Professional Documents
Culture Documents
Language, Learning, and Technology Journal - Volume 16 Number 3 October 2012
Language, Learning, and Technology Journal - Volume 16 Number 3 October 2012
October 2012
Articles Columns
Eye Tracking as a Measure of Noticing: Emerging Technologies
A Study of Explicit Recasts in SCMC Edited by Robert Godwin-Jones
Abstract | Article PDF
Robot Assisted Language Learning
Bryan Smith, Arizona State University
Article PDF
pp. 53–81
by Jeonghye Han
pp. 1–9
Promoting Learner Autonomy through
Multiliteracy Skills Development in
Action Research
Cross-Institutional Exchanges
Edited by Greg Kessler
Abstract | Article PDF
Carolin Fuchs, Teachers College, Promoting EFL Students’ Inferential Reading Skills
Columbia University through Computerized Dynamic Assessment
Mirjam Hauck, The Open University Article PDF
Andreas Müller-Hartmann, University of by Adeline Teo
Education, Heidelberg pp. 10–20
pp. 82–102
Announcements &
Call for Papers
Announcements
News From Sponsoring Organizations
Article PDF
pp. 32–34
Learning & Technology is a refereed journal that began publication in July 1997. LLT disseminates
research to foreign and second language educators worldwide on issues related to technology and
language education.
Language Learning & Technology is sponsored and funded by the University of Hawai'i National
Foreign Language Resource Center (NFLRC) and the Michigan State University Center for
Language Education and Research (CLEAR).
Language Learning & Technology is a fully refereed journal with an editorial board of scholars in
the fields of second language acquisition and computer-assisted language learning. The focus of
the publication is not technology per se, but rather issues related to language learning and
language teaching, and how they are affected or enhanced by the use of technologies.
Language Learning & Technology is published exclusively on the World Wide Web. In this way,
the journal seeks to (a) reach a broad audience in a timely manner, (b) provide a multimedia
format which can more fully illustrate the technologies under discussion, and (c) provide
hypermedia links to related background information.
Back and current issues of Language Learning & Technology are indexed in the Current
Contents/Social and Behavioral Sciences, ISI Alerting Services, Institute for Scientific
Information's (ISI) Social Sciences Citation Index (SSCI), Linguistics Abstracts, PsycINFO, and
Social Scisearch databases.
Language Learning & Technology is currently published three times per year (February, June,
and October).
SPONSORS
University of Hawai‘i National Foreign Language Resource Center (NFLRC)
Michigan State University Center for Language Education and Research (CLEAR)
Editorial Board
Nike Arnold Portland State University
Robert Blake University of California, Davis
Klaus Brandl University of Washington
Thierry Chanier Universite Blaise Pascal
Scott Crossley Georgia State University
Tracey Derwing University of Alberta
Lara Ducate University of South Carolina
Robert Godwin-Jones Virginia Commonwealth University
Regine Hampel The Open University
Debra Hardison Michigan State University
Claire Kennedy Griffith University, Brisbane
Markus Kötter University of Siegen
Eva Lam Northwestern University
Jenifer Larson-Hall Kyushu University
Joshua Lawrence University of California, Irvine
Lina Lee University of New Hampshire
Meei-Ling Liaw National Taichung University
Lara Lomicka University of South Carolina
Paul Kei Matsuda Arizona State University
Jill Pellettieri Santa Clara University
Mark Pegrum University of Western Australia
Hayo Reinders Middlesex University, London
Bryan Smith Arizona State University
Patrick Snellings University of Amsterdam
Susana Sotillo Montclair State University
Editorial Staff
Editors Dorothy Chun University of California, Santa
Barbara
Mark Warschauer University of California, Irvine
Associate Editors Trude Heift Simon Fraser University
Philip Hubbard Stanford University
Rick Kern University of California, Berkeley
Marie-Noëlle Lamy The Open University
Carla Meskill State University of New York,
Albany
Glenn Stockwell Waseda University
Managing Editor, Incoming Gerriet Janssen University of Hawai‘i at Manoa
Managing Editor, Outgoing Daniel Jackson University of Hawai‘i at Manoa
Web Production Editor Carol Wilson-Duffy Michigan State University
Book & Multimedia Review Paige Ware Southern Methodist University
Editor
Action Research Column Editor Greg Kessler Ohio University
Emerging Technologies Editor Robert Godwin-Jones Virginia Commonwealth University
Social Media Director Chin-Hsi Lin University of California, Irvine
Copy Editors Rebecca Estes University of California, Davis
Jake Kletzien Daegu National University of
Education
Troy Rubesch Kanda University of International
Studies
EMERGING TECHNOLOGIES
ROBOT ASSISTED LANGUAGE LEARNING
Jeonghye Han, Cheongju National University of Education
Along with the rapid development of information and communication technologies, educators are trying
to keep up with the dramatic changes in our electronic environment. These days mobile technology, with
popular devices such as iPhones, Android phones, and iPads, is steering our learning environment
towards increasingly focusing on mobile learning or m-Learning. Currently, most interfaces employ
keyboards, mouse or touch technology, but some emerging input-interfaces use voice- or marker-based
gesture recognition. In the future, one of the cutting-edge technologies likely to be used is robotics.
Robots are already being used in some classrooms and are receiving an increasing level of attention.
Robots today are developed for special purposes, quite similar to personal computers in their early days.
However, in the future, when mass production lowers prices, robots will bring about big changes in our
society. In this column, the author focuses on educational service robots. Educational service robots for
language learning and robot-assisted language learning (RALL) will be introduced, and the hardware and
software platforms for RALL will be explored, as well as implications for future research.
These robots not only have a tablet interface but are also capable of autonomous movement, visual
recognition through a camera, voice recognition through a microphone, and physical interaction based on
various sensors, such as a touch sensor, infrared sensor, ultrasonic sensor, bumper sensor, or floor
detection sensor. Computers or mobile devices also can be said to be capable of nonverbal
communication employing cyber characters or videos. However, robots are notable in their capacity for
nonverbal communication, such as facial expressions, gestures and actions, while coexisting with users in
a real environment, such as the home or the classroom. Also, robots are different from computers and
mobile devices in that they have a friendly appearance, a name, a birth story, a personality, and are
capable of social relations. Moreover, robots with a computer display on their body can provide mobile
services just the same as computer and mobile devices.
Educational service robots have started to appear in Canada, Japan, South Korea, Taiwan, and the United
States. To take a closer look at the use of educational robots in each case, there are: Canada’s PEBBLES
(Fels & Weiss, 2001); Japan’s ROBOVIE (Kanda, Hirano, Eaton, & Ishiguro, 2004) as a peer tutor for
elementary English, PAPERO (NEC, 2004) for child-care, Keepon (Beatbots, 2007) as an interaction
inducer for autistic children, and Saya (Shuster, 2010) for use in Japanese preschools; South Korea’s
preschool teaching assistant, IROBI (Han, Jo, Kim, & Park, 2005), IROBIQ (Hyun, Kim, Jang, & Park,
2008), primary school English teaching ROTI and ENGKEY (Yonhapnews, 2009), and primary school
English teaching assistant, ROBOSEM (Park, Han, Kang, & Shin, 2011); Taiwan’s elementary English
teaching assistant ROBOSAPIEN (You, Shen, Chang, Liu, & Chen, 2006); and, the United States’
preschool peer tutor, RUBI and QURIO (Movellan, Tanaka, Fortenberry, & Aisaka, 2005; Movellan,
Eckhardt, Virnes, & Rodriguez, 2009), and VGO (Vgo communications, 2011) for young patients.
Educational service robot hardware platforms
Table 1 shows a brief description of the hardware platforms for the major education service robots.
Most tele-presence robots in Table 1 feature a simple Wi-Fi video conferencing function and remote
controlled movements. Also, to cater to the convenience of the remote controller, these robots are usually
equipped with compulsory wheels for travel, but have no arm or head movement. They are mostly
developed at a low-cost. Tele-presence robots are predominantly used by native speakers or English
teachers, and the interaction is comparatively limited since they only offer video conferencing and limited
mobility. They can, however, sustain long-term interaction, but when there is a breakdown in the network,
all interactions stop.
The autonomous robots in Table 1 feature complex body and arm movement, but they are quite expensive
for the general market and mostly used for research. Also, for the robots to interact autonomously, they
need preprogrammed scenarios. With current artificial intelligence technology, however, seamless
interaction is unattainable, thus the novelty effect can wear off quickly. Still, the advantage is that
interaction is possible regardless of network problems.
In the case of the transformed type of robots in Table 1, these robots have been developed to perform an
assisting role rather than to interact autonomously. The teacher can control the robots by remote control
or voice recognition, and when they are in tele-operated mode, a native speaker can control them. If
network problems occur during the tele-operated mode, they can swiftly be switched to the autonomous
mode. Hence interaction is always possible regardless of network stability, and continuous interaction is
supported.
Educational service robot software platforms
Most operating systems for intelligent robots are based on MS Windows and Linux. Since the majority of
robot companies own unique robot hardware and exclusive software platforms, developing robot
applications is difficult. This has decelerated the popularization of robots and the opening up of robot
markets. To address this predicament, the worldwide movement towards developing an integrated
software platform started in earnest in 2000. Since then, many countries have developed their own robotic
software platform with slightly different characteristics, such as Microsoft Robotics Developer Studio
(Microsoft RDS), Willow Garage’s ROS (Robot Operating System), the EU’s RT CORBA-based
standard platform OROCOS (The Open Robot Control Software), Japan’s National Institute of Advanced
Industrial Science and Technology (AIST) driven RT middleware, and Korea’s OPRoS (Open Platform
for Robotic Service), and so on.
MS RDS is software oriented towards developers who are restricted to developing robot applications
using preset robot hardware. In 2010, Willow Garage provided open source robotic software called ROS.
ROS is engaged with a community of one hundred thousand robot developers worldwide, and from this
foundation, various robot applications can be programmed, leading to the effective promotion and
expansion of robot proliferation (http://www.ros.org). Recently, Willow Garage has received attention for
having integrated their software with the Smartphone OS Android.
These robot software platforms enable programmers to develop various applications so that robots can
perform multiple functions. By integrating a robot’s modular features, such as voice- and sound-
recognition, face-detection and recognition, gesture- and object-recognition, synthetic speech, driving,
space recognition, and position recognition, these software platforms enable the robot to provide various
services. For example, in an instance of RALL service, the robot can easily look at a learner who is
calling its name and move to a corresponding location. Most robots shown in Table 1 have their own
robot software platforms. Efforts are being made, however, to integrate these robots to an open platform,
such as ROS, and soon, when standardization is realized, the robot service market will rapidly expand.
employ native speakers, teachers can always utilize computer- or mobile-based applications. Yet, due to
limitations in current image recognition technology, most applications based on computers or mobiles
focus on voice-based verbal messages.
As an alternative, video conferencing with native speakers can be considered. Video conferencing can be
effective in demonstrating pronunciation. However, this is not popular because there are still limitations
in pronunciation diagnosis since the native speaker needs to be close to individual pupils or the classroom
as a whole. Recently, some after-schools programs in Korea have been conducting one-on-one native
speaker tutoring. It is true that the educational benefits of individualized interaction are substantial, but
there also are several drawbacks: it is costly for learners; when miscommunications occur there is no help
available in the learner’s mother tongue; and, finally, different native speakers may be assigned to
different classes according to their schedules and this would impede learners’ progress-tracking.
Figure 3. A traditional collaborative class, a video conferencing class, and a one-on-one conference in
Korea.
Among the various instructional models in language learning, we should consider RALL, employing
currently emerging robot technology. This anthropomorphized version of existing mobile devices is
autonomous, with features such as image recognition through camera, voice recognition through
microphone, and interaction based on various sensors. Robot-assisted learning (RAL) can be defined as
learning assisted by robots with such features (Han, 2010). In the domain of RAL, RALL can be defined
as targeting language learning in particular. Most robots in RALL are interlinked with instructional
material, and can perform the role of the native speaker to interact with learners. The following scenes in
Figure 4 show English classes being conducted with IROBIQ and ROBOSEM.
Figure 4. RALL with IROBIQ (left) & RALL with ROBOSEM (right from Park, Han, Kang & Shin,
2011).
RALL shares the merits of the conventional collaboration model—the face-to-face, physical interaction
with native speakers—but it is easier to recruit native speakers for tele-conferencing. Furthermore, RALL
also has the advantage of an instruction model employing applications, by sharing instructional material
over a TV or display device, such as a projector beam. Class activities, such as English chanting, can be
recorded. Using radio frequency identification (RFID) tags, individual progress can be logged and
tracked. Figure 5 shows a sample of a RALL lesson plan. When compared to common computer-assisted
language learning (CALL) classes, robots need to be programmed to interact with the teacher and pupils.
Grade 3th grade students on a Korean elementary school Period Formal learning
Unit 4. Happy birthday! (3/3)
Theme Celebrating birthday, expressing thanks
The students are able to express birthday greetings and thanks.
The students are able to give and receive presents (cards) with birthday
The purpose of the lesson greetings.
The students are able to make conversation, using expression of birthday
greetings and thanks.
Materials ROBOSEM, Gift-picture cards, Bracelet RFID tag
Treatment time 40 minutes
ROBOT TIME
STEP PROCEDURE TEACHING & LEARNING ACTIVITIES
CONTROL
Teacher: Good morning everyone.
Students: Good morning teacher, Kim.
Recognition of
Teacher: Let's ask ROBOSEM together, ‘How are
voice localization
you?’
Students: How are you?
Intro-
Greeting ROBOSEM: I'm good. I'm excited today because 3 mins
duction Voice recognition
today is my birthday.
Teacher: I've heard that today is your birthday,
Recognition of
ROBOSEM. Boys and girls, let's say happy
voice localization
birthday to langbot.
Voice recognition Students: Happy birthday ROBOSEM!
The most popular service was TIRO’s praising and cheering, followed by face-to-face conversation and
role-play with TIRO. Han (2010), based on previous studies, found that in Japan and the United States,
robots are mostly given the role of a peer tutor, while in Korea, they are given the role of a teaching
assistant and friend. Japanese robots provide service focusing on voice- and gesture-based interaction,
while Korean and American robots base their service on visual instruction materials, such as Flash
contents. It was also concluded that the subjects most suitable for RAL would be language learning, such
as in English education, where robots are effective in inducing motivation and enhancing achievement,
synchronizing visual contents with instruction, and playing the teaching assistant role necessary for long-
term interaction. Park, Han, Kang, and Shin (2011) designed class materials to evoke children’s
motivation, to sustain their interaction with ROBOSEM, and to facilitate English learning. Their study
showed meaningful results after four weeks of English classes conducted using ROBOSEM.
CHALLENGES OF RALL
As promising results of the educational effectiveness of RALL have appeared in numerous studies, many
countries that teach English as a foreign language have begun to take the first step in RALL’s research
and commercialization. Korea is one such country where RALL is in full swing, with already over 1,500
robots employed in preschool play activities and attitude training, as well as over 30 English education
robots currently in active use in elementary school afterschool programs. As RALL expands, there will be
more emerging challenges for language educationalists to explore and address.
First of all, research on the system framework of RALL is necessary. The system framework of RALL
consists of elements such as robot hardware, robot applications and visual contents. Each element needs
to be designed in consideration of the language learning goals. Also, not only language educationalists but
also educational service robot developers need to conduct collaborative research on designing guidelines
for each element and on integrating these elements to carry out RALL successfully.
Secondly, more theoretical research on the current status of educational service robot technology, such as
the RALL instruction model, is desirable. There needs to be consideration of how the teacher and robot
would cooperate in the classroom and of what kind of teaching and learning model would be used in
RALL. Both You et al. (2006) and Han and Kim (2009) have suggested several models, but a more
specific teaching and learning model needs to be developed, reinforced with the language-learning
experts’ perspective regarding the most suitable type of robot (tele-operated, autonomous, and
transformed type).
Third, more field experiments and studies on educational effectiveness are necessary to compare RALL to
the conventional instruction methods of the multimedia program application model, native speaker
collaboration model, and one-on-one videoconferencing model. Comparing educational effectiveness of
such instruction models is very difficult to carry out, due to difficulties in conducting controlled
experiments and its high cost. A Delphi survey with an expert group might be performed as an alternative.
Fourth, research on teacher education is essential since teachers are the ones who will be installing RALL.
We need to find out what teachers need as prior knowledge (for example, basic knowledge of robot
hardware and software), and develop a teacher education program. Also, consideration needs to be given
to areas of possible concern for teachers when conducting RALL, such as whether a child is becoming too
dependent upon the robot.
Fifth, research on various moral and technological issues arising through RALL applications are
necessary. For example, RALL raises several problems, including: the exposure of the learning
environment to outsiders through tele-conferencing; sharing data on class activities; students trusting a
robot over the teacher; and the possibility of obsession with robots. Furthermore, solutions to class
management problems, such as network breakdowns or robot hardware failure, are required.
Recently, with the emergence of an open robot software platform, robot applications enabling a variety of
functions are being developed. Just as many people now have a personal computer, in the near future,
personal robots (PR) may become the next paradigm-shifting tool for everyday life. Thus, as the computer
sparked revolutionary changes in learning environments, studies needs to be conducted on how robots
will bring about changes in education. Especially important will be the preparation of various approaches
for RALL from the perspective of language learning and pedagogy, with a vision to evolve in tandem
with already advanced developments in applications, programs, and services.
REFERENCES
Fels, D.I., & Weiss, P. (2001). Video-mediated communication in the classroom to support sick children:
A case study. International Journal of Industrial Ergonomics, 28, 251–263.
Godwin-Jones, R. (2011). Mobile apps for language learning. Language Learning & Technology, 15(2)
2–11. Retrieved from http://llt.msu.edu/issues/june2011/emerging.pdf
Han, J. (2010). Robot-aided learning and r-learning services. In D. Chugo (Ed.), Human-Robot
Interaction. Retrieved from: http://sciyo.com/articles/show/title/robot-aided-learning-and-r-learning-
services
Han, J., Jo, M., Park, S., & Kim, S. (2005). The Educational Use of Home Robots for Children. In
Proceedings of the 14th IEEE International Workshop on Robot and Human Interactive Communication
(RO-MAN 2005), 378–383. Piscataway, NJ: IEEE.
Han, J., & Kim, D. (2009). r-Learning services for elementary school students with a teaching assistant
robot. In Proceedings of the 4th ACM/IEEE Human Robot Interaction, 255–256. New York, NY: ACM
Hyun, E., Kim, S., Jang, S., & Park, S. (2008). Comparative study of effects of language education
program using intelligence robot and multimedia on linguistic ability of young children. Proceedings of
the 14th IEEE International Workshop on Robot and Human Interactive Communication (RO-MAN 2008).
Piscataway, NJ: IEEE.
Kanda, T., Hirano, T., Eaton, D., & Ishiguro, H. (2004). Interactive robots as social partners and peer
tutors for children: A field trial. Human-Computer Interaction, 19(1–2), 61–84.
Movellan, J.R., Eckhardt, M., Virnes, M., & Rodriguez A. (2009). Sociable robot improves toddler
vocabulary skills. In Proceedings of the 4th ACM/IEEE Human Robot Interaction, 307–308. New York,
NY: ACM.
Movellan, J.R., Tanaka,F., Fortenberry, B., & Aisaka, K. (2005). The RUBI/QRIO Project: Origins,
principles, and first steps. In Proceedings. The 4nd International Conference on Development and
Learning, 80–86. Retrieved from http://doi.ieeecomputersociety.org/10.1109/DEVLRN.2005.1490948
Park, S., Han, J., Kang, B., & Shin, K. (2011). Teaching assistant robot, ROBOSEM, in English class and
practical issues for its diffusion. Proceedings of IEEE A Workshop on Advanced Robotics and its Social
Impacts, http://www.arso2011.org/papers
Shuster, R. (2012, April 1). English teaching robot comes with a cost. Korea IT Times. Retrieved from
http://www.koreaittimes.com/story/8216/english-teaching-robot-comes-cost
Sopher, J. (2011, February 2). Student uses telepresence robot to attend school. Retrieved from
http://geekbeat.tv/student-uses-telepresence-robot-to-attend-school/
Yonhapnews. (2009, December 22). English tutoring Telepresence robot: Roti. Retrieved from
http://www.youtube.com/watch?v=-_wa_SZaqQU
You, Z., Shen, C., Chang, C., Liu, B., & Chen G. (2006). A robot as a teaching assistant in an English
class. In Proceedings of the 6th IEEE International Conference on Advanced Learning Technologies, 87–
91. New York, NY: IEEE.
RESOURCE LIST
students. Providing one-on-one mediation to individual students has become a challenging and
unmanageable task for many EFL teachers. The researcher in this action research project decided to take
advantage of the technology available today and use it to design a computerized dynamic assessment (C-
DA) program to overcome the time-constraint challenge many teachers face in their classrooms. The goal
of the project was to promote her Taiwanese college EFL students‘ inferential reading ability. The
researcher focused on the inferential reading ability because it is an important skill that involves multi-
faceted reading skills. According to Harvey and Goudvis (2000), the inferential reading ability is a
cognitively-demanding skill which requires learners to read between the lines and make educated guesses
on certain outcomes, events, or actions based on their understanding of the reading materials. Therefore,
an EFL student‘s ability in inferential reading not only reflects whether he understands the overall ideas
of the reading materials, but also his own critical thinking skills.
BACKGROUND LITERATURE
Theoretical Concept
The central concept of the computerized dynamic assessment (C-DA) is grounded in Vygotsky‘s
theoretical framework (1978). To challenge the appropriateness of using a static IQ score to predict a
child‘s capabilities in succeeding at school, Vygotsky and his colleagues developed a dynamic assessment
by providing the test-takers with meditational prompts during the testing procedures (van der Veer &
Valsiner, 1991). They concluded that the children‘s reactions to the mediation were not the same, and
different types of mediation benefitted children in different ways. They also claimed that while a static IQ
test helped indicate the children‘s present level of ability, providing mediation to the test-takers during the
assessment allowed teachers to determine the learners‘ assisted performance, which indicated their
potential future IQ.
Vygotsky (1978) also introduced the concept of Zone of Proximal Development (ZPD), which
emphasized that a learner should receive structured intervention from mediators that aimed at promoting
his development. With the assistance of a mediator, a learner can move from his current level where he
can‘t work independently to a level where he can accomplish the same task on his own. The focused
intervention, which comes from a more mature counterpart, can be from a human, a cultural artifact, an
object, or a tool. With the advancement of technology, Dixon-Krauss (1996) suggested the use of
technology to realize Vygotsky‘s vision of designing lessons in a way that facilitates instruction that is
slightly ahead of the learner‘s development. Crook (1991) also claimed that computers could act much
like a human partner or classroom teacher within the ZPD, and technology makes the computerized tool
pertinent to the mediation periods associated with internalization. Therefore, the C-DA program in the
current project was designed based on the belief that when human mediators are not accessible, the social
process necessary for development can be facilitated through computer assisted devices.
Metacognition and Reading Comprehension
Metacognition is defined as thinking about one‘s thoughts (Harris & Hodges, 1995). Perkins (1992)
identified four types of readers based on four different levels of metacognitive knowledge the readers
employ in their reading strategies. The four types of readers are: (1) tacit readers; (2) aware readers; (3)
strategic readers; and (4) reflective readers. Tacit readers are not aware of how they think when they read;
aware readers do not know how to fix their problems although they know that they do not comprehend the
meaning from their reading materials. Strategic readers are those who are capable of using effective
reading strategies and repairing meaning on their own when necessary. Reflective readers are able to
reflect on the strategies they use in reading and can apply and revise the strategies flexibly in different
contexts. Researchers such as Pressley and Afflerback (1995) found that high-skilled readers were able to
use specific and appropriate metacognitive strategies before, during and after reading to help their
comprehension of the texts they read. In Perkins‘ terms (1992), these types of readers would be
categorized as the strategic and reflective readers. Pressley and Afflerback as well as other researchers
such as Block and Mangieri (2003) suggest that it is crucial for teachers to integrate metacognitive
strategies in their literacy instruction to help the learners become independent and self-regulated readers.
Therefore, the C-DA program consisted of mediation that was designed to improve the learners‘
metacognitive strategies, especially with the intention of training them to be strategic and reflective
readers. One of the main purposes in the present study was exploring the effect of the C-DA program on
the participants‘ metacognition.
Research Questions
The current action research was conducted to answer the following two research questions:
What are the effects of the C-DA program on promoting the participants‘ metacognition in their
inferential reading skill?
Is there a difference in the participants‘ performance before and after the use of the C-DA
program?
METHODOLOGY
Participants and Setting
The action research project was carried out at a university in Taiwan. The participants were 68 EFL
college freshmen who were enrolled in a required two-semester course titled ―Freshman English for Non-
English Major Students‖ (FENEMS) taught by the researcher. FENEMS was held two hours weekly for
eighteen weeks each semester. At the beginning of the semester, the participants were assigned to the FENEMS
course by the university‘s General Education Department based on their scores in the English subject of the
Taiwanese national college entrance exam. Their scores ranged from 38 to 90 out of 100, which placed their
English proficiency levels from intermediate low to advanced. There were 33 males and 35 females
participants whose ages ranged from 18 to 19 in this research project. The participants were from five
different departments at the university, namely the Statistics Department, the Information Management
Department, the Public Health Department, the Occupational Therapy Department, and the Respiratory
Treatment Department. None of the participants had experienced computerized dynamic assessment (C-DA)
prior to the current project. The C-DA program was implemented as a part of the curriculum in the FENEMS
course, which took place in a computer lab at the university where each participant had his own computer
and could get access to the Internet to work on the C-DA program individually.
Method Design and Procedure
The research method design took the form of a ―sandwich‖ format (Sternberg and Grigorenko, 2002).
This means that the participants took a traditional static test as a pre-test where no mediation was
provided. Then, in the intervention phase, the computerized mediation was provided for the participants
through the computerized dynamic assessment (C-DA) program. Afterward, each participant took a post-
test. This design allowed the researcher to compare the participants‘ performance before and after the
mediation intervention. Also, during the intervention phase, each participant was given time to record
their reading strategies and reflections in their working portfolio.
The action research project lasted 10 weeks. In the first week, the participants took a pre-test created
specifically for the current research project. The reading passages used for the pre-test were selected from
previous years‘ TOEFL samples. Twelve reading passages were in the pre-test that consisted of 12
multiple-choice inferential reading questions. In the pre-test, the participants worked independently
without receiving any mediation. The pre-test helped the researcher get a sense of the participants‘ current
level in their English inferential reading skills. Based on the mistakes the participants made in their pre-
test, the researcher designed a set of mediation for the C-DA program that would be used in the
mediational phase. The meditational phase lasted eight consecutive weeks, beginning from the second
week. The mediation sessions were held once a week. During each weekly mediation session, the
participants worked on three reading passages. Each reading passage involved an inferential question. In
the tenth week, the participants took a post-test, in which 12 reading passages and inferential questions
were designed in the same format as the pre-test. Since the passages used in the pre-test, mediation, and post-
test stages were adapted from the TOEFL exam samples, which were designed and tested in advance by a team
of professional language test-designers, their validity and reliability were believed to be highly acceptable.
There was no time limit for each C-DA session. The participants were allowed to work on the activities at
their own pace during or after the class session. Although they were informed that they could continue
working on the activities after class if they could not complete the task within the two-hour class time, all
students were able to complete the C-DA sessions during class time throughout the current ten-week
research project. Besides reading passages and answering questions posted on the C-DA program, for
each mediated session, all participants were asked to write down their reflections on the mediation
provided by the C-DA program, record their reading process and use of reading strategies, and supply
concrete self-selected evidence of their growing/changing reading abilities. The participants were
encouraged to do so during—not after—the mediated session when they were presented each slide. The
reason was to allow actual documentation of one‘s reading process in the C-DA program to take place,
rather than solely relying on introspective accounts of perceived strengths, weaknesses, and affective
reactions to the C-DA program. Reflections anchored in concrete evidence of learning experiences
recorded in the participants‘ working portfolio served as a benchmark for the participants to think about
their metacognitive reading development and monitor their performance.
At the beginning of each session before the participants started working on the C-DA program, the
researcher modeled how to record reflections and comments in the working portfolio. However,
participants were not given any specific guidelines concerning what to write; the evidence to be submitted
was controlled and selected by the participants themselves. They were also asked to explain why they
thought the evidence they provided helped them learn the reading strategies. Since there were no fixed
guidelines for the recording of their working portfolio, the participants were provided with unlimited
freedom for self-reflection, critical analysis, and the discovery of new strategic orientations towards their
L2 reading processes and development.
C-DA Mediation Designs
The researcher developed the C-DA program by using the user-friendly Viewlet Quiz 3 software to
integrate mediation with assessment. The software uses Adobe Flash technology and allows educators to
create dynamic and interactive programs that can be saved as executable files and accessed through a web
browser on the Internet. The C-DA program created by the researcher also stored the students‘ responses
and recorded the number of incorrect responses as well as the mediation that were activated by each
student.
The C-DA program allowed the learners to interact with and respond to the preprogrammed computerized
intervention. There were four levels of mediation in the C-DA procedure, and it progressed gradually
from implicit to explicit. After a learner finished reading a passage, he would be asked an inferential
question, followed by five multiple choices. He was asked to select one correct answer from the choices
given. Each time a learner answered a question incorrectly, the computerized mediation would be
presented to him in order of increasing explicitness. The computerized mediation ended automatically
when the learners found the correct answer to the question. The maximum level of mediation each student
received was four. Appendix A shows a flowchart of the arrangement of the mediation slides in the C-DA
program. Also, the following website presents one of the C-DA activities the learners in the current
project worked on: http://tinyurl.com/ch4ws8h.
The following section details the types of assistance each level of mediation provided for the learners.
Mediation Level 1
This level provides the most implicit mediation. The mediation focuses on explaining what inferential
reading means. Then, the mediation involves asking a general question to guide the learners to find the
main idea in the passage, which is an essential step prior to helping them to read between the lines. At this
level, definitions of keywords in the passage as well as some common places where main ideas can be
found are provided to help the learners identify the main idea.
Example: To make inferences successfully, you need to read between the lines. This means that an
inferential question cannot be answered by looking at the text itself. Instead, you will need to use the
information stated in the passage to infer what is not stated. Your job is to make your best guess based on
what you read. First, think about what the main idea of the passage is. The list below gives you the
meanings of the keywords and places where you can possibly find the main ideas. Use the information to
help you.
Mediation Level 2
This level provides more explicit mediation than the one in Level 1. Hints are narrowed down to guide the
learners to focus on certain paragraphs, or sentences, while looking for the correct answer. This more
specific information is followed by an explanation of the overall meaning of the specific context.
Example: Read the third sentence through the last sentence of the passage very carefully. The author is
trying to point out something serious about the society at that time. What is it?
Mediation Level 3
This level provides very explicit mediation. Therefore, the mediation focuses on ONE sentence, phrase, or
word. The explanation given at this level is very context-specific, instead of emphasizing the overall
meaning of the entire passage or specific paragraphs/sentences. The goal is to pinpoint for the learners
how the specific information can lead to the correct answer.
Example: Pay attention to the second sentence. The author said two things “bore little relation.” So, ask
yourself: according to the author, what are the two things that were not related to each other? Since these
two things were not related, it caused some serious societal problems. Think about it, and then answer the
inferential question.
Mediation Level 4
At this level the correct answer is provided and then is followed by a step-by-step explanation. When a
learner arrives at this stage, it usually shows that he is still far from fully mastering the reading strategy
required to understand the concept tested. Thus, this stage focuses on explaining ―how‖ the answer is
obtained so that the learner can follow the instructions to justify the correct answer.
Example: The word “critical” in the 5th sentence gives out a sense of urgency. The author used this word
to express how serious the air pollution problem is. Although she did not say it directly in the passage,
she was implying that the problem needed to be solved urgently because of its critical condition.
RESULTS
Two types of data were collected: (1) the participants‘ written reflections in their working portfolio, and
(2) the p value based on a paired samples t-test. The participants‘ working portfolio was collected to
answer the first research question, investigating the effects of the C-DA program on promoting the
participants‘ metacognition in their inferential reading ability. The p value was calculated to determine
whether there was a statistically significant difference in the participants‘ performance before and after
the use of the C-DA program.
Baker and Brown (2002) stated that we should not exclusively rely on a reader‘s self-report techniques,
since research has found that sometimes one is less able to be introspective about one‘s cognitive
knowledge than one would like. Thus, IBM SPSS Statistics 19 was used to run the paired samples t-test to
determine whether the difference between the pre- and post-test scores was significant as a result of the
computerized mediation. The result shown in Table 1 indicates that there was a significant difference
between the pre-test and post-test scores, t (67) = -2.70, p = 0.009. This means that the participants scored
significantly higher on the post-test (M = 7.28, SD = 3.00) than on the pre-test (M = 6.04, SD = 3.00) in
their inferential reading skill after the computerized mediation.
The qualitative data, which was based on the participant‘s self-reflection in their working portfolio,
generated rich information showing their metacognition (the process of thinking about thinking) in their
reading processes. The participants were encouraged to use their L1, Chinese, to write in their working
portfolio so that they could easily express themselves clearly and in depth. Their writing was then
translated into English and analyzed. According to Farris, Fuhler, and Walther (2004), one‘s awareness of
one‘s own reading strategies is essential for reading success. Appendix B shows several essential
characteristics of metacognition that emerged from the learners‘ responses in their working portfolio.
These characteristics indicated that the computerized mediated assistance helped the participants become
consciously aware of their metacognition and actively monitor and regulate their reading process.
REFLECTION
This action research enlightened the researcher/instructor in several essential ways. First, computerized
dynamic assessment (C-DA) is a powerful tool to help her understand her students‘ potential in learning
to read in English by examining their reaction to different types of mediation. The number of attempts the
participants made and the corresponding answers they chose helped the researcher determine which types
of mediation worked and which did not. The results generated in the C-DA program then helped her
modify and fine-tune mediations to accommodate the learner‘s needs in one-on-one human interactions.
For example, when one participant, while working on the C-DA program activities, repeatedly arrived at
Mediation Level 3 (as described in the above section ―C-DA Mediation Designs‖) before being able to
provide the correct answer, it indicated that he needed more elaborate mediation at this level than what
was provided originally in the C-DA program. In this case, the learner would likely need extra help from
the teacher on a one-on-one basis. When the teacher, who served as a mediator, worked with this
particular learner, she would observe closely what was lacking in the Level 3 mediation for this learner,
and then modify it by adding a more explicit, more powerful, and ultimately more understandable
intervention based on every interaction between herself and the learner. As a result, the information
generated from the C-DA program became a valuable resource for the teacher to create an effective one-
on-one mediated learning environment that meets individual learners‘ needs. Secondly, the C-DA
program is not merely an assessment tool but also an effective tool to teach the learners useful reading
strategies through the test. The participants were given an optimal amount of assistance based on their
individual needs when the mediation was presented in a gradual progression from implicit to explicit
design.
Furthermore, the participants benefitted greatly from the C-DA program where instruction and assessment
were unified into a single activity. Based on the four types of readers identified by Perkins (1992), it is
noteworthy that many of the participants in the present action research demonstrated rather high levels of
metacognitive knowledge in their reading strategies. The C-DA program helped the participants move
beyond merely answering the questions in the assessment and transcend into what Perkins called
―strategic‖ and ―reflective‖ readers. As shown in the evidence listed in Appendix B, the participants were
capable of using effective reading strategies and repairing meaning on their own when necessary. They
also demonstrated profound awareness of their reading behaviors. In addition, their effective
metacognitive strategies increased their meaning construction, their monitoring of text, and their ability to
evaluate their own performance. During the C-DA process, the participants became strategic and
reflective readers who are aware of knowledge, procedures, and controls of their reading processes.
The C-DA program has also met the researcher‘s teaching needs very well. First, it does not have to be
restricted to the classroom only. Instead, it can be used at home or outside the classroom. Since the
procedures can be posted on the Internet, burned onto CDs, or stored on portable flash drives, they can be
implemented anywhere and at the students‘ convenience as long as they can get access to a computer. The
C-DA program thus places the classroom teacher in the role of a guide or facilitator, who intervenes only
as necessary. Secondly, the C-DA program can also be simultaneously administered to large numbers of
learners; individuals may be re-assessed as frequently as needed; and reports of learners‘ performances
can be automatically generated. It allows the researcher to make good use of computer technology as a
mediator to help support her students‘ learning process in and outside the classroom more efficiently.
End End
After reading the instructions given, I learned that very often I read too much
into what I read. I could have answered the question correctly the first time.
The computerized activities helped me change my bad habits in reading. I
got caught up by trying to understand every sentence when I read. The
instruction guided me to focus on the main points. I spent too much time
reading sentences that are just examples or details.
REFERENCES
Baker, L., & Brown, A. L. (2002). Metacognitive skills and reading. In P.D. Pearson (Ed.), Handbook of
reading research (pp. 353–370). Mahwah, NJ: Lawrence Erlbaum Associates, Inc.
Block, C. C., & Mangieri, J. N. (2003). Exemplary literacy teachers: Promoting success for all children
in grades k-5. New York, NY: Guilford.
Chan, Y. C. (2006). Elementary school EFL teachers' beliefs and practices of multiple assessment
[Electronic version]. Retrieved November 1, 2010, from http://www.nus.edu.sg/celc/publications/
RELT71/37to62-chan.pdf
Crook, C. (1991). Computers in the zone of proximal development: Implications for evaluation.
Computers and Education 17(1), 81–91.
Dixon-Krauss, L. (1996). Vygotsky in the classroom: Mediated literacy instruction and assessment. New
York, NY: Longman.
Farris, P. J., Fuhler, C. J., & Walther, M. P. (2004). Teaching reading. A balanced approach for today’s
classrooms. New York, NY: McGraw-Hill.
Feuerstein, R., Feuerstein, R. S., & Falik, L. H. (2010). Beyond smarter: Mediated learning and the
brain’s capacity for change. New York, NY: Teachers College Press.
Feuerstein, R., Rand, Y., & Hoffman, M. B. (1979). The dynamic assessment of retarded performers: The
learning potential assessment device, theory, instruments, and techniques. Baltimore, MD: University
Park Press.
Harris, T.L., & Hodges, R.E. (1995). The literacy dictionary: The vocabulary of reading and writing.
Newark, DE: International Reading Association.
Harvey, S., & Goudvis, A. (2000). Strategies that work: Teaching comprehension to enhance
understanding. Portland, ME: Stenhouse Publishers.
Lantolf, J. P., & Poehner, M. T. (2004). Dynamic assessment: Bringing the past into the future. Journal of
Applied Linguistics, 1, 49–74.
Perkins, D. (1992). Smart Schools: Better thinking and learning for every child. New York, NY: Free
Press.
Pressley, M., & Afflerback, P. (1995). Verbal protocols of reading: The nature of constructively
responsive reading. Hillsdale, NJ: Erlbaum.
Sternberg, R. J., & Grigorenko, E. L. (2002). Dynamic testing. The nature and measurement of learning
potential. Cambridge, UK: Cambridge University Press.
van der Veer, R., & Valsiner, J. (1991). Understanding Vygotsky: A quest for synthesis. Oxford, UK:
Blackwell.
Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes. Cambridge,
MA: Harvard University Press.
wealthy enough to be using Wi-Fi enabled, touch screen, application integrated mobile phone technology.
Indeed, Stockwell and Sotillo (2011) identify 2009—the year in which Stockwell’s (2010) study ceased
collecting data—as the beginning of a new MALL research area, apps for language learning. Thus, given
the pace at which the technologies concerned are developing, the issue of technology’s pace of change
would appear to be an issue which longitudinal research designs cannot avoid. Which is to say, this is not
a criticism of Stockwell’s 2010 research.
Looking at Stockwell’s (2010) research design from another angle, however, an ungenerous observer
might conclude that he was guilty to some extent of the ―conservatism and unimaginativeness‖ Fox
complained of back in 1991. Describing the vocabulary activity system he used to collect the data he says:
―Each lesson included 13–17 vocabulary items which were selected from the commercially produced
textbook‖ (p. 100). As mentioned above, Stockwell was no doubt constrained by the technology
commonly available at the beginning of the study, and probably also by the context his study was
conducted in: the need to integrate the MALL activity with the participants’ expectations and course of
study. However, his description of the learning activity students engaged with makes it sound as if it
would have been just as suitable for pen and paper completion as CALL or MALL. This may have been
because it was a format that he felt could be realized through a web browser interface, and his decision to
employ a browser-based interface and stick with it across all three years of his study would appear to be
for sound methodological reasons: to maintain comparability between the three cohorts he collected data
from. However, though he does indicate that he found it necessary to simplify some facets of the mobile
interface to accommodate the technological limitations of the platform, it appears the approach adopted
was to realize the activity on both computer and mobile platforms as similarly as possible, presumably to
increase comparability. However, to some extent, this would seem to somewhat invalidate his findings
vis-à-vis his declared aim: to investigate the effects of the platform.
Stockwell (2010) appears to assume that decisions about the design of the learning software can be
separated from the platform they will be delivered on, but approaching the issue like this could be seen as
predetermining the outcome. For example, Web pages are convenient to read on computers, but their
layout and design would make little sense if transposed to the traditional print medium. Thus, the study’s
conclusions seem to apply most properly to the effects of completing pen and paper type vocabulary
activities on mobile phones and computers. However, it is not clear that this says much about the potential
of the MALL platform to engage learners in vocabulary activities. Provide a learner with a pen and paper
type activity, and ask them if they would like to complete it with pen and paper, through CALL or
through MALL, and they may well choose CALL, or even perhaps pen and paper, but this seems to be, to
some extent, missing the point of MALL.
Given recent technological advances, MALL activities should be app based, exploiting touch screen
technology and designed to be used in the situations MALL has opened up (Godwin-Jones, 2011).
Regarding this last point, Stockwell (2010) himself points out that a possible confound in the study is the
effect of environment on MALL use. He hypothesizes that some of his participants may have utilized the
mobile phone activities in preference to computer activities to take advantage of ―dead time,‖ for
example, when commuting. He suggests that such an environment may have had an effect on users, an
effect which would also be present for computer users if their computers were in such an environment. He
also quite correctly identifies what is probably the core attraction of MALL: the potential to turn such
―dead time‖ into useful study time. However, his research design investigates using this time for learning
activities which were not designed for this kind of environment or this kind of platform. There is of
course some value in exploring the effects of completing pen and paper type activities on your phone,
possibly even on a crowded train, but it is also possible to imagine other types of MALL-enabled
activities designed explicitly to be conducive to study in just these kinds of unconventional study
situations. Stockwell’s research is enlightening in many respects, but in some ways, the empirical findings
report on the least justifiable exploitation of MALL’s potential. It is not just the rapid pace of
technological development which limits the utility of Stockwell’s findings; they are also limited in terms
of their scope. Stockwell’s study answers questions regarding the potential of MALL to compete with
CALL as a substitute for a textbook well, but a more important question remains: does MALL have the
potential to supplement our existing language learning resources with something else, something designed
to exploit the ―dead spaces‖ which MALL has brought to life and, if so, what effect would the platform
have then?
REFERENCES
Beatty, K. (2003). Teaching and researching Computer-Assisted Language Learning. Harlow: Pearson
Education Limited.
Godwin-Jones, R. (2008). Emerging Technologies Mobile-Computing Trends: Lighter, Faster, Smarter.
Language Learning & Technology, 12(3), 3–9. Retrieved from http://llt.msu.edu/vol12num3/
emerging.pdf
Godwin-Jones, R. (2011). Mobile apps for language learning. Language Learning & Technology, 15(2),
2–11. Retrieved from http://llt.msu.edu/issues/june2011/emerging.pdf
Martinez, R., & Schmitt, N. (2010). Invited Commentary: Vocabulary. Language Learning &
Technology, 14(2), 26–29. Retrieved from http://llt.msu.edu/vol14num2/commentary.pdf
Stockwell, G. (2010). Using Mobile Phones for Vocabulary Activities: Examining the Effect of the
Platform. Language Learning & Technology, 14(2), 95–110. Retrieved from http://llt.msu.edu/
vol14num2/stockwell.pdf
Stockwell, G. & Sotillo, S. (2011). Call for Papers for Special Issue of LLT. Theme: Mobile Language
Learning. Language Learning & Technology, 15(3), 130. Retrieved from http://llt.msu.edu/issues/
october2011/call.pdf
INTRODUCTION
While the enthusiasm that Ballance shows in his commentary (see this issue) for mobile learning and the
possibilities that new developments in technology of mobile devices may bring about is commendable,
there are several areas of concern in his commentary which deserve mention. My first concern is that the
commentary misrepresents the vocabulary learning system described in the Stockwell (2010) article in
many respects through, among other things, describing the learning activities inaccurately (it was an
intelligent adaptive system), and suggesting that the same format was adopted for both the PC and mobile
versions of the system (which was not the case). Secondly, Ballance seems to be stating that since
problems that existed at that time were going to be solved by smart phones anyway, many of the findings
were already obsolete at the time of publication. Not only does this seem to place an exaggerated
expectation on the effects that smart phones may have, but it also fails to consider the ways in which
learners engaged in the activities and their views towards mobile learning in general. These issues will be
dealt with in more detail later, but in short, this paper will argue that (a) Ballance has failed to
comprehend the activities described and the learning environment; and (b) Ballance has unrealistic
expectations of the effect that new technologies have on learner attitudes towards using them for learning
purposes.
BALLANCE’S CLAIMS
To summarise, Ballance primarily appears to be arguing:
1. that the vocabulary activities in Stockwell’s (2010) study could have been replicated with pen and
paper and were not suitable for mobile phones; and
2. that the findings in Stockwell’s (2010) study are obsolete, as smart phones will solve the majority
of problems that were apparent in pre-smart phones and improve learner engagement in activities.
As I deal with each of these points, I would also like to point out some fundamental issues regarding
mobile learning which are relevant to my responses.
groups using an accompanying text book. Given the relatively low level of the learners, many of the
learners experienced difficulties with the vocabulary that came up in the videos, so key vocabulary items
were selected so that learners could learn them in order to better understand the videos and participate
more actively in class discussions about the content of the videos. Since class time was limited, I felt that
it would be beneficial if learners could have a means through which they could learn these difficult
vocabulary items in their own time, while at the same time to keep records where I could get an indication
of how much difficulty learners were actually having with learning the vocabulary items.
The activities were developed as part of an intelligent system that kept detailed records of what learners
knew based on their performance in the vocabulary activities, providing learners with items that the
learners had difficulty with far greater frequency until the system judged that the learners had acquired
them. This score was calculated by assigning a score depending on how many times the learner was able
to get the questions about each item correct (see Stockwell, 2007, for a description). Questions in the
system ranged from recognition-type questions in the early stages of each unit through to production—
specifically, the writing of a single word—after learners had been exposed to each item at least twice. It
should be pointed out that the learners had also been exposed to the items in context in various listening
activities covered in the class. After extensive pre-testing of the system before its implementation in early
2007, it was confirmed that the mobile version of the site showed clearly on a range of mobile phones
provided by various telecommunications companies at the time, and that selecting items in the multiple
choice and inputting of the text was not overly difficult, even on the small keypads.
In addition, the system also included push and pull modes, the push mode being where information was
sent to the learners’ mobile devices, and the pull mode being where students needed to access the site
themselves (see Mellow, 2005; Motiwalla, 2007). The push mode in this case was e-mails that were sent
to learners’ mobile phones with lists of the vocabulary items that they had experienced difficulties with
during the activities, and, if a learner had not accessed the system for three days when a unit was still
incomplete, a reminder was sent out to let them know that they should access the activities. The reason for
the use of mobile e-mail was that at the time, Short Message Service (SMS) was not commonplace, and
that each mobile phone has its own unique email address that could receive messages sent from either
mobile phones or from PCs. The messages also included a link to the site so that learners could
immediately access it from their devices without the need to retype the URL. Students were also provided
with a QR code which they could scan with their mobile phones to bring up the URL in their mobile
phone browsers without typing.
Given the complexity of the system in its adaptability to learners’ input and its integration of both push
and pull modes, I fail to see how this could be compared to simple pen and paper activities as Ballance
suggests. In saying this, however, his statement prompted me to consider the issue of whether replicating
pen and paper activities on mobile devices is necessarily the bane of mobile learning that Ballance
appears to brand it as. We already replicate many things that were traditionally done with pen and paper
on mobile devices. Many of us use our mobile devices as calendars, train schedules, and, even more so
with the spread of tablet computers, books. While of course there are some added benefits in the
functionality of using mobile devices for these purposes, one might argue that one of the primary reasons
why we opt to use a mobile device for these is that we have everything in one place on a device that we
typically carry with us anyway. Although the activities described in the Stockwell studies would be
difficult to replicate with pen and paper, there have been other examples of mobile learning could be
thought to also largely be a replication of pen and paper activities, such as flashcards (Lin, Kajita, &
Mase, 2008) or mini lessons (Thornton & Houser, 2005), which were well received by learners.
Obviously, there may be some advantage in offering these in conjunction with more interactive types of
activities, but they appeared to serve their purpose as accessible tools in learners’ hands, meaning that
there was a greater chance of exposure to the target language than if these tools were not available to
learners. Different learners have different learning styles, and providing a range of tools—some of which
are more interactive, and some of which are less interactive—means that there is greater chance of being
able to meet the needs and preferences of a wider range of learners.
Another indication of Ballance’s lack of understanding of the learning activities is his statement that
―Stockwell (2010) appears to assume that decisions about the design of the learning software can be
separated from the platform they will be delivered on‖ (p. 22), which is based on his misled assumption
that the mobile and PC activities were the same. A look through the details of the earlier articles to which
the reader is directed will show that while the content was the same, there was quite a marked difference
between the activities for the mobile platform and the PC platform, and that the design for each was quite
carefully planned to tailor it to the platform it was to be used on. For clarity’s sake, however, I will
reiterate this here.
In very broad terms, the general layout of the screen on the PC and the mobile were relatively similar in
terms of where things could be found and the colour schemes. A commonality was that the activities
could be left off at any time during a specific unit, and the system would record what the learners had
done, and they could seamlessly pick up from where they left off from either the PC or the mobile phone.
Each individual activity in the units was designed to be able to be completed in a span of 20–60 seconds
(depending on the type of activity) in an attempt to allow learners to use even very small snippets of time
that they found available. At the same time, learners could, should they so desire, focus their efforts and
complete an entire unit, which generally took around 10–30 minutes, depending on the number of items
that they experienced difficulties with and the platform that they used.
The activities for each platform were accessible from completely different URLs, and the PC version was
accessed through the Moodle site that was used for the subject, whereas the mobile site was accessed
from a separate URL that could be used without needing to go through Moodle. Students were
recommended to add the URL to their favourites list on their phones, but they could, as described earlier,
use a QR code with the barcode reader function of their mobile phones to access it. The page size, when
seen in a mobile web browser, was fit to the screen, a very common practice for mobile sites in Japan, and
the number of items per page could be controlled by the learners, from a smaller to a larger number of
items as desired. Learners could have different numbers of items presented on both the PC and mobile
platforms, but contrary to my expectations, most learners opted for the middle figure of eight items on
both platforms. Demonstrations of how to change the number of items were given in the first class, and a
help menu was provided in both the PC and mobile versions. Graphics were removed from the mobile
version given the potential for extra downloading costs and to make the most of the screen without
unnecessary cluttering. Thus, the layout and the design screens of the PC and mobile were not the same,
and a good deal of time and effort was expended to ensure that the designs were suitably adapted to each
to make them as easy as possible to use. The nature of the activities themselves was very similar, but the
reason for this is that these activities were deemed to be the most effective in learning the necessary
vocabulary given learners’ previous experiences with language learning and from discussions with
learners during the development of the system. In the same way, the screens were specifically tailored to
suit the technology of the time, and this design remained relevant across the three years of the study.
The conclusion that Ballance makes about the learning activities is that they were the cause of learners
not exploiting ―dead time‖ because the activities ―were not designed for this kind of environment or this
kind of platform‖ (p. 22). As the description above shows, the activities were most certainly designed to
suit both PC and mobile platforms respectively, and took advantage of the affordances of each to
maximise the learning experience for learners. Apart from the factual inaccuracy of the statement, his
statement also fails to account for the large number of learners (58.8% in 2007, 78.0% in 2008, and
42.2% in 2009) who did not access the mobile activities at all, and thus had no grounds to make
judgments as to whether the activities were suited to their mobile phones or not. Nor does his statement
take into consideration the number of learners who were aware that it took longer to complete the
activities on their mobile phones, yet made the decision to use the mobile phone until the end, despite the
fact that they could use the PC if they so desired. It is obvious that there are other factors at play that go
beyond the activities, and as was reflected by comments from learners in an earlier study, there were
learners who expressed that they felt that learning on a mobile phone ―didn’t feel like studying‖ and that a
mobile phone is ―not a tool for studying‖ (Stockwell, 2008, p. 260). These were learners who had decided
from the outset that they did not want to learn using mobile phones as they were already comfortable with
learning with the PC, largely due to the lack of external distractions, and to blame this lack of use on the
activities themselves seems somewhat naïve when it is obvious that there were other factors at play that
influenced the outcome (which was the objective of the study in the first place).
In conclusion of this section, I will not make the claim that there were not alternatives that could have
been used instead of the types of activities that were used with the learners in this study. However, those
that were selected were done so with an eye on the complete learning environment in which they were to
be used, the pedagogical philosophy held by the teacher-researcher, and the voices of other learners who
had taken part in the course before, and in that light, I would argue that they were appropriate for both the
PC and mobile platforms.
while in transit? A look at some of the basic differences between smart phones and pre-smart phones may
take us some way towards a response to Ballance’s claim that the results are obsolete due to the
technology used. Physically, there are some obvious differences. Being based in Japan, I’m afraid that I
am only able to explain pre-smart phones that are available in Japan, but considering that these were
devices that were used in the study, perhaps this is the most appropriate. A typical pre-smart phone in
Japan has a 3 inch screen, compared to that of the iPhone 4S, which is 3.5 inches, and of the Samsung
Galaxy S III, which is an impressive 4.8 inches. Japanese pre-smart phones are typically flip phones or
slide phones, which means that there is, in addition to the screen, a keypad and centre ―joystick‖ which
allows for moving through menus and webpages. Smart phones, on the other hand, use touch screens for
manipulating icons and virtual keyboards for inputting text, the latter taking up as much as one-third to
one-half of the screen when used. Looking at plain figures themselves would seem to suggest that smart
phones have an advantage in screen size compared to pre-smart phones, but that this advantage is
somewhat reduced when the virtual keyboard is in use. Manipulation of icons on the screen is also very
convenient with a smart phone given the touch screen, and use of the virtual keyboard is for the most part
relatively easy to input text. Scrolling is done with the joystick on pre-smart phones and through running
one’s finger along the screen on smart phones.
While these differences are certainly important ones, are they relevant in the scope of the limited amount
of information required to be put on a screen and the minimal input that would be required in activities
that could be completed in short spaces of time? One might argue that the differences are perhaps not as
large as we may imagine. Certainly, it is possible to check your email and watch YouTube videos at the
same time as accessing learning activities, which cannot be done on pre-smart phones, but is it not
possible that these could be extraneous distractions that eventually prevent learners from engaging in
activities when using them? In saying this, however, it is obvious that at some point in the not-too-distant
future, we are likely to see essentially all mobile phones carried by our learners being ―smart‖ phones, or
whatever the next generation of phones beyond those that are currently available is going to be, and
discussions of differences of what came beforehand will indeed be irrelevant. My point is, however, that
screen sizes of mobile phones are still small and inputting is still difficult compared to, for example, a
tablet computer or PC, thus the discussions of screen size and input difficulties of mobile phones for
language learning purposes are likely to continue, at least as long as mobile phones remain a comparable
size to those used today.
Another feature of smart phones that Ballance suggests is mobile apps. Based on Godwin-Jones (2011),
he argues that ―MALL activities should be app based, exploiting touch screen technology and designed to
be used in the situations MALL has opened up‖ (Ballance, 2012, p. 22). While Ballance himself fails to
give any further details of what kind of apps would be appropriate, we might consider three different
possibilities of apps that learners might use. The first of these is commercial apps, which learners can
download for somewhere ranging between two and 50 U.S. dollars, depending on what is involved in the
app. While this is a promising potential market, there have been learners that have expressed
disappointment and distrust regarding the available commercial apps for language learning (see Stockwell
& Stockwell, 2012). The majority of learners in this study indicated an unwillingness to pay for apps for
learning purposes (which might be compared to learners in earlier studies who expressed an unwillingness
to pay for Internet connectivity for learning), and a number of learners who did download language
learning apps found that the content was unhelpful or even trivial. The second type of app is one that is
created by teachers themselves. This is an issue that has been an ongoing one with regard to developing
any materials that utilise technology. Teachers have long lamented the lack of skills or funding required
to create good learning materials (e.g., see Pelgrum, 2001), and to this end it is not surprising to see the
same problem arising with regard to materials for mobile devices as well. When CALL itself was still
emerging on an international scale, the majority of the discussions centred around the selection of
Windows vs. Macintosh, and advocates on each side stood firmly by their technology of choice. For
mobile devices, we see the game opened up even more widely, where native mobile apps written for,
among others, iOS, Android, Blackberry, WebOS, and Windows Phone 7 are mutually incompatible (see
Godwin-Jones, 2011, for a discussion). With most teachers struggling to be able to gain sufficient skills to
create apps for one of these operating systems, the idea of having to make separate apps for each of the
available operating systems is, needless to say, far from practical. The last main type of app is a Web app,
which makes it possible to create apps which can be used across different platforms. In essence, Web
apps are apps that operate using web browsers. They may be downloaded onto the device itself or
alternatively they may operate from a remote server. In one sense, these are perhaps the only practical
alternative to teachers who want to develop their own apps that can be used by learners who possess a
range of different technologies, but they are of course limited in many ways to the functionality of web
browsers.
As this brief discussion of apps alludes to, apps may have the potential to solve some of the problems of
technologies that preceded them, but at the same time, they also create new problems of their own,
particularly with regard to compatibility issues. More importantly, just because something has been
written as an app does not mean that it is pedagogically sounder than something that has been written
using other technologies such as web browsers. In this way, Ballance’s claim that apps (and indeed smart
phones themselves) will solve the problems of earlier technologies fails to take into complexities of
task/activity design, and in one sense walks down the same path of those early pioneers of mobile
learning who approached mobile technologies as providing a way for learners to engage in tasks outside
of the classroom simply because the technologies existed. Very relevant to this discussion is the
observation of Bax (2003) regarding two fallacies that are often seen with regard to the use of technology
in language learning: (a) the belief that a new technology can do more than it actually can; and (b) the
belief that the existence of new technologies is the only relevant factor leading to its success. While I
don’t believe that Ballance would claim that the simple existence of smart phones can lead to successful
implementation of mobile learning, his commentary most certainly has undertones of the first fallacy,
exhibiting an inflated expectation that new technologies will solve the problems of earlier ones because
they have enhanced functionality. Regardless of what technology we used, we are likely to see
preferences that some learners will opt to use it in the manner and regularity that we had predicted, while
others will either use it in ways that are very different from what we had expected, or alternatively, fail to
use it at all (see Fischer, 2007, for a discussion of learner usage patterns). In some cases, this may well be
caused by the type of activities used, where we see some percentage of the learners who engage in
activities very successfully and others who do not—particularly in an environment where learners are
working on their own devices completely free of teacher intervention. Alternatively, the variations in the
outcomes may simply be the inevitable result of individual learner differences and preferences, that
teachers may have little choice but to accept.
CONCLUSION
I would like to close this commentary with again indicating that I am happy to see enthusiasm for the
potential of mobile learning such as that expressed by Ballance, and to that end, discussion of this type is
not only welcome but necessary. It is a shame, then, that he failed to grasp the essence of the activities
that were used across the three years, and indeed, to some degree, even the nature of the 2010 study itself.
The study was not intended as a comparison of PCs and mobile phones for learning vocabulary, but rather
as a means of trying to shed some light on to why learners opted to use PCs over mobile phones for
language learning activities, despite the fact that the activities on the mobile platform were designed for
very short periods of time to be used effectively while on the move. Ballance appears to have jumped to
the conclusion that it had to have been the activities or the technology which resulted in the learners
opting not to use the mobile phones, with insufficient understanding of what was used in the study. Such
an assumption would presumably be based on his personal attitude towards the potential of mobile
technologies, and perhaps he himself would most certainly have capitalised on the opportunities to exploit
the ―dead spaces,‖ provided he had access to a device that included apps suited to his own approach to
learning a language, but care should be taken to avoid excessive generalisations based on personal
assumptions. It is also unfortunate that if Ballance held such strong views about how learning
opportunities through mobile devices could have been better created and capitalised upon, that he failed to
provide any concrete suggestions beyond using smart phones and apps, or more importantly, any
empirical evidence to support his suggestions rather than relying solely on speculation. Should he
undertake such a study, however, I myself would most look forward to seeing the results to see how I
could apply them to my own learning environment.
As teachers, we are always faced with constraints that we must overcome. In some cases this is the range
of technologies that are available to us, and in other cases it is our own skills and time to create materials
that we believe would be beneficial for our learners. With technological advancements, there is increased
pressure on teachers to provide professional tools that learners feel are useful, and this situation can make
it easy for us to take our focus off what it available in the current environment towards what will become
available in the future. Care must be taken to ensure that pedagogy takes a primary role in the design of
activities, and that the devices, mobile or otherwise, are selected as a natural outcome of an analysis of the
learning goals and the pedagogical approach (see Colpaert, 2006), and not just because they are new
and/or popular in non-educational settings. Emerging mobile technologies certainly do have the potential
to enrich our learners’ learning experiences, but it is essential to avoid sound sacrificing pedagogy for the
sake of using technologies (Felix, 2003). Mobile learning will continue to take on new shapes and forms
as it becomes more familiar to both teachers and learners. The way in which it develops, however, will
continue to be constrained by practical factors that include but extend beyond the availability of
technologies.
Teachers who have been in the field for a long time have seen, and continue to see, a massive shift in the
tools that are used and expectations of them held by learners. Learners themselves, however, are generally
in the system for such a short time that the snapshot view of education that they have is the only model of
learning that they know, and they have little interest in what happened beforehand and really of what will
happen after they finish. Our goal as teachers is to gain an understanding as quickly as possible of the
expectations of our learners, and to move with the times, adapting to the environment of now while
keeping a keen eye on the environment of the future—including developments beyond smart phones—for
our upcoming students of years to come.
REFERENCES
Bax, S. (2003). CALL—past, present and future. System, 31, 13–28.
Colpaert, J. (2006) Pedagogy-driven design for online language teaching and learning. CALICO Journal,
23(3): 477–497.
