Selectivity Processes in and Weights for the Berlin Aging Study II (BASE-II)

Like many medical studies, the Berlin Aging Study II (BASE·II) is based on a non·random "convenience sample" of self·recruited participants. To study processes of selectivity in BASE·II, we used an identical questionnaire to compare BASE·II with a large, representative reference study, the German Socio·Economic Panel (SOEP), thereby allowing differences in characteristics of participants in BASE·II and SOEP to be analysed easily. Based on this selectivity analysis, we then generated propensityscore weights that adjust for the selectivity in the BASE·II survey. In addition, we adjusted the weights of the BASE·II sample to regional information from the Federal Statistical Office so that the BASE·II study has the same totals as the official statistics.