Methodology and Statistics

Research

Research in the unit Methodology and Statistics focuses on advanced behavioral research methods. There are four key research areas: Applied Psychometric and Sociometric Modeling, Statistical Learning and Artificial Intelligence, Neuroimaging Statistics, and Responsible Research Methods.

Four key research areas

1. Applied psycho- and sociometrics

This research team is concerned with the measurement and modelling of behaviour, cognition, and unobserved traits (e.g., latent variables, missing data). Central aims include the distillation of a complex set of features into lower-dimensional representations (e.g., factor analysis, IRT), determining nonlinear relationships between traits and categorical item responses, and modelling of change over time.

2. Statistical Learning and Artificial Intelligence

This research team focuses on the development of data-driven methods for prediction, such as regression, classification, and clustering. These methods employ resampling techniques, including cross-validation, bootstrapping, and permutation, to assess the quality of solutions and conduct statistical inference. They facilitate hypothesis generation and prediction in applied behavioral science.

3. Neuroimaging Statistics

This research team focuses on improving the valid and reliable measurement of brain representations of behaviour, cognition, and other mental states through advanced statistical modelling of signals derived from functional and structural neuroimaging (MRI, EEG, etc.). Goals include maximizing the signal-to-noise ratio of derived neuroimaging measures, validating the reliability of classification and clustering methods for multimodal MRI, and increasing reliability of methods for brain-wide association.

4. Responsible research methods

This research team focuses on improving psychological science through research on methods, statistics, and research practices. It includes (meta) research on replication, reproducibility, robustness, Open Science, and science communication that together further the transparency, integrity, and overall credibility of our field.

Collaborations

The M&S unit participates in the research school IOPS (Interuniversity Graduate School of Psychometrics and Sociometrics). Also, the M&S unit coordinates and participates in Statistical Consultation Service StatiCS for the FSW Graduate School, and has a leading role in the community management of the Open Science Community Leiden (OSCL).