The focus of this chapter is methods, which I take to be the processes of generating data; the relationships that were central to this; my position(s) within the field; the ethics of fieldwork; the particularities of researching with multiple groups within one setting, including children and young people; and the analytical protocols followed. It explains how the analyses presented were carried out, and the ethical and practical issues that surfaced during this process. Section one documents the process of undertaking Foucauldian Discourse Analysis (FDA) of government-produced texts on academy schools. Section Two considers the overlapping activities I engaged in as part of a post-structuralist ethnography of Eastbank Academy. In both cases I consider what these research activities offered and how the resulting data was selected, analysed and interpreted. In section three, I reflect on the ethical and interpersonal dimensions of this research and their impact on the resulting data.
Section One: Foucauldian-inspired Discourse analysis
Discourse analysis that takes its lead from Foucault’s work must contend with the lack of a clearly defined methodology. Foucault’s work provides tools for thinking about and questioning phenomena, rather than a strict methodological protocol (Kendall & Wickham, 1999). Indeed, a strict protocol would be distinctly non-Foucauldian, since it would operate to construct a truth about analysis (Hewitt, 2009). Instead, Foucault’s work or
“anti-method” (Grimaldi, 2012: 446) has been understood through the metaphor of the toolbox (Ball, 2013), and he encouraged researchers to apply his tools to their particular questions (Gutting, 2005). I use Foucault’s work as a guide to method (Given, 2008), as well as drawing on other studies that use Foucault’s work (Bacchi, 1999; Bailey, 2009; Ball, 2013; Hewitt, 2009).
Such flexibility in method does not sit well with the imperatives of contemporary research agendas in the UK context (Ball, 2009b). The notion of a strict operational protocol continues to be aligned with contested, but still dominant, notions of quality in social science research (Torrance, 2014).
Such protocols are depicted as being particularly attuned to making research replicable, but also attend to the demand for neutral or objective research that is useful to policy makers (May, 1997). One possible argument here is that by following a clear methodological protocol, researchers can show that they have not been ‘swayed’ in a particular direction, but have simply stuck to a pre-outlined method, and are reporting what this has generated accordingly. As Chapter Three discussed, post-structuralist approaches are critical of such arguments, which attempt to diminish the influence of the researcher. Foucault’s work offers a lens for critiquing positivistic privileging of rationality and objectivity, which are instead positioned as master discourses of truth, which must also be problematised. As Butler (1990) argues, the demand for clarity, for instance in methodological protocols, must itself be questioned about the messy realities it obscures. Analysing the production of the academy school requires tools that are capable of making sense of “an imprecise, fuzzy, woolly reality” for which we require concepts that are “polymorphic and adaptable, rather than defined, calibrated and used rigidly” (Bourdieu & Wacquant, 1992: 23).
Research is necessarily intertwined with the particular theories being put to use, and the interpretations and judgements that accompany the entire research process (Law, 2004). The mark of the researcher is never absent from this, although it may be muted in the writing of research. Rather than aspiring to be a disinterested researcher (Jones, 2014), who is nonetheless present in every decision, I make explicit my processes of interpretation and judgement. It is a task I began at the beginning of Chapter Three, were I clarified how I had arrived at important decisions in the project. My methodological protocol has not been rigid or immovable, or a way of making claim to objectivity. Instead it is a framework that prompts reflection
and invites the documenting of shifts, developments and interpretations during analysis. The aim is to enable readers to make informed judgements about the nature of the research, the use of theory to illuminate the academy school, and the resultant interpretations.
The methodological choices made were driven by the problem of interest – the academisation of ‘failing schools’ - and through the concepts and definitions being worked with. In Chapter Three I detailed the theoretical underpinnings this work draws on to inform the aims and boundaries of analysis. I use Foucault’s definition of discourse, as the rules that govern what it is possible to say, write, think and know about a particular phenomenon at a particular point in history. His analytical approach aimed to reveal the rules of operation of discourse, attending to what is said and present, but also what is not said, what is forbidden or what is relegated to the shadows of discourse (Foucault, 1969).
I particularly draw on Foucault’s later genealogical works, since it is here that power relations are more explicitly dealt with (Hewitt, 2009; McNay, 1994). My work does not provide the historical detail that would be required for a genealogy, which would constitute a study on its own without the considerable ethnographic data I am also drawing on. Instead I adopt the Foucauldian approach to discourse to interrogate what has been said and written about academies across their lifespan. Discourse analysis is a research method that involves examining communication (Hewitt, 2009). In Foucault’s studies it relies on the close analysis of texts to explore patterns and rules of how language is used and narratives are constructed. I used his work as a guide to formulating the questions I asked of texts, which shaped the way texts were filtered and connected with one another, as well as the interpretations that ensued.
Foucault’s method facilitates an analysis of “how things have come to be the way they are, how it is that they remain that way, and how else they might have been or could be” (Given, 2008: 355). It guides an analysis of the
relationships and order that underpin discursive ‘truths’, and their relationship with wider discourses and operations of power (Hewitt, 2009). I apply this understanding of discourse to the task of addressing the following question in Chapter Five: How have academy status and the academy school been produced and shaped as objects for thought through discourse? The influence of Foucault’s theory is apparent here, since this is different to asking ‘what are academies’, and instead seeks to understand the shaping of academies through language, without aiming to assess the accuracy of these representations.
