The influence of facial race and facial sex on the recognition of faces morphed to vary in emotional intensity from angry to happy

The main goal is to understand how social category information (e.g., race and gender) influences the recognition of facial expressions that vary in emotional intensity. More specifically, I examine, using a neutral baseline, whether emotion recognition is facilitated for faces that belong to social categories that are stereo-typically associated with specific emotions. So, for example, the study permits a test of whether people are more likely to categorize male vs female neutral expressions as angry. This study will extend my pre-registered secondary analyses of data that was originally collected Korb, S., & Massaccesi, C. (2020, December 29; https://osf.io/29mn7/). The study by Korb and Massaccesi was also a pre-registered replication of an earlier study by Harris et al (2016). The study by Harris et al was designed to examine the effects of face sex on forced-choice angry and happy decisions for faces morphed to vary in emotional intensity from angry to happy. The research by Harris et al was novel because the authors calculated indices of psychophysical performance (e.g., Point of Subjective Equality). In this specific study, I will 1. establish the generality of the effects across face race - the initial studies examining morph intensity have used only the faces of white individuals. I will establish if similar effects hold for the faces of black individuals. One hypothesis is that people are prejudiced towards black individuals making them more likely to respond "angry" than happy when the faces are of black individuals with a neutral expression. 2. establish whether the same effects are found for a new different stimulus set Further objectives are: 1. Bayesian Modelling. I will model choice decisions in a Bayesian Hierarchical Logistic Regression Model and RTs in a Bayesian Hierarchical ex-Gaussian Regression Model. Finally, a Bayesian Hierarchical Drift Diffusion Model in which choice decision and reaction times (RT) are modeled as a joint outcome, will also be estimated.