PhD studentship in Statistics at Maastricht University, The Netherlands.

Project: Efficient two-stage design of multilevel intervention studies: more power for the same price
Job description
This research project, funded by a grant from the Netherlands Organisation for Scientific Research (NWO), is about efficient design of intervention studies in psychology and health sciences. Such studies concern the effects of behavioral treatments, such as smoking prevention in primary school or a therapy for depression.

The studies often have a nested design (patients in health centers, pupils in schools, employees in companies) with repeated outcome measurements. Sometimes they additionally have a factorial design to study the combined effects of two treatments. This project is about optimal and robust design of such studies, in particular about the best sample size at each level (organisation, person, repeated measure), under cost and power constraints. Part of the project is about efficiency gains by two-stage and multi-stage design, where interim analysis after part of the data have been obtained, determines whether the study should be continued or stopped.

The PhD student will primarily do research, follow courses according to his or her needs (e.g. advanced statistics, computer programming, Dutch or English language), and do some teaching at bachelor level (about 10% of the time). The research is applied mathematical (analytical, numerical, simulation), not empirical. However, its theoretical results will be illustrated with data from published intervention studies.

Requirements
A master degree in statistics, biometrics, econometrics, psychometrics, or a related field with a solid training in statistics. Applicants are expected to be familiar with the general linear model, some matrix algebra, and computer programming (e.g. in Matlab). Further requirements are good (written and oral) communication skills in English, and an authorization to work in the European Union.