A little more madness in our methods? A snapshot of how the educational leadership, management and administration field conducts research

ABSTRACT The field of educational leadership, management and administration (ELMA) uses methods drawn primarily from cognate educational disciplines. But does this matter? This paper explores the methods used in recently published papers through a snapshot of six issues of six ELMA journals. The analysis showed a preponderance of survey, interview and case study methods, with one journal, JEAH, also publishing papers using methods drawn from history, philosophy and sociology. The snapshot also revealed the methods that were rarely used – for example, ethnography, visual and on-line methods. Through a Bourdieusian lens, the paper argues that the ELMA field appears to be somewhat removed from methods developments and debates in the wider educational and social science fields. There may thus be mileage in the ELMA field considering the use of additional methods, including the ‘wilder’ ones. The field might also benefit from understanding methods as more than tools and as practices possessed of a social life.

whether the methods chosen by a researcher are likely to produce the desired results, and whether those results match the claims made for significance.
There are problems in this choice talk. 'Choosing a method' presumes that the researcher has at their fingertips a wide range of research tools. This is not always the case. But let me assume for the sake of argument that it is. A further implication of 'just choose your methods' is that the researcher has total control of their decision. It is as if the choices that the researcher makes are free-floating, outside of disciplinary, institutional, spatial, national and discursive framings. Choice of methods is just a technical matter. There is no need for, or benefit from, a more critical interrogation of the workings of particular tools and approaches.
Could it be useful to think about methods as socially constructed? And does it matter, and if so, to whom and why? These questions were the starting point for this paper.
The questions arose from a recent word search I conducted on 10 major journals in the educational leadership, management and administration field (hereafter ELMA). My concern at that time was the research questions that were asked in the field. I was not so interested in methods, but rather in the kinds of projects that were undertaken. The word search that I did was a late addition to this project and was a very preliminary foray. The search of major journal publisher websites used standard research methods termsinterview, survey, mixed methods, narrative, questionnaire, ethnography, case study. The search revealed a preponderance of interviews and surveys (Ch. 2, Thomson 2017). The dominance of these two methods led to my questions (above) and an interest what in a further examination of the research methods used in ELMA might yield.
I wanted to know how research methods are addressed by ELMA researchers. I began by looking at my bookshelf where there were two ELMA methods texts. One, that by Brundrett and Rhodes (2013), is organised around three stages of designing a research project and is focused on the production of 'evidence'. It pays relatively little attention to epistemology and methodology. By contrast, Research methods in educational leadership and management by Briggs et al. (2012), covers epistemological and methodological concerns in a first section that has expanded with every edition. The notions of 'culture' (Stephens 2012), and taking 'a critical stance' (Grogan and Cleaver Simmons 2012), are specifically addressed in separate chapters. These concerns are preceded by a chapter which orients the reader to research, and to ways of thinking about methods.
In this opening orienting chapter, Morrison (2012) explains the 'features' of quantitative and qualitative research. Quantitative research is concerned with the use of measurement in order to establish causality, generalisability and replicability. Morrison contrasts these 'core features' with those of qualitative research which shares concern for context, a focus on participant perspectives, an interest in words and texts, the valuing of description and detail. She addresses the use of a mix of both approaches, noting that combining aggregated statistical data with discrete, individual responses may be problematic, due to the incommensurate nature of the data (after Bryman 1988). Morrison argues that the quantitative/qualitative binary is only one way to understand methods and that what is important is that the 'choice of method is determined by the needs of the investigation and not the personal preferences or fears of the investigator', and that 'completely different methods can have the same research aim' (p. 25). She too delivers the message that it is up to the researcher to choose.
Morrison's discussion of the quantitative/qualitative binary speaks to the most visible contemporary debate about research methods and their social construction and political usethat centred on 'evidence' and the use of experimental and 'what works' models. This debate permeates Western social science research in general, and educational research is no exception. Educational researchers in the UK and USA in particular are very familiar with arguments for and against randomised controlled trials, effects studies and systematic reviews (e.g. Davies 2000, Evans 2001, Hammersley 2001, Biesta 2007. They may take a firm position for/against the epistemological view that words and numbers are both subject to researcher influence and social construction.
