Abstract
This article reports on the views of selected Southeast Asian and Mexican second language teacher education students regarding the characteristics and pedagogical behaviours of good EFL instructors. A total of 116 participants from Mexico, Brunei, Malaysia, Myanmar, the Philippines, Thailand, Indonesia, Cambodia, and Vietnam took part in the current investigation. These countries have all been characterized as pedagogically conservative and slow to adopt educational innovations. Data was gathered using the repertory grid technique and was subjected to content, cluster, and principal component analyses. Results foregrounded the importance placed on teachers’ positive personality traits. Most significantly, findings indicated a preference for modern teaching approaches among the research participants, despite the pedagogic conservatism of their home countries.
Introduction
This article reports on how Second Language Teacher Education (SLTE) students from selected countries in two regions, Southeast Asia and Latin America, understand the characteristics and pedagogical behaviours of good EFL instructors.
It has been reported that a rigid adherence to traditional instructional methods (i.e. approaches which are synthetic, transmissional, teacher-centred, forms-focused, and examinations-driven) is endemic in the countries considered in this study, namely, Mexico, Brunei, Malaysia, Myanmar, the Philippines, Thailand, Indonesia, Cambodia, and Vietnam (Darasawang, 2007; Davies, 2009; Khamkhien, 2010; Liu, 2004; Maskhao, 2002; Nonkukhetkhong et al., 2006; Yamat et al., 2014). These countries are often characterized as pedagogically conservative and resistant to educational innovations. A number of reasons for why pedagogic norms are difficult to alter have been proposed (Arnett and Turnball, 2007; Benson, 2010; Farrell, 2006; Freeman, 1992; Gorsuch, 2000; Gutierrez Almarza, 1999; Hayes, 2008; Johnson, 1999; Kennedy, 1990; Liu and Fisher, 2006; and Zeichner and Gore, 1989). Taken together, the sheer number of impediments to change paint a rather grim picture of what might realistically be achievable within the educational systems of pedagogically conservative countries such as those considered in this study.
At the same time, a number of developments challenge the assumption that pedagogic traditionalism is ineluctable: in many of the countries considered here, governments and educational institutions are currently in the process of developing new curricula and piloting new instructional approaches; the Internet now allows instructors global access to state-of-the-art information; and professionalization within the field is on the rise, with increasingly more students enrolling in SLTE programmes and a rising number of teachers exchanging ideas at educational conferences and events. There is considerable evidence to suggest that advances such as these are spurring reconceptualizations of professional practice (Chepetla et al., 2008; Darasawang, 2007; Khamkhien, 2010; Othman, 2010; Ramírez-Romero and Pamplón, 2012).
In light of ever-heightening tensions between entrenched practices and educational innovation, this seems an appropriate moment to ask what contemporary SLTE students from pedagogically conservative countries think about second language teachers and teaching. First, these students will find themselves at the forefront of any educational reforms carried out in their countries. Second, cultural and institutional forces are not deterministic but rather interactive: teachers act on their environments as much as they adapt to them (Andrews, 2006); instructors, therefore, play an important role in how pedagogy is understood and practiced in their home countries. Thus, how SLTE students view teacher attributes and instructional practices is of great importance.
Background
EFL in Educationally Traditional Countries: Pedagogy in Flux
Generally speaking, the adoption of pedagogical innovations in language teaching has been slower in EFL contexts than in ESL ones. In an international survey of 448 English language teachers, Liu (2004) found that of the respondents who identified themselves as non-native speakers working in EFL environments, 75% favoured the Grammar Translation Method as the primary pedagogical approach to teaching reading and writing, while their native-speaker ESL colleagues favoured communicative and ‘eclectic’ approaches by roughly the same margin.
