1. Introduction
The question of how we understand and measure quality in educational provision has always remained an important area of interest, both for researchers and for wider social and political discussion around education. Research investigating the impact or utility of specific teaching methods, techniques or practices has tended to dominate this literature, both in language teaching (e.g., Murphy et al., Reference Murphy, Arndt, Briggs Baffoe-Djan, Chalmers, Macaro, Rose, Vanderplank and Woore2020) and mainstream education (e.g., Hattie, Reference Hattie2009). And while the use of experimental, cross-sectional and correlational study designs has often offered useful evidence to inform the quality debate, because such research necessarily tends to isolate and compare individual variables or specific interventions, it rarely provides us with holistic understandings of educational quality. As a result, we may be seen to lack comprehensive descriptions of what good teachers do, how they do it and why – descriptions that are capable both of ‘joining the dots’ between the different elements investigated and of providing useful models or exemplars that teachers, teacher educators and wider educational systems may be able to learn from.
Research involving expert teachers is capable of doing just this, and doing it in a way that has high ecological validity and context-specificity – two important prerequisites for any research in education to be credible and relevant to real classrooms and real teachers. In this regard, it is revealing that a recent scoping review (Murphy et al., Reference Murphy, Arndt, Briggs Baffoe-Djan, Chalmers, Macaro, Rose, Vanderplank and Woore2020), conducted to assess (intra alia) the effectiveness of different approaches and strategies in foreign language teaching, notes in its executive summary that ‘more important than the specific method used is the way in which it is delivered and by whom’ (p. 4); that is, ‘teacher expertise’ seems to matter more than method (p. 108). Yet, despite the promise that such research may hold, it is notable that only a handful of studies of expert language teachers have ever been conducted (discussed in section 2).
With these concerns in mind, this article aims to offer a concrete basis upon which to build an agenda for teacher expertise research in language teaching over the next 10–20 years, offering solid theoretical and methodological foundations for the field as well as a range of broad research tasks that I hope will interest and inspire both novice researchers and experienced scholars alike to conduct studies involving expert teachers. The article begins with a brief historical background to teacher expertise research. It then presents a theoretical justification for this important domain of enquiry, identifying the high ecological validity of expert teacher studies, the practical applicability of the findings and the ‘enhancement’ perspective involved. The article then discusses important theoretical and methodological considerations for researchers, particularly concerning how we define expertise and identify expert teachers for studies. It introduces a framework for future research on language teacher expertise as a heuristic tool for identifying potential tasks that span both the qualitative–quantitative methodological continuum (rather than dichotomy) and the alleged paradigm divisions between positivist, critical realist and interpretivist epistemologies. Finally, the article proposes six broad research tasks, located within the research framework, from small-scale studies that are potentially feasible for Ph.D. students to larger-scale research designs that may require experienced teams and access to large datasets.
2. Historical background
Since its inception in the 1960s, expertise research has expanded slowly but steadily from studies of specific skills (e.g., chess playing) to domains of increasing complexity (see Ericsson et al., Reference Ericsson, Hoffman, Kozbelt and Williams2018; Glaser & Chi, Reference Glaser, Chi, Chi, Glaser and Farr1988). Building on work by the Dreyfus brothers (Reference Dreyfus and Dreyfus1986), Chi and Glaser (Chi et al., Reference Chi, Glaser and Farr1988) and Bereiter and Scardamalia (Reference Bereiter and Scardamalia1993), it moved from studying primarily cognitive ability and behaviour to focusing on more complete descriptions of expertise, linking aspects of the practice, cognition and personal characteristics of experts to try to understand not only what expertise is, but also who experts are as people (e.g., Bullough & Baughman, Reference Bullough and Baughman1995) and how professional expertise develops (Schön, Reference Schön1983, Reference Schön1987).
In most domains of professional practice (e.g., healthcare, legal practice, business), ‘expertise’ is generally accepted as an appropriate indicator of professional competence. Yet, it has a more uneasy relationship with the field of education – including language teaching (Hirvela & Belcher, Reference Hirvela and Belcher2022) – perhaps because many teachers are somewhat wary of the notion of expert teachers, even those who are studied as such (see Goodwyn, Reference Goodwyn2011; Sorensen, Reference Sorensen2014). This may be due to its potential associations with elitism or exclusivity, something that is likely to be antithetical to the values and mission of many teachers. Further, because of the complexity of theorising and studying expertise in education, as Berliner observes (Reference Berliner2004), the link between expert teachers’ practices or nature and their impact or ‘achievements’ has been much more difficult to establish than in other domains.
Yet, despite these challenges, teacher expertise research has expanded steadily from its beginnings in the work of United States researchers Gaea Leinhardt (e.g., Reference Leinhardt1983a, Reference Leinhardt1983b) and David Berliner (e.g., Reference Berliner1986) in the 1980s. Today, there are hundreds of studies available across numerous national contexts, albeit with a strong Western/Northern and anglophone bias; see Anderson and Taner (Reference Anderson and Taner2023) and Sternberg and Horvath (Reference Sternberg and Horvath1995) for overviews. However, this bias is being countered by a rapid increase in expertise studies in China, where two terms (专家型教师, lit. ‘expert type teacher’ and 名师, lit. ‘famous teacher’) are used and where there seems to be less social stigma towards the application of such terms in education. Recent examples in language teaching include studies by Lee and Yuan (Reference Lee and Yuan2021), Li and Zou (Reference Li and Zou2017) and Yuan and Zhang (Reference Yuan and Zhang2020).
Two recent systematic reviews have investigated commonalities among expert teachers in K12 education (Anderson & Taner, Reference Anderson and Taner2023) and expertise frameworks in university teaching (van Dijk et al., Reference van Dijk, van Tartwijk, van der Schaaf and Kluijtmans2020), revealing findings that are likely to be of use to a wide range of stakeholders at different levels of education. Unfortunately, there is still comparatively little research available on language teacher expertise. For example, of the 106 studies on K12 teacher expertise reviewed by Anderson and Taner (Reference Anderson and Taner2023) only eight involved foreign or second language teachers, notably fewer than mathematics teachers (n = 25).
Nonetheless, renewed interest in the issue of language teacher quality is detectable in recent publications. In addition to the studies from China mentioned above, this includes practically-oriented contributions to an edited volume entitled Lessons from good language teachers (Griffiths & Tajeddin, Reference Griffiths and Tajeddin2020), interest in the development of language teacher expertise (see Johnson et al., Reference Johnson, Verity and Childs2020 and contributions to Maley, Reference Maley2019), and a clear revival of interest in teacher expertise in one sub-field of language teaching – second language writing (e.g., Christiansen et al., Reference Christiansen, Du, Fang and Hirvela2018; Eick et al., Reference Eick, Fields and Matsuda2017; Hirvela & Belcher, Reference Hirvela and Belcher2022; Lee & Yuan, Reference Lee and Yuan2021) – which builds on earlier research on highly experienced teachers of writing by Alister Cumming (e.g., Reference Cumming1990, Reference Cumming, Belcher and Braine1995).Footnote 1
It can be surmised from this brief history that while there is clear and recent interest in language teacher expertise, the research itself is still fairly limited. But why might such research be useful?
3. A case for language teacher expertise research
Contrary to some beliefs, teacher expertise research does not involve research into ‘outstanding’ or ‘exceptional’ teachers of the type that are sometimes popularised in national and international teacher awards (e.g., the Global Teacher PrizeFootnote 2), although the individuals celebrated through such awards are also likely to be expert practitioners. When appropriately defined, the construct of teacher expertise is capable of describing the kind of fully qualified, highly experienced, caring and competent professionals that are likely to be widespread across all education systems, regardless of systemic challenges. Such expertise is potentially accessible to all teachers, providing the necessary requisites for them to achieve it are present (e.g., time, motivation, reflexivity and support) (Anderson, Reference Anderson2023c; Hirvela & Belcher, Reference Hirvela and Belcher2022).
Because it involves real teachers working in real classrooms in a given context, research involving expert teachers can claim the highest level of ecological validity; there is no need to isolate a specific variable, nor to measure the impact of a specific intervention that may or may not prove implementable and effective when rolled out across multiple classrooms. Instead, teacher expertise research is motivated by a need to understand how those individuals who have achieved professional competence teach, why they do so, what underpins this ability and what impact their work has on their learners and wider community. It provides opportunities to unlock the relationships between previously disparate areas of teacher development, such as experiential learning, reflection, teacher caring and professional competence (see Anderson & Taner Reference Anderson and Taner2023). This broad focus on diverse aspects of teacher experience, personality and identity means that it is capable of offering empirical support, or critical feedback, for extant frameworks of teacher competence and professionalism (e.g., BALEAP, 2008; British Council, 2015).
Perhaps most obvious among its contributions, teacher expertise research is able to identify important contingent commonalities among expert teachers (e.g., Anderson & Taner, Reference Anderson and Taner2023; Sternberg & Horvath, Reference Sternberg and Horvath1995) that are of obvious interest and utility in varied domains, such as initial teacher education, in-service teacher development, teacher evaluation and educational policy making. Yet, it is also able to spotlight difference among expert teachers. It can shed light onto areas where there seem to be greater differences, enabling us to understand what many experienced educators have long known – that there are many ways to be ‘good’ at teaching, even in comparable contexts (see Anderson, Reference Anderson2023c, Ch. 8), reaffirming the important point that there is never a ‘one size fits all’ model of best practice for any subject (Seloni, Reference Seloni2022).
Finally, because it adopts an enhancement rather than a deficit approach towards teachers and their practices, teacher expertise research is capable of celebrating and rewarding worthy practitioners within a given context or system (i.e., through recognition, rather than anonymised study; see Anderson, Reference Anderson2023a). It documents and spotlights examples of successful practice and appropriate role models that novice teachers can identify with and look up to. Further, because expertise research starts in the classroom and shows a fundamental interest in practitioners and their behaviour, it is able to break down – or avoid altogether – the often perceived divide between academic and practitioner interests in applied linguistics (e.g., Rose & McKinley, Reference Rose and McKinley2022; Sato & Loewen, Reference Sato and Loewen2022). This enhancement approach presents a useful foil to what can, at times, seem to be a constant barrage of criticism of teachers and education in the popular press and social media in some countries (e.g., India, UK, USA). In this sense, it may contribute towards improved teacher well-being – a focus of increasing concern in language teaching (e.g., Talbot & Mercer, Reference Talbot, Mercer, Mohebbi and Coombe2021).
4. From theoretical to methodological considerations
In principle, teacher expertise research is simple in design: find one or more expert teachers and study one or more aspects of them – typically their practices, cognition, development and/or personal characteristics. They can be studied in isolation or in comparison, either with each other or with ‘non-expert’ teachers (either novices or experienced non-experts; see Tsui, Reference Tsui2009). Studies can be qualitative (most common), mixed methods, or quantitative in design, with sample sizes varying from one-participant ethnographies (e.g., Traianou, Reference Traianou2006) to comparative studies with several participants (e.g., Anderson, Reference Anderson2021) and even larger correlational studies attempting to investigate, for example, the impact of expert teachers on learner academic outcomes (e.g., Hattie, Reference Hattie2003). Providing one has found expert teachers appropriately and successfully (discussed in sections 4.1 and 4.2), the findings will be potentially beneficial in a range of areas, including:
• teacher education (e.g., informing pre-service curriculum design);
• teacher professional monitoring and evaluation (e.g., informing teacher observation instruments or professional development frameworks);
• in-service teacher development (e.g., offering role models for novice teachers).
However, this apparent simplicity belies a more complex theoretical issue that penetrates right to the heart of the question of quality in education: the interrelated challenges of how we understand and define ‘expert’ or ‘expertise’, and the need for our methods for identifying and recruiting study participants to be consistent with these definitions. Figure 1 displays this relationship diagrammatically, the components of which are explored in sections 4.1 and 4.2.
From theory to selection – defining and finding expert teachers
4.1. Defining expertise
A major theoretical challenge when defining expertise relates to the often-referenced tension between the temptation to value that which we can easily measure and the need to measure that which we value (e.g., Alexander et al., Reference Alexander, James and Glaser1987). Early research in this field tended to do the former, equating teacher expertise with teacher ‘effectiveness’ (i.e., impact on learner academic outcomes; e.g., Leinhardt et al., Reference Leinhardt, Weidman and Hammond1987). This narrow focus can be seen to be problematic – arguably naïve – particularly today, when we recognise a much wider range of beneficial outcomes of teacher expertise beyond academic achievement (UNESCO, 2017). This includes impacts and influences on learners that are not typically or easily measured in summative assessment (e.g., social and emotional learning, personal well-being, and critical and digital literacies), as well as the range of benefits expert teachers bring to their colleagues, institution and wider communities. Thus, researchers have recently proposed more multifaceted working definitions of teacher expertise, capable of capturing more of what we value in expert teachers. In the field of second language writing, Hirvela (Reference Hirvela, Seloni and Henderson Lee2020) defines teacher expertise as ‘the instructional beliefs, knowledge and skills that may be considered as essential at a certain level of proficiency in order for teachers to guide students towards the acquisition of beneficial L2 [second language] writing ability’ (p. 17). In my own research (Anderson, Reference Anderson2021), which sought to develop a socioculturally responsive definition of teacher expertise, I defined it as ‘an enacted amalgam of learnt, context-specific competencies (i.e., embodied knowledge, skills and awareness) that is valued within an educational community as a source of appropriate practice for others to learn from’ (p. 44). Numerous other definitions and discussions exist in language teaching alone (see, e.g., Johnson, Reference Johnson2005; Lee & Yuan, Reference Lee and Yuan2021; Tsui, Reference Tsui2003), some of which also recognise a process orientation to teacher expertise – what Burns (Reference Burns2022) calls ‘a dual focus on emergent and extant expertise’ (p. 1).
Working definitions of expertise may also benefit from preliminary research within a given context of interest (as suggested in Figure 1; also see Research task 5). They may need to be sensitive to national norms, educational stage, type of instruction or data collected. Consider, for example, the difference between a small-scale study of expert teacher classroom practice conducted in Malawian primary schools with a statistical analysis of expert teacher impact in Chinese higher secondary education, or one investigating teacher–student relationships in pupil referral units in inner-city London.
Perhaps most important for researchers to bear in mind when developing a working definition of expertise is the need for reflexive awareness of the personal biases that permeate our decision-making when theorising and operationalising any construct of quality in education. This is particularly true when it comes to the complex issue of selecting participants for expertise studies.
4.2. Identifying participants
A generally agreed a priori criterion for participants for expertise studies is the need for sufficient time for expertise to have developed. This is generally agreed to be at least five years (sometimes seven) of full-time teaching experience and is typically seen as necessary, but not sufficient on its own, to identify participants as experts (e.g., Berliner, Reference Berliner2004; Palmer et al., Reference Palmer, Burdenski and Gonzales2005; Tsui, Reference Tsui and Johnson2005). Thus, a wide range of additional criteria have been used for identifying participants for teacher expertise research (for overviews, see Anderson, Reference Anderson2023c, pp. 80–84; Palmer et al., Reference Palmer, Burdenski and Gonzales2005). These criteria include:
• nomination by relevant stakeholders (e.g., school inspectors or headteachers);
• higher qualifications (e.g., National Board Certification in the USA);
• student performance-based criteria (e.g., exam scores);
• receipt of teaching awards (discussed critically by Tsui, Reference Tsui and Johnson2005);
• institutional leadership roles and experience as teacher educators or mentors.
While this diversity of criteria may be seen to reflect a range of understandings of expertise itself, it is not always the case that researchers link definition and selection criteria together explicitly, as recommended in Figure 1. Perhaps the single most important recommendation, stressed by both Palmer et al. (Reference Palmer, Burdenski and Gonzales2005) and Anderson (Reference Anderson2023c), is the need for the use of multiple appropriate criteria when identifying participants to enable triangulation or consilience between different indicators of expertise, thereby increasing the reliability of suitable participant identification, particularly when broader definitions of expertise are invoked. For example, advanced teacher qualifications are likely to offer evidence of an extensive knowledge base, student exam achievement of effective curricular instruction and leadership roles of wider impact and value to relevant communities of practice.
In language teaching research, there are a number of studies that, while insightful, identify participant teachers as experts based on criteria that would be considered insufficient by the above authors. These include, for example, studies by Cumming (Reference Cumming1990), Han (Reference Han2021) and Shin et al. (Reference Shin, Lee, Brawn and Do2021), which mention only experience as a basis for characterising participants as ‘expert teachers’ and Farrell's well-cited study (Reference Farrell2013), involving participants who are described as being selected based on experience and initial qualifications alone.Footnote 3
Two further validity threats relating to participant selection frequently impact on the utility of a study's findings. These are:
1. researchers influencing aspects of their participants’ expertise prior to, or during, the study;
2. researchers cherry-picking participants to confirm their prior personal theories of expertise.
An example of the former is Tsui's (Reference Tsui2003) choice to select one of her own faculty students (both former and contemporaneous) as the expert language teacher in her study. The teacher's beliefs and practices were evidently influenced both by her education in Tsui's department (see, e.g., pp. 90–91, p. 96) and by Tsui's writings (p. 226), making it difficult to assess which aspects of her practices and beliefs were the manifestation of her own experientially developed expertise. Concerning the latter, some researchers have made the problematic decision to ‘cherry-pick’ study participants whose practices (observed during the selection process) meet their own personal criteria for good teaching. They have then reported these practices as evidence to validate their theories of good teaching (i.e., circular reasoning) (e.g., Sabers et al., Reference Sabers, Cushing and Berliner1991; Westerman, Reference Westerman1991). Such examples offer further evidence for the need for researcher critical reflexivity.
4.3. The limitations of subject-specific research
A final issue of theoretical importance concerns the extent to which we theorise and research issues of quality through subject-specific or generalist lenses. Applied and educational linguistics are fields founded largely on the assumption that language teaching and learning are fundamentally different to other types of teaching and therefore must be studied and theorised in isolation (e.g., Dörnyei, Reference Dörnyei2005). While there are aspects of subject-specific pedagogy that must always be researched and discussed separately, there is also ample evidence that many aspects of appropriate good practice in education apply widely across many subjects (Freeman, Reference Freeman2016). Both systematic reviews of expert teacher research (Anderson & Taner, Reference Anderson and Taner2023; van Dijk et al., Reference van Dijk, van Tartwijk, van der Schaaf and Kluijtmans2020) and analyses of ‘good’ (e.g., Korthagen, Reference Korthagen2004) or ‘effective’ (e.g., Campbell et al., Reference Campbell, Kyriakides, Muijs and Robinson2004) teacher practices have identified a large number of areas of importance that are not subject specific (e.g., interpersonal relationships, classroom management, lesson planning and organisational skills, in-class awareness of learner behaviour, use of formative assessment, pastoral support, etc.). Yet, because of the likely subject-specific interest of researchers in any given field of education (including language pedagogy), there is an ever-present danger of overlooking or underemphasising the fact that many of the factors that make ‘our’ expert teachers effective may not actually be subject-specific, and are linked – both to each other and the expert teacher's mission and ethos – as a holistic, embodied, situated model of quality in the classroom.
5. Mapping out the field of future teacher expertise research
Figure 2 provides an overview framework for how teacher expertise research may develop in the forthcoming 10–20 years. It offers one vision of how researchers working at different scales (macro, meso and micro), potentially approaching the topic from different epistemological perspectives, can contribute to complementary research goals. Taken together, these goals (identified in the ‘Focuses’ and ‘Broad RQs’ rows) provide a means by which we can build our understanding of teacher expertise, both in the round and in specific instances, thereby enabling us to flesh out a much more detailed version of the ‘differentiated framework of teacher expertise’ offered by Anderson (Reference Anderson2023c, pp. 224–225) as a ‘work in progress’. In the ‘Methodological options’ row, the framework distinguishes between single participant and small-n studies (c. <10 participants), mid-n studies (c. 10–30 participants) and large-n studies (c. >30 participants). It also identifies a range of possible research designs, some of which are likely to be appropriate for only one paradigm perspective and others for more than one. Systematic reviews, which can adopt quantitative, qualitative and mixed methods approaches, stretch across the three paradigms (see Research task 6). The utility column identifies areas of practice, policy and theory for which such research may offer useful insights. In all rows of this framework, the horizontal location of each item reflects the most likely scale. Both scales and paradigmatic positions provided should be seen as indicative only, rather than prescriptive.
Teacher expertise research task framework
Note: RQs: research questions; ET: expert teacher; NET: non-expert teacher; EXP: expertise.
6. Research tasks
Here I present six example research tasks – indicating in each case a likely scale from Figure 2, moving from micro- to meso- and then macro-scale. These should be seen as examples only; the framework itself offers opportunities for researchers to identify, or add, other tasks suited to the specific contexts, areas of focus and needs of communities around the world. Further, the example tasks in question have been included to be representative of the wide range of understandings of expertise in the literature and are not necessarily all consistent with a single definition or epistemological perspective, including mine.