But, while Morrison's discussion can be read against these contemporary debates about methods and methodology, she does not fully enter this territory. Yet, as educational researchers (e.g. Grek 2009, Ozga 2016 suggest, it is the promise of certainty offered by numbers that she discusses that constitutes the appeal of this kind of research to policy-makers. The corollary is that it is the very uncertainty and foregrounding of particularity and researcher interpretation of the qualitative that makes it politically unpopular. This methods debate has its own turn within the ELMA field, where there have been energetic discussions about the apparent congruence of school effectiveness and school improvement research with the interests of conservative policy-makers (critics include Thrupp andWilmott 2003, Wrigley 2004). In this debate, the SESI research designnamely, the use of test and exam data that substitutes for a deeper appreciation of wholistic educationalis particularly at issue. However, despite the lack of deep engagement with this contemporary methods argument in the two methods texts on my shelf, I could not begin my investigation into methods thinking that there was no discussion in the ELMA field. The methods used to establish 'truth', as well as the very possibility of truth, are the focus of some debate. But I could perhaps consider how this debate was taken up, by whom and to what ends. Did it appear in a corpus of recent publications? I therefore decided to take a 'snapshot' of the field to see what methods were used, and how they were discussed.
The remainder of the paper proceeds as follows: I outline my theoretical approach and the 'snapshot' method that I used. I then report the results, noting the frames that were apparent. I document what methods were less used or nor used at all and discuss potential questions arising from the analysis. I conclude with the implications of the snapshot for the field.
A snapshot: the method used in this paper A snapshot has a common sense meaning. It is a picture, taken at one point in time, perhaps carefully, perhaps not, of a specific event, place, thing or person. The person making the snapshot has chosen where to point the lens, what to include and exclude, what to put in the foreground and background. How well their picture turns out is dependent on external factors such as light, the technical capacities of the tool they are using, and their own skill. Shaking hands combined with low light and a slow auto-focus inevitably leads to a worthless image, usually discarded. The image, supposing it meets the maker's intentions, has perhaps been further processed/edited/manipulated. Editing in the digital era is likely to require proprietary software which crops, removes red eyes, re-colours and so on. The resulting image is a representation, not an exact replica of the original: it is not a picture which never lies, something that is 'true' (Tagg 1988, Chaplin 1994, Wells 2000. The snapshot is a helpful analogy for a type of research which offers a particular and limited view of a phenomenon (Hall et al. 2010, Thomson andHall 2017). Its boundaries and framing are explicit. The data are generated at a particular time; it is time specific and time limited. The researcher decides what to include and exclude and what to foreground and background. They also decide how to analyse, interpret and theorise the data. The process that they use to generate the data is not neutral, any more than a camera or software is. The researcher is not a passive agent in snapshot making, they make decisions which are subject to their own socio-cultural positioning and reflexive research practice. A research snapshot is not mimetic, but is a representation resulting from serial selections and exclusions.
The empirical work reported in this paper is a snapshot. It is limited and has been subject to researcher interpretation, as will be explained further. Like a photograph, this research is intended to perform a function in its field (Bourdieu et al. 1990). In this case, the purpose of the snapshot is to raise questions for the ELMA field.

A snapshot of the field
The paper uses a Bourdieusian framing, but lightly. My methods question addresses practices within a field (Thomson 2017). I take ELMA as a field of knowledge production, a field with blurred borders, containing particular positions. Field positions stand in relation to each other and agents who occupy each position have competing or complementary truths, narratives and practices geared to obtain status and credibility within the field as a whole. In this instance, a position is taken to be a journal and its agents, the community of editors, reviewers, writers and readers.
