-----Original Message-----
From: luisbotella@usa.net [SMTP:luisbotella@usa.net]
Sent: Tuesday, October 07, 1997 3:17 AM
To: pcp@mailbase.ac.uk
Subject: Grounded Theory Methodology
** In reply to request to say more about Grounded Theory Methodology
and QSR-NUD·IST Dr Stephen Tagg writes;
Grounded Theory, I am told by
our resident expert Andy Lowe, requires emergent methodology. This means a
flat first pass of the data. QSR NUDIST can only manage this with
difficulty. Most of it's examples involve a pre-existing hierarchical
coding
system being applied to texts. This is the basis for my claim that it is
not designed for grounded theory methodology.
** Dear Stephen, what
about a firts paper-and-pencil pass on the data as a basis for the
hierarchical coding system before entering anything into the computer?
That's
what we're doing here in Barcelona and we're quite satisfied with the
result.
Perhaps using qualitative data research mainly for representational
purposes makes sense in your application. I can think of no other reason
why you would do all the hard work with paper and pencil on the data (or do
you develop the hierarchy with a subset) then to do it again with NUDIST.
How amenable do you find NUDIST to changing your hierarchy once it's in?
Can you do maps of relationships between codes as Atlas/ti seems to offer?
I should also say that my imagining and teaching of how to do qualitative
data analysis are not based in my own experience, rather on distilling
discussion list comments I see from users of nudist and atlas and from
discussion with my colleagues. I feel that to be a good teacher of
multivariate methods, I should appreciate other ways of handling complexity
and be able to advise researchers from a wider appreciation than just my
area of expertise.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%