Monday, July 14, 2008

Semi-structured interviewing tips

A colleague of mine is doing his first qualitative study. He is interviewing lower-income couples about their views of what makes a successful or unsuccessful marriage.

He had some questions for me to pass along to his interviewers:

1. Any tips for bringing an interviewee back on track? We worry that people will start telling us irrelevant stories, and we want tools for politely interrupting and getting back to our interview.

2. We would like to, at the end, come up with a polite and nonthreatening way to ask about income. This is supposed to be a study of low-income individuals, and we want to ask at the end -- by the way, what is your income? -- just to confirm that we got who we were going for. Reviewers will want to know this, I think. But how do we ask about this sensitive topic?

I came up with a few things off the top of my head and thought I would post them here in case they are helpful to others:


For your first question, I always tell my interviewers that they will have two major difficulties in doing semi-structured interviews: getting respondents who won't shut up to focus on talking about one thing at a time and getting respondents who won't talk to open up. I don't think there are simple solutions to either problem. It has a lot to do with the talent of the interviewer to connect with the person they are interviewing and to sense what kinds of techniques would work better than others. I think when people are starting to go on and on about unrelated issues, you just have to recognize it is happening, wait until they pause briefly and then jump in with a transition statement, like "Let me ask you about something different" or "what I am really interested in is...". You could throw in something like, "What you are saying is very interesting and important but I don't want to take up too much of your time. I'd really like to know what you think about..." When I have monitored my interviewers, there have been times when they have very tactfully managed to connect the tangential conversation to the topic of the interview. This requires some good thinking on your feet. It is hard to give a formula for doing this because there is no way to anticipate a tangential topic, but that is a great way to steer them back on track. You can even act dumb and ask them how what they are talking about relates to the interview topic and that will reinforce to them that there is a topic to the interview and they are bringing up unrelated things.

Being an active listener is helpful. For interviewees who don't talk much and give simple answers, silence is sometimes a great way to probe them to talk more. The silence after they stop giving and answer is a subtle hint that the interviewer is expecting them to say more because the answer they just gave was not sufficient. However, when the interviewee is more verbose, interviewers should be throwing in things to show that they are getting the information they are looking for and want more, like "Ok" or "uh-huh" or "I see" or repeating back to them what they just said with a questioning tone ("So you think that the most important aspect of marriage is ....?). If the interviewer is doing this along the way, they will more easily be able to say something that steers the discussion in the direction they want it to go and it won't seem abrupt or like they are cutting off the interviewee.

When you are freelisting, interviewees often forget that they are supposed to be listing things and start giving stories. A good way to get them back on track, and to train them to think about lists, is to keep going back to the list once they stop talking. Say something like, "Ok so you said that good marriages have X, Y, and Z. What else do they have?" Keep doing that and most respondents will get the point after doing that a few times.

There is always a danger in sending a signal to the interviewee that they should restrain their answers in certain ways, which is why cutting them off is usually a bad idea. If you stop them mid-sentence that is probably going to make them shut down and start giving you more simple answers. Sometimes they can get very insulted and will start giving yes/no answers.

On the other hand, there is a type of interviewee who will not stick to a topic unless they are repeatedly cut off. An interviewer has to decide that they have one of these people after their other tactics have proved futile. Usually if it is obvious that someone is like this, they are used to being cut off because this is how they have conversations all the time and other people become annoyed at them also. I have heard interviews that would not have been completed unless the interviewer was vigilant about making sure the interviewee did not start going down the road to talking about tangential topics. I have also seen interviewers not cut off interviewees who were going off on tangents early in the interview and then the interviewee gets annoyed when they get tired of talking and then notice that the interview is going beyond the time they were told it would last.

For questions like the income question that may be sensitive, a good technique is to give them a long preamble before the question (it is important to know information about _____, people often don't like to talk about that kind of thing but it would really help us if we knew something about ______ for our study...) and then ask the question in as brief a way as possible.

You could say, "We realize that often people don't like to discuss the money they make. It is a very sensitive issue for a lot of people. Some people are embarrassed if they think that they don't make enough money or other people might be embarrassed that they are bragging about their income. People also don't usually discuss that kind of thing with strangers, so we understand. However, for our study it is very important to understand how marriages are affected by people's income, so it is something that we have to ask about. What is your income?"

You could also include the specific question about details about how much they make as a probe within a general discussion about how their marriage is affected by their income. Once you get them started about the issue, it would be easy to jump in and say, how much do you make? The interviewer could also be looking for a good time to ask that question somewhere along the way in the interview and then throw it in there. It is much easier if they bring up the topic on their own and you ask a follow-up question. You could be asking them about the problems that couples face in their marriages. Once the inevitably say that they get in fights about money or when they don't make enough money, the interviewer can say "so how much do you make?" and it won't seem so awkward.

0 comments: