobs, G. J., & Durandt, R. (2017). Attitudes of pre-service mathematics teachers towards mod-
elling: A South-African inquiry. EURASIA Journal of Mathematics Science and Technology
Education, 13(1), 61–84.
Meyer, W. J. (2012). Concepts of mathematical modeling. Mineola: Dover.
Ng, K. E. D. (2013). Teacher readiness in mathematical modelling are there differences between
pre-service and in-service teachers? In G. A. Stillman, G. Kaiser, W. Blum, & J. P. Brown
& M. Niss (Eds.), Modelling and applications in mathematics education (pp. 3–32). New York:
Springer.
Schoenfeld, A. H. (2014). What makes for powerful classrooms, and how can we support teach-
ers in creating them? A story of research and practice, productively intertwined. Educational
Researcher, 43(8), 404–412.
Stillman, G., Galbraith, P., Brown, J., & Edwards, I. (2007). A framework for success in imple-
menting mathematical modelling in the secondary classroom. In J. Watson & K. Beswick
(Eds.), Proceedings of MERGA 30 (Vol. 2, pp. 688–697). Hobart: MERGA.
Tan, L. S., & Ang, K. C. (2012). Pedagogical content knowledge in mathematical modelling
instruction. In J. Dindyal, L. P. Cheng, & S. F. Ng (Eds.), Proceedings of MERGA 35 (pp. 712–
719). Singapore: MERGA.
Chapter 38
Mathematical Modelling: A ‘Growing
Tree’ for Creative and Flexible Thinking
in Pre-service Mathematics Teachers
Rajendran Govender
Abstract This qualitative case study was guided by the research question: To what
extent are pre-service mathematics teachers competent in invoking and using mod-
elling processes to solve a real practical problem? They were required to determine
the height of a tree. In-depth analysis of the written reports and presentations of two
groups showed they were not only able to experience the modelling process but also
able to creatively invoke and apply necessary skills and abilities to enable them to
move from one step to the next. They were able to make assumptions, set up a math-
ematical model through mathematizing, use their mathematical knowledge to solve
the model and interpret and critically reflect on their solutions to see they made
sense. The kinds of novel models built and used in creating unexpected solutions
suggest the tree modelling task stimulated creative and flexible thinking.
Keywords Mathematical creativity · Creative and flexible thinking · Modelling

process · Mathematical modelling competency · Pre-service mathematics teachers
38.1 Introduction
Poincaré (1948) describes mathematical creativity simply as discernment or choice.

This suggests that mathematical creativity is the ability of an individual to analyse
a given problem from different perspectives, generate multiple unusual ideas and
ultimately produce insightful and novel solutions, despite the level of complexity
(Sriraman 2004). The inherent ability to generate a variety of ideas and unexpected
R. Govender (*)
e-mail: rgovender@uwc.ac.za

444 R. Govender
ones from different perspectives defines flexible thinking (Chamberlin and Moon
2005). In this study creative and flexible thinking is considered as an ability to think
in novel ways and create unexpected solutions to a real problem using mathematical
modelling.
According to Oberholzer (1992), modelling is the process of taking any problem
for which the modeller does not have a ready-made method or clear-cut algorithm
that can or may be used to obtain an answer and then building towards a plausible
answer acceptable to the community of practice in a systematic manner. Thus, we
see mathematical modelling playing a vital role in developing students to think
creatively and flexibly in their endeavour to solve real-world problems in unfamiliar
situations (Doerr and English 2003). Mousoulides et al. (2010) assert that modelling
activities set within authentic contexts provide opportunities for students to stretch
their minds as they attempt to solve challenging problems in multiple creative ways.
As teaching is an iterative process, it is imperative that pre-service mathematics
teachers are engaged with typical mathematical modelling tasks and processes so
that they can gain a deeper understanding of the modelling processes and competen-
cies (Schorr and Koellner-Clark 2003). Such experiences can enable pre-service
mathematics teachers to gain the necessary insight and modelling experiences,
which they can later pursue in their own classrooms (NCTM 2000). Hence, this
chapter focuses on pre-service mathematics’ modelling processes and mathematical
modelling competency as they worked on a Tree Problem. It is hoped that this could
provide information that could elucidate the modelling processes and the range of
mathematical competencies that one ought to expose pre-service mathematics
teachers to. More importantly, the study aims to determine the extent to which pre-
service mathematics teachers are competent in invoking and using the mathematical
processes to solve real practical problems.
38.2 The Theoretical Framework
As shown in Fig. 38.1, the process of mathematical modelling is complex. This begins
with identifying the problem inclusive of conditions and constraints and then simpli-
fying, structuring and idealizing the problem to determine a real model follow. Next,
the task solver engages in mathematically interpreting the real model and transform-
ing it into a mathematical model. Invoking and using relevant mathematical tech-
niques and theories to develop a solution occur next. Finally, the task solver engages
in reflecting and checking if the solution makes sense in the context of the problem
and reviewing or refining the mathematical techniques employed to obtain a more
reasonable solution to the problem if necessary (Maaß 2006; Swetz 1991).
Mathematical modelling competency is defined as one’s autonomous insightful
readiness to manoeuvre across all phases of the modelling process in a given prob-
lematized context (Blomhøj and Jensen 2003; Jensen 2007). Blum and Kaiser
(1997, p. 9 as cited in Maaß 2006, pp. 116–117) provide the following detailed list
of sub-competencies (A–E) that permeate the modelling process:
38 Mathematical Modelling: A ‘Growing Tree’ for Creative and Flexible Thinking… 445
B: Set up
Mathematical Model
Real Model
A: Understand Mathematizing
C: Solve
mathematical
Real life practical question
Problem
E: Validate
Interpreted
Solution Mathematical
D: Interpret Solution
Reality Mathematics
Fig. 38.1 Modelling process. (Adapted from Maaß 2006, p. 115 and Niss 2010, p. 44)
A. Competencies to understand the real problem and to set up a model based on reality:
Competency
• To make assumptions for the problem and simplify the situation.
• To recognize quantities that influence the situation, name them and identify the key
variables.
• To construct relationships between variables.
• To look for available information and to differentiate between relevant and irrelevant
information.
B. Competencies to set up a mathematical model from reality: Competency
• To mathematize relevant quantities and their relations.
• To simplify relevant quantities and their relations if necessary and to reduce their
number and complexity.
• To choose appropriate mathematical notations and to represent situations graphically.
C. Competencies to solve mathematical questions within this mathematical model:
Competency
• To use heuristic strategies such as division of the problem into part problems, estab-
lishing relations to similar or analog problem, viewing the problem in a different
form, varying the quantities or the available data, etc.
• To use mathematical knowledge to solve the problem.
D. Competencies to interpret mathematical results in a real situation: Competency
• To interpret mathematical results in extra-mathematical contexts.
• To generalize solutions that were developed for a special situation.
• To view solutions to a problem by using appropriate mathematical language and/or
communicate about the solutions.
446 R. Govender
E. Competencies to validate the solution: Competency

• To critically check and reflect on found solutions.
• To gain review some parts of the model or again go through the modelling process if
solutions do not fit the situation.
• To reflect on other ways of solving the problem or if solutions can be developed
differently.
• To generally question the model.
38.3 Study Design
This study, which was underpinned by a constructivist perspective and developmen-

tal approach to teaching and learning, sought to examine pre-service mathematics
teachers’ modelling competencies to perform modelling processes. Hence, a quali-
tative approach in the form of a case study research design was adopted to allow the
researcher to extract detailed in-depth and rich information (Barbie and Mouton
2010). The sample for this study was taught by the researcher. It consisted of a sin-
gle class of 68 third year pre-service mathematics teachers studying towards the
Bachelor of Education Senior Phase (grades 7–9) degree. They were divided into
ten groups consisting of about seven students per group to work on the Tree Problem
in their own time and do their project write-up. Data were collected from a combi-
nation of group written project submissions and PowerPoint presentations during
class time.
In this study, modelling is thought of as the creative and diligent use of mathe-
matics to solve problems built around real-life objects located in our environment
rather than pseudo-practical applications which typically dominate our mathematics
textbooks (Blomhøj 2008). Students were shown an old tree on the university cam-
pus and were required to determine the height of the tree. Students were required to
select suitable mathematics to model or explain their responses, and in creating and
using their model, they were required to (a) show their understanding of the prob-
lem, (b) state their assumptions, (c) create the mathematics needed to solve the
problem, (d) interpret their solution and (e) check their answers produced by the
model against reality.
38.4 Main Results and Analysis
The ‘tree’ modelling task required ‘out-of-the-box thinking’ for which there was no
exact correct answer. Data analysis was governed by an interpretive paradigm using
the modelling competency framework put forth by Blum and Kaiser (1997). As each
group debated how to go about finding the tree height, they decided on the
assumptions, diagrammatic representations and what mathematics and processes

were necessary to determine this. In the process they argued about the most feasible
approach which included doing research. Approaches ranged from the use of scale
to trigonometric methods as will be evidenced in the following discussions of Cases
1 and 2. These two cases, which stand out from the other eight groups, have been
selected as they demonstrate the flexibility in pre-service mathematics teachers’
thinking in the development of non-standard solutions using a modelling approach.
38.4.1 Case 1: Group A
The group initially expressed in their own words what the given task means for
them, which signals their ability to understand the real problem. For example, the
group stated: ‘What we understand about the question is that we have to find the
height of the tree from the ground up, which means it would exclude the roots of the
tree’ (Group A: Written report).
The group first used a scale model to establish an estimate of the height of the
tree. In doing so they first stated their assumptions as follows:
1. We assume that the position where student A2 is standing (see Fig. 38.2) is an
appropriate position in order to scale the tree.
2. The tree branches out, making the branch at the highest point, not 90° relative to
the ground. However, we assume that it is; thus, we assume that the highest
branch is perpendicular to where student A2 is standing, making it an appropri-
ate position regarding the scale (see Fig. 38.3) (Group A: Written report).
Fig. 38.2 Student A2 adjacent to tree

448 R. Govender
Fig. 38.3 Unit of measure stacked vertically
By making the assumptions, the students have recognized that they must deter-
mine vertical height of the tree. This mathematical knowledge pushed them to con-
struct a real model in their mind, where a vertical line was dropped from the vicinity
of the highest branch to a position on the assumed level ground, where student A2
was standing (Fig. 38.3). In doing so, students have identified the significant vari-
ables such as the upright (in this case the tree) must be perpendicular to the horizon-
tal ground to establish the required height. This has enabled the students to simplify,
structure and make the situation more precise. Subsequently, students proceeded to
set up a Picture-Real Object Scale model through mathematizing related quantities
and their relations. This entailed group members proceeding to convert the physical
situation into a mathematical model by using the real height of ‘student A2’, which
was measured using a tape measure to be 1.65 m, as the unit of measure to deter-
mine the real height of the tree. To facilitate this move, the students took a photo-
graph of the tree with ‘student A2’ standing adjacent to the tree (see Fig. 38.2). Then
they took the height of ‘image of student A2’ as the unit of measure and then built a
stacked ladder made up of 7¾ copies of the unit ‘image of student A2’ onto the
photograph (see Fig. 38.3). This suggests that students were able to represent the
situation using a pictorial representation (which is graphical in nature). They inter-
preted the fit of 7¾ units to mean that the height of the real tree is equal to 7¾ times
the real height of ‘student A2’. Hence, they obtained the real height of the tree to be
7.75 × 1.65 m = 12.79 m. The latter move signals that students were able to work
mathematically through using the concept of ratio to establish the relation between
the height of the tree and height of student A2.
However, the group members were keen to use a trigonometric method to verify
the reasonableness of their answer. Initially, the group members assumed that
Fig. 38.4 Student A3 measuring angle of elevation, building and using a trigonometric model to
find the height of tree
student A3, who was using a clinometer to determine the angle of elevation to the
tree top, was standing upright relative to the ground, which in mathematical terms
means standing perpendicular to a horizontal piece of ground. This enabled the
students to simplify the situation and build a real model as indicated in Fig. 38.4a.
This included the use of trigonometry after determining the angle of elevation and
distance from the tree. Through using a clinometer (Fig. 38.4a), the students esti-
mated the angle of elevation to be 41°. As shown in Fig. 38.4b, the students used
relevant quantities (like the height of student and distance between student and foot
of the tree), notations and objects to construct a two-dimensional trigonometric
model to mathematically represent the situation. This model made it possible for
students to invoke necessary mathematical knowledge to develop a possible solu-
tion. Through using the definition of tangent of an angle located within a right-angle
triangle, students were able to firstly calculate DE (see Fig. 38.4b, c) and then use it
to establish the height of tree to be 15.44 m as shown in Fig. 38.4c.
However, on comparing their answers obtained via the Picture-Real Object
Scale and trigonometric methods, respectively, the students argued why they prefer
to accept 12.79 m as the more reliable answer:
Both methods used gave us different answers. Method 1 was the most accurate method
therefore we will stick to this answer (12.79 m high). The method using trigonometry may
not be that accurate as the exact angle could not be obtained and also when looking at the
height of the surrounding buildings, which is approximately the height of the tree, we
noticed that the tree was close to the height of the surrounding buildings, who’s [sic] height
was ±13 m [sic]. (Group A: PowerPoint).
In the case of the trigonometric method, students interpreted their answers in

terms of the prevailing conditions of the real situation from which the problem
emerged, namely, the height of the tree cannot be higher than the height of the sur-
rounding buildings. This signals that students were able to critically check and
reflect on found solutions to the extent they were able to review their mathematical
model and see that their estimated measure of the angle using the clinometer could
be a possible cause of the inflated height of the tree.
450 R. Govender
38.4.2 Case 2: Group B
38.4.2.1 Method: Using the Bending Principle
In proceeding to determine the height of the tree, firstly group B showed understand-
ing of the question by reaffirming the given task as follows: ‘We would need [to]
measure from the tip to the base of the tree’ (Group B: PowerPoint). In order to move
forward, the group stated their assumption, namely: ‘the tree stands upright at a
perpendicular angle to the ground’. This assumption not only simplifies the given
situation but also demonstrates that they can see the prospective influence of the
identified perpendicular on a prospective calculation method. This group had done
some Internet research and identified a creative way to determine the height of the
tree, which I now call the ‘Bending Principle’ for calculating the height of a tree
(https://www.wikihow.com/Measure-the-Height-of-a-Tree). This model provides an
alternative way to calculate the height of a tree through a process of incremental unit
moves away from the base of the tree by a person bending and looking through their
legs at an angle of 45° to the top the tree. As soon as the person sees the top of the
tree, the person will stop. Then the horizontal distance between the person’s foot and
the base of the tree is measured. The students argued this horizontal distance gives
the height of the tree (see Fig. 38.5a) as follows: ‘In any right-angled triangle with
one of its angles measuring 45°, you find the lengths of the sides adjacent to the
right-angle always equal to each other’ (Group B: PowerPoint). This can be easily
verified by using the fact that sides opposite equal angles in any triangle are always
equal or by using the definition of the tangent of a given angle within a right-angled
triangle.
The students proceeded to try the method out practically. In their first attempt,
student B2 volunteered to walk forward in a bending position from the base of the
tree whilst trying to look at the tree at an angle of 45°. Student B2 (see Fig. 38.5b),
in the bending position (and thinking she is bending at an angle 45°), walked at
intervals of 450 cm until she could see almost the top of the tree, after which she
adjusted her final move and stopped. It was then recorded using a tape measure and
ground markings (intervals of 450 cm) that the student had covered four measure-
ments of 450 cm with the final move being 128 cm. This then provided the horizon-
tal distance traversed by the student to be calculated as (4 × 450 cm) + 128 cm
= 1928 cm. The students interpreted the 1928 cm to represent the height of the tree
a b
The height of the tree is

approximately equal to the
distance between the foot of the
tree and foot of student
Fig. 38.5 Bending principle – application (Source: Author)

and hence stated: ‘The tree was in fact 19.28 m tall’. Immediately on reflecting on
their answer, the group stated: ‘… not true given the real life situation’. The stu-
dents’ rejection of their answer was premised on the grounds that the tree was not
very much taller than the adjacent building which was three stories high (i.e. 12 m
high) and research confirming the height of the tree species ranges between 11 and
16 m. The latter incident illuminates students’ competency to perform modelling
processes like interpreting and validating.
As the prior solution of 19.28 m did not fit the situation, the students decided to
relook at the strategy they used. In doing so, they came up with the following
position:
We realised that the point at which the bending process was happening needed to be a true
45° angle. Therefore we decided to follow the same steps but to pay more attention to angle
at which we bend whilst trying to see the tree top/tip. We will repeat the task using the same
interval of measurement which was 450 cm. (Group B: PowerPoint).
This move by the students confirms that they have developed a sub-competency
of the validating process, namely, the ability to review parts of the model and again
go through the modelling process if solutions do not fit a situation. After repeating
the bending movement with a better estimation of a 45° angle, they calculated a
horizontal distance of 12.82 m. This led them to conclude the tree was approxi-
mately 13 m high. The group accepted the answer as the adjacent building was
approximately 12 m high and the tree was slightly taller than this building. This
latter move once again asserts that students have developed a sub-competency of the
validating process, namely, critically checking and reflecting on found solutions.
Pre-service mathematics teachers engaged in the process of understanding through

trying to make sense of what the real problem means to them. In most instances they
articulated in their own words what the problem requires them to calculate in the
given context. This was often proceeded by students articulating their assumptions,
which enabled them to find a real model through structuring and simplifying.
Permeating the cases was the assumption that the tree is vertical and is perpendicu-
lar to the ground, which represents the horizontal. This signals that the students
were able to recognize a quantity, namely, the right angle, which could influence the
situation, and use it to build a relationship. The aforementioned actions, skills and
abilities of the students suggest that they understood the real problem and this
helped finding and/or visualizing a real model through simplifying and structuring
(Maaß 2006). Each student group looked at their real model from a different per-
spective and mathematized relevant mathematical quantities, variables and relations
to set up a mathematical model from reality as espoused by Maaß (2006). The
Picture-Real Object Scale model and trigonometric models were expressed using a
variety of written symbols, photographs and diagrams.
452 R. Govender
Group A constructed the Picture-Real Object Scale model and used their math-
ematical knowledge associated with measurement, estimation, ratio and proportion,
rearrangement of formula and operations to produce a mathematical result (Blum
and Kaiser 1997). Their competency to interpret a mathematical result in a real situ-
ation was evident when they announced the height of the tree as 12.79 m. In addi-
tion, the students proceeded to validate their solution by building a two-dimensional
trigonometric model and used relevant mathematical knowledge to obtain a differ-
ent height of 15.44 m (see Fig. 38.4c). Through reflection and comparison (Sect.
38.4), students did not accept the mathematical result of 15.44 m to be the height of
the tree. In questioning the solution to the trigonometric model (Maaß 2006; Swetz
1991), students ascribed the reading of 15.44 m to an error of physical measurement
when using the clinometer. This suggests that Group A have been able to compe-
tently invoke sub-competencies associated with validation of the modelling process
(Blum and Kaiser 1997).
Group B students constructed a trigonometric model and applied the Bending
Principle to interpret the height to be 19.28 m (Sect. 38.4.2). In the validation pro-
cess, students refuted 19.28 m through comparison with the height of an adjacent
building of approximately 12 m high (Sect. 38.4.2). This shows students were able
to critically reflect on their solutions and question their model to the extent they
realized that the angle of bending was the causal factor (Maaß 2006; Swetz 1991).
Hence, they re-did the measurements, which resulted in a more realistic answer of
12.82 m. The latter action suggests that the students were able to identify the dis-
crepant part of the model and then go through the modelling process again if solu-
tions do not fit the situation (Blum and Kaiser 1997).
Consistent with the notions of creative and flexible thinking (Chamberlain and
Moon 2005; Poincare 1948; Sriraman 2004), results of this study showed that as
students worked in groups on the open-ended Tree Problem, they analysed the prob-
lem from different perspectives, generated unusual ideas and produced insightful
novel solutions (Doerr and English 2003). In achieving this, students were able to
experience mathematical modelling and sufficiently invoke and put into action the
necessary modelling sub-competencies (skills and abilities) to move from one step
of the modelling process to another (Blum and Kaiser 1997).
References
Barbie, E., & Mouton, J. (2010). The practice of social research (10th ed.). Cape Town: Oxford
University Press.
Blomhøj, M. (2008). Different perspectives in research on the teaching and learning mathemati-
cal modelling. In M. Blomhøj & S. Carreira (Eds.), Mathematical applications and model-
ling in the teaching and learning of mathematics (pp. 11–13). Proceedings of Topic Study
Group 21 at ICME11 in Monterrey, Mexico, July 6–13, 2008. Roskilde: IMFUFA, Roskilde
Universitetscenter.
Blomhøj, M., & Jensen, T. H. (2003). Developing mathematical modelling competence:

