AbstractsPsychology

The present text examines and applies methods for modeling recurrent event data, focusing on whether to use global time or recurrence time as the basic time scale. The purpose is to gain an insight into how easy/difficult it is in practical situations to discriminate between the various modeling approaches, and to discover whether unobservable heterogeneity between individuals should be handled by a frailty model or whether one should use a dynamic modeling approach. Our case study is based on data from a study concerning the influence of improving sanitation facilities in Salvador, Brazil. While the challenging issue is that our observation pattern is subject to significant intermittent missingness and drop out.