Teaching and Teacher Education 71 (2018) 190e205

Contents lists available at ScienceDirect

Teaching and Teacher Education

journal homepage: www.elsevier.com/locate/tate

The influence of teachers’ attitudes and school context on

instructional practices in integrated STEM education
Lieve Thibaut a, *, Heidi Knipprath a, Wim Dehaene b, Fien Depaepe c
a
Research Group Education and Labour Market, Research Institute for Work and Society (HIVA), KU Leuven, Parkstraat 47 - Box 5300, 3000 Leuven, Belgium
b
MICAS, Microelectronics and Sensors, Department of Electrical Engineering (ESAT), KU Leuven, Kasteelpark Arenberg 10 - Box 2443, 3001 Leuven, Belgium
c
Centre for Instructional Psychology and Technology, Faculty of Psychology and Educational Sciences, KU Leuven, Dekenstraat 2 - Box 3773, 3000 Leuven,
Belgium

h i g h l i g h t s

A framework with five principles for integrated STEM is proposed.

Teachers' attitudes influence their instructional practices for each STEM principle.

Management support is the most important factor of school context.

a r t i c l e i n f o a b s t r a c t

Article history: This paper uses structural equation modelling (SEM) to examine the influence of teachers' attitudes and
Received 21 June 2017 school context on reported instructional practices in integrated STEM (Science, Technology, Engineering
Received in revised form and Mathematics). Instead of using an overall measure for integrated STEM, five specific characteristics
12 December 2017
(integration, problem-centered, inquiry-based, design-based, and cooperative learning) are examined to
Accepted 23 December 2017
get a more in-depth and nuanced insight into the factors influencing implementation. For each STEM
characteristic, teachers’ attitudes are positively linked with instructional practices Moreover, different
aspects of school context influence instructional practices either directly or indirectly. Opportunities for
Keywords:
Teacher attitudes
ameliorating the implementation of integrated STEM are discussed.
Instructional practices © 2018 Elsevier Ltd. All rights reserved.
School context
Integrated STEM

1. Introduction
Scientific and technological innovations are assumed to be
crucial to deal with the challenges of globalization and a
knowledge-based economy (National Academies of Science, 2007).
In order to provide all citizens with the competences necessary to
succeed in this new information-based and highly technological
society, education in the fields of science, technology, engineering
and mathematics (STEM) is becoming increasingly more important
(National Society of Professional Engineers, 2013). A promising
approach in this regard, is the use of an integrated curriculum,
which provides opportunities for “more relevant, less fragmented,
* Corresponding author.
E-mail addresses: lieve.thibaut@kuleuven.be (L. Thibaut), heidi.knipprath@
kuleuven.be (H. Knipprath), wim.dehaene@kuleuven.be (W. Dehaene), Fien.
depaepe@kuleuven.be (F. Depaepe).
and more stimulating experiences for learners” (Furner & Kumar,
2007, p. 186). Studies in a broad range of disciplines have shown
that students involved in an integrated curriculum perform as well
or even better than their peers in traditional instruction with
separate disciplines (Czerniak, Weber, Sandmann, & Ahern, 1999;
Hinde, 2005). Moreover, the use of an integrated curriculum has
been found to improve students’ non-cognitive learning outcomes,
such as interest and motivation (Erlandson & McVittie, 2001; Vars,
2001; Weilbacher, 2001).
When integrating science, technology, engineering and mathe-
matics, this approach is called ‘integrated STEM education’. While
different definitions exist, integrated STEM education is usually
used to designate an instructional approach in which students
participate in engineering design and/or research and experience
meaningful learning through integration and application of math-
ematics, technology and/or science (Moore & Smith, 2014).
Empirical research has indicated that integrated STEM education
has a positive impact on students' achievement (Austin, Hirstein, &

https://doi.org/10.1016/j.tate.2017.12.014
0742-051X/© 2018 Elsevier Ltd. All rights reserved.
 L. Thibaut et al. / Teaching and Teacher Education 71 (2018) 190e205
Abbreviations
STEM Science, Technology, Engineering and
Mathematics
SEM Structural Equation Modelling
CFA Confirmatory Factor Analysis
EFA Exploratory Factor Analysis
RMSEA Root Mean Square Error of Approximation
CFI Comparative Fit Index
Df Degrees of freedom
INT Integration of STEM content
PCL Problem-centered learning
IBL Inquiry based learning
DBL Design based learning
COL Cooperative learning
Walen, 1997; Hurley, 2001), their attitudes and interest in school
(Tseng, Chang, Lou, & Chen, 2013) and their motivation to learn
(Gutherie, Wigfield, & VonSecker, 2000). Moreover, integrated
STEM education has been reported to improve students' higher
order thinking skills and technological literacy, making them better
problem solvers, innovators and inventors (Morrison, 2006;
Stohlmann, Moore, Roehrig, 2012).
However, to optimally profit from the possible benefits of inte-
grated STEM education, it is important to ascertain that it is
implemented as intended (Stohlmann, Moore, Roehrig, 2012). Prior
research has identified two broad categories of characteristics that
affect the implementation of an educational program: teacher-level
characteristics (e.g., attitudes) and teaching context (Chen, 1998;
Henderson, Beach & Finkelstein, 2011; Henderson & Dancy, 2007;
Lund & Stains, 2015; Stains & Vickrey, 2017; Sunal et al., 2001).
Teachers' attitudes are found to be related to their instructional
practices since they determine a teacher's level of commitment to
infuse the new principles into daily classroom practices (Rockland
et al., 2010). Features of teaching context are important since
they create (or prevent the creation of) the necessary conditions for
teaching practices of high quality (Martin & Mullis, 2013; Stoll,
Bolam, McMahon, Wallace, & Thomas, 2006).
While prior research has emphasized the importance of teach-
ers' attitudes in the implementation of a new instructional strategy,
measuring teachers' attitudes toward teaching integrated STEM has
proven to be a difficult task. Some researchers resort to interviews
and focus groups, which is time consuming and only applicable to
small numbers of teachers (e.g. El-Deghaidy & Mansour, 2015;
Goodpaster, Adedokun & Weaver, 2012). Moreover, existing in-
struments for measuring teachers' attitudes toward STEM, only
assess attitudes toward each of the different STEM disciplines
individually, without taking the specific features of integration into
account (e.g. Friday Institute for Educational Innovation, 2012; Ho,
Yang & Yang, 2016) or use statements which include the term
‘interdisciplinary STEM’ without explaining the specific instruc-
tional strategy underlying this term (e.g. Al Salami, Makela & de
Miranda, 2017). This paper aims at contributing to this challenge
by proposing a questionnaire for measuring teachers' attitudes
toward teaching integrated STEM, which looks in more detail at the
different characteristics of the instructional approach. Through a
narrative review of existing literature, a conceptual framework for
integrated STEM, consisting of five key principles was developed.
Based on this framework, the questionnaire was constructed,
containing items for each of the key principles, thus allowing a
more detailed analysis of teachers' attitudes toward teaching in-
tegrated STEM.
191
Furthermore, to the best of our knowledge, no prior research
about the influence of teachers' attitudes and school context vari-
ables on their instructional practices in integrated STEM exists.
Nonetheless, insight into this specific relation could provide op-
portunities to improve the implementation of integrated STEM
education, ultimately contributing to the challenge of increasing
students’ interest in STEM fields. Therefore, this study aims at filling
this gap. To do so, the relationship between attitudes, school
context and implementation is examined separately for each of the
five key principles. By using individual measures for each key
principle rather than an overall measure for integrated STEM, the
findings of the study give a more in-depth and nuanced insight into
the factors influencing implementation. Examining this differential
impact of school factors and attitudes on the implementation of
integrated STEM is important, since it provides valuable informa-
tion to educators and teachers who want to target specific aspects
of their teaching practice.
In the next part, first, the conceptual framework for integrated
STEM, consisting of five key principles is described and the specific
instructional practices for each principle are defined. Next, struc-
tural equation modelling (SEM) is used to examine the relationship
between attitudes, practices and context. Direct, indirect and total
effects of the two predictor variables, attitudes and school context,
on reported instructional practices are presented. Finally, implica-
tions of these results are discussed and recommendations for
educational practice and future research are made.
2. Theoretical background
2.1. Integrated STEM education
While the importance of integrated STEM education in our
current society has been stressed, different definitions of integrated
STEM exist (Kelley & Knowles, 2016). According to Sanders (2009),
integrated STEM education includes all “approaches that explore
teaching and learning between/among any two or more of the
STEM subject areas” (p. 21). He suggests that each STEM course
should be designed to include learning outcomes for at least one of
the other STEM courses, e.g. a math or science learning outcome in
a technology or engineering class (Sanders, 2009). However, this
definition does not explicitly include the focus on engineering
design and real-world contexts that is stressed by other re-
searchers. Therefore, in this study, the definition of Moore and
Smith (2014) is adopted. They define integrated STEM education
as an instructional approach in which students participate in en-
gineering design and/or research and experience meaningful
learning through integration and application of mathematics,
technology and/or science (Moore & Smith, 2014).
2.1.1. Narrative review of existing literature
Since many researchers have already stipulated different oper-
ationalizations or conceptual frameworks for integrated STEM, a
narrative review (Rother, 2007) of existing literature was con-
ducted. This review focused specifically on papers that provided a
clear description of teaching practices for integrated STEM and that
identified multiple aspects within the instructional strategy for
integrated STEM, instead of reducing it to either inquiry-based
learning, problem-based learning or design-based learning. In to-
tal, nine papers were reviewed. For each of these nine papers, the
instructional elements were listed and reorganized in an effort to
synthesize them. Through grouping of similar aspects, five
distinctive but related key principles for integrated STEM were
discerned: integration of STEM content, problem-centered
learning, inquiry-based learning, design-based learning and coop-
erative learning. An overview of the five principles and their
192 L. Thibaut et al. / Teaching and Teacher Education 71 (2018) 190e205