Through analysis the following set of sub-questions emerged and were refined, which were used to clarify subsequent analysis and writing:
• How are academies made compelling?
• What representations have come to be associated with academies and what do these perpetuate, enable, and constrain?
• How are these representations sustained and why have they been possible at this time?
Method
The ‘decision trail’ in this work clarifies and draws together analytical method, theory, questions and sampling. The analysis of discourse began with the literature review when a broad sweep of literature was first encountered and a sense of the dominant themes, contentions and representations emerged. These initial readings and understandings prompted analysis. It was during this phase of the research that I became aware of the repetitive presentation of academies as transformative, which became central to analysis in Chapter Five.
I then read the texts more thoroughly, in light of the literature review, and initial ideas and perceptions were trialled more systematically. I experimented with narrative theory as an initial framework to guide analysis
(Hewitt, 2009: 10). At this point the emphasis moved to government-produced texts as I began to realise the significance of discourse analysis as a stage of analysis in its own right. This second read supported the selection of texts for close analysis and coding.
During stages one and two a list of codes was created and refined. Some of these codes related to the types of statements being made about what academies are, and what they are expected to achieve. Others related to my emerging sense of an academy narrative, through which ‘truths’ about academies were created and perpetuated. Drawing on narrative theory, I investigated the extent to which narrative concepts such as ‘character’,
‘narrator’, ‘metaphor’ and ‘storyline’ were illuminating. I used the qualitative data package NVIVO to store and manage data (Gibson, 2010). I did not use any of the wider theory-building functions of NVIVO, and coding remained researcher driven. However, coding is problematic, since it can be positioned through positivist ideals of sorting, counting and organising data in such a way that “themes ‘emerge’ as if data and the interpreter are not always already theory-laden” (Adams St.Pierre & Roulston, 2006: 677). Instead, the view taken was that themes are shaped through my reading of theory, experiences, characteristics and aims, which I discuss in the final section of this chapter.
Coding continued to be refined through subsequent readings of texts, and through a consideration of the context of production for each text. Through this some texts emerged as key moments in the bid to establish academies as the unequivocal future of education in England, or as clearly illuminating the use of a particular narrative technique. This led to decisions about which texts to include in the analysis and which to draw on as examples in writing (see below). The emergence of narrative suggested the importance of scanning a wider set of texts to pick up on repeated and nuanced aspects of this.
By setting the analytical process out as a list of stages I am fashioning a level of clarity and linearity that was not present during the process, in the interests of readability. This masks the iterative nature of analysis, through which methodological protocol, theory, questions and sampling informed one another. Below I clarify particular issues relating to sampling, which should be envisaged as happening in tandem with one another and with the stages outlined above.
Text Selection
Some of the parameters of text selection were more obvious than others, because they stemmed more straightforwardly from the nature of the phenomena of interest. This project is an analysis of the academies policy, which was first mentioned by David Blunkett in 2000, and which remains an education policy at the time of writing. The timespan of the policy is therefore straightforward in one sense, although as noted in Chapter Two, the ancestors of the policy can be traced much further back, and to other national contexts. More specifically, this thesis is a close analysis of the strand of the academies policy that concerns ‘failing’ schools being turned into academies in order to improve. Thus texts that say something about these schools were the sampling pool. Since policy ideas do not “have a single starting point” but are “the product of the blending and clashing of other ideas, the origins of which are, in many cases, lost in time“ I am obliged to select a starting point whilst recognising that others would have been possible (Ward et al, 2016: 47).
I wanted to understand how successive governments have constructed a particular set of representations and arguments for action around schools in challenging contexts. I therefore prioritised the analysis of texts that have been produced by governments since 2000. These may be considered as dominant policy discourses, although they are certainly not the only available discourses, for instance an analysis of counter and critical academy discourses would also be possible. The discursive outputs of key political figures in the academy movement include a range of text types, including
written texts such as legislation, policy documents, opinion pieces, blog posts, books, and spoken texts (which have been turned into written texts) such as speeches and interviews. In line with Foucault’s work, the emphasis is less on ‘who’ speaks and more on what is spoken about academies, the positions it is spoken from, and “how this is mediated by the speaking positions of others; an architecture of policy positions” (Gale, 2001: 389).
Some sampling decisions were more difficult. My analysis draws on a wide body of texts, but conveys the points of this analysis by directly referring to only a fraction of these texts. This is common in discourse analysis, and qualitative researchers are always faced with important decisions about what will and will not be directly represented in written outputs. These are decisions to be wary of and to trouble (Butler, 1994). Two distinct sampling decisions emerge here: how to sample texts for analysis and how to select texts to develop understandings through writing?
