PhD defence A. (Anoek) Castelein

In this thesis, I develop approaches to accurately estimate the population distribution of responses. The proposed approaches overcome important limitations of the existing approaches by allowing for more realistic behavior. That is, they allow for many different forms of response distributions, including those were some individuals may have no response to certain variables (Chapter 2). Also, the proposed methods allow for the responses of individuals to change over time (Chapters 3 and 4). In the applications in this thesis, I find that our proposed approaches lead to improved predictions of individual outcomes. Also, the approaches lead to interesting insights into how individuals respond. These improvements can lead to the design of more effective policies.

Due to corona, the PhD defences do not take place publicly in the usual way in the Senate Hall or in the Professor Andries Querido Room. The candidates will defend their dissertation either in a small group or online.