Conceptual clarification and educational planning. Teaching Mathematics and its Applications,
22(3), 123–139.
Blum, W., & Kaiser, G. (1997). Vergleichende empirische Untersuchungen zu mathematischen
Anwendungsfähigkeiten von englischen und deutschen Lernenden. [Unpublished application
to Deutsche Forschungsgesellschaft.]
Chamberlin, S. A., & Moon, S. M. (2005). Model-eliciting activities as tool to develop and identify
creativity gifted mathematicians. Journal of Secondary Gifted Education, 17(1), 37–47.
Doerr, H., & English, L. D. (2003). A modelling perspective on students’ mathematical reasoning
about data. Journal of Research in Mathematics Education, 34(2), 110–136.
Jensen, T. H. (2007). Assessing mathematical modelling competency. In C. Haines, P. Galbraith,
W. Blum, & S. Khan (Eds.), Mathematical modelling (ICTMA12): Education, engineering and
113–142.
Mousoulides, N., Pitttalis, M., Christou, C., & Sriraman, B. (2010). Tracing student’s modeling
processes in school. In R. Lesh, P. L. Galbraith, C. R. Haines, & A. Hurford (Eds.), Modelling
students’ mathematical modeling competencies (pp. 119–129). New York: Springer.
National Council of Teaching of Mathematics. (2000). Principles and standards for school math-
ematics. Reston: NCTM.
Niss, M. (2010). Modelling a crucial aspect of students’ mathematical modelling. In R. Lesh,
P. L. Galbraith, C. R. Haines, & A. Hurford (Eds.), Modelling students’ mathematical modeling
competencies (pp. 43–60). New York: Springer.
Oberholzer. (1992). Mathematical modelling. In M. Moodley, R. A. Njisane, & N. C. Presmeg
(Eds.), Mathematics education for in-service and pre-services teachers (pp. 193–209). Shuter
and Shooter: Pietermaritzburg.
Poincaré, H. (1948). Science and method. New York: Dover.
Schorr, R. Y., & Koellner-Clark, K. (2003). Using a modeling approach to analyze the ways in
which teachers consider new ways to teach mathematics. Mathematical Thinking and Learning,
5(2–3), 191–210.
Sriraman, B. (2004). The characteristics of mathematical creativity. The Mathematics Educator,
14(1), 19–34.
Swetz, F. (1991). Incorporating mathematical modeling into the curriculum. Mathematics Teacher,
84(5), 358–365.
Chapter 39
The Design of a Successful Teacher
Training to Promote Interdisciplinary
STEM Modelling Projects
Jean-Marie Lantau, Martin Bracke, Wolfgang Bock, and Patrick Capraro
Abstract This chapter describes the first year of a PhD project considering the
question of how a teacher training focusing on the implementation of interdisciplin-
ary STEM modelling projects based on the context of a Segway can be designed
successfully. It is quite ambitious to answer this question, since some deeper knowl-
edge about physical, technical and mathematical concepts is needed. Nevertheless,
based on 2-day teacher trainings, four interdisciplinary STEM modelling projects
were implemented, in which grade 12 students have gained insight into the control
of a Segway. After a short explanation of the mathematical background, we describe
the design of the teacher training as well as the resulting modelling projects. The
aim of this research is to identify successful factors of the teacher training which
will be analysed in a forthcoming period of the project.
Keywords Teacher training · STEM modelling projects · Segway
The working principle of a Segway was first treated in school by Papert (1972) who
gave school students the task to control an inverted pendulum on a truck aiming to
teach young students in the mathematical framework of control theory. Papert used
a very easy approach to solve the control problem, as the students would only move
the truck if a small angle was exceeded. In addition, Papert considered “recogniz-
able phases” of project work, such as planning, choosing simple strategies and
J.-M. Lantau (*) · M. Bracke · W. Bock · P. Capraro

Competence Center for Mathematical Modelling in STEM Projects at School (KOMMS),
Technische Universität Kaiserslautern, Kaiserslautern, Germany
e-mail: lantau@mathematik.uni-kl.de; bracke@mathematik.uni-kl.de; bock@mathematik.
uni-kl.de; capraro@mathematik.uni-kl.de

456 J.-M. Lantau et al.
finding of simple solutions, which might to be modified, to finally reflect the process
of working towards his goals. In our first interdisciplinary STEM modelling project
which focused on the comprehension of the functional principle of a real Segway,
students required knowledge about modelling a Segway as an inverted pendulum to
detect physical forces and energies leading to a mathematical model of a state space
system. The linear state space system is controllable and therefore stabilizable by
feedback control or PID control1 which can be implemented practically to control a
Lego Mindstorms® Segway (Lantau 2016). In a pilot project (i.e. Lantau 2016), the
participating students gained insight into the important role of STEM education for
daily life applications such as a Segway.
Tseng et al. (2011) noted that first year Taiwanese tertiary students recognized
the important role of STEM in science and engineering disciplines after participat-
ing in an interdisciplinary STEM modelling project. They pointed out that these
students had a positive attitude concerning the combining of STEM disciplines in
project-based learning. Thomas (2000) highlighted that students can establish their
own know-how in a project-based learning environment by practical activities,
interactive discussions and team cooperation. In this sense the teacher acts as a
facilitator of the learning process. The teacher is essential in supporting students’
learning through modelling processes (Moore et al. 2015). Doerr (2007) stresses
that teachers need to be aware that students might use a broad diversity of approaches
while doing modelling. Blum and Borromeo-Ferri (2009) indicate similar results
advising that teachers should encourage the finding of multiple solutions for a mod-
elling problem and thus be aware of some strategic interventions during modelling
projects to promote mathematical modelling in the classroom. In order to teach
mathematical modelling processes, Blum and Borromeo-Ferri (2009) suggested
that teachers need to be aware of a balance between maximal independence of stu-
dents and minimal help of teachers. The modelling project described in this chapter
requires not only mathematical concepts and tools but also a more holistic interplay
between different mental and manual skills. A concrete path to solve the problem is
not given but, due to the advice of Blum and Borromeo-Ferri, the teaching process
is steered by the action of the students during their project work.
In addition, the results of the LEMA project show that a professional develop-
ment course has a positive effect on self-efficacy and on pedagogical content knowl-
edge of teachers (Maaß and Gurlitt 2011). Gastón and Lawrence (2015, p. 9)
concluded in their work that “research does not reveal one perfect method or set of
techniques that can prepare each individual teacher to effectively teach mathemati-
cal modelling as a transferable process”. Most of the professional development
courses for mathematical modelling concentrate on preparing teachers to use math-
ematical modelling processes in regular lessons or small tasks. In contrast, the long
tradition of mathematical modelling in Kaiserslautern focuses on interdisciplinary
and real-life STEM problems which students usually investigate in project-based
learning environments such as modelling days and weeks (Bock and Bracke 2015).
It was not clear how a teacher training has to be designed such that teachers are able
1
PID control is the abbreviation for proportional-integral-derivative control.
39 The Design of a Successful Teacher Training to Promote Interdisciplinary STEM… 457
to supervise complex interdisciplinary STEM modelling projects. In order to inves-

tigate successful factors of a teacher training, Lipowsky (2010) set up a model to
describe the efficacy of teacher trainings. Concerning the design of a successful
teacher training, Lipowsky states that a research-based learning environment should
be aimed such that the participants recognize the intent of active learning which
increases the teachers’ contentment and acceptance of the teacher training.
39.2 Research Focus
The aim of the first year of the PhD project was to implement a successful design
for a teacher training concerning the implementation of interdisciplinary STEM
modelling projects at school. Four school projects were implemented by eight par-
ticipating teachers of a 2-day teacher training. These allow the assessment of the
design of successful training as teachers provide an interdisciplinary project-based
learning environment for their students. Since the main observation from the pilot
project (Lantau 2016) as well as from the first year of this project was that teachers
need to attend a respective training first to implement modelling projects on their
own, the research focus of the project is the analysis of key factors for a successful
teacher training so that modelling projects can be implemented at school with stu-
dents from grades 10 to 12.
39.3 Methodology
The first year of the project was necessary to explore a successful design of a teacher
training aiming at modelling projects to be implemented at school. In order to obtain
some first impressions of successful factors of the teacher training, the teachers
were asked to indicate components of the teacher training which they wanted to
include in their school projects. This was done by electronic surveys. The aim of the
survey was to gain a first insight of those components of the teacher training which
were translated into the school projects. Since the teacher training gave the teachers
the opportunity to create their own modelling project, another aim of the first year
was to observe how the resulting projects could promote interdisciplinary teaching.
Similar to the teacher training, the teachers were also asked to assess some factors,
which could be responsible for the success of their school projects. Based on a first
implementation of the project design, the research focus for the second year will be
the key factors for a teacher training, such that modelling projects are implemented
at school. In order to analyse these factors, a guided interview is set up based on the
model of Lipowsky (2010) which describes key factors of successful teacher train-
ings. The interviews were conducted with 11 teachers and 3 teacher students who
attended the teacher training in school term 2017/2018. Based on the 2018 teacher
trainings, five different modelling projects were implemented in summer 2018. The
analysis of the successful factors will be done by the method of content-structuring

qualitative content analysis (Kuckartz 2018).
39.4 Physical and Mathematical Background of a Segway
A Segway is modelled as an inverted pendulum, which contains two centres of mass

(those of the wheels M and of the Segway m) connected by a massless pole of length
l. The aim is to steer the inverted pendulum by a motor force u such that the pendu-
lum is positioned vertically upright in its rest position. In order to describe the
inverted pendulum by physical forces, two approaches can be chosen. The first one
analyses acting tangential and radial forces, since the control of the centre of mass
of the inverted pendulum is done in a circular movement. This approach leads to a
nonlinear, inhomogeneous ordinary differential equation of second order:
mg sin ml u.
The second approach is to describe the movement of the inverted pendulum by the
Euler-Lagrange formalism (Goldstein 1980) in which the Lagrange function (dif-
ference of kinetic and potential energy) for the generalized coordinates position s
and angle α is set up:
1 1
 t ,s t ,s t , t , t M m s2 m 2 l 2 ml cos  s mgl cos us
2 2
In a second step, the Euler-Lagrange equation is calculated for both generalized

coordinates:
d   d  
0 0
dt  dt s s
This process leads to two equations, one describing horizontal forces and the other
describing tangential forces of an inverted pendulum:
0 M m 
s ml cos ml 2 sin u
0 ml 2 ml cos 
s mgl sin
A mathematical model for both physical approaches is set up by using linear Taylor
approximations for the position s and the angle α in the equilibrium:
x T s s  0 0 0 0 . T
T
The mathematical model is described by a linear state space system x Ax Bu :

0
s
¨ 0 1 0 0 s 1
s
s mg M
0 0 0 u
 M 0

¨
0 0 1  1
0
M m g Ml
0 0 0
Ml
in which the state-space vector x contains the generalized coordinates angle α and
position s and their time derivatives. In order to stabilize the linear state space sys-
tem, two control techniques can be applied. The first one describes the PID control
(Sontag 1998) in which the motor force u is chosen as a weighted sum of the gener-
alized coordinates (proportional control), its time integral (integral control) and its
time derivative (derivative control). The parameters for the PID control part are
usually determined experimentally regarding some conditions that have to be ful-
filled to generate asymptotic stable solutions for the linearized system. The second
control technique can be implemented by feedback control in which the motor force
u is chosen as u = Fx to generate a linear system of ordinary differential equations
x A BF x . Regarding the theorem of Wonham (Sontag 1998), we can choose
a monic polynomial P such that there exists an F for which P is the characteristic
polynomial of the matrix A + BF since the time-invariant system is controllable.
The aim is now to place the poles of P into the left-half complex plane ℂ_ to calcu-
late F by the formula of Ackermann F = − e4 ⋅ K(A; B)−1 ⋅ P(A) (Ackermann 1968),
where K(A,B) is the Kalman matrix of the system K(A, B) = (B, AB, A2B, A3B) which
is invertible, since the linear state space system is controllable and P(A) describes
the characteristic polynomial of A + BF which is evaluated in A.
39.5 Project Conceptualization
The first iteration of the project took place in the school term 2016/2017, starting
with an introductory meeting in which the attending teachers gained an insight into
the results of a pilot project (Lantau 2016), concerning the mathematical modelling
of a Segway. Four introductory meetings took place during December 2016 until
January 2017 and were attended by 23 teachers. These meetings were followed by
two similar teacher trainings of 2 days’ duration each in March and April 2017
which were attended by eight teachers. The aim of the teacher training was to pre-
pare teachers to implement a modelling project over several days with their grade 12
students. In addition, an online course was set up in which the participants perused
material that was elaborated before and during the teacher trainings (mainly scripts
and GeoGebra simulations). Just before the start of the school projects, preliminary
meetings took place to discuss the project plans individually. In June 2017 four dif-
ferent school projects took place, lasting from 2 to 4 days. The projects were imple-
mented by 9 teachers and 51 students of grade 12 (16–18 years old).
39.5.1 Implementation of a Two-Day Teacher Training
The 2-day teacher training followed the modelling cycle of Blum and Leiss (2007).
Since the attending teachers indicated that they had less experience in implementing
modelling projects, the teacher trainings were designed as action-orientated with
only small phases of lecturer-centred teaching. Figure 39.1 shows the design of the
2-day teacher training.
At the beginning of the teacher training, teachers experimented with a Ninebot
Mini Segway. After their first driving experiences, the teachers could explore how
the Ninebot Mini Segway is controlled to drive safely. The exploration of the
Ninebot Mini Segway’s control is very important, since it facilitated the subsequent
Fig. 39.1 The design of the 2-day teacher training

steps of the teacher training. Due to their own driving experience, the teachers fig-
ured out that the inclination angle α influences the driving behaviour of a Segway.
Based on that knowledge, the teachers were able to set up a first model for the con-
trol process of a Segway without additional help. Furthermore, the phase of experi-
encing the control of a Segway specifies the aim of the modelling project. Based on
the model of an inverted pendulum, two different kinds of physical descriptions
were introduced to the teachers. The teachers judged the physical concept of describ-
ing the movement of an inverted pendulum by the Euler-Lagrange formalism as too
difficult to teach in a modelling project for grade 12. The teacher training was con-
tinued by setting up a linear mathematical model to develop the control techniques
of P(I)D and feedback control. Furthermore, an explicit Euler method was elabo-
rated in GeoGebra to generate parameters for the PD control in the nonlinear math-
ematical model. In addition, a stability criterion for the linearized homogenous
system of ordinary differential equations based on the eigenvalues of the matrix
A + BF was elaborated, leading to the derivation of Ackermannn’s formula. In order
to emphasize the product-based and interdisciplinary design of the subsequent
school projects, the teachers were introduced to the construction and practical con-
trol of a Lego Mindstorms® Segway, which is based on a PID approach. The con-
nection between a Lego Mindstorms Segway and a Ninebot Mini Segway was
given, since the Ninebot Mini Segway is also controlled by a PID control, using 200
gyro sensors in the vehicle. Finally, the teacher training was summarized by pre-
senting the modelling cycle of Blum and Leiss (2007), firstly to reflect on the train-
ing and secondly to show how their own project could be implemented following
the steps of the modelling cycle.
39.5.2 Implementation of Four School Projects
In June 2017 the projects were implemented in four different schools in Rhineland-
Palatinate, Germany. The projects were planned individually by the teachers, such
that the thematic focus of the projects was set differently. However, all projects had
a common setting: At the beginning of each project, the students were provided with
the opportunity to do experiments with the Ninebot Mini Segway, and the project
days were closed by a final presentation of the students, summarizing the main
aspects of the project. As three of four projects took place at the same time, the
Ninebot Mini Segway was not always available during the projects. The following
sub-sections will briefly describe the implementations of the project.
39.5.2.1 Gerolstein: 21/06/2017–23/06/2017 – 4 Teachers and 20 Students
The first task for the students was to set up their own model autonomously. Since
students were able to include their own experiences while driving with the Ninebot
Mini Segway, it was easier for them to set up the model of an inverted pendulum. In
a following step, the teachers set up the main question of how to stabilize a Segway.
In order to answer this question, students described acting tangential forces for the
inverted pendulum. The physical description was followed by a phase in which the
teachers guided the students to develop a PD control. Parallel to the elaboration of
a theoretical PD control, the students created a video analysis of the stabilizing
Ninebot Mini Segway to compare their real observations to their theoretical results.
The phase of elaborating a PD control technique was designed by following the
principle of minimal help. The students worked in an action- and research-orientated
manner by using video tapes to analyse the control of a Ninebot Mini Segway.
During their mathematical working phase, the students elaborated an explicit Euler
method to solve both the nonlinear and the linearized ordinary differential equation.
Furthermore, they implemented a GeoGebra simulation to visualize their mathe-
matical results. In a final part of the project, some students used their theoretical
results to implement a PD control in leJOS EV32 to control a self-constructed Lego
Mindstorms® Segway.
39.5.2.2 haleischweiler-Fröschen: 26/06/2017–27/06/2017 – 2

T
Teachers-12 Students
The teachers in Thaleischweiler-Fröschen used the modelling project to introduce

the concept of ordinary differential equations to their students. Therefore, the phase
of doing experiments with a Segway was followed by the physical description of
analysing tangential forces for the inverted pendulum. The main focus of the project
in Thaleischweiler-Fröschen was to promote action-orientated and research-based
lessons. Therefore, the students used the Segway to develop their own questions and
answer them for themselves by doing experiments and using video analysis to
explain relevant parameters. As an example, the students asked themselves what is
the maximal speed one can drive in a circle with given radius. Interdisciplinary
learning was promoted while working on this question, since the students developed
a model involving circular forces to express the maximal speed. A video analysis of
the movement was used to express the circular movement mathematically by using
trigonometric functions and their derivatives.
39.5.2.3 Trier: 26/06/2017–29/06/2017 – 1 Teacher-9 Students
The main focus of the project in Trier was to compare the control process of a
Segway to a similar physical control process, namely, Robert Pohl’s stabilization of
a monorail vehicle using an induction brake. The students in Trier were taught to
apply the physical concepts of momentum of inertia and torque to the stabilizing
process of a Segway. In Trier, the students also used video analysis to describe the
2
leJOS is a tiny Java virtual machine which is used with the Lego Mindstorms® EV3 brick.
stabilizing process of the Ninebot Mini Segway. In their phase of working mathe-
matically, the students distinguished three cases for stabilization depending on the
choice of the parameters of the PD control. They visualized their theoretical results
by setting up a GeoGebra simulation. Furthermore, the students constructed and
implemented a Lego Mindstorms® Segway based on their theoretical results.
39.5.2.4 Zweibrücken: 26/06/2017–30/06/2017 – 2 Teachers-11 Students
The project in Zweibrücken was also designed to be research-orientated, since the

students elaborated their own questions. As one example, the students investigated
how gyro sensors are constructed to calibrate the angle and the angular velocity of
a Segway. Furthermore, the students built two Lego Mindstorms® Segways in order
to implement a P and a PD control to distinguish that only a PD control stabilizes
the Segway asymptotically. In this project, the inverted pendulum was also described
in physics terms by calculating kinetic and potential energy. In Zweibrücken, the
concept of a PD control was elaborated theoretically in order to solve the linearized
ordinary differential equation of second order.
39.6 First Impressions
The teacher training and school projects were evaluated by more informal teacher
surveys to analyse aspects of the teacher training which were implemented by the
teachers at their own modelling project. The electronic teacher survey was filled out
by every participating teacher of the teacher training and the school projects to gain
an insight into first impressions of successful factors for the teacher training and the
school projects.
A first insight from the teacher survey shows that teachers think that the phase of
doing experiments with the Segway and experiencing the control of a Segway
increases the motivation of the students to answer the question how to control a
Segway safely. This observation is connected to theoretical work of Bock et al.
(2017), who claim that a product increases the motivation and the success of model-
ling tasks in interdisciplinary modelling projects. In the analysis of the preparation
of the project, teachers wanted to design their project as open- or semi-structured.
More precisely, most of the teachers wanted to implement a research-based learning
environment in which the students could elaborate the answers to their own ques-
tions alone by doing experiments and collecting important data. Some teachers
planned the project semi-structured, following the approach of Thomas (2000) in
which teachers act as moderators of the learning project such that project-based
learning can be promoted, but short interventions by the teachers are planned to
guide the process of learning. One last aspect of the teacher questionnaire shows
that teachers assessed several aspects of their project design as a good basis to
enhance learning of their students. These aspects were first the interdisciplinary
connection of concept and the implementation in a project based on an authentic

context such that theoretical results and practical experiments are connected.
39.7 Outlook
In the first year of the project, the focus was on the implementation of a teacher
training aimed at the implementation of interdisciplinary STEM modelling projects
with students of grade 12. The structure of the teacher training followed the model-
ling cycle of Blum and Leiss (2007) such that teachers gained an insight into pro-
moting interdisciplinary modelling tasks. In order to evaluate the teacher trainings
and the school projects, a more informal survey was set up to obtain first insights
into the conditions which are responsible for the success of the teacher training
concerning the implementation of modelling projects. As a next step, the project
conceptualization will be evaluated by setting up guided interviews for the partici-
pating teachers of the second year of the project in order to analyse key factors of
the teacher training more precisely.
Acknowledgement This work was partially supported by the European Social Fund (ESF) proj-
ect SchuMaMoMINT and by U.EDU-Medienbildung entlang der Lehrerbildungskette (Technische
Universität Kaiserslautern) as part of the national and federal project Qualitätsoffensive
Lehrerbildung, funded by the Federal Ministry of Education and Research.
References
Ackermann, J. (1968). Time-optimal multi-input sampled-data systems. Proceedings of IFAC

