Almut Schumann , German Federal Institute of Population Research (BiB)
Detlev Lück, Federal Institute for Population Research (BiB)
The assessment of relationship quality relies on indicators which reflect several dimensions of relationship quality, such as satisfaction scales, conflict frequencies or subjective instability of the relationship. Since measurements based on sensitive questions are strongly influenced by the interview situation and therefore the mode of data collection, the switch from face-to-face to web mode or to mixed-mode designs in panel studies can challenge the comparability of measurements. The same is true for cross-national comparisons and for differences in modes of data collection in international surveys. It is uncertain to which degree such mode effects also apply for indicators on relationship quality. Therefore, a sound understanding of the sensitivity of indicators on relationship quality and of the impact of the mode of data collection on such items is needed. We use German data from a GGP Pilot study (2018) with an experimental design, using a real survey instrument of the Generations and Gender Survey. At first, we assess mode differences by comparing distributions of indicators for various dimensions of relationship quality. In order to control for effects of selectivity among socio-demographic characteristics, post-stratification weighting and multivariate regression analysis are applied. Findings reveal differences between web and face-to-face mode for almost all items. Mode differences persist when controlling for socio-demographic characteristics. Respondents in web mode evaluate their relationship with a higher degree on self-disclosure and more honest in thinking significantly more than twice as often about breaking up with their partner and show a significant lower satisfaction with their relationship in general.
Presented in Session P1. Postercafe