A study led by researchers at the Hebrew University of Jerusalem suggests that autism may affect how the brain evaluates feedback and updates expectations rather than how accurately it initially perceives sensory information. The findings were published by scientists from the Edmond and Lily Safra Center for Brain Sciences and the university’s Department of Psychology.
The research involved 33 autistic and 35 non-autistic adults who completed an auditory pitch-discrimination task. Participants listened to pairs of tones, identified which had the higher pitch and then received visual feedback indicating whether their answer was correct. Their brain activity was monitored continuously with 64-channel electroencephalography (EEG).
Both groups achieved approximately 70% accuracy, indicating comparable performance in basic pitch discrimination. The researchers nevertheless identified three differences in the way participants processed feedback and adjusted later decisions.
After incorrect responses, autistic participants showed a weaker feedback-related negativity, or FRN, an EEG signal associated with processing negative feedback. Their feedback-related neural responses also changed less according to whether a perceptual judgment was difficult or relatively obvious.
The study further found that autistic participants made less use of recent, easy and correct trials when adjusting subsequent responses. In comparison, non-autistic participants placed greater weight on those high-confidence outcomes when updating their perceptual decisions.
The researchers repeated the testing with 51 participants one to two years later. The follow-up reproduced all three patterns: reduced FRN responses, limited neural sensitivity to task difficulty and weaker behavioral adjustment after reliable successful trials.
Because the three measures were not significantly correlated within participants, the authors concluded that they likely represent separate neurocognitive processes rather than one general impairment. When combined in a machine-learning model, the measures classified autistic and non-autistic participants with approximately 80% accuracy. The researchers emphasized that the model is intended to describe neurocognitive variation and is not a stand-alone diagnostic tool.
According to the study, these differences may help explain why adapting to changing situations can be more challenging for some autistic people. The findings point to differences in the evaluation of experiences and in how past information guides future decisions, while basic auditory accuracy remains comparable between groups.
Sources: Neuroscience News.