Descriptions of the pedagogical practices in just a few of the countries considered in this study can be taken as representative. In Thailand, for instance, the educational system continues to focus on teacher-dominated rote learning. Most Thai EFL teachers use a textbook-based approach in which lessons and tests primarily focus on grammar structures, vocabulary, and reading; ‘chalk and talk’, Grammar Translation and Audiolingual methods are the norm (Khamkhien, 2010; Maskhao, 2002; Nonkukhetkhong et al., 2006). In Malaysia, Yamat, Fisher and Rich (2014) observe that classroom practice primarily focuses on scoring well on high-stakes examinations. In Mexico, Davies (2009) reports on an exploratory study conducted during 2001-2002 in over 100 state secondary schools in Mexico. The most common teaching practices found were reading aloud, translating, making lists of vocabulary, and repeating in chorus. Davies concludes that traditional forms of teaching are one of the main reasons for the failure of the educational system in Mexico.
At least three interrelated forces militate against pedagogic innovation in these countries: macro-cultural understandings of teaching (Gorsuch, 2000; TALIS, 2009); micro-cultural influences at the institutional level (Benson, 2010, Calderhead, 1988, 1993, 1996; Farrell, 2006; Liu and Fisher, 2006); and teachers’ own experiences as language students (Freeman, 1992; Kennedy, 1990; Johnson, 1994; Lortie, 1975).
The forces outlined above would seem to discourage educational change in pedagogically traditional countries. However, an impartial observer of these countries in Southeast Asia and Latin America would be hard-pressed to argue that the educational situation in these regions remains completely static. In response to the low success rate of many of its ESL students (see, for example, Anderson, 2012), governments in Southeast Asia and Latin America have attempted to update teaching practices, adopting new curricula and pedagogical approaches. Topics at regional TESOL conferences reflect this change in focus, with more emphasis placed on such subjects as communicative teaching, technology, teacher development, strategy training, and affective factors (Othman, 2010). In Thailand, there has been a recent emphasis on Task-Based Learning (TBL), Content and Language Integrated Learning (CLIL) and Communicative Language Teaching (CLT) (Khamkhien, 2010), as well as on strategy training (Darasawang, 2007). In Malaysia, too, there have been efforts to introduce CLIL approaches to the teaching of Mathematics and Science (Yamat et al., 2014). In Vietnam, the latest 10-year national plan for teaching and learning foreign languages emphasizes a focus on communicative competence, adopts the Common European Framework of Reference for Languages as the standard for assessing the quality of English learning, and encourages Vietnamese educational institutions to actively develop and carry out bilingual programmes (Hoang, n.d.). And in Mexico, Ramírez-Romero and Pamplón (2012) remark on the growth of teacher preparation programmes at the undergraduate and graduate level, and on the proliferation of professional associations and forums that introduce and advocate state-of-the-art approaches to language pedagogy.
Teacher Characteristics and Practices
The current study is situated within a large and growing literature on the characteristics and practices of teachers, both inside and outside the field of ESL. Findings from a large number of studies suggest that, first and foremost, the possession of positive personality traits is seen as the sine qua non of all good instructors. Teachers and students agree that instructors should be enthusiastic, energetic, patient, motivating, humorous, warm, kind, and flexible (Broder and Dorfman, 1994; Chong et al., 2005; Feldman, 1986; Holt-Reynolds, 1992; Minor et al., 2002; Morine-Dershirner et al., 1992; Mowrer-Reynolds, 2008; Murray, 1985; Proctor, 1989; Walls et al., 2002; Weinstein, 1989). In terms of their pedagogical roles, it is felt that teachers should be organized, knowledgeable, and act professionally; they should be skilled in the classroom, particularly in terms of their ability to manage their classes and explain concepts clearly (Emmelman and DeCesare, 2007; Forston and Brown, 1998; Helterbran, 2008; Minor et al., 2002; Proctor, 1989; Strage, 2008; Walls et al., 2002).
