Washback: Exploring what constitutes “good” teaching practices

Journal of English for Academic Purposes 30 (2017) 1e12

Contents lists available at ScienceDirect

Journal of English for Academic Purposes journal homepage: www.elsevier.com/locate/jeap

Washback: Exploring what constitutes “good” teaching practices Melissa Barnes Monash University, Faculty of Education, McMahons Road, Frankston, VIC 3199, Australia

a r t i c l e i n f o

a b s t r a c t

Article history: Received 10 February 2017 Received in revised form 6 October 2017 Accepted 6 October 2017

Realising the social consequences involved in language testing, many researchers have investigated “washback”, or the influence of testing on teaching and learning. While it is widely acknowledged that the nature of washback is dependent on context (Cheng, 2005; Tsagari, 2011), the very definition of washback is problematic due to its reliance on what constitutes “good” teaching and learning practices, which can differ from one educational context to another. This article explores the attitudes, beliefs and teaching practices of four teachers teaching both TOEFL iBT preparation courses and general English classes in Vietnam with the aim to better understand how washback (positive or negative) is realised in these contexts. In this qualitative case study, the participating teachers were observed and interviewed in both their TOEFL iBT preparation and general English class. The study revealed that teachers’ beliefs about what constitutes “good” teaching practices differed depending on the teaching context, or more specifically, the role and purpose of the course (test preparation versus general English). © 2017 Elsevier Ltd. All rights reserved.

Keywords: Washback CLT TOEFL iBT Vietnam English for academic purposes

Washback, or “backwash,” is used in applied linguistics to refer to the influence of testing on teaching and learning and has been described by researchers as a complex phenomenon consisting of numerous mediating factors. While most researchers agree that washback exists, they also acknowledge that there are varying degrees (Alderson & Hamp-Lyons, 1996; Spratt, 2005; Watanabe, 2004), intensity (Cheng, 2005; Green, 2007) and direction (Green, 2007) of washback. Washback is often characterised as either positive or negative and is recognized as playing an important role in the relationship between testing, teaching and learning. Contrasting negative and positive washback, Taylor (2005) argues that negative effects occur “when a test's content or format is based on a narrow definition of language ability” while positive effects occur when the testing procedures encourage “good” teaching practices (p. 154). While Taylor's definition of positive and negative washback appears reasonable, what constitutes “good” teaching practices is often contextually defined. The reality is that teaching approaches evolve and change and therefore caution must be taken in defining phenomenon such as washback in connection with a preferred teaching paradigm (e.g. Communicative Language Teaching). Similar to Bailey (1999), Taylor (2005) suggests that tests, which are not aligned with or “run contrary” to the principles and practices of Communicative Language Teaching (CLT), generate negative washback (p. 259). She goes on to argue that “it is unlikely that a test based on outmoded theoretical constructs will lead to positive washback. Since, in many parts of the world, a narrow view of linguistic competence has been replaced by a broader perspective on communicative competence” (p. 276).

E-mail address: [email protected]. https://doi.org/10.1016/j.jeap.2017.10.003 1475-1585/© 2017 Elsevier Ltd. All rights reserved.

2

M. Barnes / Journal of English for Academic Purposes 30 (2017) 1e12

The views of Taylor (2005) and Bailey (1999) suggest that the way in which a test is designed, particularly if it aligns itself with either CLT or traditional teaching practices can determine whether the test's washback is positive or negative. This can then be realised through “good” teaching practices. A look into several prominent washback studies will attempt to uncover how the language testing community has come to define washback by the presence or absence of CLT. 1. CLT & high-stakes language testing CLT emerged from a paradigm shift in the 1970's when linguists and language educators began to view language as a system for the expression of meaning rather than a system of syntactic rules. This shift began in the late 1960's when Hymes (1967; 1972) introduced a theory or model of communicative competence within a first language framework. Hymes distinguished between actual language used, language knowledge and the “ability for use.” Using Hymes' first language model to inform their own, Canale and Swain (1980) adapted the model to suit a second language framework and created more specific domains of language knowledge: grammatical, sociolinguistic, and strategic competence. They argued that communicative competence consists of the interaction between knowing the rules of grammar (grammatical competence), the rules of language use (sociolinguistic competence) and the skills to overcome problems in communication (strategic competence). It was Canale and Swain's model of communicative competence that would became the theoretical backbone of CLT and acceptance became widespread. Nunan (1998) described it as the most pervasive change to teaching practice over the past twenty years. According to Richards and Rodgers (2001), the aim of CLT is to promote real-life, authentic communication tasks which are meaningful to the learner and support their learning processes. Holliday (1994) argues there is a strong and weak version of the communicative language approach. The weak communicative approach focuses on language use with an emphasis on student talk time and pair and group work. This approach is underpinned by the belief that communication facilitates learning (Holliday, 1994, p. 170). On the other hand, the strong version focuses on how language works in discourse or how students engage with the text (Holliday, 1994, p. 171). Given the focus on the student and his or her learning in a communicative approach, many teachers mistakenly associate the weak version (e.g. pair and group work) as central to this approach. In classroom contexts in which student numbers are much larger, CLT and a student-centred approach appears contextually inappropriate. Not surprisingly, many claim that transferring communicative teaching methods to other parts of the world can be problematic because they are not always appropriate to the local context (Ellis, 1996; Holliday, 1994; Kramsch & Sullivan, 1996; Lewis & McCook, 2002; Pham, 2007; Spicer-Escalante & deJonge-Kannan, 2014). On the other hand, however, Pham (2007) argues that, “undoubtedly, CLT originates in the West, but to decide a priori that this teaching approach is inappropriate to a certain context is to ignore developments in language teaching” (p. 196). This proves interesting for two reasons. First, there is an assumption that CLT is “Western” and second that to ignore CLT is to ignore language teaching development. Several empirical studies have been conducted in Vietnam to examine Vietnamese teacher's beliefs about CLT (e.g. Ellis, 1996; Lewis & McCook, 2002; Pham, 2007). In Lewis and McCook’s (2002) study, Vietnamese secondary teachers who were attending a workshop on CLT were asked to reflect on teaching practices in Vietnam in the form of journal entries. Though their research may have uncovered some interesting perspectives on how the 14 participating teachers perceived teaching and learning, there was little evidence to suggest that these beliefs were aligned with actual practice. Similarly, Pham (2007) conducted a study on teachers’ beliefs on the use of CLT, in which data is collected from both conversations and classroom observations. Pham selected three female teachers, who had completed graduate degrees from Australian universities, and were teaching at a university in Vietnam. Pham (2007) claims that the teachers in his study embrace CLT but when it comes to applying its principles in practice, they encountered many problems. The teachers did not perceive themselves as being successful, especially with pair and group work. While the teachers expressed their support for CLT practices, their attempt to appropriate them to a Vietnamese classroom offered many contextual and cultural constraints. 2. Washback studies on high-stakes national english examinations With the growth of communicative language teaching (CLT) in the 1970's, the need for practical measures of language performance rose. In contrast with traditional pencil and paper tests, performance testing requires a test taker performance which is observed and judged. McNamara (1996) argues that this shift was in response to two main needs: the need to develop selection procedures (e.g. foreign students studying at English-medium universities) and to align testing with developments in language teaching which had been influenced by theories of communicative competence. Language proficiency, which is currently defined in terms of communicative competence in most high-stakes international language tests (e.g. TOEFL & IELTS), is what is predominately measured in most language tests. High stakes tests have often been used in language education to change teaching and learning practices as intended by policymakers and test designers (Qi Luxia, 2005). A number of empirical studies have discovered that while test designers have intended for the test to introduce a shift to communicative language teaching practices, they have fallen short of their aim (Andrews, 1994; Chen, 2006; Cheng, 2004, 2005; Wall & Alderson, 1993). Wall and Alderson (1993) used classroom observation and teacher and student interviews to investigate the washback of the O-Level English examination in Sri Lanka. This test was linked to a textbook series that introduced new ideas in terms of content and methodology (Wall & Alderson, 1993, p. 44). The test and the textbook series was underpinned by a