correspondence to the instructional elements, as named in the nine
papers, is given in Table 1.
The first key principle refers to the integration of content from
different STEM disciplines. In all papers instructional elements
were found that could be linked to this key principle. Nonetheless,
different aspects are stressed in the different papers. While some
researchers focus specifically on using content from science and
mathematics (e.g. Bryan, Moore, Johnson, & Roehrig, 2015; Moore,
Guzey, & Brown, 2014), others especially underscore the impor-
tance of technology (e.g., Kelley & Knowles, 2016; Stohlmann,
Moore & Cramer, 2013). Moreover, a final group (e.g., O'Neill,
Yamagata, Yamagata & Togioka, 2012; Pinnell et al., 2013) empha-
sizes that it is not only important to use content from the different
STEM-disciplines, but that this content should be purposefully in-
tegrated across the disciplines. In our framework, this last aspect
was chosen as a main focus. Focusing on the importance of content
from the different STEM-disciplines has important consequences
for the role of technology within integrated STEM education.
Integrating content from technology means that technology should
not simply be used in education, but instead, education should be
organized such that students' knowledge of technology and tech-
nological artefacts is developed, together with technological skills
and technological literacy (Sade & Coll, 2003).
The second principle, which was mentioned in all papers, entails
the use of real-world problems tied to an engaging and motivating
context. To designate this, the term problem-centered learning was
chosen. Although problem-centered and problem-based learning
both advocate the use of authentic real-world problems and are
often used synonymously, problem-centered learning has a stron-
ger focus on teacher guidance (Merrill, 2007; Van Merrie €nboer &
Kirschner, 2007). Since this study is set in secondary education,
where some form of teachers' guidance is desirable, due to stu-
dents’ lack of experience and knowledge (Purser & Renner, 1983),
problem-centered learning was chosen in favour of problem-based
learning.
A third principle, described in six of the nine papers, is the use of
inquiry-based learning, i.e. learning that is driven by questioning,
thoughtful investigating, making sense of information and devel-
oping new understandings (Diggs, 2009). While some of the papers
focus specifically on science inquiry (e.g. Kelley & Knowles, 2016;
O'Neill et al., 2012), which usually entails that students have to
plan and design experiments, collect data and reflect on the results
by providing explanations for scientific phenomena (Capps &
Crawford, 2013), inquiry-based learning is not restricted to this
domain. It also occurs in mathematics, through questioning, chal-
lenging, discussing, interpreting and exploring mathematical ideas
(Menmuir & Adams, 1997) and in technology, through using, dis-
cussing and systematically assessing current and emerging tech-
nologies in an effort to understand them (Reed, 2003).
The fourth principle, design-based learning, refers to the use of
technological or engineering design. This was described explicitly
in six of the nine reviewed papers and implicitly eby naming
hands-on learning and the use of manipulatives-in one paper. In
our framework, design-based learning is used to describe an
instructional strategy in which students actively work on multi-
disciplinary design problems (Perrenet, Bouhuijs & Smits, 2000).
Students are challenged to thoroughly analyze a problem and in-
vent a solution for it. This solution is then designed and the final
product evaluated (Ke, 2014).
The final principle, which was explicitly listed in five of the
papers, entails the promotion of teamwork and collaboration with
others. To designate this, the term cooperative learning was chosen,
since it has a stronger focus on teacher guidance compared to
collaborative learning (Matthews, 1995). In our framework, coop-
erative learning refers to an instructional method that requires
students to work in small, mixed-ability learning groups, while the

Table 1
Overview of the instructional elements named in the nine papers.

Integration of STEM content Problem-centered Inquiry-based Design-based Cooperative Other aspects

learning learning learning learning

Moore et al. Standards-based S or M content Meaningful and Engineering Promote Student-centered pedagogy
(2014) engaging context design communication Learn from failure
Problem-solving skills and
teamwork
Stohlmann, Integrate technology Use a problem Discussion and Use Cooperative Teacher as facilitator Assessment
Moore, and solving approach inquiry Questioning manipulatives learning within instruction. Justification
Roehrig and conjectures and hands-on of thinking
(2012) learning
Sanders (2009) Purposefully teach a S or M Problem-based Authentic scientific Technological Engaging teams of
learning outcome in a T lesson learning inquiry design students
Wang, Moore, Content knowledge from S, T, E Problem solving by Inquiry-based
Roehrig & and M developing solutions instruction
Park (2011)
Stohlmann, M and/or S content as the main Motivating, Engineering Cooperative Student-centered pedagogy
Moore & objectives of the activity meaningful, and design process learning Teacher as facilitator
Cramer engaging context
(2013)
O'Neill et al., Trans-disciplinary Real-world, problem Science inquiry Project-based
(2012) based Self-directed inquiry
Kelley and Technological literacy Situated learning Scientific inquiry Engineering
Knowles Mathematical thinking design
(2016)
Pinnell et al. Integration of knowledge and Problem-solving Higher order thinking Engineering Collaboration with Performance-based tasks
(2013) skills from S, T, E and M challenges skills through inquiry design process. others
Interpersonal skills
Bryan et al. Learning goals, content and Solving real-world Engineering Teamwork Development of 21st century
(2015) practices from one or more S problems or tasks practices & skills
and M disciplines design
Design
justification
 L. Thibaut et al. / Teaching and Teacher Education 71 (2018) 190e205
teacher moves from team to team, observes the interactions and
intervenes when he feels it is appropriate (Matthews, 1995; Slavin,
1987).
As can be seen in the last column of Table 1, several other
instructional elements were named in the papers that were not
explicitly included in the framework. Nonetheless, many of these
elements are overarching aspects that are inextricably linked to one
or more of the discerned key principles. For example, both
problem-centered learning and inquiry-based learning are student-
centered pedagogies. In addition, the teacher as facilitator is an
important aspect of problem-centered learning, inquiry-based and
design-based learning. Moreover, the 21st century skills named by
Bryan et al. (2015) include four different components that are
linked to the five key principles. For example, one of the compo-
nents refers to the use of real-world themes, such as global
warming and health care, which is an important characteristic of
problem-centered learning. In addition, another component of the
21st century skills emphasizes the importance of creativity and
innovation, critical thinking, problem solving, communication and
collaboration, which are all aspects of problem-centered learning
(problem solving), inquiry-based learning (critical thinking),
design-based learning (creativity and innovation) and cooperative
learning (communication and collaboration). The other two com-
ponents also refer to aspects that are characteristic for one or more
of the key principles, such as dealing with information (inquiry-
based learning) and social skills (cooperative learning). Hence,
although the importance of developing these skills is not denied,
they are already largely present in the other key principles, defying
the need for a separate category. Other elements that were not
explicitly incorporated in the framework are aspects that are
named only once and are more general elements of quality in-
struction, rather than specifically important for integrated STEM,
such as letting students learn from failure or including assessment
into instruction.
2.1.2. Defining instructional practices
For each of the STEM principles, in Table 2, a brief description is
provided. Moreover, for each STEM principle, a literature research
was conducted to determine the defining instructional practices
within the context of integrated STEM education. These instruc-
tional practices are shown in the third column of Table 2 and will
serve as the main indicators for developing the measuring in-
struments (for attitudes and classroom practices) in this study.
As mentioned above, the first principle, integration of STEM
Table 2
Core characteristics for integrated STEM education.
STEM principle Brief description
Integration of STEM Content from different STEM disciplines is aligned and connected
content
(INT)
Problem-centered Students are involved in authentic problems
learning
(PCL)
Inquiry-based learning Students are actively engaged in posing questions, planning and designing
(IBL) experiments.
Design-based learning Students are actively engaged in creating some kind of artefact (e.g. robot,
(DBL) computer program)
Cooperative learning Students work together in teams
(COL)
193
content, refers to the alignment of content from different courses
(Choi & Pak, 2006; Drake & Burns 2004; Pettus, 1994). Several
levels of integration exist: (1) multidisciplinary integration, in
which concepts and skills are learned separately in each discipline
but within a common theme; (2) interdisciplinary integration,
where closely linked concepts and skills are learned from two or
more disciplines with the aim of deepening knowledge and skills;
and (3) transdisciplinary integration, where knowledge and skills
learned from two or more disciplines are applied to real-world
problems and projects (English, 2016). Since the proposed con-
ceptual framework focuses on the last two approaches to integrated
STEM, the defining instructional practices of this STEM principle
include aligning the learning objectives and learning activities from
different STEM disciplines. This allows for connections to be made
between closely linked concepts in the different disciplines.
Moreover, to comply with the transdisciplinary nature of integra-
tion, content from different disciplines should be applied jointly to
solve real-world problems.
Defining instructional practices for the second principle,
problem-centered learning, include the use of real world problems,
the use of a problem in its totality rather than only some compo-
nents, and the use of ill-defined problems that allow for multiple
solution paths and even multiple answers (Jonassen, 2000; Merrill,
2007, 2009; Merrill & Gilbert, 2008; Van Merrie €nboer, 1997; Van
Merrie €nboer & Kirschner, 2007).
For, inquiry-based learning, the third principle, the key
instructional practices encompass stimulating learners to question,
challenge, discuss, interpret, and explore mathematical and scien-
tific ideas (Menmuir & Adams, 1997). Moreover, students should be
actively involved in designing experiments and collecting data and
teachers should encourage their students to reflect on the research
results (Capps & Crawford, 2013).
Similarly, important instructional practices for design-based
learning include encouraging students to formulate hypotheses
for possible solutions, involving them in the design process, and
stimulating them to reflect on partial solutions in connection with
the original problem (Ke, 2014).
Finally, crucial instructional practices for successful cooperative
learning include creating positive interdependence between the
group members, providing opportunities for face-to-face interac-
tion between students, assuring individual accountability of all
group members and encouraging students to reflect on the group
process (Johnson, Johnson, & Smith, 1998).
Instructional practices
- Aligning learning objectives and learning activities from
different disciplines
- Emphasizing connections between content of different
disciplines
- Applying content jointly
- Using realistic contexts
- Presenting problems in their totality
- Providing ill-defined problems
- Stimulating active involvement of students through questions
- Letting students conduct research, experiments
- Encouraging students to reflect on research (results)
- Actively engaging students in design
- Encouraging students to reflect on design (process)
- Providing opportunities for face-to-face interaction between
students
- Creating positive interdependence between students
- Making students individually accountable
- Encouraging students to reflect on team work
194 L. Thibaut et al. / Teaching and Teacher Education 71 (2018) 190e205
2.2. Implementation of integrated STEM education
Several studies have reported positive effects of integrated
STEM education on students' achievement in STEM subjects (e.g.,
Austin et al., 1997; Hartzler, 2000; Hernandez et al., 2014; Hurley,
2001), their attitudes and interest in school (Tseng et al., 2013),
and their motivation to learn (Gutherie et al., 2000). However, to
optimally profit from the possible benefits of integrated STEM ed-
ucation, it is important to ascertain that it is implemented as
intended (Durlak, 1998; Stains & Vickrey, 2017). Prior research has
identified two broad categories of characteristics that affect the
implementation of an educational program, curriculum, or
instructional strategy: teacher-level characteristics (e.g., attitudes
and values of individual teachers) and contextual characteristics of
the teaching setting (e.g., school expectations, time, resources)
(Chen, 1998; Henderson et al., 2011; Henderson & Dancy, 2007;
Lund & Stains, 2015; Stains & Vickrey, 2017; Sunal et al., 2001).
Therefore, in this study, the influence of teachers' attitudes and
school context on their reported instructional practices in inte-
grated STEM is examined. In the next sections, we will focus on the
specific definitions of teachers’ attitudes and school context used in
this paper. Moreover, empirical results confirming the existence of
a relationship with instructional practices, both within STEM and
within other fields of study, are provided.
2.3. Teachers’ attitudes
2.3.1. Definition of attitudes
Historically, the term ‘attitude’ has been defined in many ways
(Maio & Haddock, 2014). For example, Ajzen (1988, p. 4) defined
attitude as “a disposition to respond favourable or unfavourable to
an object, person, institution or event.” Eagly and Chaiken (1993, p.
1) on the other hand, described it as “a psychological tendency that
is expressed by evaluating a particular entity with some degree of
favour or disfavour.” Despite some small differences, both defini-
tions of attitude emphasize evaluative judgements about an object.
Therefore, in this paper, the term ‘attitude’ is defined as the overall
evaluation of an object on several dimensions (good/bad, pleasant/
unpleasant) (Ajzen, 2005; Maio & Haddock, 2014).
To define these dimensions, van Aalderen-Smeets, Walma van
der Molen and Asma (2012) proposed a conceptual framework for
teachers' attitudes toward teaching science. Results of their vali-
dation study provided evidence for a framework consisting of three
dimensions: cognition, affect and perceived control. Within the
field of integrated STEM education, research about teachers' atti-
tudes is relatively scarce. Al Salami et al., (2017) reported a measure
for teachers' attitudes toward interdisciplinary teaching. However,
they did not provide a conceptual model for teachers' attitudes and
no explanation about the specific subdimensions taken into ac-
count was given. Mobley (2015) described the development of the
SETIS instrument to measure teachers' attitudes to teach science in
an integrated STEM framework. In this instrument, teachers' self-
efficacy was the only subdimension of attitudes measured. Due to
the non-existing or limited frameworks for teachers' attitudes to-
ward teaching integrated STEM found in research literature, a new
model was created based on the tripartite framework proposed by
van Aalderen-Smeets et al., (2012). This new conceptual framework
for teachers’ attitudes toward teaching integrated STEM is shown in
Fig. 1.
Within this framework, the cognitive dimension contains the
opinions a teacher has about the relevance and importance of
teaching integrated STEM. The affective dimension entails the
feelings of anxiety a teacher has when teaching integrated STEM.
Finally, the dimension of perceived control includes statements of
self-efficacy, which refers to a person's conviction that he/she can
Fig. 1. Framework for teachers' attitudes toward teaching integrated STEM.
successfully execute the behaviour (Beck & Ajzen, 1991).
2.3.2. Relation between attitudes and instructional practices
The idea that teachers' instructional practices are related to their
attitudes is not new and has been examined in multiple fields of
study. According to Bandura (1986), the behaviour of a person is
strongly influenced by his/her attitudes. Similarly, Pajares (1992)
asserts that beliefs, which represent the cognitive component of
attitudes, are the “best indicators of the decisions that individuals
make throughout their lives” (p. 307). Hence, teachers' attitudes are
believed to play a fundamental role in their actual classroom
practices, their decision making about curriculum and their
acceptance of new instructional approaches, techniques, and ac-
tivities (Donaghue, 2003; Nespor, 1987; Pajares, 1992). Next, a non-
exhaustive overview of empirical research about the relation be-
tween teachers’ attitudes and their classroom practices is given.
Empirical research about the relation between teachers' atti-
tudes and their classroom practices is widespread (e.g. Hashweh,
1996; Roehrig & Kruse, 2005; Savasci-Acikalin, 2009). However,
as mentioned above, many different conceptions of attitude exist.
Research about the relation between classroom practices and
teachers' perceived relevance, anxiety or self-efficacy, the three
subdimensions of attitude, is less frequent. Nonetheless, for each of
these subdimensions, empirical research exists that confirms the
relation with teachers' classroom practices. Studies examining the
influence of teachers' perceived relevance are mostly situated
within the field of reading instruction. For example, Bunt (2008)
examined 159 elementary teachers' perceptions about the rele-
vance of reading instruction and found a strong positive relation
between these perceptions and teachers’ classroom practices.
Likewise, Mehr, Kazemi and Omidvari (2015) measured the
importance attributed to reading strategies by 20 high school
teachers and reported a strong correlation with their classroom
teaching. Only one study was found within the field of science
education. Cronin-Jones (1991) conducted a case-study of 2
middle-level science teachers who implemented a 20-lesson cur-
riculum package and concluded that perceptions about the relative
importance of the content influenced the implementation of the
curriculum.
Research about the relationship between feelings of anxiety and
classroom practices has mostly been conducted among student
teachers. Nnenna and Olanrewaju (2015) investigated perceptions
of anxiety in 1416 student teachers in Southwest Nigeria and re-
ported that these perceptions were significant predictors of their
teaching performance. Preece (1979) measured anxiety about
teaching in 100 student teachers at the beginning and at the end of
a term's teaching practice. They concluded that for student teachers
of science, but not for teachers of non-science subjects, anxiety was
a cause, rather than a consequence, of class-control problems,
therefore confirming the relation between anxiety and teaching
behaviour. Bush (1989) investigated math anxiety in elementary
teachers and found that math anxious teachers spent a greater
amount of class time devoted to whole class instruction, while they
fielded fewer student questions during lessons.
In comparison with perceptions about relevance and anxiety,
teachers' self-efficacy has been studied more frequently. For
 L. Thibaut et al. / Teaching and Teacher Education 71 (2018) 190e205
example, Marshall, Horton and Switzer (2009) measured beliefs
about and use of inquiry in the classroom in 1222 K-12 mathe-
matics and science teachers. They reported that self-efficacy for
teaching inquiry was significantly correlated with the percentage of
time that students are engaged in inquiry during a typical lesson.
Dilekli and Tezci (2016) examined 1003 elementary teachers'
practices aiming at teaching thinking skills and their self-efficacy
towards teaching thinking skills. The results showed that self-
efficacy was a meaningful predictor of teachers' teaching thinking
practices. In an effort to summarize the extensive research about
teachers' self-efficacy, Zee and Koomen (2016) conducted a meta-
analysis consisting of 165 articles. Results suggested that teachers’
self-efficacy shows positive links with patterns of teacher behav-
iour and practices related to classroom quality.
2.4. School context
2.4.1. Definition of school context
A universally agreed-upon definition of school context does not
exist (Cohen, McCabe, Michelli, & Pickeral, 2009). Since many re-
searchers focus on promoting and measuring various aspects of
schools, a variety of definitions and frameworks have been devel-
oped. Within the field of STEM education, the National Research
Council (NRC) report (2011) has identified elements that are shared
by schools that showed improvements in student learning in
mathematics and science. In this report, school leadership was
named as the driver for change. Other aspects that were found to be
important included the capacity of the staff to work together, the
organization of the curriculum and the tools teachers have to
advance learning (such as instructional materials). Another attempt
to denote important components of STEM-schools was done by
Peters-Burton, Lynch, Behrend and Means (2014). Based on a
literature review, they identified ten critical components for suc-
cessful inclusive STEM schools. These components include real-
world STEM partnerships and well-prepared STEM teaching staff.
Nonetheless, this framework focuses specifically on inclusive STEM
schools, i.e. schools that explicitly target underrepresented stu-
dents and provide them with the means to succeed in STEM majors,
rather than focusing on the school context needs for the imple-
mentation of integrated STEM. Specifically within the field of in-
tegrated STEM, no framework for school context was found.
2.4.2. Relationship between school context and instructional
practices
Despite the absence of a theoretical framework, some studies
have focused on the relationship between teaching context and
classroom practices within the fields of science, technology, engi-
neering and mathematics. For example, Appleton and Kindt (1999)
asked teachers to indicate which context factors would make it
harder or easier for them to teach science and found three impor-
tant factors: collegial support, availability of resources and priority
given to science by the school system. Kynigos and Argyris (2004)
investigated teaching practices in an innovative school program
in which students were involved in mathematical investigations
through the use of exploratory software. They interviewed and
observed eight teachers and concluded that their instructional
practices were influenced by the classroom context and the school's
pedagogical priorities. Bingimlas (2009) conducted a meta-analysis
of the relevant literature about technology integration in science
education. Results of the study showed that lack of access to re-
sources was an important barrier to exhibiting the desired teaching
behaviour. Roehrig, Kruse and Kern (2007) examined the role of
school level factors on the implementation of a reform-based
chemistry curriculum. Analysis of the data revealed that imple-
mentation of the curriculum was strongly influenced by the
195
presence of a supportive network at the school site.
2.5. Relationship between school context and teachers’ attitudes
According to prior research, context factors not only affect
teachers' instructional practices, but also their attitudes
(DeChenne, Koziol, Needham & Enochs, 2015). For example,
DeChenne et al. (2015) investigated self-efficacy beliefs in 128
graduate teaching assistants in STEM and found teaching self-
efficacy to be strongly affected by a variety of environmental fac-
tors. Ayub, Bakar and Ismail (2015) explored factors influencing 187
mathematics teachers' attitudes towards the integration of ICT in
the teaching and learning process. Their findings showed that
teachers' attitudes towards using ICT in teaching and learning were
positively correlated with the school culture, school support and
access to ICT resources. Therefore, this relationship between
teachers’ attitudes and school context will also be taken into ac-
count in this study.
2.6. Purpose and research questions
The goal of this study is to provide insight in the reported
instructional practices of integrated STEM teachers. Moreover, two
groups of predictor variables for these instructional practices are
examined: teachers' attitudes toward teaching integrated STEM
and school context factors. To do so, the study first aims at devel-
oping and validating measurement instruments for teachers' atti-
tudes towards and instructional practices in each of the five core
characteristics of integrated STEM. Next, the relationship between
teachers’ attitudes, school context variables and their instructional
practices for each of the five STEM principles is examined. The
specific research questions are:
1. What is the influence of STEM teachers' attitudes on their
instructional practices for each of the five defining characteris-
tics of integrated STEM?
2. What is the influence of school context factors on STEM teach-
ers' instructional practices and their attitudes for each of the five
defining characteristics of integrated STEM?
3. Method
3.1. Sample
Data were gathered as part of a larger study investigating STEM
education in Flanders. An online questionnaire was administered to
595 secondary (K6-K12) schools between January and March 2017.
For this study, all teachers who indicated that they were involved in
teaching integrated STEM were selected. This led to a sample of the
study consisting of 244 in-service teachers from 121 schools. The
participants were approximately evenly divided in terms of gender
(51% male) with a mean age of 40 years (range 22e65 years) and an
average teaching experience of 14.4 years (range 1e43 years). Of
the 244 teachers, 14.7% were science (i.e. geography, biology,
chemistry or physics) teachers, 4.1% technology/engineering
teachers and 5.7% mathematics teachers. By contrast, all other
teachers (75.5%) were involved in teaching two or more STEM
disciplines. Teachers also differed in terms of their training: 41.6%
obtained an undergraduate (bachelors) degree in teaching, while
the others held a graduate (masters) degree in the fields of science
(25.8%), mathematics (2.0%) or technology/engineering (30.6%).
196 L. Thibaut et al. / Teaching and Teacher Education 71 (2018) 190e205