Symposium Multivariable Control, Düsseldorf.
Blum, W., & Borromeo-Ferri, R. (2009). Mathematical modelling: Can it be taught and learnt?
Blum, W., & Leiss, D. (2007). How do students and teachers deal with mathematical modelling
problems? In C. Haines, P. Galbraith, W. Blum, & S. Khan (Eds.), Mathematical modelling:
Education, engineers and economics (pp. 222–231). Chichester: Horwood.
Bock, W., & Bracke, M. (2015). Applied school mathematics – Made in Kaiserslautern. In
H. Neunzert & D. Prätzel-Wolters (Eds.), Currents in industrial mathematics: From concepts
to research and education (pp. 421–456). Berlin: Springer.
Bock, W., Bracke, M., & Capraro, P. (2017). Product orientation in modelling tasks. In T. Dooley
& G. Gueudet (Eds.), Proceedings of CERME10 (pp. 1028–1030). Dublin: DCU Institute of
Education and ERME.
Doerr, H. M. (2007). What knowledge do teachers need for teaching mathematics through appli-
cations and modelling? In W. Blum, P. Galbraith, H.-W. Henn, & M. Niss (Eds.), Modelling
and applications in mathematics education – The 14th ICMI-study (pp. 69–78). New York:
Springer.
Gastón, J., & Lawrence, B. (2015). Supporting teacher’s learning about mathematical modeling.
Journal of Mathematics Research, 7(4), 1–11. https://doi.org/10.5339/ijsp.v4n4pl.
Goldstein, H. (1980). Classical mechanics (2nd ed., pp. 352–353). San Francisco: Addison Wesley.
Kuckartz, U. (2018). Qualitative Inhaltsanalyse. Methoden, Praxis, Computerunterstützung (4th

ed.). Weinheim: Beltz Juventa.
Lantau, J. (2016). Mathematische Modellierung eines Segway mit Umsetzung in der Schule. Eine
interdisziplinäre Projektarbeit. Saarbrücken: AV Akademikerverlag.
Lipowsky, F. (2010). Lernen im Beruf. Empirische Befunde zur Wirksamkeit von
Lehrerfortbildungen. In F. Müller, A. Eichenberger, M. Lüders, & J. Mayr (Eds.), Lehrerinnen
und Lehrer lernen – Konzepte und Befunde zur Lehrerfortbildung (pp. 51–73). Münster:
Waxmann.
Maaß, K., & Gurlitt, J. (2011). LEMA – Professional development of teachers in relation to math-
ematics modelling. In G. Kaiser, W. Blum, R. Borromeo Ferri, & G. Stillman (Eds.), Trends
in the teaching and learning of mathematical modelling (pp. 629–639). Dordrecht: Springer.
Moore, T. J., Doerr, H. M., Glancy, A. W., & Ntow, F. D. (2015). Preserving pelicans with models
that make sense. Mathematics Teaching in the Middle School, 20, 358–364.
Papert, S. (1972). Teaching children to be mathematicians versus teaching about mathematics.
International Journal of Mathematical Education in Science and Technology, 3(3), 249–262.
Sontag, E. (1998). Mathematical control theory: Deterministic finite dimensional systems.
New York: Springer.
Thomas, J. (2000). A review of research on project-based learning. San Rafael: Autodesk.
Tseng, K., Chang, C., Lou, S., & Chen, W. (2011). Attitudes towards science, technology, engineer-
ing and mathematics (STEM) in a project-based learning (PjBL) environment. International
Journal of Technology and Design Education, 23(1), 87–102. https://doi.org/10.1007/
s10798-011-9160-x.
Chapter 40
Using Mathematical Modelling
in the Teaching and Learning
of Mathematical Literacy: Perspectives
on the South African Context
Duncan Mhakure
Abstract This chapter provides a theoretical perspective of how the South African
Literacy given by the Organisation for Economic Co-operation and Development.
Through unpacking this definition, and the South African Mathematical Literacy
curriculum definition, I demonstrate through the use of examples that mathematical
modelling is the missing cog in the South African Mathematical Literacy curricu-
lum context. I argue that this difference is significant, insofar as students’ acquisi-
tion of mathematical skills and competencies needed in the real world of work and
as active participating citizens of a democratic country. I also argue that the teaching
of Mathematical Literacy should focus on the development of key mathematical and
statistical skills.
Keywords Curriculum · Mathematical competencies · Mathematical Literacy ·

Mathematical modelling · Mathematical skills · Statistical skills
40.1 Introduction
One of the flashpoints in the incendiary debates about Mathematical Literacy, at

least within the South African context, is that it tends to focus on the ability of stu-
dents to understand and do mathematical calculations. This notion is often derived
from “word problems” which do not represent authentic everyday life situations. In
this chapter, I seek to provide a theoretical perspective of how the South African
D. Mhakure (*)
Centre for Higher Education Development, University of Cape Town,
e-mail: duncan.mhakure@uct.ac.za

468 D. Mhakure

Literacy given by the Organisation for Economic Co-operation and Development
(2013). Within the South African mathematics education context, students can
choose to study either mathematics or Mathematical Literacy as subjects during
their last 3 years of high school (Years 10–12). The difference between Mathematical
Literacy and mathematics is that mathematics has a vertical development and it
focuses on moving up the ladder of abstraction, whereas Mathematical Literacy is
about seeing everyday contexts through a mathematical lens.
Education reform through the work published by the Organisation for Economic
Co-operation and Development (2013) has identified Mathematical Literacy as one
of the key literacy developmental areas and as a possible strategy for promoting
widespread implementation of reforms in mathematics education. The rationale
behind focusing on Mathematical Literacy by the Organisation for Economic
Co-operation and Development (2013) is that Mathematical Literacy is seen as a
cornerstone to the development of mathematical competencies and skills that stu-
dents need beyond school (Best 2008). While this chapter is anchored in the
Organisation for Economic Co-operation and Development (2013) definition of
Mathematical Literacy, which has evolved over the last decade, the construct of
Mathematical Literacy is contested. This is so because the role that Mathematical
Literacy plays and the form it takes to develop mathematical competencies and
skills vary from one country to another (Frejd and Geiger 2017) and can also be
given different interpretations within one country. In some countries, such as the
United States of America, Mathematical Literacy is referred to as quantitative lit-
eracy (De Lange 2003) or as numeracy (Gardiner 2004) in the countries which were
historically under the influence of the British Empire.
The Organisation for Economic Co-operation and Development (2013 p. 25)
definition of Mathematical Literacy is as follows:
Mathematical Literacy is an individual’s capacity to formulate, employ, and interpret math-
ematics in a variety of contexts. It includes reasoning mathematically and using mathemati-
cal concepts, procedures, facts, and tools to describe, explain, and predict phenomena. It
assists individuals to recognise the role that mathematics plays in the world and to make the
well-founded judgements and decisions needed by constructive, engaged and reflective
citizens.
This definition gives a clear framework on how Mathematical Literacy can be oper-
ationalised and new insights on what should constitute Mathematical Literacy prac-
tices (Stacey and Turner 2015). A close analysis of the Organisation for Economic
Co-operation and Development (2013) Mathematical Literacy framework shows
that first, the definition stresses competencies that are applicable to all ages and all
spheres of everyday life from schooling and technical and vocational and other lev-
els of professional expertise. Hence, this new definition seeks to address some of the
perceived criticisms that Mathematical Literacy is a low-level kind of mathematics
targeting the teaching and learning of basic mathematics equivalent to the type of
mathematics curriculum taught in elementary schools. Second, the Organisation for
Economic Co-operation and Development (2013) document aligns itself with
notions of mathematical modelling which is critical to mathematical thinking as
40 Using Mathematical Modelling in the Teaching and Learning of Mathematical… 469
evidenced by its focus on formulating, employing and interpreting mathematics in

a variety of unfamiliar contexts. In this chapter, a variety of unfamiliar contexts
constitute personal, societal, scientific and occupational ones, among others. Last,
the definition also stresses the importance of communicating and social interacting
with others as one explains, describes, engages and makes informed decisions about
issues of public interest and citizenry.
In addition, and apart from using critical insight into formulating, employing and
interpreting mathematical concepts in a variety of unfamiliar contexts, Mathematical
Literacy involves a further understanding of algorithms (Hughes-Hallett 2001). The
understanding of these algorithms necessitates an understanding of algebra and of
the identification and interpretation of quantitative relationships in unfamiliar con-
texts (De Lange 2003). This complex analysis I have outlined here is hardly part of
the teaching of mathematics and mathematics curriculum in schools, not least in the
teaching of Mathematical Literacy, in the context of the South African mathematics
school curriculum.
40.2 Cognitive Apprenticeship Theoretical Framework

to Guide Teaching and Learning
of Mathematical Literacy
Cognitive apprenticeship theory thrives in an authentic learning environment where

knowledge is constructed within social contexts (Lyons et al. 2017). Authentic
learning environments use teaching and learning activities that focus on the applica-
tion of mathematical concepts to real-world complex problems whose solutions are
aimed at improving or understanding societal challenges, such as environmental
issues, social justice and global warming among others. Cognitive apprenticeship
theory (Oriol et al. 2010), as a pedagogical model, is foregrounded within the situ-
ated learning paradigm and the traditional apprenticeship model. As a theoretical
framework, cognitive apprenticeship theory emphasises the “development of cogni-
tive and metacognitive knowledge rather than the development of physical skills”
(p. 210). Cognitive apprenticeship theory, as a framework, can make important con-
tributions to the pedagogy of Mathematical Literacy. First, through the acquisition
of higher-order thinking skills, cognitive apprenticeship theory allows Mathematical
Literacy students to gain access to expert thinking – this is facilitated by the use of
teaching strategies, such as mathematical modelling, problem-solving and coaching
among others (Bouta and Paraskeva 2013). Given that the long-term goal of transfer
of knowledge in mathematics is perceived difficult, particularly where learning
using unfamiliar complex contexts is envisaged, cognitive apprenticeship theory
can be used to make the students’ thinking processes visible to the teachers (Jürges
et al. 2012). Caspersen and Bennedsen (2007) assert that cognitive apprenticeship
theory is designed to develop the students’ higher-order thinking skills with help
from the teacher. Second, in addition to promoting student-centred learning, and the
470 D. Mhakure
development of higher-order thinking skills, through interacting as communities of

practice, students work together, thus creating opportunities “for peers to serve as
cognitive resources for each other” (Bouta and Paraskeva 2013, p. 162).
40.3 he Context of Mathematical Literacy

T
as a Mathematics Subject in South Africa
40.3.1 hat Constitutes Mathematical Literacy Curriculum

W
in South Africa
Mathematical Literacy as a school subject was introduced in South Africa at the

beginning of 2006 at the Further Education and Training phase – it was intended for
Years 10–12. Currently, within the context of South African mathematics education,
students are allowed to study either mathematics or Mathematical Literacy. The
decision to study Mathematical Literacy instead of mathematics is largely based on
the following: first, low mathematical competencies and skills and the assumption
that there is a better chance of achieving better grades in Mathematical Literacy, and
second, mathematics not being perceived as necessary for a career path (Department
of Basic Education 2011). Before the introduction of Mathematical Literacy in
2006, South Africa’s mathematics education at the Further Education and Training
phase was divided into two tiers: higher-grade mathematics which was intended for
students who were higher achievers in mathematics and standard-grade mathemat-
ics for students who were lower achievers in mathematics. Before the implementa-
tion of Mathematical Literacy, about 40% of the students studied neither higher-grade
mathematics nor standard-grade mathematics, 50% studied standard-grade mathe-
matics, and under 9% of the students studied higher-grade mathematics
(Venkatakrishnan and Graven 2006). Within these two tiers, there was no direct
mentioning of how literacy issues were dealt with; however, one can assume that
literacy was covered within the mathematical content of the tiers. Post 2006, the
rationale for the implementation of Mathematical Literacy foregrounds access
(Graven and Buytenhuys 2010) and empowerment.
In terms of access, the concerns before 2006 were that the majority of students
were not studying mathematics in Years 10–12 and among those that studied math-
ematics, a paltry pass rate of approximately 50% was registered. The issue was that
many students graduated from high school without the necessary numeracy and/or
quantitative skills required of them in the real world of work and as active citizens.
So the introduction of Mathematical Literacy was intended to empower students
with the necessary elemental mathematical and statistical skills that will allow them
to function as critical citizens in social and political discourses (Department of
Basic Education 2011).
This statement from the Department of Basic Education (2011) aligns itself with
five key tenets about Mathematical Literacy. First, the content of Mathematical
Literacy is restricted to elementary mathematical concepts and competencies –
there is no focus on abstract mathematical conceptions. Second, the mathematical
concepts taught in Mathematical Literacy should be located in authentic real-life
situations, rather than those located in semi-real, contrived, and/or fictitious situa-
tions. Third, in Mathematical Literacy students are expected to solve problems from
both familiar and unfamiliar contexts, for example, problems with unfamiliar con-
texts are problems that are related to workplaces which students are not part of.
Fourth, to enhance the development of students as critical citizens, Mathematical
Literacy learning activities should give students the opportunities to engage with
contexts where they practice decision-making and communicating skills particu-
larly where national issues are being discussed, for example, social justice, political,
economic and environmental issues to mention a few. Last, in Mathematical
Literacy, the learning activities are derived from real-life contexts where solutions
require the use of integrated mathematical content from different topics. These key
tenets of Mathematical Literacy distinguish it from mathematics as a high school
subject.
Hughes-Hallett (2001) makes two important differences between Mathematical
Literacy and mathematics: first, that mathematics has a vertical development in that
its focus is about moving up the ladder of abstraction, while Mathematical Literacy is
about seeing everyday contexts through a mathematical lens. Second, research has
shown that at school level, increased mathematics learning does not increase the stu-
dents’ facility in applying mathematical concepts to unfamiliar everyday life contexts
(Hughes-Hallett 2001). Research by Chapman and Lee (1990) has shown that stu-
dents studying mathematics were not necessarily applying their mathematical skills
to other school subjects. On the contrary, Chapman and Lee (1990) also observed that
students from less extensive mathematical backgrounds had the tendency to apply
mathematical skills they possessed well in other contexts they encountered.
The current South African Mathematical Literacy curriculum consists of two
categories: the basic skills topics and the application skills. These are shown in
Table 40.1. Students are expected to integrate elementary mathematical content/
skills from the basic skills topics to make sense of contexts and content embedded
Table 40.1 Summary of the basic skills topics and application topics of Mathematical Literacy
Basic skills topics Application topics
Interpreting and communicating answers and Finance
calculations
Patterns, relationships and representations Data handling
Numbers and calculations with numbers Probability
Measurement
Maps, plans and other representations of the
physical world
472 D. Mhakure
in application topics. The students would have been exposed to elementary mathe-
matical content from the basic skills topics in Years 8 and 9. The determination of
which basic skills are used during problem-solving is driven by the context where
the problem is situated.
40.3.2 ifferences Between Mathematical Literacy

D
Curriculum in South Africa and Organisation
for Economic Co-operation and Development Definition
The South African Mathematical Literacy curriculum differs from the definition of
Mathematical Literacy given by the Organisation for Economic Co-operation and
Development. Table 40.2 shows the differences between Organisation for Economic
Co-operation and Development (2013) and Department of Basic Education (2011)
definitions of Mathematical Literacy.
Mathematical Literacy curriculum in the context of South Africa does not define
the role of the use of technology or the use of computer-based environments for
learning and teaching Mathematical Literacy. Recent lines of research investigating
the Mathematical Literacy skills required in workplaces have shown that there is an
interdependence between Mathematical Literacy skills and computer-based skills.
In addition, the computer technology through a range of its software packages, for
example, Excel, provides opportunities for students to work more interactively as
they engage in the analyses of large real-world data sets from complex contexts. The
use of computer-based tools in Mathematical Literacy goes beyond the latter to
include computer-based assessments of Mathematical Literacy skills (Bardini 2015)
such as making charts, producing graphs of functions after completing modelling
processes to answer questions and sorting and analysing raw data, among others
(Greiff et al. 2013).
Table 40.2 Summary of the differences between Organisation for Economic Co-operation and
Development and Department of Basic Education definitions of Mathematical Literacy
Organisation for Economic Co-operation and Department of Basic Education
Development definition definition
Mathematical Literacy is operationalised through Mathematical modelling is not
mathematical modelling explicit, though it is implied
Understanding algebra is key Elementary algebra expected
Taught using ill-structured challenging tasks that help Taught using “word problems” or
students to rehearse for complex ambiguities of the “pseudo-contextualisation” – No real
world of work – Authentic real contexts are used which contexts are used
are appropriate for mathematical modelling
Stresses the interdependency between Mathematical Does not define the role of the use of
Literacy and the use of technology. Technology, through technology environment for the
the use of appropriate software packages, can be used to teaching and learning of
analyse large amounts of data during mathematical Mathematical Literacy
modelling
40.4 sing Mathematical Modelling in Mathematical

U
Literacy in South Africa
In this section, first, I present an example of an authentic real-world context to illus-

trate how mathematical modelling can be used as an instructional strategy in the
teaching of Mathematical Literacy in the South African mathematics education con-
text. Second, I illustrate how technology can be used in the teaching and learning of
Mathematical Literacy in South Africa.
40.4.1 xample 1: Applications of Modelling to an Authentic

E
Real-World Context
The following task is an excerpt from an article in the Cape Times Newspaper. It
offers the task solver a real-world context.
Instructions to the students: Using the excerpt below as a guide, describe the
growth pattern of the City of Cape Town from 1945 to 2030.
Development land poser in Cape Town as city size expected to double

by 2030
The City of Cape Town, estimated to be almost 35,000 hectares in size in

2007, will double by 2030. This means it is expanding by 650 ha annually,
increasing by what is the equivalent of two rugby fields in size every day. The
city in 2007 was seven times larger than it was in 1945. Since 1977, the city
has been growing at a fairly constant 2% until 1988, with the rate declining to
1.5% thereafter. City and population growth were similar until 1988, when the
population growth outstripped the city’s growth.
According to the Organisation for Economic Co-operation and Development