Colpaert, J. (2010). Elicitation of language learners’ personal goals as design concepts. Innovation in
Language Learning and Teaching, 4(3), 259–274.
Felix, U. (2003). Pedagogy on the line: Identifying and closing the missing links. In U. Felix (Ed.),
Language learning online: Towards best practice (147–170). Lisse: Swets & Zeitlinger.
Fischer, R. (2007). How do we know what students are actually doing? Monitoring students’ behavior in
CALL. Computer Assisted Language Learning, 20(5), 409–442.
Godwin-Jones, R. (2011). Mobile apps for language learning. Language Learning & Technology, 15(2),
2–11. Retrieved from http://llt.msu.edu/issues/june2011/emerging.pdf
Lin, N., Kajita, S., & Mase, K. (2008). Mobile user behavior and attitudes during story-based Kanji
learning. The JALT CALL Journal, 4(1), 3–18.
Mellow, P. (2005). The media generation: Maximize learning by getting mobile. In ASCILITE 2005
Conference: Balance, fidelity, mobility: Maintaining the momentum? Conference Proceedings. Retrieved
from http://www.ascilite.org.au/conferences/brisbane05/blogs/proceedings/53_Mellow.pdf
Motiwalla, L. F. (2007). Mobile learning: A framework and evaluation. Computers & Education, 49,
581–596.
Pelgrum, W. J. (2001). Obstacles to the integration of ICT in education: Results from a worldwide
educational assessment. Computers & Education, 37, 163–178.
Stockwell, G. (2007). Vocabulary on the move: Investigating an intelligent mobile phone-based
vocabulary tutor. Computer Assisted Language Learning, 20(4), 365–383.
Stockwell, G. (2008). Investigating learner preparedness for and usage patterns of mobile learning.
ReCALL,20(3), 253–270.
Stockwell, G. (2010). Using mobile phones for vocabulary activities: Examining the effect of the
platform. Language Learning & Technology, 14(2), 95–110. Retrieved from http://llt.msu.edu/
vol14num2/stockwell.pdf
Stockwell, G., & Stockwell, E.S. (2012, August). Do smart phones affect learners’ preferences for
language learning? Paper presented at the EUROCALL 2012 Conference, Gothenburg, Sweden.
Thornton, P., & Houser, C. (2005). Using mobile phones in English education in Japan. Journal of
Computer Assisted Learning, 21, 217–228.
The University of Hawai„i National Foreign Language Resource Center engages in research and materials
development projects and conducts workshops and conferences for language professionals among its
many activities.
Did you know that the NFLRC has its own Facebook page with over 1,000 fans? It‟s one of the best ways
to hear about the latest news, publications, conferences, workshops, and resources we offer. Just click on
the “Like” button to become a fan. For those who prefer getting up-to-the-minute “tweets,” you can
follow us on our Twitter page. Finally, NFLRC has its own YouTube channel with a growing collection
of free language learning and teaching videos for your perusal. Subscribe today!
Using group dynamics as an umbrella term for overlapping literatures on community, cooperative, and
collaborative practices, I introduce the concept of PCOIz which might help teachers to better
conceptualize their classes and the time-frame influences on their students. I position PCOIz as
complementary to but different from communities of practice and imagined communities, and in some
ways overlapping, with an emphasis on imagining and re-imagining. We look at how teachers can
systematically organize activities to look at students‟ pasts, presents, and futures. I also present several
mixed method studies done in Japan which indicate that PCOIz, when well developed, can nurture the
aspirations, resilience, learning strategies, beliefs, motivations, and possible selves of its members
through critical dialogue and collaboration. View/download free.
Save the trees! Check out our other two online journals:
CONFERENCES
CLEAR exhibits at local and national conferences year-round. We hope to see you at ACTFL, CALICO,
MIWLA, Central States, and other conferences.
We will be presenting a session at ACTFL 2012. We hope to meet you in person!
PROFESSIONAL DEVELOPMENT
Our webinars continue to be a popular professional development option for educators who are not able to
travel to MSU for our workshops. These low-cost, online, interactive workshops allow language teachers
from around the country (and the world) to access CLEAR‟s professional development on demand. The
currently available webinars focus on CLEAR‟s Rich Internet Applications and language assessment. We
hope to add to the collection in the coming months. Webinar participants are able to apply for State Board
Continuing Education Units. Detailed information on all webinars is posted on the webinars page of our
Web site.
MATERIALS DEVELOPMENT
CLEAR is developing several new products during our fifth funding cycle. Check our Web site for
updates on new products and services. Some of our upcoming projects include:
Professional development webinars on diverse topics (see above)
Online videos for language teaching techniques
Online listening and speaking tests for LCTLs
NEWSLETTER
CLEAR News is a free bi-yearly publication covering FL teaching techniques, research, and materials.
Download PDFs of back issues and subscribe at http://clear.msu.edu/clear/newsletter/.
2011
ISBN: 9780415889964
US $ 39.95 (paperback)
216 pp.
Routledge
New York, NY
heretofore entrenched oppositions and move the conversation in productive new directions.
Professor Buckingham‘s foreword is, honestly, the review I would like to have written myself. His
concise remarks cut to the heart of the issues and problems at hand, rightly asserting for instance that
―technologies have possibilities and limitations (or ‗affordances‘), but they do not produce social change
in themselves‖ (p. x). Yet, he also admits the potential utility of thinking in generational terms about
technological and social change, setting an appropriately even tone for the succeeding chapters. In
Chapter One, Thomas examines the discourse of techno-evangelism, its implications—including the
contagious appeal of the notions of the digital native and net generation and consequent calls for new
pedagogical priorities—and the positions of notable ―dissenters‖ (p. 1). In so doing, he lays further
necessary groundwork for the chapters that follow, which, Thomas explains, set out ―a number of
research directions for future studies shown in the concerns of the contributors‖ (p. 7).
Part One, Reflecting on the Myth, begins with Chapter Two, a version of a recent article by Marc Prensky
himself, in which he addresses what, to his mind, have been misconceptions and exaggerations on the
parts of his critics. Prensky then reframes his original formulation of the digital native/digital immigrant
distinction so as to highlight the necessity of cultivating ―digital wisdom‖ (p. 20), the profit of enhancing
natural human intellectual capacities through digital technology. Concluding Part One is Chapter Three,
Jones‘s critical analysis of several conceptual underpinnings of the net generation and the digital
native/digital immigrant divide (e.g., technological and generational determinism). Jones regards
Prensky‘s updated formulation, focused on digital wisdom, as persistently deterministic: ―even though it
moves beyond a straightforward divide between immigrants and natives, the argument retains a simple
moral imperative: digital enhancement has to be accepted in order to succeed‖ (p. 38). Jones suggests that
the changing ways that young people adopt and use technology are certainly worthy of attention from an
educational perspective, but also asserts that assumptions that digital technologies might ―determine the
outlook of an entire generation‖ (p. 43) are wrongheaded and unhelpful.
Part Two (Perspectives) consists of Chapters Four through Ten, variously contributing empirical support
to the process of deconstruction. Banaji‘s paper is centered on the European CivicWeb project, ―the
largest existing cross-national study of young people, the Internet, and civic participation‖ (p. 52).
Findings from this project suggest that any general presumption of young people being digitally
networked, civically and politically active, and so also primed to develop digital wisdom, is misguided,
especially in that it ―sidesteps issues of power, ideology, and privilege‖ (p. 62). In Chapter Five,
Takahashi looks at mobile-internet use among youth in the Tokyo area and demonstrates in vivid
ethnographic terms the inherent diversity of Japanese young people and the roles digital technologies play
in the dynamic performance of varied individual and cultural identities, to both reproductive and
innovative effects. Chapters Six (M. Levy and Michael) and Seven (Erstad), respectively referring to
work in Australian and Nordic contexts, further develop what emerge as key themes in the volume: that
young people‘s relationships to and uses of new media are both complex and diverse and that
automatically competent—or fluent to apply Prensky‘s metaphor—use and production of digital texts
cannot be assumed in young people, nor are such capacities well understood as yet. Kennedy and Judd, in
Chapter Eight, review research on tertiary students‘ use of Google for information gathering, concluding
that criticality and digital wisdom are often subordinated to expediency, ―somewhat counter-intuitively‖
(p. 132), the authors point out—referring to the expectation of masterful, conscious use of technologies
that the digital native concept would evoke—putting ―more not less onus on faculty to support ‗Digital
Native‘ students in the development of their information literacy skills‖ (ibid.). In Chapter Nine,
recruiting Jenkin‘s participatory culture framework, Zimic and Dalin analyze the types and levels of
participation of youth and younger adults on the Internet in relation to the participants‘ perceptions about
their participation. The authors find that individual sense of involvement in the information society varied
in degree according both to type of internet-enabled activity and age, suggesting that frequent use of the
Internet does not imply a feeling of positive participation in the broader digital culture. With Chapter Ten,
Rachael Levy brings an interesting close to Part Two inasmuch as her case studies of three primary-age
school children both acknowledge the digital heterogeneity that other contributors have highlighted, and
illustrate the homogenization of digital expression that may occur within educational institutions,
frustrating the development of digital wisdom, which resonates with core arguments of Prensky, Tapscott,
and likeminded others.
Part Three (Beyond Digital Natives), paralleling Part One, comprises two chapters. Bennett and Maton are
the authors of Chapter Twelve, in which they reassert and elaborate their earlier arguments (see Bennett,
Maton, & Kervin, 2008) as to the lack of evidentiary support for Prensky‘s claims. Quite helpfully, I
believe, these authors emphasize in this chapter the importance of ―sound and transparent research‖ (p.
181) around these contentious issues and questions and recommend a shift in focus away from
generational differences and effects of technologies, and toward ―studies of young people‘s experiences
with technology‖ and ―how these are integrated into the array of contexts and practices in daily life‖ (p.
180). In the concluding piece, Chapter Twelve, Palfrey and Gasser ask whether the term digital natives
can, in fact, be productively reclaimed for use in moving the conversation forward. While supporting the
critical perspectives in the preceding chapters (e.g., by roundly dismissing any technological and
generational determinism), these authors also raise important questions about the dimensions of safety,
privacy, and responsibility in communicating and creating with new media. ―Most important,‖ Palfrey
and Gasser recommend to concerned readers, ―is that we share a common commitment to understanding
of what is going on with new media practices and, in turn, that we work together to seize the opportunities
and mitigate the challenges associated with media practices of youth and adults alike‖ (p. 201). The
simplest advice is often the best, I believe.
Still, one might ask if we couldn‘t have just begun with such a commonsense prescription. Why such ado
over the digital native, which is, after all, only a metaphor? And doesn‘t the illustrative force of any
metaphor naturally depend on its simplified comparative representation of only selected qualities of
things and aspects of experience, as Prensky himself is quick to point out in Chapter Two?
Reductive though such tropes may necessarily be, they may also be anything but simple, or innocuous, as
scholars of metaphor George Lakoff, Mark Johnson, Mark Turner, and colleagues remind us. Lakoff and
Johnson (1980) cogently demonstrate the fundamental organizing function of metaphor in human
cognition, communication, and action, explaining in particular how so-called structural metaphors admit
other extended meanings, or entailments, that give shape to our thoughts, language, and behavior (pp. 61–
68). The common metaphorical mapping of characteristics and functions of machines onto those of the
human mind, for instance, has profoundly influenced educational theories and approaches, compelling
teachers and students to think, speak, and act in terms of input and output, online processing, and the like.
On the specific entailments of the digital immigrants metaphor, media scholar Henry Jenkins (2007)
colorfully remarks,
[It] tends to exaggerate the gaps between adults, seen as fumbling and hopelessly out of touch,
and youth, seen as masterful. It invites us to see contemporary youth as feral, cut off from all
adult influences, inhabiting a world where adults sound like the parents in the
old Peanuts cartoons—whah, whah, whah, whah—rather than having anything meaningful to say
to their offspring. In the process, it disempowers adults, encouraging them to feel helpless, and
thus justifying their decision not to know and not to care what happens to young people as they
move into the online world (n.p., italics in original).
Jenkins‘ comments implicitly support Lakoff and Johnson‘s conclusions, that such metaphors are not
mere illustrations or analogies. In concrete social realities, metaphors do matter. Notwithstanding
Prensky‘s intent to usefully proffer only an intuitively persuasive ―broad generalization‖ (Chapter 2, p.
16) and his admitted confusion over why ―so many supposedly well educated, thinking people ‗just can‘t
take a metaphor‘‖ (p. 15), I hope that he and other so-called techno-evangelists come to see ardent
skepticism and careful examination of that metaphor, which Deconstructing Digital Natives provides and
invites, as healthy and necessary.
My criticisms of this volume are few, but worth noting here. First, as persuasive as the included
perspectives are, I think that a somewhat more balanced representation of voices might make the
deconstructing seem a bit less like debunking, which I do not imagine was the intent. At a recent
conference, I ran into Sue Bennett, a contributor here and co-author of one of the most widely cited
critiques of Prensky‘s work (Bennett, Maton, & Kervin, 2008). When I expressed my admiration for that
2008 paper, she explained how difficult it was to get the article published, having received a number of
reviews that assessed their arguments as simply ―wrong‖ (personal communication). The problems
associated with the digital native/digital immigrant formulation are quite evident, and the critics are
legion; but there are other scholars who would make more supportive arguments, clearly, and they might
also be heard.
Two other criticisms I would are offer relate to the format of the volume itself. At certain points I felt that
the absence of any images or figures was conspicuous, if not problematic. For instance, in Chapter Five,
which reports on students‘ multimodal composing practices, the analysis would be much clearer and more
cogent if the students‘ works were visible. Another issue is with the index, comprising only a sparse and
rather random collection of terms and names, which I found particularly odd given the relatively narrow
ranges of topics discussed and works referenced. In a volume such as this that aims to circumscribe the
array of considerations, concepts, and contributors associated with a particular complex of issues, a more
complete, better organized index would be helpful to the reader in making connections among the various
positions within the conversation.
Overall, however, Deconstructing Digital Natives does fulfill the promise of its title. It helpfully breaks
down the assumptions, metaphors, mythology, and ―moral panic‖ (Bennett, Maton, & Kervin, 2008, pp.
782–3) according to which relations between youth, digital technologies and education are commonly
understood and renders sensible and comprehensible the particular value and shortcomings of the digital
natives/digital immigrants distinction and the social and pedagogical exigencies that this and similar
explanatory frameworks would convey. This book should be of significant interest and use to scholars,
university students, teachers, and policy makers alike, in any area of education or communication. It
would be most enlightening, I imagine, to all those who regard the metaphorical digital chasm between
digital natives and digital immigrants—with tech-savvy children, teens and twenty-somethings on one
side and their quaintly outmoded elders on the other—as somehow self-evident.
REFERENCES
Bennett, S., Maton, K., & Kervin, L. (2008). The ―Digital Natives‖ debate: A critical review of the
evidence. British Journal of Educational Technology, 39(5), 775–786.
Jenkins, H. (2007). Reconsidering digital immigrants. Retrieved from
http://henryjenkins.org/2007/12/reconsidering_digital_ immigran.html
Lakoff, G., & Johnson, M. (1980). Metaphors we live by. Chicago: University of Chicago Press.
Prensky, M. (2001). Digital natives, digital immigrants. On the Horizon, 9(5), 1–6.
Tapscott, D. (1998). Growing up digital: The rise of the net generation. New York: McGraw-Hill.
Tapscott, D. (2009). Grown up digital: How the net generation is changing your world. New York:
McGraw-Hill.
2011
ISBN: 190671620X
US $37.00
300 pp.
SECTION 2: ASSESSMENT
Chapter 5 begins with an accessible description of assessment theory followed by the caveat that
traditional classroom assessments should not be applied to the online setting without substantial
modification. When developing an assessment strategy, they advise that teachers develop multiple forms
of assessment that maximize on the affordances of the tools used. A course assessment strategy that
provides students with enough opportunity to receive feedback and gauge progress is recommended.
Teacher feedback, they explain, should feed ―forward‖ (p. 63) and give students advice that will shape
future performance instead of simply giving grades or evaluative feedback. With respect to assessment
design, they warn that the skills assessed need to match the affordances of the online tools selected. In
such cases, the authors recommend that teachers generate student buy-in by creating authentic tasks and
assessments that simulate real-world applications of web-based or written communication.
Chapter 6 explores assessment in greater depth and presents it as a vehicle for language instruction.
Following up on earlier comments, the authors stress the need to provide feedback that is useful to the
student and can improve future performance. This chapter further discusses the nature of usable feedback,
specifically addressing teacher’s tone, depth of comments, as well as feedback on productive language
tasks specifically. Although cognizant of its time demands, they point to the value of individualizing
instruction by combining computer-based assessments with spoken and written feedback.
rely heavily on the use of the telephone, particularly for one-on-one sessions and tutoring. In rural areas
or where Internet access may be inconsistent, telephone land lines are often used. Many adult students
who do not have home Internet have cell phones with data plans that allow them to communicate with
teachers via email and enter course sites (The Parthenon Group, 2003). Challenges posed by the absence
of visual cues in telephone teaching are discussed along with strategies for maximizing on this tool. A
sample lesson plan of a telephone session is included to illustrate specific ways teachers can prepare
students, build community, and guide a class through a variety of communicative activities in the
telephone context. Consistent with the book’s practical focus, the stages of the telephone session and
common pitfalls are subsequently presented in detail.
Chapters 9 and 10 explore synchronous teaching. For the language educator striving to create an online
environment in which language interaction and negotiation of meaning can take place, student
participation is often a top concern. Chapter 9 focuses on planning to teach through several synchronous
online conferencing tools with attention to pedagogy, task design, and teacher roles during the session. As
in previous chapters, the authors emphasize the importance of a less dominant teacher role and provide
advice on how to maintain this balance. The authors note that students who might otherwise fade into the
background can become empowered by taking on leading roles with technology even when their language
skills do not match those of their in-class peers.
Chapter 10 takes a more practical approach to the delivery of synchronous teaching online. Guidelines
and principles for effectively managing groups, interactions, technology, and time during synchronous
sessions are offered. Noting the artificiality to which online interactions can become captive, the authors
discuss options for communication flow and provide ideas and techniques for giving feedback and
technical support with minimal intrusion. In their view, the teacher should be the architect of the learning
space working behind the scenes to ensure that students are able to interact with one another with few
interruptions or glitches. Other important topics covered here include the role of silence, peripherality,
and learner identities in the online context.
Chapter 11 presents the face-to-face component as a way to help overcome the shortcomings of
technology-mediated distance learning modes. Whereas online and telephone sessions can require a great
deal of planning and attention to detail, the face-to-face session is characterized by its fluidity. It can
serve as an important medium for reducing student anxiety levels through direct social interaction and
language games and can help establish a foundation for and lend authenticity to follow-up activities
online. Sample teaching plans are provided, and models that encourage the use of the physical space,
movement, spontaneous groupings, and the use of props are emphasized.
Chapter 12 focuses on asynchronous teaching, as many blended courses are taught using learning
management systems in which students complete assignments and interact primarily through
asynchronous discussion boards. Aiming to increase learner autonomy and expand opportunities for
language use in this mode, the authors propose blogs and wikis, providing explanations of each of these
tools and how to motivate students to take part in them. They also express concern for the social isolation
that is common in the asynchronous mode and advise that teachers create online spaces for students to
interact informally. The authors seem to understand the challenge that unstable student participation can
pose to language teachers in adult and continuing education programs where traditional grades or credits
may not be offered or used as an incentive and students’ motivations are often susceptible to shifts in
personal or external circumstances.
typical in adult education programs, these classes are sometimes diverse not only in student language
level, but also in native language and educational background.
Chapter 13 includes a rich discussion on language varieties and ways of leveraging technology to access
language resources and individualize input for diverse, multi-leveled learner populations. Even chapter
14, in which the history and nature of Celtic languages in the UK are discussed in further detail, there is
much that can be applied to a broader blended language teaching context, particularly with respect to
learner motivations and instructional design. There is also a section suggesting ways to further enrich the
blended language learning experience through community resources and volunteer participation.
REFERENCES
The Parthenon Group. (2008). Innovation in ESL education: Mobile-learning technology. Melinda Gates
Foundation. Retrieved from http://www.gatesfoundation.org/learning/Documents/innovation-esl-
education-mobile-learning-technology.pdf
Smith, C., & Gillespie, M. (2007). Research on professional development and teacher change:
Implications for adult basic education. Review of Adult Learning and Literacy, 7.
INTRODUCTION
Rocket Chinese is a straightforward, and self-contained language learning website. Rocket Chinese has
been developed specifically for English speakers who would like to learn Mandarin Chinese as a foreign
or second language. Interactive content, intelligent dashboard displays, personalized customer support,
and audio-based learning are all basic features of Rocket Chinese. With spot-on cultural information and
insightful audio lessons, this program is ideal for beginning to low-intermediate adult learners. As an
Internet-based language learning environment, Rocket Chinese combines up-to-date features of
interactive tools, such as a voice comparison system and interactive audio files with the benefits of social
utilities, such as Facebook, Twitter, and on-line forums. Users have lifetime access and get free upgrades
when the program updates.
NAVIGATION BAR
Once users have used their account, the opening portal at each subsequent log-in will provide them with
pie charts to orient them in their previous progress. Above these pie charts is the navigation bar which
consists of five principal components: Interactive Audio Lessons, Language and Culture Lessons,
Writing, Survival Kit, and Games. Each component has a drop-down list which shows different aspects of
the program’s language learning resources and materials. Rocket Chinese presents an enhanced user
PERSONALIZED DASHBOARD
Below the navigation bar is a personalized dashboard which guides users on the recommended learning
path by Rocket Chinese and allows users to see all the activities at a glance. This easy-to-manage
interface contains several sections: Tour Video, My Overall Progress, Recent Activity, My Proficiency,
Interactive Audio Lessons, Language and Culture Lessons, Writing, Survival Kit, and Games.
Several of these sections are straightforward. An orientation “tour video” as well as additional strategy
videos are offered. A self-rating system called "My Level" is introduced in the “My Overall Progress” area.
This progress tracking system enables users to determine how well they know a lesson and whether they
should restudy it or move on. Users can test their comprehension of words and phrases associated with
each lesson and get an overall rating for future reference. Users may also simply click the “My Level” icon
at the top right of every lesson to manually rate the lesson overall. Under “Recent Activity,” users can find
a hyperlinked list of lessons they recently viewed and rated. The proficiency tests draw from the quizzes in
each lesson. When a user has successfully completed each test, a certificate will be emailed for
encouragement.
There are, however, two comments about these sections. First, users need to notice that their progress does
not automatically update when they finish a lesson, and they must rate a lesson before the pie charts reflect
that change in progress. Also, the quizzes are shown in written form, which may confuse beginners who
have no idea about Pinyin (Pinyin is a kind of phonetic alphabet that uses the Latin letters to standardize
pronunciation of Chinese characters). It would be valuable if audio files could be provided at the same
time so the learners have better idea about the quiz questions.
The leading part of the program is the section “Interactive Audio Lessons,” which provides abundant
material, including lesson content, conversation transcripts, audio files, practice record, notes, tour videos,
lesson feedback, level rating system, and quizzes. The primary aim is to cultivate learners’ listening and
speaking ability. Learners at level #1 can access thirty-one lessons, and level #2 learners can enjoy a more
advanced selection of thirty-two courses in this section. The interactive audio course is designed to get
learners participating in a series of conversations between Mandarin Chinese speakers. It covers practical
topics that one may encounter in a Chinese-speaking environment, from self-introductions to various
points of discussion. The conversation page displays transcripts in Pinyin, Chinese script, and the English
translation. Previous research has shown that the use of similar transcripts can facilitate second language
comprehension (Chen, 2011). While the interactive audio course's primary aim is to cultivate learners’
listening and speaking ability, these trilingual transcripts may also enhance learners’ reading acquisition.
Additionally, the English translation helps learners keep up with the lesson and get comfortable with
written Chinese. The Pinyin transcripts provide users with useful pronunciation practice.
The “Practice” page offers users a voice comparison feature, “Rocket Record.” This voice-mapping tool is
designed to allow learners to record their voice and compare their pronunciation with that of a Chinese
tutor. Users can record and repeat until their accent, tone, and rhythm match that of the native host.
Strategies such as evaluating and monitoring can be important for language learners who do not have much
exposure to the target language (Li, 2010), but users do not receive any feedback on their pronunciation
from the program, which may raise questions about the adequacy of self-monitoring without recourse to
any authority (Engwall & Bälter, 2007).
The “Language and Culture Lessons” section attempts to address the need to integrate culture by
describing cultural issues, facts, and social interactions; it offers resources, such as lesson content,
transcripts, practice records, notes, tour videos, lesson feedback, a level rating system, and quizzes. The
courses teach users the grammatical concepts behind the conversation as well as give them insights into
different situations they are likely to encounter. Realistically portrayed impressions of Chinese society
(Chinese in this context meaning Chinese society as a linguistic group) help learners get the most out of
their experience. Many of the topics in the language and culture section correspond to the ones in the
interactive audio course. By linking these two sections, users have a chance to take a closer look at many
of the words and phrases they learned earlier. Each lesson in this section provides audio sounds for
individual sentences, but it would be even more helpful for learners if completed audio files for course
content were offered, just like those in the “Interactive Audio Lessons” and “Survival Kit.”
In the “Writing” section, there are easy-to-follow videos that teach learners how to write stroke by stroke.
However, several problems surface here. There are two different kinds of Chinese (see above note)
characters: simplified and traditional ones. Simplified Chinese is introduced throughout the program, but
traditional Chinese characters are left out, which limits learners’ ability to understand different cultural
dimensions of the target language. Moreover, the course content in this section does not relate to the
“Interactive Audio Lessons” and “Language and Culture Lessons” and does not take a practical, usage-
oriented approach. Words and phrases that appear in other sections or that are used in daily conversation
should be put in the writing section.
Rocket Chinese also includes a series of learning games which makes learning fun. These games are self-
paced and can be repeated, which gives users an incentive to keep improving their scores and language
skills. There are four kinds of learning games in this section: MegaVocab, MegaAudio, MegaCards
Beginners, and MegaCards Advanced. MegaVocab and MegaAudio are word games that build users'
Chinese vocabulary and test their audio recognition of Chinese words. These two games aim to increase
users’ reading, listening, and comprehension skills. Both of them are presented in Pinyin and also come in
a version with Chinese characters. Another feature of MegaVocab is the MegaVocab Creator, which
allows users to add their own extra words and pictures. MegaCards Beginners and MegaCards Advanced
are memory games that enable learners to test their knowledge of Chinese words and build familiarity with
common Chinese phrases. Two of the concerns are the absence of audio to assist learners to recognize and
speak those phrases, and the lack of alignment in the content of the four games with previous lessons.
Users can get instant feedback from quizzes, tests, and games so they can see how much their Chinese has
improved. Lesson content, mini games and audio tracks are downloadable. The on-average 28-minute
audio files are compatible with computers and iPod/MP3, so learners can listen and practice easily through
these portable and flexible files. The downloadable audio files are appealing, but additional files would
also be useful, potentially with video as well as audio.
ROCKET TOOLBOX
On the left frame of the homepage is an area where users have access to additional features: My
Community, My Motivation, My Notes, My Vocab, and Latest News.
“My Community” is a virtual space for learners to interact with the Rocket Chinese teachers, native
speakers, and fellow Chinese learners. Users can have the system email them automatically when
someone replies to a community topic they have commented on. The personal support is with learners
every step of the way, which adds a human touch to virtual learning. It would be useful if, possibly for an
extra fee, users could schedule a live tutor so that users could practice Chinese in real time. In “My
Motivation” users find guidance on how to advance their language learning with a range of techniques,
tips, tools, and motivational ideas with twenty-three lessons. The “My Notes” page displays the notes that
have been saved, and users can click on the lesson name to go straight back to the lesson.
“My Vocab” is a learning tool that makes it easy for users to build their own vocabulary list, practice and
remember important vocabulary, and rehearse pronunciation. Any word from the course materials can be
added with its meanings and notes. Users are able to listen to the word or phrase to improve recognition
and to practice their pronunciation as well. All the words in the “My Vocab” list are highlighted whenever
they appear in the course, and the meanings and notes will pop-up by clicking the words. Users need to
add the English translation by themselves when they build their vocabulary list. It would therefore be
beneficial to the learners to have a Chinese-English dictionary which can automatically translate the
words for users.
The “Latest News” announces new features of Rocket Chinese, such as introducing Rocket’s Facebook or
Twitter account, changing the layout, and adding more materials. The dearth of posts shows that the news
function is not frequently used. It would be desirable to have a moderator who could help to keep the site
active by posting very important or fun news to all users in order to make fuller use of this announcement
function.
CONCLUSION
Rocket Chinese presents an Internet-based language learning environment which is extremely friendly,
from the helpful words and phrases used in language instruction to the simple user interface. Rocket
Chinese offers a favorable language learning environment in a self-contained website that integrates
interactive features and inspires users with appropriate resources and materials to foster their learning.
The strengths of this program lie in the continuity of native presenter, the voice-recognition technology,
the variety of linguistic and cultural tips, the lifetime access, and the downloadable audio files. Rocket
Chinese allows users to learn Chinese at their desired pace, place, and time. Beginning or low-
intermediate adult learners of Chinese will find this program insightful and achievable. Overall, Rocket
Chinese makes learning Chinese a rewarding and interesting experience.
REFERENCES
Chen, H. J. (2011). Developing and evaluating SynctoLearn, a fully automatic video and transcript
synchronization tool for EFL learners. Computer Assisted Language Learning, 24(2), 117–130.
Engwall, O., & Bälter, O. (2007). Pronunciation feedback from real and virtual language teachers.
Computer Assisted Language Learning, 20(3), 235–262.
Li, F. F. (2010). Relationship between EFL learners' belief and learning strategy use by English majors
in vocational colleges. Journal of Language Teaching & Research, 1(6), 858–866.
Randi Reppen
2010
ISBN: 9780521146081
US $35.00
118 pp.
Chapter Three makes the jump from providing general information about corpus linguistics and its
classroom application, to detailing practical steps for exploring free online resources. The book outlines
some resources that are designed for language educators and others that can be approachable for students
as well. Reppen begins this chapter with a discussion of how to assess the quality of online resources so
that educators may feel free to explore the Internet—and the wide variety of resources available to
them—with a critical eye. From there, she introduces several approachable corpus-based online resources
that educators and students can use to learn about English. Finally, she introduces three online corpora
(Michigan Corpus of Contemporary American English; Corpus of Contemporary American English; and
the Time Corpus) by describing their composition (spoken English, multiple register English, and
academic English), as well as the resources that are available on their associated webpages. After
providing a checklist of points to keep in mind when developing materials from corpus information,
Reppen then highlights the strengths of each of the three corpora by leading readers through searches that
highlight teachable points for classroom materials.
By the fourth chapter, readers are aware of a number of both non-technical and technical methods for
finding corpus-based materials. While Reppen briefly lists several types of lessons that can be created for
students and offers some guidelines for making materials, she quickly turns the focus toward building a
personal corpus. This chapter explains a number of decisions that need to be made when building a
personal corpus and how to go about collecting and preparing texts. Language educators will also find
information not only for corpora of written language, but advice and useful software for creating spoken
corpora as well.
