Research Methodology and Statistics
Since 2012, 8 PhD students from M&S successfully defended their theses. Our department currently includes 2 PhD students. Furthermore, we are currently co-promotor of more than 17 PhD students in various research schools (CAPHRI, GROW, NUTRIM and MHeNs).
Our innovative statistical ideas were awarded by diverse grants as main applicants (NWO Open competition, Veni, Aspasia, ZonMW, KULeuven PhD partnership) and our statistical expertise by grants as co-applicants (ZonMW, Dutch Cancer Society, ESCRS).
Research themes
Our main research themes include:
- Bayesian methods and missing data handling in cost-effectiveness trials (including cluster randomised and multicentre trials)
- Causal inference and causal mediation analysis
- Intensive longitudinal data analysis
- Longitudinal data analysis (mixed/multilevel regression)
- Machine learning
- Multiple imputation methods for missing data in nested study designs and in meta-analyses
- (Multivariate) (Bayesian) longitudinal data analysis with latent structures
- New methods for detecting person by situation interactions
- Optimal design for regression-based norming of tests and questionnaires
- Optimal design and sample sizes for nested and cross-classified trial designs
- Optimal design and sample sizes for reliability and agreement studies
- Propensity scores
- Reliability and agreement studies for intensive longitudinal data
- Structural equation modeling and growth curve models
R packages and Shiny apps
To help researchers applying our new methodologies we develop R packages and Shiny apps:
Software |
Name |
Link |
Description |
R |
missingHE |
|
Handling of missing data in cost-effectiveness analysis through a Bayesian approach |
R |
multiagree |
|
Compare several (multilevel) kappa coefficients |
R |
REMAXINT/E-ReMI |
|
Studying row by column interaction by means of two-mode clustering |
R |
SamP2CeT |
Available as supplement to: |
Power and sample size calculation for two-level cost-effectiveness trials |
R |
simpleagree |
|
Statistical inference for agreement on binary scales |
R shiny |
Simple agree |
|
Statistical inference for agreement on binary scales |
Collaborations
We are also interested to collaborate with other researchers from in- and outside the statistical field about the development of methods to answer their particular research questions. The result of such collaboration can go from co-authorship to joint grant applications and promotions.
We are part of the FHML research institutes CAPHRI (programs PHPC and ALTC), SHE and of the Graduate School of Psychology & Neurosciences.