Compared to studies in other academic disciplines, there have been relatively few studies profiling teacher characteristics in the field of EFL. Overall, the vast majority of studies on L2 teacher characteristics have been focused on student perceptions of their instructors. These studies suggest that while language teachers are expected to embody the same positive characteristics as instructors in other fields, they are also judged in terms of a number of domain-specific traits and pedagogical skills (Borg, 2006a). Several investigations have underlined the fact that both students and language teachers tend to put a premium on L2 proficiency and language knowledge (Barnes and Lock, 2010, 2013; Brosh, 1996; Chacón, 2005; Chen, 2012; Ghasemi and Hashemi, 2011; Johnson, 2004; Khodadady et al., 2012; Park and Lee, 2006; Shishavan and Sadeghi, 2009; Thompson, 2008; Tsai, 1999; Wichadee, 2010). It is generally believed that language teachers should: possess cultural awareness that complements their language knowledge, be able to use a variety of techniques and approaches specific to ESL, provide practice opportunities that cover the four language skills, and connect language to real life (Bell, 2005; Brosh, 1996; Brown, 2009; Çelik et al., 2013; J. Chen, 2012; X.R. Chen, 2008; Cortazzi and Jin, 1996; Çubukcu, 2010; Ghasemi and Hashemi, 2011; Khodadady et al., 2012; Korkmaz and Yavuz, 2011; Park and Lee, 2006; Shishavan and Sadeghi, 2009; Thompson, 2008; Tsai, 1999).
The question of what attributes successful language teachers should possess has generally been approached from the points of view of English language students and ESL teachers. Only a small number of studies have looked at how teachers in training define the qualities and practices of a successful ESL teacher (see Korkmaz and Yavuz, 2011; Mullock, 2003; Thompson, 2008). This is surprising, since it seems self-evident that understanding teacher characteristics is important to understanding teaching practice. As Borg points out:
language teacher education presupposes an understanding of what specifically it means to be a language teacher, and therefore insight into the distinctive characteristics of language teachers is central to the work of language teacher educators (2006a: 3).
Methodology
The Repertory Grid Technique
The repertory grid technique (RGT) is a data elicitation instrument used to examine the structure and content of the implicit theories through which people construe reality. The RGT is the most famous of the methodologies associated with George Kelly’s (1955) theory of personal construct psychology (PCP). Although PCP was initially developed by Kelly for use within the field of clinical psychotherapy, scholars and practitioners in various other disciplines have adopted its premises and employed its methods. Currently, ‘rep grids’ are used as a research instrument in a wide range of fields and have appeared in more than 2,000 journal articles, books, book chapters, and doctoral dissertations; over the course of the last decade, more than 100 works utilizing the rep grid technique have been published each year (Saúl et al., 2012). In the field of general education, there are numerous studies of teacher development and cognition based on rep grid data. However, only a very few investigations in the field of applied linguistics have made use of the repertory grid technique (Bodycott, 1997; Murray, 2003, as cited in Borg, 2006b; Richter, 2012; Richter and Derry, 2013; Sendan and Roberts, 1998).
The repertory grid technique actually refers to a number of different data elicitation methods. The ‘standard’ version of a repertory grid is usually understood to consist of a set of elicited elements, a set of elicited constructs, and a set of elicited numerical ratings that join them together (see Figure 1).

A Standard Repertory Grid.
Regardless of the form a grid takes, all grids are essentially sorting tasks which enable a psychotherapeutic client or research participant to explain the way in which he or she orders the world (Fransella et al., 2004). The general procedure for completing a standard repertory grid is as follows:
A number of elements are elicited. Elements are people, objects, events, or even concepts that are representative of the topic of investigation.
A construct is elicited. Constructs are the ways in which elements are seen by a participant to be similar or different and are expressed in bipolar forms (i.e. happy/sad, bright/dark, etc.). Elicitation of constructs generally consists of arranging elements in groups of two (dyadic elicitation) or three (triadic elicitation) and having participants comment on their similarities or differences. The first pole elicited is called the ‘emergent’ pole. Once an emergent pole is generated, the participant is asked to think of a ‘contrast’ pole.
All elements are rated on the construct, generally using a 5- or 7-point scale, with 1 point reflecting that an element lies perceptually close to the emergent pole, and 7 points reflecting its closeness to the contrast pole. Steps (2) and (3) are repeated until the participant can no longer offer new constructs.
The grid is analyzed, either qualitatively, through content analysis, or statistically. Since most of the data in a grid is numerical, it is amenable to various types of multivariate statistical procedures, such as two-way cluster analysis or principal component analysis (see below).
Participants
A total of 116 participants took part in this study, 59 from Latin America and 57 from Southeast Asia. In all, nine countries were selected on the basis of convenience sampling, according to where the authors of this article were working: Mexico (59 participants), Brunei (5), Malaysia (2), Myanmar (11), the Philippines (2), Thailand (12), Indonesia (16), Cambodia (4), and Vietnam (5). At the time data was collected, all the participants were enrolled in second language teacher education courses. The Mexican participants were students in a four-year university SLTE programme. The participants from Southeast Asia studied at a teacher training institute in Singapore. All the participants, both in Mexico and in Singapore, are practicing teachers who returned to school to update their pedagogical knowledge and skills.
Procedure
Participants were asked to select both grid elements and constructs, allowing them maximum freedom to express themselves. First, elements were elicited according to the following prompts: ‘a great language teacher’ (two elements); ‘an average language teacher’ (two elements); ‘a poor language teacher’ (two elements); ‘you, the teacher you are now’; and ‘you, after some professional growth’.
Constructs were obtained through dyadic elicitation; the elements were then rated along each of their constructs using a 7-point rating scale. The contents of the individual grids were subsequently subjected to content analysis. Using a ‘bootstrapping’ approach (Jankowicz, 2004), four major categories were identified, as well as 24 sub-categories (Table 1).
Emergent Constructs.
The individual emergent constructs of all the participants were then grouped according to standard content analysis procedures (Jankowicz, 2004; Maykut and Morehouse, 1994; Wright, 2004). Once the most important collective constructs were determined, the ratings from the original individual constructs were averaged. In this way, three ‘collective grids’ were constructed: one Mexican grid, one Southeast Asian grid, and one ‘super grid’ representing the common construing of all the participants in the study (see Figure 2). The averaged ratings were weighted proportionally to reflect the number of constructs generated within each major category and then used as the input for the Rep 5 (v. 1.05) suite of programmes. Rep 5 generated display grids, principal component analyses, and FOCUS analyses.

Display Grid: Mexican and Southeast Asian SLTE Students’ Collective Grid.
Results
Content Analysis
In all, the 116 repertory grids generated 1206 constructs and 9648 ratings (Table 1). Counting in the eight elements, the repertory grid interviews produced a total of 10,862 pieces of data. Overall, the largest number of constructs had to do with pedagogical concerns (484 constructs, or 40.13% of the total). The single largest sub-category of constructs concerned the positive personality traits of teachers; participants produced 215 such constructs, which accounted for 17.83% of the total. Typical bipolar constructs in this latter category included sense of humour/boring; energetic/dull; patient/too demanding; friendly/strict; and mature/immature. The second largest sub-category encompassed those constructs having to do with modern pedagogical approaches and perspectives. 1 Typical emergent constructs in this category included uses formative assessment, trains student for self-learning, prepares authentic activities, uses technology in the classroom, creates classroom as a communicative experience, gives students assessment criteria, understanding of different learning styles, student centred learning, class is based on student needs, use of peer work to help students, lets students solve problems in their own ways, teaches inductively, noticing work, task-based teaching, and embraces creative thinking. The Mexican students produced a slightly higher number of constructs about modern pedagogy than they did about personality traits (17.11% and 15.57%, respectively), whereas the Southeast Asian students generated more personality-related constructs than they did constructs related to modern approaches (21.1% and 15.62%, respectively). The third largest category for both groups was ‘rapport with students’; these constructs accounted for 9.45% of the total number generated. The fourth largest category was ‘techniques’, which included any constructs involving the quality, variety, or creativity of activities and materials. Constructs in this category accounted for almost 7% of the total. Finally, the fifth largest category was ‘lesson delivery’, which included such emergent constructs as explain in detail, able to communicate clearly, can explain in different ways, speaks clearly, good presenter, and good storyteller.
Statistical Analysis
One method of uncovering relationships among constructs and elements is the use of a hierarchal cluster analysis of correlations. Correlations are represented by a dendrogram. The more that constructs or elements are alike, the closer they approximate a score of 100, which would signify a perfect correlation. Thus, in Figure 3, the emergent constructs ‘modern pedagogy’ and ‘techniques’ are a 97.3% match.

