New Publication: Contrast-Analysis for competing hypotheses

Excited to share our new open-access publication in Behavior Research Methods: We present a tutorial on how to compare predictions from two competing theories using contrast analysis.

The project was led by Mirka Henninger and Simone Malejka — I was happy to contribute as third author. I truly appreciated the collaboration and learned a lot from working closely with both of them.

If you work with factorial designs where two theories predict different mean patterns across groups, this method lets you directly test the difference between patterns in a single statistical test … and yes, this approach is literally — and statistically — powerful.

The tutorial shows the entire workflow step by step and includes reproducible examples in R.

For immediate application: analyses can be performed with our R package cofad (contrast analysis for factorial designs).

Article (open access): https://link.springer.com/article/10.3758/s13428-025-02833-w

R package: https://github.com/johannes-titz/cofad

Happy to hear how others are analyzing predictions from competing theories (Bayesian anyone?) or whether contrast analysis plays a role in your research.