Within the education field more broadly, agents are engaged in contests for and the creation of symbolic capitalnetworks, credentials and knowledges. These not only (re)produce the field of education itself, but can also be 'cashed in' in other fields for economic capital (money and other financial assets), cultural capital, or socially valued knowledge and knowhow; and social associations and social group membership, or social capital. A journal practice can be understood as the production and reproduction of knowledge about and for the field (Fitzgerald and Gunter 2008), and through this, the production and reproduction of positional advantage/disadvantage for associated agents.
Knowledge production via a journal is not a simple processit is rather the result of the combined strategies of various actors (playing the journal game). Writers choose a particular journal on the basis of what it has already published, reviewers make judgements based on their own positioning and what they understand of the journal, the editor and editorial board can steer what is published through editorials, the declaration of particular types of papers it would like to receive, and through special issues. What ends up being published produces and reproduces a dominant journal genre.
But distinctive practices between journals might be expected in relation not only to differences in normative truths (doxa) and practices, but also in the wider context in which the agents exist. The field of ELMA knowledge production is not divorced from the wider educational field (including that of schools, colleges and higher education) nor other fields such as politics, economics and media. The position of each journal is also likely to have homologies with other positions in other fieldsdisciplines, institutions and so on.
Publications might be seen to reflect some aspects of the wider field. Agents in the ELMA field have variable access to economic capital (funding and resourcing), often directed via the political/policy field. Its cultural capitalknowledges such as research findings and teaching materialsare contingent on the credibility of its practices within and beyond the field, as well as the social capital it accrues via membership of networks, the affiliations and connections of agents, and formal partnerships within and beyond the field. This paper is a small step in locating practices that contribute to this field formation.
Methods too can be understood through the Bourdieusian metaphor of a game. Methods are integral to knowledge creating practices in the field, geared to producing knowers and ways of knowing, being and acting (i.e. truths but also misrecognitions). Choices of methods can be understood as relational moves made to create and acquire symbolic capital, status and associations.
And there is no avoiding the positioning of the researcher in the field. My own social positioningwhite Western second wave feministand my researcher actions, are written all through this paper, from its inception to its current textual form. Additionally, the project is conceived from a particular position within the ELMA fieldthat concerned with two of the journals under investigation. Not only do I publish in the International journal of leadership in education and the Journal of educational administration and history but I am also on their editorial boards. I gain status in the field through my association with these journals, and through the publication of papers in them. My own possible blind spots and taken-for-granted truths and values undoubtedly shape my discussion of ELMA journals, and their relationship with others in the field. In Bourdieusian fashion I have attempted to exercise considerable critical reflexivity in relation to the data and this analysis, but nevertheless, I do stand in a particular place and for particular kinds of scholarship that are already evident.
In this paper, Bourdieusian thinking is used to generate questions about the results of the snapshot. There is no attempt to offer a full analysis, but rather to think with Bourdieu's tools.

Taking a snapshot of ELMA methods
The data reported in this paper are drawn from 208 papers in six ELMA journals: Leadership and policy in schools, International journal of leadership in education, Educational administration quarterly, Journal of educational administration and history, School leadership and management, and Educational leadership, administration and management (see Figure 1).
As is the case in any snapshot, what is included in the corpus was governed by both reason and practicality. At the outset, there were some differences within this selection in both geography and orientations. The first three journals ( Figure 1) are edited in the North Americas, and the latter three in the UK. According to their mission statements, one in each trio is also of a more critical orientation than their two geographical colleagues -IJLE and JEAH. I also had full access to these journals through my university library, whereas other ELMA journals were only available to me after a one-year period. This was important as I also decided that I would look at the most recent six issues, thinking that if this was neither a census study which covered all published content, nor a representative corpus (how would I decide on inclusions and criteria?), the most recent issues would perhaps indicate what research was being reported now.
As noted, this boundary setting establishes some things I cannot say. I cannot for instance argue that my analysis has coverage of the complete ELMA field. From a Bourdieusian perspective I can however suggest that even six issues of six journals might say something about the game afoot. Looking at the different methods used by particular journal communities, as well as indicating what might be common across them, might reveal some further questions and steps to be taken.