(2013) definition of Mathematical Literacy, mathematical modelling is represented
as an individual’s capacity to “formulate, employ, and interpret mathematics in a
variety of contexts” (p. 25). Using this example of the City of Cape Town, I will
illustrate how these key tenets of formulate, employ and interpret are operation-
alised. First, formulating situations mathematically (referred to as formulate) – stu-
dents convert a real-world context on the “Development land poser in Cape Town”
into a mathematical problem through identifying key mathematical conceptions.
These mathematical conceptions include, but are limited to, percentages, percent-
age changes, growth factors, areas as of rugby fields – as measured in hectares, City
of Cape Town area – measured in hectares, and compound growth. In addition, the
474 D. Mhakure
student at this stage starts making or identifying the relationships between these
mathematical conceptions, for example, how consecutive percentage increases are
related to compound growth and how many hectares are in one rugby field. Second,
employing mathematical concepts, facts, procedures and reasoning (known as
employ) – at this stage, the student is expected to carry out the calculations and
arrive at a mathematical solution. Important in this example, though, students are
expected to do mathematical calculation using procedures requiring understandings
of percentages, growth factors and physical areas of shapes. Last, interpreting,
applying and evaluating mathematical outcomes (referred to as interpret) – the
example on the development of the city of Cape Town requires the students to make
predictions on the size of the city in two ways. For example, given a year before
2007, students are required to predict the size of the city in 2030. Or given the year
ahead – 2007, the students are required to predict the size of the city in 1945. In both
calculations, students have to evaluate whether their answers derived from the real-
world context make sense. If the solutions make sense – can the same procedures of
solving this problem be applied to finding solutions to real-world problems in other
unfamiliar contexts? Using this example of the City of Cape Town provides a social
setting in which to apply the cognitive apprenticeship. In addition, cognitive appren-
ticeship will allow the teacher to use the strategies of coaching, scaffolding, articu-
lation and exploration, to ensure that students move beyond the declarative
knowledge (such as facts, concepts and procedures of the subject) to the acquisition
of cognitive process and skills required in expert practice (Lyons et al. 2017).
40.4.2 xample 2: Use of Technology in the Teaching

E
and Learning of Mathematical Literacy
Within the context of South Africa, social grants, for example, child grants, are the
subject of considerable debates among different communities. Hence, students will
find analysing and solving the question which follows using technology interesting.
Child Grants
Instructions to students: Describe the trends in the number of beneficiaries
accessing child grants and grant expenditure patterns in South Africa between
2006 and 2013.
Table 40.3 Summary of the distribution of number beneficiaries accessing child grants and the
child grant budget per year
2006 2007 2008 2009 2010 2011 2012 2013
Number of beneficiaries accessing child 7 7.9 8.2 8.8 9.4 10.2 10.9 11.3
grant (millions)
Child grant expenditure (R billions) 14 18 20 22 27 30 34 38
Unpacking example 2 using the three key tenets of mathematical modelling of

formulate, employ and interpret from PISA 2012, the following are observed. First,
the formulate process is truncated because the mathematical model in this example
is already presented in a simplified format. However, students need a clear under-
standing of how the three variables – years, number of beneficiaries accessing child
grant and child grant expenditures – in the context are related. Second, in the employ
process, students are expected to use Excel to generate scatter plots and/or time-
series charts to show relationship between the three variables in Table 40.3. Last, the
interpret process is about students’ understanding of the trends of expenditure and
numbers of child grant beneficiaries – including what could happen beyond 2013.
In addition, students can also calculate correlations between number of beneficia-
ries and child expenditures per year. What could be interesting is to find out how
many students will attempt to derive a formula which shows the relationship
between the two variables – assuming that it is linear, which could be used to predict
future growths of the two variables. During the modelling process, through the use
of technology, cognitive apprenticeship offers a flexible learning environment where
students can develop lifelong skills. Cognitive apprenticeship also provides a frame-
work for feedback for Mathematical Literacy teachers by making students’ thinking
processes “visible” to the teachers through interactions (Knobelsdorf et al. 2014;
Lyons et al. 2017).
40.5 Conclusion
This chapter offered a comparison between the South African Mathematical Literacy
curriculum and the Mathematical Literacy definition given by the Organisation for
Economic Co-operation and Development (2013). While it is generally accepted
that Mathematical Literacy is a key literacy developmental area, and a possible
strategy for promoting widespread implementation of reforms in education
(Organisation for Economic Co-operation and Development 2013), Mathematical
Literacy implementation in many countries has been different. In South Africa, the
construct of mathematical modelling and the use of technology, though very criti-
cal, are not used in the teaching and learning of Mathematical Literacy. Mathematical
modelling and technology as instructional models are important because they
address the increasing demand from employers of students entering the world of
work to possess “new kinds of mathematics knowledge that are shaped by systems
that govern their work” (Noss and Hoyles 2010, p. 76). It is perhaps worth noting
that in the high-stakes examination environments currently prevailing in South
Africa, it is a challenge to implement mathematical modelling and the use of tech-
nology in the day-to-day teaching and learning of Mathematical Literacy. However,
I argue that if the use of technology and mathematical modelling cannot be done in
476 D. Mhakure
high-stakes examination environments, then project-based learning should be

implemented as a parallel instructional model (Antonius 2007).
References
Antonius, S. (2007). Modelling based project examination. In W. Blum, P. L. Galbraith, H.-W. Henn,
& M. Niss (Eds.), Modelling and applications in mathematics education (pp. 409–416).
New York: Springer.
Bardini, C. (2015). Computer-based assessment of mathematics in PISA 2012. In K. Stacey &
R. Turner (Eds.), Assessing mathematical literacy (pp. 173–188). Cham: Springer International.
Best, J. (2008). Beyond calculation: Quantitative literacy and critical thinking about public issues.
In B. L. Madison & L. A. Steen (Eds.), Calculation vs. context: Quantitative literacy and its
implications for teacher education (pp. 125–135). Washington, DC: Mathematical Association
of America.
Bouta, H., & Paraskeva, F. (2013). The cognitive apprenticeship theory for the teaching of math-
ematics in an online 3D virtual environment. International Journal of Mathematical Education
in Science and Technology, 44(2), 159–178.
Caspersen, M. E., & Bennedsen, J. (2007, September). Instructional design of a programming
course: a learning theoretic approach. In Proceedings of the third international workshop on
computing education research (pp. 111–122). ACM. https://doi.org/10.1145/1288580.1288595.
Chapman, A., & Lee, A. (1990). Rethinking literacy and numeracy. Australian Journal of
Education, 34(3), 277–289.
De Lange, J. (2003). Mathematics for literacy. In B. L. Madison & L. A. Steen (Eds.), Quantitative
literacy: Why numeracy matters for schools and colleges (pp. 75–89). Princeton: The National
Council on Education and the Disciplines.
Department of Basic Education. (2011). Curriculum and assessment policy statement: National
curriculum statement – Mathematical Literacy, Further Education and Training phase grades
10–12. Pretoria: Republic of South Africa.
Frejd, P., & Geiger, V. (2017). Exploring the notion of mathematical literacy in curricula docu-
ments. In G. A. Stillman, W. Blum, & G. Kaiser (Eds.), Mathematical modelling and applica-
tions: Crossing and researching boundaries in mathematics education (pp. 255–263). Cham:
Springer.
Gardiner, A. (2004). What is mathematical literacy? Paper presented at the ICME -10, 4–11.
Copenhagen.
Graven, M., & Buytenhuys, E. (2010). Mathematical literacy in South Africa: Increasing access
and quality in learners’ mathematical participation both in and beyond the classroom. In
B. Atweh, M. Graven, W. Secada, & P. Valero (Eds.), Mapping equity and quality in math-
ematics education (pp. 493–508). Dordrecht: Springer.
Greiff, S., Wüstenberg, S., Holt, D. V., Goldhammer, F., & Funke, J. (2013). Computer-based
assessment of complex problem solving: Concept, implementation, and application.
Educational Technology Research and Development, 61(3), 407–421.
Hughes-Hallett, D. (2001). Achieving numeracy: The challenge of implementation. In L. A. Steen
(Ed.), Mathematics and democracy, the case for quantitative literacy (pp. 93–98). Princeton:
The National Council on Education and the Disciplines.
Jürges, H., Schneider, K., Senkbeil, M., & Carstensen, C. H. (2012). Assessment drives learning:
The effect of central exit exams on curricular knowledge and mathematical literacy. Economics
of Education Review, 31(1), 56–65.
Knobelsdorf, M., Kreitz, C., & Böhne, S. (2014, March). Teaching theoretical computer science
using a cognitive apprenticeship approach. In Proceedings of the 45th ACM technical sympo-
sium on Computer science education (pp. 67–72). ACM.
Lyons, K., McLaughlin, J. E., Khanova, J., & Roth, M. T. (2017). Cognitive apprenticeship in
health sciences education: A qualitative review. Advances in Health Sciences Education, 22(3),
723–739.
Noss, R., & Hoyles, C. (2010). Modeling to address techno-mathematical literacies in work. In
R. Lesh, P. L. Galbraith, & A. Hurford (Eds.), Modeling students’ mathematical modeling com-
petencies (pp. 75–86). New York: Springer.
Organisation for Economic Co-operation and Development (OECD). (2013). PISA 2012 assess-
ment and analytical framework: Mathematics, reading, science, problem solving and financial
literacy. Paris: OECD Publishing. https://doi.org/10.1787/9789264190511-en.
Oriol, M. D., Tumulty, G., & Snyder, K. (2010). Cognitive apprenticeship as a framework for
teaching online. MERLOT Journal of Online Learning and Teaching, 6(1), 210–217.
Stacey, K., & Turner, T. (2015). The evolution and key concepts of PISA mathematics frame-
works. In K. Stacey & R. Turner (Eds.), Assessing mathematical literacy: The PISA experience.
New York: Springer.
Venkatakrishnan, H., & Graven, M. (2006). Mathematical literacy in South Africa and func-
tional mathematics in England: A consideration of overlaps and contrasts. Pythagoras,
64(Dec), 14–28.
Chapter 41
A Lecturer’s Learning Goals for Teaching
Mathematical Modelling
Kerri Spooner
Abstract This chapter presents the preliminary findings to the research question:
What student learning goals does a lecturer have for his university first year math-
ematical modelling course? Hypothetical learning trajectories were adopted as a
framework for the study. Data were collected from interviews and teaching observa-
tions with one mathematical modelling lecturer. The results show that the lecturer
had clear student learning goals in place for teaching mathematical modelling to
solve real-world problems. These include exposing students to modelling tech-
niques and processes side by side and providing opportunities for students to model.
The importance of the lecturer and some approaches taken within other tertiary
modelling courses will also be discussed.
Keywords Hypothetical learning trajectory · Lecturer · Learning goals ·

Mathematical modelling · Teaching mathematical modelling · Tertiary
modelling course
41.1 Literature Background
With the establishment of the teacher/lecturer as an important variable in the learn-

ing of mathematical modelling (Blum 2011; Blum and Leiß 2007; Crouch and
Haines 2004; Lipowsky 2006; Niss 2001; Soon et al. 2011), a doctoral study is cur-
rently being undertaken by the author investigating how lecturers create mathemati-
cal modelling learning experiences for students at tertiary level. Part of the creation
of a learning experience is the establishment of purposeful learning goals (Stronge
2018). This chapter addresses the importance of the lecturer and presents a selection
of approaches taken within tertiary modelling courses. Preliminary findings for the
K. Spooner (*)
School of Engineering, Computer and Mathematical Sciences, Mathematical Sciences
Department, Auckland University of Technology, Auckland Central, New Zealand
e-mail: kspooner@aut.ac.nz

480 K. Spooner
research question: What student learning goals does a lecturer have for his univer-
sity first year mathematical modelling course? will be presented and discussed.
Central to the teaching and learning of mathematics and mathematical modelling
is the teacher (Blum 2011; Lipowsky 2006). Lipowsky’s (2006) meta-analysis of
factors affecting learning showed that the teacher is the most important variable
when learning mathematics, with the manner in which the subject is taught being a
key factor in student learning outcomes. Blum (2011) supports this view for math-
ematical modelling, believing that the decisive variable for successful learning of
modelling is the teacher and the quality of his/her teaching. From reviewing earlier
existing research, Niss (2001) concluded that students’ capabilities with modelling
tasks were influenced by the teaching approach. This supports the assertion that the
teacher plays a crucial role in mathematical modelling. Evidence from studies
points towards teacher delivery, at both school and university, having an influence in
removing or creating student difficulties with modelling (Blum 2011; Blum and
Leiß 2007; Crouch and Haines 2004; Soon et al. 2011). Research is beginning to
reveal teaching strategies for modelling, although “it is still disputed how to inte-
grate mathematical modelling into the teaching and learning processes” (Kaiser
2014, pp. 396). Some of these strategies include discussion cycles (Legé 2005;
Spooner 2012), reflection on approaches (Soon et al. 2011; Stillman et al. 2010) and
strategic interventions (Blum and Leiß 2007; Kaiser and Stender 2013).
There has been a variety of different approaches implemented and suggested for
tertiary modelling courses. Some of these include the use of case studies (Fu and
Xie 2013; Heilio 2011), simulated work placement experiences (Alpers 2011),
guided modelling projects (Blomhøj and Kjeldsen 2013) and open-ended modelling
projects (Caron and Bélair 2007). Heilio (2011) suggested the use of case studies to
establish links between the real world and mathematics and to demonstrate the
range and complexity of different model types. Fu and Xie (2013) used student-
relevant problems where first year students thought about the problem themselves,
followed by lecturer led class discussion of student-generated ideas and concluded
with the presentation of different well-established models for the problem. A final
year course at a German university used industry-type problems to simulate tasks
highly likely to be encountered in an engineering workplace (Alpers 2011). These
tasks are carefully designed to enable students to apply and understand models
learnt in class in a simulated real-life problem. Supplementary to a modelling first
year course in Denmark, a guided modelling project was used to develop mathemat-
ical concepts as well as modelling processes (Blomhøj and Kjeldsen 2013). Caron
and Bélair (2007) carried out research on a final year modelling course using open-
ended projects. It was found that for students to have successful modelling experi-
ences with these types of projects, time spent discussing the purpose of the model
in conjunction with explicit teaching of, and frequent alluding to, the modelling
process is needed within the course.
According to Blum (2011, 2015), there are teacher qualities and certain situa-
tions that promote development of modelling competency in students. A teacher
needs to be able to give appropriate support, balancing this between minimal teacher
guidance and maximum student independence. The use of a broad variety of
41 A Lecturer’s Learning Goals for Teaching Mathematical Modelling 481
examples and contexts is necessary to make transfer of situations and domains eas-
ier for students. A teacher’s openness to promote and accept multiple solutions is
key, along with the ability to support individual student modelling routes. These
characteristics support and endorse Blum’s (2011, 2015) belief that modelling is
learnt by doing. This view is also recommended by Blomhøj and Kjeldsen (2006)
promoting context-based project work as one situation for students to actively be
modelling. Blomhøj and Kjeldsen (2006) recommend a scaffolded approach with
students eventually taking responsibility for the entire modelling process. For suc-
cessful learning experiences, teachers need some personal modelling experience
and knowledge of how to best support students through modelling sub-processes
(Blomhøj and Kjeldsen 2006; Blum 2011, 2015). Effective learning experiences are
ones where teachers focus on learning objects and goals for student learning as the
initial part of their planning (Blomhøj and Kjeldsen 2006).
What learning goals do mathematical modelling lecturers in engineering have for
their first year students? To help answer this question, a case study involving a lec-
turer of a New Zealand university first year engineering mathematical modelling
course was carried out using design research (Gravemeijer and van Eerde 2009) to
establish the hypothetical learning trajectory learning goals of the lecturer.
41.2 esign Research and Hypothetical

D
Learning Trajectories
Design research is an approach taken within Realistic Mathematics Education that

allows for the development of domain-specific theory (Gravemeijer and van Eerde
2009). Design research aims to develop a local instructional theory about how stu-
dents learn a particular topic and how best to support this learning process
(Gravemeijer and van Eerde 2009). Design research uses a hypothetical learning
trajectory as the lens for collecting data. A hypothetical learning trajectory consists
of a set of learning goals, planned instructional activities and a conjectured learning
process (Simon 1995).
A hypothetical learning trajectory is normally developed by a teacher and/or
researcher based on the current understanding of curriculum content, recommenda-
tions for teaching the mathematical topic (Andrews-Larson et al. 2017; Bakker and
van Eerde 2015; Scott-Wilson 2014), known student difficulties for the chosen con-
tent area (Andrews-Larson et al. 2017; Bakker and van Eerde 2015) and learning
characteristics of the student body (Scott-Wilson 2014). This hypothetical learning
trajectory is about learning the process of mathematical modelling as learning con-
tent in its own right (Galbraith 2012).
Hypothetical learning trajectories more commonly exist for primary and second-
ary school mathematics topics than for undergraduate mathematics areas (Andrews-
Larson et al. 2017). Due to the different culture of tertiary education compared to
primary and secondary education, it is unsurprising that there are limited examples
482 K. Spooner
of hypothetical learning trajectories in undergraduate mathematics (Andrews-

Larson et al. 2017). Indeed, it is highly likely that many tertiary educators are
unaware of the hypothetical learning trajectory concept itself. This was the case for
the lecturer participant in this study. Due to this, and data collection occurring after
the lecturer had planned his lessons, the hypothetical learning trajectory was estab-
lished retrospectively by deductively analysing interview transcripts and lecturer
observations for evidence of the components of the lecturer’s hypothetical learning
trajectory.
41.3 Case Study
An exploratory case study (see Yin 1981 for method) was conducted in New Zealand
to provide insight into how a lecturer creates student learning experiences in math-
ematical modelling. The research aim was to establish the hypothetical learning
trajectory learning goals of the lecturer for teaching the process of mathematical
modelling. The context in which these hypothetical learning trajectory learning
goals are developed involves a first year summer school university mathematical
modelling course designed for engineering students. The lecturer has received four
excellence in teaching awards recognising him as an expert lecturer of engineering
and mathematical modelling courses. He is particularly interested in developing
modelling competencies; views modelling as holistic, that is encompassing the full
process of modelling (Haines et al. 2003); and runs full-day modelling competitions
for secondary school students. The course runs for 5 weeks with 2-h classes four
times per week. Each class is a blend of lecture and tutorial. An assessed open-
ended modelling project is undertaken by students grouped in fours. The 1-day proj-
ect occurs after the last class and before the final examination. Seven hours is
allocated to the project, allowing the modelling process to be worked through. At
the end of the 7 h, a written report is submitted showing the model produced and the
outline of the process students have gone through to produce the model. The project
is worth 5% of the students’ final grade. An assessed pre-modelling project tutorial
is held to prepare students for the project.
41.4 Data Collection and Analysis
Data were collected through video recordings and field notes of a lecturer interview
and a mathematical modelling tutorial. The lecturer interview contained questions
to enable comprehension of the participant’s philosophy for teaching mathematical
modelling and to begin to establish the participant’s hypothetical learning trajectory
for the modelling component of the course. Sample interview questions included:
• What is the objective of the mathematical modelling course you teach?