While the first four chapters introduce educators to resources they can use to inform their lesson plans and
activities, Chapter Five puts the corpora in the students’ hands. This chapter makes the case for student
use of corpora in the classroom by presenting five specific classroom activities that target a variety of
English proficiency levels and topics. The activities are laid out in step-by-step form, much like mini
lesson plans, and are designed with the assumption that students have access to computers and are
prepared to do their own simple searches. Reppen emphasizes the need to have a back-up plan whenever
asking students to access anything online, so a lesson is not ruined by a faulty link or site crash. If
educators are not in a teaching context that allows students to do their own searches in class, activities
could be adapted into handout-based lessons by the instructor. Reppen finishes the chapter by describing
several ways that corpora and corpus-based materials are being used in a variety of different teaching
contexts and offers some additional food for thought on how these resources may be of use to both
students and educators.
Further information is available in the appendices and warrants reading. Appendix A includes a handful
of photocopiable pages that were referenced in the first chapter and that can be used directly in a
classroom. Additional activity ideas for using corpora in the classroom are tucked in here as well,
including information about MonoConc Pro, a specific concordancer that can be used to search any
corpus of text files. The tutorial for MonoConc Pro will only be of value for educators who own a copy of
the software or are willing to acquire a copy. The book gives solid and easy-to-follow instructions for
doing a variety of tasks using MonoConc Pro and is well worth the read for anyone starting out with the
software. Appendix B includes a list of trusted online resources that language educators can explore.
Resources range from highly educator-centered corpus-informed materials, to more research-centered
materials that encourage independent corpus searches and evaluations.
In addition to the materials in the book, Using Corpora in the Language Classroom offers a companion
website that, while very slow, is worth a visit. Users will have to register on the Cambridge website with
an email address before accessing any materials. The website offers a handful of additional lessons and
worksheets. While the majority of them are geared toward high-level learners of academic English, they
do offer some insight into how activities might be designed and how corpus-based information may be
presented as a worksheet. The website also offers access to the KWIC results for three words in a readily
printable format, though it may be more valuable to look the same words up in one of the corpora
described in the book as a way of familiarizing oneself with the resources.
As with any book that relies on technology, and especially one that relies on online links, resources may
not be reliable. For example, at the time of writing this review, one of the resources described in the book,
the Birmingham Kibbitzers, was not available, though a subset of the Kibbitzers could be found hosted on
another site after a brief Internet search. Also, several screen shots in the book no longer match the most
updated website. The ever-changing nature of the Internet demands that educators be prepared for
hiccoughs when reading a book, such as this one, that relies so heavily on web-based links. Links may
change or move, sites may crash, and information may be taken down. Reppen appears to be aware of this
and does offer the list of links from the book on the book’s companion website. At this time, the list is
identical to the list in the book, but one might hope it will be updated periodically as sites inevitably
change. Many of the websites described in the book are hosted by universities. These sites have been
available for some time and are still being maintained and expanded. These links, therefore, appear to be
fairly stable and hopefully will continue to be available for some time to come.
A troublesome point that Reppen fails to discuss in her book is the increasingly strict protections being
placed on these bodies of language. Many of the hosting sites generously provide a concordancer to use
with their corpus, but several restrict any copying and pasting of language from their sites. Educators may
run up against this problem if they try to cut and paste lines from some of the online corpora into their
handouts, as suggested in the book. This does not at all limit educators and students in terms of language
exploration, and language samples can still be read and hand-typed into handouts, but these restrictions
may make things a bit more time consuming when creating classroom materials.
It is also important to note that Using Corpora in the Language Classroom is designed with an English
language teaching context in mind, so the materials and corpora described in the book are all based only
on English. The book does still give an accessible introduction to using concordancers in the classroom
that may be of interest to educators of other languages as well, provided they have access to a corpus of
their target language or are interested in spending time with Chapter 4 to create their own corpus. Activity
ideas are also adaptable to other languages.
In spite of these potential drawbacks, Using Corpora in the Language Classroom is an informative read
for those who are willing to set the book down periodically to pick up their computer and explore the
software being described. This is especially true in Chapter Three, in which many of the corpus searches
described are much easier to understand while manipulating them directly. Reppen gives directions that
are relatively easy to follow, and new users will almost certainly come away with a better sense of the
tool and what can be done with it if they take the time to play with the software online. For educators who
need visual reinforcement to make the jump from the text description to the actual online software, there
is a PowerPoint presentation available on the companion website that provides visual steps for the
educator of several corpus searches using the same corpora described in the book.
The breadth and depth of information that can be gathered from corpora and other corpus-based resources
is impressive and is becoming easier for educators to access. This book is designed as an introduction for
language educators who are truly starting at the beginning, or who are familiar with corpus linguistics but
are not sure how to make the transition to using it in their classrooms. For educators who are already
familiar with corpora and/or concordancers and are already using it in their classrooms, the book may be
too lightweight to be of much interest. However, the strength of Using Corpora in the Language
Classroom is the highly approachable tone used to introduce corpora and their uses in classrooms. Above
all, Using Corpora in the Language Classroom brings together a wealth of interesting websites for
language educators and their students to explore further.
Email: jmull@riele.org
INTRODUCTION
The construct of noticing can be said to be key in all cognitively-oriented approaches to SLA. There have
been several approaches to explaining this construct in the SLA literature (Gass, 1988; Robinson, 1995;
Schmidt, 1990, 1995, 2001; Tomlin & Villa 1994; Truscott & Sharwood Smith, 2011), and noticing is
widely viewed as a critical factor that mediates L2 input and interaction driven learning (Gass, 1997; Gass
& Varonis, 1994; Long, 1996). Schmidt (1990) operationally defined noticing as the availability for
verbal report. According to Schmidt, attention controls access to awareness and is responsible for
noticing, which is necessary and sufficient for the conversion of input to intake. Schmidt (1995) suggests
that input that is not attended to cannot be held in short-term memory and, therefore, is not available for
further processing. Researchers espousing a cognitive interactionist approach (IA) to SLA have borrowed
widely from Schmidt‘s notion of noticing. Gass (1997), for example, argues, ―an initial step in grammar
change is the learner‘s noticing (at some level) a mismatch between the input and his or her own
organization of the target language‖ (p. 28). That is to say, linguistic input from an interlocutor must first
be noticed before it may begin to be integrated into the learner‘s interlanguage (IL). Additionally,
noticing this gap may represent a higher mental activity than the simple noticing of L2 exemplars, in that
the identification of a gap between two forms involves a deeper linguistic analysis.
Measures of Noticing
In exploring the relationship between noticing and SLA, it is crucial that we have an adequate measure of
what learners notice and when. Three methods have been employed most widely. One is the analysis of
immediate uptake in the discourse (Braidi, 2002; Lyster & Ranta, 1997; Mackey & Philp, 1998; Tarone &
Bigelow, 2007). The other two methods are types of verbal reports, including concurrent verbal reports
such as think alouds (Alanen, 1995; Lai, Fei, & Roots, 2008; Leow, 1997, 2000; Rosa & O‘Neill, 1999;
Sachs & Suh, 2007) and retrospective verbal reports such as stimulated recall (Egi, 2007; Gass &
Mackey, 2000; Mackey, 2006; Mackey, Gass, & McDonough, 2000; Mackey, Philp, Egi, Fujii, &
Tatsumi, 2002).
Stimulated Recall
Stimulated recall is an introspective/retrospective means of exploring L2 learning processes which
otherwise are not observable (Faerch & Kasper, 1987). In stimulated recall, prompts such as audio- or
video-taped events are used to stimulate the participants' recall of their mental processes during the
events. A well-executed stimulated recall approach to determining what learners notice (awareness at the
level of noticing) in the input seems to be a strong choice for researchers in that it provides
complementary data for understanding, for example, the effect of corrective feedback (Egi, 2007;
Mackey, 2006; Mackey et al., 2002; Mackey et al., 2000; Egi, 2010).1
Recasts
The present study focuses on corrective feedback in the form of recasts. Therefore, a brief discussion of
this important construct as it relates to the current study is in order. Brown and Bellugi (1964), first
referred to recasts in the context of child L1 acquisition. And over the past fifteen years recasts have
become one of the most widely studied types of corrective feedback in the second/foreign language (L2)
classroom (Braidi, 2002; Doughty, 1994; Lyster & Ranta, 1997; Nabei & Swain, 2002; Oliver, 1995;
Sheen, 2004). Though there are several competing definitions of recasts in the instructed SLA literature,
recasts are essentially discourse moves (by an interlocutor) that, within the context of a communicative
activity, correctly restate or rephrase (modify) all or part of a learner‘s utterance to be more target-like by
changing one or more sentence components while still retaining its central meaning (Lyster & Ranta,
1997; Nicholas, Lightbown, & Spada, 2001; Sheen, 2006; Trofimovich, Ammar, & Gatbonton, 2007).
Recasts are generally theorized to be a powerful and potentially effective form of corrective feedback
because they simultaneously provide both negative feedback and positive input (Leeman, 2003), arguably
heightening the salience of targeted items (which may otherwise lack salience for learners) in a way that
does not disrupt the flow of communication.
Though there are studies that show limited or no effects of recasts on SLA (Lyster, 2004; Lyster & Ranta,
1997; Ellis, Loewen, & Erlam, 2006; Loewen & Erlam, 2006; Sauro, 2009), a more compelling case can
be made for their generally facilitative role in SLA (Goo & Mackey, in press), with their effectiveness
being more pronounced in lab over classroom settings (see Ellis & Sheen, 2006; Li, 2010; Long, 2007;
Lyster & Saito, 2010; McDonough & Mackey, 2006; Mackey & Philp, 1998; Mackey & Goo, 2007;
Nicholas et al., 2001). Indeed, Nicholas et al. (2001) suggest that recasts appear to be most effective in
contexts where it is clear to the learner that the recast is a reaction to the accuracy of the form, and not the
content of the original utterance, something more likely in lab settings.2
Problematizing Recasts
Comparing the research on recasts is problematic not least of all because recasts are a dynamic and
multifaceted, rather than a one-dimensional construct. For example, Sheen (2004) points out that recasts
are a bit more complex than most definitions suggest. Indeed, there are many different types of recasts,
including corrective, non-corrective, intensive, explicit, implicit, full, partial, simple, complex, adjacent,
and non-adjacent, among others. That is to say, recasts vary in their form and presentation thus making
general conclusions about their efficacy challenging.
Factors Affecting the Noticing of Recasts 3
Current research suggests that many factors affect the noticing of recasts. At a very basic level several
researchers have argued that there is a great degree of ambiguity in teacher-provided recasts and
therefore, recasts may be less effective than other types of feedback moves, such as prompts (Chaudron,
1977; Fanselow, 1977; Lyster, 1998; Lyster & Ranta, 1997; Nabei & Swain, 2002). One of the main
sources of this ambiguity—especially in the communicative and meaning-based classroom—is that
students are expected to sort out whether the teacher‘s intentions are focusing on form or meaning. In
addition to the argued general ambiguity of recasts, other factors that may affect the noticing of recasts
include the nature of the recast itself (Mackey et al., 2002; Long, Inagaki, & Ortega, 1998; Trofimovich et
al. 2007), learner-internal variables such as proficiency level and/or learner readiness (Ammar & Spada,
2006; Mackey & Philp 1998; Philp, 2003), working memory capacity (Mackey et al., 2002; Sagarra,
2007; Tarone & Bigelow, 2007), and learner anxiety (Sheen, 2008). Factors related to the pedagogical
and social context such as learner familiarity with the type of feedback provided and setting are also
important to consider.
When considering the relationship between the nature of the recast itself and noticing the following
aspects are apparent: the form targeted by the recast (the linguistic category) and the structural
composition and linguistic context of the recast. In terms of the former, there is evidence that learners first
focus on (and arguably notice) the semantics of interactional exchanges and only later focus on form
(Mackey et al., 2000; Tarone & Bigelow, 2007) despite the fact that teachers seem more inclined to
provide recasts more often to grammatical than to lexical errors (Lyster, 1998; Mackey et al., 2000; Nabei
& Swain, 2002). This is likely the explanation for the higher effectiveness of the lexical recasts over
grammatical recasts (Mackey et al., 2000; Sheen, 2006; Trofimovich et al., 2007; Williams, 1999).4
In terms of the structural composition and linguistic context of recasts, several variables are apparent,
which may affect their effectiveness. For example, Ellis et al. (2006) suggest that recasts may lie on an
explicit-implicit continuum. Recasts may also be declarative or interrogative (Kim & Han, 2007; Loewen
& Philp, 2006). Finally, the length, number of changes, and number of recasts on the same target seem to
be important factors related to effectiveness (Loewen & Philp, 2006; Philp, 2003; Tarone & Bigelow,
2007).
Setting and pedagogical context are extremely important in recast research (Ellis, Basturkmen, &
Loewen, 2001; Morris & Tarone, 2003; Nabei & Swain, 2002; Nicholas et al., 2001; Oliver & Mackey
2003; Sheen, 2004, 2006). Factors such as learners‘ prior familiarity with the input (Sheen, 2004), task
type and task content (Nicholas et al., 2001; Révész, 2009), and learner perception (Morris & Tarone,
2003) have been argued to be key. Morris & Tarone (2003), for example, suggest that recasts may
provide greater opportunities for language acquisition in contexts where both teachers and students are
oriented to attending to linguistic form as opposed to focusing exclusively on meaning.
While it is certainly true that many learner internal and external variables will combine to affect the
likelihood that recasts will be noticed by learners and, therefore, be potentially effective, methodological
issues surrounding the capturing, coding, and evaluation of recasts must be accounted for as well. Indeed,
Trofimovich et al. (2007) suggest that any lack of consensus regarding the relationship between recasts,
noticing, and SLA is largely due to the different methodologies employed by researchers to measure
noticing, as well as the use of various measures of working memory, proficiency level, etcetera.
Written Recasts
Seedhouse (1997) suggests that teachers often prefer recasts to other forms of corrective feedback because
they are relatively non-threatening, mitigated, unobtrusive, and implicit in nature. However, this
assumption is by no means absolute. The assumption that recasts are necessarily implicit in nature, for
example, is perhaps too simplistic for an instructed SLA context (Ellis & Sheen, 2006; Sheen, 2006) since
they may surely lie at various points on a continuum of linguistic implicitness-explicitness (Ellis et al.,
2006). For example, in an oral-aural mode accurately positioning a recast on this continuum may be
affected by things like repetition and stress (Doughty & Varela, 1998). One might also argue that written
corrective recasts in a synchronous computer-mediated communication (SCMC) environment are
necessarily more explicit than those found it its oral counterpart (yet still less explicit than, say, direct
error correction), all other things being equal (see also Dasse-Askildson, 2008). This is due to the written
nature of the medium itself as well as the message permanence afforded. Similarly, the argued ambiguity
(see discussion above) of oral recasts seems less convincing in the written mode. That is to say, written
recasts may be more forgiving of the learner‘s need to simultaneously distinguish features of the
interlocutor‘s responses that sustain the interaction from those that provide a means of focusing on the
language used in the interaction (see Nicholas et al., 2001). It also seems unlikely that learners will
interpret a written recast as simple repetition of their previous (well-formed) utterance.
Recasts in Text-Based SCMC
The research on recasts in a text-based CMC environment is quite limited. Loewen and Erlam‘s (2006)
investigated the relative effectiveness of recasts and metalinguistic prompts administered during small
group text-chat interaction. They found no significant advantage for either feedback type over the control
condition and no significant advantage for one corrective feedback type over the other. Sachs and Suh
(2007) compared the efficacy of textually enhanced and unenhanced CMC recasts in the development of a
targeted form (back shifting of verbs from simple past to present perfect in reported speech). They found
that textual enhancement was related to reported awareness and that higher levels of reported awareness
showed stronger correlations with posttest performance. However, there was no direct significant
relationship between enhancement and posttest performance. Sauro (2009) used full recasts in comparison
with metalinguistic feedback. Neither feedback type was significantly more effective than the other in
either the immediate term or over time. Lai et al. (2008), using think aloud and stimulated recall to
measure noticing, suggest that the contingency of CMC recasts affect whether or not they are noticed.
This explanation is related to the notion of enhanced salience (Leeman, 2003) with more contingent
recasts (perhaps) being more salient than non-contingent recasts. In Lai et al., 53% of the contingent
recasts were noticed as opposed to 35% of the non-contingent recasts.5 In a study employing a similar
methodology to that reported in the current one, Smith (2010) showed that learners noticed about 60% of
the intensive recasts they received. Lexical recasts were found to be much easier than grammatical recasts
for students to notice, retain, produce more accurately on a written posttest, and use more productively in
subsequent chat interaction. Similarly, Smith and Renaud (in press) found that learners were able to pay
careful attention to both lexical and grammatical targets in recasts. Learners focused on close to 75% of
teacher recasts with between 20% and 33% of these resulting in posttest gains.
There are many potential benefits of text-based CMC for SLA. From an interactionist approach, which
informs the present study, these benefits include increased student participation, an increased quantity and
heightened quality of learner output, an enhanced attention to linguistic form, and an increased
willingness to take risks with their second language (Smith, 2004; Pellettieri, 1999). The slower speed of
typing as well as a software- and network-induced lag time between turns, coupled with the heightened
salience of input (including corrective feedback) and output afforded by the permanence of the written
message on the screen means that interlocutors have more time to both process incoming messages
(Pellettieri, 1999) and produce and monitor their output (Smith, 2008; Sauro & Smith, 2010). Essentially,
then, from a cognitive interactionist perspective, the potential advantages afforded SLA by text-based
SCMC comes down to the construct of noticing—especially the noticing of corrective feedback, as well
as non-target like input and output (Smith, 2004; Izumi, 2003; Salaberry, 2000). More specifically, text-
based SCMC may be said to afford more online planning time, which is argued to contribute to L2
performance by freeing up attentional resources (Yuan & Ellis, 2003). Thus learners are able to more
closely attend to linguistic form while formulating messages and also engage in more focused pre- and
post-production monitoring (Yuan & Ellis, 2003). Sauro and Smith (2010), using screen capture
technology in an SCMC study, convincingly show that learners do appear to use the increased online
planning time afforded by chat to engage in careful production that results in more complex language.
Though the actual online planning itself remains an unobservable activity (Skehan & Foster, 2005), their
findings support previous (face-to-face) studies (Ellis & Yuan, 2004; Yuan & Ellis, 2003), which also
found a positive effect for increased online planning time. Recasts, then, in an SCMC environment might
be expected to be especially salient to learners, though the research to date seems mixed (Lai et al. 2008;
Sauro, 2009).
Eye Tracking
Before moving on to a discussion of the role eye tracking played in the current study, a brief examination
of eye tracking technology in general is in order. Eye tracking technology has been employed as a tool in
psychological reading research for over 100 years. It consists of a suite of techniques whereby an
individual‘s eye movements are measured so that the researcher knows both where a person is looking at
any given time and the sequence in which their eyes are shifting from one location to another (Poole &
Ball, 2006).
Eye movements during reading—such as gaze duration, saccade length, occurrence of regressions, and
search time—can be used to infer moment-by-moment cognitive processing of a text by the reader (Just &
Carpenter, 1980) without significantly altering the normal characteristics of either the task or the
presentation of the stimuli (Dussias, 2010). These eye movements are considered empirical correlates of
processing complexity, which allow us to make inferences about perceptual and cognitive processes. For
example, the analyses of eye-movement patterns have been used to explore the cognitive processes of
pronoun resolution and co-reference, word frequency, lexical ambiguity, syntactic ambiguity, and
discourse factors (for a review, see Rayner, 1998).
The two most widely used measures of eye movements are eye fixations and saccades. Eye fixations are
those moments when the eyes are relatively stationary and reflect when information is being encoded. It is
eye fixations that allow readers to extract important and useful information about the text (Dussias, 2010).
Saccades are the fast movements of the eyes between fixations. No encoding takes place during saccades,
so they are not helpful in interpreting a target‘s complexity or salience.6 In general, eye fixations during
(L1) reading in English last approximately 200–250 milliseconds (Rayner, 2009), though there is
considerable within- and between-reader variability. For example, a fixation may range from just under
100ms to more than 500ms. This variability in fixation duration is believed to vary as a function of the
relative cognitive difficulty in comprehending a text (Rayner, 1998). However, L1 readers do not fixate
on every word in a text, but rather they fixate on about two-thirds of the total words (Just & Carpenter,
1980).
Context is extremely important in interpreting the meaning of fixations, which cannot be done in a
vacuum. For example, when browsing a Web page (an encoding task), higher fixation frequency on a
specific area can mean greater interest in the target, such as when one views a picture in an advertisement.
Higher frequency can also be a sign that the target is complex in some way and more difficult to encode
(Jacob & Karn, 2003; Just & Carpenter, 1976). Poole, Ball, and Phillips (2004) suggest that more
fixations on a particular area indicate that it is more noticeable, or more important, to the viewer than
other areas (reported in Poole & Ball, 2006). However, in search tasks, a higher number of single
fixations, or clusters of fixations, are often an indicator of greater uncertainty in recognizing a target item
(Jacob & Karn, 2003). The duration of a fixation is also linked to the processing-time applied to the object
being fixated with a longer fixation duration indicating either difficulty in extracting information, or that
the object is more engaging in some way (Just & Carpenter, 1976). The so-called eye-mind assumption
holds that the reader's eyes remain fixated on a word as long as the word is being processed.
Up to now, eye-movement records in L2 research have been used to investigate two major areas: (a) the
way in which L2 speakers recognize words when they are spoken in each language; and (b) the question
of whether monolingual and L2 speakers process various syntactic sub-processes similarly during
sentence comprehension tasks (see Dussias, 2010, for an overview). Only a handful of published studies
have so far employed eye tracking to explicitly examine L2 noticing (Godfroid, Housen, & Boers, 2010,
Kuhn, 2012; O‘Rourke, 2008, in press; Smith, 2010; Smith & Renaud, in press). Each of these studies
suggests that eye tracking is suitable to use as an instrument for measuring the noticing of written text.
METHODOLOGY
Participants
Potential participants were recruited by a general call for volunteers circulated among the non-native
speaker sections of freshman composition as well as among the university‘s intensive English program
and international student groups at a large university in the United States. Volunteers were paid a small
amount for their participation in the study. Participants‘ (n = 18) self-reported L1s were as follows:
Chinese (10), Cantonese (2), Korean (2), Turkish (2), Mandarin (1), and Thai (1).
Instruments and Procedures
Pre-Treatment: Pre-Test
A link to an online test of general English proficiency was sent to each participant prior to the treatment
(http://www.transparent.com/learn-english/proficiency-test.html). Participants completed 40 four-item
multiple choice questions about English grammar and vocabulary. Raw scores are automatically
calculated by the website upon completing the test and scores are immediately available to the user. These
proficiency test scores, which ranged from 63–98 (out of 100) were then recorded and their overall score
was used in testing for normal distribution across this variable, which was sufficiently met (n = 17,7
M = 86, SD = 11).8 Figure 1 shows the order of each of the procedures discussed in this section.
Figure 1. Procedures.9
Recast Training
I discussed earlier the multiple potentially confounding variables in interpreting recast research. Among
these variables is the notion of ambiguity, which is inextricably intertwined with pedagogical and social
context. One major design aspect of the current study was to remove this variable to the extent possible.
To this end, participants were explicitly trained on the nature and intent of recasts, which in this case were
designed to orient participants to form rather than meaning.10
A few days prior to their treatment session, a sheet detailing the concept of recasts in language teaching
was sent to each participant. This sheet also included examples of several types of recasts, lexical,
morpho-syntactic, semantic, etcetera. Participants were asked to review this sheet and familiarize
themselves with the concept of recasts as well as with the examples provided. A separate recast
evaluation sheet was also sent to each participant. This sheet consisted of six items that reflected a short
fictitious teacher-student classroom exchange. Participants were asked to determine whether each item
reflected a recast or not.
The conscious choice to explicitly introduce learners to the form and function of recasts was based on
research that shows learners‘ orientation toward and familiarity with specific types of corrective feedback
is an important variable in their ability to utilize the feedback (Morris & Tarone, 2003; Philp, 1999;
Sheen, 2004), and also on the suggestion that recasts appear to be most effective when it is clear to
learners that the recast is a reaction to the accuracy of the form and not the content of their original
utterance (Nicholas et al., 2001). It is my view that in an instructed L2 setting, learners do indeed
gradually (or even quickly) pick up on the instructor‘s intent when providing corrective feedback of
various types, as well as when they engage in questioning, praising, silence, etcetera. As such, the notion
that a particular form of corrective feedback is implicit, explicit, or somewhere in between is largely one
stemming from the instructor‘s point of view rather than that of the learner. The trade-off in making this
methodological decision, of course, is that the relevance of any findings to studies of purely implicit
recasts or those conducted in markedly different pedagogical (and social) contexts may be limited.
Nevertheless, I argue that the approach taken here in no way compromises the argued benefit of recasts,
which essentially comes down to teachers heightening the salience of a targeted form by providing both
positive and negative evidence in a way that is palatable to the communicative classroom.
Day of Treatment
Participants handed in their pretest sheets as well as their recast evaluation sheet at their scheduled
session. If participants scored less than 100% correct on their recast sheet, the researcher reviewed those
incorrect items together with them and explained those items that were incorrect. Learners were allowed
to ask questions in order to clarify the nature and purpose of recasts in their minds.
Next, the researcher verbally demonstrated two additional recasts for students and discussed with each
participant why each of these was a recast. The participants were then asked to think for a moment and
come up with a short dialog where participant A was the student and B was the teacher. This dialog was
to contain one error of some sort and one recast by the teacher. Participants were then asked to identify
the error and the recast. This part of the procedures concluded with a short discussion of whether recasts
are common in language learning classrooms in the participant‘s home country as well as their experience
with recasts in ESL classrooms in the United States. In this way, it was reasonable to assume that
participants had a good grasp of the nature and purpose of recasts, therefore, removing a potentially
confounding variable from the research design. That is to say, every effort was made to ensure that
participants were aware of the existence, intent, format, and function of recasts as a form of corrective
feedback.
Calibrating the Eye Tracker
A MangoldVision portable eye tracker from EyeTech Digital Systems (model TM3) was used in this
study along with the MangoldVision Gaze Data Analysis Software Suite. This remote eye tracker fastens
to the bottom edge of any computer screen.11 An infrared camera follows one‘s pupil movements and can
record one‘s eye gaze in real time. It also allows for a normal range of head movements (25cm x 16 cm x
19 cm) and will find one‘s pupils again even if one is temporarily out of range. The eye tracker must be
individually calibrated to each learner. Learners follow a red ball displayed on the screen with their eyes
for about 30 seconds until the software provides a score reflecting the calibration for each eye. For this
study, the acceptability level was set at 75%. In order to provide an additional source of calibration,
learners were instructed to slowly focus on a test line that consisted of three specific letters (spaced apart
from one another), which appeared as the first message from the researcher at the beginning of the chat
session. They were also asked to read these letters aloud in order to provide additional verification
regarding each learner‘s eye gaze patterns. That is, it was possible to match the eye gaze record with the
audio record of where participants were focusing their attention.
In addition to the eye tracking data captured, the Mangold software includes a video (and audio) screen
capture capability, which was used to play back the chat interaction (with the eye gaze tracking ball
disabled) in real time during the stimulated recall sessions. Camtasia 6.0 was used to capture the audio
and video playback of the stimulated recall sessions. The text-based chat software used in this study was
PSI, a jabber-based compose and post software. The spell check (check as you go) function was disabled.
The Task
Learners were asked to view a short clay animation video clip (Plonsters) of about 2½ minutes in
duration. The Plonsters characters use no real language in the clip, though they do communicate with one
another using an indecipherable gibberish. These characters also use a wide range of facial expressions
and extra- and para-linguistic vocalizations.12
Immediately after the video ended, participants were asked to chat with their interlocutor online. The task
goal was to retell the story as accurately as possible to allow the interlocutor to put a sequence of still
images taken from the video into the proper order (sequential ordering task). Learners were told that their
interlocutor may ask questions as well as provide comments and corrective feedback including recasts.
In additional to providing continuers (such as ―OK‖), comments, and follow-up questions on the content
of the learner‘s previous message, the researcher (interlocutor) provided full recasts when it seemed
natural to do so.13 These recasts were provided at the earliest possible point in the chat interaction after
the researcher noted a problematic utterance, thus resulting in recasts that were primarily contingent in
nature (Lai et al., 2008). Further, since the interlocutor stressed the naturalness of the recasts, many of
these recasts targeted more than one linguistic item in the participant‘s previous message. During the task,
all text-based chat interaction was logged automatically to a text file. Once the task was declared finished
by the learner, the chat file was saved and the MangoldVision Gaze program stopped. A maximum of 12
minutes were allowed for the chat task.
Immediate Posttest (IPT)
Immediately following the chat session, learners were asked to write (in MS Word) a narrative retelling of
the video they had just seen. They did so using a laptop PC set up in the same room. Spell and grammar
checking were disabled on the computer. Participants were given 12 minutes to complete this test. No
outside resources were allowed.
Stimulated recall
Upon completing the immediate posttest, participants engaged in a stimulated recall with the researcher.
The chat interaction was replayed using the MangoldVision Gaze Data Analysis Software, without sound
and with the eye tracking ball disabled. As part of the stimulated recall procedure, the video playback was
paused after each incoming message from the researcher. Camtasia Studio 6 was used to record the
stimulated recall session, including the screen capture of the chat interaction and the audio produced
during the stimulated recall questions and answers. At each message the researcher asked ―What were you
thinking when I sent you this message?‖ This was often followed up by the question ―At the time, did you
notice anything in particular about my message?‖ When learners indicated that they had in fact noticed
some difference between their preceding message and the recast (where applicable), they were asked to
specifically identify this difference verbally. The researcher also asked specifically if they noticed that
difference ―just now‖ or ―during the chat.‖
Delayed Posttest (DPT)
The tailored delayed posttest consisted of a participant-specific chat transcript adapted to isolate the
participant‘s output. For the DPT, the chat transcripts could be reviewed by the researcher and each
targeted (recast) item could be presented to learners, something not possible in the immediate posttest.
Asking participants to complete a written story retelling similar to that in the IPT was deemed unrealistic
given the short exposure to the story (video) and the time lag between posttests. All messages from the
chat session that were sent by the researcher/interlocutor were removed. That is to say, the DPT was a
transcript of the learner‘s half of the chat interaction. Each substantive line of learner output was
numbered. Complete sentences from the chat transcript that were spread across multiple lines during the
chat interaction were combined into one cohesive line for the purposes of the delayed posttest. The
number of entries on the delayed posttest varied for each participant based on the length of their chat
transcript. Since all lines of learner transcript text were included in the delayed posttest, those lines that
were not the target of a recast during the chat interaction (well-formed utterances) served as distractors.
The delayed posttest asked learners to determine whether each sentence was correct as written or if there
were one or more errors in the sentence. If participants felt the sentence was correct as written, then they
were instructed to simply write ―OK‖ under the sentence. If they felt that there were one or more errors,
they were instructed to write ―No‖ and then rewrite the sentence in its entirety repairing the error(s).
Thus, the DPT accounted for the target items of every recast offered by the researcher. The DPTs were e-
mailed to each student one week after the student‘s treatment session. They were asked to complete the
posttest within one day and e-mail it back to the researcher.14 Once the DPT file was opened participants
were given one minute per item to complete the DPT. That is, if a learner‘s DPT consisted of seventeen
items (say, ten target items and seven distractors), this learner would have seventeen minutes to complete
the DPT.15
Morphological errors:
All errors in verb tense or form; plural or possessive ending incorrect, omitted, or unnecessary;
subject-verb agreement errors; article or other determiner incorrect, omitted, or unnecessary.