M. Barnes / Journal of English for Academic Purposes 30 (2017) 1e12

3

communicative language approach and aimed to influence both how and what the teachers taught. However, while the content in the classrooms changed, teachers did not change the way in which they taught. While it was intended that the introduction of the test and textbooks would encourage communicative practices in the classroom (and therefore be evidence of positive washback), change was solely evident a change in content and not in a change of teaching methodology. Cheng (2004; 2005) and Chen (2006) examined the washback of a national English test in Hong Kong and Taiwan, respectively as part of a government initiative to encourage more communicative teaching practices in the classroom. However, it was discovered in both studies that a gap existed between teachers' perceptions of what was expected of them with their actual actions within the classroom. Though the teachers acknowledged that a change in teaching methodology must occur due to the changes in the examination and curriculum, few actually implemented these methods. Chen (2006) reports that even though the Basic Skills Test (BCT), used within Taiwan junior high schools as a high school entrance exam for English, aimed to encourage CLT and students' communicative competence, teaching remained test-oriented with a focus on de-contextualised language points rather than communication. The teacher who was observed chose methods aimed at teaching toward the test rather than implementing the curriculum's goal of communicative competence. Likewise, Cheng (2004; 2005) discovered that the introduction of the Hong Kong Certificate of Education Examination (HKCEE) did not show a significant amount of change to the teachers' current teaching methods. While the exam aimed to facilitate more integrative and task-based approaches to teaching, teaching remained test-oriented, highly controlled, and content-based, thus exhibiting negative washback. In a more recent washback study on the English National Examination (or in Bahasa Malay Ujian Nasional) in Indonesia, Furaidah, Saukah and Widaiti (2015) explored how the test influenced “high-achieving” and “low-achieving” schools. One area that they examined was the amount of class time that was dedicated to preparing students for the test. They argue “it seems that only the washback manifested in the increased amount of time allocation for UN [Ujian Nasional or National Examination] subjects is likely to be beneficial as it can potentially support the development of communicative competence as stated in the standards of competences of the English subject” (Furaidah, Saukah, & Widait, 2015, p. 16). Again, beneficial washback is linked to communicative competence and there is a tenuous link between spending time on preparing students for the test and their development of communicative competence. In addition to washback studies associated with modifications or changes to national tests, there are a number of washback studies exploring the influence of international language proficiency tests which measure students' academic English proficiency. While there are a number of tests that measure students' academic English (e.g. Test of English for Academic Purposes [TEAP]) and English for the workplace (e.g. Test of English for International Communication [TOEIC]), the following section explores the washback literature relating to two international language tests, the International English Language Testing System (IELTS) academic module, and the Test Of English as a Foreign Language (TOEFL), which provide benchmarks for the level of ‘academic English’ students are to demonstrate for entry into English-speaking universities. 3. Washback studies on high-stakes international proficiency tests The IELTS academic module, aims to measure students' academic English proficiency with several studies exploring the test's washback on test preparation courses (e.g. Hawkey, 2006; Hayes & Read, 2004). Looking beyond just test preparation courses, Green (2007) investigated the washback of IELTS in both test preparation courses and courses with a focus on English for Academic Purposes (EAP). More specifically, Green (2007) focused on learning performance on IELTS writing tests administered at the start and finish of three different types of courses: test preparation courses, pre-sessional courses in EAP and courses with features of both. His results suggest that test preparation courses did not improve scores to a greater extent than the other courses. “There is no evidence here that course providers were able, through dedicated test preparation practices, to exploit test design characteristics to boost scores,” (p. 93). This, then, suggests that focusing on test-taking strategies is not more advantageous than focusing on academic English skills more holistically. Similar to Alderson and Hamp-Lyons (1996), Green (2007) observed classes and interviewed students and teachers from both TOEFL test preparation courses and general English language classes in order to investigate the washback of the TOEFL Paper-Based Test (PBT). They found a discrepancy between the percentage of time spent on various activities (e.g. pair work) between the TOEFL and non-TOEFL classes, suggesting that the TOEFL did not encourage communicative practices in the classroom. However, they also found that teachers differed greatly from one another in their personalities and teaching styles (Alderson & Hamp-Lyons, 1996). They argued, “While the nature of the TOEFL test, consisting as it does primarily of discrete items focusing on language below discourse level, and occurring independent of each other, may not lend itself immediately to discourse-based communicative teaching, it does not by its nature preclude this” (Alderson & Hamp-Lyons, 1996, p. 295). While they found that many of the TOEFL test preparation courses lacked communicative teaching practices, they cautioned placing blame solely on the design of the test as their data suggested that the textbooks and teacher personality and beliefs also have an influence. 4. The development of the TOEFL iBT Given the mounting criticism of the TOEFL, a call for change was initiated in 1995 by various constituencies, who believed that the test should reflect communicative competence models, include more constructed-response tasks, directly measure writing and speaking, integrate language skills, and measure a student's ability to communicate in an academic setting

4

M. Barnes / Journal of English for Academic Purposes 30 (2017) 1e12

(Educational Testing Service, 2007). Educational Testing Service (ETS), the governing body of TOEFL, claimed that one of the reasons for the introduction of a new TOEFL test was “to keep up with the best practices in language teaching by using a communicative and integrated skills approach” (Educational Testing Service, 2006, p. 8). The language “best practices” combined with “communicative and integrated skills approach” suggests that the test designers felt (or hoped) that these approaches would be embraced by the classroom, in contrast to methods aligned with discrete-point testing that the PaperBased test (PBT) had once been criticised for. According to Jamieson, Jones, Kirsch, Mosenthal, and Taylor (2000), who created a framework of the new TOEFL, language teachers felt that TOEFL's use of discrete-point testing had a negative effect on teaching and learning. “ESL/EFL teachers are concerned that discrete-point test items, and the exclusive use of traditional, multiple-choice items to assess receptive skills, have a negative impact on instruction” (Jamieson et al., 2000, p. 3). Therefore, the development of a new TOEFL focused on creating a test, which would have a positive washback effect on the classroom through a communicative and integrated skills approach. The TOEFL iBT differs from previous formats (PBT and CBT) in that it focuses on all four skills and academic communication and is underpinned by an integrated approach. First, all four language skills (speaking, listening, reading and writing) are included in the TOEFL iBT. TOEFL iBT is the first format to include a speaking section, in which structure (grammar) is assessed through the speaking and writing sections rather than as a separate subtest. Although the TOEFL iBT shares a computermediated format with its predecessor the CBT, its introduction of a semi-direct speaking sub-test requires test-takers to speak into a microphone attached to their headset so that a digital file can be recorded and transmitted to the administrators. Another main feature of the iBT is its focus on academic communication. For example, students listen to longer conversations and lectures set in an academic context (e.g. a student asking a librarian a question, a Geography lecture, etc.) and are encouraged to take notes, which was not allowed in previous formats. Integrated tasks, in which students gather information from a variety of sources and respond with a written or spoken response, reflect authentic academic communication and skills needed to be successful in an academic setting. The writing and speaking sections consist of both independent tasks (based on test taker's opinion and background knowledge) and integrated tasks (based on written and spoken texts provided within the test). Now 12 years after the TOEFL iBT's 2005 rollout, Education Testing System (ETS), the test's administrator, argues that TOEFL “is the most widely respected English-language test in the world, recognized by more than 10,000 colleges, universities and agencies in more than 130 countries, including Australia, Canada, the U.K. and the United States” (ETS, 2017; para. 2). While research into the washback effects of the TOEFL iBT is still limited, Wall and Horak’s (2006; 2008; 2011) 5-year longitudinal study provides insight into how the TOEFL iBT, which was designed to espouse a communicative and integrative skills approach, was realised in test preparation classrooms in Central and Eastern Europe. The first phase was dedicated to discovering what type of washback the developers of the iBT intended, in addition to, describing what TOEFL preparation courses in Central and Eastern Europe looked like. Phase 1 acted as a baseline study for the broader longitudinal study. After they completed Phase 1, they commented that: … There was a general hope that the new TOEFL would lead to a more communicative approach to teaching and that preparation classes would pay more attention to academic tasks and language, there would be more speaking, there would be integrated skills work, and some aspects would change in the teaching of other skills (Wall & Horak, 2006, p. 17 as cited in Wall & Horak, 2008, p. 3). Their findings from the longitudinal study align with the findings from other washback studies, suggesting more evidence of influence on content rather than on methodology (e.g. Alderson & Hamp-Lyons, 1996; Stetcher, Chun, & Barron, 2004) and the power of coursebooks in defining what is taught (e.g. Alderson & Hamp-Lyons, 1996; Author, 2016; Cheng, 2004, 2005; Hayes & Read, 2004; Tsagari, 2011; Wall & Alderson, 1993). Wall and Horak (2011) also discovered the difficulty in defining the types of washback that were expected by the test designers and thus what defines positive and negative washback. They felt that there were statements made in the TOEFL framework documents suggesting that the test aimed to take a communicative language approach (Wall & Horak, 2011, p. 124). However, the test designers and experts did not comment specifically on the teaching methods that should be employed in future TOEFL preparation courses (Wall & Horak, 2011, p. 124) or the types of impact/washback desired (Wall & Horak, 2011, p. 135). The problem exists, however, that while all tests presumably aim to produce positive washback, it is unclear what positive washback looks like in different classrooms and to what extent this is caused by the test, teaching materials or teachers. Moreover, it is suggested in the studies mentioned in this article that positive washback is defined by the presence of a communicative language approach in the classroom and the test's ability to encourage such a paradigm in more traditional classrooms. In light of this, a study exploring the beliefs, attitudes and teaching practices of English language teachers in Vietnam aims to reveal the nature of washback in this particular education context. 5. Background of study In 2009, a broader study was conducted to investigate the washback of the TOEFL iBT on general English and TOEFL iBT preparation courses in Vietnam (Withheld, 2016). Vietnam was chosen as the context for this study because of its growing market for study abroad programs in English-speaking countries and history of more traditional teaching methods. While there were several research questions that shaped the broader washback study on the TOEFL iBT, this paper will only focus on one of these research questions:

M. Barnes / Journal of English for Academic Purposes 30 (2017) 1e12

5

What are the washback effects on how TOEFL iBT test preparation courses are taught)? In other words, to what extent did the 2007 rollout of the TOEFL iBT in Vietnam influence teaching methods and activities? 6. Methods As the classroom is an intricate social construction, a case study, underpinned by an interpretivist approach, was chosen to allow for a more in-depth investigation of the classroom activities and the meanings they construct. This case study incorporated both qualitative and quantitative perspectives to allow for a more robust data set. In order to explore the washback effects on how the TOEFL iBT was taught, data were collected though interviewing four teachers and observing their classrooms. These teachers taught in both the private and public sectors of Vietnamese education, representing a national university (now referred to as NU), an American language centre (AL), a Vietnamese language centre (VL) and a home-based course (PC). The teachers ranged in age from 24 years of age to 60; two were Vietnamese, one American and the other British. Table 1 provides a brief profile of the four participating teachers, illustrating the wide range of backgrounds and educational and professional experiences present among them. All participating teachers were given pseudonyms to protect their identities. 6.1. Classroom observations The four participating teachers were observed in an iBT preparation class and a general English language class. Given that the rollout of the TOEFL iBT had already begun at the time of the study and a baseline study could not be conducted, the general English classes acted as a point of comparison. These classes ranged from 1.5 to 3 h per lesson with 2e3 classes per week. The TOEFL iBT preparation course, in particular, ranged from 72 to 85 contact hours and lasted from 3 to 5 months in the four institutions observed. While three of the teachers were observed twice, once in an iBT course and once in a general English course, the fourth teacher was observed teaching in an iBT course at both a language centre and a national university and a general English course in the national university. The rationale was to distinguish between a teacher's individual teaching style and the methods they use to adapt to their teaching context. Due to ethical considerations, the lessons were not audio recorded or video recorded but an observation scheme was used to collect data in real time. The Communicative Orientation of Language Teaching Observation Scheme “Part A” (COLT; Spada & Frohlich, 1995) was employed to analyse classroom events by identifying methods used and recording the various applications of content materials in real time. Part A of the COLT requires the observer to make detailed notes in real time on the activities and episodes that occur during the course of the lesson. Classroom activities, or events, are viewed as the unit of analysis, rather than a time period, and are analysed by a set of themes or codes established by Spada and Frohlich (1995). In the broader study, the coding conventions were used to identify five main components: time, participant organisation, content, student modality and materials. In addition to these five components, an analysis of activity type as a percentage of class time is employed (see also Cheng, 2005) as an adjunct to the participant organisation and content components. However, for the purpose of this paper, only the participant organisation and activity types will be discussed. 6.2. Teacher interviews After each observation, the teacher was asked to recount the lesson from his/her perspective, explicating thoughtprocesses and decisions that informed his/her teaching. The teacher was asked to comment and explain why he/she utilized certain methods and explain the ideology behind decisions made. A semi-structured interview (see Appendix) was used to extend the themes identified in the COLT (e.g. content and activities) and underpin the teacher's beliefs and knowledge about teaching and learning and the TOEFL iBT. The interviews were audio recorded and transcribed. The participants were given copies of the transcripts to check for accuracy, ensuring that the data collected was an appropriate representation of their attitudes, beliefs and perspectives. All the interview data was checked by participants with no amendments made.