Cluster Analysis of Mexican and Southeast Asian SLTE Students’ Collective Grid.
This suggests that for the study participants, these concepts are largely isomorphic: an instructor who uses modern approaches will also tend to employ worthwhile activities and materials, and vice versa. Similarly, ‘class management’ and ‘delivery’ match at 96.9%; ‘personality’ and ‘rapport’ match at 96.7%; and ‘rapport’ and ‘delivery’ match at 96.5%. Teacher experience is not highly correlated with teacher effectiveness, an 83.5% match. The construct that least correlates with any other is the construct concerned with teachers’ moral and ethical behaviour.
Correlations between elements can also be represented in a dendrogram. Looking at Figure 2 again, one can see that the SLTE students in the study generally identify themselves as average teachers (‘you as you are now’ and ‘an average language teacher 1’ match at 89.3%). The participants view professional growth as the pathway to becoming good teachers (‘you with professional growth’ and ‘a good teacher 1’ match at 93.3%).
Study participants tend to judge poor teachers according to the same criteria (‘a bad language teacher 1’ and ‘a bad language teacher 2’ match at 93.5%). An eyeball analysis of the cluster analysis grid reveals that, according to the participants, bad teachers are particularly bad in terms of social rapport, lesson delivery, class management, expectations for students, pedagogic effectiveness, personal motivation, and their lack of interest in professional development. Poor teachers are also unlikely to employ effective or modern pedagogical techniques.
While bad language teachers all appear to be incompetent in similar ways, in the minds of study participants, there appears to be many varieties of ‘average language teachers’.‘An average language teacher 1’ and ‘An average language teacher 2’ only match at 72.6%.
Principal components analysis (PCA) is another common type of statistical analysis carried out on rep grid data. Principal components analyses calculate the similarities and differences between the elements and constructs in a given grid. Elements and constructs can be plotted in a geometrically graphic form as variables in a system of coordinates that represent a person’s ‘psychological space’ (Kelly, 1955) or ‘cognitive map’ (Wright, 2004). Figure 4 shows a principal component analysis graphing the study participants’ elements and constructs. The vertical and the horizontal lines represent maximally distinct patterns within the ratings and are called the principal components. Each component is a statistical invention the purpose of which is to represent the most significant patterns in the grid (Jankowicz, 2004).

Principal Component Analysis (PCA) of Mexican and Southeast Asian SLTE Students’ Collective Grid.
Constructs are placed in relation to the two main components, their poles linked by a line. The closer constructs lie to one of the components, the more similarities there are between them. Here we can see another representation of the same information imparted by the cluster analysis, above: the emergent constructs ‘modern pedagogy’ and ‘techniques’ lie almost directly on top of each other. Similarly, the elements representing bad language teachers are situated very close to one another, while those representing average teachers lie on opposite sides of the biplot. As in the cluster analysis, teacher morality is an outlier.
The contrast poles having to do with grammar are clustered together in the top left portion of the quadrant and reside at some distance from the other constructs. The contrast pole associated with teacher experience also lies at a distance from the other constructs represented in the biplot.
The main value of PCA is the ability to visualize relationships between constructs and elements. In Figure 4, the element ‘you as you are now’ is at a significant distance from any of the constructs, but is closest to those having to do with ‘general knowledge’, ‘formal education’, ‘punctuality’, and ‘personality’. The element ‘you with professional growth’ lies closest in space to the constructs concerning teaching technique, modern methods, and teacher effectiveness.
Finally, further statistical analysis makes it possible to compare the responses of the Mexican participants and the Southeast Asian participants. In Figure 5, one can see that the two groups largely view the constructs they generated in the same way.

Comparison Grid of Mexican and Southeast Asian SLTE Students’ Collective Grids.
