PUB - Publikationen an der Universität Bielefeld

Drawing on data from two multitrait multimethod experiments carried out in the context of the German Socio-Economic Panel Study (SOEP), this paper identifies questionnaire designs that minimize measurement error in reports of subjective well-being. Among the survey instruments most often used to measure well-being, the analysis focuses on three response formats (11-point, 7-point and magnitude satisfaction scales) and three modes of data collection (self-administered paperand-pencil questionnaires (SAQ), personal paper-and-pencil interviews (PAPI) and computer-assisted personal interviews (CAPI)). Results show that both the choice of a response format and the choice of a mode of data collection make a difference in terms of measurement error: The 11-point satisfaction scale and both CAPI and PAPI improve the quality of subjective well-being data. The paper also reports differences between response formats in terms of their ease of administration and illustrates that the choice of a survey instrument affects conclusions drawn from applied well-being research.