Table 1 Background of participating teachers. Tuan Gender Nationality First language Academic qualifications

Male Vietnamese Vietnamese Bachelors in TESOLdVietnam Masters in LinguisticsdAustralia Years teaching English 15 Institution of employment NU Self–PC

Mike

David

Ly

Male American English Bachelors in EconomicsdUSA TEFL certification–Thailand 5 months AL NU

Male British English High school diplomadUK Online TEFL certification

Female Vietnamese Vietnamese Bachelors in EducationdVietnam

3 AL

2 VL NU

6

M. Barnes / Journal of English for Academic Purposes 30 (2017) 1e12

The COLT, field notes and transcriptions were then used to perform a content analysis and determine key themes (e.g., view of self, the TOEFL iBT, teaching TOEFL iBT, and the Vietnamese context). The raw data were first organized into preliminary categories, which were then given codes or themes. This process was ongoing until the researcher was content with the amount of description the data had generated and the themes that have surfaced. 7. Results and discussion 7.1. A look into the classroom Before the results of classroom observations are discussed, it is important to mention the type of materials used to shape and guide the classroom practices and activities observed. Table 2 below shows the four general English classes that were observed and the main textbook used for each course. In both Mike and Ly's courses at NU, while they incorporated various materials at different times in the course, they were not tied to one particular textbook or bound collection. More importantly, students in both of these classes played a large role in deciding what materials would be used in their classes. For example, Ly's “Speaking” course focused mostly on pronunciation and oral communication. In the class observed, four students were scheduled to give a presentation on a topic of their own choice, in which they researched and wrote a paper. Following each presentation, the other students asked questions and discussed the topic as a class. The only materials used in the class were those created by the students, which were then presented to the other students in both spoken and written form. In contrast to the general English class materials, all four institutions used a set textbook, or a set collection in their TOEFL iBT courses as is shown in Table 3 The Vietnamese Language centre (VL) and Tuan's private class were based around a collection of photocopied tasks from commercial TOEFL iBT textbooks, which were given to the students as a bound set (VL) or given out separately to students at each class (PC). The textbooks or collections used by all of the participating institutions were from commercial TOEFL iBT textbooks published primarily in the years 2006 and 2007. These textbooks followed a similar organisational pattern; the chapters were organized into subtests or macro skills, such as listening, speaking, writing and reading and then each subset was divided into academic skills required for that subset. These academic skills were identified by textbook designers as necessary skills needed for taking each subtest. Examples of academic skills in the listening chapter include “listening for details” or “listening for main ideas.” Each academic skill is first introduced and then followed by several practice exercises with questions for students to complete. After these exercises are completed, students move on to a different skill within that subtest. The materials collected from the five lessons observed, which were primarily from the textbooks or set collections described above, consisted of “tasks.” Examples of these tasks include: an introduction to an academic skill with practice exercises (e.g. “recognising coherence” in the reading subtest) or a writing practice test using a computer software program in the computer lab. Fourteen of the total sixteen tasks or activities employed in the classroom during the observations were from TOEFL iBT-specific materials, twelve of which came from commercial TOEFL iBT textbooks and the remaining two from the Educational Testing System (ETS) website and TOEFL iBT software. Of all the materials collected, only one teacher, David, incorporated non-TOEFL materials in his lesson, totalling only two out of the total sixteen tasks. The classroom observations attempted to address if and to what extent the introduction of the TOEFL iBT influenced how teachers taught in the classroom. The data collected from the Communicative Orientation of Language Teaching Observation Scheme (Spada & Frohlich, 1995) during the classroom observations provides insight into the participant organisation and activity types of the classes observed. The COLT, in addition to the work of Cheng (2005) and her analysis of “activity types,” helped define the following two categories as a way to describe the classroom activities in this study: Participant organisation: Who is talking and how are students participating in classroom activities? Activity Type: What types of classroom activities are employed? These categories help in describing how classroom activities were realised in the classrooms observed and how much time was spent on each activity as a percentage of class time. First, by analysing classroom activities by way of participant organisation, the roles of teacher and student in the classroom are explored and defined. Participant organisation helps to identify who is participating and how they are participating in classroom activities. Patterns of participant organisation in the COLT are described by Spada and Frohlich (1995) in terms of: 1) is the teacher working with the whole class, 2) are students working in groups, or 3) are they working individually? Patterns of participant organisation found in both the iBT preparation and general English classes are reported in Table 4 below (the shaded columns highlight the TOEFL iBT preparation classes), which reports the participant organisation patterns in the lessons observed in this study as a percentage of total class time. The

Table 2 General english materials. AL Course

Advanced Listening and Speaking Main textbook Cutting Edge: Advanced Who decides on textbook or materials? Centre

NU (Mike)

NU (Ly)

Private class (PC)

Writing

Speaking

English for Academic Purposes (EAP)

None None Collectiondgeneral & test prep Teacher & Students Teacher & Students Teacher

M. Barnes / Journal of English for Academic Purposes 30 (2017) 1e12

7

Table 3 TOEFL iBT preparation materials. American Language Centre (AL) Main textbook

Developing Skills for the TOEFL iBT: Intermediate (Edmunds & McKinnon, 2006) Mastering Skills for the TOEFL iBT: Advanced (Macgillivray, Yancey, & Malarcher, 2006) Who decides on Centre textbook?