• What do you think are the key points for teaching formulating a mathemati-
cal model?
• What are your ideas/theories about how students learn to formulate a mathemati-
cal model?
• Can you define what your student learning goals are for the lecture(s)?
• What is your planned instructional delivery for the lecture(s)?
• What is your vision for the learning process the students will take part in?
To begin to establish the lecturer’s hypothetical learning trajectory, the data were
coded both inductively, with codes derived from the data, and deductively, with
specific codes defined before coding (Braun and Clarke 2006). Both descriptive
codes, which summarise the primary topic, and analytical codes, which capture
deeper meaning (Gibbs 2007), were used for the inductive coding process. Learning
goals, instructional activities and learning processes were used as the predeter-
mined deductive codes. Subsequently, the learning goals were re-examined for
clues on what these meant for learning and how the lecturer knew that learning had
taken place.
The emerging themes that captured important notions in relation to the research
question (Braun and Clarke 2006) were solving real-world problems; modelling
process; mathematical techniques; modelling difficulties; teaching strategies; mod-
elling strategies; modelling preparation; course structure; timing; group work; and
balance. These have been verified by an independent researcher.
41.5 Preliminary Findings
The first stages of establishing the lecturer’s hypothetical learning trajectory have
resulted in some initial learning goals. His learning goals can be classified into three
different areas: general learning goals for the course, learning goals for the pre-
modelling project tutorial and learning goals for modelling day project. This chap-
ter presents the preliminary findings relating to these and what they mean for student
learning.
41.5.1 General Learning Goals for the Course
The overarching theme is solving real-world problems. To enable this to be met, the
following goals were in place:
484 K. Spooner
41.5.1.1 Develop Techniques for Modelling
The application of techniques is at the heart of forming a model. However, the use
of techniques on their own is not modelling. According to the lecturer, “You’ve got
the mathematical techniques on one hand and then you’ve got the modelling”. He
illustrates the difference between mathematical techniques and modelling, stating
that techniques are “how to get the equations…how to solve those equations” and
modelling is the use of “those techniques for solving real-world problems and to
feed into the modelling process”. Without techniques to draw on, students would
flounder to create models. According to the lecturer, techniques are implied to be
mathematical techniques, methods and tools. Specific tools mentioned were appli-
cation of proportionality relationships, dimensional analysis, optimisation, rates of
change, physical laws including Newton’s second law and curve fitting.
41.5.1.2 Acquire Knowledge of the Process of Mathematical Modelling
The course is embedded within “the context of the modelling process and how these
techniques [modelling techniques] feed into that” with the modelling process inter-
woven throughout the course. This enables continual student exposure to the com-
plete process of modelling, an approach known as holistic modelling (Haines et al.
2003). The lecturer believes that students would find it difficult to model without
first observing that modelling is taking place. “You can’t just throw them into it. I’ve
got to model it enough so they can see what’s going on”. As well as role-modelling
the process, the lecturer acknowledges the importance of allowing students to expe-
rience the process themselves with statements such as “asking them [the students]
to do parts of the process along the way” and “so they have got to see it done and
then they have to get a chance to do it”. Embedding the course in the “process” and
giving students a full-day experience of modelling developed student confidence in
their own modelling ability.
41.5.1.3 Engage in Independent and Critical Thinking
The development of critical and independent student thinking as a hypothetical

learning trajectory learning goal was evident from the lecturer’s comments such as
“I am not going to be chipping in on this” and “I want them to realise that they
shouldn’t just get an answer and stop”. Nurturing the skills of independence and
critical thinking develops students’ confidence and self-assurance. In the context of
modelling, this is the ability of a student to find the right tools for a problem and the
capability to determine if they are using the tools correctly or not.
41.5.2 Learning Goals for the Pre-modelling Project Tutorial
These learning goals all relate to student preparation for the modelling project day.
41.5.2.1 Become Familiar with Skills for Effective Group/Teamwork
The lecturer expressed concerns about the inexperience of students with group work,
stating “if they can’t work on the day [in a group] they are not going to accomplish”.
The lecturer intimated that students can potentially achieve more in an effectively
working group, helping individual students to not “get bogged down”. To access the
benefits of group work desired for the project day, a learning goal for the tutorial was
for students to become familiar with the know-how of group work, including roles
within a group. Interesting to note, one of the main group work benefits perceived by
the lecturer was the effective use of time. He said, “if I ran this as an individual proj-
ect, I’d have to give them a week…by putting them in a group they can achieve a lot
more”. The project is run as a group “because modelling is difficult and time-con-
suming”. This makes the process of modelling more accessible to students.
41.5.2.2 Understand the Criteria for Assessment of Modelling Project
Understanding of assessment criteria can significantly improve student learning and

preparation. Discussion of the assessment criteria was used by the lecturer as a
learning experience to “recap the modelling process” and to explain how the group
work would be assessed. An apparent concern of the lecturer was anxiety of stu-
dents concerning group work. He stated “they are very hesitant about being assessed
in a group”.
41.5.3 Learning Goals for Modelling Day Project
41.5.3.1 Relate Mathematics to Genuine Contexts
By applying “the modelling process to a real-life problem, not just an in-class exam-
ple”, students learn to relate mathematics to genuine contexts, according to the lec-
turer. Students get to experience “this messy process, that is modelling”.
486 K. Spooner
41.5.3.2 articipate in the Modelling Process as a Group, Independent

P
of the Lecturer
One of the benefits of group work is learning to utilise each other’s strengths. “The
group is going to have different strengths and different skills, different background
knowledge and [individuals will benefit by] being able to draw on this”. Learning
outcomes for students are development of confidence and ability in moving through
the process of modelling. To work in a group also mimics the behaviour of research
and industry modellers.
41.5.3.3 Communicate Thought Processes
Clear communication of thought process is a valued modelling skill. “Graphs, dia-

grams and that kind of thing” can help to structure and communicate ideas when
you are modelling. Being able to accurately convey ideas enables students to work
collaboratively and to share their ideas and creations with a wider audience.
It is recognised that the teacher is an important variable in the learning of mathemat-

ical modelling in an educational setting (Blum 2011; Soon et al. 2011) with effec-
tive learning experiences having teacher-determined learning goals underpinning
them (Blomhøj and Kjeldsen 2006). There is a known connection between these
learning goals and learning experiences produced (Stronge 2018). A global learning
goal for mathematical modelling teaching is to develop modelling competency
(Blum 2011). Research by Blomhøj and Kjeldsen (2013), Caron and Bélair (2007),
Fu and Xie (2013) and Heilio (2011) revealed some approaches taken at tertiary
level designed to provide opportunities to meet this goal. A discussion concerning
the subsidiary learning goals to meet the global learning goal that underpins the
teaching approaches appears to be lacking.
This study identifies a lecturer’s subsidiary goals, as part of a hypothetical learn-
ing trajectory, which inform his approach to teaching modelling competency. The
lecturer’s teaching approach to utilising these goals involves demonstrating the
modelling process during lectures and using an open-ended project to develop stu-
dent independence with modelling. This approach meets the lecturer’s learning
goals by providing opportunities for students to first observe, and then participate,
by using techniques and processes for modelling. Soon et al. (2011) and Haines
et al. (2003) support this approach, and Caron and Bélair (2007) recommended that
students will be more successful with open-ended projects when the modelling pro-
cess is explicitly taught and frequently alluded to throughout a course. As the proj-
ect is limited to 1 day in length, the lecturer’s approach differs from those
recommended by Caron and Bélair (2007) and Blomhøj and Kjeldsen (2013), who
suggested a more scaffolded approach for projects undertaken over a longer period
of time. Support suggested for longer projects includes guided instruction in the
form of checkpoints (Caron and Bélair 2007) or mandatory course time lecturer
consultations (Blomhøj and Kjeldsen 2013). Caron and Bélair (2007) and Blomhøj
and Kjeldsen (2013) also use their approaches to develop mathematical competen-
cies alongside modelling competencies. This differs from the way the lecturer uses
his project, which is to provide opportunities for students to independently learn
modelling by doing (Blomhøj and Kjeldsen 2013; Blum 2015), and it demonstrates
the openness of the lecturer to multiple solutions, both of which are characteristics
of quality modelling teaching (Blum 2015).
The results of this case study show that this lecturer has some clear supportive
student learning goals for the process of mathematical modelling. By teaching
together modelling processes and techniques for modelling, his students observe
and participate in a comprehensive experience of modelling. The inclusion in the
course of the modelling project day creates an opportunity for students to have an
autonomous experience of modelling. The anticipated student learning is that they
gain an appreciation of the process and the need for techniques, development of
group work skills and familiarisation with real-world modelling. Student under-
standing of assessment criteria is also an important learning goal for this lecturer.
Students should leave the course having utilised the modelling process and the tech-
niques associated with it.
References
Alpers, B. (2011). The mathematical expertise of mechanical engineers: Taking and processing
measurements. In G. Kaiser, W. Blum, R. Borromeo Ferri, & G. Stillman (Eds.), Trends in
teaching and learning of mathematical modelling (pp. 445–456). Dordrecht: Springer.
Andrews-Larson, C., Warro, M., & Zandieh, M. (2017). A hypothetical learning trajectory for
conceptualizing matrices as linear transformations. International Journal of Mathematical
Bakker, A., & van Eerde, D. (2015). An introduction to design-based research with an exam-
ple from statistics education. In A. Bikner-Ahsbahs, C. Knipping, & N. Presmeg (Eds.),
Approaches to qualitative research in mathematics education: Examples of methodology and
methods (pp. 429–466). Dordrecht: Springer.
Blomhøj, M., & Kjeldsen, T. (2013). Students’ mathematical learning in modelling activities. In
Blum, W. (2011). Can modelling be taught and learnt? In G. Kaiser, W. Blum, R. Borromeo Ferri,
& G. Stillman (Eds.), Trends in teaching and learning of mathematical modelling (pp. 15–30).
do? In S. J. Cho (Ed.), The Proceedings of the 12th International Congress on Mathematical
Education (pp. 73–96). Cham: Springer.
488 K. Spooner
Blum, W., & Leiß, D. (2007). How do students and teachers deal with modelling problems. In
C. Haines, P. Galbraith, W. Blum, & S. Khan (Eds.), Mathematical modelling ICTMA 12:
Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in
Psychology, 3(2), 77–101.
Caron, F., & Bélair, J. (2007). Exploring university students’ competencies in modelling. In
C. Haines, P. Galbraith, W. Blum, & S. Khan (Eds.), Mathematical modelling ICTMA 12:
Crouch, R., & Haines, C. (2004). Mathematical modelling: Transitions between the real world
and the mathematical model. International Journal of Mathematical Education in Science and
Technology, 35(2), 197–206.
Fu, J., & Xie, J. (2013). Comparison of mathematical modelling skills of secondary and tertiary
students. In G. A. Stillman, G. Kaiser, W. Blum, & J. P. Brown (Eds.), Teaching mathematical
Galbraith, P. (2012). Models of modelling: Genres, purposes or perspectives. Journal of
Mathematical Modelling and Applications, 1(5), 3–16.
Gibbs, G. (2007). Analyzing qualitative data. Los Angeles: Sage.
Gravemeijer, K., & van Eerde, D. (2009). Design research as a means for building a knowledge
base for teachers and teaching in mathematics education. The Elementary School Journal,
109(5), 510–524.
Haines, C., Crouch, R., & Fitzharris, A. (2003). Deconstructing mathematical modelling:
Approaches to problem solving. In Q. Ye, W. Blum, S. Houston, & J. Jiang (Eds.), Modelling
and mathematics education and culture (pp. 41–53). Chichester: Horwood.
Heilio, M. (2011). Modelling and the educational challenge in industrial mathematics. In G. Kaiser,
W. Blum, R. Borromeo Ferri, & G. Stillman (Eds.), Trends in teaching and learning of math-
ematical modelling (pp. 479–488). Dordrecht: Springer.
Kaiser, G. (2014). Mathematical modelling and applications in education. In S. Lerman (Ed.),
Encyclopedia of mathematics education (pp. 396–404). Dordrecht: Springer.
Kaiser, G., & Stender, P. (2013). Complex modelling problems in co-operative, self-directed learn-
ing environments. In G. A. Stillman, G. Kaiser, W. Blum, & J. P. Brown (Eds.), Teaching math-
Legé, J. (2005). Socrates meets the 21st century. Teaching Mathematics and its Applications,
24(1), 29–36.
Lipowsky, F. (2006). Auf den Lehrer kommt es an. Empirische Evidenzen für Zusammenhänge
zwischen Lehrerkompetenzen, Lehrerhandeln und dem Lernen der Schüler [It depends on
the teacher. Empirical evidence for the relations between teacher competence, teacher behav-
ior, and student learning]. In C. Allemann-Ghionda & E. Terhart (Eds.), Kompetenzen und
Kompentenzentwicklung von Lehrerinnen und Lehrern: Ausbildung und Beruf (pp. 47–70).
Weinheim: Beltz.
Niss, M. (2001). Issues and problems of research on the teaching and learning of applications and
modelling. In J. Matos, S. Houston, W. Blum, & S. Carreira (Eds.), Modelling and mathemat-
ics education: ICTMA-9 applications in science and technology (pp. 381–390). Chichester:
Horwood.
Scott-Wilson, R. (2014). An analysis of learning characteristics, processes, and representations in
mathematical modelling of middle school learners with special needs (Unpublished doctoral
thesis). Stellenbosch: Stellenbosch University.
Simon, M. A. (1995). Reconstructing mathematics pedagogy from a constructivist perspective.
Journal for Research in Mathematics Education, 26, 114–145.
Soon, W., Tirtasanjaya Lioe, L., & McInnes, B. (2011). Understanding the difficulties faced by
engineering undergraduates in learning mathematical modelling. International Journal of
Mathematical Education in Science and Technology, 42(8), 1032–1039.
Spooner, K. (2012). A glimpse of reality (Unpublished master’s thesis). Auckland: University of
Auckland.
Stillman, G., Brown, J., & Galbraith, P. (2010). Identifying challenges within transition phases
of mathematical modeling activities at year 9. In R. Lesh, P. L. Galbraith, C. R. Haines, &
A. Hurford (Eds.), Modeling students mathematical modeling competencies (pp. 385–398).
New York: Springer.
Stronge, J. (2018). Qualities of effective teachers. Washington, DC: ASCD.
Yin, R. K. (1981). The case study as a serious research strategy. Knowledge, 3(1), 97–114.
Part VI
Examples of Modelling and Applications
in Practice
Chapter 42
Embodied Phenomenology
in Mathematical Modelling of Sailing
for Integrated STEM Learning
Celil Ekici and Cigdem Alagoz
Abstract Mathematical modelling is under-investigated in complex learning envi-

ronments where embodied knowledge is a natural and essential part of learning,
such as sailing. Mathematical modelling is used here as an ongoing process in build-
ing and testing models by the participants to make sense of their sailing practices.
We demonstrate how a culturally responsive modelling practice can build on learn-
ers’ experiences, their funds of knowledge, and existing practices as young sailors.
Participants’ embodied knowledge of sailing is used in modelling of tacking to
travel towards the wind. Participants engage in experimental and active mathemati-
cal modelling of sailing in its own lifeworld in this integrated STEM learning con-
text. We demonstrate and discuss how an embodied phenomenology approach
works as learners go through modelling phases from understanding the tacking
problem to validating their models.
Keywords Aerodynamics · Culturally responsive pedagogy · Embodied

phenomenology · Funds of knowledge · Interdisciplinary · Mathematical
modelling · STEM learning · Hydrodynamics · Sailing
C. Ekici (*)
Department of Mathematics and Statistics, Texas A&M University-Corpus Christi,
Corpus Christi, TX, USA
e-mail: celil.ekici@tamucc.edu
C. Alagoz
Division of Research and Analysis, Texas Education Agency, Austin, TX, USA

494 C. Ekici and C. Alagoz
42.1 Introduction
Mathematics education research has not fully attended to the ways in which math-
ematical modelling and culturally responsive pedagogy can be united (Anhalt et al.
2018). Building on Ladson-Billings (1995), cultural responsive pedagogy requires
teaching and learning mathematical modelling to be grounded in students’ lives and
their experiences in their communities (Greenstein and Ekici 2017). Many young
islanders, for example, already have embodied knowledge of sailing that can be
used as funds of knowledge to access and build motivation towards mathematical
perspectives connected to their lives (Moll et al. 1992). With a culturally responsive
pedagogy lens, lived or living mathematics in the local sailing practices can be
examined via mathematical modelling. As pointed out by Stender and Kaiser
(2016), modelling competencies can best develop when students engage with com-
plex situations where they tackle modelling of realistic problems.
While studies have investigated mathematical modelling in the workplace (Frejd
and Bergsten 2016; Noss and Hoyles 1996), informal and complex learning envi-
ronments have been little studied with relevant modelling practices. Mathematical
modelling processes with embodied phenomenology take an active and experimen-
tal approach, drawing on modellers’ funds of knowledge and experiences incorpo-
rating their prior and ongoing embodied sense and knowledge of sailing, for
example, which they gained individually or as a community of practice. Research is
lacking where modellers have prior embodied knowledge before they study a phe-
nomenon with mathematical modelling but not non-existent (see, e.g. Fernandes
and Matos 1998). Such contexts include sailing, tennis, figure skating, dancing, and
basketball, among others.
Sailing practice is full of rich problems for mathematical modelling such as tack-
ing which is investigated in this chapter (Anderson 2008). Sailboats can travel
towards wind but not directly into the wind. The tacking manoeuvre is used to chart
an upwind course in zigzag patterns by turning a boat’s head through the wind as
seen in Fig. 42.1a. The diagram in Fig. 42.1b depicts variations on the tacking
manoeuvres to sail from A to B where the wind is blowing from B to A. Only three
paths are depicted in Fig. 42.1b, but there are infinitely many variations, still all
under the assumption that the wind direction is not changing. For tacking, the angle
of attack here is shown as 45°, which is common for sailboats, yet it can vary.
The coordinated manipulation of objects such as sails, body, boat, and wind
direction is a part of the repertoire of a sailor’s knowledge during sailing manoeu-
vres. Perception of a situation from the learner’s perspective is not a passive recep-
tion of information but instead implies activity and bodily movement with an end in
mind (Husserl 1970). Achieving and maintaining the functional coordination of
actions is the ultimate learning motive for Dewey (1935/1987) during an inquiry
where each goal “marks out milestones but does not show the road” (p. 511). We
build on Dewey in our study of inquiry-based modelling of sailing, as it requires a
purposeful coordination of tools as means to an end (Dewey 1916/1980).
42 Embodied Phenomenology in Mathematical Modelling of Sailing for Integrated… 495
Fig. 42.1 Tacking

manoeuvre to sail upwind
where angles for turns and
lengths for runs can vary as
seen in (a) and (b). (b)
shows alternative tacking
scenarios from A to B
42.2 Problem and Theoretical Framework
Towards building culturally responsive mathematical modelling practices with a

community, this study taps into the local funds of knowledge of sailing where learn-
ers have prior embodied knowledge in practice. A phenomenological approach is
applied for living/lived mathematical modelling of sailing through its embodiments
(Roth 2010; Roth and Thom 2009). Embodied phenomenology in mathematical
modelling incorporates tacit knowing in the modelling process and is an innovation
in research about teaching and learning mathematical modelling.
We are seeking to understand how embodied knowing plays a role in mathemati-
cal modelling of tacking. Mathematical modelling processes include the phases of
understanding the task, simplifying/structuring, mathematising, working mathemat-
ically, interpreting, and validating (Blum and Leiß 2007). The validation stage also
requires models to be translated and interpreted in context with their embodiments
(Watson 2002). We investigate how embodied phenomenology aligns with learners’
experience with modelling stages when learners have living/lived embodied knowl-
edge of sailing.
The phenomenological approach is descriptive of a lifeworld in its own terms
(Husserl 1970) to understand and interpret the meanings of sailing actions and prac-
tices as experienced. Recent research indicates that tacit knowing in sports such as
sailing is not merely a collection of motor skills, but rather the grasping, in the
practical embodied way, of physical principles (Andersson et al. 2015). Sailing
requires sustaining a functional alignment with boat, water, body, and sails in coor-
dination with a series of bodily actions towards a goal.
As the practitioners manipulate their objects (body position, sail position, sail
shape, tiller position, gazing the angle of attack to the wind) which they consider
essential to a tacking manoeuvre, they experience instances of tacking in different
ways with each configuration. Phenomenologically, the tacking is not just the expe-
rience with a set of objects; it emerges from the multiplicity among configurations
as experienced with the bodily objects from tacking attempts. Each attempt pro-
vides the ground for a subsequent understanding of the tacking concept. Tacking is
an embodied process where the connection between multiple (perceptual) experi-
ences that your senses, including your body, realize to occur as a result of bodily
movements. Embodied perception of tacking extracts sensory information from the
environment by touching and manipulating. The practitioners therefore know the
tack as such precisely at the moment when they can anticipate through sense impres-
sions upon moving and manipulating (their bodies, gaze, sails) for the desired
results (Merleau-Ponty 2012; Roth and Thom 2009).
In a sailing practice, the sailors often struggle to describe how and why they
execute a series of bodily actions during sailing. Focusing on the lifeworld of young
sailors, we aim to build on their lived/living experiences during the mathematical
modelling process to facilitate reflective representations of mathematics behind
their intentional practices and assumptions during sailing acts such as tacking.
Tacking experience involves intentional coordination of sailing objects such as tiller
to control angle of attack, shape of sails, and positioning of their body on the sail-
boat. Modelling knowledge for sailing is acquired through embodied learning where
learners focus on their decisions during manoeuvres and critically represent a model
to express how it works and use the model to anticipate and validate how it works
across its different instances. With regard to mathematical modelling competencies
(Blum and Leiß 2007), embodied phenomenology helps to connect the processes of
understanding the task, simplifying/structuring, interpretation, and validation as
critical competences involved in complex sailing situations. Mathematising and
working mathematically through multiple representations are part of debriefing and
briefing sessions in connection with the experiments. Students develop conjectures
about how their initial models would work in practice, devise an experiment, collect
data, apply and interpret their models in context, and validate or modify the model
as needed.
42.3 Method and Data
The embodied phenomenology approach to sailing is adopted here with our inquiry-
based modelling of sailing sessions with junior sailors in an informal learning envi-
ronment at a local sailing club with and without a sailboat, in and out of the sea. We
observed that mathematical modelling of sailing provides learners with motivation
towards integrating STEM knowledge involved in the mathematics and science
behind the sailing. Tapping into their rich funds of knowledge of sailing, we wanted
to help junior sailors focus on how tacking works and why it works with a proper
understanding of mathematical and scientific models underpinning the accom-
plished practices. Similar to complex modelling problems’ training by Stender and
Kaiser (2016), here we provide after-school sessions on mathematical modelling for