Semantic errors:
Errors in word choice, including preposition and pronoun errors; omitted words or phrases,
unnecessary words or phrases. Spelling errors included only if the (apparent) misspelling resulted
in an actual English word.
Syntactic errors:
Errors in sentence/clause boundaries (run-ons, fragments, comma splices), word order, other
ungrammatical sentence constructions.
Figure 2. Procedures for marking errors. From Liu, 2008.
Measures of Noticing
There were up to two different measures of noticing recorded for each recast item for every participant. I
borrow the term noticing event from Godfroid et al. (2010) since both eye gaze and stimulated recall data
are indirect indicators of noticing and tell us little about the nature of cognitive processing that ensues
from such an event. Noticing events are essentially evidence of increased visual attention to the target,
which suggests that the reader was cognitively engaging with the target item. In terms of the eye gaze
data, this increased visual attention is marked by the overall dwell time, which consists of the sum of all
fixations on a word, including first run dwell time and all regressions.
In contrast, noticing events during stimulated recall are defined here as cases where learners indicated that
a specific recast was targeting an earlier non-target-like item and they were able to verbally identify the
noted difference. I signify a noticing event with the symbol ―+‖ and the absence of a noticing event with
the symbol ―-‖. For the statistical analysis discussed later, those items reflecting a noticing event were
coded as 1 whereas those items where no noticing event is evident were coded as 0. Further, since there
will be many references to noticing events across both measures of noticing employed in this study, I will
abbreviate these as follows:
In many cases it was necessary to create several heat maps for a single recast item to accommodate these
shifts in the screen. Figure 3 (above) reflects a single item recast. The recast in this case was targeting the
verb ―got,‖ recasting it as ―jumped.‖ Based on the heat map data, this item was coded as ―0‖ (not noticed)
for the variable heat map since there was no blue shaded area on any part of the word jumped.
In contrast, Figure 4 (below) shows a heat map for the single recast item targeting the form of ―to throw‖
used by the learner. Since there is blue shaded area touching part of the word ―threw,‖ this recast item
was coded as ―1‖ (reflecting a positive noticing event) for the variable heat map.
Stimulated Recall
Like the heat map measure, stimulated recall was coded as either 0 or 1. In cases where learners indicated
that a recast was targeting an earlier non-target-like item and were able to specifically identify verbally
the noted difference, that item was coded as 1 (SR+). Again, such items were only coded as SR+ if
learners specifically indicated (upon prompting) that they had noticed a specific difference between their
previous output and the interlocutor‘s recast ―during the chat.‖ Those items where learners indicated that
they just noticed the difference during the stimulated recall were coded as 0 (SR-). As will be discussed
below, this presents some difficulty in analyzing the delayed posttest data for stimulated recall since
stimulated recall essentially becomes part of the treatment (see also Swain, 2006).
Immediate Posttest (IPT)
Each IPT was reviewed for productive use of the targeted recast items from the chat interaction. Since
recasts were only provided for non-target-like utterances, each targeted item was considered unknown (or
at least not mastered) prior to the treatment. It is important to note that in cases where learners initially
used an item in a target-like way, but then later in the chat used this same item in a non-target-like way
(resulting in a recast from the interlocutor), this item was excluded from the data since it may reflect some
degree of competence with that item. That is to say, only those items that showed only non-target-like use
upon their first attempt by learners during the chat interaction are included in the data below. Further, if
there was only one attempted use at an item and this use was also non-target-like and during the
stimulated recall the learner revealed that they had known the correct form of the targeted item
(performance error), then this item was also removed from the data.
Since the IPT essentially asked learners to repeat the treatment task, it was expected that a large
percentage of the structures and items used during the chat interaction would be repeated during the IPT.
Only those items that were recast during the chat were candidates for consideration. In cases where a
single target item was attempted several times in the IPT, only those cases where all attempts were target-
like or non-target like were considered and issued a score of 1.0 and 0.0 respectively. In cases where the
IPT showed a mix of target-like and non-target-like use of a recast item, this item was removed from the
IPT data. Not all targeted items were attempted by learners during the IPT. In the very few instances
where learners made grammatically correct changes to the targeted item(s) that differed from the form
suggested in the interlocutor‘s recast, these cases were omitted from the data. These coding decisions
were made in view of the types of statistical analysis chosen, (which require binomial data) and are
considered a limitation of the study.
Delayed posttest (DPT)
The DPT score for each targeted item was calculated in the following way: If learners failed to correct a
sentence that was recast in the treatment phase, then this item was scored 0 (incorrect). Likewise, if they
made non-target-like changes to a recast item, then this was also scored 0 (incorrect). Changes (correct or
incorrect) made to the distractor items were ignored. Only those recast sentences that were rewritten to
follow the form suggested by the interlocutor in the recast were scored as 1 (correct). In the very few
instances where learners made grammatically correct changes to target items in a recast sentence that did
not mirror that offered by the interlocutor during the chat session, these cases were omitted from the data.
RESULTS
The data discussed below are made up of 156 total target items across 18 participants. The actual number
of recasts (114) was smaller than the 156 total target items since many of the recasts contained multiple
targeted items. Though the actual number of errors addressed in any given recast is certainly of interest,
this study focused on full recasts wherever possible and, therefore, contained both single item and
multiple item recasts. The difficulty in providing only partial recasts is that learners may take the
interlocutor‘s omission of corrective feedback to be a tacit confirmation of the accuracy of their non-
target-like use.
Table 1 below shows the noticing events across the two measures explored. This table shows that where
stimulated recall occurred for an item (where there was a valid record of stimulated recall for that item),
learners indicated noticing (SR+) that targeted item during the chat interaction in 74 of the 137 cases, or
54% of the time. No noticing SR- was reported in 63 cases, or 46% of the time. One of the 18 participants
had no stimulated recall record because of technical difficulty. All 18 participants had a valid heat map
record. Table 1 shows a noticing event (HM+) for the targeted items (after the recast) during the chat
interaction in 90 of the 127 cases, or 71% of the time. No noticing event was apparent via heat map in 37
cases, or 29% of the time.
Table 1. Noticing Events Across the Measures Stimulated Recall and Heat Map
Stimulated Recall Heat Map
-/0 +/1 -/0 +/1
n (%) 63 (46%) 74 (54%) 37 (29%) 90 (71%)
Valid cases 137 127
Table 2 shows the overall proportion of recast items scored ―correct‖ on the posttests. Of all of the non-
target-like items that were recast by the researcher and for which we have immediate and delayed posttest
data, about 59% and 41% were scored as ―correct‖ on the immediate and delayed posttests respectively.
Table 2. Overall Immediate and Delayed Post-test Scores for all Valid Cases of Recasts
Total number of recasts Number scored correct Overall % correct
Immediate Posttest 121 71 59%
Delayed Posttest 85 35 41%
Research Question 1
What is the relationship between the individual measures of noticing (heat map and stimulated recall) and
posttest scores for the targeted recast items?
Tables 3 and 4 below show the descriptive data and overall immediate and delayed posttest performance
for recast items (noticed and not noticed) based on each of the measures of noticing employed. One
notices that the stimulated recall and heat map records are quite similar to one another in terms of the
number/percentage of recasts with positive and negative noticing records as well as the respective posttest
scores for those same items.
Table 3. Posttest Scores on Individual Target Items Based on Stimulated Recall Measure
Noticing Immediate posttest Delayed posttest
record n (%) Mean score SD n (%) Mean score SD
SR- 46 (43%) .15 .36 43 (52%) .15 .35
SR+ 62 (57%) .87 .34 40 (48%) .73 .45
Total 108 (100%) .56 .50 83 (100%) .42 .50
Taken together, the descriptive data suggest that the heat map and the stimulated recall measures, though
strong predictors of IPT success (IPT = 1), and to a certain extent DPT success, were very strong
indicators of what learners likely failed to notice given the very low posttest means for those recasts
coded as HM- and SR- respectively.
Table 4. Posttest Scores on Individual Target Items Based on Heat Map Measure
Noticing Immediate posttest Delayed posttest
record n (%) Mean score SD n (%) Mean score SD
HM- 33 (32%) .09 .29 27 (40%) .15 .36
HM+ 69 (68%) .81 .39 41 (60%) .66 .48
Total 102 (100%) .58 .50 68 (100%) .46 .50
From these tables we see a considerably higher percentage of recasts with a positive noticing event record
in the heat map condition (about 68%) relative to the stimulated recall condition (about 57%). However,
the mean IPT score of noticed items is slightly higher in the stimulated recall condition and more
noticeably so in the DPT. Likewise, the mean IPT score on the recast items not noticed (HM-) is
considerably lower in the heat map condition than in stimulated recall. This suggests two things: first, it
suggests that the heat map measure is picking up on recast items that are attended to at some level, but
this engagement often does not translate into the ability to productively use (in writing) the target-like
form of the recast item a short time later. It also suggests that if the heat map record shows no evidence of
noticing (HM-), then it reasonably certain that the item was not noticed at any level. In contrast, evidence
of noticing based on the stimulated recall measure is a more solid indicator that learners have noticed a
targeted item; at least at a level that allows target-like production a short time later. When comparing heat
map and stimulated recall, the higher immediate posttest scores for items in the stimulated recall
condition that were not noticed (SR-) suggest that learners perhaps did notice many of these items at some
level, but were not able to articulate this during the stimulated recall session.
Because participants received an unequal number of recasts from the interlocutor, the 156 observations
were not independent of one another and were grouped by participant. This guided the statistical analysis
chosen as most of the more commonly used statistical analysis procedures assume an independence of
observations. In order to take this non-independence into account, a Generalized Estimating Equation for
Logistic regression (GEE logit in SAS) was used (with exchangeable working matrix) (hereafter GEE
logit). The GEE logit estimates the same model as the standard logistic regression but GEE logit allows
for dependence within clusters (participants). This analysis was conducted for each measure of noticing
(stimulated recall and heat map) for both immediate (IPT) and delayed (DPT) posttests.17
Table 5 shows the likelihood that a particular recast item would be scored correct on the posttest when it
showed a positive noticing event (SR+, HM+) by the two measures of noticing.18 For example, we can see
that during the stimulated recall session, when learners indicated that they had noticed a targeted item in a
recast (SR+), then the chances that this participant would get that same targeted item correct on the
Immediate Posttest was 38.43 times greater than in the SR- condition (p < .0001). Thus, Table 5 suggests
that both measures of noticing, when considered individually, seem to be generally favorable predictors of
immediate posttest success with heat map being the stronger of the two. In terms of delayed posttest
success the two measures of noticing seem to be much more similar in predicting the probability of
success. The DPT value for stimulated recall shows that when participants have indicated SR+ for an
item, that they are over 15 times more likely to score the item correct on the posttest than when they
indicated SR-. This is a bit higher than the same statistic for heat maps, but may be (partially) explained
by the fact that whereas the IPT occurs before the stimulated recall session, the DPT occurs afterward,
allowing for a second exposure.
Research Question 2
What is the relationship between heat map and stimulated recall to one another when used as measures of
noticing?
To help answer Research Question 2, which focused on the possible relationship between heat map and
stimulated recall, the same type of GEE logit was used as in the analysis reported in Table 5 above.
Table 6 reflects the likelihood that stimulated recall will indicate noticing (SR+) when heat map also
indicates noticing (HM+). That is to say, when the heat map indicated a noticing event (HM+), it was
11.45 times as likely that the stimulated recall measure would also indicate noticing (SR+) than when the
heat map measure did not indicate a noticing event (HM-).19
Turning our attention now to how well the combination of measures predict posttest success, Table 7
below shows that when considering heat map in conjunction with stimulated recall, both measures add
positively and significantly to the model predicting IPT success. In this case when both measures are
included in the model simultaneously, stimulated recall is a slightly stronger predictor of IPT success than
heat map. In terms of the DPT, heat map fails to add any predictive weight to determining posttest
success, whereas the stimulated recall measure continues to do so. Again, this is likely related to the fact
that in the DPT participants had the benefit of a second exposure to the corrective feedback.
Table 8 shows the same data as in Table 7, but in a slightly different way. In Table 8 we see that posttest
scores for each recast item for which there exists a valid stimulated recall and heat map record
(IPT n = 93; DPT n = 66). There are four possible combinations for the noticing record. These are listed
under the heading measure of noticing.
Table 8. 20 Posttest Scores for Individual Target Items by Heat Map (HM) Stimulated Recall (SR) Record
Measure of noticing IPT score (n = 93) DPT score (n = 66)
HM SR 0 1 0 1
Row 1 0 24 (.92) 2 (.08) 20 (.91) 2 (.09)
0
Row 2 1 6 (.86) 1 (.14) 3 (.60) 2 (.40)
Row 3 0 8 (.89) 1 (.11) 4 (.57) 3 (.43)
1
Row 4 1 2 (.04) 49 (.96) 8 (.25) 24 (.75)
As expected, the first row shows that when neither stimulated recall nor heat map indicate noticing, then
those items are typically scored incorrect on the IPT. To be clear, this means that learners in these cases
produced the same faulty formulation on the IPT that they used in the chat interaction even after receiving
a target-like recast from the interlocutor. Also as expected, row four shows an overwhelming score of
correct when both the stimulated recall and heat map records show evidence of noticing. Most interesting
are rows two and three. The second row shows seven instances of the combination HM-/SR+ (IPT score).
The data here are in the direction expected. A HM-record suggests that these items were not viewed by
the learner and, therefore, were not available for further processing. I suggest that the six cases of
incorrect despite the SR+ score indicate that learners misreported what they noticed during the task. That
is to say, that they actually noticed the recast for the first time during the stimulated recall session, but for
some reason reported having noticed it during the chat interaction. Another possible but unlikely
explanation is that in these cases they did in fact notice the targeted item (and the eye tracker did not pick
up on these fixations) and their incorrect score on the IPT was simply a performance error.
The third row shows nine instances of the combination HM+/SR- for the IPT.21 In these cases learners
gave no indication in the stimulated recall session of having noticed the recast item during the chat
whereas their heat map record suggests there was some heightened degree of attention to it. I will argue in
the next section that these cases present evidence of learners attending to the targeted items in the input at
some level (beyond that required for simply reading it, in which case we would expect a heat map color of
purple), but not at the level of noticing with awareness. The percentage of such cases scored incorrect to
correct was 89% to 11% respectively with only one instance of a positive IPT score. The data from the
DPT shows essentially the same patterns, but we must be cautious in interpreting the DPT data since
learners in this case had the benefit of re-exposure to the target items and some degree of explicit
attention to form during the stimulated recall sessions as well as additional time (one week delay).
In summary, both measures of noticing, when considered independently, are positive predictors of IPT
and DPT success. Further, where one measure suggests a noticing event, then there is a high likelihood
that the other measure will indicate a noticing event as well. Stimulated recall reflects a slightly less
powerful indication of IPT success (indeed, perhaps negligibly so, see Table 5) arguably because of
memory decay which occurs between the chat interaction and the IPT. In contrast, the slightly more
powerful indication of DPT success afforded by the stimulated recall measure may suggest one of two
things. First, this measure consistently reflects a level of noticing with at least a low level of awareness
whereas the heat map measure may not. Second, since the stimulated recall measure occurred after the
task had been completed, whereas the heat map measure is recorded online during the task, we must
assume that this fact could have had an effect on the DPT scores. That is to say, it may be that the DPT
scores for the stimulated recall condition are high (partially) because of the stimulated recall procedure,
which essentially becomes part of the treatment. Thus, unless we are to discuss the pedagogical value of
stimulated recalls, then it is best to proceed cautiously with our interpretations of any apparent differences
in the DPT scores of the two measures and stress, rather, their similarity.
Building on the results above that show the likelihood of posttest success based on noticing events, the
next research question sought to examine the possible relationship between the linguistic category of the
recast and noticing. Though because of the modest sample size these results are not explicitly tied to the
posttest measures, the assumption here is that those recasts showing a noticing event would result in
similar posttest patterns as those found in Research Questions 1 and 2.
Research Question 3
What is the relationship between the linguistic category of the recast target and noticing?
In this section I will examine the relationship between linguistic category on the noticing of recasts as
measured by heat maps and stimulated recall.
Since posttest success was often used above as the major dependent variable in evaluating the viability of
the two measures of noticing, it was important to eliminate the possibility that linguistic category itself
(rather than noticing) had an effect on the posttest gain scores. That is to say, it may be reasonable to
think that learners simply score higher on target items of one linguistic category over the others for some
reason unrelated to noticing.
Since each recast was coded for linguistic category, a GEE logit was used as in Research Questions 1 and
2 above. This analysis showed no significant main effect for linguistic category for either the immediate
posttest or delayed posttest scores. Table 9 below shows the descriptive data for the immediate and
delayed posttest scores respectively across the three linguistic categories. These data suggest that posttest
scores did not vary (statistically) as a function of linguistic category, though for both tests we do see
markedly lower scores for syntax (and also a very low n).
Table 10 shows the descriptive data by linguistic category across the measures of noticing. In order to
answer this research question a GEE logit was conducted for the variable linguistic category for the two
measures of noticing.
Results for the heat map measure showed no significant relationship between linguistic category and
noticing. However, there was a significant main effect for stimulated recall. The odds of semantic recasts
resulting in a noticing event in the stimulated recall measure was 2.12 times higher than for
morphological recasts (p = 0.006), 95% (exp) CI (1.24, 3.64). Likewise, the odds of syntactic recasts
resulting in noticing in the stimulated recall measure was 3.73 times higher than for morphological recasts
(p = 0.0002), 95% (exp) CI (1.87, 7.47). No significant relationship was found between semantic recasts
and syntactic recasts. This suggests that when using the stimulated recall methodology to assess learners‘
noticing of recasts, semantic and syntactic recasts were two and three times (respectively) more likely to
be noticed than were morphological recasts. The heat map measure of noticing (though generally in the
same direction) did not show this pattern as clearly.
In order to determine if the measures of noticing differed in how they reflected the extent to which recasts
from each linguistic category were noticed, a series of paired sample t-tests were run (see Tables 11 and
12). For morphology, the data show that the heat map measure was more likely to yield a positive
noticing event, whereas no significant difference was found for semantics. This supports the notion that
Table 11. Percentage of Target Recasts Showing Positive Noticing Events across the Linguistic Category
of Morphology
explicit written SCMC recasts targeting morphology, though attended to at some level, are more difficult
than semantic recasts to notice at any useful level as measured by stimulated recall. Further analysis of
syntactic recasts was not advisable due to the extremely low number of possible paired comparisons.
Table 12. Percentage of Target Recasts Showing Positive Noticing Events Across the Linguistic Category
of Semantics
DISCUSSION
When considered independently from one another, the two measure of noticing employed in this study
appear to be favorable predictors of noticing, with the heat map measure being slightly more
discriminating in its ability to predict the probability of IPT success than stimulated recall. This
difference, however, is a matter of degree. Discriminating in the sense that when the heat map record
indicates noticing (HM+) then the likelihood that the relevant item is scored ―correct‖ (IPT = 1) is over 48
times higher than in the HM- condition. This is in contrast to the 38 times higher for stimulated recall.
The theoretical value of employing eye tracking in explorations of learner attention is that it adds a new
and powerful methodological dimension in exploring constructs associated with attention and noticing
and their respective roles in SLA. Though the current study was not designed to specifically examine
possible levels of awareness with respect to corrective feedback, the modest data presented here might
serve as a springboard for future more in-depth investigations into this area. For example, we may
interpret Table 8 in the following way: Assuming that the eye tracker was properly calibrated, rows 1 and
2 (HM = 0) seem to show no evidence of detection since no significant eye gaze was registered over the
targeted items. We may say, then, that there seems to have been no registration of the stimuli at all. Of
course, there are three cases of IPT success for these items; however, over 90% of these cases led to an
IPT score of zero. In rows 3 and 4 we have evidence of the cognitive registration of the stimuli (Tomlin &
Villa, 1994 at some level (detection). Schmidt (2001) suggests that it is necessary to distinguish between
1. detection without awareness and 2. detection within focal attention accompanied by awareness, that is,
conscious perception or noticing. Making this distinction, as he further points out, is not an easy task. For
example, how can we know whether some stimulus (or a feature of it) has been attended? And how can
we know whether some stimulus (or a feature of it) has been noticed (Schmidt, 2001)? I argue that eye
tracking data may help us make this distinction and we can do this in a pedagogically relevant way. In
terms of the present data, items coded in Table 8 as HM = 1, SR = 0 may be viewed as having been
attended to at some level but not noticed (detected without awareness). Viewed another way, it could be
argued that these items were detected but did not reach a sufficient threshold of rehearsal in short-term
memory (Robinson, 1995). Those items coded as HM = 1, SR = 1 may be viewed as evidence of
detection with awareness (noticed), at least a low level of awareness (see Leow, 2000). In both cases the
IPT scores confirm this view.
The results in terms of which linguistic category of recast was noticed more readily are mixed, and vary
as a function of the measure employed. There was no significant relationship found for linguistic category
and noticing in the heat map condition. For the stimulated recall condition, however, learners were much
more likely to report having noticed semantic recasts than morphological recasts. Syntactic recasts also
hold the same advantage over morphological recasts, though the extremely low numbers of syntactic
recasts in the data make any real claims in this regard problematic. This is consistent with previous
research by Mackey et al. (2000) and Nabei and Swain (2002) who had positive findings for noticing
lexical feedback over other types. This evidence is also compatible with an input processing theoretical
perspective (Van Patten, 2007), which argues that learners prefer to process information lexically over
other features in the input.
CONCLUSION
This study offers a foray into the use of eye tracking technology as a measure of noticing corrective
feedback. Though measuring what learners attend to in the input in any precise way remains a difficult
challenge for researchers, the use of eye gaze data seems to be potentially valuable in helping to
determine which features of the input are likely to be noticed and which are not since we can see precisely
what learners view and arguably attend to. Used in conjunction with stimulated recall, the eye tracking
methodology may help researchers untangle the lower levels of the noticing continuum. This, of course,
remains to be seen. Finally, eye tracking may provide an avenue whereby many of the argued benefits of
text-based SCMC, such as its facility to enhance the salience of input, can be put to the test.
One practical conclusion of the study is that stimulated recall and heat maps are highly (and positively)
correlated with one another and in cases where there is a positive record of both, then the likelihood that
the targeted item has reached a higher level of learner awareness (ultimately confirmed by their IPT
scores for that same item) is extremely high. For SLA researches this is good news given the
pervasiveness of stimulated recall as a methodological technique.
In an email that I will send a little later you will receive a copy of what
you wrote the other day in the English study chat session. I have removed my
half of the dialog so you will only see what YOU wrote.
When you have a minute over the next day or so I would like you to look over
these sentences you wrote. Please do NOT use any outside sources.
Please don‘t open the second email as I am timing how long it takes for each
student to complete this activity based on when they SEND me the completed
email. That is, when you actually open the email I will receive a message
that tells me that you have viewed or opened it. I will give you one minute
to complete each sentence. This means that if I send you 15 sentences, then
you should complete the activity and send it back to me within 15 minutes.
In the second email ―subject line‖ I will write how many sentences you have
to look at. Please finish the activity once you begin since you cannot
return to it later.
Directions:
Simply decide if the sentences you wrote are correct or not as they are
written. That is, are they grammatical as they are written here or are there
one or more errors in the sentence? If the sentence is OK, then simply write
OK under it. If there is an error, please write ―No‖ and then re-write the
sentence and FIX the error. Please and email me this back as soon as you
can.
Thank you!
Here is an example:
Researcher‘s Comment: See the past tense ―-ed‖ added after ―jump‖ from line 1?
Your answer: OK
NOTES
1. However, a failure to notice a target cannot be inferred by a failure to verbalize something.
2. It is possible that these positive findings could be attributed to the narrow discourse contexts often
present in experimental environments. That is to say, recasts in these (largely) experimental studies are
often isolated from other feedback strategies and are frequently provided intensively by the interlocutor.
Oliver (1995), for example, suggests that the interpretation of recasts is greatly influenced by the
discourse context.
3. Though the learner internal variables of proficiency level and working memory capacity are viewed as
important factors in explaining what learners notice in the linguistic input, a discussion of the respective
roles played by these two variables is beyond the scope of this paper.
4. However, some research shows recasts to be effective for other aspects of language as well (Doughty &
Varela, 1998; Sagarra, 2007; Sheen, 2006).
5. A contingent recast in this context referred to recasts that immediately followed the trigger turn and
appeared just below the trigger turn on the screen.
6. However, regressive saccades (i.e., backtracking eye-movements) can act as a measure of processing
difficulty during encoding (Rayner & Pollatsek, 1989). Indeed, between 10–15% of the time readers
perform regressive saccadic movements to return to material that has been already read (Dussias, 2010).
any two observations. This option also minimizes the number of parameters that need to be estimated.
The exchangeable matrix requires only calculation of two variance components. Finally, there is some
evidence from the literature that GEE is robust to the structure of covariance matrix chosen (see Park and
Shin, 1999 for a discussion of the use of working correlation matrices in the GEE approach for
longitudinal data).
18. Cases where a targeted item was recast during the task but where there was no attempted use apparent
from the immediate posttest were omitted from the analysis. All items recast were included in the delayed
posttest.
19. Table 6 reflects the GEE analysis when the variable Heat Map is entered into the statistical model first
and Stimulated Recall second. When Stimulated Recall is entered into the model first, we have a similar
finding (Exp(log(Odds Ratio)) = 9.10; SE 1.60; p < .0001; C.I. of odds ratio 3.62–22.89).
20. Tables 1–8 do not distinguish between semantic, morphological, and syntactic targets. Rather they
reflect the number of recasts in general. Data for the linguistic category of the recasts is reflected in
Tables 9–12.
21. There were a few ambiguous cases in this category, which were removed from the data for this
analysis (n = 5). For example, in one case a possessive pronoun was omitted in an obligatory case (its).
The recast provided the required pronoun, in the IPT, however, the learner used the masculine possessive
pronoun his rather than the neuter possessive pronoun. Though this could be viewed as a positive change,
the threshold for scoring an item correct on the posttests was a target-like change that reflected the exact
language from the recast.
REFERENCES
Alanen, R. (1995). Input enhancement and rule presentation in second language acquisition. In R.
Schmidt (Ed.), Attention and awareness in foreign language learning and teaching (pp. 259–302).
Honolulu: University of Hawai‗i Press.
Ammar, A., & Spada, N. (2006). One size fits all? Recasts, prompts and L2 learning. Studies in Second
Language Acquisition, 28, 543–574.
Braidi, S. M. (2002). Reexamining the role of recasts in native-speaker/nonnative-speaker interactions.
Language Learning, 52(1), 1–42.
Brown, R., & U. Bellugi (1964). Three processes in the child's acquisition of syntax. Harvard
Educational Review, 34, 133–151.
Carroll, S., & Swain, M. (1993). Explicit and implicit negative feedback: An empirical study of the
learning of linguistic generalizations. Studies in Second Language Acquisition, 15, 357–386.
Chaudron, C. (1977). A descriptive model of discourse in the corrective treatment of learners‘ errors.
Language Learning, 27, 29–46.
Dasse-Askildson, V. (2008). How learners‘ affective variables impact their perception of recasts in the
acquisition of grammatical gender in L2 French. Arizona Working Papers on SLA and Teaching, 15, 1–
35. Retrieved from http://w3.coh.arizona.edu/awp/AWP15/AWP15%5BDasse%5D.pdf
Doughty, C. (1994). Fine-tuning of feedback by competent speakers to language learners. In J. Alatis
(Ed.), Strategic interaction and language acquisition: Theory, practice, and research. GURT 1993 (pp.
96–108). Washington: Georgetown University Press.
Doughty, C., & Varela, E. (1998). Communicative focus on form. In C. Doughty & J. Williams (Eds.),
Focus on form in classroom second language acquisition (pp. 114–138). New York: Cambridge
University Press.
Dussias, P. (2010). Uses of eye-tracking data in second language sentence processing research. Annual
Review of Applied Linguistics, 30, 149–166.
Egi, T. (2007). Interpreting recasts as linguistic evidence: The roles of length and degree of change.
Studies in Second Language Acquisition, 29, 511–537.
Egi, T. (2010). Uptake, Modified Output, and Learner Perceptions of Recasts: Learner Responses as
Language Awareness. The Modern Language Journal, 94, 1–21.
Ellis, R., & Yuan, F. (2004). The effects of planning on fluency, complexity and accuracy in second
language narrative writing. Studies in Second Language Acquisition, 26, 59–84.
Ellis, R., & Sheen, Y. (2006). Reexamining the role of recasts in second language acquisition. Studies in
Second Language Acquisition, 28, 575–600.
Ellis, R., Basturkmen, H., & Loewen, S. (2001). Learner uptake in communicative ESL lessons.
Language Learning, 51, 281–318.
Ellis, R., Loewen, S., & Erlam, R. (2006). Implicit and explicit corrective feedback and the acquisition of
L2 grammar. Studies in Second Language Acquisition, 28, 339–368.
Færch, C., & Kasper, G. (Eds). (1987). Introspection in second language research. Clevedon, UK:
Multilingual Matters.
Fanselow, J. F. (1977). The treatment of error in oral work. Foreign Language Annals, 10, 583–593.
Gass, S. M. (1988). Integrating research areas: A framework for second language studies. Applied
Linguistics, 9, 198–217.
Gass, S. M. (1997). Input, interaction, and the second language learner. Mahwah, NJ: Erlbaum.
Gass, S. M., & Mackey, A. (2000). Stimulated recall methodology in second language research.
Mahwah, NJ: Lawrence Erlbaum Associates.
Gass, S. M., & Varonis, E. M. (1994). Input, interaction, and second language production. Studies in
Second Language Acquisition, 16, 283–302.
Godfroid, A., Housen, A., & Boers, F. (2010). A procedure for testing the noticing hypothesis in the
context of vocabulary acquisition. In M. Pütz & L. Sicola (Eds.), Cognitive processing in second
language acquisition (pp. 169–197). Amsterdam: John Benjamins.
Goo, J., & Mackey, A. (in press). The case against the case against recasts. Studies in Second Language
Acquisition.
Izumi, S. (2003). Comprehension and production processes in second language learning: In search of the
psycholinguistic rationale of the output hypothesis. Applied Linguistics, 24, 168–196.
Jacob, R. J. K., & Karn, K. S. (2003). Eye tracking in Human-Computer Interaction and usability
research: Ready to deliver the promises, In J. Hyönä, R. Radach, & H. Deubel (Eds.), The mind’s eye:
Cognitive and applied aspects of eye movement research (pp. 573–605). Amsterdam: Elsevier.
Just, M. A., & Carpenter, P. A. (1976). Eye fixations and cognitive processes. Cognitive Psychology, 8,
441–480.
Just, M. A., & Carpenter, P. A. (1980). A theory of reading: From eye fixations to comprehension.
Psychological Review, 87, 329–354.
Kim, J. H., & Han, Z-H. (2007). Recasts in communicative EFL classes: Do teacher intent and learner
interpretation overlap? In A. Mackey (Ed.), Conversational Interaction in second language acquisition: A
series of empirical studies (pp. 269–297). Oxford: Oxford University Press.
Komínková, B. (2008). Comparison of two eye tracking devices used on printed images. Unpublished
master‘s thesis. Department of Graphic Arts and Photophysics, University of Pardubice, Poland.
Kotrlik, J. W., & Williams, H. A. (2003). The incorporation of effect size in information technology,
learning, and performance research. Information Technology, Learning, and Performance Journal, 21(1),
1–7.