3.2. Measures integrated STEM. Furthermore, to test the reliability of the ques-
tionnaire, the internal consistency was determined by computing
3.2.1. Measures of attitudes toward teaching integrated STEM the Cronbach's alpha coefficient for each subscale (Table 3). Cron-
To determine secondary teachers' attitudes toward teaching bach's alpha values range from 0.74 to 0.94, therefore exceeding
integrated STEM, a questionnaire was developed in alignment with 0.70, which is generally used as the threshold for an acceptable
the theoretical framework. Items were created based on the reliability coefficient (Nunnaly, 1978).
defining instructional practices for each of the five distinguished
STEM principles: integration of STEM content, problem-centered
3.2.2. Measures of instructional practices
learning, inquiry-based learning, design-based learning and coop-
To examine teachers' classroom practices in integrated STEM, a
erative learning. In line with the theoretical framework for teach-
second questionnaire based on the theoretical framework was
ers' attitudes of van Aalderen-Smeets et al. (2012), items about
developed. For each of the five STEM principles, items about the
teachers' perceived relevance, anxiety and self-efficacy were
frequency of certain instructional practices were generated. Re-
developed. Next, a pilot study with 135 teachers was conducted.
spondents were asked to indicate the level of occurrence of these
Moreover, interviews with three experts in the field, two physics
items on a 5-point Likert-scale (1 ¼ never, 5¼ (almost) always).
teachers, one mathematics teacher, one chemistry teacher and one
Again, interviews with experts in the field and teachers with
engineering teacher were held. Based on the pilot study and the
experience in mathematics, physics, chemistry and engineering
interviews, several items were adapted or removed in order to
were held to assure the validity of the questionnaire. This proced-
assure the construct validity of the questionnaire. Finally, 75 items
ure led to a total of 25 items measuring teachers’ classroom prac-
were retained (a complete overview of the items translated into
tices in integrated STEM (a complete overview of the items
English is shown in Appendix A). Respondents were asked to
translated into English is shown in Appendix B). Sample items,
indicate their level of agreement with these items on a 5-point
mean values and standard deviations for each of the STEM princi-
Likert-scale (1 ¼ totally disagree, 5 ¼ totally agree). All items
ples are given in Table 4.
about Perceived Relevance were formulated as ‘How important do
Confirmatory Factor Analysis was used to examine the construct
you think it is to … ’, items about Self-Efficacy as ‘How capable do
validity of the questionnaire. As separate CFA models for each STEM
you feel to do this while teaching integrated STEM?’ and items
principle were necessary to measure teachers' attitudes, separate
about Anxiety as ‘How anxious do you feel to do this while teaching
CFA models for instructional practices were constructed as well.
integrated STEM?’. Sample items for all STEM principles are given
The results of the validation tests are shown in Table 4 and indicate
in Table 3. Moreover, mean values and standard deviations for the
that the questionnaire is a valid instrument for measuring sec-
three subscales (Perceived Relevance, Self-efficacy and Anxiety) in
ondary teachers' instructional practices in integrated STEM.
each of the STEM principles are given in this table.
Chronbach's alpha values range from 0.74 to 0.86, confirming the
To confirm construct validity of the questionnaire, Confirmatory
reliability or internal consistency of the classroom practices
Factor Analysis (CFA) was performed. Conventional rules of thumb
questionnaire.
in CFA dictate that a sufficient model fit is achieved when the c2/df
value is lower than 2, CFI is higher than 0.90, and the RMSEA is at
least lower than or equal to 0.08 (Marsh, Balla & Hau, 1996). 3.2.3. Measures of school context
Separate CFA models were constructed for each of the STEM prin- Since no theoretical framework for school context within the
ciples. Model fit indices for these CFA models are given in Table 3 field of integrated STEM education was found, a combination of
and indicate that the questionnaire is a valid instrument for exploratory and confirmatory factor analysis was used to construct
measuring secondary teachers' attitudes toward teaching different school context factors. Based on prior research within
other fields of study (e.g. Appleton & Kindt, 1999; Lumpe, Haney, &

Table 3
Sample items, Chronbach's alphas, means, standard deviations and CFA metrics for the attitudes questionnaire.

STEM principle Items N Subscale alpha Mean s.d. CFA metrics

Items

Integration - Combine math, science and/or technology content in one course 4 Perc. Relevance 0.74 4.23 0.58 c2 42.0
- Connect mathematical, scientific and/or technological concepts Self-Efficacy 0.86 3.71 0.71 df 32
Anxiety 0.92 2.93 1.03 CFI 0.994
RMSEA 0.036
Total 3.34 0.63
Problem-centered learning - Link STEM concepts or procedures to realistic scientific situations 4 Perc. Relevance 0.74 4.08 0.58 c2 38.1
- Let students work on complex assignments with multiple solutions Self-Efficacy 0.88 3.69 0.70 df 31
Anxiety 0.92 2.98 1.02 CFI 0.996
RMSEA 0.031
Total 3.26 0.63
Inquiry-based - Instruct STEM concepts by means of research assignments 5 Perc. Relevance 0.75 4.32 0.48 c2 101.6
learning - Let students reflect on the research process Self-Efficacy 0.87 3.95 0.59 df 65
Anxiety 0.93 2.61 0.92 CFI 0.983
RMSEA 0.048
Total 3.55 0.53
Design-based learning - Instruct STEM concepts by means of design assignments 6 Perc. Relevance 0.86 4.12 0.56 c2 146.9
- Let students reflect on the design Self-Efficacy 0.92 3.82 0.67 df 99
- process Anxiety 0.94 2.78 0.94 CFI 0.986
RMSEA 0.045
Total 3.39 0.61
Cooperative learning - Guide students while they're working in teams 6 Perc. Relevance 0.75 4.17 0.50 c2 173.8
- Let students reflect on their team work Self-Efficacy 0.88 3.88 0.60 df 100
Anxiety 0.94 2.76 0.96 CFI 0.974
RMSEA 0.055
Total 3.43 0.44
 L. Thibaut et al. / Teaching and Teacher Education 71 (2018) 190e205 197

Table 4
Sample items, Chronbach's alphas, means, standard deviations and CFA metrics for the instructional practices questionnaire.

Principle Items N Alpha Mean s.d. CFA metrics

items

Integration In integrated STEM … 6 0.86 3.20 0.82 c2 9.17

- Connections are made between related ideas or concepts from different STEM principles df 7
- Examples from different STEM disciplines are used CFI 0.997
RMSEA 0.036
Problem-centered learning When students get an exercise in integrated STEM … 4 0.74 3.73 0.59 c2 2.37
- There are multiple possible solutions df 3
- There are multiple solution strategies CFI 1.00
- The exercise depicts a realistic situation RMSEA 0.000
Inquiry-based learning During a research process in integrated STEM … 6 0.78 3.44 0.67 c2 4.60
- Students develop their own experiments df 6
- Students communicate their research results CFI 1.00
- Students explicitly evaluate their research process RMSEA 0.000
Design-based learning During a design process in integrated STEM … 5 0.84 3.43 0.78 c2 3.82
- Students have to motivate their design choices df 4
- Students plan their design work CFI 1.00
- Students explicitly evaluate their design process RMSEA 0.000
Cooperative learning During teamwork in integrated STEM … 4 0.85 3.83 0.66 c2 1.00
- All students are actively involved in the task df 1
- Conversations between the students deal with the task CFI 1.00
RMSEA 0.000

Czerniak, 2000) and experience within an integrated STEM df ¼ 131, p ¼ .00; CFI ¼ 0.96, RMSEA ¼ 0.054) show that the four-
research program, 19 items referring to different elements of school factor-model is a good fit.
environment were constructed. Teachers were asked to indicate to
which extent they felt the different elements were present in their 3.3. Data analysis
school on a 4-point Likert scale (1 ¼ totally absent; 4 ¼ more than
adequately present). Through exploratory factor analysis (EFA) with Structural equation modelling (SEM) analyses were conducted
oblique rotation on half the dataset, four categories of school using SPSS Amos software version 23.0. SEM was chosen over
context were defined: social, technical, management and organi- multivariate regression, since it has the advantages of simulta-
zational context (Table 5). These categories roughly correspond neously estimating relationships between multiple independent
with the factors named in the theoretical framework. Next, the and dependent variables and of incorporating measurement errors
other half of the dataset was used to validate the model through due to scale unreliability (Gerbing & Anderson, 1988; Malaeb,
confirmatory factor analysis. Model fit indices (c2 ¼ 222.46, Summers, & Pugesek, 2000). For the attitude questionnaire, there
were no missing values. In the classroom practices and school
context questionnaire, there were only few missing items
Table 5 (approximately 1,8%). These missing values were automatically
Factor loadings presented in the pattern matrix obtained by factor analysis using replaced by means of maximum likelihood estimation available
direct oblimin rotation on 19 items (n ¼ 122). from the AMOS software (Enders & Bandalos, 2001). The normality
1 2 3 4 assumption for each item was tested in terms of skewness index
Social context
(SI) and kurtosis index (KI). The SI ranges between 1.580 and
Consultation meetings at fixed times 0.82 0.462, the absolute value of which is less than the recommended
Multidisciplinary STEM team 0.77 level of 3 (Kline, 2010). The values for KI lie between 1.491 and
Consultation meetings when necessary 0.73 3.839, whose absolute value is less than 10 (Kline, 2010). Therefore,
Shared vision among colleagues 0.70
normality of the data is assumed.
‘free’ time for preparation 0.66
Five separate SEM models were constructed, one for each STEM
Technical context principle. Fig. 2 shows the hypothesized model for one principle.
Functional classrooms -0.89 Unobserved latent variables are represented by ellipses, observed
Access to technical facilities -0.85 variables by rectangles and circles indicate measurement error. In
Techn. material -0.84
line with the hypotheses proposed in the theoretical framework,
IT material -0.75
Techn. support by experts -0.61 instructional practices are the dependent variable in the model,
Budget -0.43 while teachers' attitudes and school context factors are the pre-
Management context
dictor variables. As mentioned above, according to prior research
(e.g. DeChenne et al., 2015) context factors not only affect teachers’
Involvement of management in implementation -0.83
instructional practices, but also their attitudes, therefore indirectly
Appreciation of management for STEM teachers -0.83
Clear vision of the management about STEM -0.77 influencing instructional practices. These hypothesized relations
education between context and attitude are represented by dotted lines in
Confidence of management in STEM teachers -0.67 Fig. 2.
Organizational context

Subsequent teaching hours 0.82 4. Results

Teaching hours for integrated STEM 0.76
Small class groups 0.45 4.1. Correlation
Co teaching 0.41

Note: Factor loadings <0.30 are omitted and factors loadings are sorted. Bivariate correlations were used for preliminary examination of
198 L. Thibaut et al. / Teaching and Teacher Education 71 (2018) 190e205

Fig. 2. Hypothesized SEM model for one STEM principle.

the pattern of associations between variables. Table 6 shows high
correlations between the attitudes towards the different STEM-
principles, indicating that teachers' feelings and beliefs about the
different aspects of integrated STEM education are highly related.
Although high correlations can be found between all constructs, in
general, the correlation between attitudes and practices regarding
the same STEM principle is stronger than for non-matching
principles. For example, the correlation coefficient between
teachers' attitudes towards cooperative learning and their
instructional practices in cooperative learning is 0.46, whereas the
correlation coefficient between these attitudes and teachers’
instructional practices in design-based learning is only 0.37. One
exception can be found: attitudes toward problem-centered
learning are more strongly correlated with instructional practices

Table 6
Bivariate correlations between manifest variables.

Attitudes Classroom Practices Context

Int Pcl Ibl Dbl Col Int Pcl Ibl Dbl Col Org. Tech Social Man.

