But we have still not yet provided

But we have still not yet provided a definition of what we mean by analysis in the
context of qualitative research. This is because the contexts, problems, questions and
issues that constitute analysis are necessary parts of the definition of what analysis is.
Any generic definition will be so general as to be of no particular help in defining it,
and will likely result in the types of confusion that we have identified. If this is con-
sidered too much of a cop-out, then we would like to offer ‘using data to deal with
some problem, issue or other’ as a definition.
A part of what we would like to accomplish with this book is to provide something
of an account of how analysis relates to the other practices of social research – what
we call contextualized analysis. Our definition of analysis is about the relationship
between data and conceptual problems, and our aim is to explore this relationship as
a feature of all social research work. We are interested in looking at the ways in which
researchers use this basic issue of the relationship between ‘data’ and ‘problem’
throughout their research as a means to, or as an aspect of, undertaking their research
work. Our basic thesis is that one way to think about data analysis is as one compo-
nent of a broader analysis of a problem in relation to data. What we hope to shows
through this book is that when analysis is considered in this more general way, it
becomes clear that the distinction between data work and other types of work is in
many ways unhelpful, and is part of the reason why people find qualitative analysis
so opaque. The situated approach to analysis helps to show, for example, how
research problems are developed through data work; how literature is used to con-
struct research problems and to think about and even work with data; how research
plans and designs are produced and worked through in relation to data and the ana-
lytic work it is supposed to do; how ‘gathering’ data through research always involves
a simultaneous analysis of that data. When viewed like this, ‘data’ and ‘analysis’
becomes much less abstract, and more tightly integrated into research as a whole.
But this may raise a question: there may be nothing different about it conceptually
and at this general level, but surely there is something distinctive about data work as
a set of practices? Surely there is something that constitutes data work? Well, the answer
is both ‘yes’ and ‘no’. We will show through this book that, in fact, when you reflect on
the research process many of the problems that people face when thinking abstractly
about data work disappear, as the issues to which analysis is directed become much
more visible. However, the practices of dealing with data are different from, say, deal-
ing with literature or planning a research project, and there is a lot to say about the

Introduction: qualitative data analysis in context
7
particular things that get done during data work. In addition to working through our
approach to contextualized analysis, then, we will also be addressing some key issues
related to data work, such as the use of computers in relation to research, the ways that
audio and video data can be handled, and the issues of transcription in qualitative
enquiry. Given what we have said about the contextual nature of analysis, our discus-
sion of these matters is not in any sense complete. We could not possibly show, for
example, how all researchers ought to analyse or deal with their video data or what a
good transcription should look like. Our discussions should be taken as restricted
(how could they be otherwise?), and as offering ideas and illustrations rather than firm
and generalizable methods of working.
But what we have said so far does not take account of the fact that when people
talk about qualitative data analysis, they often do so in relation to some more or less
formal ‘approach’. Discourse analysis, thematic analysis, rhetorical analysis, conversa-
tion analysis, narrative analysis, critical incident analysis, semiotic analysis, cross case
analysis, grounded theory analysis, ethnographic analysis – these are just a few of the
terms that are often used when talking about qualitative data work. This extreme
diversity, and the wide range of theoretical and disciplinary perspectives that feed
into it are another one of the reasons why qualitative analysis is so difficult to address
or to make sense of. Wolcott provides a list of more than 50 different distinctive
approaches to analysis (1994: 27), many of which could easily take up a book in their
own right. It would be impossible for this or any book to provide a thorough guide
to this immense body of work. While we will be looking in detail at a number of
them, our purpose in doing so is to exemplify the ways that particular forms of analy-
sis direct enquiry and data analysis. In this way we hope to raise people’s interests in
enquiring about different approaches or modes of analysis, and to encourage an atti-
tude of critical reflection in relation to them. This should not be seen in any way to
retract or distract from our arguments about the situated nature of qualitative data
work. On the contrary, it is precisely by working with data in context that the rele-
vance or otherwise of these diverse perspectives and approaches becomes evident.
We hope that these opening pages have provided some clarity as to our purposes
and general approach. But there are a few more issues to clear up before we launch
into the more focused discussions of the book’s constituent chapters. In particular, we
would like to say something about the process of qualitative enquiry in general, and
about the role of theory within that process.
The notion of ‘qualitative’ in
qualitative data analysis
Already in this opening chapter we have been implying and occasionally actively using
a distinction between ‘qualitative’ and ‘quantitative’ research, and qualitative and quanti-
tative data analysis. However, providing definitions to support this well used distinction
is a notoriously difficult thing to do (see Snape and Spencer, 2003). A part of the diffi-
culty is that the methodological debates, epistemological positions and research prac-
tices to which the distinction pertains are not easily divided into two separate camps,

8
Working with qualitative data
but are areas of discourse that have a complex relation to one another. It is common for
the aims of qualitative research to be defined in the following ways:
Examining the construction of meaning
Understanding the details of peoples’ lives or frames of reference
Reflecting on the role of the researcher in the generation of data
The practices of qualitative research are often described as being flexible, iterative,
naturalistic, and as resulting in thick descriptions that are reflexive about the ways in
which research data is constructed. All of these characterizations are appropriate as
general descriptors, but they hide significant variations.
As the ‘other’ in the dichotomy, quantitative research is often described as involving
an interest in the correlation between variables, and with the uses of scientific methods
and statistical procedures to generalize findings – we have described it that way our-
selves earlier on in this chapter. Again, though, such definitions invariably gloss differ-
ent practices, methodologies and commitments, and oversimplify a complex interplay of
ideas and traditions. It is, then, a characteristic of the labels ‘qualitative’ and ‘quantitative’
that they perform crude glosses. They divide up the social research community in a way
that many researchers would not themselves choose. With this caveat in place, we will
invariably, and frequently, make use of the loose distinction implied by these terms.
Box 1.1
Key concepts in qualitative research
Reflexivity is a key issue in social research that refers to the process of reflect-
ing on the role of the researcher in the construction of meaning and, critically,
of data. The ‘reflexive turn’ has been particularly visible in ethnographic
research, and is exemplified nicely in the writing of Clifford and Marcus
(1986) and of Clifford Geertz (1990).
Thick description is a term made famous by Clifford Geertz (1973) and
involves the production of rich descriptions that outline the details of the con-
texts of people’s actions and practices so that they become intelligible in their
own terms.
The term naturalism is particularly difficult to define as it refers to a set of
debates about the socially constructed nature of the social world and the impli-
cations of these characteristics for social research practice. Lincoln and Guba
(1985) provide a very influential paradigm for thinking about these issues that
draws attention to the multiplicity of perspectives in social life, their negotiated
character, and the requirement for contextual explanation and understanding.
Quantitative data is usually thought of as that which can be coded numerically
for the purposes of statistical analysis. By this definition, qualitative data can be
characterized as ‘everything else’. It is common for quantitative research to produce
some qualitative data (i.e. things that can’t be numerically coded, like descriptions

Introduction: qualitative data analysis in context
9
of experiences), and for qualitative research to generate data that can be described
numerically and analyzed statistically. Such data forms are often entirely complemen-
tary, and illustrate the oversimplicity of the qual/quant distinction. Indeed, the differ-
ence is often not actually in the data itself, but in the uses to which it is put (on this
point, see Wolcott, 1994: 4). In writing a book about qualitative analysis we are, by
implication, focusing on the ‘everything else’ that is left over from numerical analysis.
In spite of the title, this book is not just aimed at ‘qualitati