I undertook a semiotic content analysis (Manning and Cullum-Swan 1998) of the corpus. I read each paper, asking first of all what kind of paper it was. Was it an empirical study, a think piece, a state of the art commentary, a report of practice or a problematisation of field thinking (categories from Petrie and Rugg 2011, Alvesson and Sandberg 2013)? I then asked what kind of research design was used and what methods were deployed. I also looked for time-space dimensionsthe duration of data collection and its locationas well as the scale of the data, the type of data generated and modes of analysis. I noted what conceptual or theoretical resources were used. I read editorials as well as documenting any specific paper 'types' allowed in the publication.
The resulting information was collated in table form with one table per journal. This format allowed me to count methods types across all of the 208 papers, as well as to see some common threads within a journal (a position) and points of comparison between them. I then examined this count through a Bourdieusian lens, generating a partial view of methods practices in the ELMA field.
I now turn to the results of that analysis.
A first look at the snapshot of ELMA methods As my original word count had suggested, there was indeed a preponderance of interview and survey methods used in these papers. But there was also a large number of case studies, many of which used either interviews, or surveys, or both, as part of their data generation process. However, these methods were not evenly distributed across journals (see Figure 2.) While these data confirm the initial word search I had conducted, they also complicate the picture. In all, 13.5% of the corpus were studies based only on interviews, 14% reported only surveys and 4.8% reported on the basis of survey and interview. Case studies were 20.7% and they used multiple methods, often adding in document analysis and observation. It was these latter tools that provided some of the variation in case studyfor example, documents included school plans and organisation charts, mission statements, standards documents, media, instructional materials, policy texts, strategies maps and online material.
Overall, 26% of the total corpus used some form of quantitative approachsurveys, effects studies, model testing and secondary data analysis. But there were very few effects and model testing studies across the journals -2 in EAQ, 3 in SLAM and 13 in ELMA, the latter having the highest percentage of this type of research, with 25% of its total papers. The count of quantitative research may mean either that the field is biased towards the interpretive, or that researchers concerned with models and effects publish in specialist journals as well as the more general high-status educational research journals owned by national/international learned societiesor both. Here is one of the disadvantages of the snapshotit is not a comprehensive corpus. As noted, it is important to see the possible interpretations of these percentages as raising questions, rather than providing answers. It is equally important to acknowledge that qualitative data can also be dealt with very differently. While all of the interviews in JEAH had been subject to some kind of critical discourse analysis which connected individual statements to wider policy and social concerns, this analytic approach was used rarely in the other journals. A more general emic approachtaking an 'internal' perspective generated through thematising the data, occasionally combined with etic, 'external', literatures-based approacheswas the norm in the remaining five journals. The emic qualitative approach contrasts in particular with the effects and models based studies, where etic approaches are necessary in order to generate statistically testable propositions. The use of pre-existing literatures is one of the critiques made of this approachit is additive, rather than breaking new ground.

Differences between journals
There were also some distinctive methods used in particular journals: . IJLE offers writers the option of contributing opinion pieces and descriptions of practice. During the last twelve months it has also devoted a special issue to discussing the general state of education and education research. This led to six papers that could be seen as 'state of the art' or 'think pieces'. LPS had one such paper, and JEAH three. . EAQ published five papers which used secondary data analysis of large North American state-developed/managed longitudinal data basesfor example, labour market statistics and surveys. LPS also had two papers which also used state-generated longitudinal data. These seven papers mobilised a range of statistical approaches, including those drawn from economics and demography. . As its name indicates, JEAH had a relatively high proportion of articles that used historical methods and where archival or oral history material was the primary datasome 13, or 38% of the papers. (LPS and IJLE also had one each of historical methods). JEAH also had five papers that were philosophical in orientation, not surprising given a special issue on Hannah Arendt. This leaning suggests that JEAH is positioned nearer to journals in the Humanities; only 11 of the papers in the journal could be said to be 'pure' Social Sciencebut these 11 were located within Sociology and Politics. These disciplinary orientations stand in sharp contrast to the five other journals which were dominated by generic social science empirical approaches, that is, they largely reported on field work in schools, colleges, universities and communities. . JEAH also had the strongest orientation to theory, with Bourdieu and Foucault prominent, as well as Arendt. Marcuse was used in one paper. Bourdieu made an appearance in a paper in ELAM, and in IJLE. Nancy Fraser appeared in LPS, Critical Race Theory in LPS and EAQ and Sen, Nussbaum, Mouffe and Ranciere in IJLE, together with more general discussions of social justice. The JEAH theory use is congruent with the sociological, political and philosophical basis of the majority of non-historical papers. . All journals but JEAH had a minority of papers which were based in formal literature reviews. ELAM had 8 such papers, and SLAM 4, making 12 literatures papers in UKedited journals compared to the North American 5. This may be an accident of the snapshot or suggest a slight preference from the North American journals for empirical field and desk work, rather than library-based work.