Vietnamese Language Centre (VL)

National University (NU) Private class (PC)

Collection of several TOEFL iBT textbooks

Mastering Skills for TOEFL Collection of several TOEFL iBT textbooks iBT: Advanced (Macgillivray et al., 2006)

Centre

Teacher

Teacher

whole class represents interactions between the teacher and student or the entire class (T to S/C), and student to students or class (S to S/C). Examples of student to student or class interaction (S to S/C) include oral presentations and class discussions, which are initiated and controlled by the students. As can be seen from Table 4, all but one of the five TOEFL iBT classes observed (Mike AL) spent more than half of the class time on teacher and student or class interaction (T to S/C), indicating a focus on teacher instruction. The general English classes were not much different with two of the four general English classes spending more than half of the class on teacher instruction. With the exception of Mike's TOEFL iBT lesson at AL, all of the TOEFL iBT classes had a higher percentage of teacher instruction than in their general English class. Interestingly, however, Mike and David incorporated group work in their general English classes but Mike did not use group activities in either of his TOEFL iBT lessons and David's use of group work was a result of using the only non-TOEFL source (conversation questions). Group work, a characteristic of Holliday’s (1994) “weak” CLT, was only observed in the two non-Vietnamese teachers' classrooms and when TOEFL-iBT was not the focus. Also, all of the TOEFL iBT lessons spent the majority of their class time on teacher instruction and individual student practice. Ly and Tuan spent all of their class time on teacher instruction and individual student practice while Mike spent 88% (AL) and 89% (NU) and David spent 65% due to his use of the non-TOEFL related conversation questions. Overall, the TOEFL iBT classes focused primarily on teacher instruction and individual student practice, which would suggest that the majority of time consisted of teachers explaining the test tasks in the TOEFL iBT textbooks they were using and then setting students the task of individually completing the test tasks. By investigating the type of activities observed in this study, a clearer picture develops of how teaching and learning is realised in the classes observed. Patterns of participant interaction are further explored by identifying teacher activities (e.g. greeting students, explaining, lecturing, guiding, etc.) class activities (e.g. checking answers together [not just teacher giving answers], class discussion, student presentations) and student activities (e.g. individual practice, group/pair work, and computer exercises/activities). In Table 5 below, the activities are reported as a percentage of class time so as to make judgements of what activity types are given priority in the classroom and the shaded areas highlight the TOEFL iBT preparation classes. Overall, the table reinforces the predominant role of the teacher and the role of individual work in TOEFL iBT preparation classes. As was reported in the participant organisation Table 4, teacher instruction, or teacher activities more broadly, occupied more than half of the total class time in all of the TOEFL iBT preparation classes (with an overall average of 57.2% between the five TOEFL iBT classes in “teacher activities”), with the exception of Mike's AL class which spent only 24% of total class time on teacher activities. Given that Mike's AL class was the only class which had scheduled time in the computer lab during the lessons observed, which took 40% of class time, the percentage of teacher activities in his AL class was replaced by student practice time. It should also be noted that the computer lab session was treated as its own activity type and not as an independent student activity in order to distinguish between individual practice with durable textbooks and with software that more accurately reflects the test format and mode. Overall, four out of the five TOEFL iBT preparation courses had a higher percentage of teacher activities than their corresponding general English class. It was only Mike's AL course that had a slightly lower percentage (24%) than his general English course (29%), which again, may have been a result of having spent a large part of the lesson in the computer lab in which students were working independently and teacher activities were kept to a minimum. According to the data collected from the COLT, not only were opportunities limited for the students to speak to one another in English (pair and group work) but also in front of the class, suggesting that there was limited time for students to “interact” Or “use” the language as encouraged in a classroom, underpinned by a CLT approach. While all of the teachers explained or gave the answers to the practice TOEFL iBT tasks, Mike and Ly engaged students by calling on them to answer the questions instead of just giving them the answers. However, the time spent was limited, with Mike spending 4% of class time in his AL lesson and 12% with his NU class and Ly spending 5%. Tuan and David tended to walk the students through an exercise, such as a listening text, and gave the answers as they went. However, another way in which speaking opportunities were provided to students in class was observed in Mike and David's classes in which they attempted to engage students in speaking activities by providing opportunities for students to present their constructed responses to the entire class, devoting 12% and 9% of total class time respectively. While Mike gave every student in the class an opportunity to respond in front of the class and receive feedback, David only chose one student. Though David's feedback was thorough, only one of his five students received a chance to respond and receive feedback on that particular day. As discussed before, group and pair work were also limited in

8

M. Barnes / Journal of English for Academic Purposes 30 (2017) 1e12

Table 4 Participant organisation as a percentage of class time. Percentage of class time (based on total of 100%) Teacher

Tuan Mike

David Ly

Course

TOEFL iBT General English TOEFL iBT AL TOEFL iBT NU General English TOEFL iBT General English TOEFL iBT General English

Whole class T to S/C

S to S/C

81 67 24 60 29 57 52 66 27

0 0 12 11 19 9 0 0 73

Group

Individual

0 0 0 0 52 26 39 0 0

19 33 64 29 0 8 9 34 0

the classrooms observed. In contrast to the limited amount of whole class activities (e.g. student discussion, etc.) in the other nine lessons observed, Ly's general English class dedicated 60% of class time to whole class activities, as oral presentations were an important feature of her speaking course. In the TOEFL iBT preparation courses, students spent an average of 21% of total class time on individual student work (e.g. practice exercises), while the general English classes averaged about 18.5%. If teachers do in fact rely heavily on TOEFL iBT textbooks as their primary teaching source in TOEFL iBT preparation course and these textbooks consist primarily of practice exercises, it might lead one to believe that the time allocated to student practice between the TOEFL iBT preparation class and the general English class would be more significant than 2.5%. More important, however, is that group and pair activities were incorporated in general English classes for 24% of total class time (averaged among the four classes) in contrast to 5.2% in TOEFL iBT preparation classes, with only one class (David's) of the five TOEFL iBT classes incorporating group and/or pair work. Overall, from these findings, it does not appear that the TOEFL iBT preparation classes were “underpinned” by a CLT approach. Overall, most of the TOEFL iBT preparation courses observed spent the majority of class time on teacher instruction or explanations, in addition to incorporating individual practice rather than pair and group work. Like several other washback studies (Alderson & Hamp-Lyons, 1996; Cheng, 2004; 2005; Wall & Alderson, 1993; Wall & Horak, 2008; 2011) this study found that teachers primarily used TOEFL iBT textbooks for their main source of content in TOEFL iBT preparation courses. This suggests that the teachers' reliance on TOEFL iBT textbooks, which consists of test-like tasks, have the potential to limit one's approach to teaching. While there were opportunities for students to interact with one another through pair and group work allowing for more individual feedback and the co-construction of language which would be characteristic of CLT, seldom were these opportunities given. It could be argued, however, that the format of the TOEFL iBT textbooks may encourage teachers to adapt an approach to teaching that focuses primarily on instruction and individual response (e.g. no group or pair activities suggested), albeit this type of input and output may be appropriate for test-taking it may not be appropriate for the classroom. While the findings from the classroom observations suggest that the teachers' methods were limited by the over-reliance on TOEFL iBT textbooks, it is through the teachers' voices that we begin to understand how teachers’ attitudes and beliefs towards teaching, learning and testing shape how they teach.

7.2. From the teachers’ perspectives While several themes emerged from the teacher interviews (e.g., view of self, view of teaching in a Vietnamese context, strengths and weaknesses of TOEFL iBT, etc.), this paper will focus on the teachers’ views on teaching TOEFL iBT preparation classes in Vietnam.

7.3. Teaching the TOEFL iBT: does it change how I teach? From Tuan's perspective, he felt that teachers must change their teaching methodology to teach TOEFL iBT preparation courses. With the introduction of the TOEFL iBT, Tuan claimed that he had to change his teaching methods to accommodate for the changes in the new test format (in contrast with the previous format the Paper-Based Test (PBT): I had to change my teaching style. For the TOEFL iBT, they have the speaking test so I have to spend quite a lot of time on speaking. When I taught PBT, I didn't have to teach speaking. I had to change my methodologies in order to help my students get familiar with the new questions types. For speaking I have to teach pronunciation and how to organise their answers. Similar to the studies by Cheng (2004; 2005) and Chen (2006), Tuan perceived a change in his teaching methodology, but in fact, his examples primarily reflect a change in what he taught (e.g. the speaking subject) and less about how they changed

M. Barnes / Journal of English for Academic Purposes 30 (2017) 1e12

9

Table 5 Activity types as a percentage of class time. Activity type

Tuan iBT

Tuan Gen

Mike iBT AL

Mike iBT NU

Mike Gen

David iBT

David Gen*

Ly iBT

Ly Gen

Teacher: pre-lesson/greeting Teacher: lecturing/explaining/ Guiding Class: check answers to exercise together (both teacher and students talk) Class: discussion Class: presentation (w/teacher feedback) Student: individual work Student: group work Student: Use of TOEFL iBT software in computer lab

0

0

4

22

13

4

4

0

6

81

67

20

41

16

53

26

61

21

0 0 0 19 0 0

0 0 0 33 0 0

4 0 12 20 0 40

12 0 0 25 0 0

0 0 5 9 57 0

0 0 9 8 26 0

0 0 0 22 48 0

5 0 0 34 0 0

0 13 60 0 0 0

his teaching practices. From the data collected from the classroom observations, he still relied on teacher instruction to “teach” speaking strategies rather than give students speaking opportunities as reflected in a CLT approach. Like Tuan, Ly had some experience with the previous TOEFL formats but not as a teacher, as a student. However, Ly understood the changes that had been made and the implications it would have on Vietnamese test takers. When asked if the introduction of a speaking subtest would change the way she taught, she explained that she did not have to adapt the way she taught at all but “just started teaching.” She did not need to adjust the way that she taught because she just taught what was in the book. Thus, a change of what was taught (content) did not require teachers to change their teaching methods or style. Ly argued that while the changes in the testing format required teachers to become familiar with the format, it did not necessarily change the way in which students learnt or teachers taught. She argued “We, ourselves, have to make sure we understand the process in the test. What will appear on the computer screen at first is quite challenging but in fact there is nothing different because for example in listening, students listen and teacher comments in the classroom like in Vietnamese context. You see nothing can be different.” Therefore, according to Ly, while the TOEFL iBT influenced what was taught, it did not influence how teachers taught the new content. Moreover, Ly argues that in a “Vietnamese classroom,” students listen and teachers comment (high percentage of teacher instruction) which is a similar approach adopted in test preparation courses. 7.4. TOEFL iBT preparation courses vs general english classes: does the nature of the course change how I teach? In articulating his approach to teaching TOEFL iBT courses, Mike argued that there are three important ingredients: formula, application, and correct answer. He argued that TOEFL iBT preparation courses were very different from his general English classes which were more “organic.” In his general English classes, he did not want his students to focus on a formula or a correct answer, but rather on creativity. While David also agreed that TOEFL iBT preparation classes lacked an atmosphere that encouraged creativity and described them as being “rarely fun,” he claimed that this was partly due to his students' having little room for imagination and stating opinions as this was something that was not “encouraged” in a Vietnamese educational context. Due to this, he argued he had to find creative ways to take students away from the TOEFL iBT content to broaden their perspectives. This might explain why he was the only teacher to use non-TOEFL iBT materials and to incorporate small group discussions even though the interactive component of speaking is not a skill that is tested on the TOEFL iBT. From David's perspective, he was teaching skills in which students could learn to be more imaginative and forthcoming with their opinions, which could later be applied to TOEFL iBT test tasks. Unlike David, Mike felt that TOEFL iBT courses and general English classes were different teaching contexts and therefore were approached differently pedagogically. For his TOEFL iBT courses, formulas were used for teaching structured responses, which Mike argued “is the majority of what I teach because it is the root of what they do not get in their educational system. They are not taught a formulaic, structured analytical writing or speaking style.” He felt that his students lacked the skills to structure a spoken and written response because they were not taught these skills in their education system. He argued that due to the difference between Vietnamese and Western discourse, students were not accustomed to writing a linear and structured response. For example, he explained that he had to tell his students that he did not want them to write as much as they could for the sake of writing, but to say only what was pertinent to their argument. He felt that this was difficult for his students as being direct and concise were Western attributes that they were not accustomed to and were not encouraged to do in their own education system. Mike also felt that students needed formulas for creating responses on the TOEFL iBT in order to include the necessary elements for a structured response. With the limited time given, students did not have time to be creative but must rely on a formula or structure to help them organise their ideas quickly. Mike argued: To them [students] it's an English class and to me it's a strategy course. I'm not going to teach how to talk in a TOEFL class and I'm not really going to teach how to think either and in an English class I will. In a TOEFL class, I'm going to give you a formula to apply and make sure you can apply it ….I'm not going to teach the creative method, I'm going to teach

10

M. Barnes / Journal of English for Academic Purposes 30 (2017) 1e12

them how to do it in 30 seconds. So I don't really care if their answer is creative. I don't really care if their answer is right. But what I need them to be able to do is to stick it in the right places and have an answer. Thus, Mike felt that the purpose of the TOEFL iBT preparation course was not to develop students’ English language skills but to teach them how to take a test. This is central to understanding his approach to teaching TOEFL iBT preparation courses and particularly his approach to teaching in a Vietnamese classroom where he argues that Vietnamese students are not equipped to provide a direct, linear spoken response. If the goal or aim is to teach students how to take the test, all other skills, whether important or not, will not be taught. In other words, it is not about promoting real-life authentic language opportunities through communicative language teaching (Richards & Rodgers, 2001) but teaching the needed strategies to pass a test. While Mike criticised the TOEFL iBT for not incorporating an interactive component in both the speaking test and within the computer format (e.g. interactive reading), his lack of class discussions and group or pair work revealed in the classroom observations reflect his aim to only teach the skills on the test. Though he personally felt these skills were important, he would include them in his general English classes but not in his TOEFL iBT preparation courses. In a similar vein, David states: “We're not teaching them general English, we're not teaching them survival skills, we are teaching them to pass or to get the highest possible score they can get ….We are coaching them to pass an exam.” He explained that this was one of the main factors in why teaching TOEFL iBT preparation courses was so “boring.” These courses were a means to an end rather than a process of development. Like Mike and David, Ly argued that teaching TOEFL iBT preparation courses was very different from other classes because while general English courses focused on skill development, TOEFL iBT preparation course only focused on the test tasks. Thus, she was not teaching students to develop English language skills but to become acquainted with the test tasks by practising them. Ly argued that TOEFL iBT preparation courses were “test-oriented” and were not intended to have attributes of general English classes or vice versa. Therefore, is it even possible for TOEFL iBT preparation to promote “good” teaching practices if the goal is to pass a test and not necessarily develop language skills? Moreover, why should teachers try “to keep up with the best practices in language teaching by using a communicative and integrated skills approach” (Educational Testing Service, 2006, p. 8), if they only require that their students apply formulas to pass the test?

