Proceedings of the Annual Meeting of the Cognitive Science Society

Joint Nonnegative Matrix Factorization (JNMF) is a methodfor factor analysis that is capable of simultaneously decom-posing two datasets into related latent state representations.Enabling factor analysis for contrasting applications, i.e., tofind common and distinct structural patterns in data, JNMF hasgreat potential for use in the field of cognitive science. Appliedto experimental data, JNMF allows for the extraction of com-mon and distinct patterns of behavior thereby extending theoutcomes of traditional correlation-based contrasting methods.In this article, we introduce JNMF to the field of cognitive sci-ence and demonstrate its potential on the exemplary domainof syllogistic reasoning by comparing reasoning patterns fordifferent personality factors. Results are interpreted with re-spect to the theoretical state of the art in syllogistic reasoningresearch.