Iris Willigers

Iris Willigers

Methodology and Statistics
Tilburg School of Social and Behavioral Sciences
Tilburg University

Email
Website

Project
Examining Variation in Causal Effects in Psychology

Statistical tools for the investigation of causal effect variation assume normality of a sample. However, using psychological scales as covariate or outcome variable could create variation because of scaling issues and failures of measurement invariance across groups and/or conditions possibly reflecting multidimensionality (Millsap, 2011; Stroevenbelt et al., 2022; Wicherts et al., 2005). Distributional assumptions underlying statistical tools for studying effect variation are often violated for outcomes composed of sum scores (Micceri, 1989). In addition, measurement invariance is often not tested in psychological studies as been found by Maassen et al. (2022). They also found that only 29% of the investigated studies reached the required level of measurement invariance.

This IOPS project will investigate current practices of scaling problems and measurement invariance from a psychometric perspective. We will also work on resolving the problems that arise with the variation in causal effects due to scaling problems (e.g. diminished observed Cohen’s d) and measurement invariance.

Supervisors
Prof. Dr. Jelte Wicherts
Dr. Marjan Bakker

Financed by
Nederlandse Organisatie voor Wetenschappelijk Onderzoek (NWO)

Period
1 September 2024 – 1 September 2028