Attitudes

Int 1
Pcl 0.80** 1
Ibl 0.62** 0.73** 1
Dbl 0.60** 0.72** 0.77** 1
Col 0.49** 0.59** 0.77** 0.72** 1

Classroom Practices

Int 0.46** 0.44** 0.44** 0.40** 0.44** 1

Pcl 0.33** 0.41** 0.34** 0.40** 0.28** 0.34** 1
Ibl 0.30** 0.36** 0.47** 0.45** 0.44** 0.43** 0.44** 1
Dbl 0.34** 0.44** 0.41** 0.46** 0.36** 0.40** 0.51** 0.62** 1
Col 0.36** 0.37** 0.40** 0.37** 0.46** 0.40** 0.30** 0.45** 0.50** 1

Context

Org. 0.27** 0.28** 0.21** 0.27** 0.20** 0.31** 0.30** 0.14** 0.21** 0.31** 1
Tech 0.18** 0.26** 0.19** 0.34** 0.21** 0.32** 0.35** 0.25** 0.21** 0.19** 0.51** 1
Social 0.21** 0.27** 0.23* 0.26** 0.24** 0.49** 0.12 0.11 0.17** 0.19** 0.48** 0.38* 1
Man. 0.25** 0.28** 0.30** 0.30** 0.24** 0.40** 0.31** 0.29** 0.28** 0.30** 0.47** 0.46** 0.54** 1

Int ¼ Integration, Pcl ¼ problem-centered learning, Ibl ¼ Inquiry-based Inquiry-based learning, Dbl ¼ Design-based learning. Col ¼ cooperative learning, Org ¼ organizational,
Tech ¼ technical, Man ¼ Management.
* Correlation is significant at the 0.05 level (2-tailed).
** Correlation is significant at the 0.01 level (2-tailed).
 L. Thibaut et al. / Teaching and Teacher Education 71 (2018) 190e205
related to integration (r ¼ 0.44) and design-based learning
(r ¼ 0.464), than with instructional practices related to problem-
centered learning (r ¼ 0.41).
4.2. SEM analysis
To investigate the relation among teachers' attitudes, school
context and teaching practices, a SEM model was constructed for
each STEM principle. Model fit indices indicated that the proposed
model, as shown in Fig. 2, does not provide a good fit for the
observed data (e.g. for the principle of integration: c2 ¼ 1109.005,
df ¼ 580, p ¼ .00; CFI ¼ 0.897, RMSEA ¼ 0.061). As shown in Table 6,
the four school context variables are highly correlated. When co-
variances between the context variables were included, model fit
improved substantially (c2 ¼ 866.870, df ¼ 574, p ¼ .00; CFI ¼ 0.943,
RMSEA ¼ 0.046 for the principle of integration). This indicates that
the different context factors are strongly related and this relation
needs to be taken into account to accurately investigate the influ-
ence of school context on teachers’ attitudes and instructional
practices.
The final SEM models, which include covariances between the
school context factors, are shown in Fig. 3. For each model, the
goodness of fit indices indicate a sufficient fit. The squared multiple
correlations show that about 30% of the variation in instructional
practices can be explained by the structural equations (ranging
from 24% in the SEM model for design-based learning to 33% in the
model for integration of STEM content), whereas 15% of the varia-
tion in teachers’ attitudes is explained (ranging from 13% in the
SEM model for cooperative learning to 20% in the model for design-
based learning).
4.2.1. Attitudes and instructional practices
An overview of the direct, indirect and total effects of teachers'
attitudes and school context variables on instructional practices for
each STEM principle is given in Table 7. For each model, a significant
correlation between teachers’ attitudes and teaching practices is
found. Nonetheless, correlation strength differs among the STEM
principles. While the standardized regression weights well exceed
0.40 for inquiry-based learning (b ¼ 0.54, p < .001), design-based
learning (b ¼ 0.43, p < .001) and cooperative learning (b ¼ 0.46,
p < .001), this value only amounts to 0.30 for integration of STEM
content (b ¼ 0.32, p < .001) and problem-centered learning
(b ¼ 0.29, p ¼ .008).
4.2.2. School context and instructional practices
The effect of school context on teachers' instructional practices
differs depending on the STEM principle. While management
context has a significant total effect on instructional practices for all
STEM principles except cooperative learning (range b ¼ 0.20 to
0.34), social context only significantly influences teachers'
instructional practices for integration (b ¼ 0.23, p ¼ .018) and
problem-centered learning (b ¼ 0.31, p ¼ .012). For organizational
context only one significant total effect was found, i.e. for instruc-
tional practices in problem-centered learning (b ¼ 0.23, p ¼ .017).
Nonetheless, organizational context also directly influences
teachers' cooperative instructional practices (b ¼ 0.27, p ¼ .047),
even if no significant total effect was found. For technical context,
no significant total effects on teachers’ instructional practices were
found, although a significant indirect effect on design-based
learning practices (b ¼ 0.11, p ¼ .010) exists. Moreover, a negative
direct effect of technical context on instructional practices for
cooperative learning (b ¼ 0.21, p ¼ 0.019) was found.
199
5. Discussion
The goal of this study was to find variables that can explain
variation in teachers’ instructional practices in integrated STEM.
Examination of these instructional practices was based upon the
five distinctive but related core characteristics for integrated STEM
education, defined in the theoretical framework. These character-
istics, derived through a narrative review of existing literature,
include: integration of STEM content, problem-centered learning,
inquiry-based learning, design-based learning and cooperative
learning. Prior research has shown that two groups of variables can
affect the implementation of an educational program: teacher-level
characteristics, such as attitudes, and contextual characteristics of
the teaching setting (Chen, 1998; Henderson et al., 2011;
Henderson & Dancy, 2007; Lund & Stains, 2015; Stains & Vickrey,
2017; Sunal et al., 2001). Therefore, the effects of these two
groups on the teaching of integrated STEM were examined.
A first challenge in this regard is the measurement of teachers'
attitudes toward teaching integrated STEM. Existing instruments
for measuring teachers' attitudes toward STEM, often only assess
attitudes toward each of the different STEM disciplines individually,
without taking the specific features of integration into account (e.g.
Friday Institute for Educational Innovation, 2012; Ho et al., 2016) or
use statements which include the term ‘interdisciplinary STEM’
without explaining the specific instructional strategy underlying
this term (e.g. Al Salami et al., 2017). Therefore, this paper has
proposed and validated a questionnaire for teachers' attitudes to-
ward teaching integrated STEM, which specifically measures five
core characteristics of integrated STEM. This questionnaire is a
valuable contribution to the field of STEM education research,
because it allows for a more detailed analysis of teachers' attitudes
toward teaching integrated STEM.
To investigate the influence of teachers’ attitudes and school
context variables on their instructional practices in integrated
STEM, a differential approach was chosen. Five structural equation
models were composed, one for each STEM principle. By examining
each key principle separately, rather than using an overall measure
for integrated STEM, a more in-depth and nuanced insight into the
factors influencing implementation is provided.
5.1. Main findings
5.1.1. Teachers’ attitudes and instructional practices
Results of the SEM analyses show that, for each of the STEM
principles, teachers’ attitudes are significantly related to their
classroom practices. These findings are in line with the expecta-
tions. Although no studies were found assessing every one of the
different core characteristics, prior research has confirmed the
existence of a relationship between attitudes and classroom prac-
tices for most of the separate principles (e.g. De Hei, Strijbos, Sjoer
& Admiraal, 2015; Xie & Sharif, 2014).
Although positive correlations between teachers' attitudes and
their teaching practices were found for all STEM characteristics, the
strength of this correlation differs among the five principles. While
the correlation is relatively strong for inquiry-based, design-based
and cooperative learning, it is much lower for integration of STEM
content and problem-centered learning. An explanation for this
observation might be found in the specific nature of the principles.
Both integration and problem-centered learning rely on making
connections between different disciplines. For the principle of
integration this is inherent to its definition; for problem-centered
learning this is dictated by the use of authentic problems, which
are multi-disciplinary in nature. Therefore, implementation of
these instructional practices is not only dependent on the teacher
him-or-herself, but is also influenced by collaboration with
200 L. Thibaut et al. / Teaching and Teacher Education 71 (2018) 190e205

Fig. 3. Structural equation modelling results using full information maximum likelihood with standardized parameter estimates.

colleagues. A statement that is supported by the significant effect
that social context has on teachers’ instructional practices for both
integration and problem-centered learning (see further). It is
possible that, due to the dependence on external factors, such as
collaboration with colleagues, the influence of internal factors, such
as attitudes, becomes less pronounced. Nonetheless, as mentioned
above, no prior research was found comparing the relationship
between attitudes and instructional practices across the different
STEM principles. Therefore explanations for this observation
remain hypothetical and further qualitative research (e.g., through
interviews or focus groups with integrated STEM teachers) is
necessary to clarify the specific reasons behind this difference in
correlation strength.
5.1.2. School context and instructional practices
In addition to teachers' attitudes, the effect of different aspects
of school context on teachers' instructional practices in integrated
STEM was examined. These effects differ depending on the STEM
principle. For integration of STEM content and problem-centered
learning, the collaborative aspects of school context, regarding
both peer teachers (i.e., social context) and management (i.e.,
management context), directly influence teachers’ instructional
 L. Thibaut et al. / Teaching and Teacher Education 71 (2018) 190e205 201

Table 7
Direct, indirect and total effects of teachers’ attitudes and school context variables on instructional practices for each STEM principle.

Attitudes Management Social Technical Organizational

b p b p b p b p b p

Integration of STEM content

Direct effect 0.32** <.001 0.17* .048 0.23* .018 0.14 .069 -0.13 .148
Indirect effect // // 0.03 .410 0.07* .043 0.00 .931 0.05 .072

Total effect 0.32** <.001 0.20** .043 0.30** <.001 0.14 .100 -0.08 .437

Problem-centered learning

Direct effect 0.29** .007 0.33** .004 0.31* .012 0.09 .275 0.17 .105
Indirect effect // // 0.01 .946 0.07 .074 0.02 .446 0.06 .091

Total effect 0.29** .007 0.34** <.001 0.38** <.001 0.11 .176 0.23* .017

Inquiry-based learning

Direct effect 0.54** <.001 0.15 .201 -0.08 .508 0.01 .993 -0.04 .646
Indirect effect // // 0.14* .045 0.07 .212 0.03 .642 -0.00 .847

Total effect 0.54** <.001 0.29* <.001 -0.01 .876 0.03 .766 -0.04 .751

Design-based learning

Direct effect 0.43** <.001 0.14 .124 -0.02 .767 -0.09 .370 0.09 .506
Indirect effect // // 0.09 .068 0.05 .410 0.11* .010 -0.02 .711

Total effect 0.43** <.001 0.23* .019 0.03 .889 0.02 .762 0.065 .573

Cooperative learning

Direct effect 0.46** <.001 0.16 .163 -0.14 .236 -0.21* .019 0.27* .047
Indirect effect // // 0.06 .246 0.08 .228 0.07 .212 -0.00 .982

Total effect 0.46** <.001 0.22 .080 -0.06 .591 -0.14 .205 0.27 .091
*
Correlation is significant at the 0.05 level (2-tailed).
**
Correlation is significant at the 0.01 level (2-tailed).
Italics ¼ total effect, bold ¼ effect is significant at the 0.05 level, bold and italics ¼ total effect is significant at the 0.05 level.

