Very interesting thread this.
As the project I have been working on becomes more rooted in
quantitative work, I've gone with it and have noticed a shift in my
attitude to qual work.
What a lot of worry over substantively very little! At the beginning of
the thread, John Seidel offered an example of the 'dangers' of
autocoding using text search of the word 'pain'- a very disingenuous
example of using 'autocoding' to say the least. I doubt anyone for whom
pain was an issue would ever dream of doing this unless the intention
was then to go back through the text search and look at all the
contexts in which respondents use the word 'pain'.
However, I refuse to accept that there is any problem in 'auto-coding'
responses to a question, socio-demographic data of respondents and so
forth. Qual researchers seem to worry too much about being seen to do
rigorous analysis. It is unsurprising that much qual analysis is rooted
in iterative operations which in many ways simulate positivist methods
of statistical data analysis; assuming nothing to start with and
gradually refining one's understanding of data.
There's nothing wrong with this of course. But this is not the root
cause of successful data analysis per se, rather it imposes a framework
on the analyst's thinking and allows him/ her to spend a lot of time
amid his/ her data engendering a thorough understanding of it.
There are many ways of achieving this- and all are consonant with the
analyst's style of thinking. Autocoding may not work for John Seidel.
However it will work for other people whose approach is quite
different. And that is because the concerns he has over 'autocoding'
are of no relevance at all to the way others go about analysis.
Dave Thomson,
Research Fellow,
University of North London.
--- Lyn Richards <[log in to unmask]> wrote:
> Love the metaphor, Birrell, backhoes and camelhair
> brushes are
> henceforth in the methodological glossary! To Doug's
> point, I absolutely
> agree: we have to be able to see data always in
> context. (John's right,
> we all I think agree there.) But what's the relevant
> context?
>
> Don't we have to be able to see in many contexts and
> out of context too?
> (It's what I was getting at in my "Closeness to
> data" paper in QHR last
> year - closeness can mean many things and some,
> particularly immediate
> access to coded chunks, make the distancing that
> theorizing requires
> very difficult.) So we need computer retrievals to
> offer us immediate
> recontextualizing in several ways - show the
> (relevant) context, code
> it, go back to the whole document, go elsewhere to
> understand...
> Character based text search, for example, gives you
> zero context unless
> you specify a context; so you can do a pincer search
> in finest possible
> context or a real backhoe job, and things move in
> different ways
> depending on which you did. And Birrell's crucial
> point is that
> backhoes themselves crudely recontextualize! Just
> seeing all this
> different material together helps you re-see it.
> Strauss taught me that
> re-seeing is sometimes half the battle - and then,
> the other half is not
> clinging to the original context.
>
> (btw Birrell, I always thought till today of the
> scoop searches in terms
> of drag net fishing. Bad for the fishing fields,
> just because you get so
> much that you weren't aiming to get. But you do get
> a lot of fish, and
> you find out a lot about other critters out there
> too. So long as you
> don't think you can catch river trout that way...)
>
> cheers
> Lyn
>
> Lyn Richards,
> Research Professor of Qualitative Methodology,
> University of Western
> Sydney,
> Director, Research Services, Qualitative Solutions
> and Research.
> (email) [log in to unmask]
> (Ph) +61 3 9459 1699 (Fax) +61 3 9459 0435
> (snail) Box 171, La Trobe University PO, Vic 3083,
> Australia.
> http://www.qsr.com.au
>
>
_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|