Skip to content

Publication

Collecting life-course data in a panel design

Abstract

"The National Educational Panel Study (NEPS) has to combine the retrospective collection of life-course data with repeated competence measurements in a panel design by updating life-course information on an ongoing basis. The greatest challenge to updating life courses in a panel study is ensuring the overall consistency and completeness of the life course across multiple waves and preventing seam effects. These effects occur in the transitions between different states of interest from one panel wave to the next, and their number is much higher when the data for each period come from two different interviews than when the reports come from the same interview. To minimize this effect and to ensure that episodes collected in different panel waves are connected with each other, NEPS researchers use dependent interviewing techniques that draw on information collected in previous panel waves in order to phrase questions and direct respondents through the questionnaire. Proactive Dependent Interviewing -- whereby information from the previous interview (named preload) is used to stimulate the memory as part of the questioning process -- is particularly widely used because of its potential to lower respondent burden, increase efficiency, and reduce measurement errors, such as seam effects. Against the background of findings from cognitive psychology, we describe how we implemented this technique in the NEPS Starting Cohort 6-Adults. We then evaluate the quality of this kind of 'anchoring' by empirically analyzing the conditions under which respondents disagree with preloaded data." (Author's abstract, IAB-Doku) ((en))

Cite article

Trahms, A., Matthes, B. & Ruland, M. (2016): Collecting life-course data in a panel design. Why and how we use proactive dependent interviewing. In: H.- P. Blossfeld, J. von Maurice, M. Bayer & J. Skopek (Hrsg.) (2016): Methodological issues of longitudinal surveys : the example of the National Educational Panel Study, p. 349-366. DOI:10.1007/978-3-658-11994-2_20