. ELAM had by far the most contributions outside of the country of editorship. Only 26% of its contributions came from within the UK with 55% coming from countries where English is not the majority language. Some of the editorials also took a very international stance, drawing together comparative threads around a core theme.
What are we to make of this? Three things stand out. First of all, secondary data analysisor 'big data'is now increasingly advocated by UK policy-makers. The absence of secondary data analysis in the UK journals could suggest that researchers in the UK, or those who choose to publish in UK journals, are not interested in this kind of research. Or perhaps researchers in North America want to communicate the results of their secondary data analysis to their local readers. In the case of the UK it is more likely, given the nature of the policy environment, that such ELMA-oriented databases either do not exist, or that their government owners are not willing to allow researchers to access them. Some evidence about the political sensitivity of research does exist, see for instance, a recent edict that researchers must give two days' notice before they release results of research using national pupil databases (Dickens 2016). This political nerviness may also be the case in Australia and New Zealand.
Secondly, the position of disciplines and foci is significant. The different disciplinary makeup of JEAH suggests that it provides a home and conversation for scholars from disciplines that have been marginalised in UK schools of educationhistory, philosophy, sociology, politics. It also offers a place for ELMA critical policy scholarship. The strong focus in IJLE on social justice and discussions of the purposes of educational change, as well as its reports of change and improvement, suggest that IJLE provides a place where ELMA scholars concerned about the inequities resulting from current education practices can publish their concerns to a sympathetic audience. The geographical diversity in ELAM attests to its global visibility; the presence of some papers which address cross cultural and potential Western imposts of ELMA theory suggests that its writers are not unaware of this as both benefit and potential problem. However, the consistency of methods may suggest that there is still something homologous about what is sent to the journal and is selected through the reviewing process.
Thirdly, this small snapshot of journal methods resonates with Gunter's research (e.g. 2012, 2016) of knowledge production in the UK ELMA field which showed that a majority of researchers are focused on improvements in schools, colleges, universities and systems while a minority focus on more critical concerns. My snapshot adds corroborating data about the narrow range of methods used in this work.

What's marginal or missing from this ELMA journal snapshot?
Across the corpus there were very few studies which took a temporal perspective, that is, they were longitudinal in design. By their nature, all six secondary data analyses had a time dimension. In addition, there were: two projects which lasted three years and used multiple visits and data generation waves; one case study of two years, two one-year case studies and one eight-month ethnography. Other projects obviously lasted for a period of time, but the temporal dimension of data generation did not seem to be integral to the design. In some of the effects studies it was the sequencing of data generation that was important, not the change and variety over time. This result resonates with concerns by Weindling (2004) that funding available for UK ELMA research promoted short-term projects. But the lack of temporality might equally speak to the press for doctoral completion in the UK and Australia in particular; this has now made longitudinal research very difficult.