7.5. Can a test influence teaching methodology? Many researchers claim that the introduction of a new test can influence the content of what is taught but has little to no influence on a teacher's methodology (Chen, 2006; Cheng, 2004; 2005; Wall & Alderson, 1993; Watanabe, 2004). However, this claim appears to be too simplistic because while a test may have a more profound difference on content, there still appears to be evidence of its influence on teaching practices. This influence is evident in how teachers change their teaching methodology to reflect the purposes of TOEFL iBT preparation courses. While teachers like Mike, David and Ly incorporated a high percentage of student and class activities (e.g. group and pair work, class presentations, etc.) in their general English classes, which they claimed allowed for more flexibility in the use of content and materials, they rarely incorporated these activities in their TOEFL iBT courses. However, it is important to note that Tuan's teaching was not influenced either way and may have reflected the typical “Vietnamese” classroom that Ly describes as “students listen and teacher comments in the classroom like in Vietnamese context.” Moreover, due to only being one classroom observation of Ly's general English class when class presentations were scheduled, it is difficult to argue whether Ly's classes were routinely student-centred or embraced a CLT approach, particularly when the class required “instruction.” However, from these findings it could be argued that some of the participating teachers, particularly Mike and David, changed the way in which they taught in order to meet the needs of the highly structure, goal-oriented course. This study suggests that TOEFL iBT preparation courses are influenced by the goal-oriented nature of the course which can potentially encourage teachers to place their teaching styles and beliefs about teaching aside and simply teach the content of the test. Therefore, teachers tend to adhere to a teacher-centred approach in which the focus is on the content and skills which reflect the test and not the way students learn and develop language skills or “best practices” in communicative teaching. Furthermore, it appears that TOEFL iBT preparation courses are not about “good” teaching practices that promote the process of learning and developing language proficiency but rather about providing formulas and correct answers to successfully pass a test, the product of learning the test content (e.g. the needed score). 7.6. Conclusion: positive or negative washback? If we define washback in this study based on the literature's definition that positive washback is defined by the presence of communicative language teaching (CLT), it would appear that the TOEFL iBT has had a negative effect. While this paper did not explore how the students in this study engaged with the texts in class as an indication of Holliday's (1994) strong CLT version, it did explore the role of student to student interaction (weak CLT) and the role of teacher instruction in the classrooms observed. In regards to Holliday's weak CLT or student interaction version, this study would suggest that the TOEFL iBT has had a negative washback effect on the classes observed in this study.

M. Barnes / Journal of English for Academic Purposes 30 (2017) 1e12

11

However, does this fairly represent the nature washback? Deciding whether a test has had a positive or negative washback effect becomes extremely complicated because there are other variables to consider. For example, even if a test is underpinned by a communicative approach this does not mean that this approach is transferred to the classroom. As illustrated in this study and previous studies, the textbooks used and their goal-oriented nature play a large role in how teachers teach (Alderson & Hamp-Lyons, 1996; Cheng, 2004, 2005; Wall & Alderson, 1993; Read & Hayes, 2004; Tsagari, 2011). The “tasks” or practice exercises in these textbooks are linked to the academic skills that have been identified as being assessed in the TOEFL iBT (e.g. reading for main ideas) and act as defined competencies to complete and master. A tension then exists between practising only the skills that will be tested and providing a more holistic approach to preparing students for the test in order to develop their general communication skills. The need to develop language proficiency and communication skills with an EAP focus, rather than just predetermined academic skills, would appear to be an important element in preparing students for an academic language proficiency exam. However, as the TOEFL iBT specific materials attempt to reflect the format and content of the TOEFL iBT test, a skill such as spoken interaction, is a communication skill that is pushed to the side or even more worrying, viewed as unimportant. While the designers of the TOEFL iBT attempted to provide students opportunities to demonstrate more authentic academic communication skills through an integrated approach, test preparation courses still seem to focus on somewhat decontextualised ‘strategies.’ The power of tests is that they may determine what skills will be taught in a test preparation course and deemed as important, thus also determining what skills are insignificant. Therefore, even if the test is designed to reflect a communicative approach, there is no guarantee that this will be reflected in the classroom. This study has revealed that the teaching context, including the teacher, is another important variable in test washback as each classroom is shaped and influenced by those who participate in it. Furthermore, what teachers believe “good” teaching constitutes and their beliefs regarding the purpose and role of test preparation courses in contrast with other English language classes, including EAP classes, may differ. Receiving a passing score on the exam is the ultimate aim for most students taking an English language exam and therefore the process of learning is often not deemed as important as the product (passing the exam). In other words, a “good” test preparation course may not actually reflect “best practice” in language teaching or espouse a communicative language approach but it may ensure that students receive a passing score. There were several limitations to this study in regards to its research design. First, the data collected from the classroom observations could have been more in-depth and reliable if the researcher had observed the classes several times throughout the course and had a larger sample of teacher participants with more teaching experience and teaching qualifications. This is apparent in Ly's general English class in that if class presentations were not scheduled, would her classes have been more teacher-centred? This then would have influenced the data. Second, there was no student voice in this study which limits how one can discuss washback as they are integral part of the teaching and learning process. This would also have provided another perspective on the whether the classroom was teacher or student driven, capturing students' views of the classroom outside of the classroom observations. Given the limitations of this study, this study attempts to draw attention to the need to reconsider how we define positive and negative washback. While it may appear that the TOEFL iBT has had a negative washback on test preparation courses due to the focus on teacher-centred activities (an average of 57.6% of total class time across the 5 TOEFL iBT classes) and the lack of group and pair activities (an average of 5.2% of total class time across the 5 TOEFL iBT classes), more research is needed to identify if there is in fact a link between these “negative” washback effects and student success on the TOEFL iBT. Therefore, the concept of washback and particularly how it is realised in a TOEFL iBT preparation course must be considered when attempting to define positive and negative washback effects. In addition, a better understanding of the relationship between student success (e.g. meeting the required score for entry into an overseas university) and classroom practices and activities are needed in order to determine whether there is a positive and/or negative impact on a students' test preparation success. Finally, building on Green’s (2007) research, the nature of academic English language courses (e.g. test preparation or EAP) needs to be explored in relation to student success and washback effects. While both of these courses aim to build students' academic English language skills, a test preparation course focuses primarily on the product (passing the test) while an EAP course is more concerned with the process (improving academic English skills). Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.jeap.2017.10.003. References Alderson, J. C., & Hamp-Lyons, L. (1996). TOEFL preparation courses: A study of washback. Language Testing, 13(3), 280e297. Andrews, S. (1994). The washback effect of examinations: Its impact upon curriculum innovation in English language teaching. Curriculum Forum, 4(1), 4e58. Author. (2016). Details omitted for double-blind peer review. Bailey, K. (1999). Washback in language testing (TOEFL monograph series 15). Princeton, NJ: Educational Testing Service. Canale, M., & Swain, M. (1980). Theoretical bases of communicative approaches to second language teaching and learning. Applied Linguistics, 1(1), 1e47. Chen, L. (2006). Washback effects on curriculum innovation. Academic Exchange, 206e210. Cheng, L. (2004). The washback effect of a public examination change on teachers' perceptions towards their classroom teaching. In L. Cheng, Y. Watanabe, & A. Curtis (Eds.), Washback in language testing (pp. 147e170). Mahwah, NJ: Lawrence Erlbaum Associates. Cheng, L. (2005). Changing language teaching through language testing: A washback study. Cambridge: Cambridge University Press.

