Hi.  I agree with a prior response - you demonstrate an audit trail but showing how one code or them was developed from raw data to open codes to axial ,etc.  maybe 2 examples.

Your problem is deeper, of course.  I may be hard to do so, but providing what is commonly done by consensus of qual scholars is what you should do here..

As a summary - here's a look a someone else's work (content analysis) that show the trail.  It is not rocket science.

Student's online evaluation comments included "Awesome teacher. Freakishly smart. Helpful and his lectures are informative and he has a good sense of humour. He'll challenge you and makes sure you learn the material. He should teach other instructors in the program how to teach" (2010, p. 80).  Reagan interpreted this passage as "Student is appreciative of instructor's knowledge and presentation, including humour, with the goal of challenging the student to learn. (p. 80).  She then coded this evaluation with the initial codes "Awesome," "Smart," "Humorous," "Helpful," "Informative" and "Challenging."   Students, however, did not rate all instructors so positively.  Another student's evaluation read "Does not follow curriculum, inconsistent marker, hard to follow in class" (p. 80).  Reagan interpreted this evaluation as meaning: "Student is dissatisfied with the absence of content, inconsistency in marking and lack of clarity in presenting the content" (p. 80).  She then coded this evaluation with the initial codes "Inconsistent" and "Unclear" (p. 80).  The initial codes, like open codes in grounded theory, are quite close to the specific content stated in each evaluation.   The second step of combining and consolidating codes involves more interpretation.  It may also involve a move toward greater abstraction and conceptualization. 

This is an audit trail.  No one reviews it all - it's never needed (as in quant fraud - the consistency of data shows fraud - not each specifiic.

Hang i!


On Sat, May 11, 2013 at 5:30 AM, Helen <[log in to unmask]> wrote:
Hello Everyone,

my first post!!

I need some advice.  I am doing a qualitative PhD while my supervisor is quantitative through and through.  We have been clashing since day one of starting my thesis.  So it’s been a really rocky road.  I have completed the analysis chapters and thought they were quite robust.   I have used NVivo to analyse the data.  While I have discussed the relevance of this program to my research my supervisor wants me to establish an audit trail to demonstrate the analytic logic from end to end.  My data base is approximately 8,000 pages (this is not a typo).  Given the enormity of the data base I don’t have detailed notes of each and every step I took in using NVivo.  In your own research how have you explained how qualitative software supported the analysis.  Did you describe all the steps taken.  If you can  provide advice or direct me toward useful papers which answers this issue.

Thanks



--
James W. Drisko, PhD, LICSW
Professor
Smith College School for Social Work
Lilly Hall
Northampton, MA 01063 USA