A Transitive Index to Test for Acquiescent Response Style

Publisher Website
Google Scholar
Abstract
Acquiescent response bias continues to plague survey research, especially among less educated respondents and among respondents who are less familiar with interviewing and questionnaire methods. The failure of studies to test or control for acquiescence is often due to practical and theoretical limitations of the existing acquiescent measures. A short and convenient transitive index of acquiescence which avoids many of these limitations has been proposed and illustrated with survey data collected from a sample of 591 young men in Kano, Nigeria. The index, validated on the basis of differences in mean scores on sets of Likert and matched non-Likert items worded separately in “modern” and “traditional” directions, has proven useful in testing for spurious and masked relationships due to acquiescent bias. It should be emphasized that the specific items comprising the transitive index in this study are illustrative; a different set or sets of items following the basic transitive principle could be constructed to fit other survey situations. It would be useful to investigate the possibility of constructing a transitive index general enough to be used in different studies, thus permitting comparison of patterns and effects of acquiescence. Such a general index would facilitate and encourage the inclusion of acquiescent response assessment as a routine part of survey research, as well as the study of the acquiescent phenomenon in itself. In the meantime the study specific transitive index proposed here hopefully provides a useful and convenient addition to the tools of the survey researcher.

This publication has 21 references indexed in Scilit: