Materialien Aus Der Bildungsforschung Nr. 50 an Analysis of Data Quality in a Retrospective Life Course Study

Going beyond the traditional social-psychological model of the interview, criticism of the survey method points to the importance of the organizational framework of the interviewing process. Control of interaction in the interview through standardization may lead to an alienation of interviewers and respondents who may react with withdrawal from the task, aggression, or 'output reduction.' Using a large sample of computer-assisted telephone interviews from the German Life History study, I show that a multi-dimensional conceptualization of data quality both in terms of appearances of the data and in terms of their meaningfulness is useful for evaluation. A method for measuring data quality in retrospective life course interviews is presented. Previous findings with respect to respondent effects on data quality, especially of their education, are replicated. Testing hypotheses drawn from organizational behavior perspectives on survey research with respect to the impact of interviewers reveals a complex picture. Specifically, interviewers' experience and productivity are mixed blessings and may, contrary to the hypotheses developed, decrease apparent data quality but enhance meaningfulness.