There was very little action research, practitioner research, self-study or professionally based commentary offered across the corpus. There were only four such papers, two each in SLAM and ELAM. Interestingly, three of these papers came from Nordic countries and one from Australia; in both instances there are long histories of action research used in school change programmes. However, the practice-based papers in IJLE also explicitly used professional knowledge as a resource. The absence of practitioner research is a surprising omission, given the number of practicing school leaders undertaking doctorates, particularly professional doctorates which are often based in their schools. The paucity of this kind of research points to the difficulty that practitioners might have in finding time to publish their research, or it might signal that they see little reason to write for scholarly journals. Either way, the absence of practitioner voices signals to the ELMA field as a whole a potential issue worth discussing.
There were some outlying methods that appeared in this snapshot. Papers in IJLE had the most methods variety. There was research that used collage to generate conversation, life history, self-study, network analysis and use of artefact to stimulate discussion. In JEAH, one of the authors used biographical portraiture as an historical method, and three authors used network analysis. ELAM published a paper which used approaches based in experimental psychology while SLAM had one writer whose interview analysis focused on metaphor. The presence of such papers shows that these journals are not averse to research which uses methods other than the mainstream, and perhaps that the domination of particular approaches reveals field-produced writer and reviewer strategiesthese agents see the journal and ELMA research as embodying a particular method.
There were also methods that I expected to see more. I counted only one ethnography and only a handful of papers which referred to anything online, let alone used online methods. Narrative approaches were not obvious to me, with only one mention of the term and nothing which took 'stories' or 'memories' as a primary method. I wondered whether this was indicative of a method that is currently out of favour in the field or simply an accident of the partiality of the snapshot. One paper used interviews to generate critical incidents which were then used as a basis for role play and discussion (in ELAM). As far as I could see there was no systematic shadowing a la Wolcott (1973), perhaps a 'classic' ELAM method. Surprisingly there was nothing that had media text analysis as its focus, or analysis of any other kinds of cultural texts (films, advertisements and so on), although there has been some of this work in the past. There was also almost no visual researchone paper which used a visual/creative methodcollage as elicitation (this was based on work by Gauntlett 2011)and one paper which used images as part of its description of a community (the focus of the paper). Given the increasingly interest in visual research and in online research in the wider educational research field, this was surprising.
The newer theory-driven methods derived from for example Actor Network Theory (ANT) (Latour 2007) or Practice Architectures (PA) (Wilkinson and Kemmis 2015), were also missing. Like earlier Bourdieusian and Foucauldian approaches, ANT and PA de-centre the individual leader and the individual school and focus instead on investigating socially distributed practices and relationships. A possible explanation for the absence of post-humanist and meshwork approaches might be found in the workings of the wider education field. An ongoing focus on 'the leader' and 'the school/college/university' sits neatly with current policy emphases on 'autonomy' and self-management, key doxa in an educational field governed through markets, audits and new constellations of actors (academies, free schools, consultants, philanthropists, edu-businesses and the like) (Ball 2012, Gunter 2016. And there was nothing in the corpus that came from what might be seen as the 'wilder side' of social methods. Social science and educational researchers are now experimenting with, for example, creative approaches (Kara 2015), inventive methods (Lury and Wakeford 2012), place-based approaches (Thomson and Hall 2017), post-human approaches (Tayor and Hughes 2016), mobile methods (Buscher et al. 2010), collaborative research with artists and arts practices (Barone and Eisner 2011), psycho-geographic practices (Richardson 2015), participatory approaches (Hacker 2013) and wardrobe studies (Weber and Mitchell 2004). While some of these approaches might sound a little off the wall at first hearing, and they are probably contributing to a 'method-isation' of the field (of which this paper must be seen as part), they may nevertheless have some promise for ELMA. The field will not know this unless agents are prepared/positioned to experiment.
The marginalisation of particular methods, and non-use of others, might suggest an inherent conservatism in the field, or a preference for mainstream methods which will be readily accepted not only by other scholars, but also practitioners and policy-makers there are certainly pressures from the political field to produce research directed to 'what works'. Perhaps the core set of methods are a game that works for most of the players in the ELMA field. Or perhaps the narrow range of methods is a product of the wider education field. Or all of these.
