RESEARCH

My research interest is in three main areas: ‘Predictive Psychometrics’, ‘Regression models for categorical response variables’ and ‘Longitudinal Data Analysis’.

Predictive Psychometrics
Most psychological research focusses on explanation. That is the ‘why’ question: why do things happen as they happen? In practical situations it is not only important to understand why things happen but it is also important to predict, i.e. to be able to tell what will happen in the near future for a specific person. This line of research focusses on statistical methods for prediction.

Regression models for categorical response variables
Many measurements in Psychology are categorical. Main examples are

  • correct or incorrect answer to a test item (two categories);
  • totally disagree, disagree, neutral, agree, totally agree with a statement (5 ordered categories);
  • living independently, homeless, or living in a community house (3 unordered categories)

Although these measurements are categorical often the categorical nature is not taken into account in the statistical analysis. This might lead to fallible results and wrong conclusions.

Longitudinal Data Analysis
To really understand the manner in which people are developing we need to obtain measurements over time. Having such data, however, invalidates an important assumption of most statistical techniques: the independence assumption. This line of research applies and develops methodology for the statistical analysis of repeated measurements.