Full Program »
42. Developing and validating a facial emotion recognition task with graded intensity
Facial emotion recognition (FER) biases and deficits have been theorized to account for social impairment in psychopathology. However, the FER tasks used in psychopathology research (a) have psychometric limitations and (b) provide poor tests of theories. The present study aimed to develop and validate an FER task appropriate for psychopathology research (timed, subtle expressions, etc.). Photographs from the NimStim set of facial expressions were selected, based on previous reliability studies, and then digitally altered to produce five levels of expression intensity for each emotional face, resulting in over 700 stimuli. In Sample 1 (N = 224), undergraduates viewed subsets of these faces, each for 1.0 seconds, then labeled the face neutral, angry, happy, sad, surprised, disgusted, or fearful. Item-total correlations were examined and used to produce a final task that included 380 of the original stimuli, balanced for gender and ethnicity. In Study 2, participants (N = 230) completed the final task and an array of measures aimed to demonstrate convergent (e.g., Reading the Mind in the Eyes Test) and discriminant validity (e.g., Balloon Analogue Risk Task). Preliminary analyses indicated internally consistent emotion-level indices and that, as expected, subtle emotional expressions were more difficult for participants to detect.