The lacunae might be connected with the kinds of research methods training in ELMA: methods training at doctoral level in the UK is typically offered by people trained in mainstream social science, with supervisors expected to offer the specialised support required for particular projects. While there are advanced methods resources available online and often within institutions, their timing is often past the point where a research design has been finalised. It also seems plausible that doctoral candidates and early career researchers, as they become more and more immersed in the ELMA literatures, simply see the dominance of particular methods as the 'right way' to do their research. The game in the field is interview, case study, survey, document analysis and observationthese become doxic, The Way to do ELMA research.
On the other hand, the narrow range of methods might be the glue holding together the tribes and territories that sit within the ELMA field, the big tent with small huddles around discrete campfires, as Donmoyer (1995) once described it.
A further conclusion might be that the ELMA field does not yet focus sufficiently on research methods which might lead to new ways of thinking, seeing and saying. Perhaps we have not yet imagined that the ELMA field might not only be open to critical studies, but also its own methods avant-garde?
A second look at the snapshot of ELMA methods My analysis of the journals also included looking at the ways in which methods were discussed and presented. Without exception, all of the empirical field work papers presented a specific section, usually after a discussion of literatures, which described their research design and particular methods. At no point in any of the papers was there a substantive discussion about what the methods could and could not do. On the odd occasion a note was made about methods 'limitations' in the conclusion, but this was the exception rather than the rule. There was no section which equated to my discussion of the 'snapshot' earlier in this paper. The dominant genre was to present the research design and methods as if they were transparent, obviously appropriate and relatively problem free.
On the other hand, the papers that were literature based did not discuss their method of selection or analysis at allthe systematic review was the exception. The historical papers were no different, usually proceeding straight into an interpretive account, with little discussion of sources, historiographical tradition, or any issues there may have been in the conduct of the research. And there were no methodological papers per se in the corpus, although a model testing paper could perhaps be interpreted to be such. There was thus far less detail about method in JEAH than in other journals.
The lack of methodological papers sits alongside the small number of methods texts specifically written for the ELMA field. Perhaps the field simply uses general educational research or social science texts. But which ones? There are within educational research, and in the social sciences more generally, a range of discussions and debates about methods which did not feature in the snapshot. The educational and social science fields offer more than general compendia of tools. For instance, . critical and feminist methods literatures (e.g. Alvesson and Deetz 2000, Griffiths 2003, Lather 2007) have long argued for deconstruction. Now also non-scientised approaches such as dreams, sensual and haptic knowledges, and fragments of knowing, find a place in and as research method (e.g. Pink 2009, St. Pierre and Jackson 2014). . as post-colonial scholars made the connections between the imperial mind-set and the research enterprise (Smith 1999), decolonising methodologies pointed to the value of narrative, poetic and metaphorical ways of knowing (Bishop and Glynn 1999). . creative researchers, arts based researchers, and arts practice as research have not only challenged the epistemological and methodological basis of research, but also its primary ways of knowingits methods (Leavy 2009, Back and Puwar 2013, Nelson 2013).
Some of this development was reflected in the Briggs, Coleman and Morrison ELMA methods text, but not in my snapshot.
There is also an emerging discussion in the wider social science field about 'the social life of methods' (Law et al. 2011)thinking about and investigation as to how methods actively shape what is known and what it is possible to know. Acknowledging the social or political life of data means actively rejecting the notion that research can be divided into three spherestheory, substance and methodin which method is simply a tool for finding out about the world. (I note in passing that this challenges some assumptions made in this paper.) The isolation of method from theory and substance leads, Law, Ruppert and Savage argue, to an unhelpful oscillation between instrumentalist discussions of limitations of methods and humanist concerns about what is beyond particular methods to see. Rather, they suggest, while it is not wrong to think about technique, it is crucial to consider the 'double' social life of methodthe ways in which methods have been socially produced, and the ways in which they produce and reproduce particular understandings of the world.