practices. As mentioned above, both principles rely on making
connections between different STEM disciplines, which could
explain the importance of a collaborative environment. Prior
research has confirmed these findings. Stohlmann et al. (2012)
investigated the implementation of an integrated STEM program
in middle school and reported that collaboration between teachers
with different subject backgrounds positively influenced imple-
mentation. Since teachers felt more comfortable with different
parts of the curriculum depending on their background, they could
act as a resource for each other and help colleagues with questions.
Moreover, weekly meetings to share ideas about how to teach the
classes were found to be beneficial for the implementation of the
program and helped teachers to feel more comfortable (Stohlmann
et al., 2012).
For inquiry-based learning and design-based learning, no direct
effects of school context on instructional practices were found.
Nonetheless, specific aspects of school context do influence
teachers’ attitudes toward these STEM principles, therefore indi-
rectly affecting instructional practices. Management support and
involvement was found to be an important factor for both inquiry-
based and design-based learning. Moreover, as indicated above,
management support directly influences instructional practices in
integration and problem-centered learning as well. Extensive
research evidence has pointed to the importance of leadership from
principals and other administrators in improving the quality of
teaching and learning in their schools (Fullan, 1993; Prather, 1996).
Support, guidance, and leadership are vital if teachers are to make
major shifts from a traditional didactic style to a reformed one
(NRC, 2000).
For cooperative learning, two aspects of school context were
found to directly influence instructional practices. Organizational
context, including sufficient teaching hours and co-teaching, was
found to positively influence the use of cooperative learning. Prior
research has indeed shown that having classroom-based peer
support for teachers is an important factor contributing to the
successful implementation of cooperative learning (Jolliffe, 2005).
Moreover, the same research identified insufficient teaching time
as a key factor hindering successful implementation. Interestingly,
technical context, including technical material, IT material and
budget, was found to negatively affect cooperative learning. While
no evidence from literature was found to support this observation,
it might be that the lack of sufficient resources necessitates
collaboration among students. Since working on a single product or
having group members share resources promotes positive inter-
dependence (Smith & Waller, 1997), it is possible that teachers who
lack the means to provide each student with his/her own resources,
unintentionally create more opportunities for qualitative coopera-
tive learning by having students share them.
5.2. Limitations and perspectives for further research
The results of the study are valuable since they address the lack
of knowledge about factors influencing teachers' instructional
practices in integrated STEM. Nonetheless, the current research
design has some limitations. One limitation is the fact that all
measures are self-reported. This may have provoked social-
desirable answers and could have contributed to artifactual
covariation among measures. Further research is required to
investigate the validity of the results when more objective mea-
sures, such as classroom observations, are used. Another limitation
of the study lies in the fact that only questionnaires eor quantita-
tive data-was used. Although this was a deliberate choice in order
to examine the attitudes and teaching practices of a larger dataset,
future research may benefit from the addition of teacher interviews
or other qualitative data. That way, more in-depth and detailed
information about the relationship between attitudes, school
202 L. Thibaut et al. / Teaching and Teacher Education 71 (2018) 190e205
context and instructional practices can be gathered. A third limi-
tation of the study concerns the variables taken into account to
explain the variance in teachers' instructional practices in inte-
grated STEM. Although studies have demonstrated that teachers'
attitudes and teaching context are important factors influencing
teachers' instructional practices (e.g. Henderson & Dancy, 2007;
Lund & Stains, 2015), other factors have been found to affect
these instructional practices as well. For example, Ingvarson,
Beavis, Bishop, Peck and Elsworth (2004) investigated effective
mathematics instruction in middle school teachers and reported
that teacher knowledge and educational background were posi-
tively linked to teachers' instructional practices. Faour (2003)
examined the relationship between beliefs and practices of Leb-
anese teachers and concluded that several factors, including edu-
cation, training and age, may affect both teachers' beliefs and
practices. Since the explained variance in teachers' instructional
practices and attitudes in this study is only moderate to low, the
addition of other factors, such as teachers' knowledge or back-
ground, could improve these values. However, a larger dataset is
required to accurately conduct these analyses. Other opportunities
for further research include the detailed examination of teachers'
attitudes and the factors influencing these attitudes. The ques-
tionnaire proposed in this paper could be used to investigate the
current state of teachers' attitudes toward teaching integrated
STEM and relate these attitudes to specific background character-
istics, such as gender, training, teaching assignments et cetera.
Finally, to further examine the implementation of integrated STEM
education, the effect of teachers’ attitudes and their instructional
practices on both cognitive and affective learning outcomes of
students could be studied.
6. Significance and conclusion
This research aimed at finding variables that can explain varia-
tion in secondary teachers' instructional practices in teaching in-
tegrated STEM. Instructional practices in five categories,
characteristic for integrated STEM education, were examined.
These categories are integration of STEM content, problem-
centered learning, inquiry-based learning, design-based learning
and cooperative learning. For each category, teachers' attitudes
were found to positively affect their instructional practices. More-
over, different aspects of school context were found to influence
teachers’ instructional practices, either directly or indirectly. Find-
ings of this study are important, since they provide insight into the
Perceived Relevance
How important do you think it is that:
(Totally not important- Rather not important- neutral-
Rather important- Very important)
INTa
Mathematical, scientific and/or technological content is combined in one
course.

Mathematical, scientific and/or technological concepts are linked.

The content of the separate STEM-courses (e.g. mathematics, physics, en-
gineering, …) and the integrated STEM course are aligned.

Students are provided with exercises/assignments in which they
simultaneously have to apply knowledge from different STEM-disciplines.
PCL
Students work on complex assignments that have multiple solutions.

STEM-concepts or eprocedures are linked to complex problems for which
various solutions exist.

STEM-concepts or eprocedures are linked to realistic scientific situations.

Students work on authentic/real-world problems.
opportunities for ameliorating the implementation of integrated
STEM education.
In line with research results in other fields of study (e.g. Chen,
1998; Henderson et al., 2011; Henderson & Dancy, 2007; Lund &
Stains, 2015; Stains & Vickrey, 2017; Sunal et al., 2001), findings
indicate that teachers' attitudes affect their instructional practices
in integrated STEM. Therefore, influencing teachers' attitudes can
act as a lever for improved STEM education. Although attitudes are
often regarded as stable personal ideas that are difficult to change, a
recent study by van Aalderen-Smeets and Walma van der Molen
(2015) showed that teachers' attitudes can be improved through
attitude-focused professional training. According to them, such
professional training should focus on assignments that stimulate
attitude change, rather than providing pre-structured, recipe-like
lesson examples. Hence, the results of this study call for specific
professional training, which includes assignments to create
awareness about teachers’ own attitudes towards teaching inte-
grated STEM in combination with challenges intended to change
these attitudes.
Another lever for improved STEM education, indicated by the
results of the current study, concerns school context. An important
role in this regard is reserved for principals and school adminis-
trators. Prior research has shown that support, guidance, and
leadership are vital if teachers are to make major shifts from a
traditional didactic style to a reformed one (NRC, 2000) and find-
ings of this study show that this is the case for implementing in-
tegrated STEM as well. In addition to support and appreciation from
school management, providing sufficient opportunities for collab-
oration and consultation between teachers with different STEM
subject backgrounds is also beneficial for the successful imple-
mentation of integrated STEM education. Interestingly, effects of
technical facilities and resources on teachers’ instructional prac-
tices were found to be small or even negative. While this does not
necessarily mean that the availability of technical resources is not
important for the implementation of integrated STEM education,
this result is promising, since it demonstrates that opportunities to
facilitate integrated STEM exist, even in schools with limited access
to technical resources.
Appendix A. Items of the attitude questionnaire (translated
from Dutch)
Self-Efficacyb Anxietyb
How capable do you feel to do the How stressful do you find the following
following when teaching integrated when teaching integrated STEM:
STEM: (Totally not capable- Rather not capable-
(Totally not capable- Rather not capable- neutral- Rather capable- Very capable)
neutral- Rather capable- Very capable)

Combining mathematical, scientific and/or technological content in one course.

Linking mathematical, scientific and/or technological concepts.

Aligning the content of the separate STEM-courses (e.g. mathematics, physics, en-
gineering, …) and the integrated STEM course.

Providing students with exercises/assignments in which they simultaneously have
to apply knowledge from different STEM-disciplines jointly.

Working with students on complex assignments that have multiple solutions.

Linking STEM-concepts or eprocedures to complex problems for which various
solutions exist.

Linking STEM-concepts or eprocedures to realistic scientific situations.

Working on authentic/real-world problems with students.
 L. Thibaut et al. / Teaching and Teacher Education 71 (2018) 190e205 203
(continued )

Perceived Relevance Self-Efficacyb Anxietyb

How important do you think it is that: How capable do you feel to do the
(Totally not important- Rather not important- neutral- following when teaching integrated
Rather important- Very important) STEM:
(Totally not capable- Rather not capable-
neutral- Rather capable- Very capable)
How stressful do you find the following
when teaching integrated STEM:
(Totally not capable- Rather not capable-
neutral- Rather capable- Very capable)

IBL
STEM-concepts are taught by means of research assignments
Teaching STEM-concepts by means of research assignments

Students find an answer to their own questions.
Encouraging students to find an answer to their own questions.

Students reflect critically upon the gathered (research) data.
Encouraging students to reflect critically upon the gathered (research) data.

Students are guided throughout a research process.
Guiding students throughout a research process.

Students reflect upon the research process.
Encouraging students to reflect upon the research process.
DBL
STEM-concepts are taught by means of design assignments.
Teaching STEM-concepts by means of design assignments.

Different student designs are evaluated.
Evaluating different student designs.

Students justify their design choices.
Encouraging students to justify their design choices.

Students are guided throughout a design process.
Guiding students throughout a design process.

Students reflect upon the design process.
Encouraging students to reflect upon the design process.

Students evaluate their designed product.
Encouraging students to evaluate their designed product.
COL
STEM-concepts are taught by means of teamwork.
Teaching STEM-concepts by means of teamwork (starting from 2 persons).

All students are actively involved in the teamwork.
Making sure that all students are actively involved in the teamwork.

The interaction between the students during teamwork deals with the
Making sure that the interaction between the students during teamwork deals with
content. the content.

Students are guided throughout teamwork.
Guiding students throughout teamwork.

Students reflect upon the process of their teamwork.
Encouraging students to reflect upon the process of their teamwork.

An individual student is evaluated based on teamwork.
Evaluating an individual student based on teamwork.
a
INT ¼ Integration of STEM content; PCL ¼ problem-centered learning; IBL ¼ inquiry-based learning; DBL ¼ design-based learning; COL ¼ cooperative learning.
b
For Self-Efficacy and Anxiety, the same items were used. These items have a slightly different formulation compared to the items for Perceived Relevance, but deal with the
same content.

Appendix B. Items of the implementation questionnaire

(translated from Dutch)

Items
a
INT
During integrated STEM …
(never, sometimes, regularly, often, very often)- … explicit references to the content of other STEM courses are made.
- … connections are made between related ideas or concepts from different STEM domains.
- … examples, applications and analogies from various STEM domains are used.
- … exercises/assignments are provided for which multiple STEM components are required to solve them.
- … there are agreements between the various STEM teachers about the use of symbols and names.
- … there is alignment between the content in integrated STEM and the contents of other STEM subjects (e.g. physics, mathematics, engineering).
PCL
When students receive an exercise/assignment during integrated STEM …
(never e usually not - sometimes - usually e (almost) always)- … there are several possible solutions.
- … there are several possible solution strategies.
- … this exercise depicts a realistic situation.
- … this exercise resembles a challenge encountered by real scientists/engineers.
IBL
During integrated STEM …
(never, sometimes, regularly, often, very often)- … students conduct an experiment.
- … students look up information/data (e.g. on the computer, in manuals).

During a research process in integrated STEM …
(never e usually not - sometimes - usually e (almost) always)- … students develop their own experiments.
- … students choose their own research theme or research question.
- … students compose their own hypothesis.
- … students check if the research results are realistic.
- … students communicate their results to fellow students.
- … students explicitly evaluate their research process (e.g. by discussing it with fellow students, writing a reflection, through an evaluation questionnaire).
DBL
During integrated STEM …
(never, sometimes, regularly, often, very often)- … students work on a design proposal and/or prototype.

