See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/354065550
English as a Foreign Language International Journal
Article · July 2021
CITATIONS
READS
0
4,559
1 author:
Napak-On Sritrakarn
Rajamangala University of Technology Isan Khon Khan Campus
9 PUBLICATIONS 27 CITATIONS
SEE PROFILE
All content following this page was uploaded by Napak-On Sritrakarn on 23 August 2021.
The user has requested enhancement of the downloaded file.
English as a Foreign Language
International Journal
(EFLIJ)
Volume 25 Issue 4
July 2021
Published by the Academics Education International Journals
English as a Foreign Language International Journal
A Division of AEIJ
Part of TESOL One
www.academics.education/eflij
© English as a Foreign Language International Journal 2021
This book is in copyright. Subject to statutory exception no reproduction of any part may
take place without the written permission of the Academics Education International
Journals.
No unauthorized photocopying
All rights reserved. No part of this book may be reproduced, stored in a retrieval system or
transmitted in any form or by any means, electronic, mechanical, photocopying or otherwise,
without the prior written permission of the Academics Education International Journals.
editor@academics.education
Publisher: Academics Education International Journals
Chief Editor: Dr. John Adamson
Guest and Production Editor: Dr. Jun (Scott) Chen Hsieh
ISSN: 2799-0699 (Online)
Table of Contents
Foreword by Jun (Scott) Chen Hsieh.……………..…....………………..….…...... 1-2
1. Napak-on Sritrakarn ….………………...........………………..……................... 3-25
- The Effects of Anonymity and Non-Anonymity on Students’ Writing Reviews
and Corrections
2. Omar Karlin & Sayaka Karlin ……..………………………..….…………….... 26-54
- Comparing the Effectiveness of L2 Extensive and Intensive Listening
Approaches
3. Suthathip Thirakunkovit ………………………….………..………………...…. 55-83
- Production and Perception of Geminate Consonants in English Words by
Thai Learners of English: Implications for English Teaching and Learning
Book Reviews
1. English-medium Instruction and Translanguaging ………………………………. 84-86
BethAnne Palsrud, Tian, Zhongfeng and Jeanette Toth (Eds). Bristol: Multilingual
Matters, 2020. ISBN 13: 978-1-78892-731-4 (pbk). Pp. 208.
Reviewed by Marilyn Lewis, Honorary Research Fellow,
The University of Auckland
2. Critical Thinking in ELT: A Working Model for the Classroom ………………… 87-89
Dummett and John Hughes. National Geographic Learning: Boston, USA, 2019.
Pp. viii +158.
Reviewed by Thiri Soe, Graduate School of International Cultural Studies,
Tohoku University, Japan
EFLIJ Volume 25 Issue 4 July 2021
1
July 2021 Foreword
Welcome to the July issue of English as a Foreign Language International Journal in 2021.
Including three research articles and two book reviews, this issue explores diverse topics
essential to the field of teaching and learning English as a second/foreign language. Specifically,
the research articles address anonymity and non-anonymity in writing, extensive and intensive
listening approaches, and geminate consonants in English words.
The July issue starts with “The Effects of Anonymity and Non-Anonymity on Students’ Writing
Reviews and Corrections” by Napak-on Sritrakarn. In comparing the effects of employing the
senior review activity in two modes of anonymous and non-anonymous reviews, Napak-on
Sritrakarn observed students’ positive attitudes toward the mode of reviews and more
constructive comments on junior students’ writing from senior students in the anonomous mode.
Implications of a review activity in a writing classroom as well as suggestions for future studies
are provided.
Omar Karlin and Sayaka Karlin, in “Comparing the Effectiveness of L2 Extensive and Intensive
Listening Approaches,” engaged three groups of Japanese university students in six listening
assignments adhering to either principles of extensive or intensive listening. Based on the pretest and post-test of 100-question TOEIC listening tests, the researchers found the advantages
of the extensive listening approach over the intensive listening approach. Implicatons and
suggestions include codifying extenstive listening and intensive listening methodologies,
ensuring variability in testing conditions, and scrutinizing the participants' adherence to the
intervention.
In the third article, entitled “Production and Perception of Geminate Consonants in English
Words by Thai Learners of English: Implications for English Teaching and Learning,”
Suthathip Thirakunkovit explored the production and perception of geminate consonants of
Thai learners of English. By examining the tasks as reading aloud and dictation, Suthathip
Thirakunkovit identified sifnificant differences in the production of lexical geminates between
native and non-native English speakers. Low intermediate students seem to find identifying
English geminates in regular speech challenging. This study paves the way for future research
concerning geminate consonant production and perception in ESL/EFL contexts.
1
EFLIJ Volume 25 Issue 4 July 2021
2
Marilyn Lewis and Thiri Soe, respectively reviewing “English-medium Instruction and
Translanguaging” and “Critical Thinking in ELT: A Working Model for the Classroom”,
succinctly convey the contents and value of the books with a brief summary of the book’s
content and attention to its main focus, thus facilitating the readers’ understanding of EMI,
translanguaging, and critical thinking.
We hope you find the articles in this July 2021 issue to be informative, inspiring, and
comprehensive. Bearing in mind the contribution to continuous improvement in English
language instruction around the world, particulary in times of Covid-19 pandemic, we sincerely
hope that this issue helps provide new insights into the formulation of future research and
innovations for EFL/ESL practitioners in cross-border, interdisciplinary, and collaborative
manners. We would like to express our sincere appreciation to the contributors and reviewers
of articles and book reviews who have made this issue possible.
Dr. Jun (Scott) Chen Hsieh
Department of Foreign Languages and Literature
Asia University, Taiwan
Guest and Production Editor of the English as a Foreign Language International
Journal (EFLIJ) July 2021 Issue
2
EFLIJ Volume 25 Issue 4 July 2021
3
The Effects of Anonymity and Non-Anonymity on Students’ Writing
Reviews and Corrections
Napak-on Sritrakarn
Rajamangala University of Technology Isan,
Khon Kaen Campus, Thailand
Bio data
Napak-on Sritrakarn is an Associate Professor at the Department of English for International
Communication, Faculty of Technical Education, Rajamangala University of Technology Isan,
Khon Kaen Campus, Thailand. She received her Ph.D. from Macquarie University, Australia.
Her research interests cover English Language Teaching (ELT), academic literacy, discourse
analysis, and genre studies. napak-on.sr@rmuti.ac.th
Abstract
This study compared the effects of employing the senior review activity in two modes of
anonymous and non-anonymous reviews. There were seven senior students who have made
reviews on the essay writing drafts of two junior classes on the same topic. It was found that
the two groups of junior students had positive attitudes toward the mode of reviews they
participated in and that senior students had made more constructive comments on junior
students’ writing when their names were not disclosed. The findings on senior students’
attitudes also informed that even though they had no problems with the review conditions of
either being anonymous or non-anonymous, they preferred to know whose work they were
reviewing. The paper discusses some implications for the application of a review activity in a
writing classroom as well as directions for future studies.
Keywords: writing reviews; anonymity; non-anonymity
Correspondence address:
Rajamangala University of Technology Isan,
Khon Kaen Campus
Address: 150 Srichan Road, Muang, Khon
Kaen, 40000 Thailand.
3
EFLIJ Volume 25 Issue 4 July 2021
4
Introduction
English has long been used as a global language for communication worldwide, and the
proficient use of it as a language tool is of imperative importance (Nur, 2020; Rajab et al., 2020;
Vallente, 2020). Even though the requirements of its command are high, many Thai learners at
different levels have difficulties in learning the language (Arjpru, 2017; Foley, 2005; Noomura, 2013; Sritrakarn, 2018; Sweeney, Kunyot, & Preedeekul, 2017; Viriya, 2018). Not only
that students find it difficult to master the skills in English, but teachers also find teaching
English challenging as well. Baiyaem (1997) reported that the obstacles faced by teachers
include heavy teaching loads and insufficient language skills and teaching knowledge. For
students, numerous studies claimed that the challenges for their learning are such as the lack of
opportunity to use English in real life (Noom-ura, 2013; Viriya, 2018; Wiriyachitra, 2003),
being passive learners (Arjpru, 2017; Viriya, 2018), and uninteresting lessons (Viriya, 2018).
Especially for many Thais as well as other Asian learners, the culture issue could be one of the
barriers for students to learn the English language (Chen, 2019; Raymond & Choon, 2017).
Findings from previous studies show that many foreign language learners have anxiety in
learning a language (Akbari & Sadeghi, 2013; Dordinejad & Ahmadabad, 2014). They are shy
to communicate in English with their classmates (Arjpru, 2017; Vallente, 2020). Some even
intend to make efforts to maintain friendly relationships with peers by trying to avoid
confrontation in the interaction (Chen, 2019) while others may avoid asking questions for fear
of being incorrect or feel embarrassed (Raymond & Choon, 2017). As a consequence, they
engage less and display a passive role in class. These reflect serious problems as a result of
cultural impacts. Given that exposure to the language through active participation in class is
one of the key factors for successful English language learning, by avoiding to interact or
participate, their chances for improvement could be limited. This may thereby result in
unproductive learning outcomes. In order to overcome the abovementioned problems faced by
both teachers and students in language learning and teaching, it is essential for the teacher to
design a classroom learning activity that encourages students to interact actively using the
language, take more control of their own learning, and work cooperatively to accomplish the
task (Altun & Sabah, 2020). By focusing on the improvement of Thai students’ writing, the
present study employed anonymous and non-anonymous review activities in an essay writing
classroom. The following section discusses the literature related to the current study.
4
EFLIJ Volume 25 Issue 4 July 2021
5
Literature Review
Writing
Writing is the act of putting down the graphic symbols that present a language in order to convey
some meaning so that the reader can grasp the information which the writer has tried to impart
(Sapkota, 2012). It is a key factor in students’ academic development and the primary means
by which students transform from passive recipients of knowledge to active participants in their
own education (Ruiling & Bol, 2007). Writing is a complex process that requires both language
and many other skills to accomplish. For language skills, Watcharapunyawong and Usaha
(2013) argued that writing requires both syntactic and semantic knowledge. For other skills,
writing also requires a great deal of effort to continuously practice (ibid.). This also includes
mental, psychological, rhetorical, and critical skills (Sapkota, 2012).
As far as the nature of communication is concerned, writing and speaking skills are
different in that while spoken language is context-dependent, the writer and reader share neither
verbal nor aural contact in written language. In other words, the language is used independently
from the context in writing, and the structure is organized in a monologic way with a diversity
of lexical choices relevant to the topic contents (Piriyasilpa, 2009). For these reasons, many
students find writing skill to be the most challenging in real practice and in the examination
(Sapkota, 2012; Zakaria & Hashim, 2020), and this problem exists even among native English
speaking learners (Lu & Bol, 2007). In order to support students to compose their English
essays more successfully, a number of strategies have been employed in the language
classrooms; and one of the common techniques is the peer review activity
Peer review
When students critique one another’s work with the intention of helping their peers revise and
eventually improve their writing, the process is called peer review (Ruiling & Bol, 2007). Peer
review is also defined by Waemusa (2017, p. 93) as “a learning process whereby students
engage in communication for work evaluation and reflection.” According to Hosack (2003),
peer review involves learners reading one another’s draft compositions and providing feedback
that can be used by the writer when revising the drafts. Based on these definitions, the concepts
of peer review can be summed up as a kind of cooperative learning activity in which students
read the writing drafts of their peers and provide feedback with the goal to help revise and
improve the drafts. Peer review can be both verbal and in written form. Even though the
feedback made by students is suspected by Truscot (1996), other studies have argued that peer
review has become commonplace in the writing environment (Ruiling & Bol, 2007) and the
5
EFLIJ Volume 25 Issue 4 July 2021
6
activity is a central aspect of English as a Second Language (ESL) and English as a Foreign
Language (EFL) writing programs across the world (Hyland & Hyland, 2006). Colpitts (2016)
argues that allowing students to engage in peer editing of one another’s work is surely not as
accurate as having the teacher do so, but it does reduce workloads for time-constrained
educators.
Previous studies have conducted peer review activity in the language classrooms and
claimed for numerous benefits of the activity. First of all, the peer review process promotes
critical thinking and the awareness of effective writing skills (Kunwongse, 2013). It allows
students to analyze and provide constructive feedback on their peers’ writing (Ngoc lac &
Gurung, 2015) before making a reflection on their own work (Hosak, 2003). Evidence of this
has been shown in the findings from the study by Colpitts (2016), in which student participants
considered that both giving and receiving feedback from other students helped with the act of
“noticing” and the act of “discovery.” That is to say, by analyzing their peers’ work critically,
students are encouraged to reconsider and adjust their own writing. For those who receive the
feedback, they are more attuned to the needs and expectations of the readers. This whole process
enables students to interact analytically and promotes cognitive processing, which results in the
‘durability of memory’ (Houston, 2001, p. 270) and constitutes part of the process of second
language learning (Swain, 2005). Secondly, feedback made by peers tend to be pitched close to
the learner’s own level of proficiency, so it is potentially more informative to the novice writer
than feedback provided by the instructor (Hosak, 2003). The study by Sritrakarn (2018) also
showed that students felt comfortable receiving feedback from other students while teacher
comments were sometimes confusing and abstract. Finally, peer review activity promotes social
interaction and collaborative learning (Lu & Bol, 2007). Given that language learning happens
as a result of social interactions (Vygotsky, 1986), learners need to be engaged in the writing
environments which enable them to collaborate and interact with peers or adults. By taking an
active role in analyzing their peers’ work, and by providing feedback and receiving comments
through the peer review activity, students are provided with the opportunity to communicate
linguistically and eventually acquire the target language (Swain, 2005).
Scholars have also pointed out some problems of peer review activity. Despite the
problems in terms of time consumed (Kunwongse, 2013), the two main problems of peer review
can be pointed out in terms of feedback quality and cultural issues. To begin, concerns have
been made regarding the quality of peer feedback. Studies found that some students who
provided feedback have limited knowledge of language and grammar (Wanchid, 2015) and felt
they did not have enough English ability themselves to accurately give feedback to others
6
EFLIJ Volume 25 Issue 4 July 2021
7
(Colpitts, 2016). Moreover, those who received the feedback stated that they still preferred the
teacher’s feedback to the feedback received from other students (Sritrakarn, 2018). Even though
some problems still exist in the application of peer review activity, numerous studies argue that
if designed appropriately, the peer review activity can still bring benefits to students’ writing
practices. Colpitts (2016), for example, argues that while the students lack the linguistic and
grammatical abilities to always correct one another’s work with confidence, they still felt a
sense of personal growth in terms of their own English writing ability occurring during the
process of giving and receiving peer feedback. Therefore, it is essential for the study on peer
review activity to be aware of those aforementioned problems and design the activity
systematically to overcome the problems.
Another problem that has been mentioned in a number of studies is related to cultural
issues, especially if identifiable peer review activity is conducted. Evidence from studies
conducted in Asian learning contexts show that cultural norms can restrain students’ level of
collaborative learning (Chen 2019; Cote, 2007; Lu & Bol, 2007; Waemuza, 2017; Wanchid,
2015). Scholars found that some learners have difficulty providing feedback because they are
reluctant to criticize their classmates’ work (Hosac, 2003; Waemusa, 2017). Raymond and
Choon (2007) explain that this is because Asian students are “face conscious” in nature (p. 198).
They believe that one should be modest and should not put someone else to shame (ibid.). As a
result, they tend to maintain a harmonious relationship in the classroom and avoid challenging
others by asking questions or expressing their different thoughts on the discussed issue. In
addition to these cultural factors, other related factors that have the potential to influence
students to be easily biased or not honest in giving feedback are such as friendship, gender,
race, interpersonal relationships, or personal preferences (Lu & Bol, 2007). This implies that
cultural issues as well as learners’ identities are vulnerable during the language learning process
and can be at risk if students are exposed to direct criticism or forced to be critical of their peers
(Silver & Coomber, 2010). The discussion also raises the question of whether or not the
undisclosed identity would help to promote more interaction and successful language learning
among students.
By following the concepts of using ‘senior’ as a replacement for ‘peer’ (Sritrakarn, 2018),
the present study aims to enhance the quality of feedback made in the review activity. Sritrakarn
(2018) claimed that when working on the task with the assistance of experienced adults,
students could refine knowledge and build up further from their existing knowledge and
understanding of the subject matter. In this context of the investigation, the selected senior
students have undertaken the English essay writing course. They, therefore, had the experience
7
EFLIJ Volume 25 Issue 4 July 2021
8
in learning and composing the essay to review the writing drafts of junior students. To examine
whether or not learner identities have an influence on the senior students’ reviews and the junior
students’ corrections, anonymous and non-anonymous review activities were conducted. The
following section discusses further the concepts of anonymity and its benefits.
Anonymity
Scholars define anonymity as the learning and teaching activity in which both reviewers and
reviewees are kept unknown to one another (Lu & Bol, 2007), or learners’ real identities are
kept unknown to others by using pseudonyms or not using names at all (Chen, 2009).
Anonymity has been employed in different language classrooms of both native-speaking
contexts (Chen, 2009; Lu & Bol, 2007) and non-native speaking contexts (Cote, 2014; Hosack,
2003; Jessup, Connolly & Tansik, 1990; Silver & Coomber, 2010). Based on the findings from
these previous studies, the benefits of anonymity in language learning can be pointed out in
three main aspects: deindividuation, interpersonal relationship, and more critical feedback.
First of all, anonymity contributes to deindividuation, which was defined by Lu & Bol
(2007) as the situations in which individuals in groups stop thinking of other members as
individuals and feel that they cannot be singled out by others. Anonymity fosters the level of
comfort in language learning. When students feel that they cannot be singled out by others, they
stop thinking of other members as an individual, resulting in a reduction of normal inner
restraints and enabling group members to engage in behavior that they would not ordinarily
display (Jessup et al., 1990). In this way, they participate and contribute more to the learning
process. Evidence of this has been shown in the study by Chen (2019), which investigated the
impacts of anonymity on international students’ learning performances and found that
deindividuation promotes less peer pressure and that international students who have diverse
backgrounds suffer less from the social constraints and feel more comfortable to participate.
Secondly, anonymity in peer review promotes interpersonal relationships among the
participants. By nature, Asian students tend to be reluctant to criticize peers’ work (Hosack,
2003). They make an effort to maintain friendly relationships with peer students (Chen, 2019).
So, they may try to avoid confrontations or conflicts with friends in face-to-face peer review
activities. When staying low profile in the peer review activity, students make their comments
less aggressive, allowing for a higher level of participation and production in peer interaction
(ibid.).
Finally, anonymous review enhances more critical feedback. Anonymous reviewers were
more honest with their comments and provided more critical feedback than those who
8
EFLIJ Volume 25 Issue 4 July 2021
9
conducted peer review identifiably (Lu & Bol, 2007). Cote (2014) adds that the feedback
provided anonymously is based solely on the text itself, not its author, because the reviewers
do not know whose essay they are reviewing. The activity, therefore, provides a comfort zone
for sharing critical feedback because students do not feel worried about their comments on their
peers (Waemosa, 2017), and the students thereby are more honest in their critique and provide
feedback without bias regardless of attitude toward the experience.
Even though there have been numerous studies employing peer feedback review in the
Thai language classrooms (Sapkota, 2012; Sritrakarn, 2018; Wanchid, 2015), there has still
been insufficient research comparing the use of anonymity and non-anonymity. Especially, the
use of senior students instead of junior students has been very rare. To add more findings in the
previous studies, the present study employed anonymous and non-anonymous review activities
in an essay writing classroom. The study seeks to answer the research questions below.
1.
What are the effects of anonymity and non-anonymity on the quality of junior
students’ writing drafts?
2.
How does anonymity or non-anonymity affect the reviews made by senior students?
3.
What are senior and junior students’ attitudes toward the use of anonymity in peer
review?
Method
Senior students
In this study, junior students’ writing drafts were revised by the same group of senior students.
Initially, eight senior students (one male and seven females) were selected to provide feedback
on junior students’ essay writing. These students, majoring in English and were in their fourth
year, have passed the Essay Writing Course and were undertaking the Argumentative Writing
Course. By considering the potential problems regarding the non-systemic process of senior
students’ selection (see Sritrakarn, 2018), the senior participants in the current study were
selected using the purposive-sampling technique. In so doing, the students’ TOEIC score
reports were considered, and those who have passed the TOEIC at the minimum score of 480
were recruited as the research participants. Later on, one senior student opted out. So, there
were seven senior students (one male and six females) who have been selected as the research
participants.
9
EFLIJ Volume 25 Issue 4 July 2021
10
Junior students
The two groups of junior students (Group A and Group B) majored in English. Students in
Group A received feedback on their writing in anonymous mode while Group B students
participated in the non-anonymous mode, receiving comments on their work by identifiable
senior students. Twenty-three students (four males and nineteen females) in Group A enrolled
in the Essay Writing Course in the first semester of 2019, and twenty-six students from Group
B (two males and twenty-four females) enrolled in the second semester of the same year. The
two groups of students have passed the Paragraph Writing Course, and their average level of
English proficiency was pre-intermediate. During the research activity, some of the junior
students from both groups did not participate constantly. This resulted in the number of junior
participants in this study dropping to sixteen for anonymity (four males and twelve females)
and seventeen for non-anonymity (two males and fifteen females).
The data collection process
Groups A (Anonymous Activity)
Prior to the data collection, both junior and senior student groups were trained to get familiar
with the Microsoft word ‘review’ system. The knowledge of essay structure was revised to the
senior students, and they were trained about the feedback types and how to provide feedback
using the ‘review’ system on Microsoft word program. The junior students were also trained
on how to track the changes or comments given by senior students, as well as how to accept or
reject them using the ‘review’ system. During the training sessions, the teacher ensured that
both groups of students had plenty of practice to get used to the ‘review’ tracking system on the
Microsoft word program.
To collect the data, the teacher has made an announcement of the essay writing topic
(Should same-sex marriage be legalized?) to the junior group and stated the deadline to submit
the first drafts. After all junior students submitted their first drafts to their class email, the
teacher checked emails and edited the names of the students by changing them to pseudonyms.
The teacher then emailed the drafts to the class email of the senior group. The senior students
would, later on, check emails and select the drafts they were assigned to review and provide
feedback. After the reviewed drafts were returned back to the senior class’s email, the teacher
changed the draft names back to junior students’ original names and emailed the drafts back to
their class email. Junior students were told to return their edited drafts back on the deadline
before the teacher changed the edited drafts back to the pseudonym and forwarded them to
senior students for the second round of review. After junior students received the second
10
EFLIJ Volume 25 Issue 4 July 2021
11
comments, they edited the drafts before submitting the final drafts to the teacher for marking.
The data collection process can be summarized in Figure 1 below.
Figure 1. The review process (anonymous review)
Group B (Non-Anonymous Activity)
Junior students in Group B were also trained about the feedback types and negotiated about the
learning goals of the activity. This group of junior students was not trained about the ‘review’
system of the Microsoft word program because they participated in the non-anonymous mode.
However, the process of review, as well as deadlines for submission of each draft, were
identified, and students were advised that they could consult the senior students directly if there
were some questions or any further explanations needed regarding the reviews. The review
process can be summarized as follows.
Figure 2. The review process (non-anonymous review)
As shown in Figure 2, the junior students had the opportunity to further ask questions regarding
the comments or seek further advice from senior students during the review process of their
first and second drafts before submitting their drafts to the teacher. There was less involvement
from the teacher than the anonymous review during the editing process. However, students
were advised that they could always consult with the teacher if they were not certain about the
comments.
11
EFLIJ Volume 25 Issue 4 July 2021
12
Research instruments
In order to investigate the effects of anonymity and non-anonymity on students’ reviews and
writing corrections, the data were collected from three sources: students’ writing drafts, Likertscale questionnaires, and group interviews. The junior students’ writing drafts on the topic:
“Should same-sex marriage be legalized?” were collected. The drafts composed by those who
participated in the entire activity were used for data analysis.
After the review activity, the questionnaires (see Table 5 and Table 6) were distributed to
both junior and senior student groups to investigate their attitudes toward the activity. The
designed questionnaires were sent to three experts in the field for comments and rating for
content validity to validate the quality of the questionnaires prior to the distribution. Then the
index of item-objective congruence (IOC) was calculated. The reviewers were asked to rate
whether the question items were congruent (+1), incongruent (-1), or questionable (0). The
items that had scored lower than 0.5 were revised, and the items that had scores higher than or
equal to 0.5 were reserved. After changes had been made by the reviewers, they were piloted
with students who were at a similar level to the research participants. Then, the reliability values
of the two sets of questionnaires were calculated by using Cronbach’s alpha. In this study, the
two sets of questionnaires had reliability values of 7.5 (for senior students) and 7.7 (for junior
students), which were acceptable and could be used for data collection.
To ensure the validity of an interview protocol, an expert in the field was consulted for
comments and suggestions of how to make the question items precise and concise. Then
changes were made based on the comments. Finally, the senior students and the two groups of
junior students were interviewed to investigate their attitudes toward the activity and to follow
up as a clarification of some certain issues received from the findings (see Appendix A:
Interview Protocol). These questions were flexible to students’ responses in that further
questions were sometimes added and asked to clarify the initial response made by junior
students to each question.
Data analysis
The collected data were analyzed quantitatively and qualitatively to find the answers to the
research questions set above. To answer research question 1, the first and final drafts composed
by junior students were compared in terms of essay structure and changes made in their
responses to the reviews received from senior students. For research question 2, senior students’
reviews were investigated by classifying the types of feedback. Further analysis of comments
12
EFLIJ Volume 25 Issue 4 July 2021
13
made by senior students was done to see whether similar or different types of comments were
provided when senior students’ names were disclosed and hidden.
Results
Research question 1: What are the effects of anonymity and non-anonymity on the quality of
junior students’ writing drafts?
Taking a broad view, there was not much difference in the essay structures of the first and final
drafts composed by the two groups of junior students. In other words, all of the students were
aware of the essay structure and composed their drafts consisting of an introduction, body (with
arguments), and conclusion. Further analysis of the first and final drafts of junior students’
writing was made to identify the efficiency of their correction. In so doing, changes in the final
drafts as a result of senior students’ reviews in the first drafts were analyzed, and the findings
are demonstrated as follows.
Table 1
Successful responses to senior students’ comments
Total reviews
No.
383
%
100
Anonymous reviews
Successful
Unsuccessful
responses
responses
No.
%
No.
%
337
87.98
46
12.02
Total reviews
No.
586
%
100
Non-anonymous reviews
Successful
Unsuccessful
responses
responses
No.
%
No.
%
532
90.78
54
9.22
As shown in Table 1, while non-anonymity resulted in more corrections in junior students’ final
drafts, both types of review led to more successful corrections (87.98% for anonymity and
90.78% for non-anonymity). The two types of review also led to unsuccessful corrections, and
these were due to different reasons. The main reasons were tallied and calculated in percentages,
as shown in Table 2 below.
Table 2
Reasons for unsuccessful responses
Reasons
Ineffective feedback
Original language already correct
Junior students’ rejection
Junior students’ misunderstanding
Adjustment of reviews
Total
Anonymous review
17 (36.95%)
23 (50.00%)
6 (13.04%)
46
Non-anonymous review
27 (50.00%)
2 (3.70%)
18 (33.33%)
5 (9.25%)
2 (3.70%)
54
Table 2 shows that the unsuccessful corrections of junior students were influenced by different
reasons. When examined closely, however; most of the unsuccessful corrections were caused
by senior students (86.95% in anonymous review and 53.70% in non-anonymous review), by
13
EFLIJ Volume 25 Issue 4 July 2021
14
either providing ineffective feedback or making changes of the statements which were already
correct. While there were only three reasons which affected the unsuccessful corrections made
by junior students in the anonymous review activity, the reasons for unsuccessful corrections
in non-anonymity varied. This raises the question of whether or not the mode of review has the
effects on the review quality made by senior students. Further analysis would therefore be made
to examine the types and quality of reviews made in the two circumstances.
Research question 2: How does anonymity or non-anonymity affect the reviews made by
senior students?
To identify the effects of anonymity and non-anonymity on senior students’ reviews, the
researcher has initially analyzed the types of feedback made on junior students’ writing drafts.
Following Sritrakarn (2018), the reviews which affect direct responses or changes in junior
students’ final drafts were categorized into four types: direct feedback (when junior students
are provided with correct form), indirect feedback (when junior students are reminded that
errors exist but correction is not provided), metalinguistic feedback (when some kind of
metalinguistic clue as to the nature of the error is provided) and comments (other types of
feedback that do not fit in the previous three categories made by senior students and received
responses). The findings are presented in Table 3 below.
Table 3
Feedback types
Frequencies
Anonymous review (383 reviews)
Non-anonymous review (586 reviews)
Feedback types
Feedback types
Direct
Indirect
Metalinguistic
Comment Direct
Indirect
Metalinguistic
Comment
295
(77.02%)
1
(0.26%)
13
(3.39%)
74
(19.32%)
11
(1.87%)
11
(1.87%)
17
(17.17%)
547
(93.34%)
Table 3 shows that most of the reviews made by senior students, both anonymously or nonanonymously, are direct feedback. Especially for the non-anonymous mode, almost all of the
reviews were direct feedback (93.34%). In addition to the high frequencies of direct feedback,
there were significant numbers of comments made in the reviews. These comments shown in
the table (19.32% in anonymity and 17.17% in non-anonymity) are those which received
responses from junior students. There have been, however, more comments made by senior
students that did not receive explicit responses nor affect changes in the writing. These
comments may have an influence on the improvements of junior students’ writing in some way.
14
EFLIJ Volume 25 Issue 4 July 2021
15
Further analysis will therefore be made to investigate all the comments made by the senior
students. To do this, the researcher followed the analysis framework of Sritrakarn (2018), and
divided the types of comments into compliments, questions to the writers regarding the
contents, criticism, and suggestions of how the essay could be better constructed. The findings
from the analysis are shown in Table 4 below.
Table 4
Types of comments (anonymous)
Compliments or support
Anonymity
Total
anonymous
reviews = 99
Nonanonymous
reviews = 50
Nonanonymity
11 (11.11%) 10 (20%)
Comment types
Questions to the writer
Criticism
Suggestions of how the essay
regarding the contents
could be better constructed
Anonymity NonAnonymity NonAnonymity Non-anonymity
anonymity
anonymity
2 (4.00%) 7 (7.07%)
12 (24.00%) 81 (81.81%) 26 (52.00%)
Table 4 shows that senior students made more comments when they were anonymous, and most
of their comments were to provide suggestions of how the junior students could improve their
drafts (81.81%).
There must be infinitive verb after modal verb. (Narisa)
To avoid (using) the same word, this one you can cut ‘by’; and make prejudice to be
adverb ‘prejudicially’ (Chanchai)
The examples above show that senior students took more engagement in the reviewing process
and provided more explicit comments on junior students when being anonymous. Moreover,
the use of modalities (“must” or “can”) shows their confidence in delivering comments on
junior students’ writing. When their names were disclosed, the number of their comments
dropped to almost half of those made in the anonymous activity. Moreover, they made
comments by providing compliments, asking questions, and criticizing more often when being
non-anonymous.
Good job!! ☺ (Compliment- Papawadee)
And then what is next? (Asking question- Tad-dao)
This sentence is too long. (Criticism- Chanchai)
15
EFLIJ Volume 25 Issue 4 July 2021
16
Research question 3: What are senior students’ and junior students’ attitudes toward
anonymous and non-anonymous review activity?
The third area investigated in this study was the attitude of the participants toward the activity.
Questionnaires were distributed to senior students as well as the two groups of junior students,
followed by group interviews. The senior students’ attitudes are presented in Table 5 below.
Table 5
Senior students’ attitudes
Statements
1. It is useful to read and review the junior students’ work.
2. Reviewing junior students’ work helps me improve my own writing.
3. Reviewing junior students’ work helps me give some ideas for my own writing.
4. I am a better reviewer if I know whose work I am reading.
5. I am a better reviewer if I don’t know whose work I am reading.
6. It is difficult to find things to say about the junior students’ writing if my name is disclosed.
7. It is difficult to find things to say about the junior students’ writing even though it is anonymous.
8. In my non-anonymous review, I often tell the junior students about the strengths of their writing.
9. In my anonymous review, I often tell the junior students about the strengths of their writing.
10. In my non-anonymous review, I often tell the junior students about the weaknesses or problems of their
writing.
11. In my anonymous review, I often tell the junior students about the weaknesses or problems of their
writing.
12. In my non-anonymous review, I often tell the junior students if I don’t understand something they have
written.
13. In my anonymous review, I tell the junior students if I don’t understand something they have written.
Mean
4.71
4.43
4.71
3.71
3.29
3.00
2.71
4.86
4.29
4.43
S. D
0.49
0.53
0.49
1.25
1.50
1.41
1.50
0.38
0.76
0.79
4.43
0.79
5.00
0.00
4.29
0.76
As shown in Table 5 above, senior students had positive attitudes toward the activity. By
indicating the low mean scores in number 6 (It is difficult to find things to say about the junior
students’ writing if my name is disclosed.) and number 7 (It is difficult to find things to say
about the junior students’ writing even though it is anonymous.), it indicates that senior students
had no problems making reviews of junior students’ writing in either anonymous or nonanonymous mode. Even though they had no problems doing reviews with their names disclosed
or hidden, the results, however; show that senior students preferred to know whose work they
were reviewing as evidenced in their responses to number 4 (I am a better reviewer if I know
whose work I am reading. m = 3.71) and number 5 (I am a better reviewer if I don’t know whose
work I am reading. m = 3.29).
Junior students from both anonymous and non-anonymous groups had positive attitudes
toward the activity and found the senior students’ comments useful. Table 6 shows that the
anonymous group indicated a low mean score (m = 3.04) when being asked if they wanted to
know the name of the reviewer after receiving comments (no. 9). The non-anonymous group,
however, showed a high level of agreement on this point (m = 4.14). This shows that,
interestingly, each group preferred their way of receiving reviews.
16
EFLIJ Volume 25 Issue 4 July 2021
17
Table 6
Junior students’ attitudes toward anonymous and non-anonymous activity
Statements
1. I enjoy receiving the senior students’ comments on my writing.
2. I find the senior students’ comments helpful when I revise my
writing.
3. In their comments, the senior students sometimes point out
problems with my writing that I didn’t notice.
4. It is not useful if the senior students say only good things about
my writing.
5. I sometimes disagree with the senior students’ comments about
my writing.
6. The senior students’ comments help to make my ideas clearer
for the reader.
7. The senior students’ comments help me improve the
organization of my writing.
8. The senior students’ comments help me improve the language
of my writing.
9. When I receive comments on my writing, I prefer to know who
has written them.
Anonymous
Mean
S. D
3.96
0.36
4.08
0.65
Non-anonymous
Mean
S. D
4.36
0.49
4.36
0.66
4.04
0.69
4.41
0.80
3.83
1.05
4.09
0.81
3.25
1.03
2.77
1.07
4.00
0.66
4.00
0.76
4.08
0.58
4.23
0.75
4.04
0.75
4.29
0.81
3.04
1.40
4.14
0.99
Discussion
The findings from this study have provided some implications for language learning and
teaching in Thailand and other EFL learning contexts, in particular for the application of the
peer review activity. These implications will be pointed out and discussed below.
Anonymity has led to more constructive responses from junior students
As shown in Table 1 above, most of the unsuccessful corrections in both anonymous and nonanonymous activities were influenced by senior students, either providing ineffective feedback
or making changes to the statements which were already correct. When examined further into
the responses of junior students to senior students’ reviews in their second drafts, some in an
anonymous group were made to react, question back, or challenge the senior students, and these
actions were found in the drafts of junior students whose levels of proficiency were advanced
or moderate. To exemplify this, Student ‘N’ in Figure 3 below was suggested by a senior student
in an anonymous activity to combine two sentences. She challenged the senior student back by
asking for the reasons to do so.
17
EFLIJ Volume 25 Issue 4 July 2021
18
Figure 3. Sample of a junior student’s rejection and challenge to a senior student
Moreover, the advanced junior learners also rejected the comments or feedback made by senior
students more often than other students, and many of those rejections have led to successful
corrections. The findings in this area confirm the claims made in previous studies regarding the
quality of feedback made by student reviewers (Sritrakarn, 2018). The findings have also
reflected some implications for the activity. Firstly, the teacher has to ensure that the chosen
senior students have the capability to make reviews on junior students’ writing. Even though
the average score of the selected senior students was 480, the findings show that the level of
proficiency should be higher to reduce the chance to provide ineffective reviews. Secondly, the
use of review activity may be suitable to certain groups of junior students whose level of
proficiency was either advanced or moderate. For those who were poor students, the writing
drafts should be submitted to the teacher to provide feedback or comments as well as to avoid
confusion.
Anonymous review activity influences the senior students to provide more constructive
feedback
From the findings, even though the two kinds of review have led to high frequencies of direct
feedback made by senior students, when examined closely, senior students provided a greater
number of comments to help the junior student better construct their essays when they were
anonymous. When they could be identified, however, they shifted the strategies to provide
compliments or support, to question the writers, and to criticize the junior students’ writing
more often. This shows that anonymous review activity tended to provide senior students with
‘a comfort zone’ (Waemusa, 2017) in that when they provided feedback anonymously, they
may become less worried about offending others (Hosack, 2003). So, they were courageous to
take risks to provide direct suggestions when their names were hidden. When they could be
identified, however, they avoided the chance to provide wrong comments by asking questions
18
EFLIJ Volume 25 Issue 4 July 2021
19
or making criticism. In doing so, they allow junior students to take their own risks for
correction. Evidence of this could be found in the interview with senior students. Out of seven,
two senior students said that they preferred being anonymous, and one of the reasons was that
they did not want to lose their face if making wrong comments.
Being a senior student, I have to be careful not to make wrong comments and lose my
face. They could laugh at me if my name was disclosed. So, I tried to say less and
avoided making direct comments. (Tad-dao)
This finding supports the claim made by scholars in previous studies that Asian learners hold
their value of saving face (Raymond & Choon, 2017), so they tend to be compromising and
may not want to give comments sincerely for fear that it might cause (them or) their peers to
lose face (Kulwongse, 2013; Wanchid, 2015). This reflects the beliefs and attitudes of senior
students toward the activity, which could affect the opportunity of junior students in improving
their writing. Clear explanations of the goals for learning and exchanging collaboratively should
therefore be made at an early stage.
Training should be organized and activity goals should be negotiated explicitly with senior
students.
As shown in Table 6, even though most of the students had no problems providing feedback
either in anonymous or non-anonymous status, they have shown a high level of attitudes (m =
3.71) when being asked about their interest in knowing the names of the writers when making
comments (I am a better reviewer if I know whose work I am reading.). This reflects the
uncontrolled curiosity of senior students about the owners of the work. Evidence of this was
also demonstrated during the data collection stage when one anonymous-senior student posted
the writing drafts of junior students on social media (Facebook) and made complaints about
their writing problems. As the number of students in the context of the investigation was about
ninety-six from year 1 to 4, it was still a small community. The primary goal for posting the
writing drafts was, therefore, to trigger certain students who were also friends on social media
and to communicate the feedback directly to the virtual community. By doing this, it reflected
that some senior students did not seriously perform their roles of being anonymous, and this
could lead to some misunderstanding or offense between the two groups of students.
The findings also imply that training and negotiation of the activity goals should be made
at the beginning prior to the data collection. In particular, the concept of ‘deindividuation’
(situations in which individuals in groups stop thinking of other members as individuals and
19
EFLIJ Volume 25 Issue 4 July 2021
20
feel that they cannot be singled out by others- Chen, 2019) should be presented to ensure that
students did not take the review activity personally.
Furthermore, during the group interview, some senior students stated that they sometimes
wanted to communicate directly to the owner of the work as they needed the clarification of
some parts or wanted to explain some issue. Real communication, however, did not seem to
happen as junior students submitted their drafts twice, and the activity was ended. Due to
anonymity, both groups of students lack the opportunity to discuss the work or explain their
points. The finding from this study supports the claim made by previous studies on the
limitations of anonymous reviews regarding the lack of ongoing communication. Waemusa
(2019), for example, stated that anonymous peer review has a limitation in terms of ongoing
communication in that both reviewers and reviewees could not continue communication like
that in the face-to-face method. Further study may create a special channel, for example, a
Facebook group, where students use their pseudonyms to communicate. Alternatively, the
teacher may create a common email address that can be used by both senior and junior students
using their pseudonyms and let the students submit the drafts or return the comments without
teacher intervention. By allowing students to take control of their own reviewing process,
students may be able to manage the schedule and could seek further explanations from senior
students. This could thereby promote more ongoing interactions among the two groups of
learners.
Conclusion
This study investigated the effects of senior students’ reviews in anonymous and nonanonymous modes on the review quality and junior students’ corrections of their writing drafts.
The findings from this current study have contributed to previous studies and provided useful
implications on the implementation of review activity in the writing classroom. By
demonstrating how the nature of anonymity helps to promote more constructive responses from
junior students, the findings have also reflected the constructive nature of reviews made by
senior students in the same circumstance. Moreover, the attitudes of students (both senior and
junior groups) toward the activity also reflect their preferences in terms of learner identity
disclosure, which provides implications and raises awareness of the teacher when applying the
anonymous and non-anonymous review activity in other writing classrooms.
Some of the findings in this study, however, could be limited to the groups of learners
investigated in this context only. Different findings could be informed when applying the same
activity in other learning contexts. To confirm the findings from this present study, further
20
EFLIJ Volume 25 Issue 4 July 2021
21
studies could, therefore, be conducted to investigate the effects of anonymity and nonanonymity in different learning contexts. Moreover, as indicated in Table 6 that each group of
junior students preferred their individual way of receiving comments. It is important to note
here that the two groups of students participated in the consistent mode of review until the end
of the semester. These junior students did not have the chance to swap the modes of review to
experience receiving reviews in an alternative way. Their attitudes toward the activity could,
therefore, be based on a one-sided experience. To enable the more objective findings, junior
students should be exposed to both circumstances, followed by the questionnaire distribution
to compare their attitudes toward the two modes of activity.
21
EFLIJ Volume 25 Issue 4 July 2021
22
References
Akbari, M. & Sadeghi, R.M. (2013). Foreign language learning anxiety: The case of Iranian
Kundish-Persian Bilinguals. International Research Journal of Applied and Basic
Sciences, 4(0), 1-9.
Altun, M. & Sabah, R. (2020). The effect of cooperative learning strategies in the
enhancement of EFL learners’ speaking skills. Asian EFL Journal Research Articles,
27(2.3), 144-171.
