Interviewing individuals and the data

This week's subjects before the PhD seminar group were Gwendolyn Sasse and Cheryl Schonhardt-Bailey who gave contrasting accounts of quantitative and qualitative work.

Dr Sasse gave some quite useful thoughts and observations on the use of fieldwork, much of with which I could concur. But there's always the problem of knowing whether to do it or not. On the two occasions I've done fieldwork I've had different views about whether it was worth doing. In the first instance I knew I had to do it to get the information I needed; the second time I wasn't sure that it could be justified. But in both cases I realised that it was necessary, since it disabused me of my initial assumptions and ideas which had only been gleaned from reading texts in British libraries. Actually being on the ground made me realise that there were different views and perspectives and not everything I learnt was in those texts.

I recall being in Dr Schonhardt-Baily's classes while as an undergraduate. She took us for the quantitative component of a course I did back in the late 1990s here at the LSE. Then I recall getting very little from the sessions, since I've never been statistically-minded. This time was no better, but I had a greater appreciation for her presentation on an empirical assessment of ideas as words between Bush and Kerry in the election last year. We were encouraged to read her paper prior to the session and I was able to follow it.

But I did have one or two reservations. Whereas she has made available the datasets from the computer software she used to quanitify the data, I still find myself having to fall on trusting her work - since I still remain weak on the quantitative side. Which brings me to the second point, that of the results themselves. In her presentation Dr Schonhardt-Bailey noted that if the results had been counter-intuitive then she would have known the data was wrong. To another questioner she stressed the importance of knowing the context wince the language used could have contrary meanings. Which all suggests to me that this kind of empirical work is useful, but in a descriptive approach, subsequent to qualitative work. I question whether it is possible to do such work without having had prior qualitative work done (e.g. knowing the subject and how different audiences perceive it).