Working with Qualitative Data

William Gibson, Andrew Brown

• Publisher: SAGE Publications Ltd
• Year: 2009
• Print ISBN: 9781412945721, 9781412945714
• eBook ISBN: 9781446246825
• Edition: 1
• Topic: Scienze sociali
• Subtopic: Metodologia e ricerca nelle scienze sociali

Working with Qualitative Data provides a practical and accessible introduction to how to develop and apply strategies for the analysis of qualitative data by exploring the ways in which analysis is related to all aspects of research.

By situating analysis in the context of the whole research process, this book helps the reader to introduce an analytical component to every stage of doing research: from designing a project, reviewing the literature, through the various stages of gathering data, to the process of writing-up. Through practical examples the book maps out strategies for developing analytic frameworks in relation to all aspects of research, and demonstrates the ways in which such frameworks can be used in relation to various sorts of data.

In contrast to existing qualitative data analysis texts, this book offers a unified approach to the process of analysis within qualitative research. It will be of great use to students and researchers across the full range of social, health and education sciences.

---

1	Introduction: qualitative data analysis in context

---

---

This chapter discusses the following issues:

• The notion of analysis in qualitative research work

• The nature of qualitative enquiry

• Theory and qualitative enquiry

---

Introduction

The ‘success’ of a research project is very much contingent on the analysis of data: on working with data to achieve something interesting and perhaps even important in relation to the substantive focus of a research project; on successfully relating such findings to an academic or professional field; on being able to say something through engagement with the data and using it to reflect not just on the particular setting being explored, but ideally, to create some generalizable or at least ‘generally interesting’ finding or idea that can be taken forward in other contexts.

In spite of its importance, the analysis of data remains one of the most difficult aspects of social research to discuss. There is something very nebulous about analysis, which somehow seems to evade tight description. Where very detailed descriptions of analysis are given, they tend to be offered in relation to a particular example of analysis – i.e. in relation to some problem or context – or in terms of a particular approach to doing analysis, like grounded theory, or narrative analysis, or phenomenological description. Such very specific accounts of analytic work can be alienating for researchers, who can find it hard to relate their interests to working contexts that are very different from their own, or to ways of doing research with which they are unfamiliar. The problem here is not that exemplifying analysis or showing how particular approaches work is not helpful; they most certainly are. The difficulty is that doing so is not sufficient, as in their specificity such descriptions may not demonstrate more generally how analysis in different contexts, with different kinds of data, and drawing on different conceptual languages might proceed. Researchers, particularly new and inexperienced researchers, often want clear guidance on how to work with data, but the complex relation between analysis and context, research topics, theory, the everyday contingencies of doing research, the dispositions of the researcher and so on, mean that analysis resists prescriptive codification, which makes the provision of clear and generalizable guidelines hard to provide.

This book is about the ways in which data analysis relates to, impacts on and develops from the other aspects of social research practice; it is about analysis and data work as a feature of qualitative social research, and the intersection of research problems, specific approaches to social research and research data. We do not prescribe a mechanism or template for doing data analysis. Rather, we want to consider the ways that the work that people do with data relates to the other components of social research work. We want to encourage an approach to analysis that is not just about techniques for dealing with data, but is also about thinking through the relation between a particular research setting and problem, and the literary and theoretical context of research. Through this approach, we hope to provide a nuanced picture of the relationship between analysis and social research practice in general.

In this book we will be discussing particular approaches to data analysis, and working through some of the key issues related to data work. We do this not, we hope, in a dogmatic way, but as a means of showing how analysis can work when particular strategies and foci are adopted. We have attempted to address all phases in the research process, from the development of a question or research focus through to the writing and presentation of research. In each phase we have emphasized the processes of working with data, and more specifically an analytic engagement with data.

In addition, we have looked at specific forms of data, such as documents, interviews, observations, video and audio data, and explored some of the general strategies and concerns running through the processes of qualitative data analysis, such as transcription and representation of data and the identification of themes in data. In all cases, we have sought to present and discuss data work and the process of analysis in the context of specific approaches to research or specific projects.