First, the process of selecting texts for analysis is necessarily fraught because the limits of a ‘discourse’ are difficult to distinguish. One of the arguments made in Chapter Five is that the compelling nature of the academies programme has been produced, in part, by the way it meshes with other policy narratives that are flourishing. This is partly about the status of schools as a key institution for the production and reproduction of discourses, and as having a role within wider social and public policy spheres.
This makes it difficult to delineate the boundaries of an academy discourse.
As Foucault observes of a book, and we might observe of policy discourse:
The frontier of a book are never clear-cut…it is caught up in a system of references to other books, other texts, other sentences: it is a node within a network…it indicates itself, constructs itself, only on the basis of a complex field of discourse (Foucault, 1969: 25-6).
In addition to the focus on government-produced texts about schools in challenging contexts that are turned into academies in order to improve,
other theoretical parameters supported the selection of texts in this analysis.
I located these texts within a wider set of linguistic artefacts about academies. These say something about high achieving schools that become academies. Texts about academies are then situated within a wider contemporary literature about inequality, poverty and austerity. These have been purposively selected to highlight important stories that are being told about schools, young people and communities experiencing deprivation.
There are a number of repeated techniques and tropes across these texts that are indicative of the particular kind of work being attempted through the presentation of academies. The actual work that this does is the focus of analysis in Chapters Seven-Nine.
Second, the texts cited in Chapter Five are purposive and illustrative. They are selected to exemplify overarching points from the body of government texts that shape academy schools. Analysis hones in on particular examples to highlight wider discursive and representational patterns across texts.
Chapter Five focuses on identifying ‘truths’ about academies, how they are given coherence, and how they are maintained through their relationship with a wider social policy agenda. In Appendix One I document the texts that informed analysis, which have not all been directly cited, but which have each been influential in building up my sense of the ‘truths’ at work here. Documenting the texts that have been analysed was also a useful method of data management, of providing a chronology of texts and of spotting and interrogating any gaps.
Section Two: Ethnography Sampling
I selected one school to study and sampling was purposive and theoretically guided (Mason, 2002). Given my interest in the positioning of academy status as a tool to improve and transform schools in poorer communities, it was crucial that the research took place in a school that: served a community experiencing multiple deprivation, according to national measures; was
deemed to be ‘underperforming’; and became an academy as one possible measure for ‘improvement’. I therefore sampled a school that was, at least superficially, an environment where these issues would be important (Mason, 2002).
I drew on national data and categorisations to inform sampling. Throughout analysis and writing I have remained critical of the way these categorisations operate. Yet given the supremacy of key performance measures in constructions of success and failure, it is likely that a school’s position in relation to these measures will relate, albeit in complex ways, to its policies and practices. I was interested in seeing what meanings these labels came to have within a school, and how they related to academy status. Once in the school, I became interested in how academy status was being shaped.
Listening to the school guided the development of the project, as detailed in Chapter Three, and in this sense the specifics of the school are crucial to the way this project developed. This ethnography, like any, is a partial analysis and representation of the many possible stories that were available in the school, and of the ones I was able to capture. Sampling decisions have continued into the writing of this thesis in which I have picked one particular path through voluminous amounts of data.
Access
Ethnographic methods ask a lot of schools. Senior staff are agreeing to have a researcher spend a considerable amount of time in the school (Maguire et al, 2011). The experience of extra monitoring that comes with being a ‘failing’
school and a ‘turning around academy’ made some schools understandably wary of having yet another visitor. However, in Eastbank the very experience of being monitored and criticised by the government made the senior leaders interested in my research. That this was a school where senior staff were committed to education research and were critical of current education ideology and policy was clearly important in terms of access. Alongside this, the good rapport I appeared to have with the HOA undoubtedly affected the level of access I had and the frankness of our discussions. As in any
ethnography, a unique set of factors led to the eventual ‘case’ that became the focus of the project. The schools we see in detail may have particular characteristics at particular times, which make them more willing to be involved in research. However, these contexts are shifting, and it may be the case that other schools are more open for research at other points in time or with other researchers.
Ethnography requires continual sampling decisions, many of which are not in the control of the researcher. Gaining access to a school through senior
‘gatekeepers’ was stage one of an on-going process of renegotiating access (BSA, 2002), of asking “can I come to this occasion; can I join in this special activity; will I be able to participate in this conversation; can I sit here?”
(Gordon et al, 2005: 116). I wrote to all of the staff to explain what I was doing in the school (Appendix Two), although there continued to be misunderstandings about this (I return to this point). I did not just ‘turn up’
to lessons. I wanted to ensure that I was expected, as I remained concerned throughout that teachers thought I was vetting their capabilities. In email exchanges I emphasised that my concern was not with passing judgements on the quality of teaching, but rather to get a feel for the school, which lessons are clearly a central aspect of. Planning my visits to lessons in advance provided teachers with the opportunity to specify which lesson, and when. Despite this approach there were opportunities to see unplanned lessons, for instance when I was accompanying another member of staff who took me to a lesson without prior warning, or when the staff I developed relationships with invited me to observe a lesson on the spur of the moment.
The school gains umbrella permission from parents for its involvement in
The school gains umbrella permission from parents for its involvement in