Moderated (Non)-Linear Factor Analysis Now in JASP!

March 30 - 2026

To All Measurement Invariance and Item Functioning Detectives,

We have known for a while that the common approach to assess Measurement Invariance (MI) disregards quite some information, because continuous variables must often be split into discrete groups. A few years ago, Daniel Bauer and team [https://dbauer.web.unc.edu/] came to the rescue and developed moderated (non)-linear factor analysis (MNLFA), a method that takes one or multiple categorical or continuous background variables into account simultaneously, and estimates their effect in a full moderated factor model (Bauer, 2017). This allows measurement invariance to be assessed without the usual loss of information. However, up until now researchers could not apply the methodology without extensive coding in R or resorting to commercial software programs. Fortunately, that is no longer the case!! 💪🏻

We implemented MNLFA in JASP and thus provide a free and user-friendly click-and-point implementation of the methodology. Interested detectives may read our tutorial paper that guides the reader through the process of assessing measurement invariance with JASP. Preprint: https://osf.io/preprints/psyarxiv/6ftqg

References

Bauer, D. J. (2017). A more general model for testing measurement invariance and differential itemfunctioning. Psychological Methods, 22(3):507

Godmann, H. R., Molenaar, D., Ziegler, M., & Pfadt, J. M. (2025). A Tutorial on Assessing Measurement Invariance with Moderated (Non-)Linear Factor Analysis in JASP. PsyArXiv. https://doi.org/10.31234/osf.io/6ftqg_v2