But we are getting ahead of ourselves… Let’s start by thinking a bit more closely about the notion of analysis in relation to qualitative research, as it is from this that our thesis will begin to take a little more definite form.

What is qualitative data ‘analysis’ anyway?

Data analysis is an aspect of research practice that seems to create significant confusion for those new to, or working outside, qualitative research paradigms. Most areas of research work are quite intuitively grasped – generally speaking, people seem to have little trouble imagining what a literature review might involve, or what research design or writing-up are, and data collection is usually quite unproblematically understood. That is not to say that there is nothing complicated about any of these things, or that people are always right in their assumptions, but at least the general purposes of those activities, and the kinds of things that researchers might get involved in when they engage in them can be understood to some extent, or at the very least, they can be guessed at with some degree of accuracy. Very often, the issue of analysis seems to be quite different, and is seen to be rather mysterious to students, not only in terms of the practices that allegedly comprise it, but also in terms of the general aims behind it.

In the contexts of more quantitative forms of work, analysis is a little easier to conceptualize. We can point to the ways that different statistical tests work, and to some of the mechanisms for organizing data so that those tests can be performed, and that often seems to satisfy as some kind of explanation for what analysis involves. The notion that analysis will produce an explanation of the relationship between variables is also usually regarded as giving some idea as to the purposes of such analysis. In qualitative analysis, though, things are much more murky, and there are few tangible practices that can be discussed as features of work that ‘constitute’ analysis. It is also often unclear, it seems, what the purposes of analysis are and what the outcomes ought to look like. It is not uncommon for students to express the idea that there is some kind of secret that they haven’t been let in on in relation to qualitative analysis – some set of tricks or ways of working that they haven’t yet been told about.

In this brief section we would like to work through the notion of analysis in relation to qualitative research as a means of creating some kind of response to this general lack of clarity. We will start by thinking about the general usage of the term ‘analysis’. As we note in Chapter 2, many of the terms in social research have some counterpart meaning in non-research discourse, and it may therefore be useful to explore this meaning in order to create a more specific meaning that relates to qualitative social research. The New Oxford English Dictionary defines analysis as follows:

Detailed examination of the elements or structure of something, typically as the basis for discussion or interpretation.

Here, the emphasis is on the exploration of the ‘structure’ of ‘things’. Clearly, what ‘elements’ or ‘structure’ might mean depend on what the ‘something’ refers to; there is nothing specific here whatsoever as an account of what analysis is. The context in which the term ‘analysis’ is used and the ‘things’ to which it is directed are crucial to understanding what analysis might refer to. All we get from this definition is something about the examination of structure. Does this idea give us much purchase on the work of social researchers? A researcher may look at the structure of an opinion, of consciousness, of personnel in an organization, of a legal process, of communication, of experiences, of attitudes, of stories, of pictures, and so on. A part of examining structure might involve trying to understand how that structure works. This could entail explicating the constitutive components, looking at the roles of those various components, or examining the relationship between them. It might also call for some element of evaluation of the components, which could be in simplistic ‘good’/‘bad’ or ‘effective’/‘non-effective’ terms, but might be in a more complex and exploratory way. It might, though, be more straightforward and simply involve a description of those structural elements.

While this definition throws up some ideas, there is nothing tangible here – nothing that we can point to and say ‘that is what you do when you do analysis’. How does one do this ‘examination of the elements of a structure’, and what do we mean by ‘structure’ anyway? The problem here is, again, the absence of an understanding of a context in which analysis operates or an issue to which analysis is directed. But that is quite a useful step: we can begin to see that particular context and issues are key for gaining a sense of what analysis means.