Kuhn, J. (2012). The noticing of correct and incorrect forms in lengthier texts: An ESL eye-tracking
investigation. In G. Kessler, A. Oscoz, & I. Elola (Eds.), Technology across writing contexts and tasks
(pp. 237–253). San Marcos, TX: CALICO.
Lai, C., Fei, F., & Roots, R. (2008). The Contingency of Recasts and Noticing. CALICO Journal, 26, 70–
90.
Leeman, J. (2003). Recasts and second language development: Beyond negative evidence. Studies in
Second Language Acquisition, 25, 37–63.
Leow, R. P. (1997). Attention, awareness, and foreign language behavior. Language Learning, 47, 467–
506.
Leow, R. P. (2000). A study of the role of awareness in foreign language behavior: Aware vs. unaware
learners. Studies in Second Language Acquisition, 22, 557–584.
Li, S. (2010). The effectiveness of corrective feedback in SLA: A meta-analysis. Language Learning, 60,
309–365.
Liu, Y. (2008). The effects of error feedback in second language writing. Arizona Working Papers in SLA
& Teaching, 15, 65–79.
Loewen, S., & Erlam, R. (2006). Corrective feedback in the chatroom: An experimental study. Computer
Assisted Language Learning, 19, 1–14.
Loewen, S., & Philp, J. (2006). Recasts in the adult English L2 classroom: Characteristics, explicitness,
and effectiveness. Modern Language Journal, 90, 536–556.
Long, M. H., Inagaki, S., & Ortega, L. (1998). The role of implicit negative feedback in SLA: Models and
recasts in Japanese and Spanish. The Modern Language Journal, 82, 357–371.
Long, M. H. (1996). The role of the linguistic environment in second language acquisition. In W. C.
Ritchie & T. K. Bhatia (Eds.), Handbook of second language acquisition (pp. 413–468). San Diego, CA:
Academic Press.
Long, M. H. (2007). Problems in SLA. Mahwah, NJ: Erlbaum.
Lyster, R. (1998). Negotiation of form, recasts, and explicit correction in relation to error types and
learner repair in immersion classrooms. Language Learning, 48, 183–218.
Lyster, R. (2004). Differential effects of prompts and recasts in form-focused instruction. Studies in
Second Language Acquisition, 26, 399–432.
Lyster, R., & Ranta, L. (1997). Corrective feedback and learner uptake: Negotiation of form in
communicative classrooms. Studies in Second Language Acquisition, 19, 37–66.
Lyster, R., & Saito, K. (2010). Oral feedback in classroom SLA: A meta-analysis. Studies in Second
Language Acquisition, 32, 265–302.
Mackey, A. (2006). Feedback, noticing and instructed second language learning. Applied Linguistics, 27,
405–430.
Mackey, A., Gass, S., & McDonough, K. (2000). How do learners perceive interactional feedback?
Studies in Second Language Acquisition, 22, 471–497.
Mackey, A., & Goo, J. (2007). Interaction research in SLA: A meta-analysis and research synthesis. In A.
Mackey (Ed.), Conversational interaction in second language acquisition: A collection of empirical
studies (pp. 407–451). New York: Oxford University Press.
Mackey, A., & Philp, J. (1998). Conversational interaction and second language development: Recasts,
responses, and red herrings? The Modern Language Journal, 82, 338–356.
Mackey, A., Philp, J., Egi, T., Fujii, A., & Tatsumi, T. (2002). Individual differences in working memory,
noticing of interactional feedback and L2 development In P. Robinson (Ed.), Individual differences and
instructed language learning (pp. 181–208). Philadelphia: John Benjamins.
McDonough, K., & Mackey, A. (2006). Responses to recasts: Repetitions, primed production, and
linguistic development. Language Learning, 56, 693–720.
Morris, F. A., & Tarone, E. E. (2003). Impact of classroom dynamics on the effectiveness of recasts in
second language acquisition. Language Learning, 53, 325–368.
Nabei, T., & Swain, M. (2002). Learner awareness of recasts in classroom interaction: A case study of an
adult EFL student's second language learning. Language Awareness, 11, 43–62.
Nicholas, H., Lightbown, P. M., & Spada, N. (2001). Recasts as feedback to language learners. Language
Learning, 51, 719–758.
O‘Rourke, B. (2008). The other C in CMC: What alternative data sources can tell us about textbased
synchronous computer mediated communication and language learning. Computer Assisted Language
Learning, 21, 227–251.
O‘Rourke, B. (in press). Using eye tracking to investigate gaze behaviour in synchronous computer-
mediated communication for language learning. In M. Dooly & R. O‘Dowd (Eds.), Researching online
interaction and exchange in foreign language education: Methods and issues. Frankfurt am Main: Peter
Lang.
Ohta, A. (2000). Rethinking recasts: A learner-centered examination of corrective feedback in the
Japanese classroom. In J. K. Hall & L. Verplaetse (Eds.), The construction of second and foreign
language learning through classroom interaction (pp. 47–71). Mahwah, NJ: Erlbaum.
Oliver, R. (1995). Negative feedback in child NS-NNS conversation. Studies in Second Language
Acquisition, 17, 459–481.
Oliver, R., & Mackey, A. (2003). Interactional context and feedback in child ESL classrooms. Modern
Language Journal, 87(4), 519–533.
Ortega, L., & Long, M. H. (1997). The effects of models and recasts on the acquisition of object
topicalization and adverb placement by adult learners of Spanish. Spanish Applied Linguistics, 1, 65–86.
Park, T., & Shin, D-Y. (1999). On the use of working correlation matrices in the GEE approach for
longitudinal data. Communications in Statistics - Simulation and Computation, 28, 1011–1029.
Pellettieri, J. (1999). Why-talk? Investigating the role of task-based interaction through synchronous
network-based communication among classroom learners of Spanish. Unpublished doctoral dissertation.
College of Letters and Sciences, Department of Spanish, University of California, Davis.
Philp, J. (1999). Interaction, noticing, and second language acquisition: An examination of learners’
noticing of recasts in task-based interaction. Unpublished doctoral dissertation. School of Letters, Arts,
and Media, Department of Linguistics, University of Tasmania, Australia.
Philp, J. (2003). Constraints on ―noticing the gap‖: Nonnative speakers‘ noticing of recasts in NS-NNS
interaction. Studies in Second Language Acquisition, 25, 99–126.
Poole, A. & Ball, L. J. (2006). Eye Tracking in Human-Computer Interaction and Usability Research:
Current Status and Future Prospects. In C. Ghaoui, (Ed.). Encyclopedia of Human Computer Interaction
(pp. 211–219). Hershey, PA: Idea Group
Poole, A., Ball, L. J., & Phillips, P. (2004). In search of salience: A response time and eye movement
analysis of bookmark recognition. In S. Fincher, P. Markopolous, D. Moore, & R. Ruddle (Eds.), People
and Computers XVIII-Design for Life: Proceedings of HCI 2004. London: Springer-Verlag Ltd.
Rayner, K. (1998). Eye movements in reading and information processing: 20 years of research.
Psychological Bulletin, 12, 372–422.
Rayner, K. (2009). Eye movements and attention in reading, scene perception, and visual search.
Quarterly Journal of Experimental Psychology, in press.
Rayner, K., & Pollatsek, A. (1989). The psychology of reading. Hillsdale, NJ: Erlbaum.
Révész, A. (2009). Task complexity, focus on form, and second language development. Studies in Second
Language Acquisition, 31, 437–470.
Robinson, P. (1995). Attention, memory, and the noticing hypothesis. Language Learning, 45, 283–331.
Robinson, P. (1996). Learning simple and complex second language rules under implicit, incidental, rule-
search and instructed conditions. Studies in Second Language Acquisition, 18, 27–67.
Rosa, E. M., & O‘Neill, M. (1999). Explicitness, intake, and the issue of awareness: Another piece to the
puzzle. Studies in Second Language Acquisition, 21, 511–556.
Sachs, R., & Suh, B-R. (2007). Textually enhanced recasts, learner awareness, and L2 outcomes in
synchronous computer-mediated interaction. In A. Mackey (Ed.), Conversational interaction in second
language acquisition: A collection of empirical studies (pp. 197–227). Oxford: Oxford University Press.
Sagarra, N. (2007). From CALL to face-to-face interaction: The effect of computer-delivered recasts and
working memory on L2 development. In A. Mackey (Ed.), Conversational interaction in second
language acquisition: A series of empirical studies. (pp. 229–248). Oxford: Oxford University Press.
Salaberry, M. R. (2000). L2 morphosyntactic development in text based computer mediated
communication. Computer Assisted Language Learning, 13, 5–27.
Sauro, S. (2009). Computer-mediated corrective feedback and the development of second language
grammar. Language Learning & Technology, 13(1), 96–120. Retrieved from http://llt.msu.edu/
vol13num1/sauro.pdf
Sauro, S., & Smith, B. (2010). Investigating L2 Performance in Text Chat. Applied Linguistics 31(4),
554–577.
Schmidt, R. (1990). The role of consciousness in second language learning. Applied Linguistics, 11, 129–
158.
Schmidt, R. (1995). Consciousness and foreign language learning: A tutorial on the role of attention and
awareness in learning. In R.W. Schmidt (Ed.), Attention and awareness in foreign language learning (pp.
1–63). Honolulu, HI: University of Hawai‗i, Second Language Teaching and Curriculum Center.
Schmidt, R. (2001). Attention. In P. Robinson (Ed.), Cognition and second language instruction (pp. 3–
32). New York: Cambridge University Press.
Seedhouse, P. (1997). The case of the missing ―no‖: The relationship between pedagogy and interaction.
Language Learning, 47, 547–583.
Sheen, Y. (2004). Corrective feedback and learner uptake in communicative classrooms across
instructional settings. Language Teaching Research, 8, 263–300.
Sheen, Y. (2006). Exploring the relationship between characteristics of recasts and learner uptake.
Language Teaching Research, 10, 361–392.
Sheen, Y. (2008). Recasts, language anxiety, modified output and L2 learning. Language Learning, 58,
835–874.
Skehan, P., & Foster, P. (2005). Strategic and on-line planning: The influence of surprise information and
task time on second language performance. In R. Ellis (Ed.), Planning and task performance in a second
language (pp. 193–216). Mahwah, NJ: Lawrence Erlbaum Associates.
Smith, B. (2004). Computer-mediated negotiated interaction and lexical acquisition. Studies in Second
Language Acquisition, 26, 365–398.
Smith, B. (2005). The Relationship between Negotiated Interaction, Learner uptake, and lexical
acquisition in task-based computer-mediated communication. TESOL Quarterly, 39, 33–58.
Smith, B. (2008). Methodological Hurdles in capturing CMC data: The case of the missing self-repair.
Language Learning & Technology, 12(1), 85–103. Retrieved from: http://llt.msu.edu/vol12num1/smith/
default.html
Smith, B. (2010). Employing eye-tracking technology in researching the effectiveness of recasts in CMC.
In Francis M. Hult (ed.), Directions and prospects for educational linguistics (pp. 79–98). New York:
Springer.
Smith, B., & Renaud, C. (in press). Eye tracking as a measure of noticing corrective feedback in
computer-mediated instructor-student foreign language conferences. In K. McDonough and A. Mackey
(Eds.), Interaction in diverse educational settings. Philadelphia: John Benjamins.
Swain, M. (2006). Languaging, agency and collaboration in advanced second language learning. In H.
Byrnes (Ed.), Advanced language learning: The contributions of Halliday and Vygotsky (pp. 95-108).
London, UK: Continuum.
Tarone, E., & Bigelow, M. (2007). Alphabetic print literacy and processing of oral corrective feedback in
L2 interaction. In A. Mackey (Ed.), Conversational interaction in second language acquisition: A series
of empirical studies (pp. 101–121). Oxford: Oxford University Press.
Tomlin, R. S., & Villa, V. (1994). Attention in cognitive science and SLA. Studies in Second Language
Acquisition, 16, 183–203.
Trofimovich, P., Ammar, A., & Gatbonton, E. (2007). How effective are recasts? The role of attention,
memory, and analytical ability. In A. Mackey (Ed.), Conversational interaction in second language
acquisition: A series of empirical studies (pp.144–171). Oxford: Oxford University Press.
Truscott, J., & Sharwood Smith, M. (2011). Input, Intake, and Consciousness: The Question for a
Theoretical Foundation. Studies in Second Language Acquisition, 33, 497–528.
VanPatten, B. (2007). Input processing in adult second language acquisition. In B. VanPatten & J.
Williams (Eds.) Theories in second language acquisition: An introduction (pp. 115–135). Mahwah, NJ:
Erlbaum.
Williams, J. (1999). Memory, Attention, and Inductive Learning. Studies in Second Language
Acquisition, 21, 1–48.
Yuan, F., & Ellis, R. (2003). The effects of pre-task planning and on-line planning on fluency, complexity
and accuracy in L2 oral production. Applied Linguistics, 24, 1–28.
INTRODUCTION
It has been argued that the use of Web 2.0 tools and environments such as wikis, blogs, and forums in the
language classroom not only offers unlimited authentic sources and target language speaker interaction
(e.g., Guth & Helm, 2010; O‘Dowd, 2007; Pegrum, 2009; Richardson, 2006) but can also increase learner
autonomy (Hampel & Hauck, 2006). Drawing on Kress (2000), Hampel and Hauck suggest that online
environments can be conceptualized as ―packaged resource kit[s],‖ language learners as ―agents‖ or
―designers‖ and the learning process as ―a process of design‖ (2006, p. 11). Still, these tools and
environments put new demands on teachers and learners in terms of multimodal communicative
competence (Royce, 2002) and multiliteracies (Cope & Kalantzis, 2000; Kress, 2003). Consequently,
language teachers have to become agents and designers, and thus must acquire the necessary skills and
competences themselves before they can support learner autonomy in Web 2.0 contexts as understood by
Palfreyman (2006). One particularly promising field is ―telecollaboration 2.0‖ (Guth & Helm, 2010),
which stresses dialogue building and the use of commonly available social networking tools,
encompassing the development of language proficiency, intercultural communicative competence, and
multiliteracies.
In the following, we will discuss findings from two case studies which involved four groups of
telecollaborative partners from the Open University, UK (OU), Teachers College, Columbia University,
USA (TC), the Pädagogische Hochschule Heidelberg, Germany (PHH), and the College of Foreign
Languages Czestochowa, Poland (WSL). The aim of both case studies (2008–2009 and 2009–2010) was
twofold: to raise participants‘ multimodal awareness through a task-based approach, by asking them to
identify online communication modes and how they support or constrain meaning-making, and to guide
characterized as a process of design, in which the degree of multimodal communicative competence and
the degree of learner control and thus autonomy—or, as Kress claims: ―agency of a real kind‖ (2000, p.
340)—are likely to be interdependent (Hampel & Hauck, 2006). Kress describes this interrelationship as
follows:
[…] the work of design: the intentional deployment of resources in specific configurations to
implement the purpose of the designers. […] The work of the text maker is taken as
transformative of the resources and of the maker of the text. It gives agency of a real kind to the
text maker. (2000, p. 340)
Consequently, becoming aware of constraints and possibilities in terms of online modes and meaning-
making and engaging in systematic development of multiliteracy skills can potentially increase learner
autonomy understood—in line with Palfreyman (2006)—as the informed use of multimodal
environments. The autonomous learner, then, is the one who—in line with Kress—can ―choose, not
merely with full competence within one mode but with full awareness of the affordances of many modes
and of the media and their sites of appearance‖ (2003, p. 49). This involves critical use of modes
(competence in code and mode switching and familiarity with new codes, such as online speech, online
writing, and image), dealing with affective demands (engagement with unfamiliar tools leading to
affective challenges such as language anxiety and cognitive overload) and intercultural differences (the
fact that modes, making meaning, and communicating online are influenced by cultural conventions)
(Hampel & Hauck, 2006). Meeting these challenges requires the training of teachers in the design of
activities that make efficient use of multiple modalities to ensure that learners stretch and adapt all
elements available to them in any given online environment (Hampel & Hauck, 2006). Such training can
potentially contribute to what Fuchs (2006), drawing on Willis (2001), calls professional literacy, and
would allow them to systematically work up their way on Hampel and Stickler‘s (2005) ―pyramid of
skills‖ for tutoring online.
Competence Models and Teacher Education
For Hampel and Stickler (2005), the first three levels of their ―pyramid of skills‖
(http://www.llas.ac.uk/resources/gpg/2530) for teaching in online or blended contexts cover basic ICT
competences, specific technical competences when using bespoke applications, and the skill of dealing
with constraints and possibilities of the medium corresponding to the aforementioned stages of computer,
electronic, and multi-literacy. Level three is particularly relevant to the case studies here as ―teachers need
to be able to harness the potential of the medium for language learning‖ while having to deal with
learners‘ emotional responses to any given online environment and having to make ―students aware that
certain things cannot be done in a particular environment but that it is possible to make up for this in other
ways‖ (Hampel, 2009, pp. 39–40). At levels 4 and 5, teachers should be able to turn ―an online
environment into a platform where communicative competence can be developed‖ which ―is dependent
on a sense of community and trust‖ (Hampel & Stickler, 2005, p. 318). Levels 6 and 7 of the skills
pyramid relate to teachers‘ creative skills and their ability to choose the right tool for the job and develop
their own teaching style. The discussion of our findings refers to the first five levels. The following
section describes the participants, project phases, and task and research designs.
RESEARCH DESIGN
Participants
Tables 1 and 2 show the participants in case studies 1 and 2.
Institution
Number of 26 21 13 18
participants
(n = 78)
Participant Pre-service teachers Pre-service teachers Language learners Language learners
Description (master degree (state exam EFL (German B1/B2) (German B2)
candidates, candidates,
TESOL/Applied primary/secondary
Linguistics program) school)
Nature of TESOL, classroom Task-based language German language German language
Course practices learning and media course course
literacy
Modality FTF FTF FTF online only
Institution
Number of 53 20 18 10
participants
(n = 101)
Participant Pre-service Pre-service teacher Pre- and in-service In-service teacher
Description teacher (master degree (state exam EFL teacher (extramural) (EFL, ESL, EAP)
candidates – candidates –
TESOL/Applied primary/secondary
Linguistics program) school)
Nature of TESOL- classroom Task-based language ICT in ELT (blended) Staff development
Course practices learning and media (online only)
(FTF) literacy
(FTF)
online tutoring skills and e-literacy skills development (Hampel & Hauck, 2006), and was inspired by
Hoven‘s (2006) experiential modelling: participants were exploring modes available online and their
impact on meaning-making and communication by engaging in hands-on analysis of web resources and
social networking tools. Moreover, we adapted O‘Dowd and Ware‘s categories of tasks for
telecollaborative projects, which increase in levels of complexity from monologic information exchange,
to comparison and analysis, to collaborative task (2009, pp. 175–178). Specific examples are outlined
below.
Case Study 1
Teacher trainees in Germany and in the United States worked both in local and in telecollaborative groups
while the language learners at OU and WSL worked in telecollaborative groups only. Task 1 (Appendix
A) was inspired by Lamy and Hampel (2007), who suggest that first the modes involved in making up a
multimodal environment should be identified and then the meaning-making and communication
possibilities they afford the learner—both as single and as combined modes—should be considered.
Furthermore, Task 1 was informed by Halliday‘s (1989) social-semiotic framework and Kress and van
Leeuwen‘s (2001) understanding of multimodality, which includes the affordances of modes for making
meaning through the new media. This comparison and analysis task required learners to exchange
information, but also to go a step further and carry out comparisons or critical analyses of cultural
products. Participants focused on the modes featured on a web resource of their choice and on how these
modes convey information. The aim was to learn about their cross-institutional partners‘ various
backgrounds while becoming increasingly aware of how the information was communicated through the
site. Task 2 (Appendix B), a collaborative task, asked participants to exchange and compare information
and to work together to produce a joint product. The teacher trainees were a week ahead of the language
learners at the OU and WSL. Thus, the former selected and the PHH trainees translated some of the tasks
they had designed in their telecollaborative teams, and the students at OU and WSL carried them out.
Case Study 2
Teachers worked both in local and telecollaborative groups. As only few postings to the Task 1 group
forums in Study 1 (Appendix A) explicitly referred to modes and meaning-making, we reassessed the task
design for Study 2. Learners were provided with a fully worked model of a website analysis and asked to
be more explicit about meaning-making features of the web resource they examined.
Furthermore, we dedicated Task 1 of Study 2 (Appendix C) to multiliteracy skills development in more
general terms based on Pegrum (2009) before focusing on multimodality in Task 2 (Appendix D). Task 3
(Appendix E) then continued with this focus on modes.2 Based on the use of the tool analyzed in Task 1,
participants were asked to design an intercultural learning task which considered their English learners‘
multimodal competence.
analysis of qualitative data and description of events and processes‖ (Benson, 2001, p. 282). This
approach is reflected in the presentation and discussion of our findings and builds on various research
studies on telecollaboration in pre-service teacher education which have also drawn on this research
paradigm (e.g., Belz & Müller-Hartmann, 2003; Fuchs, 2006; O‘Dowd, 2007).
Allwright and Hanks (2009), however, have criticized the action research approach in that it is too
limiting and have suggested a move to exploratory practice or inclusive practitioner research instead:
Third-party research in general cannot meet our purposes, and practitioner research, the form of
AR [action research], has not yet taken us far enough away from the third-party model to
overcome these limitations. […] The first two parties for research on education are the teachers
and the learners. (Allwright & Hanks, 2009, p. 145).
Exploratory research attempts to bridge the teacher-researcher gap and to help teachers counter burnout
by focusing primarily on teachers (although with a recent shift to learners to acknowledge their
centrality), by trying to make teaching more interesting for teachers, and by emphasizing principles over
practice (Allwright & Hanks, 2009). These authors further insist that language learning and teaching and
research are social processes and thus call for learners as ―key practitioners‖ without excluding teachers.
Instead, both should be considered ―‗practitioner colleagues‘ with the teacher playing a collegial role in
helping learners develop as researchers of their own practices and as practitioners of learning‖ (p. 146).
This collegial role was taken on both by the authors of this contribution and the in- and pre-service
teachers who took part in both project cycles with all three groups reflecting on their practice (see
indicators for research questions 1 and 2 below).
The case study approach was chosen as it grasps the complexity of telecollaborative projects. A case
study investigates a single instance or phenomenon in context and focuses primarily on gaining
understanding of a context, (i.e., on the what it is and what it does) and not on generalizing results (see
Nunan, 1992). A case study is primarily ―theory-building‖ or ―data in search of a hypothesis.‖ This means
that ―[…] generalizations and hypotheses emerge during the course of the data collection and
interpretation, rather than being predetermined by the researcher‖ (Nunan, 1992, p. 56). Case studies are
―methodologically speaking […] a ‗hybrid‘ in that almost any data collection and analytical methods can
be used‖ (Nunan & Bailey, 2009, p. 157). They are characterized by the fact that ―a case is a ‗bounded
instance‘ […] whether those boundaries are physical (a certain school site), or temporal‖ (p. 161) such as
in these semester-long telecollaborations. The phenomenon in the case study ―is studied in context,
focusing on observation, description, inference and interpretation, all important facts of ethnographic and
practitioner research‖ (p. 162).
Other case study characteristics are longitudinality, multiple perspectives (by researchers and
participants), and the triangulation of data. The latter involved in both case studies gathering information
through qualitative and descriptive quantitative data from pre- and post-course questionnaires, CMC
transcripts (forum postings, wikis), learner portfolios, and journal entries. In doing so, we attempted to get
multiple viewpoints and a more in-depth understanding of the phenomenon under investigation (e.g.
Nunan, 1992). This is in line with the call for ―more description of the learners, settings, and events in
[CALL] contexts‖ (Huh & Hu, 2005, p. 17); and we especially ―need a better understanding of how
exactly all of these factors interact and operate in real pedagogical contexts‖ (Chambers & Bax, 2006, pp.
466–67; see also Müller-Hartmann & Schocker-v. Ditfurth, 2008; Hampel & Hauck, 2006). This kind of
research can offer
a broad and balanced analysis of the various factors and their interaction, it could have a local
impact, in that it could lead to the better use of CALL in the research settings themselves. It could
also have a wider impact, in that it could illuminate the ways in which these factors could be
The action research cycles helped us understand and adjust our own practice. At the same time, we model
for our learners how reflective pedagogical tools (e.g., journals and portfolios) can serve as tools for
research by engaging them in exploratory practice.
Table 3 shows the data collection instruments for both case studies (see Appendix F for a more detailed
description). Additionally, the CMC data in both studies came from participants‘ posts in the Moodle
forums, blogs, and wikis.
The authors and teacher-researchers at the OU, PHH, and at TC collaborated in designing their data
collection instruments for the first action research cycle (ARC) (case study 1) and in refining instruments
for the second ARC (case study 2). After the second ARC, they each coded their students‘ data with
reference to the research questions (see Introduction). The following indicators (IN) provided the basis for
analyzing the data in terms of learner autonomy.
For research question 1:
Indicator 1.1 Teachers describe the tool‘s multimodal potential
Indicator 1.2 Teachers describe the tool‘s potential for communication and interactivity, i.e. its
meaning-making potential, covering constraints and affordances
Indicator 1.3 Teachers describe the tool‘s potential to support EFL/ESL learner autonomy
The analysis was done by inserting color codes and comments, and by highlighting sections of the data,
(e.g., CMC transcripts, needs analyses, journal entries, questionnaires, portfolios). Next, we cross-
checked our colleagues‘ data sets against our own by coding and commenting further until categories
emerged from the data, which were then discussed in multiple online meetings.
This illustrates Kress‘ (2003) point that authorship is no longer rare, making for greater democracy and a
leveling of authority, or, in Richardson‘s words, ―[b]logs engage readers with ideas and questions and
links. (…) They demand interaction‖ (2006, p. 18). Moreover, ―[a]s students participate, they also take
ownership of the space, and (…) this can lead to a greater sense of participation‖ (p. 28). Hence,
Pegrum‘s (2009, p. 38) call for ―participatory literacy‖ as part of the skillset necessary to become more
autonomous when operating online. While Rita does not refer to autonomy development yet, Colleen,
with regard to the Chinese site, makes the connection between multimodal competence and learner
autonomy (IN 1.3) when she observes:
There is a greater chance to be an active viewer. Especially in the Culture section, there are many
videos that allow the user to click through animated pages, read information on Chinese culture,
and watch video clips. The user is responsible for controlling the pace and the direction the
information will take. […] If the learner has the necessary competence s/he can be more active,
hence more autonomous in choosing and checking out content (F-11/01/08).
Silke stresses though that ―[t]he page itself is not that useful to use in a classroom, because it is way too
overloaded with information, students could click for hours without actually working‖ (F-11/04/08),
pointing out the danger of cognitive overload through hypertext links. Both, thus describe the need for
―hypertext literacy … [the] ability to understand the rhetorical effects of links […] and to respond to their
navigational effects‖ (Pegrum, 2009, p. 38), and demonstrate their competences as to understanding the
affordances and constraints of the medium (Level 3 of the skills pyramid). Participant contributions, like
Colleen‘s observation above, led to our decision to front-load more general literacy skills development in
ARC 2 (Task 1). Next, students analyzed a web resource in terms of modes and potential for meaning-
making (Task 2).
Providing a Rationale for Negotiating and Choosing a Multimodal Website for EFL/ESL Learners
In Task 1.3 participants discuss constraints and affordances and the pedagogical potential of their two
chosen sites and thus focus on how to search information on complex websites. To offset the cognitive
overload, Silke refers to the teacher‘s role of designing tasks to support her learners‘ autonomy (INs 1.3
and 2.1):
[…] What is very good about the page is that they already divided the links into categories. So I
could choose a topic for a session and give the students some of my chosen weblinks. Then they
can work on their own, or in pairs, working with the websites (F-11/04/08).
Rita also focuses on students‘ ―search literacy‖ (Pegrum, 2009, p. 36) to support autonomy development:
―I also see this page working for a group activity, like a web quest, in that the students could do a bit of
research on their own […] without having to go all over the Internet‖ allowing them ―to play with the
computer and the Internet‖ (F-11/10/08). Rita agrees with Silke, especially for students with ―very little
experience using computers or the Internet‖ (F-11/10/08). The quotes support IN 1.3 (the website‘s
potential to support EFL/ESL learners‘ autonomy). INs 2.1 and 2.2 (awareness of the importance of
autonomy and multimodality) are also supported since participants focus on task design to support
autonomy already in this task.
Developing ESL/EFL Learners’ Skills Through a Technology-Based Task
With regard to RQ 2, the rationale of Task 2 was for participants to focus on using technology to enhance
skills development without an explicit emphasis on the importance of multimodality in this process. This
led to an underestimation of the role of multimodality in the task design (IN 2.2). The group chose the
wiki tool in Moodle to negotiate the design of the collaborative writing task. A discussion of tools though
only started when they realized that posting their final task design in the wiki ―might look messy with all
our notes back and forth‖ (Rita, W-11/23/08). Sabrina writes that she is ―really confused by this wiki
thing too,‖ and thinks that ―it would have been better to do this in the forum‖ (W-11/24/08). Susan finally
makes the connection between their own experiences in the wiki trying to write a collaborative task, and
having their learners later do the task in a wiki as well (IN 2.2): ―it seems as if the other group are going
to be using the wiki to complete the activity, so could we have them peer edit/review on the wiki and
publish on the knoll [sic] after they‘re finished?‖ (W-11/24/08). Susan brings up the Knol as an additional
tool, which the TC group had found and posted earlier in the forum (F-11/10/08). Here, they show their
competence as autonomous teachers who want to support their learners‘ autonomy (INs 2.1 and 2.2). This
also becomes evident in their search for a new tool: ―Anyone can create a knoll [sic] article but unlike the
Wikipedia, authors of knolls [sic] can take credit for their writing […]. The important thing here is that
students can take credit for their work and become the author of a written piece of work‖ (TC group, F-
11/10/08) (IN 1.3). This illustrates yet again Kress‘s (2003) observation about authorship and leveling of
authority afforded by the new media.
Hence the group demonstrates skills at level 5 of the pyramid (facilitating communicative competence
through task design); nonetheless, they do not provide a rationale for using the wiki for the drafts and the
Knol for publishing. This led to changes in ARC 2. Still, competence was developed in terms of task
sequencing and task instructions, another crucial aspect of task design. Silke writes:
[W]hat was completely new for me was to set up tasks that should be carried out online or with
the help of online tools like wiki. […] It [The task] had to be carried out in a week or two and so
had to be thought through accurately and of some kind of higher level, so it would take more than
a day to carry it out (Silke, PF entry).
She seems to have developed competences both in terms of task design (task sequencing—project task)
and how tasks can be combined with technology. Silke also reflects on the teacher and the learner role
and compares both of them and she understands the importance of task instructions: ―Sometimes […] we
had problems to figure out what exactly we were supposed to do. Experiences like this made clear to me
how important the formulation of a task is‖ (Silke, PF entry).
The great majority of participant contributions related to Task 1 in case study 1 were concerned with what
the selected sites facilitate teachers and learners in doing and what the challenges were. The number of
postings from the final part of Task 1 (part 3, step 1) which make explicit reference to modes and
meaning was unexpectedly small. To better support multimodal skills development, we re-assessed the
task design in ARC 2 by having teachers look at specific tools to develop e-literacy skills first.
