The discussion generated by Odd's paper seems to have established that
several trainers perceive a firm distinction between learning
computer software and learning approaches to qualitative data analysis.
Their reflections suggest that they experience the pedagogic problem of
clients who are at an insufficient level to appreciate the distinction.
In general, though, the discussion supports the view that QDA software
does not, by virtue of its design, forcibly impose coding according to
members' categories, but allows for coding derived from social theories
equally well. This is my experience too. QDA software, in my view, can be
used for grounded theorising, discourse analysis, conversation analysis,
narrative analysis etc. It is, at root, a sophisticated tagging and filing
system, allowing for complex retrievals of materials, which can
themselves be categorised according to a variety of conceptual schemes,
however derived. This kind of operation can, in theory, be applied to all
sorts of approach and analytic task.
Let's, then, move on to consider the empirical question of whether CAQDAS-
supported research practice reflects such diversity. Fielding and Lee's
book reports on what researchers say about what they do, and gives some
insights here. Kelle's edited collection contains several helpful examples
of different approaches, ranging from the simple to the sophisticated,
that CAQDAS has in practice supported. There are some other examples in
the literature of researchers writing about their usage of these packages
on particular projects (eg: by Lyn Richards, Alan Cartwright and
colleagues, others too that I can't at present bring to mind).
I am currently collecting materials for an article on how people have used
computer packages to analyse interview data. I have been collecting
examples from as many sources as I can, including a message asking for
material to this list a few months ago, which brought a helpful response.
I have also tried to get hold of reports that are said to use NUDIST
posted on the QSR web site, and done electronic searches of bibliographic
databases. I am beginning to get quite a collection of published pieces by
researchers reporting studies in which they mention their use of CAQDAS to
support the analysis.
I hope that the results of this survey will assist in
illuminating the empirical problem that lies at the heart of this issue,
by assessing the degree to which published work reflects methodological
diversity, or is indeed largely bound to simple coding and retrieval for
descriptive purposes (which is itself, as Lyn says, often a good thing to
be doing).
Inevitably I am snowball/volunteer sampling in this. I would be grateful
if anyone who has done work, or knows of work (especially if in the public
domain) that shows an interesting usage of CAQDAS to analyse interview
data, would email me personally. I might know about it already, but I am
sure there is a lot out there that I don't yet know about and could be
included in my survey.
Clive Seale
Department of Sociology
Goldsmiths College
Lewisham Way
London SE14 6NW
Phone: 0171 919 7729 (direct)
0171 919 7707 (office)
Fax: 0171 919 7713
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|