Assessment of emotional experience through facial expression

At present, self-reports are the most common way of assessing emotional experience, in large part because they are inexpensive, reliable, and easy for participants to understand. But self-reports induce a reflective style of judgment that can cause irrelevant information and incorrect theories to influence responses. Analysis of facial expression represents a complementary method of assessing subjective feelings that may not be subject to the same biases. This dissertation examines the relationships between self-reports, facial expression, and emotion-evoking stimuli in order to determine the feasibility and utility of assessing emotional experience through facial expression. In Study 1, participants viewed emotion-evoking films and reported feelings while their facial expressions were recorded. Relationships between facial expressions and self-reported emotion were complex, with many expressions linked to multiple emotions. In aggregate, however, facial expressions were a reliable predictor of average participant-reported valence. In addition, the history of facial expressions a participant had made predicted valence ratings better than his/her present facial expression. In Study 2, participants won $0, $4, or $15 while their facial expressions were recorded. Facial expressions distinguished between those who won different amounts at better than chance levels, outperforming the guesses of naïve coders, but were relatively weak predictors of participant happiness. Taken together, the studies suggest though there are limitations, assessment of facial expression has substantial potential as a tool for investigating emotional experience. Facial expression and self-report share significant variance, which would allow expression to validate the results of self-report, particularly in cases where the results of self-report might be called into question by known biases. The two methods also appear to have distinct influences, suggesting that facial expression analysis would provide unique insight into emotional experience.