Throughout the task design we put a stronger focus on multimodal competence development. In Task 2,
we provided participants with a fully worked example of a website analysis based on Halliday‘s
framework to model for them exactly what was expected. Teachers had to describe why and how they
would use these websites in their teaching. The aim was to support teachers in gauging the effects of
technological mediation on language teaching and in finding out how the potential of the online
environment can be used to enhance communication and interaction, thereby fostering their own and their
learners‘ autonomy. The frame spelled out in detail our expectations concerning the different modes. Task
3 ensured that teachers chose a tool they had worked on before to enable them to transfer their multimodal
competences from Tasks 1 and 2 to task design in Task 3.
Action Research Cycle: Case Study 2
Here, we present an analysis of the cross-institutional group comprising Megan (OU), Malgorzata (WSL)
Katja, and Ina (PHH). In our analysis of the forum contributions to Task 1, we also consider postings
from other participants as they were able to post in other group forums as well.
Malgorzata has tutored children and hopes that the project will help her become more proficient in using
technology in ELT. Katja has limited teaching experience and wants to learn how to get young language
learners/special needs students motivated through using technology. Ina‘s background is similar to that of
Katja‘s but she has already taken a course on teaching with technology. She feels that she has learned a
lot (about wikis, interactive whiteboards, and blogs) and thinks the project is a unique opportunity for
trying out technology in ELT. Ina also hopes to learn about telecollaboration and looks forward to share
experiences with students from other countries. Megan, an in-service tutor with several years of
experience in FTF and blended ELT settings, has worked with discussion forums (Pre-Q/5,8,9,10). While
none has much experience with social networking tools, Megan has used wikis in educational contexts
(Pre-Q/7), and Katja and Ina have used chat/forums/IM/Skype for private purposes. Malgorzata has used
chat/forums for teaching (Pre-Q/7).
Evaluating Technology Tools and their Potential to Foster Multiliteracy Skills
For RQ 1, Tasks 1.1 and 1.2 required students to analyze a tool of their choice based on Pegrum‘s (2009)
criteria. In Task 1.3, participants commented on other groups‘ results regarding Tasks 1.1 and 1.2.
Except for Malgorzata, who looked at forums, all group members analyzed the wiki. Like other teachers,
Malgorzata used forums mainly to look for resources and she is eager ―to provide knowledge in a way
easily adaptable by children‖ (F-31/10/09). This prompts the following reply by Megan, who uses forums
regularly:
[…] you took a much wider view of a forum than I had considered by considering forums that are
available to all or to bigger groups. At the OU we work on forums and they are usually restricted
either to a small tutor group, a group of tutors or a course which makes it a much different
experience […]. On the forums we use you can add pictures to your texts which also adds a
different dimension (F-2/11/09).
Megan picks up on Malgorzata‘s idea to cater for young learners suggesting that adding pictures to
complement text might be useful. Malgorzata‘s reply provides evidence of her awareness of the medium‘s
potential such as turning a forum into a platform where communicative competence can be developed (IN
1.1):
[I]f I had a chance to [use a forum as a teaching tool] my students would be very interested in the
possibility of working this way, especially one of them. He talks with people from other countries
using chats and is fascinated with vocabulary he can discover […] (F-8/11/09).
In her answer to Task 1, Malgorzata also observes the following (F-31/10/09): ―As forum is based on
writing I think that visual modes of communication such as pictures would facilitate the points. We can
send them via e-mail to chosen forum participants or post as links.‖ Malgorzata uses this point to justify
the need for multimodal literacy skills when working with forums (IN 1.1).
Another group chose to examine the wiki. Markus disagrees with Megan and Angela‘s contribution, in
particular with their ―vote against "technological literacy" in relation to the use of wikis (F-29/10/09). He
concedes that ―you don't need programming literacy like html to create a wiki, but you need to be able to
use text software for example.‖ Yet, he agrees with the remainder of their analysis saying that he ―[…]
can follow [their] idea that wikis support the democratisation of learning processes‖ (IN 2.2).
Nonetheless, Markus stresses that the implementation of wikis in the classroom is still ―hard to realize,‖ a
point which highlights the fact that teacher or learner competence of the new media nor their ability to use
them constructively should not be assumed as a given. Markus‘ post also demonstrates empathy with
learners who are not ready to embrace a more collaborative writing process with non-linear texts that
tools such as wikis facilitate. In her reply, Angela adjusts her initial decision acknowledging that she had
fallen into the trap of forgetting how this tool, or any new tool, can be difficult for those who are
not used to using computers. […] Ideally a wiki should be introduced to students through very
simple tasks to allow them to familiarise themselves with the tool (F-30/10/09). (IN 2.1)
Megan therefore suggests to ―introduce wikis right from the beginning of the course and [to] make the
task non-threatening so that the learners are not worried about making changes on each other's work‖ and
admits that she ―would find it a bit intimidating to change someone else's work‖ (F-30/10/09). Words like
―non-threatening,‖ ―not worried,‖ and ―intimidating‖ clearly indicate an awareness of the affective
demands of social networking tools (IN 2.2) and corroborate Hampel‘s (2009) point about dealing with
learners‘ emotional responses to different online environments (3rd level of the skills pyramid).
As Angela does, Ina and Katja point out ―teachers need to make sure that pupils acquire basic knowledge
about how to use this tool‖ (F-2/11/09; see also Rita in ARC 1). And—like Megan—they also refer to the
fact that the use of wikis, being a text-based tool, should be enhanced by integrating visuals—―uploading
appropriate pictures and diagrams—‖ (F-2/11/09), a fact Megan mentions in relation to forums (see
above) (IN 2.2).
In their answer to Task 1.3, Ina and Katja also observe that ―pupils are mainly working together‖ in a wiki
and that the ―teacher‘s role is to support and guide them, but the emphasis is on collaborative work‖ (F-
29/10/09) which shows their awareness of the importance of developing learners‘ autonomy (IN 2.1). The
group underlines that this tool requires intensive exchanges between users to support ―the process of
drafting, revising and editing‖ and ―to prevent misunderstandings,‖ ―in order to create a meaningful final
product all agree with‖ (F-29/10/09). They clearly see the affordances of a wiki in relation to the joint
production of a text and its meaning-making potential (IN 1.2).
Megan also sees the wiki‘s potential for fostering learner autonomy (Pre-Q/7) through leveling of
authority (Kress 2003), describing it as ―a democratic tool […] to share power‖ where ―the teacher needs
to give it over fully—not have control‖ and where there is a ―focus on the group‖ which will need to
―negotiate‖ content ―through an editing process‖ (INs 2.1 and 2.2). At the same time she highlights a
pedagogical problem, that ―often it ends up more like a blog than a piece of work that is co-authored as
learners just add to it rather than add and edit‖ (F-30/10/09) (IN 1.2).
The examples show that unlike in ARC 1, Task 1 here led to an intensive discussion of the need for
multimodal literacy. This stresses the necessity to develop teachers‘ technical competences (levels 1 and 2
of the skills pyramid). Teachers show skills in terms of the affordances and constraints of the forum and
the wiki. Through the negotiation in the forum, other participants become aware of these skills as well
(level 3 of the skills pyramid). The revised Task 1 also triggers pedagogical discussions that already
provide indicators for RQ 2 (see Megan above).
Evaluating a Website with Focus on Mode and Commenting on the Analysis
Task 2, an awareness raising task, serves as a basis for teachers to choose a website for analysis again
inspired by Halliday‘s social-semiotic framework, with the area of mode spelled out in detail. The group
chooses the Lonely Planet site for Portugal (http://www.lonelyplanet.com/portugal), which is ―mostly in
written and image mode‖ (Megan, F-20/11/09, IN 1.1). The site is linked to a series of blogs, and Megan
concludes that it ―aims at users who like to share their experience and who are willing to explore different
countries as well as cultures.‖ Megan also observes that, thanks to English subtitles and ―many beautiful
and interesting pictures within these videos, it is possible to understand the gist.‖ She is obviously aware
of the complementary function of modes in relation to making meaning, i.e., how visuals and written text
facilitate the reader‘s/viewer‘s understanding of audio input (IN 1.1).
Malgorzata draws the group‘s attention to design features and their affordances:
[The] Logo is simple and easy to remember. It contains […] a circle resembling a globe which
suggest that the website is about travelling. […] The texts are written in legible font and
characteristic words (e.g. places) are highlighted in blue so a guest can directly move to
information connected with a given concept (F-21/11/09).
She analyzes the site‘s various written modes—descriptive text provided by Lonely Planet compared to
blog entries, for example. Malgorzata differentiates between the ―very demanding‖ descriptions,
―connected with a lot of effort to read and understand‖ and ―[t]he blogs or posts of travellers [which] are
written more simple including their feelings and emotions. These texts support the lonely planet texts and
help to understand them better‖ thus making them more understandable for language learners (IN 1.1).
The focus in Task 2 is clearly different from Task 3 because the former does not ask teachers to design an
activity for English learners (ARC 1), but focuses exclusively on further awareness raising of the
multimodal potential of websites. The discussion in ARC 1 though showed that Task 2, in which teachers
consider their future teaching context, can develop awareness of the pedagogical potential of online tools
and support for developing autonomy (ARC 1), thus supporting teachers‘ competence development early
in the project.
Using Technology Tools for Intercultural Task Design for ESL/EFL
Task 2 in ARC 1 was revised as Task 3 and now focuses on designing an intercultural task by using one
of the tools/resources teachers discussed and analyzed in Tasks 1 and 2. This triggered interesting data
since teachers use one of the tools they have analyzed earlier. The task purpose is thus not just
intercultural learning, but also ―developing (an aspect of) multimodal literacy.‖
Unlike in ARC 1, the group discusses the potential and affordances of different tools from the very
beginning since it is a built-in task requirement. Ina considers the blog‘s affordances—asynchronicity
and its multimodal potential—and argues for using the tool for Task 3: ―Blogs require a discussion
beforehand, a draft of the text and an agreement of the final product. Furthermore, it is possible to
integrate (necessary in primary) pictures. Consequently, all 4 [sic] skills are trained as well as different
types of literacy for a real purpose and it helps them to develop and improve in their L2 quickly‖ (F-
6/12/09) (IN 2.2).
In Task 3, the group demonstrates that they have realized the tool‘s potential by integrating various
communication modes such as pictures, video, audio files. Moreover, they make insightful suggestions as
to the level of interactivity and intercultural learning provided for learners (i.e., for secondary learners:
creating cultural pages with factual errors the partner group has to correct by posting a revised version;
for primary learners: learning and performing a Christmas song, video-recording it and uploading it to the
blog so that the other class can learn it). Here, the group shows strong competences on levels 5 of the
skills pyramid, facilitating communicative competence through task design (INs 2.3a and b). The issue of
developing autonomy also comes up in their reflections. Commenting on her future students, Ina says that
she ―will need to be aware about the fact that they need enough freedom to try things out and to reach
their own boundaries to learn by problem-solving and active examination with the respective topic and
participants‖ (PF/7). This is a strong IN for 2.2 and 2.3b, which is further consolidated by the following
statement: ―This project helped me to realise and understand that students will grow within projects like
that because they will have to accept responsibility in this quite autonomous process of interaction and
understand the importance of their own work and contributions‖ (Ina PF/9).
When looking at the task sequence of ARC 2 it becomes clear, as Ina summarizes, that the group has
―developed an idea about teaching with the help of new media mainly through learning by doing‖ (PF/7),
or, in Hoven‘s (2006) words, experiential modeling. Based on the insight gained from the two case
studies presented above, the final section offers some preliminary conclusions.
CONCLUSIONS
In this article, we took the interrelationship between multimodal communicative competence,
multiliteracy skills and autonomy as our starting point to analyze the competencies that (future) language
teachers require to develop first their own and then their learners‘ autonomy in online and blended
settings. Our data suggest that experiential modeling (Hoven, 2006) and exploratory practice (Allwright
& Hanks, 2009) allow language teachers—both as learners and teachers—to find out about modes,
meaning making, and online communication, and help them become familiar with the mediating role of
Web 2.0 tools and environments. This approach can also contribute to learner autonomy as
conceptualized by Palfreyman: the informed use of a range of interacting resources in context (2006). The
ARC results show though that task design in this context should follow a certain sequence: First, tasks
should focus on gaining an understanding of the e-literacy skills required when working with tools such
as forums, wikis, and social bookmarking sites for language learning and teaching purposes. Ideally, this
understanding should enable teachers to provide a rationale for using bespoke tools. Next, tasks should
raise their awareness of a tool's specific affordances, i.e. the constraints and possibilities of the modes
available for meaning making and communication (Hampel & Hauck, 2006). This will allow the teachers
to move to the next level of Hampel and Stickler's (2005) skills pyramid by fostering their multimodal
communicative competence and thus their professional literacy (Willis, 2001). These steps are a
prerequisite for the subsequent phase in which teachers themselves design tasks with the goal of fostering,
in turn, their learners‘ multimodal competence and autonomy since merely equipping learners with
creative and democratic representational online resources will not necessarily result in higher student
control over the learning process or the development of autonomy (Hampel & Hauck, 2006).
This approach should become a learning goal itself both in pre- and in-service teacher training and formal
language instruction. Then, while becoming gradually more versed in multimodality and multiliteracy,
teachers as learners can take more control over and self-direct their learning in online environments
(Benson, 2001), thus becoming more autonomous and gradually gaining the competence to design tasks
that also enhance their learners' autonomy.
http://www.lonelyplanet.com/europe
Evaluate your site of choice by addressing these questions:
1. What is the site about?
2. What activities are provided?
3. How reliable is the website? Justify your answer.
4. What cultural values and beliefs are embedded in the materials?
Tenor:
1. Who is the intended audience of the site?
2. Is the site for individual use/pairs?
3. How interactive is the site?
4. What is the user's status?
Mode:
Analyze the various communication modes/channels available on the website you have chosen:
1. spoken mode
2. written mode
3. image mode
4. gestural mode
Which modes are represented and which functions do they have?
Post the answers to your group Forum.
Post at least one comment on the other groups‘ findings in one of the other forums for each part of the
task.
NOTES
1. In the following teacher trainees and tutors will be subsumed under the term teachers if not otherwise
noted.
2. Tasks 1–3 and their sub-tasks are included in Appendices A–E.
3. Abbreviations for data collection instruments are as follows:
F = Forum (plus date)
JE = Journal Entry (plus entry number plus date)
PF = Portfolio
Pre-Q = Pre-Course Question (plus question number, e.g., ―C‖)
Post-Q = Post-Course Question (plus question number, e.g., ―D‖)
W = Wiki (plus date)
ACKNOWLEDGEMENTS
We would like to express our gratitude to all participating students in this study, and to our colleagues
Dorota Ficek, Gosia Kurek, Ori Livneh, Nils Neubert, and Emmajoy Shulman-Kumin. We are
particularly grateful to Elke StJohn, who was instrumental in setting up the case studies and in
contributing to the analysis of the data.
REFERENCES
Allwright, D., & Hanks, J. (2009). The developing language learner: An introduction to exploratory
practice. Basingstoke, UK: Palgrave Macmillan.
Belz, J. A., & Müller-Hartmann, A. (2003). Teachers negotiating German-American telecollaboration:
Between a rock and an institutional hard place. The Modern Language Journal, 87(1), 71–89.
Benson, P. (2001). Autonomy in language learning. Essex, UK: Pearson Education Limited.
Chambers, A., & Bax, S. (2006). Making CALL Work: Towards Normalisation. System, 34(4), 465–479.
Cope, B., & Kalantzis, M. (for the New London Group) (Eds.). (2000). Multiliteracies. Literacy learning
and the design of social futures. London, UK: Routledge.
Fuchs, C. (2006). Computer-mediated negotiation across borders: German-American collaboration in
language teacher education. Frankfurt am Main, Germany: Peter Lang.
Guth, S., & Helm F. (Eds.). (2010). Telecollaboration 2.0 for language and intercultural learning (pp.
219–248). Bern: Lang.
Halliday, M.A.K. (1989). Part A. In M. A. K. Halliday & R. Hasan (Eds.), Language, context and text:
Aspects of language in a social-semiotic perspective (pp. 55–79). Oxford, UK: Oxford University Press.
Hampel, R. (2009). Training teachers for the multimedia age: developing teacher expertise to enhance
online learner interaction and collaboration. Innovation in Language Learning and Teaching, 3(1), 35–50.
Hampel, R., & Hauck, M. (2006). Computer-mediated language learning: Making meaning in multimodal
virtual learning spaces. The JALT CALL Journal, 2(2), 3–18.
Hampel, R. & Stickler, U. (2005). New skills for new classrooms: Training tutors to teach languages
online. Computer Assisted Language Learning, 18(4), 311–326.
Hoven, D. (2006). Designing for disruption: Remodelling a blended course in technology in (language)
teacher education. Proceedings of the 23rd Annual Ascilite Conference: Who’s learning? Whose
technology? Sydney, Australia: University of Sydney.
Huh, K., & Hu, W. (2005). Criteria for effective CALL research. In: Egbert, J. & G. Petrie G. (Eds.),
Research Perspectives on CALL (pp. 9–21). New York, NY: Mahwah. Laurence Erlbaum Associates.
Kress, G. (2000). Multimodality: Challenges To Thinking About Language. TESOL Quarterly, 34(2),
337–340.
Kress, G. (2003). Literacy in the new media age. London, UK: Routledge.
Kress, G., & van Leeuwen, T. (2001). Multimodal discourse: The modes and media of contemporary
communication. London, UK: Arnold.
Lamy, M.-N. & Hampel, R. (2007). Online communication in language learning and teaching.
Basingstoke, UK: Palgrave Macmillan.
Müller-Hartmann, A., & Schocker-v. Ditfurth, M. (Eds.) (2008). Aufgabenorientiertes Lernen und Lehren
mit Medien: Ansätze, Erfahrungen, Perspektiven in der Fremdsprachendidaktik. Frankfurt, Germany:
Peter Lang.
New Media Consortium (2005). A global imperative: The report of the 21st century literacy summit.
Retrieved from http://www.nmc.org/pdf/Global_Imperative.pdf
Nunan, D. (1992). Research methods in language learning. Cambridge, UK: Cambridge University Press.
Nunan, D., & Bailey, K. M. (2009). Exploring second language classroom research: A comprehensive
INTRODUCTION
In a review article on L2 vocabulary learning, Schmitt notes that many learners ―fail to achieve even
moderate vocabulary learning goals‖ (2008, p. 332). The present study evolved from the recognition of
this all too familiar situation, and from an attempt to develop a vocabulary learning resource which might
help to address it. Though I focus here on just one aspect, namely the use of a narrative framework, the
resource as a whole incorporates several other features based on research findings. I therefore first present
seven principles which lie behind its conception. These principles represent a personal compilation which
overlaps with, but is not identical to other sets (e.g., Hunt & Beglar, 1998).
An experiment is then presented, dealing with the use of a specific type of context, namely a story, within
which target words are presented. When stories are generated by learners, in a technique known as
narrative chaining, they improve performance in short-term and working memory tasks (McNamara &
Scott, 2001) and constitute a powerful mnemonic device for the retention of vocabulary (Bower & Clark,
1969). However, whilst mnemonic devices, such as the keyword method, can be effective (Pressley,
Levin & Delaney, 1982; Avila & Sadoski, 1996; Hulstijn, 1997; Sagarra & Alba, 2006), learners seldom
engage spontaneously in the more complex processing that the method entails compared to rote
memorization (Hulstijn, 1997; Schmitt, 1997; Atay & Ozbulgan, 2007; Barcroft, 2009). The experiment
therefore examines the effect of a teacher-provided—as opposed to learner-generated—story on the recall
of L2 word forms and meanings.
approaches are not necessarily in opposition. Extensive reading helps learners to acquire words through
incidental learning, while various forms of elaborative processing help them to consolidate specific target
words.
Principle 3 is to provide opportunities for learners to engage in deep processing (Craik & Lockhart, 1972).
Regarding L2 vocabulary, this involves at the least the mental effort of considering the semantic aspects
of a word and linking it to other words. Testing the mental effort hypothesis, Hulstijn (1992) found that
word meanings were better recalled when learners had to infer them through a multiple choice question
than when they were given by means of a synonym. In a further development, the involvement load
hypothesis (Hulstijn & Laufer, 2001; Laufer, 2001), proposes that one component of a learner‘s
involvement in processing words is evaluation, in which the word is compared or combined with other
words. Not all studies, however, have validated the hypothesis. Contrary to Hulstijn (1992), Mondria
(2003) found that in intentional learning conditions, inferring word meanings did not lead to better
retention rates than being given these meanings. In a study by Rodriguez and Sadoski (2000), inferring
from context led to good immediate recall but in a delayed test retention fell off considerably. However, a
combination of context and keyword proved to be highly effective. Other studies have found that, when
time on task is taken into account, it may not always be the case that greater cognitive involvement leads
to better retention (Webb, 2005; Keating, 2008). In sum, the findings are mixed: much depends on task
characteristics, the degree of intentionality involved, and the type of comparison being made.
Principle 4 is to provide opportunities for encountering words in context. Here, the concept of context
means the context of a single sentence as opposed to the continuous text advocated in Principle 2.
Although inferring word meanings from this type of context can be unreliable (Laufer & Sim, 1985;
Laufer, 1997; Frantzen 2003; Kaivanpanah & Alavi, 2008), sentences that are informative enough can be
an aid to vocabulary improvement (Webb, 2008), providing a useful intermediary level between words
with translations or definitions and longer texts (Laufer & Shmueli, 1997). The benefit of a context
sentence may lie more in focusing learners‘ attention on a word‘s characteristics than on simply
promoting retention (Seibert, 1930; Mondria & Wit-de Boer, 1991). In this respect, a single sentence may
not always be more effective than a translation (Webb, 2007); however, the more translation equivalents
behave differently in two languages, the more useful context is likely to be.
Principle 5 is to provide multiple encounters with target words. Estimations vary as to how many
encounters are needed for word learning, but for a stable lexical representation to be formed, a reasonable
aim is six to ten encounters (Zahar, Cobb, & Spada, 2001). This can be achieved by providing appropriate
reading texts (Cobb, 2007), but also by ensuring that target words appear in subsequent context sentences.
Principle 6 is to provide translations. Nation (2001, pp. 298–299) summarises a number of studies that
point to the effectiveness of learning words with their translations, either in lists or with flash cards.
Learning words with their L1 translations remains a favoured strategy of L2 learners (Schmitt, 1997) and
given the evidence, it might be wondered why one should bother with any other method. The reason, of
course, is that although information other than simply word meanings may be gleaned from translations
(Webb, 2007), they say less about the use of the word in L2 than L2 contexts do. Again though, there is
no reason why the use of L2 context and L1 translation should be in opposition rather than
complementary.
Principle 7 is to provide multiple aids to processing and recall. Presenting words in an enriched context,
such that different modalities (auditory and visual) or different contextual clues (e.g., translations, L2
sentences, definitions) are processed, means that more cues are available when it comes to retrieving a
memory trace for the purposes of recall. This principle applies as long as the elements present at encoding
are additive, such that the extra processing involved increases the chance of recall rather than leading to
cognitive overload (Chandler & Sweller, 1991; Plass, Chun, Mayer, & Leutner, 2003; Mayer & Moreno,
2003, Diao & Sweller, 2007). Dual modality presentation (auditory and visual) has been shown to be
effective for the recall of isolated L2 words (Nassaji, 2004) and for text comprehension, though reading
only may be just as effective for word learning (Brown, Waring, & Donkaewbua, 2008). Within the visual
modality, pictures can provide a strong cue for recall of associated stimuli (Paivio, 1969; Omaggio,
1979), both in a multimedia setting (Chun & Plass, 1996; Kost, Foss, & Lenzini, 1999; Jones & Plass,
2002; Jones, 2004; Yoshii, 2006; Sydorenko, 2010) and in a classroom (Snyder & Colon, 1988). Some
studies have also found an advantage for video as opposed to still pictures (Hanley, Herron, & Cole,
1995; Al-Seghayer, 2001; but see Chun & Plass, 1996, for an opposite finding). Within the field of
learning words as opposed to glossing text, it has been suggested that pictures are more effective than
translations in establishing links to underlying concepts (Finkbeiner & Nicol, 2003; Comesaña, Perea,
Pineiro, & Fraga, 2009; Barcroft, 2009; Tonzar, Lotto, & Job, 2009; but see Lotto & de Groot, 1998, for a
learning advantage for translations over pictures, and Boers, Piquer Píriz, Stengers, & Eyckmans, 2009,
for circumstances in which pictures were of no benefit).
Finally, in a different vein, but still to be regarded as a means to aid recall, is the use of additional links to
supplement those provided by translations or context sentences. When such links are deliberately
generated by learners, they can be classed as mnemonic devices. However, the material needed can also
be provided by the teacher or CALL designer. The device then becomes part of the presentation context,
its effectiveness depending on how willing or able the learner is to make use of it. A teacher-generated
mnemonic has been shown to be more effective (Hall, Wilson, & Patterson, 1981) in some conditions,
though not consistently so (Pressley, Levin, & Delaney, 1982). Whilst the majority of studies on
mnemonics in vocabulary learning have focused on the keyword technique, narrative chaining, in which
words to be learnt are linked in a story, is the focus of the experiment reported here. However, the aim is
not to study the effectiveness of a learner-generated story as in Bower & Clark (1969), but of one
provided by the teacher. The use of narrative has received very little attention in the research literature,
yet as noted by Bolger & Zapata (2011), stories are widely used in school textbooks. Studying the effect
of such a device is therefore of relevance to course designers.
METHODOLOGY
The Effect of Presenting Words in a Narrative Framework
It has been claimed that narrative is a fundamental aspect of cognition and that ―we see ourselves, the rest
of the world and our life story as a connected chain: a narrative‖ (Greenway, 2008, p.59). Despite its
potential, however, there has been little research into the use of narrative as a device for learning
vocabulary. Bower and Clark (1969) found that when participants constructed a meaningful story around
the words to be remembered, recall was far better than in the control condition of normal study and
rehearsal (93% versus 13%). Bolger and Zapata (2011) examined the effect of presenting semantically
related vocabulary in stories. Grouping words according to their semantic relatedness may be a hindrance
to learning rather than a help (Tinkham, 1993; Waring, 1997; Folse, 2004; Finkbeiner & Nicol, 2003;
Erten & Tekin, 2008), but Bolger and Zapata (2011) found that the contextual support provided by the
stories largely offset the disadvantage. Among the specific aspects of narratives that can facilitate recall
are narrative structure (Horiba, van den Breuk, & Fletcher, 1993) and the number of connections between
the different events within the narrative (Trabasso & van den Breuk, 1985). Unlike the confusion that may
result from semantic relatedness, the establishment of associations at a thematic level (e.g., haunted,
moonlight, yell, and ghost within a horror story framework) can be beneficial for vocabulary learning
(Tinkham, 1997). Therefore the purpose of this experiment was to see whether a teacher-provided
narrative could also help in the recall of L2 vocabulary.
This question was examined within a procedure that along with the use of context sentences (Principle 4)
incorporated several of the features described in the other principles. Firstly, the target words were of a
high enough frequency to be considered useful to know and as such figured in the vocabulary strand of
several of the university‘s English courses (Principle 1). Secondly, the presentation of the sentences was
followed by an association test devised to encourage deeper processing of the target words (Principle 3).
Thirdly, the context sentences containing the target words were accompanied by pictures (Principle 7).
These measures were adopted in order to see whether a positive effect from the story would be obtained
on top of any effect resulting from the presence of these features. As Chun and Plass (1996) have argued,
it is important to examine language acquisition in a realistic environment, where many parameters may be
present. Similarly, Bolger and Zapata (2011) stress the need to examine the relative importance of
different factors in the learning of L2 vocabulary. The purpose of the present experiment, therefore, was
not simply to see whether a story would help learners in recalling word forms and meanings, but to do so
in a situation where other aids to learning were also present. This reflects the combinatory approach set
out in the seven principles. Another, more practical reason for doing this is that a story per se does not
convey word meanings. These can be inferred from context sentences and pictures, but a clearer meaning
of a word is unlikely to be conveyed simply because the sentences are linked in a story. The effect of the
story, if any, would be to facilitate the organisation of information in memory due to the extra linking
opportunities it provides. Although this might help participants to retain word forms, it was necessary to
provide additional information that would allow them to reach an understanding of word meanings. It was
decided to use pictures for this rather than translations. The choice of words within a certain frequency
range meant that some were more abstract than others, providing instances where the meaning to be
inferred was not immediately obvious. As already noted, inferring may not in itself lead to better
retention, but the aim was to see whether there would be any difference between the two conditions when
meanings were not directly given.
The participants were not aware that their recall of the target words would be tested afterwards, but were
told that the aim of the experiment was to test the effectiveness of the presentation conditions (picture
plus sentence) by means of the association test. This raises the question of the extent to which any
learning that occurred was incidental or intentional. Hulstijn (2001) suggests that the distinction between
the two is not always easy to discern, and that what counts most is the quality of the mental processing.
The two modes of learning can be seen as being on a continuum, with at one extreme the incidental
learning that occurs in extensive reading, and at the other, the intentional learning that results from
consciously committing information to memory. Clearly, the nature of the experiment meant that
participants‘ attention was directed towards the target words, and they knew that those words would be
the focus of the association test, so any learning that might occur cannot be incidental in the same way as
it is in studies of vocabulary gains from reading of long texts (e.g., Horst, Cobb, & Meara, 1998).
However the purpose of the experiment, as explained to participants, meant that although they were
incited to understand the meanings of the target words, there was no incitement to commit the form-
meaning link to memory for purposes of production. This did not preclude intentional learning, but
restricted it to the recognition requirements of the immediately subsequent association test. Because the
experiment examined the effect of a narrative framework on initial exposure to target words, rather than
on long-term learning, the aim was thus to avoid the confounding factor of differing rehearsal strategies
that could intervene before a recall test (Laufer & Hulstijn, 2001). The hypothesis was that recall in an
immediate post–test would be better when context sentences were linked in a narrative framework than
when they were unrelated.
Participants
Forty-eight participants performed the experiment in return for a certificate contributing to course credits.
They were in their first or second year of Psychology at a French university. There were 42 female and 6
male participants, the average age being 20.6. They had been studying English for 8.5 years on average.
They were randomly assigned to one of the two conditions (sentences connected in a story or sentences
unrelated to each other). Their proficiency level, assessed by an in-house multiple choice exam, was
approximately B1 on the Common European Framework of Reference. No participant had spent more
than a month in an English-speaking country.
Materials
Sixteen semantically unrelated words were chosen, ranging in frequency rank from 1834 to 4159 in the
lemmatised list of the 6318 most frequent words in the British National Corpus (Kilgariff, 1996), with a
form frequency ranging from 15.6 per million to 50.1 per million (mean = 22.8, SD = 10.3). Two context
sentences were composed for each word, one of the sentences being inserted into a narrative framework
involving a character who lives in Japan (story condition), the other bearing no relation to other sentences
(unrelated condition). The context sentences were similar in both conditions in terms of length (mean
length: story condition = 14.8 words, SD = 2.4; unrelated condition =13.6 words, SD = 2.0), lexical
density (mean number of content words other than target words: story condition = 4.4, SD = 1.2;
unrelated condition = 4.9, SD = 1.0) and syntactic complexity (mean number of clauses: story
condition = 2.7, SD = 0.9; unrelated condition = 2.3, SD = 0.6). The critical difference was that in the
story condition they contained references to the characters and the setting. Two native English speakers
judged the informativeness of the context sentences using Webb‘s (2008) four-scale classification. In all
but two cases, sentences fell into the intermediate categories, whereby participants may gain partial
knowledge of the target word‘s meaning (inter-rater reliability = .78). To complement the sentences, a
picture was also provided in both conditions. For concrete words, the pictures gave a clear indication of
meaning, but for abstract words they provided additional support for understanding and for retrieval from
memory. All but five pictures were identical in the two conditions. Where pictures differed, it was in
order to exclude any pictorial reference to Japan in the unrelated condition (see Figure 1 and Figure 2 for
examples).
