Emotional expression recognition and attribution bias among sexual and violent offenders: a signal detection analysis

Research with violent offenders has consistently shown impaired recognition of other’s facial expressions of emotion. However, the extent to which similar problems can be observed among sexual offenders remains unknown. Using a computerized task, we presented sexual and violent offenders, and non-offenders, with male and female expressions of anger, disgust, fear, happiness, sadness, and surprise, morphed with neutral expressions at varying levels of intensity (10, 55, and 90% expressive). Based on signal detection theory, we used hit rates and false alarms to calculate the sensitivity index d-prime (d′) and criterion (c) for each emotional expression. Overall, sexual offenders showed reduced sensitivity to emotional expressions across intensity, sex, and type of expression, compared with non-offenders, while both sexual and violent offenders showed particular reduced sensitivity to fearful expressions. We also observed specific effects for high (90%) intensity female faces, with sexual offenders showing reduced sensitivity to anger compared with non-offenders and violent offenders, and reduced sensitivity to disgust compared with non-offenders. Furthermore, both sexual and violent offenders showed impaired sensitivity to high intensity female fearful expressions compared with non-offenders. Violent offenders also showed a higher criterion for classifying moderate and high intensity male expressions as fearful, indicative of a more conservative response style, compared with angry, happy, or sad. These results suggest that both types of offender show problems in emotion recognition, and may have implications for understanding the inhibition of violent and sexually violent behaviors.