Can Teacher Case Study Research be Participatory? Critical Reﬂections on the Approach Adopted for an English Language Teacher Expertise Study in India

C ase study research is a well-established, widely recognised methodology that enables us to understand phenomena of interest in social sciences through a range of data collection means, often com-bined (Flyvbjerg, 2011). 1 As such, it is recognised by scholars working across the paradigm continuum – from interpretivist (e.g

document and understand existing phenomena for their intrinsic value (Flyvbjerg, 2011).
This article reports on an attempt conducted by the author, within a PhD project, to make a comparative case study participatory (Anderson, 2021). It will also reflect on how, and in what contexts, such an approach may be of use to other educational researchers. In the context of growing calls for research that is both more relevant to teachers and more participatory and/or collaborative (see McKinley, 2019;Rose, 2019, both recently in this journal), I argue that, while fully participatory case study research is likely to be challenging for a number of reasons, there is nonetheless justification for making such projects partially participatory, when appropriate, from methodological, practical and ethical perspectives.

A PARTICIPATORY CASE STUDY OF INDIAN TEACHER EXPERTISE
As an experienced English language teacher educator from the UK who has spent much of his career working in low-income contexts across the Global South, I am very much aware of inequalities relating to issues of power, prestige and discrimination historically endemic within our profession (Anderson, 2016;Canagarajah, 1999;Holliday, 2005). Despite repeated calls for change, all too often those of us based in the Anglophone 'centre' of ELT/TESOL (Phillipson, 1992) are mistakenly perceived as (international) 'experts' while the vast majority of teachers around the world are perceived as 'local' (e.g. Carless & Walker, 2006) and their contexts as peripheral. My experience in this role has taught me not only that expertise in teaching is always contextually specific (Berliner, 2004), but also that it exists everywhere, irrespective ofindeed, as a function ofthese contextual constraints and affordances. Yet, despite the existence of over 100 empirical studies of expert teachers conducted in higher income contexts in North America, Europe and East Asia, almost no teacher expertise studies have ever been conducted in contexts in the Global South. 2 Thus, the primary aim of my PhD project was to identify and document the practices of expert teachers in a Southern context in detail, and to do so in a way that avoided explicit intervention. I wanted to document their practice as faithfully as possible; doing so would enable me to argue that Southern expertise exists and yet is overlooked in attempts to improve educational quality. However, I was concerned that my own background may lead to an inequitable relationship between myself and the research participants, and potential exploitation (even if unintended) as a result. I therefore set about finding a means to make the study as equitable and participatory as possible while also intending that the outputs of the project would be useful for all parties.

Drawing inspiration from community development literature
While my attempts to find prior non-interventionist participatory teacher case studies yielded little of interest, 3 a wider search revealed evidence of participation from case studies conducted within the community development literature (e.g. Valencia-Sandoval, Flanders, & Kozak, 2010), where a number of authors propose useful hierarchies of participation (see Bergold & Thomas, 2012;Cornwall & Jewkes, 1995;Pretty, 1995). These sources largely agree that for a study to be considered participatory both sides (participants and researcher) must benefit from the research process, and also that there needs to be meaningful interaction at the study design stage, when '[t]he most important distinctions centre on how and by whom is the research question formulated and by and for whom are research findings used [sic]' (Cornwall & Jewkes, 1995, p. 1668. Also based on this literature, I adopted a working definition of 'participatory research' as research which involves participants to varying degrees in making decisions and/or fulfilling roles traditionally assigned to the researcher(s). In agreement with Bergold and Thomas (2012), I also make a distinction between participatory and action research; while the latter typically involves a specific intervention or 'action' during the research cycle (e.g. Burns, 2009), the former does not have to.

Developing an equitable recruitment procedure
While a large number of teacher expertise studies rely on nomination by senior stakeholders (head teachers, teacher educators or inspectors) to identify participants (e.g. Hanusova, Pisova, Kostkova, Janikova, & Najvar, 2013), I avoided this approach, due, in part, to concerns about the danger of relying primarily on the opinions of individuals who may not be sufficiently well-informed (Yang, 2014), and, in part, to my intention to offer equity of opportunity to participate in the study to members of a wider teacher association community, within which I had already conducted contextualising research (see Anderson, 2020). A call for participation was shared via the association's social media channels, making both the intentions of the study and eligible inclusion criteria clear (drawn from prior expertise research; see Palmer, Stough, Burdenski Jr., & Gonzales, 2005). I invited interested parties to respond, initially with self-evaluations of these criteria that were subsequently discussed and verified in individual video interviews and/or in situ. This procedure identified eight available participants who all met at least five potential expertise criteria (see Anderson, 2021, for details).

Project planning meetings
To involve participants in the project design phase as much as possible, I organised two group meetingsthe first online and the second a whole-day face-to-face workshopto plan aspects of the study together. In the online meeting, we agreed upon the agenda for the workshop as follows: 1. Exploration of roles of participants and researcher; 2. The focus of the PhD study; 3. A co-authored publication produced by the participants; 4. Participant group reflection without the researcher; 5. Timetabling of case study visits and practical issues.
All but one of the participants were able to attend the planning workshop; the eighth provided feedback on a detailed agenda beforehand. As well as providing opportunities for participants and myself to bond and build a productive community of practice, the planning workshop achieved its aims. It was video recorded with consent and made available to participants afterwards.
We began the workshop by exploring our roles in the project; my needs with regard to the PhD study (e.g. single authorship, non-interventionist) were made transparent. We also discussed likely challenges during the study, particularly 'reactivity' (Atkinson & Hammersley, 2007)also known as the observer effectand concerns expressed by several participants that I might be 'exploited' (inverting my initial concern) by local authorities in some contexts. For example, several correctly predicted that I would be called upon to conduct workshops with teachers for district authorities.
We then considered the focus of the study itself. While it was a prerequisite on my part that it be non-interventionist (a study primarily of their expertise), I was nonetheless able to involve the teachers in two important decisions concerning the study focus (thereby ensuring meaningful participation at the design phase; Cornwall & Jewkes, 1995). The first related to the degree of involvement of their peers as potential points of comparison, with three options presented: a study focusing solely on them; a matched pairs study (involving one colleague each); or a study focusing primarily on them, but also taking advantage of opportunities to observe and interview willing colleagues whenever possible. The teachers felt that it was useful to involve peers, but only with their approval after they had got to know the researcher, leading to rapid consensus on the third option, also preferred by the absent participant.
The second decision concerned which aspects of their practice and cognition the study would focus on. I presented five options for them to discuss and select from, presented as follows: After lengthy discussion, participants agreed on the fifth option. As one explained, this allowed for the study to include 'all the multiple roles a teacher is supposed to do'. The absent teacher also preferred this option.
Discussion of the co-authored publication was more free-ranging, and while the teachers were enthusiastic about it, there was less agreement here. Ideas for this publication had emerged from a prior survey (Anderson, 2020) and ranged from a practical publication (e.g. including ready-made lesson plans) to an edited book with participants contributing diverse chapters on areas of personal interest. While some aspects were agreed upon, others were not, so we decided to continue discussion through an already established chat group, and set this aside for now.
Participants were then given an opportunity to discuss the project without my presence. I suggested they choose a group spokesperson and create a separate chat group (excluding me) to discuss any concerns that could be voiced via the spokesperson. They were also asked to discuss a number of provocative questions regarding exploitation and opportunities (e.g. 'Do we feel we are being exploited or involved?'; 'Does this project provide the opportunities for us that we had hoped for?') and encouraged to present issues of concern after I returned. When I did, the participants reported that they felt happy with how the workshop was progressing. Two issues of concern were addressed: First, they requested copies of their individual classroom and interview data to use for their own purposes, which I confirmed. Second, they requested certificates for their participation in this workshop from my university, which were also provided.
The final item for discussion concerned practical issues to ensure that timetabling of the visits to participants would be convenient for all. Issues of child protection and consent from appropriate stakeholders (headteachers, parents and learners) were also discussed.

Participant consent and ethical approval
Participant teacher consent forms were finalised after the workshop to reflect agreed choices concerning what data would be collected. These were then emailed to participants and signed when I arrived at their respective schools.
Processes for collecting informed consent of other stakeholders, particularly learners and parents were discussed at the planning meeting and also with headteachers to ensure practices were locally appropriate while also being consistent with ethical approval requirements. In some contexts, headteachers recommended collection of signed consent forms from parents and in others (e.g. where parental literacy levels were lower) spoken consent was deemed more appropriate (see Upvall & Hashwani, 2001) and the village sarpanch (community leader) was also informed about the study.

Anonymising or recognising participants
Both during the planning workshop, and subsequently, I raised participants' awareness of their rights, either to anonymity (as much as is possible today; see Walford, 2018) or to be acknowledgedif they so choosein any outputs that I produced as the researcher. Given that this was a teacher expertise study, I had some confidence that the findings would be generally positive, although I also held concerns that data might reveal issues of potential sensitivity. Thus, the participant validation (member checking) phase of the study was conducted carefully (Stake, 2006) with two feedback opportunities for participants, both on their individual case descriptions, and on the completed findings, which compared and contrasted their practices. The 'critical reflections' I had included in each participant's case description were a key focus of the validation process; on the whole (with several minor exceptions), they found them justified and useful. This did not surprise me, given that all were, to varying degrees, reflective practitioners, several much more critical of their practice than I was. The final draft of these case descriptions also included extracts from respondent validation alongside the critical reflections for full transparency (see Anderson, 2021). At the end of this process, all eight participants chose unanimously to be recognised. 4

The participants' book and other benefits they experienced
While some aspects of the participants' publication were agreed upon during the planning meeting (e.g. to contribute chapters to a co-authored volume), others took shape through subsequent discussions in our chat group during data collection, being influenced, in part, by this phase of the study. They decided to contribute reflexive descriptions of their own contexts, challenges and practices (areas that my research had also focused on), as well as offering practical suggestions to less experienced teachers who faced similar challenges. I wanted to avoid becoming the editor for this publication so we found an alternative solutionfor participants to peer-edit each other's chapters, although I offered some copyediting support at a later stage for consistency in the final work. The teacher association involved in participant recruitment agreed to publish the book (see Gode et al., 2021). Aside from its hoped-for practical utility for other educators, this publication served to offer direct voice to the participants, enabling them to speak directly to interested readers. As such, while it was never a conscious intention of the project, their accounts served both as an alternative narrative to mine and as a source of potential triangulation for my findings.
Aside from this book, I became aware of a number of other ways that the participant teachers seem to have benefitted from this project. This includes the extensive interaction, peer-support and ideas sharing that has occurred in our still-active chat group and opportunities for them to work together on several initiatives (e.g. co-mentoring on action research projects and MOOCs, or co-presenting webinars). However, the most evident positive impact was somewhat unexpected to me; the insights they gained from reading their case descriptions during participant validation. Without exception, all reported finding this very useful, both for formative and self-validatory purposes, as the following feedback reveals: This whole project has widened my experience and shaped my teaching practice a lot. Especially in a place where your strengths are often dismissed and you become the mere victim, something like this project works like a wonder. It has given me confidence to move ahead no matter what comes my way. Henceforth, whenever I'll feel low, I'll always refer to this case description (from Anderson, 2021, p. 208).

CRITICAL REFLECTIONS ON THE PROJECT
Likely due to the careful planning of the project, data collection progressed smoothly. This is noteworthy, given that it involved over 250 lesson observations and 100 interviews in eight locations across India. Both the rapport that was developed within the group and the invaluable input of the participants at the design phase contributed to this success, as did my willingness to make adjustments in response to their suggestions.
Nonetheless, critical readers of this account may have identified that my study only succeeded in becoming what we might call partially participatory. Given that several key decisions regarding the study design and roles were made in advance of participant recruitment, it would probably be assessed as achieving consultation on Arnstein's (1969) ladder of participation. On Pretty's (1995) typology, it probably achieved functional participation, in which 'involvement may be interactive and involve shared decision-making, but tends to arise only after major decisions have already been made by external agents' (p. 1252); Pretty envisages two higher stages than this-interactive participation and selfmobilisation. Within this literature, as in the participatory action research literature in education (see, e.g. Dikilitas ß & Griffiths, 2017), higher levels of participant agency are typically deemed preferable for reasons relating to issues of ownership, autonomy and sustainability. However, it is possible that offering higher levels of participation in my project may have led to it transforming into something other than a case study. As Cornwall (2008, p. 274) notes, '[p]articipatory interventions may result in effects that were never envisaged at the outset'. While this may have led to the project being even more useful for the participants, it would have lost its ability to achieve the primary goal of the study from my perspectiveto offer a descriptive account of the practices of expert teachers working in a low-income context in the Global South that is capable of leading to concrete, empirically based recommendations for policy and practice (see, e.g. Anderson, 2022;Mahapatra & Anderson, 2022 for examples of such recommendations); precisely the reason why a case study design had been selected in the first place.

VALIDITY AND TRANSFERABILITY OF PARTIALLY PARTICIPATORY DESIGNS
When considering the potential validity of attempts to make case study research participatory, it can be argued that partially participatory designs (PPDs) are justifiable when compared to fully participatory designs if, for whatever reason, certain aspects of the methodology or design need to be decided before participants are recruited (e.g. to obtain funding or ethical approval). This includes most PhD studies and many funded projects, meaning that a PPD can potentially make participatory research available to a much larger number of researchers than has previously been considered possible. All other things being equal, PPDs, are likely to be more ethically sound, more useful to teachers and less exploitative than non-participatory designs. In addition, the consultation process itself may help to facilitate the smooth progress of data collection and analysis, something that is a particular concern in teacher case study research (see Traianou, 2007).
Nevertheless, while these implications are promising, I would like to conclude with a cautionary note. I believe that one of the reasons why this study was successful as a participatory project was because it framed the participants in a positive lightas expert practitioners. Teacher expertise studies, in this sense, are likely to be suitable for PPDs. However, studies in which there are likely to be a number of critical findings may encounter more challenges when attempting to achieve a degree of useful participation. For example, if participants dispute findings at the validation stage or envisage outcomes in conflict with those of the researcher, this may lead to a breakdown in relationships and/or participant withdrawal. As such, I advise any researchers considering a PPD to read Traianou's (2007) cautionary tale alongside the account presented here.

FUNDING INFORMATION
This research was supported by the Economic and Social Research Council, UK, under grant number ES/P000771/1.

THE AUTHOR
Dr Jason Anderson is a teacher educator, author, educational consultant and researcher working in both language teaching and mainstream education. He has supported teachers in over 30 countries worldwide, pre-service and in-service, for national ministries of education and organisations including UNICEF, the British Council and the University of Warwick, particularly across the Global South (Africa and South Asia). His current roles include working as an Assistant Professor in Applied Linguistics at the University of Warwick (UK) and also as series editor for the Delta Publishing 'Ideas in Action' series.