Between the two conditions, none of the pictures differed completely in terms of composition. The
informativeness of the sentence plus picture combination was assessed by a separate group of 24 French
learners of English who were given the combinations with the target word missing in the sentences and
asked to provide words in French that could fill the gap (Taylor, 1953). Answers were scored
0 (incorrect), 0.5 (partially correct) or 1 (correct) by two English speakers. Inter-rater reliability was .83,
with differences resolved by discussion. Overall scores ranged from 0.13 (for the target word mild) to
0.96 (for sword), with mean scores showing almost no difference between the two conditions (mean
score: story condition = 0.51, SD = 0.29; unrelated condition = 0.49, SD = 0.29; t = 0.51, p = 0.6).
For each target word, an association test was devised, the task being to indicate which of two words
presented along with the target word was closest in meaning to it. The choice was binary rather than
multiple in order to keep the test as similar as possible between the two conditions. Although this meant
participants had a good chance of guessing the correct answer, the test in itself was not the focus of the
experiment and it was felt that the extra processing entailed by a multiple choice test might have led to
extraneous differences between the two conditions. A feedback sentence was created for each item,
comprising both the target word and the correct response to the question. For example, for the target word
thrust, the choice was between (A) push and (B) carry, and the feedback sentence was The gangster
pushed him back and then thrust him against the wall. The complete list of stimuli appears in the
Appendix.
Procedure
Participants were informed that the experiment comprised two phases, one on paper and one on screen.
They were then given the pre-test. In the pen and paper pre-test, participants were asked to provide a
French translation for the English equivalents of 40 words in which the 16 target words were embedded.
Participants were asked to indicate whether any of the words were familiar, and if possible to provide a
translation in French. This procedure thus encompassed levels 1 to 4 of the Vocabulary Knowledge Scale
(Paribakht & Wesche, 1997). Two raters scored the answers independently, the inter-rater reliability being
0.92, with differences resolved through discussion. Words were scored 1 for a correct translation and 0
when classified as unknown. Partially correct translations or words declared familiar but without
translations were scored 0.5. As pointed out by Waring (2002), there is a chance that learners‘ knowledge
could be overestimated this way: for example, some participants noted that they knew the word thrust, but
they gave the translation as confiance (trust), clearly a synform error (Laufer, 1997). Those who said they
were familiar with thrust but gave no translation might have been doing the same. However, not to
include the dimension of familiarity might mean participants‘ knowledge is underestimated, since the
reactivation of lexical representations, however weak, differs from the establishment of new ones. Since
there is no way of knowing if a synform error has occurred when translations are not given, scores were
therefore also calculated without the familiarity level. Pre-test scores were naturally lower but uniformly
so in both conditions (a mean loss of 0.69 in the story condition versus 0.66 in the sentence condition).
The more conservative estimate of post-test gains was retained for purposes of analysis.
Scores at pre-test were not significantly different (F(1, 47 = .17, p = .69, η² =.006) between participants
who subsequently saw words in the story condition (mean = 1.51, SD = 1.15) and those who saw the same
words in the unrelated condition (mean = 1.6, SD = 1.33). After the pre-test, participants were placed in
front of individual computers and informed that they would see 16 words from the list in a PowerPoint
display in which each word was presented in a context sentence accompanied by a picture. They were told
that the words would be presented in two blocks of eight, and after each block they would be given an
association test. This test was presented on the screen but participants answered on paper by circling A or
B for each item. They were instructed not to write anything else. They were not informed about any
further phase to the experiment, nor that some of the sentences formed a story.
The material was counterbalanced such that participants saw all 16 words, half in the story condition and
half in the unrelated condition. The presentation order was also counterbalanced, with half of the
participants seeing the story condition followed by the unrelated condition and the other half seeing the
blocks in the opposite order. The order of words within blocks was not randomised since the story
condition imposed a fixed order. Each slide was presented for 15 seconds. Afterwards, participants filled
in a sheet with biographical details for approximately three minutes. They were then informed of a third
phase. Here they were asked to write on paper as many target words as possible in whatever order they
recalled them. Participants were given four minutes for this. The sheet was removed and they were given
another and asked to write down the target words again, but this time to provide wherever possible the L1
translations. This procedure differs from the more conventional method of testing for recall and
knowledge of target items via cued recall, whether L1 to L2 or L2 to L1. The two different free recall
tests were used in order to separate the two components, form and meaning, with a view to allowing
participants to engage in a first free recall test without being encumbered by the search for meanings. This
does not mean that they did not also recall meanings, but they were not obliged to do so. A point of
interest here was whether the order of recall might differ between the two conditions due to the linear
progression of the story. After the second free recall, the sheet was removed and participants were given a
final sheet containing the L1 translations for the 16 words. They were asked to provide the L2 target
words for the translations. Thus there were four scores altogether, not including the pre-test: (a) free recall
1 = free recall of L2 word forms without the need to provide L1 translations; (b) free recall 2 = free recall
of L2 word forms plus the provision of L1 translations; (c) L2 to L1 = recall of word meanings as
evidenced by the L1 translations of words provided in the second free recall, and (d) L1 to L2 = cued
recall of L2 translations. The first three tests, with their emphasis on free recall, tap into the potential
organising linear framework of the story; in the final test, L1 words were presented in semi-random order
(12 different orders) so as not to replicate the initial presentation order. In all tests, participants were
encouraged to write partial word forms, even if they could not remember the whole word. L2 words were
scored using Barcroft‘s (2003) lexical production scoring protocol (LPSP). In this procedure, scores are
assigned according to the number of correctly produced letters of a word: .25 for approximately a quarter
of the word, .50 for a half, .75 for three quarters, and 1 for a word completely correct. Two raters scored
the answers independently, the inter-rater reliability being 0.88, with differences resolved through
discussion.
RESULTS
A two-way ANOVA was conducted with presentation condition (story or unrelated) and time (pre-test
and post-test) as independent variables and post-test scores as dependent variables. At free recall 1, there
was a significant effect of presentation (F(1, 47 = 7.67), p < .01, η² = .14), with more words recalled in
the story condition than in the unrelated condition. This advantage persisted at free recall 2, though not
significantly so (F(1, 47) = 3.69, p = .06, η² =.073). In translation from L2 to L1, where the L2 words
were those which the participants wrote down at free recall 2, there was a significant effect
(F(1, 47) = 7.51, p < .01, η² = .138), with more words from the story condition translated correctly than
from the unrelated condition. Finally, in L1-L2 translation there was no significant difference between the
two conditions (F(1, 47) = .47, p = .5, η² = .001). The means for these results appear in table 1. There
were significantly more words translated from L1 to L2 than from L2 to L1, (F(1, 47) = 54.0, p < .0001,
η² = .535), and an interaction between direction of translation and presentation, with the gain in
performance in the L1 to L2 direction compared to L2 to L1 being significantly higher for words seen in
the unrelated condition that for those seen in the story condition (F(1, 47) = 7.32, p < .01, η²=.135).
Table 1. Mean scores (and gains) in each condition over the four measures
Story condition Unrelated condition
Mean (gain) SD Mean (gain) SD
Pre-test 1.51 1.15 1.60 1.33
Free recall 1 3.96 (2.45) 2.10 3.04 (1.44) 1.72
Free recall 2 3.72 (2.21) 2.13 3.13 (1.53) 1.84
L2-L1 3.15 (1.64) 2.06 2.33 (0.73) 1.86
L1-L2 4.15 (2.64) 2.41 4.03 (2.43) 2.39
With regard to word form recall, the precision afforded by the LPSP allows partial word forms to be taken
into account, but in fact when words were recalled, they were overwhelmingly fully correct. Of all the
scores attributed, full points accounted for 85%. There was, however, considerable individual variation,
with scores at free recall 1 ranging from 1.5 to 14.
DISCUSSION
The result of this experiment confirms and extends that of Bower and Clark (1969) relating to narrative
chaining as a technique for learning vocabulary. The major difference was that the story condition in this
experiment provided participants with a ready-made framework, thus removing the need for them to make
it up themselves. The story allows for the linking of semantically unrelated words through a common
context or framework and acts as an aid to recall.
Whilst presenting words in a story appears to be effective overall, it should be noted that the advantage
for the story condition at free recall 1 did not persist to the same extent at free recall 2, which came
immediately after. On average, 0.24 words in the story condition were forgotten between the two recalls
while for the unrelated condition there was a gain of 0.09 words. A possible explanation is that at recall 1,
the story may have facilitated recall of the target words from episodic memory, while at recall 2, the
pictures and the association test may have come into play more, thus diminishing the relative
effectiveness of the story. Partial evidence for this comes from the number of words recalled
consecutively in the order in which they were presented. At recall 1 there were 39 instances of this in the
story condition compared to 17 in the unrelated condition. At recall 2, these scores drop to 14 and 6
respectively. This suggests firstly that when unencumbered by the search for word meanings, participants
made use of the story framework as an aid to recall of word forms, and secondly that when they searched
their memory a second time, albeit immediately after, they made relatively more use of other cues as well.
The experimental design adopted makes it difficult to tease out such effects, but clearly the pictures and
feedback sentences contributed to both learning and recall. As they were available in both conditions,
they do not constitute a confounding factor, but the findings suggest that the benefit of the story was due
to an episodic memory trace quickly superseded by other cues.
Interestingly, the advantage of the story condition was re-established in L2-L1 translation. Here
participants were asked to give translations of the words they had themselves written down at recall 2.
They were therefore recalling meaning as well as form, and the narrative framework appears to have been
helpful in this. One possible, though speculative, explanation is that the propositional content of the
sentences is stored in memory in L1, and the story provides a framework allowing better recall of the
content, which is then linked to the L2 words that the participant has remembered. In the unrelated
condition, this did not result in a corresponding acquisition of word meanings, although performance
improved at free recall 2. It should be noted, however, that the gains in the L2-L1 measure were the
lowest of all, indicating that providing L1 translations at some point is not just useful but necessary. This
can be seen in the final score, L1-L2 translation, where the gains were the highest overall and where there
was no significant difference between the two conditions. Participants were able to use the pictures and
the association test feedback here in order to link the L1 translations to the L2 forms whose meanings
may not have been clear up to then.
Although the overall gains appear to be small, this is in line with other studies that have focused on
incidental learning (in the sense that participants were unaware of the post-test). Hulstijn (1992) and
Hulstijn, Hollander, and Greidanus (1996) report very low retention rates of word forms and meanings,
albeit in conditions somewhat different from the present study in that participants read a complete text.
However, target words in those studies were highlighted in various ways such that participants‘ attention
was drawn to them. In the present study, target words were clearly signalled as such, but only three
participants suspected they might be tested afterwards. The low rates of retention are therefore
unsurprising, but they highlight the difference between retention derived from inferring as opposed to the
larger gains to be obtained from memorising (Mondria, 2003). Thus, although the benefit of the story
condition was significant, one may ask how meaningful this is when retention rates are so small. As
already noted with regard to the pre-test scores, a conservative estimate of gains was used here, and if
some words were falsely reported as familiar at pre-test, actual gains may have been slightly higher.
Nonetheless, although being informed of a post-test does not necessarily lead to higher retention rates
(Peters, 2007; Keating, 2008), it would be interesting to conduct a similar study in which participants are
encouraged to commit form-meaning links to memory in order to see whether the story is also effective in
helping learners to consciously organise the material to be learnt.
The absence of an intentional learning condition is thus the main limitation to this study, which gives only
a partial idea of how the narrative framework operates. When used as a conscious learner-generated
mnemonic device, as in Bower and Clark (1969), narrative chaining can be highly effective. The present
study shows that a teacher-provided narrative can be an aid to recall when there is no intention to commit
words to memory, but there is now a need to investigate a condition lying mid-way between the two, (i.e.,
when the story is provided and participants are encouraged to make use of it for long-term retention). This
would require a delayed test of vocabulary retention as part of a study looking at the effect of consciously
using the story as a means to organise target words in long-term memory. As Hulstijn (2001, p. 274)
notes, a delayed post-test becomes appropriate when one is investigating the effect of various rehearsal
strategies as opposed to the effect of initial exposure to target words. A delayed post-test could also, of
course, investigate the effect of initial exposure provided that learners did not encounter the target words
in the meantime. This may be possible with invented or low-frequency words but with the present
material could not be guaranteed. As pointed out by Bjork (1999), conditions that promote learning
during a training phase may be ineffective in the longer term, and in the absence of a delayed post-test,
the present study does not address this issue. However, in real learning conditions, the question is more
whether learners engage in rehearsal or not, and if they do, whether organising the words to be learnt in a
story can be of help.
Another limitation of this study is the absence among the post-tests of an exact replication of the pre-test
conditions (i.e., cued L2-L1 recall), since there was thus no strict comparison between equivalent
conditions. Thus, one cannot conclude from these results that the story condition is more effective in a
purely receptive test. However, the difference in scores between pre-test and post-test reflects the ability
to produce at post-test the form and/or meaning of a word that was not recognised at pre-test and thus
indicates a gain in knowledge of that word. Furthermore, it could be argued that the free recall procedure
adopted is more stringent than cued recall as it tests the effectiveness of the presentation conditions in
triggering recall, first of target word forms and then of meanings. Although there were no doubt words
that were not recalled in either condition but whose meanings might have been found in a receptive post-
test, this does not invalidate the fact that the story appears to help learners generate cues internally for
purposes of recall. This may be of benefit when no other cues are provided, as is often the case in written
or oral expression, but it remains to be seen, of course, whether the effect reported here could be
transferred to such tasks.
A final comment concerns to what extent the features proposed in the set of principles are additive or not.
Demonstrating an effect in isolation is of value, but as noted by Bolger and Zapata (2011) the relative
strength of that effect then needs to be assessed by comparing it to other effects. The present study sought
to investigate the effect of the story within a presentation procedure that combined other features, but did
not incorporate those features as experimental variables. Future research would need to address this issue
if we are to gain a clearer understanding of how the different elements of an instructional package should
be combined to ensure optimal learning. Think aloud protocols offer one way of sounding out
participants‘ processing of input and thus gaining an insight into the depth to which the various elements
of a multimedia presentation are processed (Lomicka, 1998; Bowles, 2004; Yanguas, 2009). From a
theoretical standpoint a productive framework is offered by the generative theory of multimedia learning
and cognitive load theory (Chandler & Sweller, 1991; Jones, 2004; Mayer & Moreno, 2003), but several
questions remain. For example, order of presentation is one parameter that has received little attention, yet
it clearly conditions the sort of processing that learners will engage in. In the present study context
sentence and picture appeared simultaneously, but depending on the degree of explicitness of each, it is
possible that deeper processing could be promoted by presenting them separately. Also, in the case of
abstract words, it remains to be seen precisely what role pictures can play. Though they cannot convey the
exact meaning(s) of an abstract word as well as translation(s), pictures relating to the sentence as a whole
may nonetheless be helpful. While the result of the experiment indicates that the presence of a narrative
framework was beneficial, the design did not allow the effect to be teased out with respect to the
concrete-abstract dimension. Further research is needed to see how strong this effect is in conjunction
with differing types of pictures.
CONCLUSION
The set of principles put forward in this article do not all command unanimous agreement; nevertheless,
as they have all garnered sufficient empirical support when regarded as separate features, a consideration
of their merits—when combined—should be of interest. The present article has sought to examine an
approach to vocabulary learning that combines a rich contextual presentation with a systematic
instructional programme.
One obvious pedagogical implication to this study is that course designers may be interested in making
use of stories to enhance recall of specific sets of vocabulary items. The task of generating stories
comprising clear context sentences is time-consuming, but when spread between learners who then share
the results using multimedia, a large number of words can be covered. Involving learners in the design of
certain aspects of their own course can be both motivating (Dörnyei, 2001, p.104) and effective.
Nikolova (2002) found that students who helped to create annotations for target words learned them better
than a control group with annotations already given, though not when time on task was taken into
account. When I asked my students to compose stories in pairs, each pair using ten target words,
considerable creativity was shown, even if the results had to be carefully checked and corrected. The
stories were then redistributed, with the target words in bold type, and different pairs were asked to find
pictures illustrating the context in which the words appeared. The final results were emailed to the group
as a whole and formed the basis of a follow up production activity that was engaged in with much
enthusiasm. Though little research has been conducted to assess the effect of stories on vocabulary
learning in similarly authentic conditions, the present study suggests that they provide a potentially rich
seam to tap.
REFERENCES
Al-Seghayer, K. (2001). The effects of multimedia annotation modes on L2 vocabulary acquisition: A
comparative study. Language Learning & Technology, 5(1), 202–232. Retrieved from http://llt.msu.edu/
vol5num1/alseghayer/default.html
Atay, D., & Ozbulgan, C. (2007). Memory strategy instruction, contextual learning and ESP vocabulary
recall. English for Specific Purposes, 26, 39–51.
Avila, E., & Sadoski, M. (1996). Exploring new applications of the keyword method to acquire English
vocabulary. Language Learning, 46, 379–395.
Barcroft, J. (2003). Effects of questions about word meaning during L2 lexical learning. The Modern
Language Journal, 87(4), 546–561.
Barcroft, J. (2009). Strategies and performance in intentional L2 vocabulary learning. Language
Awareness, 18,(1), 74–89.
Bjork, R. A. (1999). Assessing our own competence: Heuristics and illusions. In D. Gopher and A. Koriat
(Eds.), Attention and peformance XVII. Cognitive regulation of performance: Interaction of theory and
application (pp. 435–459). Cambridge, MA: MIT Press.
Boers, F., Piquer Píriz, A. M., Stengers, H., & Eyckmans, J. (2009). Does pictorial elucidation foster
recollection of idioms? Language Teaching Research, 13(4), 367–382.
Bolger, P., & Zapata, G. (2011). Semantic categories and context in L2 vocabulary learning. Language
Learning, 61(2), 614–646.
Bower, G. H., & Clark, M. C. (1969). Narrative stories as mediators for serial learning. Psychonomic
Science, 14, 181–182.
Bowles, M. A. (2004). L2 glossing: To CALL or not to CALL. Hispania, 87(3), 541–552.
Brown, R., Waring, R., & Donkaewbua, S. (2008). Incidental vocabulary acquisition from reading,
reading-while-listening, and listening to stories. Reading in a Foreign Language, 20(2), 136–163.
Chandler, P., & Sweller, J. (1991). Cognitive load theory and the format of instruction. Cognition and
Instruction, 8, 293–332.
Chun, D. M., & Plass, J. L. (1996). Effects of multimedia annotations on vocabulary acquisition. The
Modern Language Journal, 80(2), 183–198.
Cobb, T. (2007). Computing the vocabulary demands of L2 reading. Language Learning & Technology,
11(3), 38–63. Retrieved from http://llt.msu.edu/vol11num3/cobb/default.html
Comesaña, M., Perea, M., Piñeiro, A., & Fraga, I. (2009). Vocabulary teaching strategies and conceptual
representations of words in L2 in children: Evidence with novice learners. Journal of Experimental Child
Psychology, 104, 22–33.
Craik, F. I. M., & Lockhart, R. S. (1972). Levels of processing: A framework for memory research.
Journal of Verbal Learning and Verbal Behavior, 11, 671–684.
Diao, Y., & Sweller, J. (2007). Redundancy in foreign language reading comprehension instruction:
Concurrent written and spoken presentations. Learning and Instruction, 17, 78–88.
Dörnyei, Z. (2001). Motivational strategies in the language classroom. Cambridge: Cambridge
University Press.
Erten, I. H., & Tekin, M. (2008). Effects on vocabulary acquisition of presenting new words in semantic
sets versus semantically unrelated sets. System, 36, 407–422.
Finkbeiner, M., & Nicol, J. (2003). Semantic category effects in second language word learning. Applied
Psycholinguistics, 24(3), 369–383.
Folse, K. S. (2004). Myths about teaching and learning second language vocabulary: What recent research
says. TESL Reporter 37 (2) 1–13.
Frantzen, D. (2003). Factors affecting how second language Spanish students deriving meaning from
context. The Modern Language Journal, 87, 168–199.
Greenway, S. (2008). ID: The quest for identity in the 21st century. London: Sceptre.
Hall, J. W., Wilson, K. P., & Patterson, R. J. (1981). Mnemonic techniques: Some limitations of the
mnemonic keyword method for the study of foreign language vocabulary. Journal of Educational
Psychology, 73, 345–357.
Hanley, J., Herron, C., & Cole, S. (1995). Using video as advance organizer to a written passage in the
FLES classroom. The Modern Language Journal, 79(1), 57–66
Horiba, Y., van den Breuk, P., & Fletcher, C. R. (1993). Second language readers‘ memory for narrative
texts: Evidence for structure–preserving top–down processing. Language Learning, 43, 345–372.
Horst, M., Cobb, T., & Meara, P. (1998). Beyond A Clockwork Orange: Acquiring second language
vocabulary through reading. Reading in a Foreign Language, 11, 207–223.
Hulstijn, J. H. (1992). Retention of inferred and given word meanings: Experiments in incidental
vocabulary learning. In P. J. L. Arnaud & H. Bejoint (Eds.), Vocabulary and applied linguistics (pp. 113–
125). London: Macmillan.
Hulstijn, J. H. (1997). Mnemonic methods in foreign language vocabulary learning: Theoretical
considerations and pedagogical implications. In J. Coady & T. Huckin (Eds.), Second language
vocabulary acquisition (pp. 203–224). Cambridge: Cambridge University Press.
Hulstijn, J. H. (2001). Intentional and incidental second–language vocabulary learning: A reappraisal of
elaboration, rehearsal and automaticity. In P. Robinson (Ed.), Cognition and second language instruction
(pp. 258–286). Cambridge: Cambridge University Press.
Hulstijn, J. H., Hollander, M., & Greidanus, T. (1996). Incidental vocabulary learning by advanced
foreign language students: The influence of marginal glosses, dictionary use, and reoccurrence of
unknown words. The Modern Language Journal, 80(3), 327–339.
Hulstijn, J. H., & Laufer, B. (2001). Some empirical evidence for the involvement load hypothesis in
vocabulary acquisition. Language Learning, 51(3), 539–558.
Hunt, A., & Beglar, D. (1998). Current research and practice in teaching vocabulary. The Language
Teacher. Retrieved from http://www.jalt-publications.org/old_tlt/articles/1998/01/hunt
Hunt, A. & Beglar, D. (2005). A framework for developing EFL reading vocabulary. Reading in a
Foreign Language, 17(1), 23–59.
Jones, L., & Plass, J. (2002). Supporting listening comprehension with multimedia annotations. The
Modern Language Journal, 86(4), 546–561.
Jones, L. (2004). Testing L2 vocabulary recognition and recall using pictorial and written test items.
Language Learning & Technology, 8(3), 122–143. Retrieved from http://llt.msu.edu/vol8num3/jones/
default.html
Kaivanpanah, S., & Alavi, M. (2008). Deriving unknown word meaning from context: Is it reliable?
RELC Journal, 39, 77–95.
Keating, G. (2008). Task effectiveness and word learning in a second language: The involvement load
hypothesis on trial. Language Teaching Research, 12(3) 365–386.
Kilgariff, A. (1996). A lemmatised frequency list for the 6,318 words with more than 800 occurrences in
the British National Corpus. http://www.kilgarriff.co.uk/BNClists/lemma.num
Kost, C., Foss, P., & Lenzini, J. (1999). Textual and pictorial gloss: Effectiveness on incidental
vocabulary growth when reading in a foreign language. Foreign Language Annals, 32(1), 89–113.
Laufer, B. (1992). How much lexis is necessary for reading comprehension? In P. Arnaud & H. Béjoint
(Eds.), Vocabulary and applied linguistics (pp. 126–132). London: Macmillan.
Laufer, B. (1997). The lexical plight in second language reading: words you don‘t know, words you think
you know, and words you can‘t guess. In: Huckin, T. (Eds.), Second language vocabulary acquisition.
Cambridge: Cambridge University Press, 20–34.
Laufer, B. (2001). Reading, word-focused activities and incidental vocabulary acquisition in a second
language. Prospect, 16(3), 44–54.
Laufer, B., & Hulstijn, J. H. (2001). Incidental vocabulary acquisition in a second language: The construct
of task–induced involvement. Applied Linguistics, 22(1), 1–26.
Laufer, B., & Ravenhorst-Kalovski, G. C. (2010). Lexical threshold revisited: lexical text coverage,
learners‘ vocabulary size and reading comprehension. Reading in a Foreign Language 22, 15–30.
Laufer, B., & Shmueli, K. (1997). Memorizing new words: Does teaching have anything to do with it?
RELC Journal, 28(1), 89–108.
Laufer, B., & Sim, D. D. (1985). Taking the easy way out, non use and misuse of clues in EFL reading.
English Teaching Forum 23(2), 7–10.
Lomicka, L. (1998). To gloss or not to gloss: An investigation of reading comprehension on-line.
Language Learning & Technology, 1(2), 41–50. Retrieved from http://llt.msu.edu/vol1num2/
article2/default.html
Lotto, L., & de Groot, A. M. B. (1998). Effects of learning method and word type on acquiring
vocabulary in an unfamiliar language. Language Learning, 48(1), 31–69.
Mayer, R. E., & Moreno, R. (2003). Nine ways to reduce cognitive load in multimedia learning.
Educational Psychologist, 38(1) 43–52.
McNamara, D. S., & Scott, J. L. (2001). Working memory capacity and strategy use. Memory and
Cognition, 29, 10–17.
Mondria, J. (2003). The effects of inferring, verifying, and memorizing on the retention of L2 word
meanings: An experimental comparison of the ―meaning–inferred method‖ and the ―meaning–given
method‖. Studies in Second Language Acquisition, 25, 473–499.
Mondria, J., & Wit-de Boer, M. (1991). The effects of contextual richness on the guessability and the
retention of words in a foreign language. Applied Linguistics, 12, 249–267.
Nassaji, H. (2004). Input modality and remembering name–referent associations in vocabulary learning.
Canadian Journal of Applied Linguistics, 7,(1), 39–55.
Nation, I. S. P. (2001). Learning vocabulary in another language. Cambridge: Cambridge University
Press.
Nation, I. S. P. (2006). How large a vocabulary is needed for reading and listening? Canadian Modern
Language Review, 63(1), 59–82.
Nikolova, O. (2002). Effects of students‘ participation in authoring of multimedia materials on student
acquisition of vocabulary. Language Learning & Technology, 6(1), 100–112. Retrieved from
http://llt.msu.edu/vol6num1/nikolova/default.html
Omaggio, A. C. (1979). Pictures and second language comprehension: Do they help? Foreign Language
Annals, 12, 107–116.
Paivio, A. (1969). Mental imagery in associative learning and memory. Psychological Review, 76(3),241–
263.
Paribakht, T. S., & Wesche, M. (1997). Vocabulary enhancement activities and reading for meaning in
second language vocabulary development. In J. Coady & T. Huckin (Eds.), Second language vocabulary
acquisition (pp. 174–200). Cambridge: Cambridge University Press.
Peters, E. (2007). Manipulating L2 learners‘ online dictionary use and its effect on L2 word retention.
Language Learning & Technology, 11(2), 36–58. Retrieved from http://llt.msu.edu/vol11num2/
peters/default.html
Plass, J. L., Chun, D., Mayer, R. E., & Leutner, D. (2003). Cognitive load in reading a foreign language
text with multimedia aids and the influence of verbal and spatial abilities. Computers in Human Behavior,
19, 221–243.
Pressley, M., Levin, J. R., & Delaney, H. D. (1982). The mnemonic keyword method. Review of
Educational Research, 52, 61–91.
Rodriguez, M., & Sadoski, M. (2000). Effects of rote, context, keyword, and context/keyword methods on
retention of vocabulary in EFL classrooms. Language Learning, 50, 385–412.
Sagarra, N., & Alba, M. (2006). The key is in the keyword: L2 vocabulary learning methods with
beginning learners of Spanish. The Modern Language Journal, 90(2), 228–243.
Schmitt, N. (1997). Vocabulary learning strategies. In N. Schmitt & M. McCarthy (Eds.), Vocabulary:
Description, acquisition and pedagogy (pp. 199–227). Cambridge: Cambridge University Press.
Schmitt, N. (2008). Instructed second language vocabulary learning. Language Teaching Research, 12(3),
329–363.
Seibert, L. C. (1930). An experiment on the relative efficiency of studying French vocabulary in
associated pairs versus studying French vocabulary in context. Journal of Educational Psychology, 21,
297–314.
Snyder, H., & Colon, I. (1988). Foreign language acquisition and audio–visual aids. Foreign Language
Annals, 21(4), 343–384.
Sydorenko, T. (2010). Modality of input and vocabulary acquisition. Language Learning & Technology,
14(2), 50–73. Retrieved from http://llt.msu.edu/vol14num2/sydorenko.pdf
Taylor, W. L. (1953). ―Cloze procedure‖: A new tool for measuring readability. Journalism Quarterly, 30,
415–33.
Tinkham, T. (1993). The effect of semantic clustering on the learning of second language vocabulary.
System, 21, 371–380.
Tinkham, T. (1997). The effects of semantic and thematic clustering in the learning of second language
vocabuary. Second Language Research 13(2), 138–63.
Tonzar, C., Lotto, L., & Job, R. (2009). L2 vocabulary acquisition in children: Effects of learning method
and cognate status. Language Learning, 59, 623–646.
Trabasso, T., & van den Broek, P. (1985). Causal thinking and the representation of narrative Events.
Journal of Memory and Language, 24, 612–30.
Waring, R. (1997). The negative effect of learning words in semantic sets: A replication. System 25(2),
261–74.
Waring, R. (2002). Scales of vocabulary knowledge in second language vocabulary assessment. Kiyo, The
occasional papers of notre dame seishin univsersity. Retrieved from http://www.doc88.com/
p-51266788660.html
Webb, S. (2005). Receptive and productive vocabulary learning: The effects of reading and writing on
word knowledge. Studies in Second Language Acquisition, 27, 33–52.
Webb, S. (2007). Learning word pairs and glossed sentences: The effects of a single sentence on
vocabulary knowledge. Language Teaching Research, 11, 63–81.
Webb, S. (2008). The effects of context on incidental vocabulary learning. Reading in a Foreign
Language, 20(2), 232–245.
Webb, S., & Rodgers, M. P. H. (2009a). The vocabulary demands of television programs. Language
Learning, 59, 335–366.
Webb, S., & Rodgers, M. P. H. (2009b). The lexical coverage of movies. Applied Linguistics, 30, 407–
427.
Yanguas, I. (2009). Multimedia glosses and their effect on L2 text comprehension and vocabulary
learning. Language Learning & Technology, 13(2), 48–67. Retrieved from http://llt.msu.edu/vol13num2/
yanguas.pdf
Yoshii, M. (2006). L1 and L2 glosses: Their effects on incidental vocabulary learning. Language
Learning & Technology, 10(3), 85–101. Retrieved from http://llt.msu.edu/vol10num3/yoshii/default.html
Zahar, R., Cobb, T., & Spada, N. (2001). Acquiring vocabulary through reading: Effects of frequency and
contextual richness. Canadian Modern Language Review, 57(4), 541–572.