living and learning

Wednesday 24 February 2010

Alternative methods please

I confess I'm slightly scared of qualitative research. Something about interviews and ethnography fails to garner my trust - the general woolliness? The Hawthorne effect? Or perhaps it's just that I haven't done enough homework?

Anyhow, I've been swayed by grounded theory lately. (Apologies if this has been brought up in lectures before; it clearly didn't register!) Unlike the familiar 'literature → hypothesis → data → analysis' model, grounded theory ditches the preliminary preparations, and starts with 'data'. Ideally, you would simultaneously collect and code your data (eg different answers to one question in the interview), until you get no new responses. Then you would group your codes into categories, and use these to build a theory.

Ok I'm not entirely sure how this works (or whether it's even plausible with limited resources), but the 'no pre-research literature review' appeals to me greatly. In spades. And it's not about laziness either - the literature review stage simply comes later in the process. I mean, you don't want your preconceptions to influence your data do you? You want to be as objective as humanly possible, and that's going to be quite hard when you've devoured the opinions of everyone else in the field. Charlie happens to have an anecdote of a student who kept getting bad results in his data analysis: it turns out the results were only 'bad' because they didn't fit in with his knowledge of the subject. 'What's the point of studying what you already "know"?' quipped Charlie.

But then there's a drawback. Books on research design, our own department, and the dissertation plan provided by DARG all suggest that you 'plug your dissertation into existing literature' - can't guarantee that will happen with grounded theory. Besides this there's tonnes of criticism out there, which I really don't have the energy to read right now. Perhaps its obscure status in the undergraduate world is the best testimony? Hmm.

0 comments:

Post a Comment

  ·  Back To Top  ·  Editor's Pick