This project used a Representational Similarity Analysis (RSA) approach to test how well subject-specific models of emotion concept knowledge predicted subject-specific models of emotion perception. In Studies 1-2 the perceptual spaces were estimated using computer mouse-tracking and in Study 3 the perceptual spaces were estimated using reverse-correlation. This OSF project includes all summary data necessary to re-run our analyses. All analyses reported in the paper were multi-level GEE models implemented using the GENMOD procedure in SAS. See GEE_example.SAS in Code for example syntax. Face stimuli for Studies 1-2 were from the NimStim database of facial expressions. I do not have the rights to publish those faces here, but please see https://www.macbrain.org/resources.htm. The words and phrases used to estimate conceptual models of emotion in Studies 2 and 3 (based on a ratings task) are hosted in Materials. Materials for Study 3 hosted here include the "base face" used in the reverse-correlation task as well as the reverse-correlated classification images (CIs) from our set of subjects. The compare_CIs.m MATLAB code can be used to calculate the visual similarity (measured as Pearson correlation distance) between each subject's pair of CIs. CIs were generated using the RCICR package in R, the details of which are explained in this accessible tutorial: http://www.rondotsch.nl/rcicr/.