junior sailors.
Through the use of embodied ways of knowing, participants engage in a cycle of
sailing movement experience, reflection, formulation of alternative movement/
ideas, and further experimentation, as suggested by the experiential embodied learn-
ing cycle (Kolb 1984; Sheets-Johnstone 2011). Participants produce descriptions,
conjectures, explanations, and some embodied procedures while modelling their
sailing practices mathematically. The mathematical modelling process is used to
examine mathematics in action with interdisciplinary connections in their practice
of sailing and as an embodied knowledge building tool. We build modelling sce-
narios based on participants’ existing practices such as tacking during sailing, as
they would experience and observe them. We work with the participants first to
extract their perceived models of tacking as conjectures to be justified, modelled,
and validated with their corresponding embodiments during sailing. These conjec-
tures require the mathematical modelling of sailing. This is followed by experimen-
tal sessions at sea with a two-person sailboat. As the learner expresses his/her
thinking by talking and manipulating the tools, this critical reflection informs the
development and refinement of the mathematical models. Vygotsky (1986) and later
Sfard and Kieran (2001) suggested that speaking and communication not only
express thinking in words but also constitute a simultaneous development of under-
standing. Therefore, we adopted inquiry-based one-to-one discussion sessions with
participants as a part of our embodied phenomenological approach to mathematical
modelling of sailing manoeuvres in practice.
We anticipate that learners will develop more sense of the mathematics behind
sailing as they examine and develop mathematical models of tacking. During this
process, the learner goes through problem formulation, representation, experimen-
tation, testing their models, and revising of their assumptions.
42.3.1 Setting and Participants
This experimental study on mathematical modelling of sailing has been part of a

larger effort towards building culturally responsive STEM learning with communi-
ties and peers using locally relevant phenomenon and practices funded by a federal
grant. The participants here are junior sailors. For 10 weeks during Spring 2017, one
junior sailor, Adam, regularly teamed up with the first author in a 420 sailboat fol-
lowing the pre-session modelling discussions on sailing before experimenting and
observing their models at sea. Our key participant, Adam, is a junior in high school
who has been sailing and competing nationally for 3 years. There were three din-
ghies (420 s or similar) each manned by two other young sailors, which concur-
rently practised and provided extended opportunities for live observations and
discussions. In collaboration with his parents and his principal, Adam, as the key
participant, was given service credit and pursued project work related to modelling
of sailing, hence connecting his informal and formal experiences in and out
of school.
Inquiry-based interactive learning activities are designed for students in a
discussion-oriented setting to facilitate the mathematical modelling of critical sail-
ing manoeuvres such as tacking. The tacking manoeuvre is being modelled here
theoretically and experimentally with a two-person sailboat known as C420 (Galan
2017; The Club C420 Association 2018). The focus question is: “How does a sail-
boat move against the wind?” A cycle of briefing-experimentation-debriefing is
used in our sessions. During briefings, we conduct one-to-one discussion sessions
on modelling of sailing practices to consider the weekly modelling tasks, facilitate
the formulation of their conjectures, and explicate the connection with their relevant
embodied knowledge and mathematics. During the experimentation, participants
work on embodiments of their conjectures, examining lived and living practices by
themselves and others. The primary mode of experimentation is hands-on sailing,
hence enacting the tasks for modelling tacking by observing oneself and others in
the act. A second mode was to integrate technology using virtual simulators of sail-
ing for the tacking manoeuvres.
Data were collected over 3 months. During the 3 months, weekly sessions on
mathematical models of sailing manoeuvres such as tacking are developed. These
sessions build on scenario-based theoretical and experimental enquiry into the sail-
ing phenomena as conjectured, experienced, and interpreted by the participants.
Both researchers took notes and maintained an observation log. In addition, the
researchers developed semi-structured protocols for discussions and observations
shared with the participants ahead of the sessions.
42.4 Results and Discussions
One of the inquiry questions we present here is as follows: “What makes a sailboat
travel towards the wind?” This is an ill-posed modelling question. Far from simple,
this question provides the main motivation to guide our embodied mathematical
modelling process. The participants through their earlier trainings and experience
are able to say directly that you can sail towards the wind but not directly into it.
Adam suggested that you need to tack with an angle, depending on the boat. I asked
“How do you know the angle? How does it relate to the boat’s movement?” So
understanding the problem and the determination of variables are lengthy processes
before the quantification and mathematising processes. Adam proposed that for
tacking to occur, you need to have a stable wind and you must set your sails prop-
erly. He referred to the points of sails chart as in Fig. 42.2, which he knows from his
training. In sailing terminology, participants are quick to suggest that you cannot
sail directly into the wind called “irons” as shown in grey in Fig. 42.2. This was a
fund of knowledge Adam drew on to set the course and sails into the wind.
Working practically with Adam, we experimented with the impact of setting a
close reach which is about 45° from the north. After experimentation, in debriefing,
Fig. 42.2 Points of sail demonstrating the possible angles and sail directions
Fig. 42.3 Virtual

experimentation with
different angles of attacks
gave two examples of
forces on sails generated
by the computational fluid
dynamics applet
(Schroeder 2017)
Adam claimed that 30° does not work for the C420 boat as suggested in Fig. 42.2.
He explained that manipulating the tiller to turn, we gradually increase the turn until
we start picking up some speed at about 45°. He later further clarified that “if the
wind speed is higher, we are able to pick up some speed in Close Hauled position
with a smaller angle and tighter sails”. He added “if the wind is low, a thicker sail
shape generates more lift and driving force”. During the debriefing, wind speed was
identified as a clear variable for the angle of attack.
Adam described lift as a vertical push to the sails when the sail shape is properly
adjusted to look like an airfoil. He suggested that lift is formed on the sails when
you are going upwind in close hauled position. He expressed that there is also coun-
terpart drag along with lift, generating a force on the sails. He was able to describe
why the sail shape generates a push to the sails but was not able to connect this to
the forward motion of the boat.
Adam drew different sail shape configurations to generate lift, using a computa-
tional fluid dynamic applet (Schroeder 2017), which allowed him to test his conjec-
tures regarding the different angles, sail shapes, and the speed of wind into creating
a force on the sails. The question Adam remained pondering after this experiment
was how to connect and explain the forward driving force on the sailboat with the
force on the sails as depicted in Fig. 42.3.
Some of the conjectures we tested in water are “the flatter the sail, the smaller the
drag” and “the thicker sail shape, the less power in light winds”, to see their embodi-
ments. While Adam was figuring out the driving force, his intuitions indicated a
connection between sail area, lift force, and the wind. He experimented focusing on
sail areas with different sail shapes by easing the chords. Adam brought a quadratic
formula to the next session connecting the sail area and the lift:
1
F =C ρVA2 A
2
Here, the force (F) is the lift (L) for forces measured perpendicular to the airstream
to determine C for lateral force on a sail of area (A) and a given aspect ratio (height
to average cord width). These coefficients vary with angle of attack with the inci-
dent wind (VA for a mainsail).
During the experimentation at sea, Adam observed the embodiments of the
impact of the force on the sails. Towards further mathematising this relationship,
Adam stated “you want to maximise lift and minimise drag”. He wrote this as “the
ratio of lift over drag”. Adam worked next on the question how this generates the
forward motion for the boat. The force FT is decomposed into lift and drag, and then
the lateral force FLAT causing the boat to heeling over water and the driving force
FR. Their connection to FT representing the total force on the sails builds on the
trigonometry behind the rotation by α, which is the angle of attack. Synthesising the
virtual and embodied experiments, the variation of angles of attack produces a polar
map in sailor’s terminology as in Fig. 42.4 which Adam was already familiar with.
The resulting mathematical model explains FLAT and FR in terms of lift, drag, and the
angle of attach α as:
Fig. 42.4 As α angle of attack varies, the force on the sails changes as produced by lift and drag.
The driving force FR and the lateral heeling force FLAT are generated by the incident wind VA
FLAT = L cos α + D sin α FR = L sin α − D cos α
Adam observed that the same angle α is also between the lift and the driving force.
In Fig. 42.4, the desired driving force on the boat generates speed towards the wind
peaking around 30°. As Adam suggested, depending on the efficiency of the boat,
the map can vary from boat to boat. While it varies, Adam’s experience tells him
that a lift more aligned with the direction of travel increases the driving force and
decreases the lateral force.
These sessions ended with debriefing that our model can now explain the impact
of aerodynamic force on the sails and the drive of the sailboat. This explains the
embodied movement of heeling of the boat. Adam wanted to further connect with
the embodied practices to keep the boat stable without rolling over due to FLAT. The
choice of the angle of attack is here suggested to be critical in generating the desired
lift, yet it needed to be coordinated with bodily movements adjusting the sail shape
during the angle of attack towards the wind for the tacking manoeuvre. Adam, while
focusing on his actions during tacking, realised that he leans away from the boat out
of habit to keep the boat balanced against the heeling of the boat. He further added
that this leaning backward motion also helps the boat gain further speed. He
observed other sailors doing the same (Fig. 42.5).
Adam reflected back that they learn to do this rolling tack automatically to bal-
ance and pick up speed, since the sailors are required to counteract and balance the
forces on sails during tacking to avoid capsizing. The leaning against the other side
to balance and accelerate the boat forwards is an embodied evidence that there is a
counterpart force impacting the forward movement. The debriefing formulated the
stable balance of the boat as the goal for the next revision for our model, to incorpo-
rate the impact of water pressure on the hull and the keel as the underwater compo-
nent of the boat. This realisation is critical for building a more encompassing model
accounting for both the aerodynamics and hydrodynamics of tacking (Marchaj 2000).
Fig. 42.5 Leaning away to balance the boat between tacks to counter heeling force not to capsize
42.5 Conclusions
Our embodied phenomenological approach to mathematical modelling is a novel

contribution to research on the teaching and learning of mathematical modelling. In
mathematical modelling of complex problem situations, we portray here the case of
Adam and how he goes through the modelling process bringing his embodied
knowledge and practices to bear. During the process, he identifies the assumptions
based on his former knowledge and lived/living experiences, makes conjectures,
and contrasts them with the way he experiences sailing practice on its own life-
world. Adam’s case shows that the interpretation and validation phases of a mathe-
matical model become richer with the experience and embodied knowledge of the
learner. Through dialogue, a working configuration of tools becomes apparent to the
learner as he struggles to make a coherent sense of the model in context.
We demonstrate here how a culturally responsive modelling practice can build
on learner’s experiences, existing practices, and their funds of knowledge as young
sailors. Adopting an inquiry approach, the real-world problem of tacking has been a
driving theme for mathematical modelling activities. The goal was to figure out the
mathematics behind sailing against the wind guiding the modelling process as
Adam figured out the conceptual pieces such as lift, sail shape, wind, and the angle
of attack in the aerodynamics model of tacking. At the end, he set the next model-
ling goal as to see the bigger picture accommodating for the hydrodynamic forces.
Modelling of sailing clearly engages students where they feel empowered once
they can model and clarify core ideas behind their purposefully coordinated actions
during the sailing. These activities solidified Adam’s career disposition towards
becoming a naval architect. He felt that mathematical modelling of sailing helped
him gain deeper knowledge of sailing which provided him an edge against his com-
petitors in regattas and competitions. With support from his family and teachers, he
remained committed to pursue more modelling projects on sailing to advance his
embodied interdisciplinary knowledge of sailing as part of his formal schooling.
As demonstrated by Galbraith and Stillman (2001), the transition from real-
world problem statement to mathematical model is one of the most difficult parts of
the modelling cycle. Setting up the problem and validating and interpreting the
model require some experience of the modeller with the context of the problem.
Studying mathematical modelling of sailing with junior sailors gives embodied rep-
resentations and embodied knowledge as parts of the mathematical modelling pro-
cess especially in setting up, interpreting, validating, or revising phases of
mathematical modelling. The embodied phenomenology approach clearly demon-
strates the ongoing dynamic nature of the mathematical modelling process as con-
nected to the learner’s experiences and practices. Their funds of knowledge during
the modelling process are activated throughout the mathematical modelling. When
the learner needs most support with the mathematising phase, multiple representa-
tions are helpful including virtual manipulators.
Our embodied phenomenology approach to mathematical modelling fits well to
other areas where mathematical modelling is practised for learners who have bodily
sense and experience with the motions and movements that they are working on
understanding through mathematical models and modelling. Getting engaged in the
modelling process with the embodied approach builds on sailing phenomena as
practised and informed by a variety of disciplines including physics and engineer-
ing. The practice of sailing demands its practitioners to strategically use their tools
including their bodies, sails, and the wind effectively as they engage with problems
experienced in their communities of practice.
Acknowledgements The authors thank the St. Croix Yacht Club and Hanley Family for facilitat-
ing modelling activities with sailboats. Partial support was given by the National Science
Foundation Grant No. 1355437.
References
Anderson, K. (2008). The physics of sailing. Physics Today, 61(2), 38–43.

Andersson, J., Östman, L., & Öhman, M. (2015). I am sailing—towards a transactional analysis of
‘body techniques’. Sport, Education and Society, 20(6), 722–740.
Anhalt, C. O., Staats, S., Cortez, R., & Civil, M. (2018). Mathematical modeling and culturally
relevant pedagogy. In Y. Dori, Z. Mevarech, & D. Baker (Eds.), Cognition, metacognition, and
culture in STEM education (pp. 307–330). Cham: Springer.
engineering and economics – ICTMA 12 (pp. 222–231). Chichester: Horwood.
Dewey, J. (1916/1980). An added note as to the “practical.”. In J. A. Boydston (Ed.), John Dewey:
The middle works (Vol. 10, pp. 366–369). Carbondale: Southern Illinois University Press.
Dewey, J. (1935/1987). Liberalism and social action. In J. A. Boydston (Ed.), The later works of
John Dewey (Vol. 12). Carbondale: Southern Illinois University Press.
Fernandes, E., & Matos, J. F. (1998). Goal!!! [The mathematics of a penalty shoot-out in a football
game.]. In P. Galbraith, W. Blum, G. Booker, & I. D. Huntley (Eds.), Mathematical modelling:
Teaching and assessment in a technology-rich world (pp. 159–167). Chichester: Horwood.
Frejd, P., & Bergsten, C. (2016). Mathematical modeling as a professional task. Educational
Studies in Mathematics, 91, 11–35.
Galan, S. (2017). C420 Sailing St Croix Virgin Islands 2017 Training Up Wind. Available at https://
www.youtube.com/watch?v=GyQGEFEfxHY
Galbraith, P., & Stillman, G. (2001). Assumptions and context: Pursuing their role in modelling
activity. In J. Matos, W. Blum, K. Houston, & S. Carreira (Eds.), Modelling and mathematics
education, ICTMA 9 (pp. 300–310). Chichester: Horwood.
Greenstein, S., & Ekici, C. (2017). At the intersection of teaching and cultural diversity: Modeling
a culturally responsive mathematics pedagogy for the U.S. Virgin Islands. Journal of
Mathematics and Culture, 17(4), 39–84.
Husserl, E. (1970). The crisis of European sciences and transcendental phenomenology. Evanston:
Northwestern University Press.
Kolb, D. A. (1984). Experiential learning: Experience as the source of learning and development.
Englewood Cliff: Prentice Hall.
Ladson-Billings, G. (1995). Toward a theory of culturally relevant pedagogy. American Research
Journal, 32(3), 465–491.
Marchaj, A. C. (2000). Aero-hydrodynamics of sailing. New York: Adlard Coles Nautical.
Merleau-Ponty, M. (2012). The phenomenology of perception. New York: Routledge.
Moll, L. C., Amanti, C., Neff, D., & Gonzalez, N. (1992). Funds of knowledge for teaching. Theory
Into Practice, 31(2), 132–141.
Noss, R., & Hoyles, C. (1996). The visibility of meanings: Modelling the mathematics of banking.
International Journal of Computers for Mathematical Learning, 1(17), 3–31.
Roth, W. (2010). Researching living/lived mathematical work. Forum: Qualitative Social Research,
12(1). https://doi.org/10.17169/fqs-12.1.1604.
Roth, W.-M., & Thom, J. S. (2009). Bodily experience and mathematical conceptions: From clas-
sical views to a phenomenological reconceptualization. Educational Studies in Mathematics,
70(2), 175–189.
Schroeder, D. (2017). Fluid dynamics simulation. Retrieve from https://physics.weber.edu/
schroeder/fluids/
Sfard, A., & Kieran, C. (2001). Cognition as communication. Mind, Culture, and Activity,
8(1), 42–76.
Sheets-Johnstone, M. (2011). The primacy of movement: Expanded (2nd ed.). Philadelphia: John
Benjamins.
Stender, P., & Kaiser, G. (2016). Fostering modeling competencies for complex situations.
In C. Hirsh & A. R. McDuffie (Eds.), Mathematical modeling and modeling mathematics
(pp. 107–115). Reston: NCTM.
The Club C420 Association, Inc. (2018). C420. Retrieved from https://club420.org/
Vygotsky, L. (1986). Thought and language. Cambridge, MA: MIT University Press.
Watson, A. (2002). Embodied action, effect, and symbol in mathematical growth. In
A. D. Cockburn & E. Nardi (Eds.), Proceedings of PME 26 (Vol. 4, pp. 369–376).
Norwich: PME.
Chapter 43
Algebra Students Build Stock/Flow Models
to Study Non-linear, Dynamic Feedback
System Problems
Diana M. Fisher
Abstract Analysing personal, local, national, and global non-linear, dynamic feed-
back systems is central to addressing many current complex problems. The author
worked with a relatively broad ability cross-section of secondary mathematics stu-
dents in both algebra classes and in year-long modelling classes for 20 years in
which students built system dynamics (stock/flow) models to study these complex
problems, albeit at an introductory level. This chapter provides an example of one
such algebra lesson, the Malthus Problem partial modelling sequence, identification
of the software used by the students, and questions posed to students.
Keywords Algebra · System dynamics · Complex systems · Malthus Problem ·

Modelling
43.1 nalysing Complex Systems in Algebra: A Challenge

A
and Opportunity
Non-linear, dynamic feedback systems abound in our world. Examples include the
many systems in the human body, in ecology, in economics, in biology, in engineer-
ing, and in social systems. Preparing secondary school mathematics students to
analyse these systems presents a new challenge and opportunity.
Complex systems analysis changes what the equations are written about, shifting them from
properties of things to properties of dependencies or relationships and collective behavior.
Calculus and statistics do not have the tools to describe what we want to study.… It’s not
just about the math. We have to develop an entirely new set of concepts that enable us to
think about the behavior of the system. Different languages. Really, most importantly, it
D. M. Fisher (*)
Portland State University, Portland, OR, USA
e-mail: fisherd@pdx.edu

506 D. M. Fisher
changes the nature of the questions that we ask about the behavior of the system. It changes
how we think (Bar-Yam 2015).
Non-linear, dynamic feedback systems are a subcategory of complex systems.