During a design process in integrated STEM …
(never e usually not - sometimes - usually e (almost) always)- … students are freedom to determine the layout of their design.
- … students need to justify their design choices.
- … students plan their design process.
- … students formulate the pros and cons of their design product.
- … students explicitly evaluate their research process (e.g. by discussing it with fellow students, writing a reflection, through an evaluation questionnaire).
COL
During integrated STEM …
(never, sometimes, regularly, often, very often)- … students work per two or in small groups.

During teamwork in integrated STEM …
(never e usually not - sometimes - usually e (almost) always)- … all team members are actively working on the task.
- … the conversations between the students deal with the content of the task.

During long-term teamwork in integrated STEM …
(never e usually not - sometimes - usually e (almost) always)- … students explicitly evaluate the teamwork process (e.g. by discussing it with fellow students,
writing a reflection, through an evaluation questionnaire).
- … students are (also) individually evaluated based on the teamwork.
a
INT ¼ Integration of STEM content; PCL ¼ problem-centered learning; IBL ¼ inquiry-based learning; DBL ¼ design-based learning; COL ¼ cooperative learning.
204 L. Thibaut et al. / Teaching and Teacher Education 71 (2018) 190e205
Funding
This work was supported by the IWT-SBO [grant name
STEM@school].
Disclosure statement
No potential conflict of interest was reported by the authors.
References
van Aalderen-Smeets, S. I., & Walma van der Molen, J. H. (2015). Improving primary
teachers' attitudes toward science by attitude-focused professional develop-
ment. Journal of Research in Science Teaching, 52(5), 710e734.
van Aalderen-Smeets, S. I., Walma van der Molen, J. H., & Asma, L. J. (2012). Primary
teachers' attitudes toward science: A new theoretical framework. Science Edu-
cation, 96(1), 158e182.
Ajzen, I. (1988). Attitudes, personality, and behavior. Chicago, IL: Dorsey Press.
Ajzen, I. (2005). Attitudes, personality, and behavior. London: McGraw-Hill
Education.
Al Salami, M. K., Makela, C. J., & de Miranda, M. A. (2017). Assessing changes in
teachers' attitudes toward interdisciplinary STEM teaching. International Journal
of Technology and Design Education, 27(1), 63e88.
Appleton, K., & Kindt, I. (1999). Why teach primary science? Influences on begin-
ning teachers' practice. International Journal of Science Education, 21(2),
155e168.
Austin, J. D., Hirstein, J., & Walen, S. (1997). Integrated mathematics interfaced with
science. School Science & Mathematics, 97(1), 45e49.
Ayub, A. F. M., Bakar, K. A., & Ismail, R. (2015, October). Factors predicting teachers'
attitudes towards the use of ICT in teaching and learning. In I. Mohamed,
L. T. How, A. C. Y. Mui, & W. K. Bin (Eds.), AIP conference proceedings (Vol. 1682,
No. 1, p. 030010). AIP Publishing.
Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory.
Englewood Cliffs, NJ: Prentice Hall.
Beck, L., & Ajzen, I. (1991). Predicting dishonest actions using the theory of planned
behavior. Journal of Research in Personality, 25(3), 285e301.
Bingimlas, K. A. (2009). Barriers to the successful integration of ICT in teaching and
learning environments: A review of the literature. Eurasia Journal of Mathe-
matics, Science and Technology Education, 5(3), 235e245.
Bryan, L. A., Moore, T. J., Johnson, C. C., & Roehrig, G. H. (2015). Integrated STEM
education. In C. C. Johnson, E. E. Peters-Burton, & T. J. Moore (Eds.), STEM
roadmap: A framework for integration (pp. 23e37). London: Taylor & Francis.
Bunt, N. A. (2008). Elementary teachers' perceptions regarding four research-based
components of reading instruction. Doctoral dissertation. Vermillion, South
Dakota: University of South Dakota. Retrieved on May 17, 2017 from http://
gradworks.umi.com/33/51/3351187.html.
Bush, W. S. (1989). Mathematical anxiety in upper elementary school teachers.
School Science & Mathematics, 89(6), 499e509.
Capps, D. K., & Crawford, B. A. (2013). Inquiry-based instruction and teaching about
nature of science: Are they happening? Journal of Science Teacher Education,
24(3), 497e526.
Chen, H.-T. (1998). Theory-driven evaluations. In A. J. Reynolds, & H. J. Walberg
(Eds.), Advances in educational productivity, vol. 7: Evaluation research for
educational productivity (pp. 15e34). Greenwich, CT: Elsevier Science/JAI Press.
Choi, B. C., & Pak, A. W. (2006). Multidisciplinarity, interdisciplinarity and trans-
disciplinarity in health research, services, education and policy: 1. Definitions,
objectives, and evidence of effectiveness. Clinical and Investigative Medicine,
29(6), 351e364.
Cohen, J., McCabe, L., Michelli, N. M., & Pickeral, T. (2009). School climate: Research,
policy, practice, and teacher education. Teachers College Record, 111(1), 180e213.
Cronin-Jones, L. L. (1991). Science teacher beliefs and their influence on curriculum
implementation: Two case studies. Journal of Research in Science Teaching, 28(3),
235e250.
Czerniak, C. M., Weber, W. B., Sandmann, A., & Ahern, J. (1999). A literature review
of science and mathematics integration. School Science & Mathematics, 99(8),
421e430.
De Hei, M. S. A., Strijbos, J. W., Sjoer, E., & Admiraal, W. (2015). Collaborative learning
in higher education: Lecturers' practices and beliefs. Research Papers in Educa-
tion, 30(2), 232e247.
DeChenne, S. E., Koziol, N., Needham, M., & Enochs, L. (2015). Modeling sources of
teaching self-efficacy for science, technology, engineering, and mathematics
graduate teaching assistants. CBE Life Sciences Education, 14(3), 1e14.
Diggs, V. (2009). Ask-think-create: The process of inquiry. Knowledge Quest, 37(5),
30e33.
Dilekli, Y., & Tezci, E. (2016). The relationship among teachers' classroom practices
for teaching thinking skills, teachers' self-efficacy towards teaching thinking
skills and teachers' teaching styles. Thinking Skills and Creativity, 21, 144e151.
Donaghue, H. (2003). An instrument to elicit teachers' beliefs and assumptions. ELT
Journal, 57(4), 344e351.
Drake, S. M., & Burns, R. C. (2004). Meeting standards through integrated curriculum.
Alexandria, VA: ASCD.
Durlak, J. (1998). Why program implementation is important. Journal of Prevention
& Intervention in the Community, 17(2), 5e18.
Eagly, A. H., & Chaiken, S. (1993). The psychology of attitudes. Fort worth, TX: Har-
court Brace Jovanovich College Publishers.
El-Deghaidy, H., & Mansour, N. (2015). Science teachers' perceptions of STEM ed-
ucation: Possibilities and challenges. International Journal of Learning and
Teaching, 1(1), 51e54.
Enders, C. K., & Bandalos, D. L. (2001). The relative performance of full information
maximum likelihood estimation for missing data in structural equation models.
Structural Equation Modeling, 8, 430e457.
English, L. D. (2016). STEM education K-12: Perspectives on integration. Interna-
tional Journal of STEM Education, 3(1), 1e8.
Erlandson, C., & McVittie, J. (2001). Student voices on integrative curriculum. Middle
School Journal, 33(2), 28e36.
Faour, B. (2003). Early childhood teachers in Lebanon: Beliefs and practices. Doctoral
theses. UK: School of education, University of Leicester. Retrieved on April 27,
2017 from https://lra.le.ac.uk/handle/2381/4508.
Friday Institute for Educational Innovation. (2012). Teacher efficacy and attitudes
toward STEM survey. Raleigh, NC: Author.
Fullan, M. G. (1993). Change forces: Probing the depths of education reform. New York,
NY: The Falmer Press.
Furner, J., & Kumar, D. (2007). The mathematics and science integration argument:
A stand for teacher education. Eurasia Journal of Mathematics, Science and
Technology, 3(3), 185e189.
Gerbing, D. W., & Anderson, J. C. (1988). An updated paradigm for scale develop-
ment incorporating unidimensionality and its assessment. Journal of Marketing
Research, 25(2), 186e192.
Goodpaster, K. P., Adedokun, O. A., & Weaver, G. C. (2012). Teachers' perceptions of
rural STEM teaching: Implications for rural teacher retention. Rural Educator,
33(3), 9e22.
Gutherie, J. T., Wigfield, A., & VonSecker, C. (2000). Effects of integrated instruction
on motivation and strategy use in reading. Journal of Educational Psychology,
92(2), 331e341.
Hartzler, D. S. (2000). A meta-analysis of studies conducted on integrated curriculum
programs and their effects on student achievement. Unpublished doctoral
dissertation. Bloomington: Indiana University.
Hashweh, M. Z. (1996). Effects of science teachers' epistemological beliefs in
teaching. Journal of Research in Science Teaching, 33(1), 47e63.
Henderson, C., Beach, A., & Finkelstein, N. (2011). Facilitating change in under-
graduate STEM instructional practices: An analytic review of the literature.
Journal of Research in Science Teaching, 48(8), 952e984.
Henderson, C., & Dancy, M. H. (2007). Barriers to the use of research-based
instructional strategies: The influence of both individual and situational char-
acteristics. Physical Review Physics Education Research, 3(2), 1e14.
Hernandez, P. R., Bodin, R., Elliott, J. W., Ibrahim, B., Rambo-Hernandez, K. E.,
Chen, T. W., et al. (2014). Connecting the STEM dots: Measuring the effect of an
integrated engineering design intervention. International Journal of Technology
and Design Education, 24(1), 107e120.
Hinde, E. T. (2005). Revisiting curriculum integration: A fresh look at an old idea.
The Social Studies, 96(3), 105e111.
Ho, M. K., Yang, H. J., & Yang, H. H. (2016). Design and verify an instrument of
assessing attitude toward STEM teaching. International Journal of Education and
Information Technologies, 10, 41e50.
Hurley, M. (2001). Reviewing integrated science and mathematics: The search for
evidence and definitions from new perspectives. School Science & Mathematics,
101(5), 259e268.
Ingvarson, L., Beavis, A., Bishop, A., Peck, R., & Elsworth, G. (2004). Investigation of
effective mathematics teaching and learning in Australian secondary schools.
Melbourne: Australian Council for Educational Research.
Johnson, D. W., Johnson, R. T., & Smith, K. A. (1998). Active learning: Cooperation in
the college classroom. Edina, MN: Interaction Book Company.
Jolliffe, W. (2005, September). The implementation of cooperative learning in the
classroom. In Paper presented at the British educational research association
annual conference. University of Glamorgan.
Jonassen, D. H. (2000). Toward a design theory of problem solving. Educational
Technology Research & Development, 48(4), 63e85.
Ke, F. (2014). An implementation of design-based learning through creating
educational computer games: A case study on mathematics learning during
design and computing. Computers & Education, 73, 26e39.
Kelley, T. R., & Knowles, J. G. (2016). A conceptual framework for integrated STEM
education. International Journal of STEM Education, 3(1), 1e11.