12

M. Barnes / Journal of English for Academic Purposes 30 (2017) 1e12

Edmunds, P., & McKinnon, N. (2006). Developing skills for the TOEFL iBT: Intermediate. Woodland Hills, CA: Compass. Educational Testing Service. (2006). Propell workshop manual for TOEFL iBT. Princeton, NJ: Educational Testing Service. Educational Testing Service. (2007). Test and score data for TOEFL internet-based test. Princeton, NJ: Educational Testing Service. Educational Testing Service. (2017). The TOEFL test. Retrieved on October 6, 2017 from https://www.ets.org/toefl. Ellis, G. (1996). How culturally appropriate is the communicative approach? English Language Teaching Journal, 50, 213e218. Furaidah, Saukah, A., & Widaiti, U. (2015). Washback of English national examination in the Indonesian context. TEFLIN Journal, 26(1), 36e58. Green, A. (2007). IELTS washback in context: Preparation for academic writing in higher education. Studies in language testing, 25. Cambridge: Cambridge University Press. Hawkey, R. (2006). Impact theory and practice: Studies of the IELTS test and progetto lingue 2000. Studies in language testing 24. Cambridge: Cambridge University Press. Hayes, B., & Read, J. (2004). IELTS test preparation in New Zealand: Preparing students for the IELTS academic module. In L. Cheng, Y. Watanabe, & A. Curtis (Eds.), Washback in language testing (pp. 97e111). Mahwah, NJ: Lawrence Erlbaum Associates. Holliday, A. (1994). Appropriate methodology and social context. Cambridge: Cambridge University Press. Hymes, D. (1967). Models of the interaction of language and social setting. Journal of Social Issues, 23(2), 8e38. Hymes, D. (1972). On communicative competence. In J. Pride, & J. Holmes (Eds.), Sociolinguistics: Selected readings (pp. 269e293). Harmondsworth, Middlesex: Penguin. Jamieson, J., Jones, S., Kirsch, I., Mosenthal, P., & Taylor, C. (2000). TOEFL 2000 framework: A working paper. (TOEFL monograph series, 16). Princeton, NJ: ETS. Kramsch, C., & Sullivan, P. (1996). Appropriate pedagogy. English Language Teaching Journal, 50, 199e212. Lewis, M., & McCook, F. (2002). Cultures of teaching: Voices from Vietnam. English Language Teaching Journal, 5(2), 146e153. Macgillivray, M., Yancey, P., & Malarcher, C. (2006). Mastering skills for the TOEFL iBT: Advanced. Woodland Hills, CA: Compass. McNamara, T. (1996). Measuring second language performance. London: Longman. Nunan, D. (1998). Second language teaching and learning. Boston, MA: Heinle & Heinle. Pham, H. (2007). Communicative language teaching: Unity within diversity. English Language Teaching Journal, 61, 193e201. Qi Luxia. (2005). Stakeholders' conflicting aims undermine the washback function of a high-stakes test. Language Testing, 22(2), 142e173. Richards, J., & Rodgers, T. (2001). Approaches and in methods in language teaching (2nd ed.). New York: Cambridge University Press. Spada, N., & Frohlich, M. (1995). COLT: Communicative orientation of language testing observation scheme coding conventions and applications. Sydney: National Centre for English Language Teaching and Researching Macquarie University. Spicer-Escalante, M., & deJonge-Kannan, K. (2014). Cultural mismatch in pedagogy workshops: Training non-native teachers in communicative language teaching. Theory and Practice in Language Studies, 4(12), 2437e2444. Spratt, M. (2005). Washback and the classroom: The implications for teaching and learning of studies of washback from exams. Language Testing Research, 9(1), 5e29. Stetcher, B., Chun, T., & Barron, S. (2004). The effects of assessment-driven reform on the teaching of writing in Washington State. In L. Cheng, Y. Watanabe, & A. Curtis (Eds.), Washback in language testing (pp. 53e71). Mahwah, NJ: Lawrence Erlbaum Associates. Taylor, L. (2005). Washback and impact. English Language Teaching Journal, 59, 154e155. Tsagari, D. (2011). Washback of a high-stakes English exam on teachers' perceptions and practices. In E. Kitis, N. Lavidas, N. Topintzi, & T. Tsangalidis (Eds.), Selected papers from the 19th international symposium on theoretical and applied linguistics (pp. 431e445). Greece: Department of Theoretical and Applied Linguistics, Aristotle University of Thessaloniki, School of English, Aristotle University of Thessaloniki. Wall, D., & Alderson, J. (1993). Examining washback: The Sri Lankan impact study. Language Testing, 10, 41e69. Wall, D., & Horak, T. (2006). The impact of changes in the TOEFL examination on teaching and learning in Central and Eastern Europe. Phase 1: The baseline study (TOEFL monograph 34). Princeton, NJ: Educational Testing Service. Wall, D., & Horak, T. (2008). The impact of changes in the TOE FL examination on teaching and learning in Central and Eastern Europe. Phase 2: Coping with change (TOEFL iBT-05). Princeton, NJ: Educational Testing Service. Wall, D., & Horak, T. (2011). The impact of changes in the TOEFL exam on teaching in a sample of countries in Europe and learning in Central and Eastern Europe. Phase 3, the role of the coursebook. Phase 4, Describing change. (TOEFL iBT-17). Princeton, NJ: Educational Testing Service. Watanabe, Y. (2004). Teacher factors mediating washback. In L. Cheng, Y. Watanabe, & A. Curtis (Eds.), Washback in language testing (pp. 129e146). Mahwah, NJ: Lawrence Erlbaum Associates. Dr Melissa Barnes is a lecturer in the Faculty of Education, Monash University, Australia. While her teaching career began in the United States, it is her teaching experiences in Germany, Vietnam, Australia and Brunei that have collectively shaped her understanding of language, literacy, assessment and teacher education in diverse educational contexts. Melissa has been a primary school teacher, a high school English as a Second Language (ESL) teacher and a university lecturer. She completed her Doctor of Education at the University of Melbourne where she investigated the washback of the TOEFL iBT in English language programs in Vietnam. Melissa has a passion for teaching and assessment, particularly the use of purposeful, formative assessment tasks.