Hi, I am writing my master thesis. For my practical part I have made 30 semi-structured interviews. Some of the data form the interviews is codded and used for the statistical analysis (e,g, regression) . The affirmations that the respondents stated during the interview will be use as arguments when commenting the results.
Could you please suggest, is this a mixed congruent method ?
Thanks!
There is debate within the mixed methods community about whether "quantizing" (or some times "quantitizing") qualitative data counts as mixed methods. I have attached one of the better-known papers on this topic.
If your plan is to search for patterns in the quantitative data, and then use the qualitative data to interpret those patterns, I have an article recommending that approach:
Article Qualitative Content Analysis: A Guide to Paths Not Taken
One general issue, however, it that your regressions will be limited by your small N. In particular, it will be hard to find statistically significant effects because your analyses will lack "power" due to large standard errors.
Good response from David. Yes - in my mind Alexandrina - I'm not sure that regression will tell you much and 'seems' out of place' with such a small sample and using a semi-structured tool. Even with a larger sample - a structured tool would be a better fit.
I supervise many PhD students - and many are mixed methods. I also supervise many Master's students' - but don't encourage mixed methods with them - due to it's potential complexity, manageability - and on-going 'debate'. Personally, if you were a student of mine, I would encourage a single methodology based on your interviews - which would be qualitative alone.
Dean Whitehead Working with a quantitative analysis of qualitative data is not as demanding as collecting and integrating two different types of data in a classic mixed methods study. So, if this is primarily a question about potentially innovative ways to analyze qualitative data, that could well be workable for a thesis.
It is also worth noting that many of the qualitative data analysis software packages are making big claims about their value for mixed methods. By that, they mostly mean that their coding can be exported into spreadsheet formats, or attached to matching datasets in SPSS etc. I personally find this more than a little misleading, because there is a lot more to qualitative analysis than attaching codes to a quantitative data set.
If there is a way to calculate performance evaluation objectively, then whichever method produces the best results, should be included
- Badges
- Science topic
- Similar topics
- Linguistics
- Composition Studies
- Writing