Kline, R. B. (2010). Principles and practice of structural equation modeling. New York,
NY: Guilford Press.
Kynigos, C., & Argyris, M. (2004). Teacher beliefs and practices formed during an
innovation with computer- based exploratory mathematics in the classroom.
Teachers and Teaching, 10(3), 247e273.
Lumpe, A. T., Haney, J. J., & Czerniak, C. M. (2000). Assessing teachers' beliefs about
their science teaching context. Journal of Research in Science Teaching, 37(3),
275e292.
Lund, T. J., & Stains, M. (2015). The importance of context: An exploration of factors
influencing the adoption of student-centered teaching among chemistry,
biology, and physics faculty. International Journal of STEM Education, 2(1), 1e21.
Maio, G., & Haddock, G. (2014). The psychology of attitudes and attitude change.
London, England: Sage Publications Ltd.
Malaeb, Z. A., Summers, J. K., & Pugesek, B. H. (2000). Using structural equation
 L. Thibaut et al. / Teaching and Teacher Education 71 (2018) 190e205
modeling to investigate relationships among ecological variables. Environmental
and Ecological Statistics, 7(1), 93e111.
Marshall, J. C., Horton, R., Igo, B. L., & Switzer, D. M. (2009). K-12 science and
mathematics teachers' beliefs about and use of inquiry in the classroom. In-
ternational Journal of Science and Mathematics Education, 7(3), 575e596.
Marsh, H. W., Balla, J. R., & Hau, K. T. (1996). An evaluation of incremental fit indices:
A clarification of mathematical and empirical processes. In G. A. Marcoulides, &
R. E. Schumacker (Eds.), Advanced structural equation modeling techniques (pp.
315e353). Hillsdale, MI: Erlbaum.
Martin, M. O., & Mullis, I. V. S. (Eds.). (2013). TIMSS and PIRLS 2011: Relationships
among reading, mathematics, and science achievement at the fourth grade-
dimplications for early learning. Chestnut Hill, MA: TIMSS & PIRLS International
Study Center, Boston College.
Matthews, R. S. (1995). Building bridges between cooperative and collaborative
learning. Change, 27(4), 35e40.
Mehr, A. R., Kazemi, S. A., & Omidvari, A. (2015). The effect of teachers' beliefs on
their instructional practices regarding teaching reading strategies. Indian Jour-
nal of Fundamental and Applied Life Sciences, 5(1), 3614e3631.
Menmuir, J., & Adams, K. (1997). Young children's inquiry learning in mathematics.
Early Years, 17(2), 34e39.
Merrill, M. (2007). A task-centered instructional strategy. Journal of Research on
Technology in Education, 40(1), 5e22.
Merrill, M. D. (2009). First principles of instruction. In C. M. Reigeluth, & A. A. Carr-
Chellman (Eds.), Instructional-design theories and models: Building a common
knowledge base (Vol. III, pp. 41e56). New York, NY: Routledge.
Merrill, M. D., & Gilbert, C. G. (2008). Effective peer interaction in a problem-
centered instructional strategy. Distance Education, 29(2), 199e207.
Mobley, M. C. (2015). Development of the SETIS instrument to measure teachers' self-
efficacy to teach science in an integrated STEM framework. Doctoral dissertation.
Knoxville, Tennessee: University of Tennessee. Retrieved on May 23, 2017 from
http://trace.tennessee.edu/utk_graddiss/3354/.
Moore, T. J., Guzey, S. S., & Brown, A. (2014). Greenhouse design to increase
habitable land: An engineering unit. Science Scope, 37(7), 51e57.
Moore, T. J., & Smith, K. A. (2014). Advancing the state of the art of STEM integration.
Journal of STEM Education : Innovations and Research, 15(1), 5e10.
Morrison, J. (2006). TIES STEM education monograph series, Attributes of STEM edu-
cation. Baltimore, MD: TIES.
National Academies of Science. (2007). Rising above the gathering storm. Report from
the Committee on Prospering in the Global economy of the 21st Century.
Washington; DC: National Academies Press.
Nespor, J. (1987). The role of beliefs in the practice of teaching. Journal of Curriculum
Studies, 19(4), 317e328.
Nnenna, E. E., & Olanrewaju, M. K. (2015). Teaching practice anxiety sources as
correlates of teaching performance among student teachers in Federal College
of Education in Southwestern Nigeria. European Scientific Journal, 11(22),
181e196.
NRC [National Research Council]. (2000). Inquiry and the national science education
standards: A guide for teaching and learning. Washington, DC: The National
Academies Press.
NRC [National Research Council]. (2011). Successful K-12 STEM Education: Identifying
effective approaches in science, technology, engineering, and mathematics.
Washington, DC: The National Academies Press.
NSPE [National Society of Professional Engineers]. (2013). Science, technology, en-
gineering, and mathematics education. (NSPE position statement No. 1768).
Retrieved on June 5, 2017 from https://www.nspe.org/sites/default/files/
resources/GR%20downloadables/STEM%20Education.pdf.
Nunnaly, J. (1978). Psychometric theory. New York, NY: McGraw-Hill.
O'Neill, T., Yamagata, L., Yamagata, J., & Togioka, S. (2012). Teaching STEM means
teacher learning. Phi Delta Kappan, 94(1), 36e40.
Pajares, M. F. (1992). Teachers' beliefs and educational research: Cleaning up a
messy construct. Review of Educational Research, 62(3), 307e332.
Perrenet, J. C., Bouhuijs, P. A. J., & Smits, J. G. M. M. (2000). The suitability of
problem-based learning for engineering education: Theory and practice.
Teaching in Higher Education, 5(3), 345e358.
Peters-Burton, E. E., Lynch, S. J., Behrend, T. S., & Means, B. B. (2014). Inclusive STEM
high school design: 10 critical components. Theory Into Practice, 53(1), 64e71.
Pettus, A. M. (1994, October). Models for curriculum integration in high school. In
Paper presented at the annual meeting of the Southeastern regional association of
teacher educators, Shenandoah Valley, Virginia. Retrieved on April 2, 2017 from
http://files.eric.ed.gov/fulltext/ED389706.pdf.
Pinnell, M., Rowly, J., Preiss, S., Franco, S., Blust, R., & Beach, R. (2013). Bridging the
gap between engineering design and PK-12 curriculum development through
the use the STEM Education Quality Framework. Journal of STEM Education :
Innovations and Research, 14(4), 28.
Prather, J. P. (1996). The role of the school principal in science education reform. In
205
J. Rhoton, & P. Bowers (Eds.), Issues in science education (pp. 15e23). Arlington,
VA: National Science Teachers Association and National Science Education
Leadership Association.
Preece, P. F. W. (1979). Student teacher anxiety and class-control problems on
teaching practice: A cross-lagged panel analysis. British Educational Research
Journal, 5(1), 13e19.
Purser, R. K., & Renner, J. W. (1983). Results of two tenth-grade biology teaching
procedures. Science Education, 67(1), 85e98.
Reed, P. A. (2003). Inquiry in technology education. In K. D. Helgeson, &
A. E. Schwaller (Eds.), Selecting instructional strategies for technology education
(pp. 117e129). Peoria, IL: Glencoe/McGraw- Hill.
Rockland, R., Bloom, D. S., Carpinelli, J., Burr-Alexander, L., Hirsch, L. S., & Kimmel, H.
(2010). Advancing the ‘‘E’’ in K-12 STEM education. Journal of Technology Studies,
36(1), 53e64.
Roehrig, G. H., & Kruse, R. A. (2005). The role of teachers' beliefs and knowledge in
the adoption of a reform-based curriculum. School Science & Mathematics,
105(8), 412e422.
Roehrig, G. H., Kruse, R. A., & Kern, A. (2007). Teacher and school characteristics and
their influence on curriculum implementation. Journal of Research in Science
Teaching, 44(7), 883e907.
Rother, E. T. (2007). Systematic literature review X narrative review. Acta Paulista de
Enfermagem, 20(2), 5e6.
Sade, D., & Coll, R. (2003). Technology and technology Education: Views of some
solomon island primary teachers and curriculum development officers. Inter-
national Journal of Science and Mathematics Education, 1(1), 87e114.
Sanders, M. (2009). STEM, STEM education, STEMmania. The Technology Teacher,
68(4), 20e26.
Savasci-Acikalin, F. (2009). Teacher beliefs and practice in science education. Asia-
Pacific Forum on Science Learning and Teaching, 10(1), 1e14.
Slavin, R. E. (1987). Developmental and motivational perspectives on cooperative
learning: A reconciliation. Child Development, 58(5), 1161e1167.
Smith, K. A., & Waller, A. A. (1997). Cooperative learning for new college teachers. In
W. E. Campbell, & K. A. Smith (Eds.), New paradigms for college teaching (pp.
183e209). Edina, MN: Interaction Book Company.
Stains, M., & Vickrey, T. (2017). Fidelity of implementation: An overlooked yet
critical construct to establish effectiveness of evidence-based instructional
practices. Life Science Education, 16(1), 1e11.
Stohlmann, M. S., Moore, T. J., & Cramer, K. (2013). Preservice elementary teachers'
mathematical content knowledge from an integrated STEM modelling activity.
Journal of Mathematical Modelling and Application, 1(8), 18e31.
Stohlmann, M., Moore, T. J., & Roehrig, G. H. (2012). Considerations for teaching
integrated STEM education. Journal of Pre-College Engineering Education
Research (J-PEER), 2(1), 28e34.
Stoll, L., Bolam, R., McMahon, A., Wallace, M., & Thomas, S. (2006). Professional
learning communities: A review of the literature. Journal of Educational Change,
7(4), 221e258.
Sunal, D. W., Hodges, J., Sunal, C. S., Whitaker, K. W., Freeman, L. M.,
Edwards, L., … Odell, M. (2001). Teaching science in higher education: Faculty
professional development and barriers to change. School Science & Mathematics,
101(5), 246e257.
Tseng, K. H., Chang, C. C., Lou, S. J., & Chen, W. P. (2013). Attitudes towards science,
technology, engineering and mathematics (STEM) in a project-based learning
(PjBL) environment. International Journal of Technology and Design Education,
23(1), 87e102.
Van Merrie €nboer, J. J. (1997). Training complex cognitive skills: A four-component
instructional design model for technical training. Englewood Cliffs, NJ: Educa-
tional Technology.
Van Merrie €nboer, J. J. G., & Kirschner, P. A. (2007). Ten steps to complex learning: A
systematic approach to four-component instructional design. London, United
Kingdom: Lawrence Erlbaum Associates, Publishers.
Vars, G. F. (2001). Can curriculum integration survive in an era of high-stakes
testing? Middle School Journal, 33(2), 7e17.
Wang, H. H., Moore, T. J., Roehrig, G. H., & Park, M. S. (2011). STEM integration:
Teacher perceptions and practice. Journal of Pre-College Engineering Education
Research, 1(2), 1e13.
Weilbacher, G. (2001). Is curriculum integration an endangered species? Middle
School Journal, 33(2), 18e27.
Xie, M., & Sharif, R. T. S. (2014). The relationship between teachers' knowledge,
attitude and belief with the implementation of inquiry-based learning in
Zhengzhou, China. International Journal of Learning, Teaching and Educational
Research, 8(1), 149e161.
Zee, M., & Koomen, H. M. (2016). Teacher self-efficacy and its effects on classroom
processes, student academic adjustment, and teacher well-being: A synthesis of
40 years of research. Review of Educational Research, 86(4), 981e1015.