Mixed methods. Feel lost. My Comparison study


======= Date Modified 20 Sep 2012 02:33:46 =======

I feel a little lost as to what type of analysis to do. Thinking mixed methods. Combining either content analysis or thematic analysis with quantitative analysis. Concerned about sample sizes

I want to see if clinicians / students views of substance use are in line with the views of substance users (Indigenous and Non Indigenous). I want to see if clinicans / students who score high on warmth, openess etc have views more in line with those of substance users.

I conducted an exploratory study to compare views of health practitioners and substance users on the topic of substance use. I surveyed health pracs and students in health programs. Survey contained 7 open ended questions eg. "What influences a person to engage in substance use?"4 scales relating to personality, sensitivity etc

I surveyed 50 clinicians and 40 students. I randomly assigned students and clinicians to either an Indigenous survey or Non Indigenous.

I interviewed 20 Non Indigenous substance users and 13 non Indigenous.

I looked at the open eneded questions from clinicians and generated themes from the data. I did not have predetermined themes and I did the same with the interview data. Have I used Thematic analysis here? Can I discuss differences in themes in a critical discussion.

I then want look at health practioners personality and their views around substance use


Why do you believe that you have to do quantitative analysis?


======= Date Modified 20 Sep 2012 11:04:36 =======


I was under the impression with qualitative data (the client interviews) and small sample sizes purely quantitative would be out. Any thoughts?


I would say it's a predominantly qualitative study, based on what you've said. Yes, you have some quant data from the questionnaire (the scales) but there are always arguments that scales are more qual than quant anyway. Obviously, there'll be quant stuff like demographic information (gender, grade of job, length of time as a user/practitioner, maybe) but it doesn't sound from what you've said that you particularly need to think of it as equally qual/quant.

The truth is that, outside of pure science, it's rare to find studies that are exclusively qualitative - there is nearly always something that can be quantified, even if it's men vs women or age ranges.

I would personally keep the focus qualitative, as you seem to be suggesting. Though, of course, your supervisor is the person who really needs to help you decide this!

Hope this is of some help,



There's a school of thought, which I have a lot of time for, which says that intensive/extensive is a better dichotomy than qualitative/quantitative. This sounds like a qualitative methodology would be more appropriate. If they're Likert-type scales then that's qualitative.
Lormak, what field is your supervisor from? I know from experience with one of my supervisors that medics don't really 'get' qualitative methodology.