So what about social researchers? How do they define this notion of analysis in relation to their work with data? Does analysis take on a more certain and definitive shape when used in this domain? Marshall and Rossman define qualitative data analysis in the following way:

Qualitative data analysis is a search for general statements about relationships and underlying themes. (2006:154)

The reference to relationships and themes here implicates an interest in structure, as in the previous more generic definition of analysis. In Marshall and Rossman’s view, analysis involves using generalized themes to look at the relationships between components of a data set. Indeed, this kind of thematized comparative work is at the heart of a number of distinct approaches to qualitative data work (see, for example, Miles and Huberman, 1994; Glaser and Strauss, 1999[1967]; Boyatzis, 2008). Now, there are some techniques and procedures that we can point to here. We can describe the ways that codes can be used to categorize data, and the types of operation that researchers might perform in order to interrogate the relationships between their codes. We can discuss the difference between codes that are created prior to the analysis of data, and those that are created from data. We might also think about the ways that computers can be used as a means of facilitating such work. All of this is important, and we will deal with these matters in some detail (particularly in Chapters 8 and 11).

However, although this gives us some idea of what analysis might entail, there is a real problem with thinking of this as constituting analysis. To begin with, not all researchers think about analysis in these kinds of ways; this kind of ‘thematized analysis’, as we describe it, is not, for example, a good way to think about how conversation analysis or critical discourse analysis works. The limitation here is that while it may be broadly appropriate to describe some of what people in these areas do as being concerned with comparing data through themes, this description doesn’t tell you much about the nature of the interests that drive the enquiry. Concentrating on the processes of generating a theme, in these quite procedural ways, doesn’t explain why the theme is of interest in the first place. This problem is not just limited to disciplines like conversation analysis and critical discourse analysis, though, but is a much more general issue. Analysis is always about something or of something, and the thing that it is ‘about’ or ‘of’ is fundamental for understanding how that analysis works. In other words, thinking about analysis in a decontextualized and ‘general’ way and about ‘procedures’ to analysis does not really solve the problem of how to explain how analysis works or what it is all about.

Let’s look at another definition, this one from Harry Wolcott:

…analysis refers quite specifically and narrowly to systematic procedures followed in order to identify essential features and relationships… (1994:24).

This definition comes from a distinction Wolcott makes between ‘description’, ‘analysis’ and ‘interpretation’, which represent three components of qualitative work. Wolcott does not suggest that these are clear and mutually exclusive categories, but merely that it can be useful to make a distinction between them. Description involves producing an account that stays close to the original data. The general aim in producing descriptions is to create a narrative that presents the original data in a motivated way (i.e. that operates as a description for a particular purpose). Analysis involves going beyond these largely descriptive iterations and systematically producing an account of ‘key factors and relationships among them’ (Wolcott, 1994: 10). Again, we see some similarity with the previous discussion of themes and generalized statements here. Finally, interpretation involves trying to give sense to the data by creatively producing insights about it. A crucial difference between analysis and interpretation as used by Wolcott is that the former is constrained and conservative, and is bound by the data, while the latter is inventive and creative and less empirically cautious (Wolcott, 1994: 23).

Wolcott describes the relationship between these three elements of qualitative work through the analogy of a see-saw or ‘teeter-totter’. Description is the central part of the balance, and analysis and interpretation are the two opposite poles of the stem that balance on it. Researchers rest their analysis and interpretation (as defined above) on their description, and can give more or less emphasis to one or the other by raising or lowering one or other side of the see-saw. Wolcott’s description, and the ...

Table of contents

• Cover Page
• Title Page
• Copyright
• Contents
• 1 Introduction: qualitative data analysis in context
• 2 Theory, grounded theory and analysis
• 3 Engaging with literature
• 4 Research design
• 5 Using documents in research
• 6 Generating data through questions and observations
• 7 Transcribing and representing data
• 8 Identifying themes, codes and hypotheses
• 9 Images and texts
• 10 Video and audio data
• 11 Using technology
• 12 Writing and presenting analysis
• 13 Concluding remarks
• References
• Index