Arjpru, T. (2017). The development of task-based speaking and communication strategies
instructional model to enhance speaking competence of 1st year EIC students. Rangsit
Journal of Educational Studies, 4(2), 72-89.
Baiyaem, S. (1997). Learner training: Changing roles for a changing world, educational
innovation for sustainable development. 3rd UNESCO-ACEID international
conference, Bangkok, Thailand.
Chen, C. (2019). Using anonymity inonline interactive EFL learning: International students’
perceptionsand practices. International Journal of Education and Development Using
Informationand Communication Technology, 15(1), 204-218.
Colpitts, B. (2016). Japanese students’ perceptions of peer corrective feedback in an EFL
classroom. Humanities Series, No. 49, 345-358.
Cote, A. R. (2014). Peer feedbackin anonymous peer review in an EFL writing class in Spain.
Gist Educationand Learning Research Journal, 9, 67-87.
Dordinejad, G.F. & Ahmadabad, M. R. (2014). Examination of the relationship between
foreign language classroom anxiety and English achievement among male and female
Iranian high school students. International Journal of Language learning and Applied
Linguistics World, 6(4), 446-460.
Foley, J. (2005). English…in Thailand. RELC Journal, 36(2), 233-234.
https://doi.org/10.1177/0033688205055578
Hosack, I. (2003). The effects ofanonymous feedback on Japanese university students’
attitudes towards peerreview. Ritsumeikanhougakubessatsu, 1, 297-322.
Houston, J. P. (2001). Fundamentals of learning and memory (4th ed.). Harcourt Brace
Jovanovich College Publishers.
Hyland, K. & Hyland, F. (2006).Feedback on second language students’ writing. Lang.
Teach., 39, 77-95.
Jessup, M.L., Connolly, T. &Tansik, A.D. (1990). Toward a theory of automated group work:
The deindividuatingeffects of anonymity. Small Group Research, 21(3), 333-348.
22
EFLIJ Volume 25 Issue 4 July 2021
23
Kulwongse, S. (2013). Thammasat Review 277Peer Feedback, Benefits and Drawbacks.
Thammasat Review, 16(3), 227-288.
Lu, R. & Bol, L. (2007). Acomparison of anonymous versus identification e-peer review on
college studentwriting performance and the Extent of critical feedback. Journal of
Interactive Online Learning, 6(2), 100-115.
Ngoc Lac, H.H. & Gurung, S. (2015). Using online peer feedback in writing class: An action
research study. Proceeding of the 2nd international VietTESOL conference:
Transforming English language education in the era of globalization, 285-290.
Noom-ura, S. (2013). English teaching problems in Thailand and Thai teachers’ professional
development needs. English Language Teaching, 6(11), 139-147.
Nur, S. (2020). Students’ perception toward the use of deductive and inductive approaches in
teaching English grammar. TESOL International Journal, 15(1), 6-19.
Piriyasilpa, Y. (2009). Thematic development in EFL student online discussion postings: A
case study. The New English Teacher, 3(1/2), 17-33.
Rajab, M.F., Alzeebaree, Y., & Zebari, HAI (2020). Effectiveness of English Language in a
Globalized World: EFL Teachers of Duhok University. Asian EFL Journal Research
Articles, 27(2.3), 38-47.
Raymond, C.Y.L. & Choon, TT (2017). Understanding Asian students’ learning styles,
cultural influences and learning strategies. Journal of Education & Social Policy, 7(1),
194-210. https://www.jespnet.com/journals/Vol_4_No_1_March_2017/23
Rayupsri, K. & Kongpetch, S. (2014). Implementation of the process-genre approach in an
English in a foreign language classroom in Thailand: A case study. RJES, 1(2), 32-53.
Robillos, J.R. & Phantharakphong, P. (2020). Enhancing EFL Learners’ argumentative
abilities in written composition and critical thinking dispositions through argument
mapping within metacognitive approach. Asian EFL Journal Research Articles, 27(3),
181-208.
Sapkota, A. (2012). Developing students’ writing skill through peer and teacher correction:
An action research. Journal of NELTA, 17(1-2), 70-82.
Silver, R. & Coomber, M. (2010). How anonymity affects feedback in the peer review
process. KOTESOL Proceedings: Advancing ELT in the global context, 299-308.
Sritrakarn, N. (2018). A comparison of teachers’ and senior students’ feedback: Students’
attitudes and their writing improvement. The Journal of Asia TEFL, 15(2), 329-348.
23
EFLIJ Volume 25 Issue 4 July 2021
24
Swain, M. (2005.) The output hypothesis: theory and research. In E. Heinkel (Ed.), Handbook
of researchin second language teaching and learning (pp. 471-483). Mahwah, NJ:
Lawrence Erlbaum Associates.
Sweeney, L. A.; Kunyot, T., & Preedeekul, A. (2017). Using task-based learning approach
and drama to improve communicative competence for students at Pibulsongkram
Rajabhat University. Journal of Faculty of Education Pibulsongkram Rajabhat
University, 4(2), 112-123.
Truscott, J. (1996). The caseagainst grammar correction in L2 writing classes. Language
Learning, 46, 327-369.
Vallente, P.J. (2020). Sources of embarrassment or empowerment? Oral feedback strategies in
English language teaching classrooms. TESOL International Journal, 15(1), 31-52.
Viriya, C. (2018). Using task-based learning with students of academic English. Arab World
English Journal (AWEJ), 9(4), 337-346.
Vygotsky, L. S. (1986). Thought and Language. Cambridge, MIT Press.
Wanchid, R. (2015). Different sequences of feedback types: Effectiveness, attitudes, and
preferences. PASAA, 50, 31-64.
Wiriyachitra, A. (2003). Thai teachers’ role in teaching: Motivating students and guiding their
learning. ThaiTESOL Focus, 16(2), 25-27.
Zakaria, K.Y.N. & Hashim, H. (2020). Game-based assessment in academic writing course
for pre-service teachers. TESOL International Journal, 15(1), 65-73.
24
EFLIJ Volume 25 Issue 4 July 2021
25
Appendix
Interview Protocol
I. Interview questions for senior students
1. Did you find the anonymous and non-anonymous review activities useful?
2. Which of the two review activities did you prefer?
3. If you were not clear about the writing, what did you do?
4. Did you find any obstacles or problems during the two review activities?
5. Would you like to make any comments or suggestions for the implementation of anonymous
and non-anonymous review activities in the future?
II. Interview questions for junior students (anonymous group)
1. Did you find the anonymous review activity useful?
2. How did the activity help you improve your writing?
3. If you were not clear about the comments, what did you do?
4. Did you find any obstacles or problems due to the anonymous nature of the review?
5. Would you like to make any comments or suggestions for the implementation of an
anonymous review activity in the future?
III. Interview questions for junior students (non-anonymous group)
1. Did you find the non-anonymous review activity useful?
2. How did the activity help you improve your writing?
3. If you were not clear about the comments, what did you do?
4. Did you find any obstacles or problems due to the non-anonymous nature of the review?
5. Would you like to make any comments or suggestions for the implementation of a nonanonymous review activity in the future?
25
EFLIJ Volume 25 Issue 4 July 2021
26
Comparing the Effectiveness of L2 Extensive and Intensive Listening
Approaches
Omar Karlin & Sayaka Karlin
Toyo University & Japan Women's College of Physical Education
Bio data
Omar Karlin holds an Ed.D. from Temple University. His doctoral dissertation examined the
validity of personality questionnaires in second language research, as well as the intersection of
personality and second language development. His current research interests include the
assessment of second language listening and improving listening ability. o_karlin@yahoo.com
Sayaka Karlin holds a M.Sc. in Education from Temple University, as well as a M.Sc. in
Economics from Manchester University. Her current research interests include the effect of
authentic materials on learner motivation, task-based learning, and English for Specific
Purposes. sayakakarlin@yahoo.co.jp
Abstract
In the fall semester of 2019, 237 Japanese university students were placed into three groups, 97
in an extensive listening group, 104 in an intensive listening group, and 42 in a control group.
Participants were given a 100-question TOEIC listening test to assess L2 listening proficiency,
and then placed into their respective group. During the 14-week semester, students completed
six listening assignments adhering to either principles of extensive or intensive listening (as
well as a control group which received no L2 listening homework). At the end of the semester,
students were given another 100-question TOEIC listening test. Results indicated that the mean
difference between the pre- and posttest scores was significantly larger for the EL group than
the IL group t(193) = 2.14, p <.05. Among the suggestions for future research are a codification
of EL and IL methodologies, a linkage of testing instruments to account for variability in testing
conditions, and greater scrutiny of the participants' adherence to the intervention.
Keywords: extensive listening, intensive listening, L2 listening, audiobook, TOEIC,
Correspondence address:
Department of Sociology
Toyo University
5-28-20 Hakusan, Bunkyo-ku, Tokyo
〒112-8606
independent-sample t-test
Introduction
26
EFLIJ Volume 25 Issue 4 July 2021
27
When one thinks of the four L2 language skills of speaking, listening, writing, and reading,
listening often seems to be regarded as the least important (Vandergrift & Goh, 2012), and
usually the weakest skill for language learners (Vandergrift, 2013). To illustrate, in Japan,
English components of university entrance exams only began to include a listening assessment
as recently as 2006 (Saito, 2019), and high school English classes still focus predominately on
grammar-translation methodologies with minimal focus on listening skill development
(Asquith, 2014). Further, many classes in Japanese junior and senior high schools that are taught
by non-native speakers do not use English as the language of instruction, denying students
listening practice opportunities (Saito, 2019). Finally, textbooks are, by their nature, rife with
reading practice opportunities, but not all come with an audio component for listening practice.
Using the Japanese context, if one tallies the lack of institutional support of listening skill
development stemming from grammar-translation focused testing, the language of instruction
of the majority of teachers, and the inherent limitations of many textbooks, it is clear that
listening is prioritized below that of speaking, writing, and reading. Yet in reality, listening may
be the most important of the four skills (Flowerdew & Miller, 2010; Vandergrift, 2013;
Vandergrift & Goh, 2012). When one speaks or writes in an L2, the speaker or writer has agency
and can determine the complexity and difficulty of the language that they generate. However,
when someone is engaged in an interaction with an interlocuter, they have very little control
over the complexity and difficulty of the language that they receive from their interlocuter
(Vandergrift, 2013). Ultimately, their continued participation in the interaction relies less on
the effectiveness of their speaking than on the effectiveness of their listening, in that if they can
only produce simple sentences, the interaction can probably continue but if they can only
understand simple sentences, it is likely that the interaction will not be understood (especially
if the interlocuter is unwilling or unaware that they need to simplify their output). Reading, to
a lesser extent, also involves agency in that someone must resolve themselves to reading a text,
whereas listening is often experienced whether one wants to engage in a listening task or not.
For example, one might choose to not listen to train announcements, the conversations around
them, or television commercials, yet these sounds are still experienced nonetheless. While these
sounds can be ignored, short of wearing earplugs, they are still heard. In this regard, the
exposure to listening is the most penetrative of the four skills. Considering the importance of
listening competence in interactions, and the pervasiveness of listening opportunities, whether
sought or not, listening could be viewed as the most important of the four skills. Yet, listening
is rarely taught in language classrooms (Schmidt, 2016). Instead, what teachers may perceive
to be the teaching of listening, such as asking students comprehension questions to listening
27
EFLIJ Volume 25 Issue 4 July 2021
28
passages, is actually not the teaching of listening but, rather, the assessing of listening (Brown,
2011; Schmidt, 2016; Vandergrift & Goh, 2012). This paper will explore possibilities on how
to more effectively teach listening in (and out of) language classrooms.
Literature Review
Among the topics to be reviewed in this section are differing methods of listening instruction
(such as metacognitive, top-down, and bottom-up instruction), and how to define and
operationalize some of the concepts (such as extensive and intensive listening) associated with
these types of listening instruction. In the latter half of this section, relevant research that
showcases these concepts will be highlighted, and a possible gap in the research, which this
study hopes to address, will be explained.
The teaching of listening in language classrooms is evolving towards more effective
methods of instruction than in the past (Flowerdew & Miller, 2010). While it has been suggested
that many listening teachers are just assessing listening, rather than teaching listening (Brown,
2011; Schmidt, 2016; Vandergrift & Goh, 2012), Brown (2011) suggests that some teachers
have adopted teaching methodologies for listening instruction similar to those of reading
instruction, with a pre-listening, listening, and post listening lesson format. Others have
suggested that the key to improved L2 listening instruction lies in greater use of metacognitive
strategy instruction. Because of the effectiveness of metacognitive strategy instruction in
general L2 learning, the possibility exists that metacognitive strategy instruction is similarly
effective in L2 listening classes (Goh, 2008; Tanewong, 2019; Vandergrift, 2013; Vandergrift
& Goh, 2012). Finally, in addition to comprehension-oriented (top-down) activities often used
for listening assessment, in which understanding the meaning of the listening is prioritized,
Brown (2011) indicates that there has been increasing momentum in the field towards bottomup and blended approaches to listening instruction.
Bottom-up processing focuses on units of sound, and the combination of these sounds into
a comprehensible message (Vandergrift, 2013). With bottom-up processing, listeners rely on
their phonological, lexical, and syntactic knowledge to construct meaning (Flowerdew &
Miller, 2010; Rost, 2016), however bottom-up processing on its own might not be enough for
comprehension as listeners cannot keep up with the sound stream (which is the figurative stream
of sounds one would hear in an utterance; Vandergrift, 2013; Vandergrift & Goh, 2012).
Keeping up with the sound stream can be especially difficult for learners with lower L2
proficiency as they have to manage three tasks simultaneously; identify the boundaries between
28
EFLIJ Volume 25 Issue 4 July 2021
29
words, identify the L2 vocabulary in the listening passage, and ensure that they are not using
L1 recognition habits when trying to decode an L2 listening passage (Vandergrift, 2013).
The blended approach advocated by Jack Richards (2005) is a blending of what he terms
listening as comprehension (or meaning-focused learning) and listening as acquisition (or formfocused learning). In Richards' article, he gives examples of activities that can be used in
listening as comprehension situations, such as sequencing tasks, picture identification tasks,
and summaries, as well as examples of activities that can be used in listening as acquisition
situations, such as identifying differences between spoken and written text, cloze exercises, and
dictation (Richards, 2005). Similarly, others have also endorsed a blended top-down and
bottom-up focus for listening instruction (Brown, 2011; Flowerdew & Miller, 2010; Graham,
2006; Rost, 2016; Schmidt, 2016; Vandergrift, 2013; Vandergrift & Goh, 2012).
In Japan, educational approaches have tended to favor traditional, teacher-fronted
methodologies such as grammar-translation teaching (Mitchell, 2017; Saito, 2019), which give
students little chance for communicative or cultural learning. In addition to historical traditions,
a reason for this adherence to grammar-translation teaching is the influence of Japan's National
Center Test, a standardized test used by hundreds of Japanese post-secondary institutions to
assess student ability in a variety of school subjects, including foreign language. In the past, the
foreign language component of the National Center Test has gravitated towards grammartranslation types of questions (for example, multiple choice grammar questions). In 2006, the
test was modified to include a listening component in an effort to generate a positive washback
effect for more communicative language teaching in Japanese high schools (Saito, 2019), yet
despite this aim, there has been little progress in changing the teaching culture of foreign
languages in Japanese high schools (Asquith, 2014; Mitchell, 2017; Saito, 2019). The Japanese
government is eager to change the dominant teaching methodologies of high school to a more
communicative approach, in part, because of the increased prevalence of English-medium
instruction (hereafter referred to as EMI) at Japanese universities. Recently, EMI has increased
dramatically at Japanese universities for a variety of domestic (falling birth-rate and greater
competition for international students) and international (to produce more globally-mobile
graduates) reasons (Shimauchi, 2018). Despite the greater need for L2 listening fluency at
Japanese universities because of an increased emphasis on EMI, L2 listening at Japanese
universities has tended towards assessment through the asking of comprehension questions,
rather than instruction through alternative pedagogical approaches. In a study by Siegal (2014)
involving ten EFL instructors in Japan, in which specific instances of listening pedagogy were
coded and recorded, approximately 70% of L2 listening instruction instances were based on
29
EFLIJ Volume 25 Issue 4 July 2021
30
comprehension questions while only 15% of instances followed a bottom-up approach to L2
listening instruction (with the remaining 15% of instances split amongst other pedagogical
approaches to L2 listening, such as teaching metacognitive strategies, teacher modelling, and
predictions). In sum, listening instruction in Japan has tended towards traditional approaches
(such as grammar-translation and comprehension assessment) over emerging forms of
instruction (such as communicative language teaching or bottom-up instruction).
This paper will focus on two methods of L2 listening instruction that are closely aligned
with top-down and bottom-up orientations, namely extensive and intensive listening.
Specifically, this paper will examine how homework utilizing each of these listening concepts
can be used to develop L2 listening skills.
Extensive and intensive listening
Within the top-down and bottom-up approaches to listening instruction are types of listening
practice, extensive and intensive listening, that incorporate the principles of top-down meaningfocused learning and bottom-up form-focused learning.
Extensive listening is a relatively recent practice in language teaching, and while it lacks
a formalized framework or sizeable canon of existing research, some have used the principles
of extensive reading (ER) as a guide for implementation (Chang, Millett, & Renandya, 2018;
Day & Bamford, 1998; Mayora, 2017). While there are variations on how to define extensive
reading, one of its chief principles is “an approach to language teaching in which learners read
a lot of easy material in the new language” (Bamford & Day, 2004, p. 1). Defining what is easy
is central to ER, and the Extensive Reading Foundation (2011), defines extensive reading as
requiring 98% comprehension of a text. In their meta-analysis, Jeon and Day (2016) also note
that another key feature of ER is that readers self-select their materials. In addition to the
easiness and the self-selection of the material, Day and Bamford (2002) stress that there are
three other necessary conditions for ER; readers be free to read as much as they can or want,
that reading be done individually, and that teachers guide readers. Additionally, Jeon and Day
(2016) have suggested that the most pronounced benefits of ER can be seen in adults as opposed
to children, EFL situations as opposed to ESL situations, web-based materials as opposed to
paper-based materials, and when established as part of a curriculum as opposed to being outside
of the curriculum. Collectively, the ideal conditions for ER can be summarized as being easy,
offering freedom of choice, being done individually yet within a curriculum, utilizing webbased resources, and involving adult learners in an EFL context. These conditions can be used
to construct a framework for extensive listening.
30
EFLIJ Volume 25 Issue 4 July 2021
31
Among the conditions outlined in the previously mentioned ER guidelines (Day &
Bamford, 2002) and meta-analysis (Jeon & Day, 2016), an EL (extensive listening) framework
can be constructed around several widely-agreed upon principles. Much like ER, EL is usually
based upon students being exposed to a large amount of easily comprehensible input that they
can enjoy listening to, in the form of movies, TV shows, podcasts, audiobooks, and radio
(Chang, Millett, & Renandya, 2018). However, the Extensive Reading Foundation notes a key
difference in that reading comprehension and listening comprehension usually do not equate in
terms of difficulty level, with listening comprehension often below that of reading
comprehension (The Extensive Reading Foundation, 2011, hereafter ERF). The ERF suggests
that when using graded readers and their accompanying audiobook forms, text that is to be read
is usually one or two levels higher than text that is to be listened to (2011), with the disparity
between reading rates and listening rates most pronounced amongst lower-level students (Hirai,
1999). The reason for this disparity between reading and listening rates may be due to
listening’s continuous, variable, and transitory nature. Van Zeeland (2014) describes
continuous as words blending into each other without a clear start and finish point; variable as
the same words pronounced in different ways; and transitory as words being ephemeral and
requiring quicker processing. Van Zeeland's (2014) study indicated word listening inference
rates of 35.6% versus reading inference rates in existing research closer to 50%, which is
commensurate with the disparity suggested by the ERF. Continuing with a possible framework
for EL, Mayora (2017) defines EL along five broad principles, namely quantity,
comprehensibility, learner-centeredness, meaning orientation, and accountability. With
Mayora's construct of EL in mind, researchers must be careful of constructing strawman
definitions of IL (intensive listening), which are based upon unfavourable distinctions with EL.
For instance, it is commonly agreed upon that ER (and by extension, EL) is characterized by
reading (or listening to) large amounts of easily comprehensible material, with a focus on
meaning. As a result, Mayora's first (quantity), second (comprehensibility), and fourth
(meaning-orientation) principles of EL can be safely adopted, with IL offering a contrasted
option of less, more difficult to comprehend material, with a focus on form. However, the
strawman danger lies in the third (learner-centeredness) and fifth (accountability) principles of
Mayora's EL construct. In Mayora's article, it is noted that locus-of-control is a powerful
motivator of students, and that EL can offer students this locus-of-control as they self-select
listening materials. However, with the abundance of digital materials available online, can't IL
also offer students locus-of-control? Conceptually-speaking, if IL is viewed as how intensively
students interact with listening texts, this can be accomplished through more difficult texts and
31
EFLIJ Volume 25 Issue 4 July 2021
32
accompanying assignments, and is not contingent upon the distribution of texts (such as an open
library versus teacher-assigned texts). Simply put, if the third EL principle of learnercenteredness allows EL to be conducted out of the classroom and offer students locus-ofcontrol, IL can as well, as long as students are interacting with the text in an intensive fashion.
