Just a quick note on the issue of coding credibility... One alternative is
to have another party (a co-researcher or an "expert", depending on your
definition and needs) go over a portion of your coding just to see if it
makes sense that a coded segment of text supports the code. This strategy
may satisfy many committees...
I have an article that describes the process, but the reference is not at
hand (recently switched computers...) Will see if I can find it to post.
In general, I agree that expecting exact congruence between coders is not
appropriate for qualitative research, but that using the differences for
dialog is extraordinarily useful.
Linda Gilbert, Ph.D
The University of Georgia
[log in to unmask]
|