I work at the University of Twente, Faculty of Behavioral, Management, and Social Sciences, in the Department of Research Methodology, Measurement and Data Analysis. My research activities cover different areas in Bayesian item response modeling.
My research interest is focused on (Bayesian) statistical modeling of (clinical and educational) response data to improve statistical inferences and data-based decision making. I have developed (marginal) Bayesian models to handle high-dimensional (experimental) designs and include data from multiple sources. The marginal modeling approach facilitates the detection of personalized/indvidual (treatment) effects, and allows for statistical testing the presence of random (design) effects.
I have been involved in different medical studies including clinical trials. They covered for instance monitoring tools for cognitive function of elderly, validating instruments to diagnose smoking, and assessing (personalized) treatment effects of patients who required a new drug-eluting stent during a coronary intervention.
For many applications, I have developed software programs, mainly in the statistical software R, and I have published several software packages on popular (open) repositories. My published software comprehends novel statistical models and computational routines for data analysis.
For a longer period, I have been working on complex latent variable models in high-dimensional applications. The areas of modeling research relate to theory and methods of multivariate analysis, stochastic simulation, mixed effects modeling, among other things. Applications and data analyses are executed in the field of educational, medical, and psychological research. An application area is the development and application of Bayesian response models for large-scale comparative survey research. Specific attention is given to the Bayesian testing of measurement invariance, to test whether an instrument/measurement operates in the same way across different groups.