I am relatively new to cluster and content analysis and working on a project for which I think both are necessary. I hope to receive some advice on what procedure to follow for the analysis of my data.
I have two pieces of data:
1) A couple of hundred answers to an open question of a questionnaire. The question asks the respondent to describe the most important problem they encountered over the last 12 months.
2) Subsequently, the respondent rates his description on 8 attributes (bipolar 7-point Likert scales).
Now I want to develop a categorization of problems that the respondents face. My question is: what is the best way to do develop such a categorization?
I had the following in mind:
Step 1) Cluster analyze the open answers on the 8 attributes.
Step 2) Apply some form of content analysis on the clusters.
Step 2 is necessary because the clusters that result from step 1 need to be corrected by incorporating some kind of judgment from me as the researcher. A correction is necessary because: (a) the 8 attributes are insufficient to differentiate sufficiently between the open answers and (b) the scores on the 8 attributes are also affected by respondents' subjectivity and context and not only by the content of the open answers. Therefore, cluster analysis needs to be supplemented with another (probably more subjective and qualitative) technique.
With what technique and how should I complement cluster analysis to develop the classification? By the way, I cannot use the respondents again to do for example a card sorting exercise because I cannot bother respondents again with this study.