My research question was “What types of information behaviors do people in engage in when selecting a restaurant when they go out for dinner?” Yeah, I know – very creative and inspired. Not even sure what I picked it – since I never really go out myself. Oh well – I guess it was the best I could come up with in the middle of a brain-freeze.
Anyway – Despite the less-than-inspired subject, I did glean a few insights into the various interview techniques we discussed. Perhaps the most interesting (at least to me) of these was that, although coming up with questions mirroring each “type” (unstructured/ semi-structured/ critical incident), in the end I found myself asking for essentially the same information. So, for example, my “semi-structured” lead in question was “Please describe for me how you chose the restaurant you went to the last time you went out to dinner” – and I asked about typical problems in choosing, sources of info used for selection. My unstructured question was “Tell me how you select the restaurant you to to when you go out to dinner.” In speaking with my interviewee, I found myself almost reflexively asking the same sub-questions I had for my semi-structured and even my critical incident question. While I certainly didn’t want to “lead the witness” in that instance – and I found myself suddenly being very conscious of that problem as I started speaking with her and trying to interact with her as she answered the question – I nonetheless found it at times to be necessary to “prompt” her a bit (just by way of interacting with her). As can be imagined, because the semi-structured and critical-incident interview questions did not require this.
Speaking of the critical incident and semi-structured interview questions – I think as between the two I got basically the same “types” of response – that is, I don’t think I really got anything significant from the one that I didn’t get from the other. This could be a function of the fact that, as it turned out, at least 2 of the 3 people I interviewed are like me in that they don’t go out to dinner very often (mostly special occasions). From that perspective, I’d have to say that the unstructured interview was probably the most lively and interesting – and, although I did have to remind myself to “not lead”, I did find myself feeling freer to interact and discuss the topic with her, as opposed to sort of mechanically asking a series of questions. I guess I’d have to say I thought that the “critical incident” method was next after that in terms of feeling more “interactive” with the interviewee. I think that when you ask someone to recall a specific personal experience that something does click to help you connect with them. As you might imagine from this discussion, I found the “semi-structured” approach to be probably the most mechanical and “rote” of these three techniques.
I suppose the conclusion I arrived at is that I can see how a blend of these approaches could be useful, even in one interview. For example, you might start with some fairly unstructured questions, but then follow up later on with more specific critical-incident questions or semi-structured questions around the same subject to expound on the earlier discussion and, hopefully, serve as opportunities to ensure consistency of response. I also would note that it became clear that, just because you have an unstructured interview question, it doesn’t mean you as the interviewer shouldn’t still have some prompt questions in your back-pocket!
Just for completeness – I’ll say a few words about the “flip side” in conclusion. From the perspective of an interviewee, I think I also found the unstructured interview method a bit easier to “work with” – Sometimes when I was asked more specific/ direct questions I really couldn’t think of an example or a response right off and I could feel myself sort of tightening up a bit trying to think of “something” responsive. In the “unstructured” question format (um – we can discuss that oxymoron later…) I just feel I could sort of get myself started and warm up to the subject a bit more easily.