Complex systems contain interconnected elements that have a purpose. They con-
tain feedback (positive and/or negative) and non-linear dependencies. Complex sys-
tem problems can have multiple potential solutions and often involve multiple
disciplines and viewpoints. In addition, these problems change over time. Modelling
methods are designed to address these types of problems, which, in the absence of
closed-form solutions, need to be “solved” by simulation methods.
In secondary school algebra when students study topics like population change,
trajectories, and swinging pendulums, these topics are often stripped of important
characteristics, so students can apply simple closed-form equations as a representa-
tion for modelling the scenario, in an attempt to get the “one right answer” that is
too often the goal of the exercise. But that approach alters some of the essential
character of the original system. For example, creating an equation for population
growth with a net growth rate that is unchanged over 50–100 years is unrealistic for
almost all populations. The net growth rate over an extended period of time could
be influenced by resource availability, social pressures or changing norms, disease,
war, etcetera. Air resistance in equations for trajectories and swinging pendulums is
often ignored to keep the equations simple enough for a mathematically diverse
group of (algebra) students to attempt to solve. These simplified problems often do
not provide a problem-solving approach that allows multiple potential solutions to
become evident and ignores multi-stakeholder perspectives that could be important
in interpreting a solution. Introduction of complex system problems in secondary
school algebra can be used to enrich the breadth and depth of applications mathe-
matics students can address and have been used to reinforce the function behaviour
that is typical of traditional algebra curriculum in the United States of America
(Fisher 2011).
Feedback analysis describing the cause of systemic behaviour can be highlighted
using complex systems modelling techniques, like System Dynamics. Discussion
can ensue about which elements are most relevant to the problem and feedback
loops can be drawn within the model to capture those causal influences on systemic
behaviour. Some causal influences may be non-linear, which can be represented in
the simulation modelling software. Many complex systemic problems cannot be
represented by closed-form equations, especially in an algebra class, making simu-
lation methods a viable alternative for analysing these systems.
By allowing students to become actively involved in gaining deeper understand-
ing of complex system problems, we provide a portal for prescriptive modelling
recommended by Niss (2015). “In prescriptive modelling the ultimate aim is to pave
the way for taking actions based on decisions resulting from certain kind of math-
ematical considerations, in other words ‘to change the world’ rather than only to
‘understand the world’” (Niss 2015, p. 69). The primary goal of an education
infused with System Dynamics modelling activities is to create “systems citizens”
capable of applying what they have learned from modelling within their communi-
ties and beyond (Chichakly 2016).
43 Algebra Students Build Stock/Flow Models to Study Non-linear, Dynamic… 507
43.2 The Structure of a System Dynamics Model
The System Dynamics modelling approach was developed by Jay Forrester at

Massachusetts Institute of Technology in the mid-1950s. In 1985 one of Forrester’s
Ph.D. students developed the Stella software, giving System Dynamics modelling a
friendly icon-based user interface. Once that interface was created, System
Dynamics modelling became accessible to a broad base of secondary school stu-
dents. The author has had students in algebra classes build simple versions of such
models for 20 years. An example of an algebra modelling lesson is discussed below.
43.2.1 The Modelling Tools
There are four main icons used to create models using the Stella System Dynamics
modelling software—stock, flow, converter, and connector.
The stock is used to represent variables that accumulate over time. This is an
integral. Variables whose values we want to track over time are often selected as
stocks. Examples include population or wheat produced per annum in the world.
A flow is a variable whose value represents the rate at which the stock value
changes. A uniflow, as shown in Fig. 43.1, can be drawn with the arrowhead point-
ing towards/away from the stock, in which case it would represent a rate of increase/
decrease of the value of the stock. The flow can also be a biflow (two-headed arrow)
where a positive value (within the flow) flows in the direction of a solid arrowhead
and a negative value flows in the direction of a dotted arrowhead.
A converter is a component that holds parameters, formulas, or other ancillary
numeric information for the model. For example, in a population model, a converter
could be used to hold the value for yearly percent of deaths, or in a model containing
population and food supply, it could be used to hold a simple formula calculating
food per person.
A connector is used to show dependencies in the model. It connects stocks to
flows, stocks to converters, converters to other converters, or converters to flows. It
allows numeric information to travel from one component to another. Figure 43.3
has two connectors in the population model shown on the left of the figure. One
connector is sending information about the current World Population to the flow
(population change per year), and the other is sending information about the net
Fig. 43.1 The modelling icons used in the Stella System Dynamics software
508 D. M. Fisher
change fraction (the fraction by which the population is growing) to the flow.
Connectors do not allow material to flow, just numeric information. Only flows
allow material to move into or out of stocks.
The best practice for designing System Dynamics models is to have flows or
converters only contain constants or simple arithmetic,1 whenever possible, so mod-
els are as easy to analyse (as transparent) as possible. The objective is to create the
simplest model that contains the essential elements that produce the behaviour of
interest.
43.2.2 mphasizing Linear vs Exponential Behaviour:

E
The Malthus Problem
As an example to show how these icons are used to structure a System Dynamics
model, we will look at the Malthus Problem as used with secondary mathematics
students in algebra classes in the United States of America. Students were given a
statement from Thomas Malthus2 who predicted hardship for lower-class society
because population increased exponentially and food production only increased lin-
early. They were asked how Malthus knew this would eventually be a problem. He
did not have a calculator or a computer.
In the first part of the lesson, the students were given three grids (on paper). The
independent variable for all three grids was years (0–200) and the dependent vari-
ables were World Population, Food Supply, and food per person, respectively. No
vertical scale was designated. Students were asked to predict the shape (over
200 years) of the dependent variable on each grid. Students easily sketched graph
predictions for the first two dependent variables (83% correct) but had difficulty
sketching the third graph (63% correct) (Fisher 2008). The focus assessing the third
graph was whether the student was able to sketch and explain that the graph had to
have a decreasing slope towards the end of the 200 years.
Each student built a model (on his/her computer) containing World Population
and Food Supply segments. The values for these model segments were obtained
from the World Watch data (see www.worldwatch.org). For Food Supply, wheat
production was used as the proxy (designated by the teacher). In 1950 Food Supply
was 631 x 106 tons and was growing at a constant amount of about 22.65 x 106 tons
per year. The model was simulated for a time span of 200 years (1950–2150). The
traditional mathematical equation for this model segment would be F = 22.65 x
106(t − 1950) + 631 x 106. The mathematical engine in the Stella software calculates
the model values recursively. (See Table 43.1.) The System Dynamics model dia-
gram (Fig. 43.2a) contains a stock for Food Supply and only an inflow since the
1
There is one more acceptable converter definition. Graphically defined converters are described
(in context) after Fig. 43.4 and shown in Fig. 43.5.
2
Malthus (1766–1834) was an economist and clergyman.
Table 43.1 How the Stella software calculates the values for Food Supply
Year Food supply each year {tons}
Starting food supply 631 ∗ 106
Food supply after 1 year 631 ∗ 106 + 22.65 ∗ 106 = 653.65 ∗ 106
…
Food supply after 200 years 5138.35 ∗ 106 + 22.65 ∗ 106 = 5161 ∗ 106
Recursive formula Ft = Ft − 1 + 22.65 ∗ 106
a b Food Supply
6000.00M
22.65e6 Food Supply
3000.00M
631e6
food growth
0.00M
amount per year 1950 2000 2050 2100 2150
years
Food Supply
Fig. 43.2 Stella model (a) and graphical output (b) of Food Supply segment of Malthus Problem.
Note: the value that would be placed in each Stella model icon is superimposed on the diagram (a),
for the reader’s convenience
Table 43.2 How the Stella software calculates values for World Population
Year Population each year {people}
Starting population 2.5 ∗ 109
Population after 1 year 2.5 ∗ 109 + 0.02(2.5 ∗ 109) = 2.55 ∗ 109
…
Population after 200 years 128.64 ∗ 109 + 0.02(128.64 ∗ 109) = 131.21 ∗ 109
Recursive formula Pt = Pt − 1 + 0.02(Pt − 1)
Food Supply per year was assumed to never decrease over the 200 years (see
Fig. 43.2b).
In 1950 World Population was 2.5 x 109 people and was growing about 2% per
year. The traditional mathematical equation for this model segment is P = (2.5 x
109)e0.02(t − 1950). Exponential System Dynamics model behaviour is produced when
the flow is defined to be proportional to the stock value (World Population). The
model’s (recursive) calculations for World Population, with a 2% yearly net change
(growth) fraction, are shown in Table 43.2.
The System Dynamics model diagram for exponential World Population growth
is shown in Fig. 43.3a. The initial World Population value in 1950 was 2.5 billion
people. The flow was defined as the product of the World Population component and
the net change fraction component. The flow was a biflow, allowing World
510 D. M. Fisher
a b World Population
200.00B
World Population
* 2.5e9 100.00B
population
change per year
0.02
0.00
net change 1950 2000 2050 2100 2150
fraction
years
WorldPopulation
Fig. 43.3 Stella model (a) and graphical output (b) of World Population segment of Malthus
Problem. Note: The asterisk shown in the biflow of Stella model (a) indicates that the two compo-
nents upon which it depends (World Population and net change fraction) are multiplied together
within the biflow definition
a b
net change Food Per Person
fraction
World Population
0.1653 1.00
Minimum
food per person
population 0.00
change per year Page -3-
÷ food per
person
-1.00
1950 2000 2050 2100 2150
years
food growth
amount per year Food Supply food per person
Minimum food perperson
Fig. 43.4 Stella model (a) and graphical output (b) of food per person compared to minimum food
per person for Malthus Problem model
Population to either grow or decrease based on the sign of the net change fraction.
The net change fraction was a positive value, so World Population increased.
Students needed to calculate the food per person each year, so they created a
converter entitled food per person and divided Food Supply by World Population.
Another component representing what might be a minimum level of food needed
per person to maintain a lean body level was also added. This value was chosen
arbitrarily (by the teacher).
But there was a problem. Students were asked to identify incorrect model behav-
iour. The model in Fig. 43.4 allowed the population to continue to grow exponen-
tially for 200 years. It did not make sense that World Population would continue to
grow unabated if food per person was dropping below the minimum food per person
level (Fig. 43.4b). To fix this problem the lesson told students to construct a graphi-
cally defined converter (effect of food per person on net change fraction) that modi-
fied the net change fraction (for World Population) based on the ratio of the current
food per person compared to the minimum food per person. (See Fig. 43.5.) The
graphically defined converter used the food per person ratio as the independent vari-
able (domain 0 to 2). The dependent variable in this graphical definition had values
ranging from −10 to 1. The graphical converter output had a value of 1 if the food
per person ratio was 1 or larger, but it began to drop towards −10 as the ratio
approached 0. The graphical converter output value was multiplied by the normal
net change fraction (0.02) reducing it as the food per person dropped below the
minimum food per person value. Eventually the graphical converter output was a
negative value. When the net change fraction became negative, the biflow changed
from increasing the World Population value to decreasing the World Population
value. Students were given the entire description for creating this graphically
defined converter as it was beyond the scope of the algebra class to have them con-
struct this type of definition on their own. But they did have to understand how it
influenced the net change fraction and model behaviour.
A special note should be made at this point. Going from the model shown in
Fig. 43.4 to the one in Fig. 43.5 is a significant step. In the model (Fig. 43.5), we
have introduced a negative feedback loop. (An increase in World Population causes
less food per person which causes the ratio of food per person to minimum food per
person to become smaller – moving one left on the horizontal axis “effect of food
per person…” graph. As one moves left on the “effect of food per person…” graph,
the output value (vertical component of that graph) becomes smaller (less than 1,
eventually) causing the net change fraction to become smaller than 0.02. A smaller
net change fraction causes fewer new people to flow into the World Population
stock.) This negative feedback has the influence of curbing World Population
increase and eventually causing World Population to decrease. Moreover, the intro-
duction of the graphical converter (“effect of food per person…”) introduces non-
linear dynamic behaviour into the model. It is not possible to build the model shown
in Fig. 43.5 using closed-form equations in an algebra class in secondary school.
This model demonstrates an example of complex systems characteristics: intercon-
nected elements with a purpose, feedback, non-linearity, and change over time.
a effect of food per person b

on net change fraction
net change
fraction * 0.1653
effect of food per person
on net change fraction
World Population
Minimum
food per person
0
population
change per year
÷ food per
person
food per person

food growth Minimum food per person
amount per year Food Supply
Fig. 43.5 Stella model (a) and graphical definition of effect of food per person on net change frac-
tion component (b) for Malthus Problem model
512 D. M. Fisher
In subsequent steps, the food growth amount per year flow was stepped up at
various times during the simulation to account for technological advancements in
agricultural science, and the net change fraction was stepped down at specific times
to account for improvements in birth control. But the ultimate behaviour of the sys-
tem remained unchanged. This persistent behaviour was a surprise for students (and
a strong reinforcement of the behavioural difference between the patterns of linear
and exponential change over time). The model behaviour reinforced Malthus’ dire
prediction about exponential population growth and linear food production.
One additional change was made to the model (not shown in this chapter). Food
Supply could not grow forever. There are limitations on amount of arable land, abil-
ity to improve growing practices, effects of water and air pollution on crops and soil,
etcetera. So another graphical converter was added to the Food Supply segment cap-
ping the growth possible.
The next question presented was as follows: So what? Now that students under-
stood the problem better, how could they deal with the ultimate undesirable behav-
iour of the system? Students were asked to add policy components to this model to
try different policy alternatives of their choice. They had to show that their policy
solved the problem for the long term. They had to determine which population
groups might oppose their policy and how they might use the model to help those
people better understand the dire consequences of doing nothing. This part of the
modelling lesson highlights that complex system problems often involve multiple
disciplines and viewpoints and can have multiple solutions. Asking students to think
about the implementation of policy to address the Malthus Problem issue is part of
paving the way for “taking action” as recommended by Niss (2015). This element
of addressing real-world issues is often lacking in mathematics classes.
The Malthus Problem was used as the basis of a brief study conducted by the
author with two of her algebra classes (total of 64 students). The students were
divided into two groups, one using calculators and the other using Stella, to analyse
a simplified version of the Malthus Problem (through the model shown in Fig. 43.4).
No graphically defined components were included. In this class experiment, stu-
dents were expected to step up the food growth amount per year value, in the model,
once in 1970 and were expected to step down the net change fraction value once in
1998. Students using graphing calculators had significant difficulty building this
model using piecewise defined functions, though they were given an example of
how to define piecewise functions on their graphing calculator. Only 4.5% of the
calculator group was able to determine the second equation for the piecewise func-
tion definitions, even though the process only required determining an exponential
equation from two data points and a linear equation from one point and the slope.
Both topics had been studied previously in class. Of the students using Stella, 29%
were able to correctly modify the model to reflect new values for food growth
amount per year in 1970 and net change fraction in 1998. The results of the study
are provided in Fisher (2008). This simple experiment did not address the fact that
in the normal Malthus System Dynamics lesson, students were representing the
systemic problem more realistically by adding two non-linear definitions (one dis-
played in Fig. 43.5 and the other that capped food production).
43.3 Action Research and Development
That a broad audience of algebra students (ages 15+) can build and analyse
introductory-level complex problem models is not debatable. The author used mod-
els like the Malthus Problem in traditional algebra classes with students of many
different mathematical abilities from 1990–2010. The student response to these les-
sons was positive (Fisher et al. 2003).
In the algebra classes, students started by building simple one-function System
Dynamics stock/flow models exhibiting linear, quadratic, and exponential behav-
iour. Then they combined these System Dynamics models to study drug pharmaco-
kinetics, natural resource depletion, and predator/prey dynamics, in addition to the
Malthus Problem (Fisher 2005). Approximately eight class periods in the academic
year were devoted to lessons that helped students learn to build and analyse System
Dynamics problems.
Moreover, a year-long System Dynamics modelling course was designed and
implemented over the same 20 years for secondary school students. In the last
10 weeks of that course, students selected their own dynamic problem topic, built a
working model, wrote a technical paper explaining their model, and presented their
model to the class. Those models tended to be two to five times the size of the model
shown in Fig. 43.5. Student model diagrams, technical papers, and some presenta-
tion videos can be found at https://ccmodelingsystems.com/portfolio-type/student-
projects/. A scope and sequence set of topics for introducing students to systems
thinking and System Dynamics modelling in preK–12 has been developed in com-
munication with teachers who have taught students systems thinking or System
Dynamics modelling at these schooling levels (Fisher 2011).
There are many areas of this new System Dynamics modelling paradigm that
warrant research clarification or attention. Important questions include as follows:
At what grade level are learners able to analyse non-linear dynamic behaviour?
How well do students understand the connection between the structure of the model
and the behaviour exhibited in the model graphical output? A significant problem to
address is assessment. Teachers who have used System Dynamics modelling in their
secondary school classrooms indicate that students think about the problem being
analysed at a deeper more conceptual level (Fisher 2017). How do we capture this
“deeper” learning?
43.4 Conclusion
Modelling has been difficult to embed in secondary school mathematics classes.

There are many reasons for this difficulty that have been previously described in the
research (Doerr 2007; Stillman 2007) but will not be presented here. Yet the need
for pre-college students to engage in non-linear, dynamic, feedback model building
is growing exponentially as both decision-makers and a voting public face
514 D. M. Fisher
increasingly difficult systemic problems. Secondary students have demonstrated

that they can develop the basic skill to analyse problems containing non-linear feed-
back dynamics. If a school system were to adopt an entire preK–12 systems think-
ing and System Dynamics modelling curricular strand, this skill would develop
more deeply.
References
Bar-Yam, Y. (2015, September 26). Complex systems science: Where does it come from and where
is it going to? [Plenary Address] Conference on Complex Systems. Tempe, Arizona. Video
viewed at: www.necsi.edu/research/overview/ccs15.html
Chichakly, K. (2016). Practical systems thinking: Guidance to effect change. The Solutions
Journal, 7(1), 62–64.
Doerr, H. M. (2007). What knowledge do teachers need for teaching mathematics through applica-
tions and modelling? In W. Blum, P. L. Galbraith, H.-W. Henn, & M. Niss (Eds.), Modelling
and applications in mathematics education. The 14th ICMI study (pp. 69–78). New York:
Springer.
Fisher, D. M. (2005). Lessons in mathematics: A dynamic approach with applications across the
sciences: Teacher’s guide. Lebanon: isee systems, inc.
Fisher, D. M. (2008). Building slightly more complex models: Calculators vs. STELLA. In
Proceedings of the 26th international system dynamics conference. Athens, Greece.
Fisher, D. M. (2011). “Everybody thinking differently”: K–12 is a leverage point. System Dynamics
Review, 27(4), 394–411.
Fisher, D. M. (with Stuntz, L., Benson, T., LaVigne, A., & Farr, W.). (2017). The next 25 years
in pre-college education: A move toward global understanding of complex systems. In
Proceedings of the 34th international system dynamics conference. Cambridge, MA.
Fisher, D. M., Gallaher, E. J., & Macovsky, L. (2003). How drugs work in the human body:
Analysis of a modeling unit used in a second year algebra class. In Proceedings of 21st inter-
national system dynamics conference. New York, NY.
Niss, M. (2015). Prescriptive modelling—Challenges and opportunities. In G. A. Stillman,
Stillman, G. (2007). Implementation case study: Sustaining curriculum change. In W. Blum,
P. L. Galbraith, H.-W. Henn, & M. Niss (Eds.), Modelling and applications in mathematics
education. The 14th ICMI study (pp. 497–502). New York: Springer.
Refereeing Process
To ensure the scholastic quality of the chapters in this book, all proposed chapters
submitted to the editors for consideration in a timely manner have been subjected to
a strict impartial and peer review independent of the editors. Only those chapters
that have been accepted after this process of review and revision as being of appro-
priate quality (given differences of geographical location) and consistent with the
intended aims of the book have been included. The editors wish to thank the follow-
ing international reviewers:
Luis Albarracín Universitat Autònoma de Barcelona, Spain
Burkhard Alpers Aalen University of Applied Science, Germany
Jonas Bergman Ärlebäck Linkoping University, Sweden
Michael Besser Leuphana University of Lüneburg, Germany
Piera Biccard University of South Africa
Rita Borromeo Ferri University of Kassel, Germany
Morten Blomhøj Roskilde University, Denmark
Werner Blum University of Kassel, Germany
Martin Bracke TU Kaiserslautern, Germany
Corey Brady Vanderbilt University, the United States of America
Jill P Brown Australian Catholic University (Melbourne), Australia
Nils Buchholtz University of Oslo, Norway
Susana Carreira Universidade de Lisboa, Portugal
Dirk De Bock Katholieke Universiteit Leuven (FEB Campus Brussels), Belgium
Helen Doerr Syracuse University, the United States of America
Diana Fisher Portland State University, the United States of America
Peter Frejd Linkoping University, Sweden
Peter Galbraith The University of Queensland (Brisbane), Australia
Vincent Geiger Australian Catholic University (Brisbane), Australia
Faaiz Gierdien University of Stellenbosch, South Africa
Mellony Graven Rhodes University, South Africa

of Mathematical Modelling, https://doi.org/10.1007/978-3-030-37673-4
516 Refereeing Process
Gilbert Greefrath University of Münster, Germany

