For my survey and decision science friends...Do you think it's better to ask for multiple interviews in the household upfront (e.g., after the screener) or to wait until later (e.g., after the first interview). For simplicity, assume all interviews are done by the same respondent.
Case for early: Respondent is fresh and willing to participate. Get them committed to everything you need early. They may be fatigued if asked after the first interview particulary if it is long and difficult.
Case for later: It's too much to ask respondents for multiple interviews before they know what the experience is like. Don't want to ask too early and get a refusal.
Does the decision/commitment burden seem worse under either? Experiences with either approaches or similar decisions, even if not in surveys?
Thanks!
I suppose that it is better to postpone the request about other interviews. At first, it is more important to construct friendly and confidential conversation, and you can go to other task after it. You have to know respondent a little bit before ask her about additional time for conversation. Of course, if you ask about permission for all interview, I think it increase nonresponse rate.
And could I ask some questions about your survey? How much time do you need to fulfill all interview? How many questions in your questionnaire? Does it takes additional meetings or you plan to complete all interview in one visit?
Dmitry,
Thanks for the thoughtful response. I see your point about making the request later (e.g., end or middle of the survey) so the interviewer has more time to a) develop rapport and b) learn about the respondent and household to tailor the request and choose refusal avoidance techniques. I think this is how and why most surveys do it this way. It would obviously be too much to ask for mulitple interviews right up front.
On the other hand, if the interview is long or dull, the respondent may refuse a second interview if it's introduced late in the survey. However, if the respondent has first expressed an initial agreement (even if just tentative) to participate in the later interviews, they may be more likely to agree to more interviews even in the face of fatigue. I'm thinking of a commitment/cognitive dissonance type hypothesis here.
I was thinking about surveys in general, but my survey is a good example. It's a phone survey of about 20 min on average I think, but of course with a long positive skew. After the sample adult is interviewed, we ask to interview a parent (sometimes the same parent) about a sampled child. Then we ask to interview a sampled teen directly. I have heard parents be about to refuse the teen interview when the interviewer interrupts and clarifies that the teen is interviewed directly, at which point the parent is totally fine with it. I also just worry about getting that child interview done particularly if the interview is tough, boring, or long up to that point.
Any other thoughts? Thanks again.
I don't think that so simple and mechanic decision (ask or don't ask permission up front) can reduce non-response. It is up to interviewer. Much more important to think about the motivation of respondents who participate in the survey. Why would she spend twenty minutes to talk with you? What can you offer her? Is it understanding of the her situation, or the importance of the survey for the community or something else? At least I hope that argument of participating is more important than sequence of request to participate.
- Badges
- Science topic
- Similar topics
- Psychology
- Social Psychology