> My main concern with the validity of GTM
> is that, when it comes to the hypothesis testing stage of the NUD=B7IST
> program, you seem to get from it exactly what you've previously put into=
> it--only in a more sophisticated way, of course. Anyway, I guess the sam=
e
> could be said of any qualitative method based on hermeneutics. I'm also
> wondering what will be the best way to prove that the method have some
> external validity, I mean, that the final result is not altogether a
> product of the analyst's subjectivity.
Dear Luis,
I agree, what QSR-NUDIST does well is provide you with feedback about what=
text is accounted for by what definitional nodes and yes, this is
pretty basic stuff. The challenge is to be creative.
In terms of checks on validity a couple of strategies I use are;
(i) inter-rater reliability analysis of definitional frame (others
can understand and apply my scheme).
(ii) using new text to evaluate the comprehensivness of the scheme
(i.e., can all the text be coded according to the existing frame,
what do ambiguous cases tell us about the scheme).
(iii) triangulation with findings originating from elsewhere.
Are there more and better ways? I am keen to hear.
Kind Regards
malcolm
malcolm c. cross
department of psychology
city university
northampton square
london ec1v ohb
telephone +44 (0)171 477 8531
facsimile +44 (0)171 477 8590
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%