Chris Haines City University (London), the United Kingdom
Paul Hernandez-Martinez Swinburne University of Technology, Australia
Corrina Hankeln University of Münster, Germany
Toshikazu Ikeda Yokohama National University, Japan
Gerrie J. Jacobs University of Johannesburg, South Africa
Cyril Julie University of the Western Cape, South Africa
Azita Manouchehri The Ohio State University (Columbus), the United States of
America
Nick Mousoulides Cyprus University of Technology, Cyprus
Mdu Ndlovu Stellenbosch University, South Africa
Kit Ee Dawn Ng Nanyang Technological University, Singapore
Marc North South Africa
Daniel Orey Universidade Federal de Ouro Preto, Brazil
Catherine Paolucci State University of New York, the United States of America
Derk Pik University of Amsterdam, the Netherlands
Craig Pournara University of the Witwatersrand (Johannesburg), South Africa
Xenia Reit Goethe University Frankfurt, Germany
Milton Rosa Universidade Federal de Ouro Preto, Brazil
Akihiko Saeki Naruto University of Education, Japan
Björn Schwarz University of Vechta, Germany
Hans-Stefan Siller Julius-Maximilians-University of Würzburg, Germany
Peter Stender University of Hamburg, Germany
Hamsa Venkatakrishnan University of Witwatersrand, South Africa
Katrin Vorhölter University of Hamburg, Germany
Pauline Vos University of Agder, Norway
Dirk Wessels University of Stellenbosch, South Africa
Jinxing Xie Tsinghua University (Beijing), P.R. China
Bert Zwaneveld Open University, the Netherlands
Index
A C
Abel, T., 119–127 Capraro, P., 455–464
Aerodynamics, 501, 502 Challenge, 371–374, 376–380
Affect, 28 Challenging tasks, 372, 380
Affective domain, 97 Characteristics, 300, 302, 303, 305
Affective perspective, 18 Classroom interactions, 188
Alagoz, C., 349–358, 494–503 Cleverness, 242, 245
Algebra, 506 Cognitive diagnostic models (CDM), 351
Applications, 16, 17, 21, 22, 24, 25, 27, 237, Cognitive strategies, 313, 317
238, 240, 241, 249, 250 Collaborative analysis, 18, 28
Ärlebäck, J.B., 139–149 Compartment model(s), 398, 399, 403, 404
Assessment, 16, 23, 24, 27, 350–352, 355, Competencies, 19, 22, 24, 26, 27, 259, 262,
357, 358, 362, 365, 372–377, 380, 277, 278, 281–284, 384, 387
381, 384 Complexity, 378, 379
mathematical modelling, 357 Complex problems, 27
Authenticity, 385–389, 391 Complex systems, 505, 506
Average speed, 152–156, 159, 160 Conditional knowledge, 337
Constructivism, 332
Content analysis, 366, 367
B Cortes, D., 107–115
Bali, M., 361–369 Creative and flexible thinking, 444, 452
Basic needs, 327, 331 Critical case sampling, 365
Beckschulte, C., 129–137 Cryptography, 244
Bekdemir, M., 187–196 Cryptosystems, 244
Beliefs, 325, 326, 331 Cultural-historical activity theory, 141
Biccard, P., 223–234 Culturally relevant pedagogy, 28
Blockages, 165, 172 Culturally responsive pedagogy, 494
Blomhøj, M., 396–405 Curriculum, 468, 469, 471, 472, 475
Bock, W., 455–464
Bourdieu, 35–37, 40, 47, 48
theory of, 34, 35, 37 D
Bourdieu’s theory of practice, 17, 52–54 Data, 189, 190
Bracke, M., 455–464 DBR, see Design-based research (DBR)
Brady, C., 76–83 De Bock, D., 407–415
Brown, J.P., 371–381 Degree of contextualisation, 374, 375

of Mathematical Modelling, https://doi.org/10.1007/978-3-030-37673-4
518 Index
Degree of coverage, 399, 401, 404 Grade 5, 189

Degree of openness, 375 Grade 9 students, 270, 271
Design-based research, 213, 432, 435, Greefrath, G., 383–391
437, 440 Gurjanow, I., 265–274
de Villiers, L., 215
DGS GeoGebra, 280, 281
Dialogical approach, 108–110, 114, 115 H
Difficulties, 20, 22, 164 Habits, 299, 300, 302–304, 307, 308
Discourse analysis, 86–88 Habitus, 23, 52–55, 57, 59, 60, 299, 300,
Dominguez, A., 76–83 302–304, 306–308
Duan, X., 419–428 Hankeln, C., 277–285
Durandt, R., 432–440 Hans Freudenthal, 408
Dynamic Geometry Software (DGS), 277, 278 Hernandez-Martinez, P., 34–38, 51–60
HLT, see Hypothetical learning
trajectory (HLT)
E Holton, H., 254–263
Educational sports engineering, 202 Hydrodynamics, 501
Ekici, C., 354, 494–503 Hypothetical learning trajectory, 213,
Embodied action(s), 152, 153, 160 481–484, 486
Embodied phenomenological approach, 27, 28
Embodied phenomenology, 494–496, 502
Emic approach, 108, 112, 114 I
Emotions, 97, 98, 100, 104 ICME-1, 408, 409
Engineering technician competencies, 209, Idea development, 76–79, 83
210, 218 Identifying mathematical models, 153,
Enjoyment, 96–98, 100–104 159, 160
Ethnomodelling, 18, 28, 86, 87, 92, 107, Ikeda, T., 287–297
109–112, 114, 115 Indexical discourse analysis, 18, 28
Etic approach, 112, 114 Indexicality, 88, 89
Evaluating, 338, 340, 342, 344 Informal learning, 72
Examinations, 24 about mathematical modelling, 72
Exchange value, 36 Informal (outside-school) learning, 64
External reflections, 397, 398, 402, 404 of mathematical modelling, 64, 66,
70, 71
Interactive case teaching, 420
F Interdisciplinary, 497, 502
Facilitating, 187, 188, 191, 193 Internal reflections, 398, 402, 404
Faculty practice, 126 Interpreting, 19, 25, 131, 133, 135–137
Fiction, 242–244 Intuitions, 86
Field, 53
of (mathematics) education, 53
Fisher, D., 505–514 J
Formulation, 23, 24 Julie, C., 366
Framework, 121, 122, 124–127 Jung, H., 76–83
Frejd, P., 34–48
Function, 18, 22, 25
Function concept, 108, 110, 114, 115 K
Funds of knowledge, 494–496, 502 Kaiser, G., 315, 336–345
Kaneko, M., 176–184
Kawakami, T, 151–160, 176–184
G Klock, H., 163–172
Galbraith, P., 254–263 Komeda, S., 151–160, 176–184
Glancy, A., 76–83 Kotze, H., 299–308
Govender, R., 443–452 Krüger, A., 311–320
Index 519
L Meaningfulness, 323–330, 332

Lantau, J-.M., 455–464 of mathematics, 23
Lautenbach, G.V., 432–440 MEAs, see Model-eliciting activities (MEAs)
Learning barriers, 269, 270, 272–274 Measurement sense, 22, 266, 269, 274
Learning environment, 199–203, 206 Mentoring, 254, 260, 262
Learning goals, 479, 481–483, 485–487 Metacognition, 312, 314, 315, 317, 318, 336,
Lecturer, 479–487 338–343, 345
Level of challenge, 24 Metacognitive competency, 23
Level of intensity, 164, 165, 170 Metacognitive knowledge, 312–314, 316,
Lewis, S., 85–93 319, 337–343
Limit, 288, 290, 291, 294–297 Metacognitive skills, 312, 313
Ludwig, M., 265–274 Metacognitive strategic instrument, 130
Metacognitive strategies, 23, 312–317, 319,
320, 336–343, 345
M Meta-knowledge, 66, 71
Malthus problem, 508, 512, 513 about mathematical modelling, 71
Manouchehri, A., 85–93, 187–196 Mhakure, D., 467–476
MathCityMap (MCM), 266, 267, Mistake(s), 164, 165, 169, 170, 172
269–271, 274 Mixed methods study, 202
Mathematical competencies, 468, 470 Model development sequences, 141
Mathematical creativity, 443 Model-eliciting activities, 76, 212
Mathematical knowledge, 288–290, 293, Modelling, 383–388, 390, 391, 506, 507,
296, 297 513, 514
Mathematical literacy, 24, 26, 362–365, 369, complex systems (see Mathematical
468–473, 475 modelling)
Mathematical modelling, 16–28, 39–42, 47, Modelling challenge, 255
48, 51, 52, 54–60, 96–104, 120–122, Modelling character, 24, 28, 387, 388
124–127, 129–130, 132, 133, 199–202, Modelling competencies, 24, 25, 130–132,
206, 209–212, 216–218, 224–226, 231, 137, 397–399, 401, 402, 404, 405, 434,
234, 237, 240, 241, 278, 287, 288, 290, 437, 439, 440
296, 297, 362–365, 367, 432–435, 440, Modelling problems, 313, 314, 316, 319, 320,
468, 469, 473, 475, 479–483, 486, 487, 326–328, 330, 332, 336, 337, 340,
494–498, 502 342, 345
assessment, 350 Modelling process, 262, 444, 446, 451, 452
critical and reflective dimensions, 201 Modelling projects, 26
meta-knowledge, 54 Modelling tasks, 226, 228
sub-competencies, 130, 133–137 Models and modelling perspective, 77, 141,
teaching, 420, 423 143, 148
Mathematical modelling competencies, 211, Modern mathematics, 410–412
212, 216, 217, 362–364, 444 Monitoring, 313, 318, 319, 336, 337,
Mathematical models, 176–179, 183, 184 340–344
Mathematical skills, 471 Multidimensionality, 350, 351
Mathematical uncertainty, 77–79, 83 of mathematical modelling
Mathematics, 176–178, 183 construct, 350
Mathematics classroom, 326, 332
Mathematics “just for laughs”, 248
Mathematics teachers, 40–42, 46, 48 N
Mathematisation, 20, 24, 66, 71, 160 Narratives, 52, 54, 59
Mathematising, 19, 22, 25, 86, 133, New Math, 408, 411, 415
135–137
Maths trails, 22, 265–267, 269, 270
Mbekwa, M., 368 O
McLean, J., 80 Openness, 385–388, 391
MCM app, 266, 269–271 Orey, D.C, 113, 199–207
520 Index
P Soares, R., 199–207

Parental support, 70 Socio-cultural environment, 34
Perspectives, 330 of mathematical modelling education, 34
of modelling, 330 Solution plan, 19, 130–134, 136, 137
Planning, 313, 317, 336–338, 340–344 Speed, 20, 176–184
Preservice mathematics teachers, 26, 432, Spooner, K., 479–487
444, 446 Sports practices, 201
Pre-service teachers, 26 Statistical skills, 470
Primary school, 152, 154 Statistics, 144, 145, 147, 148
Productive struggle, 373, 380 STEM, 26, 27, 352
Professional development, 352, 357 learning, 497
Project work, 396, 397, 404 modelling projects, 457, 464
Public image, 65, 68 Stephens, M., 287–297
of mathematics, 65–66 Street activity, 199
Public understanding, 65, 70 Study
of mathematical modelling, 17, 28, 65 mixed-methods, 111
PUMM, see Public understanding of Sub-competencies, 19, 24, 384, 385, 387, 388,
mathematical modelling (PUMM) 390, 391
Support materials, 255
System dynamics, 506–509, 512–514
R
Rate of change, 191, 196
Realistic Mathematics Education (RME), 409 T
Reflections, 25 Teacher interventions, 187, 188, 190, 191, 195
Regulating, 313, 318, 319, 337, 343, 344 Teacher scaffolding, 89
Relation to reality, 384, 388–391 Teachers, 350–358
Relevance, 385–388, 391 Teachers’ perception, 339, 340, 345
Representations, 77, 78, 83 Teacher training, 456, 457, 459, 460,
Re-signification, 108, 110, 115 463, 464
Rogovchenko, Y., 95–104 Teaching, 16–23, 25–27, 120, 121, 127, 140,
Rosa, M., 107–115, 199–207 141, 143–146, 149
Royaumont Seminar, 407–410, 415 Teaching experiment, 189, 193
Teaching mathematical modelling, 482
Teaching modelling, 432
S Teaching practice, 420, 427
Saeki, A., 151–160 Technical level, 399, 400, 404
Sailing, 494–498, 502, 503 Technology, 23, 27, 152, 153, 160, 176, 177,
Salinas, T., 124 266, 277, 278, 284
Schwarz, B., 325 Tertiary modelling courses, 479, 480
Science capital, 17, 28, 35–37, 40–42, 46–48, Threshold concept, 176
51–54, 57, 58 Time-based measurement, 172
Science museums, 64–66, 70–72 Treffert-Thomas, S., 95–104
Searcy, M.E., 122 Trigonometric function, 288, 290, 294–297
Segregation, 167 Turner, R., 254–263
Segway, 455, 456, 458–463
Semantic differential questionnaire, 302, 305
Sense-making, 16, 19–21, 26, 28, 119–122, U
124, 126, 127, 153, 160, 210, 211, University education, 103
223–234, 300, 305–307, Upper secondary level, 144, 145
432–435, 438–440 Upper secondary teacher, 143, 145, 148
in mathematical modelling, 224, 227 Usefulness, 66, 68, 71, 72
Siller, H.S., 163–172, 383–391 of mathematics, 64, 66, 68, 71, 72
Simplifying, 19, 133, 135, 137 Use value, 37, 39–41, 44–47
Simulation, 20, 166, 167, 169–172 of mathematics, 39–41, 44–47
Index 521
V Wendt, L., 336–345

Validating, 19, 22, 24, 25, 27, 133, 135, 137, Wessels, D., 209–218
277–279, 281–285 Working mathematically, 25
Validating competency, 278, 281, 282, World of work, 18
284, 285 Wu, M., 419–428
Validation, 24
Viirman, O., 95–104
Vorhölter, K., 319, 323–332, Y
336–345 Yamaguchi, T, 176–184
Vos, P., 34–38, 63–72 Yao, X., 187–196
Year 8 students, 176–178, 184
W
Walking, 176, 177, 179, 181, 182, 184 Z
Wang, D., 419–428 Zwaneveld, B., 407–415

Mathematical Modeling Libro

Uploaded by

Copyright:

Available Formats

You might also like

Mathematical Modeling Libro

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Mathematical Modeling Libro

Uploaded by

Copyright:

Available Formats

International Perspectives on the Teaching and

Learning of Mathematical Modelling

Gloria Ann Stillman

Editorial Board IPTL

More information about this series at http://www.springer.com/series/10093

Christine Erna Lampen

ISSN 2211-4920 ISSN 2211-4939 (electronic)

© Springer Nature Switzerland AG 2020

Gloria Ann Stillman Gabriele Kaiser

1 Helena Wessels: Her “Life Story”���������������������������������������������������������� 1

Part I Innovative Approaches in Modelling Educational

8 Mapping Shifting Discourses in Classroom Talk

19 Concurrent Development of Engineering Technician

Part III Pedagogical Issues for Teachers and Teacher Educators

Part IV Assessment of Mathematical Modelling in Schools

Part V Applicability at Different Levels of Schooling and University

41 A Lecturer’s Learning Goals for Teaching Mathematical

Part VI Examples of Modelling and Applications in Practice

Refereeing Process ������������������������������������������������������������������������������������������ 515

© Springer Nature Switzerland AG 2020 1

Keywords Born teacher · Statistics and mathematics education · Reflective

1.2 Tertiary Education and Becoming a Teacher of Note

1.3 Helena the Devout Christian

1.4 Interest in Higher Education

to train others at other schools. She became a carrier of the problem-centered

1.5 Helena and Dirk: The Team 1

1.6 Awards and Recognition

1.6.1 Helena the Scholar

1.6.2 Dedication from Around the World

1.7 Helena and Dirk: The Team 2

Wessels, H. (2017). Exploring aspects of creativity in mathematical modelling. In G. A. Stillman,

Gloria Ann Stillman, Gabriele Kaiser, and Christine Erna Lampen

Keywords Assessment · Mathematical modelling · Modelling educational

© Springer Nature Switzerland AG 2020 15

2.2 I nnovative Approaches in Modelling Educational

coordinates a diversity of research perspectives to make sense of what was happen-

2.3  esearch into, or Evaluation of, Teaching Practice

Researching teaching practice in mathematical modelling education and evaluating

reasoning, structuring, and mathematizing processes; however, the teacher’s sug-

2.4  edagogical Issues for Teachers and Teacher Educators

The extra-mathematical domains that feature prominently for mathematical

2.5 Assessment of Mathematical Modelling in Schools

Modes of assessment in mathematical modelling fall into two categories: (1) a

approach, where the modeller is asked to demonstrate separate competencies (e.g.,

2.6  pplicability at Different Levels of Schooling

The demonstration of applicability of teaching through modelling and applications

2.7 Examples of Modelling and Applications in Practice

Pauline Vos, Paul Hernandez-Martinez, and Peter Frejd

Abstract This chapter is an introduction to the theoretical concept of science capi-

Keywords Bourdieu (theory of) · Science capital · Theoretical perspectives · Use

© Springer Nature Switzerland AG 2020 33

In the past decade, researchers of mathematical modelling education increasingly

3.2 Bourdieu and Science Capital

3.3  cience Capital as a Tool in Research on Mathematical

Keywords Bourdieu (theory of) · Mathematical modelling · Mathematics teachers

© Springer Nature Switzerland AG 2020 39

4.2  cience Capital for Mathematics Teachers and Its

Table 4.1 Science capital for mathematics teachers

As indicated in the introduction, there is a link between mathematical modelling

1 Helena Wessels: Her “Life Story”�� 1

Part I Innovative Approaches in Modelling Educational

8 Mapping Shifting Discourses in Classroom Talk

19 Concurrent Development of Engineering Technician

Part III Pedagogical Issues for Teachers and Teacher Educators

Part IV Assessment of Mathematical Modelling in Schools

Part V Applicability at Different Levels of Schooling and University

41 A Lecturer’s Learning Goals for Teaching Mathematical

Part VI Examples of Modelling and Applications in Practice

Refereeing Process �� 515

1.2 Tertiary Education and Becoming a Teacher of Note

1.3 Helena the Devout Christian

1.4 Interest in Higher Education

1.5 Helena and Dirk: The Team 1

1.6 Awards and Recognition

1.6.1 Helena the Scholar

1.6.2 Dedication from Around the World

1.7 Helena and Dirk: The Team 2

2.2 I nnovative Approaches in Modelling Educational

2.3 esearch into, or Evaluation of, Teaching Practice

2.4 edagogical Issues for Teachers and Teacher Educators

2.5 Assessment of Mathematical Modelling in Schools

2.6 pplicability at Different Levels of Schooling

2.7 Examples of Modelling and Applications in Practice

3.2 Bourdieu and Science Capital

3.3 cience Capital as a Tool in Research on Mathematical

4.2 cience Capital for Mathematics Teachers and Its

4.4.1 The Case of Abraham

4.4.2 The Case of Bert

4.4.3 The Case of Cathy

4.4.4 Summary of Results

4.5 Discussion and Conclusion

5.2 Bourdieu’s Theory of Practice

5.4.1 Scott’s Narrative

5.4.2 John’s Narrative

5.5 Discussion and Conclusions

6.2 Theoretical Frame

6.5 Conclusion, Discussion, and Recommendation

Keywords Idea development · Models and modelling perspective · Model-

7.2 Theoretical Framework

7.2.2 Mathematical Uncertainty

7.2.3 Idea Development

7.3.1 Description of the Problem

7.3.2 Student Level

7.3.3 Researcher Level

7.5 Discussion and Conclusion

8.2 Background Literature

8.2.2 Discourse Analysis and Indexicality

8.3 Context and Background

8.4 n Indexical Analysis of Profit, Cost,

8.5 Discussion and Conclusion