According to Law, Ruppert and Savage, methods … make discoveries about the world, and … those discoveries may surprise us. That's why we conduct interviews and surveys and all the rest. But also, and counterintuitively, we're saying that they also make more or less self-fulfilling assumptions about the character of the social world. And that in so doing they tend to constitute it, so to speak, below the radar in ways that we scarcely notice. In short, that they tend to produce what John Law calls collateral realities: that is, realities that we don't think about very much but that we're all busy reproducing as we go about the daily methodological work of gathering and analysing data 'about' the social. (p. 11) Methods are productive of subjectivities and social realities. They construct particular views of the world. The circulation of methods and their problems, interests, purposes, advocates, arrangements, representations and circuits must therefore be subject to critical scrutiny. It is vital, Law, Ruppert and Savage assert, that social scientists rethink knowledges, realities and methods together. It is crucial to put the social back into method, and to treat method as a social phenomenon in its own right. This kind of methods conversation seems to have little traction in the ELMA field. This perhaps suggests that ELMA agents are positioned some distance away from this part of the wider social science community and field. The lack of ELMA engagement with this conversation speaks to a general field focus on institutions and policy, rather than the conduct of knowledge production, the purpose of the field. The absence of engagement in wider social science methods conversations may also say something about the kind of border maintenance that is practiced in the ELMA field, as ELMA agents promote ELMA courses, conferences and journals as the strategy to consolidate and advance its position within higher education disciplines (Ladwig 1996). And as noted, the absence of this kind of discussion may also reflect an homology between knowledge producers and educational leadersperhaps there is a methods conservatism in the ELMA field equivalent to the suit and tie required of most designated professional 'leaders'.
However, there are germane questions that might arise from taking part in this wider social science debate. For instance, we might ask what kinds of educational subjectivities our methods assume. Do we assume in doing interviews with school leaders that they are simply telling us how it is, or do we countenance the possibility that our very semi-structured questions might contribute to shaping the ways in which they understand themselves, their practice and purposes? Can we think that when we develop categories for surveys from the literature that we may be actively inventing and bringing into being the very category we are proposing? 'It' may not exist out there at all, but be constructed in the field in part through our work of scholarly naming and framing (Hacking 1999(Hacking , 2006. In the same way that maps, flags and census data are part of the imaginary and social construction of nations (Castoriadis 1987) perhaps the ELMA field is complicit in constructing 'leadership' and 'leaders', 'management' and 'managers' and 'administration' and 'administrators', alongside policy-makers? What if the very terms at the heart of the field were a fundamental misrecognition of the social reality of education writ large, and were simply a way for a new scholarly field to establish itself and gain distinction (c.f. Bourdieu 1988)? A very discomforting idea indeed.
This possibility should surely give the ELMA field some pause for thought.

In sum
My snapshot of methods literature used in current papers in six journals in the ELMA field suggests that a limited range of methods predominate. As a result of this analysis, I do not want to simply suggest that further work is needed because a larger corpus of papers would lead to better analysisthat is certainly and unarguably the case. Nor do I want to argue for more ethnography, more longitudinal studies, more work at scale, or more experimental methodsall of which are possible implications of the analysis. Rather, I want to propose that the field simply needs to become more actively interested and engaged with questions of methods. My snapshot suggests, even more than the dominance of particular methods, that the field largely takes its tools for granted. This seems problematic to me, for reasons that I hope to have made clear through the questions I have raised in the paper. My conclusion from the snapshot is that ELMA needs to get a better grip on its tools. Just be more methods engaged.
And my provocation in and to the ELMA field is that, in order to develop a serious conversation about methods, there are benefits to be gained from some encounters with more 'far out' approaches and debates in the field. A walk on the wild side of methodsthose that are almost impossible to systematise and which relish some degree of ineffabilitywould help the ELMA field to ask questions that challenge its/our taken-for-granted practices.
It would certainly be possible to arrange special issues, symposia and conferences around questions of methods. But do we, does the ELMA field, have the resolve to put the tools that construct our social production and reproduction to the test? Can we, will we, move beyond seeing methods as simply tools that we choose, to understanding them as being social produced and (re)productive?