Sensitive question
In empirical social research, questions about critical, difficult-to-inquire areas of the respondents are referred to as sensitive questions (also sensitive or delicate questions ) . Such areas can be:
- Socially undesirable behavior , especially delinquency behavior (illegal behavior)
- Questions about private life : questions about debts and income, sometimes also about age, questions about sexuality
Techniques on Sensitive Questions
There are several techniques for getting reliable answers to sensitive questions:
- In the case of income, for example, the exact income is often not asked, which would probably overwhelm the respondent anyway, but a classification into income categories is offered. The categories are sometimes given randomly chosen letters to create a feeling of anonymity.
- Lie detectors or Bogus pipeline techniques can be used. However, these often reduce the willingness to participate.
- A really anonymous approach to such questions is the randomized response technique . Here the interviewer does not know how the interviewee answers. When using this technique, a random number generator decides whether the respondent should answer the critical question honestly, or whether, regardless of its content, e.g. B. should answer yes. In aggregated form, the proportion of people who actually answered the sensitive question with yes can be determined because the expected proportion of people who should randomly answer yes regardless of content is known from the distribution generated by the random generator and can be calculated out.
- Another method for anonymizing honest answers is the unmatched count technique . All respondents are randomly divided into two groups. While one group only answers harmless questions, the other group receives one more question, namely the sensitive question. Both groups should only give the sum of their “yes” answers. The difference is then the number of those who answered the sensitive question in the affirmative.
- The sealed envelope technique was used in the ALLBUS 2000 in order to obtain usable responses to delinquent behavior. The answers to sensitive questions were given separately from the other questions in a sealed envelope.
- Direct questioning: The Kinsey report is an example of this : questions in a certain tone were asked very quickly one after the other. It was not asked if , but when the respondent last showed a certain behavior. Kinsey (1970 (first 1948), p. 62): "Such rapid fire of questions provides one of the most effective means of exposing fraud, as detectives and law enforcement officers are well aware."
See also
Web links
- Sensitive questions - entry in ILMES
- Stefan Lang: Randomized Response: Survey techniques to avoid bias in sensitive questions (PDF file; 129 kB)
literature
- Andreas Diekmann: Empirical social research. 17th edition. Rowohlt, Reinbek 2007, ISBN 3-499-555514 , p. 418ff.
- Alan H. Barton: Asking the Embarrassing Question. In: Public Opinion Quarterly. Volume 22, 1958, pp. 67-68.
- SJ Clark, RA Desharnais: Honest answers to embarrassing questions: Detecting cheating in the randomized response model. In: Psychological Methods. Volume 3, 1998, pp. 160-168.
- Karl-Heinz Reuband , Jörg Blasius: Face-to-face, telephone and postal surveys. Response rates and response patterns in a large city study. In: Cologne journal for sociology and social psychology. Volume 48, 1996, pp. 296-318.
- Erwin K. Scheuch: The Interview in Social Research. In: Rene König (Ed.): Handbook of empirical social research. Volume 2: Basic methods and techniques of empirical social research. Ferdinand Enke Verlag, Stuttgart 1973, pp. 66-190.
- R. Tourangeau, TWSmith: Asking sensitive questions: The impact of data collection mode, question format, and question context. In: Public Opinion Quarterly. Volume 60, 1996, pp. 275-304.