Attentional shifting differences in autism: Domain general, domain specific or both?

<jats:p> Atypical attention is considered to have an important role in the development of autism. Yet, it remains unclear whether these attentional difficulties are specific to the social domain. This study aimed to examine attentional orienting in autistic and non-autistic adults from and to non-social and social stimuli. We utilised a modified gap–overlap task with schematic images (Experiment 1: autistic = 27 and non-autistic = 26) and photographs (Experiment 2: autistic = 18 and non-autistic = 17). Eye-tracking data (i.e. saccadic latencies) were then compared across condition and type of stimulus (social or non-social) using multilevel modelling. Autistic adults exhibited mostly typical gap and overlap effects, as well as a bias towards social stimuli. Yet, autistic participants benefitted from exogenous disengagement when orienting to social information more than non-autistic participants. Neither a domain general nor social domain–specific account for attentional atypicalities in autism was supported separately. Yet, subtle combined domain differences were revealed in the gap condition. </jats:p><jats:sec><jats:title>Lay abstract</jats:title><jats:p> Previous research has shown that autistic individuals look at other people less and orient to them more slowly than others. Yet, it is still unclear if this represents general visual differences (e.g. slower looking at any new information, social or not) or a uniquely social difference (e.g. only slower looking to humans but not objects). Here, we aimed to examine how quickly autistic and non-autistic adults look to and away from social (i.e. faces) and non-social information (i.e. squares and houses). We used an attentional shifting task with two images where sometimes the first image disappears before the new image appears (makes it easier to notice the new image) and other times it stays on the screen when the new image appears. In Experiment 1, we showed schematic faces and squares to 27 autistic and 26 non-autistic adults, and in Experiment 2, we showed photographs of faces and houses to 18 autistic and 17 non-autistic adults. In general, autistic adults looked at the new non-social or social images similarly to non-autistic adults. Yet, only autistic adults looked at new social information faster when the first image disappeared before the new image appeared. This shows that autistic individuals may find it easier to notice new social information if their attention is not already occupied. </jats:p></jats:sec>