Hi All, I've thought about this for a while, and finally decided to send it. Hence the reference to a thread now some weeks old (Gosh, do I really need to apologise for thinking about things for a few weeks?). Doug. RE: The Barry article I agree that the primary debates should be about the way researchers interact with available software. However, I'm not convinced that more general paradigms drawn on by the researchers (whether users or developers) should be separated from this analysis. Richards appears to set up a false dichotomy, either we focus on an almost disembodied analysis of the software and the research process, or we engage in sniping and analysis of personalities. I'm not particiularly interested in either sniping or personalities, but I am interested in a researcher's more general theoretical orientation to the research process, and how this affects the use of computing packages. One of the concerns I have with qualitative computer packages in general is the tendancy to view them as the equivalent of statistical packages available for analysing survey data. It is almost as if we have forgotten the reason for doing qualitative research in the first place. At least in the symbolic interactionist tradition that I am familiar with, the first move is to emphasise that if you want to understand why people do things, you have to understand the meanings they give to things. Meanings are not compartmentalised, but are developed as part of a person's overall self-understanding. If we apply this reflexively, to doing qualitative data analysis with computers, then a legitimate question to ask is how does data analysis fit into a person's more general approach to social research, including general theoretical assumptions (say, whether they like phenomenology or not). Don't get me wrong - I think some of the new packages are extremely helpful, and I've used them myself. I'm just trying to argue for a more complex understanding of the selection of qualitative packages for data analysis. Following the interpretative paradigm through to its full implications, the reasons for choosing a package should be based on the meanings and interpretations that researchers and developers bring to the package, not on attempts to quantify the package's features. (Put another way, for those schooled in SI heritage, I'm more influenced by the Chicago school than Iowa.) Perhaps a survey of users, if one is ever done, could include a question about the general theoretical orientation that most informs their analysis. Another way of doing it would be to do a content analysis of published papers using various packages, comparing the package and the theory. Though, things change so fast, this might not be useful. There's a wonderful article , the author of which I forget, titled "how the refrigerator got its hum". The argument is, basically, that the humming refrigerator eliminated its non-humming competition because it gained greater market share through various aggressive advertising campaigns and early entry into the market. We use VHS video rather than Beta for the same reason - it is an inferior technology, but everyone else is using it. I can't help feeling that there are all sorts of political and marketing process operating in a similar way that influence our choice of computing package, and hence, that influence the way that qualitative research is being done. I'm not wanting to suggest that the leading packages are inferior, simply that they might not be chosen, for being conducive to particular research designs or general theoretical paradigms, but for reasons associated with availability and the like. But then, what's new? Such is life! Doug Ezzy Lyn Richards wrote: > She's carefully avoided stereotypes of users and >concentrated on summarizing her conclusions about software tools and >structures, strengths and weaknesses and their appropriateness to >different research styles, not "types" of researchers. Snip >But none of these >products is perfect, and that's the point - the paper does reopen the >debate about software. What sort of access do researchers want to data, >what sort of management and what sort of linking, and are they getting >it with current software? >Barry's paper raises new ones. They are important, and they are not >about the personalities of users of 2 products - let alone the >developers - but about research goals and software tools. Douglas Ezzy PhD Sociology University of Tasmania, Hobart, 7001, Australia Ph (613) 6226 2330 Fax (613) 6226 2279 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%