There was particular agreement among the two groups in terms of their views of the importance of general content knowledge and personality (both matching at above 90%). There was a high degree of consensus (above 85%) regarding delivery, dedication, fairness, organization, preparation, student expectations, techniques, the use of modern approaches, student-teacher rapport, education, and professional development. The only constructs that do not closely match (at a threshold of 75%) are those having to do with morality and ethics, language ability, the ability to motivate, teacher experience, and grammar. It should be noted, however, that the constructs concerning these issues were very small in number compared to other construct categories. Constructs having to do with language ability comprised 2.65% of the total constructs generated. Each of the other constructs accounted for approximately 1% of the total.
Discussion
The high degree of overlap among investigations of how second language teachers and SLTE students view instructor characteristics and behaviours (Bell, 2005; Brosh, 1996; Brown, 2009; Chacón, 2005; Johnson, 2004; Korkmaz and Yavuz, 2011; Mullock, 2003; Park and Lee, 2006; Shishavan and Sadeghi, 2009; Thompson, 2008; and Wichadee, 2010) suggests that a general global consensus about EFL best practices is beginning to solidify. Findings from the current study harmonize with previous research, as well as with similar research conducted in other disciplines. This indicates that the participants in the current study are solidly situated within the global mainstream vis-a-vis their pedagogical views. This conclusion is further corroborated by the fact that, despite the geographical and cultural distance separating them, the participants from the Southeast Asia countries and Mexico demonstrated a remarkable degree of agreement about the desirable characteristics and pedagogical behaviours of language teachers.
Study participants generated more constructs having to do with personality than they did any others. If constructs concerning personality and student rapport are conflated, as they seem to be in the minds of the participants, then they would account for more than a quarter (27.28%) of the total constructs generated. The importance of personality and rapport is again seen in the participants’ views of professional development. Professional growth is more highly correlated with establishing rapport with students and with teacher personality than it is with teacher education. These findings are consonant with the many other studies which demonstrate strong associations between positive personality traits and teaching excellence. Students in educational programmes tend to believe that motivating one’s students and being warm and personable are primary characteristics of good teachers (Chong et al., 2005; Helterbran, 2008; Holt-Reynolds, 1992; Minor et al., 2002; Morine-Dershirner et al., 1992; Mowrey-Reynolds, 2008; Proctor, 1989; Walls et al., 2002; and Weinstein, 1989).
As was noted above, teacher experience is not particularly correlated with teacher effectiveness. This perhaps suggests that the study participants view those teachers with more experience as being outdated and indicates an embrace of the new and a rejection of the old. This interpretation makes sense in light of the participants’ overwhelming preference for modern approaches.
Indeed, the most surprising finding in this study was the degree to which all the participants, regardless of nationality, rejected traditional pedagogy in favour of more current approaches. None of the participants in the Southeast Asia group generated constructs having to do with grammar knowledge or instruction. Only 1.68% of the constructs created by the Mexican cohort involved grammar (i.e. six out of 713). In contrast, 16.5% of all the constructs generated indicated a preference for non-traditional approaches. No one in the study identified teacher traits or behaviours associated with traditional pedagogy (i.e. transmissional approaches, projection of teacher authority, rote learning, etc.) as desirable.
Conclusion
Given the limitations of this study, it is impossible to know to what degree the research participants operationalize their beliefs about good pedagogy in their own EFL classrooms. Hoang, writing about Vietnam, notes that ‘during training courses … teachers show great interest in new methodologies, but after they return from those courses, they continue teaching in the old methods’ (Hoang, n.d.: 16). The mismatch between beliefs and pedagogical action has oft been observed (Basturkmen, 2012; Lee, 2009; Ng and Farrell, 2003; Phipps and Borg, 2009). This is perhaps not surprising, given the many situational, psychological and cultural factors that tend to prevent innovation and the adoption of new pedagogical approaches. Despite these contravening factors, and while any generalization of findings beyond the participants in the current study must be heavily caveated, the results of this investigation do provide some evidence of a trend away from pedagogic traditionalism amongst SLTE students hailing from educationally conservative countries.
Footnotes
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Notes
References
Available at: http://r-cube.ritsumei.ac.jp/bitstream/10367/4129/1/LCS_22_1pp7-18_HOANG.pdf