Response Variation and Location of Questions Within a Questionnaire

Publisher Website
Google Scholar
Abstract
Helsing, K. J. (Dept. Epidemiology, Johns Hopkins Training Center for Public Health Research, P.O. Box 2067, Hagerstown, Maryland, USA 21740) and Comstock, G. W. Response variation and location of questions within a questionnaire. International Journal of Epidemillogy 1976, 5: 125–130. In a questionnaire administered to a general population sample of 1673 adults, a 14-item scale of social desirability was divided into two groups of questions, with the first group of seven questions appearing early in the questionnaire and the second group of seven questions toward the end. For 220 of the respondents, the positions of the two sections of the scale were interchanged, permitting analysis of the effect of location on responses. For 11 of the 14 questions, almost evenly divided between desirable and undesirable behaviour, the percentage answering ‘true’ was higher when the questions was asked early than when it was asked late. The difference was statistically significant for five of the 11, four of them relating to socially undesirable behaviour. The results indicate that responses can be dependent on the location of questions in a questionnaire, and that the effect of question location may not be the same for groups with different personal characteristics. For this reason, it seems unwise to alter the administration or design of a questionnaire in any way if the results are to be compared directly to those of other studies, unless the altered questionnaire is standardized against the original version.

This publication has 0 references indexed in Scilit: