Multivariate response patterning of fear and anger

This study examined whether differential response patterns are associated with the emotions of fear and anger as compared to neutral and physical action imagery conditions in 27 right-handed male volunteers screened for imagery ability and alexithymia. Fear, anger, neutral, and action states Were significantly differentiated on the basis of autonomic measures and facial electromyographic (EMG) responses. Discriminant function analyses (DFAs) using autonomic and facial EMG measures conducted separately on data from two experimental sessions revealed 78% and 72% correct classification of fear and anger, and neutral conditions, respectively. Cross-validation of the DFA using discriminant weights of the first session to classify the second session data led to a 66.5% correct classification of the fear and anger, and neutral imagery conditions. The combined average of both sessions revealed an 84% correct classification rate. Further, when subjective ratings were added to the discriminant function as a manipulation check for the experimental procedure, a 99% correct classification rate was obtained. These data indicate that emotion-specific response patterns for fear and anger are accurately differentiable from each other and from neutral imagery conditions. The findings provide support for the cognitive-physiological network models of physiological emotion specificity.