Objectives: We wanted to investigate the ability of children with ASD in extracting subtle emotional cues from a context and reading these expressions from dynamic facial expressions or word labels.
Methods: Two groups of 24 boys, individually matched for age (M = 11.79 y, range between 9.38 y and 14.04 y) and full scale IQ (M = 107.96, range between 83.25 and 132.25) were tested. Both groups were matched on group level for verbal IQ en performal IQ. One group had received a clinical diagnosis of ASD based on a multidisciplinary assessment and met DSM-IV-TR PDD criteria. Children with attention deficits or using medication were excluded. The typically developing group was representative for the general population. Each trial consisted of an auditory and visually presented social context (‘story’), eliciting either a subtle or an intense expression of anger or happiness. This story was followed by a test screen, consisting of either two dynamic facial expressions or two word labels (stimulus type). The answer possibilities always consisted of a match and a mismatch item. The mismatch item had either the correct intensity, but incorrect emotion (emotion-mismatch-trial) or the incorrect intensity, but correct emotion (intensity-mismatch-trial). Participants had to indicate how the protagonist in the story felt.
Results: No large group differences were found. The ASD group performed more slowly, but not worse than the typically developing group. Task difficulty largely depended on the specific task requirements. As expected, subtle emotions and intensity-mismatch-trials were more difficult than intense emotions and emotion-mismatch-trials. A trend for a three-way-interaction between mismatch, intensity and group was found. Further analysis pointed out that in the subtle trials mismatch only seemed to have an effect in the typically developing group and not in the ASD group. Happiness was responded to more accurately than anger but only in the more difficult subtle trials, not in the intense trials. Dynamic facial expressions were responded to less accurately than word labels. Older children performed better than younger children. A learning effect was observed in both groups. No correlation was found between the score on the ‘Development, Diagnostic and Dimensional Interview (3DI) subscale ‘social interaction’ and performances on this task.
Conclusions: We did not find large difficulties in extracting subtle emotions from an auditory and visually presented social context. Task difficulty and group differences depended on the specific task requirements. Slower responses and slightly less influences of task difficulty on performance could indicate rule-based response strategy in the ASD group. In the ASD group, we did not find a correlation between the score on the 3DI subscale ‘social interaction’ and performances on this task.