The conception of intensive studying should not be inherently connected to in-class studying
administered by the teacher, and should be afforded the same opportunity to utilize the vast
expanse of online digital resources that extensive listening does. Further, the fifth principle of
accountability can also be equitably applied to both EL and IL. In fact, there is some
disagreement as to how intrusive EL tasks should be, with the worry that demanding tasks could
diminish the enjoyment (and by association, the frequency) with which students engage in EL
(which is suggested in Day and Bamford's guide for extensive reading; 1998). Yet, there is
widespread consensus that IL must be accompanied with form-focused activities. It is important
to consider these issues of discrimination when constructing an EL vs IL dichotomy, and for
the purposes of this paper, any areas in which one methodology is defined favourably at the
expense of the other, has been minimized. In sum, an EL framework can be largely constructed
around the extension of ER's principles, in that practitioners be exposed to large amounts of
easily comprehensible input, with a focus on meaning. Yet it is important to remember that
listening comprehension is usually below reading comprehension, and that suggestions of
learner-centeredness and supplementary tasks should not be exclusively the domain of EL.
In contrast to EL, IL involves close listening for precise units of sound and words (Rost,
2016) at a higher difficulty threshold, with the ERF advocating 90% rather than 98%
comprehension for intensive reading activities. Another key difference between EL and IL is in
the supplementary activities. With ER (and EL), Day and Bamford (1998) suggest that students
be primarily focused on the input of as much text as possible, and not be burdened with
exercises and comprehension questions. With IL, supplementary activities, such as dictation,
are necessary and vital for permanent language acquisition (Rost, 2016). It is important to note
that some of the previously mentioned ideal conditions for ER (and by extension, EL) from
Jeon and Day's meta-analysis, such as offering learners freedom of choice, being done
individually yet within a curriculum, utilizing web-based resources, and involving adult
learners in an EFL situation, can also be applied to IL, and are not necessarily the exclusive
purview of EL. With that said, the chief distinctions between EL and IL should be in the areas
where one methodological approach is not biased over the other, such as the difficulty of the
text, the frequency of listening, the closeness of listening, and supplementary activities. The
distinction between EL and IL outlined by Chang (2012) follows similar guidelines, suggesting
32
EFLIJ Volume 25 Issue 4 July 2021
33
that EL is abundant, easy-to-comprehend, and unburdened with supplementary tasks, while IL
involves a closer listening of more difficult texts, with phonological-based supplementary
activities. Additionally, the ease of comprehension for EL should result in a large amount of
material listened to without the need for repetition, while the more difficult nature of IL and the
phonological-based supplementary activities necessitates repetition of the text. While 98%
comprehension is suggested for EL, and 80-90% for IL, the repetition of IL texts will likely
result in increased comprehension with subsequent listening (Cervantes & Gainer, 1992;
Iimura, 2007; Sakai, 2012).
Chang (2012) distinguished between EL and IL by drawing upon the suggestions of Field
(2008), Renandya (2011), and Waring (n.d.), which is shown in Figure 1.
Extensive listening
•
Intensive listening
Listening to (or being involved in)
•
Listening for specific information
•
Listening for the exact words of a phrase
massive amounts of text
•
Text which learners can understand
reasonably smoothly
or expression
•
High levels of comprehension
•
Listening for details
•
Listening without being constrained by
•
Listening to mimic a text
pre-set questions or tasks
•
Listening at or below one's comfortable
fluent listening ability
Figure 1. Differences between EL and IL (Chang, 2012)
EL and IL research
While ER is associated with a significant body of research, EL is relatively under-researched
(Chang, 2012; Chang & Millett, 2016). Of the limited research involving EL and IL, many
studies tend to focus on either EL or IL, with very few examples of direct comparison between
the differing methodologies.
One study focused on EL (Chang & Millett, 2016) suggested that when EL is
supplemented with ER, significant listening gains can be observed in just 15 weeks. In the
study, 76 Taiwanese university students read and listened to one graded reader per week, over
a 15-week semester. Those that completed more follow-up assignments, which constituted
answering 200 comprehension questions per graded reader, scored significantly higher on the
listening component of a TOEIC practice test, as well as on teacher-constructed assessments,
33
EFLIJ Volume 25 Issue 4 July 2021
34
leading the authors to advocate for a combined ER and EL approach with many supplemental
activities. Worth noting is that treatment groups were determined based on the number of
assignments that students completed over the semester, so students who completed the most
assignments were assumed to have received the most treatment, thereby resulting in the highest
listening scores. While completing more comprehension assignments may have led this group
to higher listening scores, other factors such as greater self-efficacy, confidence, and motivation
could have also played a role. Another study that advocated for a combined ER and EL
approach (Chang, Millett, & Renandya, 2018) compared the listening gains of 69 Taiwanese
college students split into three treatment groups; an ER only group, an EL only group, and a
combined ER and EL group. The results of this study indicated that a combined ER and EL
treatment yielded the greatest increase in listening comprehension scores on a teacher-created
listening comprehension test. However, it should be noted that the combined ER and EL group
received twice as much listening practice time as the other two groups, which may have led to
their higher listening comprehension scores. Nevertheless, both studies suggest that EL is best
when done simultaneously with ER.
In one study focused on IL (Siegal & Siegal, 2015), 44 Japanese university students were
placed into two groups, a treatment group receiving bottom-up instruction with six unique
bottom-up activities, and a control group that received regular English instruction with no
emphasis on bottom-up instruction. Results indicated that the treatment group outperformed the
control group on a dictation test and on the listening component of the CASEC English
proficiency test, suggesting that some degree of bottom-up instruction can be beneficial to
students. The CASEC English proficiency test has four parts, two vocabulary and two listening,
and is taken online. Usually lasting about 40-50 minutes, the website platform uses item
response theory to continuously adjust the difficulty of the questions in order to match the testtaker's ability level, thereby creating a more efficient testing instrument. The two listening
components of the CASEC test focus on general (17 multiple-choice questions) and specific
(11 dictation questions) comprehension of a listening passage (CASEC, n.d.).
To date, there have been very few studies that have directly compared the effectiveness of
EL against IL in developing L2 listening ability. The first noteworthy study in this area was
conducted by Chang (2012), and involved 55 Taiwanese adult students. In the study, the EL
group (n = 31) listened to 15 audiobooks, while the IL group (n = 24) listened to three
audiobooks over the course of a 26-week period. Both groups received several interventions
from the teacher during class time, such as teacher explanations, vocabulary instruction,
comprehension checks, and, solely for the IL group, dictation tests. Participants also had
34
EFLIJ Volume 25 Issue 4 July 2021
35
repeated listening opportunities, as well as the opportunity to read their corresponding graded
reader in tandem with listening to the audiobook. Results indicated significant gains by the EL
group on a post-intervention listening test (a modified TOEIC listening test) and by the IL group
on a vocabulary test (VLT and AWL), however there are several caveats with this study that
make it difficult to generalize these results with other EL and IL comparison studies.
Specifically, the graded readers were assigned to both groups, minimizing the autonomy of
choice that is often a hallmark of ER programs (and by extension, EL). Moreover, the three
audiobooks used by the IL group were also studied by the EL group (in addition to 12 other
texts that were mostly easier). Typically, one of the conditions of EL is that texts be easy, yet
if the IL group was using books that EL group was also using, there would be less differentiation
in this aspect of the intervention. Lastly, with regard to the texts, there is conflicting research
on the pace with which learners should advance in an ER program, with some suggesting that
students read five books per level before moving up to the next level (Nation & Wamg, 1999)
while others suggest that even 10 books per level may not be enough for lower-level students
(Webb & Chang, 2015). In the Chang (2012) study, it is acknowledged that the learners'
“English competence was low” (p. 32), which would presumably mean in an ER program they
would read at least five, but likely 10, graded readers before moving up, yet the EL
intervention's progression of audiobook texts on the Oxford Bookworms series scale was as
follows: levels 2, 1, 1, 1, 5, 1, and 2 in the first half of the research period, and levels 2, 2, 3, 4,
2, 3, 3, and 4 in the second half of the research period. It must be noted that books 1, 11, and
14 were not from the Oxford Bookworms series but their respective difficulty was equated to
the Oxford Bookworm levels of 2, 4, and 3 (these books appeared to be from the Cambridge,
Scholastic, and MacMillan series, respectively). The relatively quick advancement of students
to more difficult texts would not seem to suggest a true EL intervention, and may, at least in
terms of the text difficulty level, be similar to that of an IL treatment (which is underscored by
the EL group using the same texts as the IL group). In addition to this, ER (and EL) is predicated
on participants choosing their own texts and reading at their own pace, with little teacher
intervention, however the texts in this study were chosen for the EL group, listening was done
during class at a fixed pace for everyone, and teacher instruction was frequent. Again, this
intervention would seem to be more congruous with traditional intensive methodologies, with
little student autonomy and a sameness of experience for all. It should be noted that this paper
will advocate that EL and IL follow a similar protocol in the area of learner-centeredness, with
both the EL and IL groups enjoying autonomy and a uniqueness of experience. Recalling
Mayora's (2017) five principles of EL of quantity, comprehensibility, learner-centeredness,
35
EFLIJ Volume 25 Issue 4 July 2021
36
meaning orientation, and accountability, the boundaries between the EL and IL interventions
in the Chang (2012) study are not easily discernable in the areas of quantity (in terms of inclass listening time), comprehensibility, and learner-centeredness, so while this study offers
some valuable insights into the direct comparison of EL and IL, there are caveats that must be
acknowledged that mitigate the generalizability of the findings.
Similarly, another study with several caveats that directly compared EL and IL
methodologies in developing L2 listening ability was conducted by Karlin and Karlin (2019).
This study involved 80 first-year Japanese university students, who were split into EL and IL
groups, with each receiving listening homework that adhered to either EL or IL conventions.
Results indicated that the IL group significantly outperformed the EL group on a listening exam,
which included seven 5-minute listening passages, with six comprehension questions per
listening passage (three focused on general comprehension, and three on specific details from
the listening passage). Among the chief caveats with this study is a research design which
utilized an unbalanced intervention between the groups. While the IL group was prescribed an
intervention that required several hours of listening practice every two weeks in order to
complete a dictation assignment, the EL group was prescribed an intervention that required a
much smaller amount of time to complete. The reason for this discrepancy was that the
participants in this study were a sample of convenience, from two different listening classes,
conducted by two different teachers, with two different views on assigning homework. While
the teacher coordinating the EL group assigned homework that most would characterize as
exceedingly easy, the opposite was true of the teacher coordinating the IL group, indicating a
bias in the interventions upon which this study was based. Future research will need to ensure
that the EL and IL interventions are sufficiently differentiated (unlike the Chang study), and
that EL and IL interventions are unbiased and roughly equal in the burden placed upon students
(unlike in the Karlin and Karlin study).
While not measuring L2 listening ability, Ahmadpour and Asadollafam's (2018) study
involved 60 Iranian EFL learners and compared extensive and intensive listening approaches
in developing verb tense usage proficiency. The authors concluded that both EL and IL helped
develop proficiency in verb tense usage, however this finding comes with a few caveats. First,
the control group, which received no special treatment, improved in almost equal measure as
the intervention groups. Second, the authors concluded that there was no significant difference
between the mean change of the EL and IL groups' test scores, thus no difference in the
effectiveness of EL and IL at developing verb tense usage proficiency. However, the mean
difference in favor of the IL group almost reached significance with a p value of .059 suggesting
36
EFLIJ Volume 25 Issue 4 July 2021
37
that if the research period was slightly longer or had slightly more participants (which was
already quite low), significance could have been achieved. Further, when the mean difference
of the two treatment groups were compared with the control group, the IL group scored
significantly higher than the control group, while the EL group did not, which can be viewed
as an indirect measure of a significant difference between the EL and IL groups. Finally, when
making direct comparisons between the two interventions, it is essential that both interventions
exert an equal burden upon participants, in order to avoid bias. It seems, according to the
authors, that the EL group's intervention was conducted both at home with homework and inclass with teacher instruction, yet the IL group's intervention was only conducted in-class, with
no mention of a homework assignment to equate the study-time with the EL group, seemingly
putting the IL group at a disadvantage. Despite this, the IL group significantly outperformed
the control group while the EL group did not. These findings are consistent with Chang's (2012)
study that indicated an IL group outperformed an EL group in the area of vocabulary
acquisition. In sum, while limited research favours EL in terms of L2 listening ability
development, either explicitly or implicitly when considering research design bias (Chang,
2012; Karlin & Karlin, 2019), other areas of language development, such as vocabulary
acquisition and correct verb tense usage, seem to favor IL (Chang, 2012; Ahmadpour &
Asadollafam, 2018).
While not a direct comparison between EL and IL, Chang (2010) compared EL with a
methodology she terms as formal instruction, which is modelled after traditional teacherfronted foreign language teaching. Involving 92 Chinese college students, Chang measured the
L2 listening proficiency and anxiety of groups receiving an EL or formal instruction
intervention over a 1-year research period. Results indicated that both groups improved in
measures of L2 listening proficiency, namely a narrative listening test and a conversational
listening test, and the EL group increased significantly in listening anxiety. It should be noted
that the EL group improved to a greater degree than the formal instruction group in measures
of L2 listening proficiency, with mean gains of 7.87 and 5.79 on the narrative and
conversational listening tests, respectively. The gains for the formal instruction group were 2.66
and 4.21 on the narrative and conversational listening tests, respectively. It should also be noted
that there appears to be a mistake in the text of the article, suggesting the opposite findings,
indicated in the study's Table 1, with the formal instruction group out-performing the EL group
on the conversational listening test, with mean scores of 5.79 and 4.21, respectively. Chang
concluded that the surprising increase in anxiety for the EL group may have been due to the
novelty of EL learning activities, in which traditional educational tools of tests and grades are
37
EFLIJ Volume 25 Issue 4 July 2021
38
minimized, which might have caused apprehension in students who may have felt as if they
were doing something against educational norms. While the formal instruction group was not
explicitly identified as utilizing an IL methodology, on the surface, there appeared to be
similarities with traditional interpretations of intensive learning in that there was a focus on
teaching vocabulary and grammar, regular testing, and a shared text and learning pace amongst
all students. Despite this seeming affirmation of EL, this study has some of the same caveats
identified in the Chang (2012) and Karlin and Karlin (2019) studies. In the case of the former,
the specific nature of the EL intervention is not clearly defined in the article, and is described
as the class using the same 25 audiobooks (it is unclear if these texts were utilized individually
or collectively, in class or at home) with some teacher-led discussion and comprehension
testing. In the case of the latter, there is an acknowledgment that a convenience sample was
used involving different instructors with different degrees of strictness and leniency, which may
have resulted in unintended bias within the sample. However, Chang does indicate that both
groups received about 100 minutes of listening instruction per week, suggesting an effort was
made to minimize bias resulting from time inequalities. Results were consistent with Chang's
later study in 2012, which affirmed that EL was more effective at boosting L2 listening
proficiency when compared to other methodologies (formal instruction in this case, and IL in
the 2012 study).
Research question
To what extent does EL and IL homework improve the L2 listening comprehension of Japanese
university students?
Participants
Initially, 398 students participated in this study (246 female, 152 male), drawn from 16 classes
taught by the researchers. Of these students, 234 were first-year students, 149 were second-year
students, nine were third-year students, and two were fourth-year students. Additionally, four
Masters students participated in this study. The year of study, gender, and proficiency levels
for participants is shown in Table 1.
Table 1
Proficiency, gender, and age of initial participants, by class.
Class
Proficiency
University
n
Male students
Female students
Year of study
Avg. age
1
Intermediate
One
28
18
10
2
20.07
38
EFLIJ Volume 25 Issue 4 July 2021
39
2
intermediate
one
29
7
22
2
19.72
3
beginner
one
27
1
26
2
19.96
4
beginner
one
25
13
12
1
18.88
5
beginner
one
28
17
11
1
18.71
6
advanced
three
23
11
12
varied
20.04
7
advanced
three
12
4
8
varied
21.25
8
intermediate
two
21
6
15
1
18.67
9
intermediate
two
22
10
12
1
18.73
10
advanced
three
24
9
15
1
18.96
11
advanced
three
25
11
14
1
18.80
12
advanced
one
25
5
20
2
20.32
13
beginner
one
26
12
14
2
19.92
14
advanced
one
25
8
17
1
19.52
15
intermediate
one
28
11
17
1
18.75
16
beginner
one
30
9
21
1
18.53
398
152
246
Varied
19.37
Totals
The 16 classes used in this study were from three major Tokyo-area private universities, and as
a result, there were curriculum differences amongst the classes, yet all followed a
communicative teaching methodology. For university one, classes met once a week for 90
minutes, and used a four-skills textbook to learn and practice new grammar in communicative
situations, such as role-plays and conversations. Students in the classes at university one
completed the textbook's online homework component every two weeks, and had a speaking
assessment at the end of the semester. For university two, classes met once a week for 200
minutes, and used a four-skills textbook to learn and practice the new grammar in
communicative situations, such as role-plays and conversations. Students in the classes at
university two did not have an online homework component, however, during the semester they
completed five quizzes, two group presentations, and a final exam, all based on the textbook's
content. For university three, classes met once a week for 90 minutes, and students were drawn
from three types of classes, one that used a four-skills textbook to learn and practice the new
grammar in communicative situations, one that used a reading textbook to practice reading
skills and discuss questions related to each chapter's reading passages, and one that used a
writing textbook to clarify and practice the components of an academic essay. The first class
had online homework every two weeks and two written exams, based on the course textbook.
The second class had weekly vocabulary quizzes, a small online homework component, and
two exams during the semester. Finally, the third class had regular writing and revision
39
EFLIJ Volume 25 Issue 4 July 2021
40
assignments in which they would practice the various parts of a standard academic essay.
Despite the differences in curriculum, the classes were similar in that they were communicative
in nature and the workload associated with the homework was roughly equivalent amongst all
classes.
Due to attrition, the final number of participants in the analysis was reduced to 237 from
398. The chief reason for the attrition of participants was not closely adhering to the study's
research interventions (explained further in the analysis section). The final tally of participants
is shown in Table 2.
Table 2
Proficiency, gender, and age of final participants, by class.
Class
Proficiency
University
n
Male students
Female students
Year of study
Avg. age
1
Intermediate
One
18
10
8
2
20.00
2
intermediate
one
24
5
19
2
19.71
3
beginner
one
16
0
16
2
19.81
4
beginner
one
11
5
6
1
19.00
5
beginner
one
13
9
4
1
19.08
6
advanced
three
13
4
9
varied
19.54
7
advanced
three
11
3
8
varied
21.00
8
intermediate
two
8
1
7
1
18.25
9
intermediate
two
14
8
6
1
18.79
10
advanced
three
18
5
13
1
18.94
11
advanced
three
21
8
13
1
18.81
12
advanced
one
14
3
11
2
20.21
13
beginner
one
18
8
10
2
20.00
14
advanced
one
18
6
12
1
19.61
15
intermediate
one
10
3
7
1
18.40
16
beginner
one
10
1
9
1
18.30
237
79
158
Varied
19.41
Totals
Instrument
This study used an assessment based on the listening component of the TOEIC proficiency test.
The TOEIC test's listening component is comprised of 100 questions spread across four parts;
six questions for part one (understanding the descriptions of photographs), 25 questions for part
two (choosing the correct response to a question), 39 questions for part three (understanding
conversations), and 30 questions for part four (understanding announcements; Cid, Wei, Kim,
& Hauck, 2017). It was believed that the standard TOEIC format might prove to be too difficult
40
EFLIJ Volume 25 Issue 4 July 2021
41
for the lower-proficiency participants in this study, and the resulting demotivation from a
difficult testing experience might affect their feelings towards the course. As suggested by
Brown (2011), it is advantageous to adapt the difficulty of listening texts and tasks to the ability
of students. As a result, three versions of the listening assessment were created for this study;
one test for participants in beginner-level classes, one for those in intermediate-level classes,
and one for advanced-level classes. In a study involving 3673 Japanese and Korean test-takers
(Cid, Wei, Kim, & Hauck, 2017), it was determined that the difficulty level of the four sections
of the TOEIC listening component followed an ascending order of difficulty, with part one
representing the easiest section, followed by part two, part three, and part four, with p-values
of 0.80, 0.67, 0.66, and 0.57, respectively, in one analysis of the TOEIC, and p-values of 0.82,
0.70, 0.62, and 0.62,respectively, in another analysis (p-value was defined as the proportion of
test-takers who could correctly answer a question within the population). Accordingly, the
beginner and intermediate-level versions of the listening assessment used in this study included
more questions from parts one and two, and fewer questions from parts three and four. The
advanced-level test was kept the same as the standard version of the TOEIC listening
component. The breakdown in questions from each part of the listening assessment for
beginner, intermediate, and advanced versions of the test is shown in Table 3.
Table 3
Number of questions for each part of the listening assessment, by difficulty level
difficulty level
Part 1 questions
Part 2 questions
Part 3 questions
Part 4 questions
Beginner
40
30
21
9
Intermediate
18
35
33
14
Advanced
6
25
39
30
This study included both a pretest and a posttest to assess English listening proficiency. While
both the pretest and posttest followed the TOEIC format outlined above, different questions
were used for the pretest and the posttest, resulting in six different versions of the instrument.
Procedure
In the first class of the semester, the research project was explained in English to the
participants. An explanation of the research project was also given to participants in written
form, in both English and Japanese. Participants were then asked to sign a consent form
acknowledging their willing participation in the study. After completing the consent form,
participants took the pretest, which lasted for 40 minutes.
41
EFLIJ Volume 25 Issue 4 July 2021
42
After completing the pretest, participants were randomly placed into one of three groups;
two intervention groups (an EL and an IL group) and one control group. Once students were
placed into groups, the homework expectations were explained to each group.
The EL group was advised to choose three audiobooks from the class' online library of
audiobooks, listen to their three chosen audiobooks, and to write a brief summary of each
audiobook (one paragraph) in English. Collectively, the three summaries from this group were
expected to fill an A4-sized worksheet. The EL group was instructed to only choose audiobooks
for which they had approximately 98% listening comprehension.
The IL group was advised to choose to one audiobook from the class' online library of
audiobooks, listen to it, and write a one-page dictation from any portion of the audiobook on
an A4-sized worksheet. The IL group was instructed to only choose audiobooks for which they
had approximately 80-90% listening comprehension.
The control group was told to not listen to any of the audiobooks, but instead to listen to
three texts in their native language, including, movies, television shows, and podcasts, and write
a brief summary of each (one paragraph) in English. Additionally, they were instructed to write
a one-page dictation of one of their chosen listening texts on an A4-sized worksheet in their
native language. The reason the control group was given both an English writing assignment
(summaries) and a native language writing assignment (dictation) was to equalize the English
writing component between the control and EL groups, and the time burden amongst all groups.
If the control group had no English writing component, the mean difference in pre- and post
TOEIC scores between the control and EL group could be partially reflective of the writing
component that the EL group had to complete. Additionally, it was thought that the control
group should spend roughly the same amount of time completing the homework as the other
groups, otherwise the two intervention groups might view their heavier workload as unfair.
This homework cycle was repeated every two weeks with new listening texts, with
participants ultimately completing six assignments over the course of the 15-week semester.
An effort was made to give assignments in two-week intervals; however, this was not always
possible. For example, the third class for some courses had to be cancelled because of Typhoon
Hagibis. In this case, there was a three-week gap from when the homework assignment was
given and when it was collected. Similarly, near the end of the fall semester there was a twoweek break in classes so students could celebrate year-end and new-year holidays. For some
courses, the homework assignment was collected in the week preceding this break, and the new
assignment was collected immediately after the break; a three-week gap from when the
assignment was given and when it was collected.
42
EFLIJ Volume 25 Issue 4 July 2021
43
In the final class of the semester, participants completed the 40-minute posttest.
Audiobook library
The two intervention groups were asked to self-select their own listening texts during the
semester. The listening texts were available online through the media-hosting website
Mediafire. The listening text library consisted entirely of the audiobooks of graded readers from
a variety of publishers, as shown in Table 4.
Table 4
Graded reader audiobook series used in study
Series
Level
headwords
Titles
ERF grading scale
Popcorn ELT
Starter
150
12
Beginner
Popcorn ELT
1
200
18
Beginner
Popcorn ELT
2
250
20
Beginner
Oxford
Starter
250
26
Beginner
Pearson
1
300
25
Beginner
Popcorn ELT
3
300
16
Beginner
Scholastic
Starter
300
8
Beginner
Oxford
1
400
39
Elementary
Pearson
2
600
48
Elementary
Scholastic
1
600
13
Elementary
Oxford
2
700
39
Elementary
Oxford
3
1000
27
Intermediate
Scholastic
2
1000
31
Intermediate
Pearson
3
1200
49
Intermediate
Oxford
4
1400
20
Intermediate
Scholastic
3
1500
33
Intermediate
Pearson
4
1700
30
Upper-intermediate
Oxford
5
1800
12
Upper-intermediate
Scholastic
4
2000
5
Upper-intermediate
Pearson
5
2300
32
Upper-intermediate
Oxford
6
2500
9
Advanced
Pearson
6
3000
21
Advanced
Total tiles
533
Audiobook files were in MP3 format, and participants were able to stream the audio files
through their smartphone, or download them to their computer and listen with a media player.
Analysis
43
EFLIJ Volume 25 Issue 4 July 2021
44
L2 listening proficiency data was collected on paper response sheets from participants on the
first and last days of class. Participant responses were entered into a text file, and processed
through the Rasch measurement analysis software, Winsteps. This software allows for the
creation of a variety of statistical data, such as raw scores, Rasch person measures, person fit,
and item fit. Rasch reliability and separation data for persons and items is shown in Table 5.
Table 5
Rasch reliability and separation for pre- and post-tests
Person
n
Item
reliability
separation
SD
n
reliability
separation
SD
Pre-test
237
.93
3.60
.92
144
.96
4.68
1.14
Post-test
237
.93
3.70
.89
144
.96
4.73
.98
This study used the person raw scores to determine mean difference between the pre- and posttests. The person raw scores were entered into the software IBM SPSS Statistics ver. 22, and an
independent-sample t-test was conducted on the pre- and post- mean difference for the two
intervention groups.
Findings & Discussion
Results
Initially, this study included 398 participants, however many participants did not adhere to the
prescribed intervention, and as a result, were removed from the study. To elaborate, the 398
participants were categorized into four streams of based on their degree of intervention
completion. Those participants that completed all homework assignments, without any late
assignments were categorized as stream 1 and kept in the study. Those participants that did all
homework assignments but submitted an assignment late once or twice were categorized as
stream 2, and kept in the study. However, while one or two late assignments was viewed as a
minor disruption to the prescribed intervention, three or more late assignments was viewed as
a significant disruption to the intervention. For example, some participants did not submit
homework for six weeks, but then did all of their assignments in one day and submitted multiple
assignments at the same time. This would indicate that participants were not getting regular L2
listening practice, which is a primary condition for improving L2 listening proficiency. As a
result, participants who submitted 3-5 late assignments or did not submit an assignment, were
categorized as stream 3 and removed from the study. Finally, participants that were late with
all of their assignments, or did not submit two or more assignments were categorized as stream
44
EFLIJ Volume 25 Issue 4 July 2021
45
4, and removed from the study. The descriptive statistics for the remaining 237 participants are
shown in Table 6.
Table 6
Descriptive statistics for remaining participants
Group
Pretest
n
score
Posttest
SD
score
SD
Mean
difference
SD
EL
91
56.00
17.00
57.51
14.87
1.35
8.45
IL
104
54.56
16.48
53.07
15.08
-1.49
9.93
42
53.81
12.11
52.50
12.70
-1.14
10.25
237
54.98
15.96
54.67
14.72
-0.34
9.50
Control
Total
Note: the maximum score is 100.
Recalling the intervention for each group, the EL group was instructed to listen to three texts
for which they had 98% comprehension, while the IL group was instructed to listen to one text
for which they had 80-90% comprehension. This difference in comprehension of selected
listening texts between the EL and IL groups is reflected in the average headwords of the
selected listening texts. As the EL group was selecting more comprehensible texts, the
headwords of these texts was considerably lower than that of the listening texts selected by the
IL group. The average headwords for each group on each task, is shown in Table 7.
Table 7
Average headwords for each group, by task
EL
91
Task 1
headwords
286.40
IL
104
636.06
622.60
658.65
652.88
690.87
687.91
42
0
0
0
0
0
0
237
389.08
384.59
405.85
404.86
427.00
430.06
Group
Control
Total
n
Task 2
headwords
290.08
Task 3
headwords
304.24
Task 4
headwords
308.25
Task 5
headwords
325.11
Task 6
headwords
334.79
An independent-sample t-test was conducted for the mean difference in TOEIC listening scores
between the pre- and post-test, for the EL and IL groups. Levene's test for equality of variance
indicated homogenous groups, with the EL group's subsequent mean difference significantly
larger than the IL group; t(193) = 2.14, p <.05. Results from the independent-sample t-test are
shown in Table 8.
Table 8
Independent-sample t-test of mean difference between EL and IL groups
45
EFLIJ Volume 25 Issue 4 July 2021
Raw score
Levene's test for
equality of variance
F
sig.
2.22
.14
46
T
df
2.14
193
sig.
.03
Mean
difference
SE
difference
2.84
1.33
95% CI
Lower
.22
Upper
5.47
Discussion
The results of this study were consistent with those of Chang's study (2012) in that EL proved
to be better at developing L2 listening proficiency, albeit to a lesser degree than was observed
in the Chang study. The results run contrary to the findings in Karlin and Karlin's study (2019),
and presumably the chief reason is that study failed to equate the EL and IL interventions, with
the IL group doing much more L2 listening practice, hence their superior L2 listening
proficiency development. Regarding the difference in the magnitude of L2 listening proficiency
gains between the Chang (2012) study and this study, the reason perhaps lies in the difference
in class time spent doing L2 listening practice and teacher guidance between the two studies.
The Chang study used a lot of class time with consistent teacher guidance, which presumably
enhanced the learning of participants, whereas the Karlin and Karlin study used no class time,
and had minimal teacher guidance beyond the initial explanation of the assignment. Presumably
if the two studies had equal measures of class time and teacher guidance, and if the Karlin and
Karlin study had a more balanced research design (with equal weight on the EL intervention),
the L2 listening proficiency gains would have been more consistent between the two studies.
The discrepancy in the results of these two studies stems from the inconsistent interpretations
of EL and IL interventions within the research field, particularly with widely disparate
implementations of EL research designs. Because ER relies on empowering students to choose
their own materials, read at their own pace, and be unencumbered with tasks, future research
should ensure that class time and teacher guidance for EL research be underpinned with similar
principles. Considering these caveats, despite the Chang (2012), Karlin and Karlin (2019), and
this study differing in their results, from strongly supportive of EL, mildly supportive of IL,
and mildly supportive of EL, respectively, it appears that these differences stem from research
design incongruities.
Another noteworthy feature of the data was the importance of intervention adherence in
the sample. As mentioned above, principles of ER (and, by extension, EL) are based on
empowering students to study by themselves, however under such circumstances, the risk of
students not effectively studying or doing their homework is high. Without teacher intervention
to ensure students complete every assignment properly and on-time, it is likely that a great deal
of students will not meet expectations, which is what happened in this study. Within just a
46
EFLIJ Volume 25 Issue 4 July 2021
47
single semester, over 40% of the participants in this study had to be removed because of a
failure to adhere to the prescribed intervention. Over a longer timeline, such as a full year, or
even multiple years, the attrition rate could be so high that the credibility of the study could be
threatened. Perhaps the way to prevent participants from undermining the intervention is to add
greater incentives into the research design for participants to adhere to the intervention, such as
grade penalties for failure to submit listening homework on time, and greater teacher scrutiny
on the proper completion of homework, such as with regular teacher consultations with students
or devoting a portion of class time to oversee the completion of L2 listening practice and
accompanying tasks. Doing so, however, would seem to violate one of the principal tenets of
EL in that it should be done independently and without much teacher involvement. As a result,
future research will need to address this tension between teacher oversight and student
empowerment, and negotiate a resolution.
It should also be noted that while the gains observed by EL over IL were modest, the
timeline of this study was shorter than that of other studies. The full length of this study was
only 14 weeks, while that of the Chang (2012) study was 26 weeks, Chang and Millett (2016)
was 15 weeks, and Siegal and Siegal (2015) was 15 weeks (but 30 classes). It is unclear how a
longer timeline might have affected these results, but it is assumed that gains would have been
amplified. The reasoning for this lies in the affective benefits of EL. As students do more EL
practice, the hope is that the empowerment inherent in the activity and the variety of the
listening texts engenders a listening habit in which students enjoy listening, resulting in a more
accepting stance towards listening practice and more listening opportunities as they begin
listening to audiobooks in their free time, for fun rather than out of a homework obligation.
Thus, a two-semester timeline, which would be around 28-30 weeks, might yield EL gains
closer to those observed in the Chang (2012) study.
For Japanese universities, the implications of these findings would suggest that more EL
resources be available to students. As it stands, many universities in Japan have ER resources
such as graded reader libraries, however few have EL resources such as online audiobook
libraries. With the proliferation of smartphones that are easily capable of playing mp3 audio
files, as well as the widespread availability of internet access through 4G and 5G data networks
and free campus wifi, not fully utilizing audiobooks in English-language curriculums seems
like a wasted opportunity. Never before has it been this easy for students to access such a diverse
array of L2 listening content, not just with the audiobooks that are paired with graded readers,
but also video streaming services (such as Netflix, Hulu, and Disney+), music streaming
services (such as Spotify, Apple Music, and Pandora), podcasts, online lectures (such as through
47
EFLIJ Volume 25 Issue 4 July 2021
48
the TED website), and online news broadcasts (such as through the BBC). Curriculums are
going to have to adapt from traditional models of in-class textbook-based instruction to a model
that better incorporates these resources. The audiobooks associated with graded readers, such
as the ones used in this study, are particularly effective because they are modified to suit a
variety of different proficiency levels, allowing students with intermediate proficiency or
below, to find engaging listening materials, whereas in the past, the majority of listening
materials were limited to advanced students. Japan seems particularly well-suited to the
adoption of EL in that most students rely on public transportation to attend school. Public
transportation gives students dedicated time twice a day, where there is little else that can be
done besides sitting (or more likely, standing) during a commute, representing an ideal time to
listen to five minutes of an audiobook and to establish a daily listening habit. In 2020, the
importance of developing a self-study listening habit was even more acute, as most universities
in Japan relied fully on online lessons because of COVID-19. Having dedicated online L2
listening resources can prove to be a boon to education when students must adopt more
independent learning behaviours because of external factors. Further, if the Japanese
government's goal is to transition to more EMI at universities, having online resources that offer
students long-form L2 listening practice may help with the transition for some students from
grammar-translation learning in high-school to EMI learning at universities.
Limitations
There were several limitations with this study that need to be considered when evaluating the
findings, some of which are correctable in future research and some of which are inherently
difficult to resolve. Among the issues that are difficult to resolve is determining interventions
for EL and IL that are fairly balanced, in order to get an equitable comparison between EL and
IL. For instance, it is suggested in the existing research that EL be easier so that students will
be more likely to practice EL, when compared with IL. However, because of a lack of research
and codified definitions regarding EL and IL, researchers must use their judgment when
establishing an equitable division of work between EL and IL. In this study, EL participants
were instructed to listen to three times as many audiobooks as IL participants (3 audiobooks vs
1 audiobook) every two weeks. The decision to give the EL group three times as many
audiobooks per assignment was based on roughly how long it would take to listen to three easier
audiobooks versus one more difficult audiobook (for example, an hour of listening). Yet,
because students were free to choose their own audiobooks, many students in the EL group
especially, seemed to choose audiobooks that were too easy and fully-within their level of
48
EFLIJ Volume 25 Issue 4 July 2021
49
comprehension, as evidenced by the wide disparity in headwords between the two groups
(shown in Table 7). The IL group ultimately chose audiobooks with twice as many headwords,
far more than the 8-18% increased difficulty that was explained to them in the initial
explanation of the assignment (i.e. 98% comprehension for EL vs. 80-90% comprehension for
IL). Nevertheless, future researchers will need to do their best to equate the EL and IL
interventions so that they represent roughly the same work burden for students, even though
several variables will be quite different, such as the difficulty of the audiobooks, the length of
the audiobooks, the frequency of listening practice, and the type of follow-up tasks (if any).
Bias exists in all of these variables and researchers will need to earnestly evaluate all
intervention choices in order to minimize biases in the research design, and unfortunately, will
never fully eliminate bias or achieve universal consensus within the research field.
Among the correctable limitations in this study, one would be giving students regular
feedback on the level of audiobook that is appropriate for their proficiency level. It was
observed in the study that occasionally students who were of very high proficiency, almost
native-level, were choosing the easiest audiobooks because they wanted to reduce their
homework burden. Even though these students did all of the homework, and submitted the
assignments on time, they found a way to circumvent the goals of the intervention, and future
research would be advised to limit audiobook options for these students or to suggest that they
choose more appropriate audiobooks if it appears they are choosing audiobooks well-below
their L2 listening level.. At the same time, it is important to preserve some freedom of choice,
since freedom of choice is a central ER principle and student empowerment has been shown to
have numerous benefits in the area of developing independent and motivated learners (Dörnyei
& Csizér, 1998). Again, the lack of research for EL and IL will require a great deal of
negotiation on behalf of the researchers in regard to how much teacher intervention can be
implemented while still maintaining adequate learner autonomy (Tholin, 2008).
Another limitation related to this study's interventions involves the handwritten
completion and in-person submission of homework assignments. With 237 participants
completing six assignments each, the number of total assignments completed in the study was
in excess of 1400. For the researcher, it is crucial that each assignment receive some degree of
scrutiny. For example, the IL group was required to complete a dictation of about one A4 page
in length. Yet, some students proved adept at finding ways around this requirement by
increasing the line spacing or font size in their assignment, resulting in vastly disparate
assignments between students who studiously completed their dictation and students who were
trying to evade their homework. A way to combat this might be to introduce a minimum word
49
EFLIJ Volume 25 Issue 4 July 2021
50
count for dictations (for example, 650 words), and in order to police this requirement, it would
be necessary for students to submit their assignments as a Word file (which displays the number
of words in each document). The 1400 homework assignments in this study were hand-written,
making it impossible to monitor word counts. Further, with 237 participants, the total number
of selected texts was almost 3000. It would be virtually impossible to effectively monitor 3000
assignments for plagiarism without the use of plagiarism software. In order to process
assignments through plagiarism software, assignments need to be in a digital form (such as a
Microsoft Word or pdf file). Obviously students engaged in plagiarism are not following the
prescribed intervention, and would need to be removed from the sample. Lastly, if there are
problems with the homework or the level of the text(s) chosen by the participant, the researcher
needs to be able to give feedback, and with classes only once a week, it would not be
advantageous for the researcher to have to make notations in a log-book on who to give
feedback to, when to give it, and what feedback to give (this would likely result in forgotten
details and missed consultations). A more direct and immediate line of communication, such as
through email, would reduce the complexity of this feedback process and the burden on the
researcher. As a result, for future research, participants should submit assignments via email,
which would allow the researcher to check each assignment and provide immediate feedback
(without the cumbersome process of recording details in a log for a later consultation). Suffice
it to say, monitoring the quality of completed assignments (such as word count for dictations),
catching plagiarism, and providing guidance would all be enhanced through the digital
completion and submission of assignments.
Another limitation with this study was the seeming variability in test performance between
the pre- and post-tests, which resulted in a relatively small improvement for the EL group. The
pretest, conducted at the beginning of the fall semester, was given to students who were wellrested and not burdened with other homework assignments. Conversely, the posttest, conducted
in the last class of the fall semester, was given to students who were, in many cases, fatigued
because of the long semester, not sleeping well because of the time demands put on them by
the final assignments from other classes, and in some cases, hung-over because Japan's comingof-age celebration commemorating the start of adulthood occurred the day (and night) before
the posttest (for some classes). While it was assumed that the pre- and post-tests were roughly
equal in difficulty level, the testing conditions for the posttest were more onerous, possibly
resulting in diminished scores. To overcome this, future research would be well-served to link
the pre- and post-tests with several shared items, and thus, when analysing the data through
Winsteps, the shared items can serve as a way to calculate the relative difficulty of all items in
50
EFLIJ Volume 25 Issue 4 July 2021
51
the pre- and posttest. Thus, the resulting weighted test data would likely be much more accurate,
particularly the Rasch person measure scores produced by Winsteps. It has been shown that
Rasch person measure scores can be significantly more accurate than raw scores in assessing
person ability (Weaver, Jones, & Bulach, 2008), and linking these two tests with shared items
would improve testing accuracy. The result would be mean differences in testing scores that
would be less susceptible to variations in testing conditions, such as student fatigue.
Conclusions
Hopefully, this paper has helped to clarify some of the benefits of EL and IL in developing L2
listening proficiency. While there are only a few studies that have directly compared EL and IL
in terms of developing L2 listening proficiency, it would seem that the results of this study are
consistent with existing research, especially when considering the contextual factors
surrounding each study. Perhaps of even greater value than the results of this study, is the raising
of the issues of codified frameworks for EL and IL, the need for consistency in research design
in EL and IL studies, and the elimination of bias when comparing these variables. While the
results of this and previous studies in EL and IL are relatively consistent when considering
contextual factors, the need to account for contextual factors and differences in perceptions of
EL and IL muddies the waters when attempting to come to a consensus on the findings. Until
research addresses these contextual factors, consensus may prove to be elusive. Nevertheless,
this study represents a step in that direction.
51
EFLIJ Volume 25 Issue 4 July 2021
52
References
Ahmadpour, S., & Asadollafam, H. (2018). A study of the effect of extensive and intensive
listening on the tense use among EFL learners in Iran. The Journal of Applied
Linguistics and Applied Literature: Dynamics and Advances, 6(2), 141-161.
Asquith, S. (2014). Integrating a functional approach with Japanese junior high school
teaching practices. In P. Clements, A. Krause and H. Brown (Eds.), JALT2014:
Conversations across borders. Ibaraki: JALT.
Bamford, J., & Day, R.R. (Eds.). (2004). Extensive reading activities for teaching language.
New York: Cambridge University Press.
Brown, S. (2011). Listening myths: Applying Second Language Research to Classroom
Teaching. Ann Arbor, MI: University of Michigan Press.
CASEC global website. (n.d.). Retrieved February 04, 2021, from https://global.casec.com/
Cervantes, R., & Gainer, G. (1992). The effects of syntactic simplification and repetition on
listening comprehension. TESOL Quarterly, 26(4), 767-770. doi:10.2307/3586886
Chang, A. (2010). Second-language listening anxiety before and after a 1-yr intervention in
extensive listening compared with standard foreign language instruction. Perceptual
and Motor Skill, 110(2), 355-365. doi:10.2466/PMS.110.2.355-365
Chang, A. (2012). Gains to L2 learners from extensive listening: Listening development,
vocabulary acquisition and perceptions of the intervention. Hong Kong Journal of
Applied Linguistics, 14(1), 35-47.
Chang, A., & Millett, S. (2016). Developing L2 listening fluency through extended listeningfocused activities in an extensive-listening programme. RELC Journal, 47(3), 349362. doi:10.1177/003368821663117
Chang, A., Millett, S., & Renandya, W. A. (2018). Developing listening fluency through
supported extensive listening practice. RELC Journal, 49(3), 1-17.
doi:10.1177/0033688217751468
Cid, J., Wei, Y., Kim, S., & Hauck, C. (2017). Statistical analyses for the updated TOEIC
Listening and Reading test (Research Memorandum No. RM-17-05). Princeton, NJ:
Educational Testing Service.
Day, R., & Bamford, J. (1998). Extensive reading in the second language classroom.
Cambridge, U.K.: Cambridge University Press.
Day, R., & Bamford, J. (2002). Top ten principles in teaching extensive reading. Reading in a
Foreign Language, 14, 136–141.
52
EFLIJ Volume 25 Issue 4 July 2021
53
Dörnyei, Z., & Csizér, K. (1998). Ten commandments for motivating language learners:
Results of an empirical study. Language Teaching Research, 2, 203-229.
Field, J. (2008). Listening in the language classroom. Cambridge, UK: Cambridge University
Press. doi: 10.1017/CBO9780511575945
Flowerdew, J., & Miler, L. (2010). Listening in a second language. In A. D. Wolvin (Ed.),
Listening and Human Communication in the 21st Century. Oxford, UK: Blackwell
Publishing. doi:10.1002/9781444314908
Goh, C. (2008). Metacognitive instruction for second language listening development:
Theory, practice and research implications. RELC Journal, 39(2), 188-213.
doi:10.1177/0033688208092184
Graham, S. (2006). Listening comprehension: The learner's perspective. System, 34, 165-182.
Hirai, A. (1999). The relationship between listening and reading rates of Japanese EFL
learners. Modern Language Journal, 83(3), 367-384. doi:10.1111/0026-7902.00028
Iimura, H. (2007). The listening process. Effects of question types and repetition. Language
Education & Technology, 44, 75-85. doi: 10.24539/let.44.0_75
Jeon, E-Y., & Day, R. (2016). The effectiveness of ER on reading proficiency: A metaanalysis. Reading in a Foreign Language, 28(2), 246-265.
Karlin, O., & Karlin, S. (2019). L2 listening homework: Intensive vs. extensive. Korea
TESOL Journal, 14(2), 97-115.
Mayora, C. A. (2017). Extensive listening in a Colombian university: Process, product, and
perceptions. HOW, 24(1), 101-121. doi: 10.19183/how.24.1.311
Mitchell, C. (2017). Language pressures in Japanese high schools. JALT Shiken, 21(1), 1-11.
Nation, P., & Wang, M-T. (1999). Graded readers and vocabulary. Reading in a Foreign
Language, 12(2), 355-380.
Renandya, W. A. (2011). Extensive listening in the second language classroom. In H. P.
Widodo & A. Cirocki (Eds.). Innovation and creativity in ELT methodology (pp. 2841). New York, NY: Nova Science Publishers.
Richards, J. (2005). Second thoughts on teaching listening. RELC Journal, 36(1), 85-92.
doi:10.1177/0033688205053484
Rost, M. (2016). Teaching and Researching Listening, 3rd ed. New York: Routledge.
Saito, Y. (2019). Impacts of introducing four-skill English tests into university entrance
examinations. The Language Teacher, 43(2), 9-14. doi:10.37546/JALTTLT43.2-2
Sakai, H. (2012). Effect of repetition of exposure and proficiency level in L2 listening tests.
TESOL Quarterly, 43(2), 360-372. doi:10.1002/j.1545-7249.2009.tb00179.x
53
EFLIJ Volume 25 Issue 4 July 2021
54
Schmidt, A. (2016). Listening journals for extensive and intensive listening practice. English
Teaching Forum, 54(2), 2-11.
Siegal, J. (2014). Exploring L2 listening instruction: Examinations of practice. ELT Journal,
68(1), 22-30. doi: 10.1093/elt/cct058
Siegal, J., & Siegal, A. (2015). Getting to the bottom of L2 listening instruction: Making a
case for bottom-up activities. Studies in Second Language Learning and Teaching,
5(4), 637-662. doi:10.14746/ssllt.2015.5.4.6
Shimauchi, S. (2018). English-medium instruction in the internationalization of higher
education in Japan: Rationale and issues. Educational Studies in Japan: International
Yearbook, 12(1), 77-90. doi: 10.7571/esjkyoiku.12.77
Tanewong, S. (2019). Metacognitive pedagogical sequence for less-proficient Thai EFL
listeners: A comparative investigation. RELC Journal, 50(1), 86-103.
doi:10.1177/0033688218754942
The Extensive Reading Foundation. (2011). The Extensive Reading Foundation's guide to
extensive reading. Retrieved August 6, 2019, from
erfoundation.org/guide/ERF_Guide.pdf
Tholin, J. (2008). Learner autonomy, self-directed learning and assessment: Lessons from
Swedish experience. Retrieved from
http://bada.hb.se/bitstream/2320/3572/1/Tholin%20190508%20learner%20autonomy.
pdf
Vandergrift, L. (2013). Teaching listening. In C. A. Chapelle (Ed.), The Encyclopedia of
Applied Linguistics. Oxford, UK: Blackwell Publishing.
doi:10.1002/9781405198431.wbeal1169
Vandergrift, L., & Goh, C. (2012). Teaching and Learning Second Language Listening. New
York, N.Y.: Routledge Press.
Van Zeeland, H. (2014). Lexical inferencing in first and second language listening. Modern
Language Journal, 98(4), 1006-1021. doi:10.1 111/modl. 12152
Waring, R. (n.d.). Starting extensive listening. Retrieved from
http://www.robwaring.org/er/ER_info/starting_extensive_listening.htm
Weaver, C., Jones, A., & Bulach, J. (2008). Comparing placement decisions based on raw test
scores and Rasch ability scores. The Language Teacher, 32(6), 3-8.
Webb, S., & Chang, A. (2015). How does prior word knowledge affect vocabulary learning
progress in an extensive reading program? Studies in Second Language Acquisition,
37(4), 651-675. doi: 10.1017/S0272263114000606
54
EFLIJ Volume 25 Issue 4 July 2021
55
Production and Perception of Geminate Consonants in English Words by
Thai Learners of English: Implications for English Teaching and Learning
Suthathip Thirakunkovit
Mahidol University, Thailand
Bio data
Suthathip Thirakunkovit is Assistant Professor in the Department of Applied Linguistics at
Mahidol University, Thailand. Her current research interests cover language test development,
test validation, assessment literacy, corpus linguistics, language teaching and learning, and
second language writing. Her recent work can be found in the Asian EFL Journal, English for
Specific Purposes, and the Journal of Asia TEFL. suthathip.thi@mahidol.edu
Abstract
The present study aims to explore the production and perception of geminate consonants of
Thai learners of English. This study seeks to answer three research questions. First, are there
any significant differences in the production of different types of English geminate consonants
among five groups of participants (four groups of non-native speakers of English and one group
of native speakers of English)? Second, does the level of English proficiency of the English
learners affect the production and perception of English geminate consonants? Third, do
English orthographic forms affect the pronunciation of Thai learners of English? To answer
these three questions, the researcher compared the performance of 90 students from three
different levels of language proficiency (low intermediate, high intermediate, and advanced)
and of 15 Thai university instructors with the performance of eight native speakers of English.
The researcher examined the performance of the participants by using two tasks: reading aloud
and dictation. Even though the statistical results showed a significant difference only in the
production of lexical geminates between native and non-native English speakers, it should be
noted that the students from the lower levels of language proficiency seemed to have greater
difficulty in identifying English geminates in regular speech. Ultimately, the researcher would
like to argue that teaching pronunciation is not just about teaching students how to produce the
Correspondence address:
999 Faculty of Liberal Arts, Mahidol University
Phutthamonthon Sai 4 Rd. Salaya Sub-district,
Phutthamonthon District, Nakhon Pathom
73170, Thailand
55
correct sounds. It is also about helping
them to hear the sounds correctly, so that
they can understand what they hear.
EFLIJ Volume 25 Issue 4 July 2021
56
Keywords: English geminates, English teaching, EFL learners, English pronunciation
Introduction
When learning the pronunciation of a new language, it is possible that the sound patterns of
foreign words are adapted to conform to those existing in the native phonological system
(Thirakunkovit, 2020; Ying, 2016). The result of the researcher’s pilot study previously
published in Thai in the Journal of Language and Culture, an academic journal in Thailand,
has provided a piece of strong evidence to support this claim. In this pilot study, the majority
of her participants who were Thai undergraduate students pronounced the English word happy
/ˈhæpi/ as [ˈhæppi], for example. Therefore, she concluded that this mispronunciation came
from the influence of the Thai syllable structure which requires two separate sounds for a coda
consonant and an initial consonant of another syllable (Thirakunkovit, 2020). Yet at the same
time, some people may think that the mispronunciation could come from a [pp] input of the
English orthographic representation that could be seen as two distinctive geminate consonants
[pp] by many second language learners. Therefore, the researcher decided to conduct this
follow-up study in order to validate the results of her previous study and explore in depth
whether or not English orthography has an effect on the production of English geminate
consonants of Thai learners of English, and whether or not the mispronunciation of English
geminate consonants has an effect on the perception of English geminate consonants.
In the researcher’s previous study, she found a number of flaws with her word list and data
collection procedures. Moreover, some questions related to second language acquisition and
language teaching have not been addressed. Therefore, she decided to extend the scope of
previous study and gather more data in order to answer her research questions in a more specific
way.
In this current study, the researcher has collected more data by including two more groups
of participants in order to explore the effects of language proficiency on overall learner
performance. The English word and sentences on the previous list have also been adjusted, so
that the researcher can find more concrete evidence to answer her research questions. Moreover,
in order to shed light on the perception of English geminate consonants of Thai learners, she
has added a dictation task to determine whether or not Thai learners of English from four
different levels of language proficiency have difficulty in identifying English geminates in
natural speech. At the same time, the researcher has investigated how the discrepancy between
orthographic representations and acoustic realizations influences the way Thai learners perceive
English geminate consonants.
56
EFLIJ Volume 25 Issue 4 July 2021
57
This language feature is chosen for investigation because it has rarely been mentioned in
language classrooms. Although some may argue that mistakes in the pronunciation might not
cause serious misunderstanding, learning to correctly say words or phrases that contain
geminate consonants can result in more fluid and fluent sounding speech (Thirakunkovit, 2020).
Moreover, flipping through more than a dozen commercial English textbooks currently
available in the market, the researcher found only a few textbooks that mentioned the
pronunciation of English geminates in their content (Thirakunkovit, 2020). The researcher also
noticed that when she talked to her colleagues who are English teachers in college, there was
only one teacher who said he tried to teach the concept of English geminates to his students;
the rest have never done it at all!
Therefore, the researcher feels that it is important for language teachers to guide students
to this language feature. If students are aware of the characteristics of English geminate
consonants, they will understand what native speakers say more easily and accurately, and at
the same time they can sound more natural, which is an important goal in teaching speaking
skills and successful communication (Thirakunkovit, 2020).
The Effects of Orthographic Representation on the Pronunciation of L2 Learners
A number of studies on the acquisition of second language pronunciation have focused on the
effects of L2 orthographic representations. Many of these studies were led by Dr. Bene Bassetti
from the University of Warwick. In 2015, Bassetti and Atkinson (2015) conducted an
experiment to investigate whether the written representation of English words (L2) affects the
pronunciation of high-school learners of English whose L1 was Italian, which is a language that
makes a clear distinction between non-geminate and geminate consonants. In the experiment,
Bassetti and Atkinson examined the effects of vowel and consonant spelling on their duration.
To be more specific, they explored whether or not the Italian learners produced the same target
vowel and consonant as longer when they are spelled with two vowel and consonant letters than
with a singleton letter, i.e., whether the learners pronounced a longer [t] in kitty than in city.
Their analyses revealed that the Italian EFL learners pronounced vowels and consonants longer
in words spelled with double letters rather than with single letters. Another similar study was
published by Bassetti, Sokolović-Perović, Mairano, and Cerni (2018). This study also
investigated orthographic effects on the pronunciation of Italian learners of English. In this
study, the participants were asked to produce 33 English orthographic minimal or near-minimal
pairs, for example, finish and Finnish. The results also confirmed the conclusion from the
previous study. Italian learners produced English word pairs in a distinctive way by a long or
57
EFLIJ Volume 25 Issue 4 July 2021
58
short sound (both consonant and vowel) when the target sound was spelled with a single letter
in one word and a double in the other. Therefore, it can be seen that the alternation between
single and double consonants in these speakers’ production is clear evidence of the effects of
orthographic representation on the pronunciation of L2 learners.
Based on these studies, it can be argued that there are potential effects of L1 orthography
on L2 learners’ pronunciation. That is, the errors in L2 pronunciation can be traced back to L1
orthography-phonology correspondence. Therefore, potential effects of L1 orthographic forms
on the pronunciation of L2 should be considered in order to improve the teaching of L2
pronunciation. Therefore, in the current study, the researcher would like to hypothesize that
English orthography might be a factor that affects the production and perception of English
consonant geminates produced by Thai learners.
Background Information about Geminate Consonants
Geminate Consonants in Different Languages
According to Crystal (2008), gemination is “a term used in phonetics and phonology for a
sequence of identical adjacent SEGMENTS of a sound in a single morpheme, e.g. Italian /nɔtte/
(‘night’)” (p. 206). In phonetics and phonology, gemination in a number of languages is a
process in which an articulation of a sound has been prolonged for a certain period of time
compared to that of its single instance (Crystal, 2008). Geminate consonants in some languages
can be understood as a double consonant. For example, in Japanese, geminate consonants are
seen as a long sequence of two consonants that acts as the same segment (Davis, 1999).
Gemination can occur with either consonants or vowels. In this study, however, only the
gemination process occurring in consonants will be studied.
Gemination of consonants occurs in a number of languages. Some are distinctive, while
others are not. In some European languages such as Finnish as well as some Asian languages
such as Japanese, geminations of consonants are distinctive. In other words, consonant length
is necessary to distinguish words. In the Finnish language, gemination can occur with both
consonants and vowels. For example, taka /taka/ means back, while takka /takka/ means
fireplace, and taakka /taakka/ means burden. In the Japanese language, there is a strong contrast
in consonant duration between a singleton and a geminate or a double consonant. For example,
そっと /sotto/ means quietly, while そと /soto/ means outside. As shown in Figure 1, the
geminate /t/ as in /sotto/ is not perceived as two sounds from two separate syllables. Shibatani
(1990) described this language phenomenon as a homorganic consonant sequence which
58
EFLIJ Volume 25 Issue 4 July 2021
59
consists of a coda in the first syllable and onset of the second syllable (as cited in Hardison &
Saigo, 2010).
Figure 1. The Phonetic Structure for the Japanese Geminates and Singletons
(σ = syllable, μ = mora)
Different from the aforementioned languages, gemination in English is not a distinctive feature.
English gemination is usually indicated in writing by double consonants. Even though there are
two letters, they are pronounced together as one sound (one singleton). In other words, there is
only one onset with one release. Some examples of geminate consonants in English include
/ˈhæpi/ happy, /ˈbɒsi/ bossy, /ˈrʌnər/ runner, and /ˈhæmə/ hammer.
For English gemination, it is particularly important to note that it is still controversial
whether the single /p/ in happy and the single /m/ in hammer belong to the coda of the first
syllable of the word or the onset of the second syllable (McCully, 2009). In other words, should
we syllabify the words as ha.py and ha.mer or hap.y and ham.er, respectively? Even though
many L2 learners might be tempted to syllabify the words as hap.py and ham.mer, they are, of
course, impossible and considered ill forms for native speakers of English.
McCully (2009) has tried to solve this issue by proposing the term “ambisyllabicity” to
refer to a process when an intervocalic consonant belongs to both syllables simultaneously.
Based on the concept proposed, the structure for the word hammer looks like the following:
59
EFLIJ Volume 25 Issue 4 July 2021
60
Figure 2. The Sound Structure of the Word hammer
This language feature can pose a problem to non-native speakers of English because they may
intuitively pronounce these double consonants with the length that is twice as much as one
singleton [ˈhæppi] due to the orthographic representation happy. Even though it is true that the
production of these geminates with double consonant length may not affect the meaning in most
instances, they may sound odd and confuse English-speaking listeners momentarily
(Thirakunkovit, 2020). In the worst-case scenario, a communication breakdown or
misunderstanding can occur if the learners cannot perceive geminate consonants in normal
English speech. Therefore, it is important for language teachers to teach this language feature
explicitly in order to increase the level of production and perception accuracy of learners from
all levels (Thompson & Gaddes, 2005).
Classification of Geminate Consonants in English
In English, there are three kinds of geminates: (1) lexical, (2) assimilated, and (3) concatenated
(Lahiri & Hankamer, 1988). Lexical geminates occur in the lexicon. In other words, they are
part of the phonemic inventor (e.g. /ˈhɒli/ holly and /ˈgræmər/ grammar). Similarly, assimilated
geminates occur within the same words, but through morphological processes i.e., affixation.
For example, in the word unknown the affixing of un with its final [n] and known with its initial
[n], the consonant /n/ becomes /nn/ (Ball & Rahilly, 2014). On the other hand, concatenated
geminates occur between words when two adjacent consonants are identical such as the phrase
bad day. In this case when the two /d/s are adjacent, it is possible that the speakers will merge
them together with a long hold phase in a natural conversation speech. In other words, the first
/d/ becomes unreleased and the second /d/ has merged with the first /d/. When two identical
plosives have merged, a sequence of homorganic plosives is said to have occurred (Knight,
2012).
60
EFLIJ Volume 25 Issue 4 July 2021
61
Gemination in Thai
Thai is a language which contrasts only short and long vowels (e.g. /la/ means to omit vs. /la:/
means a donkey). In other words, there is no gemination for Thai consonants. Even though the
lengths of the same consonants might vary in natural speech, this is simply due to the manner
of articulation of a particular consonant (e.g. a plosive vs. a lateral), the voicing of a particular
consonant (e.g. /b/ vs. /p/), and the tone and stress of a syllable. In any case, they do not affect
the meaning. As there is no geminate consonant in the Thai language, Thai learners might not
be aware of this linguistic feature and produce it differently from native speakers of English.
Research Aims
The present study specifically investigates the duration of geminate consonant production in
native Thai learners of English. Specifically, the researcher would like to investigate timing
differences in the production of different types of English geminate consonants between native
and non-native speakers of English. Moreover, this study explores whether the level of English
proficiency of Thai learners affects the production and perception of English geminate
consonants. These objectives have addressed the following research questions:
1.
Are there any significant differences in the production of different types of English
geminate consonants among five groups of participants (four groups of non-native
speakers of English and one group of native speakers of English)?
2.
Does the level of English proficiency of Thai learners affect the production and
perception of English geminate consonants?
3.
Do English orthographic forms affect the pronunciation of Thai learners of English?
Research Methodology
Participants
This study includes three major groups of participants – 90 Thai EFL students, 15 Thai
instructors of English, and eight native English speakers. The total number of participants was
113. All potential participants were approached via email to ensure voluntary participation.
The first group of participants included 90 undergraduate students from three sub-groups:
English majors, non-English majors of low intermediate proficiency, and non-English majors
of high intermediate proficiency. The researcher first contacted course coordinators of several
English courses and asked whether she could come to their classes to recruit their students. The
courses were chosen from the requirement of minimum English scores from the national
admission test, which could be further classified as low intermediate, high intermediate, and
61
EFLIJ Volume 25 Issue 4 July 2021
62
advanced. For students who expressed their interest in participating in the study, they were
given a direct link to the consent form. After consenting, they were asked to complete a short
demographics survey concerning their educational background and provide contact
information. The researcher tried her best to have an equal number of participants in each group.
The second group included 15 university instructors teaching in the English or Applied
Linguistics Programs at the Faculty of Liberal Arts. All of them were native speakers of Thai.
The average length of their teaching career was ten years. About half of them graduated from a
university in the United States of America.
The third group included eight native speakers of English currently teaching English at the
university level. Four of them were American and the other four were British. The data elicited
from the eight native English speakers were used as a benchmark for performance. The reason
why the number of native speakers of English in this study was quite small is because only
eight native English-speaking instructors were available at the time of data collection.
Research Tool
The present study examines the performance of the participants in two tasks: reading aloud and
dictation.
Reading Aloud
A list for the reading aloud task includes 24 English single words, eight English noun phrases,
and five English sentences. The prepared list is shown in Appendix A. The list was prepared
solely by the researcher based on previous studies of English geminations. Since no studies
have previously been conducted to explore the pattern of geminate consonants of English words
produced by Thai learners, as mentioned earlier, the researcher previously did a pilot study on
a group of approximately 60 Thai participants who were undergraduate students in order to
ensure the validity of the instructions and the research tool. After the pilot study, the instructions
and the words on the list had been slightly adjusted for the current study.
Twenty three word stimuli contained two identical geminate consonants in their
orthographic representation. Twelve words were lexical geminates, and 11 words were
assimilated geminates. Seven words contained /p, t, k, b, d/ stop geminate consonants. Six
words contained /m, n/ nasal geminate consonants. Four words contained /s, f/ fricative
geminate consonants. Six words contained /l, r/ liquid geminate consonants. The last word copy
does not contain any geminate consonants. All of them were words with relatively high
frequency in both spoken and written language. For the stimuli at the phrasal level, three phrases
62
EFLIJ Volume 25 Issue 4 July 2021
63
contained /p, d/ stop geminates, three phrases contained /m, n/ nasal geminates, one phrase
contained /s/ fricative geminates, and one phrase contained /l/ liquid geminates. Finally, stimuli
at the sentential level consisted of five sentences that contained all three kinds of geminate
consonants -- lexical, assimilated, and concatenated -- with a heavier focus on word-boundary
levels in the analysis.
For the group of Thai participants, a list of 6 two-syllable Thai words was also added into
the current study for a comparison between Thai and English pronunciation. The first syllable
of every word ended in a consonant sound and the following sound began with the same
consonant sound. The reason why a list of Thai words has been added into the current study is
because the researcher would like to explore whether the native language of the participants
might influence the consonant length of English words. In Thai, when the final consonant of
the first syllable and the initial consonant of the second syllable are the same, they are always
pronounced as two separate sounds (CVC.CVC).
Dictation
After the first task, all participants in the current study were required to do a dictation task. This
task consisted of eight sentences. All of them were relatively short, ranging from three to eight
words. The first sentence I love you was a warm-up item that was used to make sure that the
participants were ready for the dictation task and could hear the sound clearly. The rest of the
sentences had at least one word that contained geminate consonants. The dictation task
contained all three kinds of geminates -- lexical geminates e.g. puppy and summer; assimilated
geminates e.g. illegal and immigrant; and concatenated geminates e.g. one night and lab boy.
Every target word and phrase is presented in a meaningful sentence and can be interpreted
within the sentence.
63
EFLIJ Volume 25 Issue 4 July 2021
64
Table 1
Items Tested in Each Sentence
Sentence (Number of Words)
Items Tested
Type of Geminates
warm-up item
—
one night
Concatenated
loves signing
Concatenated
4. The lab boy has an irregular work
schedule. (8)
lab boy
irregular
Concatenated
Assimilated
5. My dad gave me a puppy last
summer. (8)
puppy
summer
Lexical
Lexical
6. It takes time to love someone. (6)
It takes
Concatenated
John needs
needs support
support
Concatenated
Concatenated
Lexical
illegal
immigrant
Assimilated
Assimilated
1. I love you. (3)
2. I ran into him one night. (6)
3. Bill loves singing. (3)
7. John needs support. (3)
8. He is an illegal immigrant. (5)
Research Procedures
This study took place at a large public university in Thailand. All participants were required to
complete two task types: reading aloud and dictation. The two tasks given to every participant
followed the same order. The researcher hired two research assistants to help ensure that data
collection followed the same procedures.
In the first task, all participants were given a printed list of material (See Appendix A).
The participants were given approximately 10 to 15 minutes of silent reading to make sure that
they could do the reading aloud task correctly and confidently. Then, during the recording
session, participants were asked to produce each word, phrase, and sentence twice. The words,
phrases, and sentences on the list came in a different randomized sequence in order to prevent
order effect. For each recording, the participants were instructed to produce the speech at a
natural speaking rate. The participants were asked to put each word and phrase in the carrier
sentence, “Please say the word/phrase __________ again.” and were asked to read it two times.
Between the two readings, there was a three-second pause. Only the second production of each
pair was used in the analysis. The researcher demonstrated what the desirable speech rate was
for the reading aloud task, which is approximately 130 words per minute for a normal speaking
64
EFLIJ Volume 25 Issue 4 July 2021
65
rate. She also asked the participants to imitate the speech rate. During the recording session,
one of the research assistants was asked to monitor the consistency of the participants’ speech
rate, pitch, and tempo. The responses of all participants were recorded in a soundproof
recording studio. The utterances were recorded onto audio files using a professional stereo
microphone.
In the second task, the participants were asked to listen to eight English sentences produced
by two native speakers of English: a male and female. In this task, all participants were asked
to write down what was being said as accurately as possible. Each sentence was said twice.
After the participants heard a sentence, they were given 20 seconds to write it down. The first
four items were spoken by a male voice, and the last four items were spoken by a female voice.
The speech obtained from the native speakers was tested for auditory intelligibility by the
researcher and a native speaker of English who was not a participant in this study. The purpose
of the second task is to test whether they could understand the native speakers’ pronunciation
that contained geminate consonants and then write down the sentences they heard. Even though
the participants were supposed to write down exactly what they heard, neither punctuation nor
capitalization was scored. Both tasks were completed on the same day with a five-minute break
in between.
Data Analyses
The data were analyzed with two computer programs. For acoustic analyses, Praat, a publicly
available software for the analysis of speech, was used to identify the acoustic duration of
geminate consonants. The duration of geminate consonants was measured from the offset of a
preceding vowel to the onset of a vowel that immediately followed the geminate consonants
(Thirakunkovit, 2020). If the speaker paused between words, the total duration between the
preceding and following vowels was calculated. For fricative consonants, the total duration was
measured from the start of the formant to the end when the formant suddenly disappeared or
became remarkably weaker.
For each word, the mean duration of the target sound was obtained from the measures of
the second production. Twenty percent of the data were double-measured by a second person
who was a phonetician to ensure the accuracy of Praat analysis. All disagreements were
resolved during discussion.
Next, SPSS software was used to calculate the means and standard deviations of the
duration of geminate consonants produced by the participants, and a one-way ANOVA was run
to determine whether there were any differences in significant timing for the mean durations of
65
EFLIJ Volume 25 Issue 4 July 2021
66
different types of English geminate consonants between native and non-native speakers of
English.
For the data from the dictation task, the responses were analyzed for errors and the causes
of those errors attributed to the misperception of English pronunciation. A one-way ANOVA
was run to investigate whether there were any significant differences among the five groups of
participants.
The researcher’s marking scheme consisted of marking a word as either correct (one full
mark) or incorrect (no mark) on the basis of whether the participants could correctly identify
words or phrases that contained geminate consonants. In other words, the participants did not
receive any points if they were unable to identify those words and phrases. The researcher
ignored minor spelling errors. For example, if a student wrote puppi for puppy, no mark was
deducted since it does not require a substantial amount of judgment. No half marks were
awarded for this task. To put the current marking system into perspective, an example of
marking results is given below.
Correct answer: My dad gave me a puppy last summer.
Participant 1: My dad gave me a puppi(1) last summer(1).
Correct answer: The lab boy has an irregular work schedule.
Participant 2: The last boy(0) has an regular(0) work schedule.
With the system of scoring set at 0 or 1 for each word or phrase containing geminate
consonants, Participant 1 would receive a score of 2 even though she made a spelling mistake.
However, Participant 2 would receive a zero because it appears that Participant 2 was not able
to identify the word and phrase that contained English geminates. With this marking system,
the researcher believes that the score differential can reflect the ability of the learners in
identifying geminate consonants.
Results
The Results from the Reading Aloud Task
In the analysis, sounds that came from mispronunciation of the target words or phrases were
eliminated from the study. The average duration of each word and phrase produced by each
group of participants is shown in Appendix B. Figure 3 shows the results of mean duration
across groups. The error bars indicate standard deviations of the dataset.
66
EFLIJ Volume 25 Issue 4 July 2021
67
Figure 3. Mean Duration of Three Kinds of Geminates across Groups
For all three kinds of geminates, only the mean duration of lexical geminates indicated
significant differences among groups, F(4,108) = 3.17, p = .01. Post hoc analyses using Tukey
HSD for significance indicated that the mean duration for native speakers of English was
significantly different from those of low and high intermediate students (p < .01 and p = .03,
respectively). Even though the means of advanced students and English instructors did not show
significant differences from that of native speakers, the mean duration of these two groups was
higher than that of native speakers. It should be noted that the findings of the ANOVA test
should be interpreted with caution because the sample sizes of each group are unequal.
If we examine the waveforms and spectrograms in Praat of some lexical geminates of the
native and non-native speakers, some differences in the articulation can be found. One obvious
difference can be seen from the plosive p in the word pippin. The non-native speaker produced
it with a much longer duration of geminate and a complete closure in between (0.179 vs. 0.097).
See Figure 4. One interpretation can be that this speaker saw the [pp] input as two separate
consonants, thereby producing the sound relatively longer than it should be. However, for the
native speaker of English, the duration is shorter, and some voicing can still be seen, as
indicated by greater darkness between the two vowels. See Figure 5.
67
EFLIJ Volume 25 Issue 4 July 2021
68
Figure 4. Waveform and Spectrogram Display of a Non-Native Speaker of English for pippin
Figure 5. Waveform and Spectrogram Display for a Native Speaker of English for pippin
Another obvious difference can be seen from the production of the nasal formant. From Figure
6, it seems that the Thai learner produced the word innate with a double n sound because the
spectrogram shows a quick closure between the two syllables. On the other hand, the native
speaker of English pronounced innate with one [n] (Figure 7).
68
EFLIJ Volume 25 Issue 4 July 2021
69
Figure 6. Waveform and Spectrogram Display of a Non-Native Speaker of English for innate
Figure 7. Waveform and Spectrogram Display for a Native Speaker of English for innate
When investigating the mean durations of assimilated and concatenated geminates, even though
the results of the one-way ANOVA did not indicate any significant differences between the
native and non-native speakers of English, the non-native groups had a tendency to pronounce
them with an average of 0.04 seconds longer than the native speakers of English.
Moreover, the results from the English reading aloud task has revealed the effects of
orthography on the pronunciation of consonant durations in the L2 English of L1 Thai speakers.
The mean lengths of the [p] sound in the word floppy being pronounced longer than that of [p]
69
EFLIJ Volume 25 Issue 4 July 2021
70
in the word copy in every group of Thai learners. However, this claim should be further
investigated by including other minimal pairs in future studies.
Apart from the investigation of gemination in English words, the current study also
investigates the closure duration of Thai words that contain the same consonant between two
syllables (e.g. /khan-naa/ คันนา, /paak-kaa/ ปากกา, /ŋoŋ-ŋuai/ งงงวย) and only one consonant in
their orthography between two syllables (e.g. /kin-na-rii/ กินรี and /tham-ma/ ธรรมะ). From the
results of mean durations, we can see that the means of all kinds of English geminate
consonants, especially concatenated geminates and Thai words are actually very similar across
proficiency groups (See Figure 8 and Appendix C). Therefore, we might be able to see some
L1 influence on L2 pronunciation.
Figure 8: Mean Durations of Thai words when compared with English Geminates
The Results from the Dictation Task
In order to explore whether the level of language proficiency has a significant effect on the
perception of English geminates, responses from the dictation task were investigated. Table 2
shows the mean scores and standard deviations across five groups. A one-way ANOVA showed
that there is a significant effect of the level of language proficiency on the participants’
performance, F(4,108) = 30.23, p < .00. Post hoc analyses using Tukey HSD post hoc criterion
for significance indicated that every pairwise comparison of mean scores among the groups of
lower proficiency (Low Intermediate and High Intermediate) were significantly different from
the higher (Advanced, English Instructors, and Native Speakers of English)ones with p-values
less than .00, except for the comparisons between the advanced and English instructor groups
(p = .46), the advanced and native speaker groups (p = .45) and the English instructor and native
speaker groups (p = .99)
70
EFLIJ Volume 25 Issue 4 July 2021
71
Table 2
Results of the Dictation Task
Students
Native
Speakers of
English
Low
Intermediate
High
Intermediate
Advanced
English
Instructors
Means
7.5
9.5
11.17
12.73
13.0
SD
2.10
2.3
2.07
0.59
0
Table 3 shows the percentages of participants who could write down the words or phrases
correctly. Even though the percentages of correct answers for each test item varied widely
among the five groups, the overall accuracy rates increased as proficiency levels increased. The
participants from the low intermediate level had difficulty in identifying English geminates,
especially when the vocabulary or expressions were unfamiliar. Even though their responses in
general matched the original at both the syntactic and semantic levels, they showed poor results
in geminate perception with less than 40% accuracy rates. However, the participants from the
high intermediate and advanced levels had better performance perceiving English geminates
correctly, at approximately 60% and 80% accuracy, respectively, during the second hearing.
For the English instructors, their accuracy rates approached 100% for most of the dictation items
during the second hearing. For native speakers of English, their responses were 100% accurate
during the second hearing.
71
EFLIJ Volume 25 Issue 4 July 2021
72
Table 3
Percentages of Correct Responses of the Participants
Students
Items
Tested
High
Intermediate
Low Intermediate
Advanced
English
Instructors
Native Speakers
of English
1st
2nd
1st
2nd
1st
2nd
1st
2nd
1st
2nd
Warm up
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
one night
52.5%
57.5%
60.5%
73.5%
85%
95%
93%
100%
100%
100%
Bill loves
82.5%
90%
81.5%
97%
83%
83%
93%
100%
100%
100%
loves
singing
35%
35%
63%
66%
73%
76%
86%
100%
100%
100%
lab boy
0%
2.5%
2.5%
10.5%
7%
19.5%
40%
46%
87.5%
100%
irregular
10%
20%
34%
55%
44%
54%
53%
66%
100%
100%
puppy
45%
50%
63%
79%
85%
95%
100%
100%
100%
100%
summer
80%
85%
87%
97%
100%
100%
100%
100%
100%
100%
It takes
65%
90%
79%
90%
97.5%
100%
93%
100%
100%
100%
John needs
35%
50%
47%
58%
73%
80%
93%
100%
100%
100%
needs
support
17.5%
25%
50%
53%
71%
80%
93%
100%
100%
100%
support
87.5%
82.5%
81.5%
84%
90%
97.5%
100%
100%
100%
100%
illegal
5%
7.5%
26%
24%
49%
56%
53%
66%
87.5%
100%
7.5%
15%
39.5%
37%
85%
88%
93%
100%
100%
100%
immigrant
In the second step of the analysis of the dictation task, the answers of the participants were
analyzed according to the errors made. The following examples are the misperception
commonly found in the answers:
Original
Misperception
irregular
regular
illegal
legal
immigrant
migrant
lab boy
last boy / rap boy
one night
night / tonight
72
EFLIJ Volume 25 Issue 4 July 2021
73
In investigating the errors the participants made in taking dictation, it is quite clear that a number
of participants could not accurately perceive the variation of in- geminates and geminates in
word boundaries that are not commonly found in everyday speech. “He is an illegal
immigrant.” and “The lab boy has an irregular work schedule.” are two items from the dictation
section which clearly demonstrate this argument. These poor accuracy rates might result from
the fact that the production of consonant geminates by native speakers of English is subtly
different from the sound sequences in the Thai language, which are strung together with obvious
boundaries between them.
When looking at the errors made by the Thai participants, the non-salient nature of
unstressed English syllables seems to be another factor that affects their perception
(Suntornsawet, 2019), making them unaware of the geminate sounds. A number of participants
misperceived unstressed geminate consonants as singleton consonants, i.e., illegal as legal and
irregular as regular. As the Thai language does not have unstressed syllables/words like
English does, it is especially difficult for Thai English learners to notice these geminate sounds.
In order to improve the perception of English geminates, the researcher would like to suggest
the use of explicit instruction and a dictation task with input enhancement to draw the attention
of the learners to this language feature.
Conclusions
The results of the reading aloud and dictation tasks reported in this study seemed to show the
influence of both English orthography and L1 interference on the production and perception of
English geminate consonants of Thai learners. Moreover, the use of Thai reading aloud task has
shown a possible influence of the native language because the closure duration of Thai words
that contain the same consonant between two syllables are very similar to those of the mean
durations of English words that contain geminate consonants. The spectrogram of the
pronunciation of some English words such as innate (Figure 6) shows a quick closure between
the two syllables, which resembles the pronunciation of Thai syllable structure (e.g. /paak-kaa/
CVVC.CVV). Finally, the researcher believes that these two assumptions can help explain why
a number of Thai learners in this study could not accurately identify English geminate
consonants in natural speech.
Implications for English Teaching
Research related to teaching pronunciation has started to gain more attention in recent years
(Lightbown & Spada, 2013). One question is whether intelligibility or native-like production
73
EFLIJ Volume 25 Issue 4 July 2021
74
should be a standard that L2 learners strive for. Even though a number of studies show that the
presence of foreign accents does not necessarily result in reduced intelligibility or
comprehension (Breitkreutz, Derwing, & Rossiter, 2001; Jenkins, 2006), some research
evidence shows that some listeners negatively react to second language learners’ pronunciation
(Breitkreutz, Derwing, & Rossiter, 2001; Munro & Derwing, 2011). On many occasions, we
cannot deny the fact that accent can serve as a marker of group membership and may be used
as a basis for discrimination. Therefore, the researcher believes that second language learners
should be motivated to learn the correct pronunciation right at the start of their study.
This has become a challenge. How can learners change their view of speaking in English
if they view target-like pronunciation as an unrealistic or unnecessary goal at the moment?
Should a combination of instruction that aims at a target-like level, exposure, and motivation
be used in classrooms? And how can language teachers provide guidance and emphasize the
importance of native-like pronunciation when preparing their students for interacting with
native speakers of English or speakers who use English as a lingua franca?
Although the teaching of English pronunciation might not be a top priority for some
language teachers, pronunciation is an important aspect of learning English due to the growing
needs of cross-cultural communication and social integration. Some language learners may find
it useful to imitate pronunciation samples of native speakers’ speech. Therefore, classroom
teachers should ask their students to spend time carefully listening to native speakers’ speech
and then imitate what they have heard. However, in doing this, the researcher has no intention
of sending a strong message that language teachers have to use only native-like pronunciation
as the sole model. Instead, the researcher wants to point out that language teachers should at
least use native-like pronunciation as general guidelines to help their students correct or adjust
their pronunciation in order to increase intelligibility and improve their listening skills.
In this case, the researcher would like to recommend the use of explicit instructions to
elicit correct pronunciation. If the teachers can draw their students’ attention to this specific
feature, the researcher believes that it can result in permanent changes in their productions. In
the case of teaching English geminates, teachers should explicitly tell their students that they
should avoid the temptation to lengthen a double consonant letter because, most of the time, it
is simply pronounced as one short consonant. Even though there might be some exceptions to
geminates in the morpheme and word boundary patterns such as immigrant and prime minister,
the teachers should emphasize that those lengthened consonants rarely occur in casual speech.
For example, the word prime minister is usually pronounced as [prʌɪ mɪnɪstə] rather than [prʌɪm
mɪnɪstə], and the expression gimme (and informal version of give me), is actually pronounced
74
EFLIJ Volume 25 Issue 4 July 2021
75
[gɪmi] with only one [m]. Moreover, consonants after the [i] in words such as illogical, immoral,
irregular are pronounced as singletons.
For pedagogical pronunciation approaches, there are several techniques that language
teachers can incorporate in their classes. They can choose to integrate this language feature into
their curriculum, or to devote 10 to 15 minutes of class time to this specific pronunciation
instruction. The teachers can begin by introducing the concept of English geminates. This
allows the students to realize that the occurrence of a double consonant letter in English spelling
does not correspond with single consonant articulation. When the teachers want to integrate the
theoretical concept of English geminates into pronunciation practice, their activities need to be
based on real-life situations, so that the students can transfer their linguistic behavior beyond
the classroom. For example, the teachers may start from contexts that are familiar to students,
such as using the sentence “This is a good dress.” as an example of English geminates when
teaching students how to give a compliment to someone’s dress, or they can use pop songs to
help students hear geminate consonants better. In the 2013 song ‘Happy,’ Pharrell Williams
sings:
Because I'm happy =>/ˈhæpi/
Clap along if you feel like that's what you want to do =>/fiːlaɪk/, /wɑːntə/
From the song, the teachers may ask the students to identify how many consonants they have
heard in happy, feel like, and want to. After that, the students can be encouraged to produce
English germinate consonants through the teacher backchaining models for pronunciation. Once
the concept of English geminates is firmly established, the teachers can introduce English
geminates that occur in less familiar contexts or those in unstressed syllables such as illegal.
Nevertheless, the teachers should be reminded that the native speakers’ pronunciation
should be held as a reference point only. The goal of their pronunciation instruction is not to
achieve a native-like accent, but to help their students become more comprehensible, especially
when interacting with native speakers of English.
Even though some may argue that the mispronunciation of English geminates, such as
mispronouncing /ˈhæpi/ as [ˈhæppi], should not create misunderstandings in communication in
most cases, the mistake could make them sound strange and definitely not English-like.
Therefore, the researcher believes that explicit instructions on both production and perception
should go hand in hand in order to enhance the learners’ acquisition of English geminates.
Teachers should recognize pronunciation errors and expose students to the correct
75
EFLIJ Volume 25 Issue 4 July 2021
76
pronunciation through extensive drills and exercises, so that students will have opportunities to
improve their pronunciation and perception.
Limitations and Directions for Future Research
There are two major limitations of this study that could be addressed in future research. First,
the number of native English speakers was very small, and they were sampled from only two
English-speaking countries. Therefore, the results gained from this group might not be
representative of the population. Second, there is still a lack of prior research on the production
and perception of English consonant geminates by Thai learners of English. Therefore, some of
the phonological issues related to Thai learners of English might not be fully addressed.
Acknowledgments
This study is fully supported by a research grant of Mahidol University, Thailand.
76
EFLIJ Volume 25 Issue 4 July 2021
77
References
Ball, M. J., & Rahilly, J. (2014). Phonetics: The Science of Speech. New York: Routledge.
Bassetti, B., & Atkinson, N. (2015). Effects of orthographic forms on pronunciation in
experienced instructed second language learners. Applied Psycholinguistics, 36(1), 6791.
Bassetti, B., Sokolović-Perović, M., Mairano, P., & Cerni, T. (2018). Orthography-induced
length contrasts in the second language phonological systems of L2 speakers of
English: Evidence from minimal pairs. Language and Speech, 61(4), 577-597.
Breitkreutz, J., Derwing, T. M., & Rossiter, M. J. (2001). Pronunciation teaching practices in
Canada. TESL Canada Journal, 19(1), 51-61.
Collins, B. S., & Mees, I. M. (2013). Practical Phonetics and Phonology: A Resource Book
for Students. New York: Routledge.
Crystal, D. (2008). A Dictionary of Linguistics and Phonetics (6th ed.). Oxford: Blackwell.
Davis, S. (1999). On the representation of initial geminates. Phonology, 16(1)93-104.
Jenkins, J. (2006). The spread of EIL: A testing time for testers. ELT Journal, 60(1), 42-50.
Hardison, D. M., & Saigo, M. M. (2010). Development of perception of second language
Japanese geminates: Role of duration, sonority, and segmentation strategy. Applied
Psycholinguistics, 31(1), 81-99.
Hazan, V., Sennema, A., Iba, M., & Faulkner, A. (2005). Effect of audiovisual perceptual
training on the perception and production of consonants by Japanese learners of
English. Speech Communication, 47(3), 360-378.
Knight, R. A. (2012). Phonetics: A Coursebook. Cambridge: Cambridge University Press.
Lahiri, A., & Hankamer, J. (1988). The timing of geminate consonants. Journal of Phonetics,
16, 327-338.
Li, F. (2016). Contrastive study between pronunciation Chinese L1 and English L2 from the
perspective of interference based on observations in genuine teaching contexts.
English Language Teaching, 9(10), 90-100.
Lightbown, P. M., & Spada, N. (2013). How Languages are Learned 4th edition - Oxford
Handbooks for Language Teachers. Oxford: Oxford University Press.
McCully, C. (2009). The Sound Structure of English: An Introduction. Cambridge: Cambridge
University Press.
Munro, M. J., & Derwing, T. M. (2011). The foundations of accent and intelligibility in
pronunciation research. Language Teaching, 44(3), 316-327.
77
EFLIJ Volume 25 Issue 4 July 2021
78
Pojprasat, S., & Thirakunkovit, S. (2018). An exploratory study of Thai natives’ gemination
in English words. Veridian E-Journal, 11(1), 1103-1119.
Suntornsawet, J. (2019). Problematic phonological features of foreign accented English
pronunciation as threats to international intelligibility: Thai EIL pronunciation
core. Journal of English as an International Language, 14(2), 72-93.
Thirakunkovit, S. (2020). Production of English geminate consonants by Thai learners.
Journal of Language and Culture, 39(1), 19-36.
Thompson, T., & Gaddes, M. (2005). The importance of teaching pronunciation to adult
learners. The Asian EFL Journal, 2(1), 1-11.
Tsukada, K. (2009). Durational characteristics of English vowels produced by Japanese and
Thai second language (L2) learners. Australian Journal of Linguistics, 29(2), 287-299.
Ying, L. (2016). Degree of Foreign Accent in English Production by Japanese, Thai and
Italian Adults and Children. The Asian EFL Journal, 95, 75-90.
78
EFLIJ Volume 25 Issue 4 July 2021
79
Appendix A
Test Materials for the Reading Aloud Task
Task 1: Reading Aloud
Instructions: Look at the following words and phrases. Read each word and phrase twice
within the sentence “Please say the word/phrase __________ again.” at a normal speed.
happy
hobby
support
spelling
floppy
runner
slipper
misspell
cancellation
bossy
pippin
ladder
effect
irresponsible
illegal
yellow
unknown
ammonia
curriculum
differ
innate
immoral
unnamed
copy
top pick
one nurse
nice sock
bad day
good dress
fun name
beautiful life
dim morning
Instructions: Look at the following words. Read each word twice within the sentence
“กรุณาพูดคาว่า __________ อีกครัง้ ” at a normal speed.
คันนา
/khan-naa/
กินรี
/kin-na-rii/
กินนอน
/kin-nɔɔn/
ปากกา
/paak-kaa/
ธรรมะ
/tham-ma/
งงงวย
/ŋoŋ-ŋuai/
Instructions: Look at the following sentences. Read each sentence twice at a normal speed.
1. He became the prime minister in 2000.
2. This address seems wrong.
3. It makes sense to me.
4. Jill loves to go jogging in the morning.
5. Johnny is a calm man.
Task 2: Dictation
79
EFLIJ Volume 25 Issue 4 July 2021
80
Instructions: Listen to the recording and write down what you hear. You will hear each item
twice. You have 20 seconds to write down your answer for each hearing.
1.
First hearing: …………………………………………………………………..…..
Second hearing: …………………………………………………………………….
2.
First hearing: ……………………………………………………………………....
Second hearing: …………………………………………………………………….
3.
First hearing: …………………………………………………………………..…..
Second hearing: …………………………………………………………………….
4.
First hearing: …………………………………………………………………...….
Second hearing: …………………………………………………………………….
5.
First hearing: …………………………………………………………………...….
Second hearing: …………………………………………………………………….
6.
First hearing: …………………………………………………………………...….
Second hearing: …………………………………………………………………….
7.
First hearing: ………………………………………………………………...…….
Second hearing: …………………………………………………………………….
8.
First hearing: ……………………………………………………………...……….
Second hearing: …………………………………………………………………….
80
EFLIJ Volume 25 Issue 4 July 2021
81
Appendix B
Means and SD Durations of Items Tested
Students
Items Tested
Low
Intermediate
High
Intermediate
Advanced
English
Instructors
Native Speakers
of English
Mean
SD
Mean
SD
Mean
SD
Mean
SD
Mean
SD
happy
0.166
0.039
0.151
0.028
0.159
0.030
0.136
0.033
0.103
0.008
pippin
0.178
0.047
0.179
0.043
0.168
0.036
0.173
0.047
0.105
0.025
support
0.199
0.036
0.189
0.034
0.179
0.033
0.290
0.442
0.152
0.040
slipper
0.197
0.038
0.197
0.038
0.214
0.031
0.219
0.036
0.095
0.012
floppy
0.198
0.061
0.177
0.052
0.173
0.036
0.150
0.085
0.085
0.076
ladder
0.113
0.048
0.095
0.046
0.083
0.036
0.069
0.025
0.047
0.021
hobby
0.164
0.038
0.151
0.046
0.134
0.029
0.122
0.028
0.071
0.022
runner
0.144
0.031
0.116
0.027
0.117
0.025
0.111
0.024
0.050
0.006
innate
0.159
0.032
0.141
0.036
0.131
0.028
0.125
0.034
0.067
0.007
unknown
0.164
0.039
0.144
0.026
0.153
0.027
0.159
0.022
0.176
0.027
unnamed
0.184
0.034
0.161
0.025
0.158
0.027
0.163
0.029
0.171
0.025
immoral
0.147
0.028
0.144
0.026
0.141
0.029
0.155
0.038
0.116
0.026
ammonia
0.128
0.031
0.118
0.025
0.125
0.023
0.122
0.026
0.087
0.006
differ
0.176
0.027
0.207
0.231
0.161
0.024
0.179
0.043
0.129
0.011
copy
0.160
0.055
0.162
0.054
0.157
0.044
0.123
0.066
0.099
0.050
effect
0.182
0.033
0.159
0.032
0.151
0.027
0.159
0.015
0.140
0.024
misspell
0.130
0.039
0.135
0.042
0.137
0.045
0.153
0.062
0.153
0.019
bossy
0.166
0.030
0.158
0.029
0.154
0.028
0.164
0.029
0.151
0.010
irresponsible
0.087
0.032
0.070
0.012
0.080
0.032
0.074
0.029
0.077
0.027
spelling
0.085
0.033
0.087
0.025
0.087
0.026
0.088
0.031
0.054
0.000
yellow
0.115
0.101
0.091
0.030
0.095
0.028
0.087
0.021
0.057
0.009
illegal
0.098
0.043
0.102
0.035
0.108
0.031
0.114
0.023
0.088
0.008
cancellation
0.075
0.031
0.080
0.020
0.074
0.020
0.079
0.025
0.062
0.012
curriculum
0.053
0.019
0.061
0.019
0.064
0.024
0.061
0.017
0.039
0.004
top pick
0.225
0.069
0.237
0.080
0.213
0.078
0.261
0.103
0.229
0.045
bad day
0.157
0.030
0.162
0.061
0.156
0.038
0.172
0.031
0.177
0.029
81
EFLIJ Volume 25 Issue 4 July 2021
82
Students
Items Tested
Low
Intermediate
High
Intermediate
Advanced
English
Instructors
Native Speakers
of English
Mean
SD
Mean
SD
Mean
SD
Mean
SD
Mean
SD
good dress
0.172
0.041
0.179
0.066
0.158
0.045
0.189
0.066
0.177
0.019
one nurse
0.185
0.032
0.192
0.081
0.164
0.057
0.200
0.099
0.164
0.029
fun name
0.204
0.042
0.207
0.114
0.170
0.055
0.186
0.058
0.181
0.040
dim morning
0.170
0.064
0.181
0.068
0.165
0.045
0.225
0.094
0.199
0.023
nice sock
0.190
0.079
0.215
0.133
0.163
0.043
0.240
0.119
0.211
0.060
beautiful life
0.078
0.022
0.093
0.024
0.093
0.030
0.101
0.041
0.078
0.008
He became
the prime
minister in
2000.
0.133
0.105
0.086
0.022
0.088
0.026
0.153
0.208
0.096
0.013
Johnny is a
calm man.
0.150
0.032
0.183
0.059
0.147
0.022
0.191
0.091
0.151
0.018
This address
seems
0.312
wrong.
0.134
0.323
0.115
0.257
0.151
0.309
0.143
0.200
0.033
It makes
sense to me.
0.234
0.085
0.234
0.085
0.189
0.072
0.222
0.075
0.173
0.024
Jill loves to
go jogging in 0.180
the morning.
0.087
0.157
0.082
0.124
0.052
0.143
0.070
0.107
0.047
82
EFLIJ Volume 25 Issue 4 July 2021
83
Appendix C
Means and SD Durations of Thai Words
Students
Thai words
Low
Intermediate
High
Intermediate
Mean
Mean
SD
English
Instructors
Advanced
SD
Mean
SD
Mean
SD
คันนา
/khan-naa/
0.193
0.030
0.180
0.030
0.178
0.022
0.195
0.042
กินนอน
/kin-nɔɔn/
0.205
0.038
0.200
0.050
0.191
0.026
0.229
0.064
กินรี
/kin-na-rii/
0.106
0.032
0.101
0.030
0.093
0.025
0.102
0.022
ธรรมะ
/tham-ma/
0.200
0.032
0.182
0.045
0.177
0.022
0.200
0.032
ปากกา
/paak-kaa/
0.194
0.044
0.182
0.043
0.179
0.024
0.203
0.033
งงงวย
/ŋoŋ-ŋuai/
0.182
0.042
0.219
0.065
0.192
0.037
0.280
0.128
83
EFLIJ Volume 25 Issue 4 July 2021
84
English-medium Instruction and Translanguaging
BethAnne Palsrud, Tian, Zhongfeng and Jeanette Toth (Eds). Bristol: Multilingual Matters,
2020. ISBN 13: 978-1-78892-731-4 (pbk). Pp. 208.
Reviewed by Marilyn Lewis, Honorary Research Fellow, The University of Auckland
The status of the English language in the world and, more specifically in classrooms
internationally, has been the focus of several recent articles and books. In a volume edited by
Sifakis and Tsantila (2019), contributors from countries where English is not the first language,
such as Germany, Greece, Portugal and Turkey, as well as writers from English language
countries, reported on the place of English in their classrooms. Marr and English (2019) also
investigated questions relating to TESOL teachers and the language they use with their students,
building partly on their own experiences. Most recently, in this journal, Tamin (2021) discussed
differences between policies and practices in Pakistan. Perhaps these sources came out too late
to be referenced by the present editors, since none of them are listed here.
This most recent book also draws on contributors’ experiences in many countries. It has
11 chapters plus a foreword, introduction, epilogue and conclusion. One of the collection’s
refreshing aspects is its inclusion of teachers’ voices from countries whose voices we hear less
often. (I use the words “less often” to be safe, but in my case the truer word would be “never”.)
First, a couple of definitions. The term EMI is used for contexts where English is used
exclusively as a medium of instruction. When it comes to the title’s term of ‘translanguaging’
the writer of the foreword speaks of “the multiple ways in which it is understood” (p. xvi). More
than one article in the book mentions that the word was originally coined in Welsh in the 1980s
to describe what happened in Welsh-English bilingual classrooms. The varied descriptions from
chapter to chapter confirm the difficulty of coming up with one, simple definition for this word
which, very briefly, involves the use of multiple languages.
Since the book is not divided into sections nor arranged in any special chronology, the
reader can feel free to pick chapters in any order. In my case, curiosity had me starting with
countries I knew least about. In the Maldives, a Muslim country, children are exposed to Arabic
and English as well as their local language. The widely published writer, Naashia Mohamed,
starts by outlining how and why English came to have status in the Maldives to the point where
it is now the medium of instruction for all subjects except the other two language classes. Given
84
EFLIJ Volume 25 Issue 4 July 2021
85
the careful details about this case study of four friends aged between six and eleven telling
stories, readers wanting to replicate it in their own country have a strong starting point.
For a study of a different age group and country there is the chapter on Khazakhstani
students at an English-medium university, One of the three authors, Sulusash Kerimkulova,
speaks from her starting point of more than 40 years’ experience of teaching and research in
that country. The chapter opens with some more definitions of translanguaging, which the three
authors describe as “both an approach to teaching and learning, and the application of that
learning” (p. 141). The ’trans’ part of the term must have special meaning in a country which,
apart from Khazak and Russian, has 130 ethnic groups. One interesting feature of this study is
that it reports students’ own views about which languages they chose to study and why.
Next I turned to places that were familiar to me, starting with a Cambodian study, the
country where I once taught at a university in the early 70s. At the unnamed Phnom Penh
university in this 2009 study, bilingual students and staff in the M.Ed programme were surveyed
for their translanguaging attitudes, practices and views. A sense of that country’s international
contacts came in the list of countries from which the staff involved in the study had gained their
PhDs: four from Japan and one each from the United States and Germany.
The other seven studies are from South Africa (2), Kenya, Malawi, Hong Kong, Italy and
Japan. The classes investigated in the book include primary, secondary and tertiary levels, the
last being in the majority. They vary in their methods and, not surprisingly, in their results.
It would be interesting to hear how many of the groups mentioned in the book are amongst
its readers. Will it be the policy makers whose decisions determine which languages are used
in their country’s education systems? Probably not, because such choices are usually made for
political rather than educational reasons. Will it be researchers? Probably, because the wideranging topics and methods give them a good starting point. Hopefully, classroom teachers will
also get hold of the book so they can try out some of the many interesting tasks described here.
Also, if they happen to have students from any of the countries mentioned, then the contents
will help them understand those people’s backgrounds. The book is recommended to all these
groups.
References
Marr, T., & English, F. (2019). Rethinking TESOL in diverse global settings: The language and
the teacher in a time of change. London: Bloomsbury.
Sifakis, N. C., & Tsantila, N. (Eds.). (2019). English as a lingua franca for EFL contexts.
Bristol: Multilingual Matters.
85
EFLIJ Volume 25 Issue 4 July 2021
86
Tamin, T. (2021). Monolingual Policies and Plurilingual Practices in English Language
Classrooms: Addressing Shared Guilt and Threats. Asia EFL Journal, 25(2), 2-28.
Bio data
Marilyn Lewis is an Honorary Research Fellow at Auckland University and may be e-mailed
at mn.lewis@auckland.ac.nz
86
EFLIJ Volume 25 Issue 4 July 2021
87
Critical Thinking in ELT: A Working Model for the Classroom
Paul Dummett and John Hughes. National Geographic Learning: Boston, USA, 2019.
Pp. viii +158.
Reviewed by Thiri Soe, Graduate School of International Cultural Studies, Tohoku University,
Japan
Incorporating critical thinking in all different subject areas is highly recommended as an
educational goal. In the field of English Language Teaching, Dummett and Hughes’s (2019)
Critical Thinking in ELT: A Working Model for the Classroom is a comprehensive introductory
book with guidelines and models for integrating critical thinking activities in language skills.
The book delivers ELT teachers step by step explanations from the operational definitions and
concepts of critical thinking to the application of critical thinking activities in ELT materials in
its seven chapters.
Chapter 1 explains readers why critical thinking has to be a central role in education. The
increasing access of internet and information from digital media is the main reason that requires
21st century learners to evaluate and resonate the information before them. From the subjectspecific perspective, there is a strong bond between critical thinking and language acquisition.
Deeper processing and production of language is achieved by means of critical thinking. It is
also stressed in this chapter that two types of critical thinking skills, higher order thinking skill
and lower order thinking skill are neither inferior nor superior to each other. Having the balance
of both thinking skills is crucial in effective language learning.
Chapter 2 states how critical thinking benefits learners in such factors as academia,
business, early education, and everyday life and internet. Learners can get more engagement
and learning autonomy by exercising thinking activities from ELT textbooks. Critical thinking
activities can be added into various stages of lessons from the smallest unit of discourse like
words and syllables to larger discourse texts like paragraphs and articles. Not only that, critical
thinking activities can be applied to receptive skills as well as productive skills. Dummett and
Hughes’ innovative lesson planning ideas for inclusion of critical thinking into different
language levels and skills can be learnt in Chapter 3, 4, and 5. Their examples of critical
thinking activities are built on the language tasks and skills that ELT textbooks cover: grammar,
vocabulary, pronunciation, reading, listening, speaking, and writing.
Inductive reasoning, comparative analysis, rephrasing and reformulation, translation,
rearranging word order, identifying word order, and identifying idiomatic usage are example
87
EFLIJ Volume 25 Issue 4 July 2021
88
activities applicable in grammar teaching. For teaching vocabulary, comparing words,
considering the word scope, understanding grammatical function of words, considering
emotional, social, and cultural aspects of words, and understanding literal and figurative uses
of words are critical thinking encouraging activities. Analyzing pronunciation patterns to see
how they convey meaning, analyzing one’s own pronunciation to remove areas of ambiguity,
and conscious awareness of the pronunciation patterns of one’s interlocutor are example
activities for teaching pronunciation critically.
Another crucial idea the authors state is that critical thinking activities can be infused with
both receptive and productive skills. Identifying authors’ aims, examining the reliability of
arguments, separating facts from opinions, and looking for bias from texts are critical thinking
encouraging activities in teaching receptive skills. Three factors constitute critical language
productive skills: generating ideas, setting criteria for language output, and evaluating the
output. In setting up the criteria, learners need to pinpoint two sub-goals: identifying the goal
of the text, and the perspective of their text readers.
Chapter 6 engages ELT teachers to the extensive need of critical thinking skills beyond
classrooms, by suggesting activities learners should apply in confronting different types of
literacy. Critical thinking is an essential armor for the learners to wear in exposing themselves
to twenty-first century literacies such as visual literacy, information literacy, media literacy,
and cultural literacy. The reason for an ELT teacher to focus on these literacies is that they
appear as language input to the learners. Learners need to perceive the information judiciously
to evaluate their factual accuracy.
The last chapter, chapter 7, motivates ELT syllabus writers and teachers for the balanced
implication of lower-order and higher-order thinking skills because both of the skills involve in
the working model of critical thinking in ELT presented by Dummett and Hughes. This
concluding chapter gives readers acute discussion on three topics that are commonly addressed
and asked by English language teachers around the world. These topics are evaluating lessons
for critical thinking, teaching critical thinking at lower language levels, and promoting critical
thinking.
Overall, the book Critical Thinking in ELT: A Working Model for the Classroom is a worth
reading textbook and it bridges critical thinking and English language education. Having
explained concepts and definitions of critical thinking activities followed by example lessons,
the book is an evidence-based manual for the incorporation of critical thinking in ELT
curriculum. At the end of the last chapter, there are suggested answers to sample exercises and
ask yourself question. The book will surely stand as a successful course book in training ELT
88
EFLIJ Volume 25 Issue 4 July 2021
89
teachers for increasing awareness to the application of critical thinking activities in their
classrooms.
Bio data
Myanmar academic, Thiri Soe studies at the Graduate School of International Cultural Studies,
Tohoku University, Japan. thirisoe6@gmail.com
89
View publication stats