Research task 1 (micro)
Investigate the full ‘value’ of an expert (language) teacher
Much of the research literature on teacher and teaching quality, particularly that which is based on quantitative and large-scale studies, tends to characterise teacher ‘impact’ narrowly as impact on learning outcomes, typically operationalised through measures of learner academic achievement (UNESCO, 2017). Aside from the immense difficulty of measuring the so-called ‘value-added impact’ of an individual teacher (see, e.g., Berliner, Reference Berliner, Fan and Popkeitz2020; Darling-Hammond, Reference Darling-Hammond2012; Kane & Cantrell, Reference Kane and Cantrell2010), it should be noted that any teacher's potential impact on learners goes far beyond academic achievement (see, e.g., Brophy & Good, Reference Brophy, Good and Wittrock1986; Bucci, Reference Bucci2003; Goe et al., Reference Goe, Bell and Little2008; Muijs et al., Reference Muijs, Campbell, Kyriakides and Robinson2005). Indeed, the impact and value of expert teachers may extend well beyond the classroom to include impact on colleagues, institutional well-being and even the local community (e.g., Amrein-Beardsley, Reference Amrein-Beardsley2007; Campbell, Reference Campbell1991; Gode et al., Reference Gode, Khomne, Lingala, Mukherjee, Naik Khwaja, Prathikantam, Sagrolikar and Tayade2021; Goe et al., Reference Goe, Bell and Little2008). As Padwad (Reference Padwad, Gode, Khomne, Lingala, Mukherjee, Naik Khwaja, Prathikantam, Sagrolikar and Tayade2021) observes, ‘these teachers manage to address challenges and impact their learners’ lives by going ‘out of the way’, by going beyond the classroom, the curriculum and the system’ (p. vii).
With this in mind, an ethnographically-oriented single site study involving an expert teacher would be well suited to demonstrating this wider impact. Such a study could offer detailed insights on exactly how, when and why the expert teacher in question exerts a positive influence or impact on learners (including beyond academic achievement), colleagues and community. Such a study would need to be medium- to long-term in duration (e.g., three months to one year) – well within the scope of a carefully planned Ph.D. study. Useful guidance for ethnographic research in education is offered, for example, by Mills and Morton (Reference Mills and Morton2013) or Pole and Morrison (Reference Pole and Morrison2003). Data collection may include observational data on interactions with learners (e.g., building learner self-esteem, facilitating social and emotional development, etc.), peers (e.g., novice teacher support, resource and ideas sharing, staffroom community, etc.) and local community (e.g., pastoral relationship with students, students’ families and their well-being, participation in local community events, identifying and supporting vulnerable and out-of-school children; see, e.g., Lingala, Reference Lingala, Gode, Khomne, Lingala, Mukherjee, Naik Khwaja, Prathikantam, Sagrolikar and Tayade2021). It would also include interview data from members of these groups, eliciting their opinions, stories and beliefs concerning the role and influence of the expert teacher. Such studies could aim to offer useful recommendations for institutions and educational systems seeking to understand, encourage and even reward those teachers who demonstrate a wider impact (e.g., through the use of impact portfolios). There are very few, if any, prior studies on the wider impact of expert teachers in any field of education, although Ladson-Billings’ (Reference Ladson-Billings2009) account of successful teachers of African American children offers numerous relevant insights of the wider impact of the teachers involved and Traianou's (Reference Traianou2006, Reference Traianou2007) ethnographic case study of an expert primary school teacher offers a useful methodological template.
Research task 2 (micro)
Investigate the longitudinal development of (language) teacher expertise
Because of the challenges associated with longitudinal research (e.g., participant attrition, incomplete datasets, unplanned events), there is a well-acknowledged lack of such studies in social science (e.g., Barry, Reference Barry2005; Keeves, Reference Keeves and Psacharopoulos1987), and expertise research is no exception here. Yet, such studies, if well conducted, are immensely useful for understanding aspects of development, particularly in areas of cognition (e.g., knowledge, beliefs, interactive awareness), but also in professionalism, identity and pedagogic practice; all are of key importance to expertise research.
While early theorisation of the development of professional expertise assumed that it would follow a smooth progression (e.g., Berliner, Reference Berliner1988; Dreyfus & Dreyfus, Reference Dreyfus and Dreyfus1986), subsequent scholarship (Bereiter & Scardamalia, Reference Bereiter and Scardamalia1993), including in language teaching (e.g., Atkinson, Reference Atkinson2021; Hirvela, Reference Hirvela, Seloni and Henderson Lee2020; Tsui, Reference Tsui2003), has suggested a much more complex pathway requiring the emergence of what Hatano and Inagaki (Reference Hatano, Inagaki, Stevenson, Azuma and Hakuta1986) coined ‘adaptive expertise’ and contrasted with ‘routine expertise’. Both Hatano and Inagaki and Bereiter and Scardamalia (Reference Bereiter and Scardamalia1993) argue that, in complex areas of social practice, expertise does not simply involve procedural fluency, but also the ability to adapt to, and (critically) learn from, unfamiliar situations of practice – something that Schön (Reference Schön1983, Reference Schön1987) also highlighted. While more cross-sectional than longitudinal, Tsui's (Reference Tsui2003) well-known language teacher expertise study identifies Bereiter and Scardamalia's (Reference Bereiter and Scardamalia1993) construct of ‘progressive problem solving’ as key to expertise development (see Chapter 10). Other scholars have extended the contrast between routine and adaptive expertise (see Riel & Rowell, Reference Riel, Rowell, Rowell, Bruce, Shosh and Riel2017; Schwartz et al., Reference Schwartz, Bransford, Sears and Mestre2005) to identify and contrast different potential pathways to expertise (see Figure 3). Although interesting in itself, much of this scholarship requires further empirical support and, for this, longitudinal studies are essential.
Three paths to expertise (from Riel & Rowell, Reference Riel, Rowell, Rowell, Bruce, Shosh and Riel2017, p. 673)
A number of longitudinal studies of early teacher development offer useful methodological designs that may be appropriate to the proposed task (e.g., Hong et al., Reference Hong, Greene and Lowery2017) but there are very few involving expert teachers. The challenge of finding potential future experts and retaining them for the duration of such a study is likely to be an inhibiting factor here. It would require a larger initial cohort (e.g., 10–20 teachers) from within which – accounting for participant attrition in longitudinal studies – different levels of expertise would likely emerge with development over a 5–8-year period.
An alternative design that offers useful insights into the early development of aspects of language teacher expertise in university graduate teaching assistants is reported by Christiansen et al. (Reference Christiansen, Du, Fang and Hirvela2018). Adopting a narrative enquiry approach, data was initially collected over two years (2008–2010) and then returned to much later (in 2017) when the participants, who were also co-authors of the study, were able to reflect insightfully on their own developmental challenges, needs and successes as they first developed routine expertise and then moved towards adaptive expertise. It also offers a useful example of a participatory teacher expertise study. Also see Tardy et al. (Reference Tardy, Buck, Jacobson, LaMance, Pawlowski, Slinkard and Vogel2022) for a similarly-insightful participatory study of early language teacher expertise development, albeit over a shorter time period (one year).
It is also useful to investigate how established expert teachers adapt to new contexts and related challenges; such studies would be able to begin with expert teachers and would thus require a smaller initial cohort. A useful example of this is the participatory study by Bullough with the expert teacher Kerrie Baughman (Bullough, Reference Bullough1989; Bullough & Baughman, Reference Bullough and Baughman1993, Reference Bullough and Baughman1995). The study is useful for its holistic representation of Kerrie's cognition, personality and practice as she moves from pre-service to experienced practice and then experiences challenges moving from one school to another (Bullough & Baughman, Reference Bullough and Baughman1995). It offers numerous useful insights, including support for Bereiter and Scardamalia's (Reference Bereiter and Scardamalia1993) construct of progressive problem solving and for Berliner's observation (e.g., Reference Berliner1988, Reference Berliner2004) that teacher expertise is highly context-specific (also see Lee & Yuan, Reference Lee and Yuan2021) and not necessarily transferrable. Such research is of immense importance in systems where regular teacher transfers between institutions are required or encouraged by educational authorities (e.g., in China and India) to understand the potential negative impact that such policies may have on teaching quality. For example, a researcher or research team could recruit participants who are coming to the end of a posting and then conduct some longitudinal ethnographies or case studies. They could collect initial data at the current institution and further data at two subsequent points in time at the new institution to understand which aspects of participant expertise were transferrable, which were not, and how their adaptive expertise enabled them to manage the changes involved.
Research task 3 (micro-meso)
Conduct comparative case studies investigating (language) teacher expertise in previously unresearched contexts
Given the general paucity of expert language teacher studies, further exploratory case studies are likely to be useful, especially in previously unresearched contexts, both geographically (e.g., Africa, Latin America, parts of Asia) and systemically (e.g., basic education, low-cost private school education). There is a particular need for such studies to be conducted in low- and lower-middle income national contexts, where conditions, constraints and challenges are likely to influence teacher expertise in complex ways (Anderson, Reference Anderson2023c).
In recent years, comparative and multiple case studies have become popular as research methodologies (e.g., Bartlett & Vavrus, Reference Bartlett and Vavrus2017; Stake, Reference Stake2006), including in education. They involve researchers conducting several case studies (usually 3–10), either simultaneously or in sequence, to identify both similarities and differences between the cases in order to understand what Stake (Reference Stake2006, p. vi) calls the ‘quintain’ – the phenomenon in question. They also shed useful light on how different causal factors influence variations within the sample (Bartlett & Vavrus, Reference Bartlett and Vavrus2017). There are already several examples of comparative case studies in teacher expertise research, either comparing several expert teachers (e.g., Milstein, Reference Milstein2015; Sorensen, Reference Sorensen2017) or comparing expert teachers with their novice or experienced non-expert peers (e.g., Li & Zou, Reference Li and Zou2021; Tsui, Reference Tsui2003). To my knowledge, only two such studies have been conducted in low- or lower-middle income national contexts: Toraskar's (Reference Toraskar2015) and my own (Anderson, Reference Anderson2021, Reference Anderson2023c), both in Indian secondary education. Researchers interested in conducting such further studies could focus on a specific national context of interest. Alternatively, they could conduct an international study of expert teachers working in comparable institutions or levels in different national contexts (e.g., a study of expert primary teachers of English in Francophone Africa). Comparative case studies can also focus on a specific area of expertise, such as expert teacher improvisation (Sorensen, Reference Sorensen2017), scaffolding (Li & Zou, Reference Li and Zou2021) or reflection (Gross, Reference Gross2014).
Methodologically speaking, there is much flexibility regarding how case studies can be conducted, and this will be influenced by logistical constraints, particularly the time available, number of participants, contextual constraints, as well as the specific focus of the study. Particularly in underfunded educational systems, only a limited range of data to inform participant selection criteria may be available, meaning that researchers should make use of expertise indicators judiciously and flexibly (Anderson, Reference Anderson2023c). Case study data collection typically involves multiple data sources (lesson observation, participant interviews, document analysis, stakeholder interviews), which are triangulated to offer individual case descriptions that are then compared with one another and to prior studies in other contexts, enabling the researcher to identify the specific influence of the context(s) in question on the quintain (i.e., teacher expertise). Researchers interested in conducting comparative case studies of language teacher expertise in new contexts can draw upon any of the above prior examples for guidance and may elect to involve the participants in planning the study through, for example, choosing the focus or deciding upon the outputs it produces (see Anderson, Reference Anderson2023a).
Research task 4 (meso)
Investigate the relationship between advanced language teacher qualifications and teacher expertise and/or effectiveness
Useful research has been carried out in the USA to investigate the relationship between one particularly prominent advanced teacher qualification, National Board (NPBTS) certification,Footnote 4 and teacher effectiveness and/or expertise (e.g., Bond et al., Reference Bond, Smith, Baker and Hattie2000; Hattie, Reference Hattie2003; Smith, Reference Smith, Baker and Hattien.d.; What Works Clearinghouse, 2018). While findings are mixed, the balance of evidence generally supports the qualification as an indicator of both teacher effectiveness and expertise (see, e.g., Smith & Strahan, Reference Smith and Strahan2004). A number of advanced language teacher qualifications exist, both international (e.g., in English language teaching (ELT): Cambridge DELTA, Trinity DipTESOL, Trinity CertPT) and national (e.g., Masters-level qualifications). However, there is little research on how either completing or having such qualifications impacts on teacher quality – an issue that should be of much greater interest to the profession than this lack of prior research suggests. A number of potential research designs could be adopted for studies in this area, such as before–after studies examining the extent to which gaining the qualification leads to changes in teacher knowledge, attitudes and beliefs, or teaching practice. Alternatively, case studies or larger scale correlational designs could investigate the extent to which teachers who have a given qualification meet criteria for expertise or effectiveness, particularly in comparison with peers who do not possess the qualification.
Two studies of relevance shed initial light onto these two areas, respectively. First, Borg's (Reference Borg2011) qualitative study into the impact of the Cambridge DELTA on six English language teachers’ beliefs, which found ‘considerable, if variable’ (p. 370) evidence of positive impact. Second, Andrews and McNeill's (Reference Andrews, McNeill and Bartels2005) mixed methods investigation into the language awareness (i.e., aspects of knowledge) of what they call the ‘Good Language Teacher’ (pp. 161–162), which focused on the subject knowledge of three L2 teachers of English who received distinction grades on an advanced qualification in Hong Kong. They concluded that there was sufficient evidence to ‘suggest that in most respects all three subjects are experts according to Tsui's criteria’ (p. 174; citing Tsui, Reference Tsui2003). Both studies, therefore, offer promising initial evidence to support claims for positive relationships. However, both are limited by issues of scale and the narrow focus involved (on beliefs and knowledge, respectively). Similarly-designed studies with larger cohorts and a focus also on classroom practices would potentially shed greater light onto how such qualifications impact on broader understandings (e.g., multi-componential or holistic) of teacher expertise.
An appropriate design that would succeed in investigating the extent to which an advanced teacher qualification constituted a valid indicator of language teacher expertise would be a matched pairs study, involving 8–10 pairs of teachers (i.e., 16–20 participants). Each pair would have similar experience and would work in the same institution, making them as comparable as possible. The ‘independent variable’ would be whether they have a given advanced teacher qualification or not, with one member of each pair having the qualification and the other not. A number of potential outcome measures (i.e., ‘dependent variables’) could be investigated. These include knowledge and beliefs that could be measured through the use of testing instruments and (semi-)structured interviews, respectively. Aspects of teaching practice could also be chosen as outcome measures. They could be investigated by recording lessons that are then subsequently evaluated, either for evidence of expertise (e.g., when compared with the findings of systematic reviews of expertise studies) or using locally-defined measures of expertise, such as local school inspector, or teacher educator evaluations, if deemed appropriate. These different outcome measures may be then brought together to examine the extent to which the teacher with the qualification demonstrates greater expertise or not.
Research task 5 (meso-macro)
Investigate perceptions/understandings of teacher quality in (language) teaching in under-researched communities
While the term ‘expert teacher’ and how it is understood may be both language and community specific (see section 2, regarding Chinese terms), it can be assumed that there are understandings of teacher-embodied quality in any teaching community, either implicit or explicit. It can be expected that these are likely to vary between such communities, particularly in different national, stage and curricular contexts (Sternberg, Reference Sternberg2007). As such, in order to build a richer and more diverse database on what quality means to key stakeholders in education, there is a clear need for further studies investigating their understandings of ‘good teaching’ or ‘good teachers’. This is particularly important in language teaching, where so much of the discourse on quality has emanated from ELT norms in the Anglophone ‘centre’ (Phillipson, Reference Phillipson1992), based on competence-oriented models of good practice in English as a second language (ESL) instruction (see Anderson, Reference Anderson2023b; Kramsch, Reference Kramsch, Gass, Bardovi-Harlig, Sieloff Magnan and Walz2002).
In order to avoid being influenced by such biases, studies may prioritise qualitative data to allow participants to express their opinions without being led by the assumptions implicit in specific constructs (e.g., ‘communicative’, ‘learner-centred’, ‘competence’) that might be present in quantitative items. An example study (macro-scale) was conducted by the author (Anderson, Reference Anderson2020) and focused on Indian secondary education as its primary context. A wholly qualitative survey instrument was developed for the study. It invited respondents to (a) imagine an effective teacher's lesson, and (b) describe an effective teacher through the use of open questions and prompts for description. Seventy-five responses were analysed qualitatively and quantitatively to inform a 214-word description of the Effective Indian Secondary Teacher of English as a ‘Shared Beliefs’ prototype (p. 15). This is an example of the kind of exploration of context-specific understandings of expertise suggested as optional in Figure 1.
Future studies may choose to use similar tools to the above study (adapted to context). Alternatively, if larger samples are considered useful, researchers may choose to adopt what Dörnyei (Reference Dörnyei2007, p. 171) calls a qual → QUAN design, starting with focus groups or interviews to build appropriate items for a statistical questionnaire inductively.
Research task 6 (macro)
Conduct a systematic review of prior studies of (language) teacher expertise, competence and effectiveness
This article has, I hope, revealed both the complexity of teacher expertise as a construct and the wide potential range of aims, means and methods for researching it. As such, given the diverse range of expertise studies discussed above, readers may wonder how, if at all, it is possible to conduct a systematic review of these studies.Footnote 5 To date, only one large-scale systematic review of teacher expertise studies has been published (Anderson & Taner, Reference Anderson and Taner2023). The challenges presented by the above-discussed variety of study types and designs required the authors of this study to innovate methodologically, adapting Sandelowski's metasummary method (e.g., Sandelowski et al., Reference Sandelowski, Barroso and Voils2007) from healthcare research for the purpose.Footnote 6 They chose to focus on curricular subjects at primary and secondary (K12) levels, analysing findings from 106 qualitative and quantitative research reports in total. While the initial literature search carried out for this study identified 16 studies that involved language teachers (author's unpublished data), only eight of these met pre-specified inclusion criteria, which were developed in consideration of the aims of the metasummary. Researchers interested in conducting future systematic reviews of language teacher expertise research may opt to enlarge this fairly small sample by one of several potential means:
1. using an alternative systematic or narrative review approach (e.g., qualitative metasynthesis; see Sandelowski et al., Reference Sandelowski, Docherty and Emden1997; Thorne et al., Reference Thorne, Jensen, Kearney, Noblit and Sandelowski2004);
2. expanding the focus to include studies conducted in adult and tertiary education;
3. widening the scope of the construct investigated to make it inclusive of a wider range of measures of teacher-embodied quality.
While these three changes may impact in complex ways on the reliability and validity of the findings, providing limitations are made clear and findings are presented with awareness of this revised scope in mind, such studies would still help to build a more complete understanding of ‘language teacher quality’ as the underlying construct under investigation. Ultimately, however, if this article inspires academics and Ph.D. students to conduct more such research, there may be a larger database of potential studies to work from in future. As such, this may be a research task better postponed until there is a larger number of studies to analyse and report on.
7. Conclusion
In this article, I have presented evidence of insufficient research into language teacher expertise. I have also offered suggestions for how the theoretical and methodological underpinnings of expertise studies can be strengthened. I have also aimed to make a convincing case as to why such research can inform a range of areas of policy, practice and theory in educational provision with a view to impacting positively on learners, teachers and broader institutional communities in which we live and work.
The framework provided in Figure 2 is offered as a heuristic tool to map out the potential scope of expertise research by drawing eclectically upon different paradigmatic positions and methodologies available to social scientists. While some methodological purists may see this as inappropriate, as Stake (Reference Stake2006, p. 7) has observed, ‘the pursuit of science seems to place the highest value on the generalisable, and the pursuit of professional work seems to value the particular most, but they both need both’ – this is particularly true in teacher expertise research. The six research tasks proposed are examples from across the framework and are far from exhaustive; many others could be suggested. For example, given the potential ability of language teacher research to cross national borders, there are opportunities for cross-cultural studies investigating aspects of expert teacher practice or cognition that may vary between cultures (see, e.g., McIntyre et al., Reference McIntyre, Mainhard and Klassen2017; Rollett, Reference Rollett, Shelton Mayes and Banks2001, for studies in other subject fields). Such research could inform the extent to which strategies, methodologies or practices may be transferrable between national educational systems. Alternatively, it would be possible to use the findings of teacher expertise studies to empirically evaluate different frameworks used in language teacher education (e.g., qualification assessment criteria), teacher professional development (e.g., British Council, 2015), or teacher appraisal; something that, to my knowledge, has never been done, yet is so obviously useful. Burns (Reference Burns2022) and Hirvela and Belcher (Reference Hirvela and Belcher2022) offer further useful ideas for future teacher expertise studies.
I would like to conclude with a reflexive observation. Some readers may be concerned by my apparently uncritical use of terms such as ‘educational provision’, ‘value-added impact’ and ‘stakeholders’ as representative of neoliberal agendas. I acknowledge this danger, but also argue that the issues and terms in question are central to mainstream discourse on education in the media, politics and the third sector. Academics who fail to engage with these debates run the risk of leaving key decisions regarding teaching quality to those who would seek to reduce it to an academic achievement numbers game in which PISA-type (Programme for International Student Assessment) tests dictate national policy, research agendas and government objectives (see Berliner, Reference Berliner, Fan and Popkeitz2020). In contrast, if defined in all its complexity and researched appropriately, I believe that the construct of teacher expertise can provide a suitably powerful alternative; a complex, pluralist, context-sensitive understanding of quality in education that may help to convince decision-makers to value teachers for their full worth and invest in them appropriately.
Acknowledgements
Many thanks to Meifang Zhou and Zening Yang for assistance with Chinese language terminology. Thanks also to the three anonymous reviewers for insightful comments and critique on earlier drafts of this piece.
Jason Anderson (University of Warwick, UK) is a teacher educator, author, educational consultant and researcher, who works in both language teaching and mainstream education. He has supported teachers in over 30 countries worldwide, both pre-service and in-service, for national ministries of education and organisations including UNICEF and the British Council. He has published award-winning books and research on aspects of language teaching, multilingualism, teacher reflection, teacher expertise and teacher education. His interests include teaching methodology, multilingualism and translanguaging, and education in the global South, where he has spent much of his career as a teacher educator. His latest book, Teacher expertise in the global South (Cambridge University Press), investigates what quality in education means in the global South and why it is often misunderstood or overlooked by educational reform initiatives.
Open access
