Abstract/Summary

Many theories have been proposed to explain the social characteristics of autistic people. In this thesis, I aimed to test claims of two theoretical accounts: one that proposes domain-specific differences in social reward responsivity, and another that proposes broader, domain-general differences in predictive processing. In addition, given the increased risk of autistic people experiencing anxiety, and the often-reported overlap between autistic social characteristics and features of social anxiety, I aimed to examine whether features of anxiety moderated either of these constructs as assessed through behavioural tasks. To address these aims, in Chapter 1, I begin by reviewing the evidence for atypical social reward responsivity and predictive processing in autistic people and how these associate with anxiety. In Chapter 2, I outline my positionality and the methodological approach of this thesis, as well as the challenges and mitigations that occurred in the conduct of the research. In Chapter 3, I sought to understand social reward responsivity in neurotypical adolescents and its relationships to autistic traits and anxiety. I found increased social reward responsivity in neurotypical adolescents and reduced social orienting and simple social choice in adolescents with elevated autistic traits, however autistic traits were not related to making less effort to choose social stimuli. Additionally, social orienting and seeking behaviours were not associated with features of social anxiety. In Chapter 4, I aimed to examine both social reward responsivity and predictive processing in autistic and non-autistic adolescents, and the relationships of these constructs to autistic traits and anxiety. I used a modified social orienting task to index a component of social reward responsivity, and a non-social associative learning task to index predictive processing. I found that autistic adolescents and those with higher autistic traits showed reduced social orienting, and some evidence that autistic and non-autistic adolescents had similar predictive processing. Again, in both tasks, there were no significant moderating effects of anxiety on task behaviour. In Chapter 5, I sought to investigate whether there were differences in social predictive processing between autistic and non-autistic adults. I used a social associative learning task and found that autistic adults adjusted their behaviour based on environmental context to a lesser extent than non-autistic adults. However, this was not the case for learning about social information. There were some differences in autistic adults’ predictive processing, but these were strongly influenced by the type of information and task design. Once more, there were no significant effects of anxiety. Finally, in Chapter 6, I discuss findings from Chapters 3-5 and how they relate to these theories and what the implications are for future research. In sum, this PhD presents some evidence for reduced social reward responsivity in autistic adolescents and atypical predictive processing in autistic adults. Yet, evidence for each theory is notably influenced by task design and there was strikingly no significant effect of anxiety across tasks. Therefore, several recommendations are made on how future studies should consider social reward responsivity, predictive processing, and anxiety in relation to autistic social differences. Namely, (1) carefully consider component processes of reward responsivity when designing tasks and follow-up findings with autistic adolescents to see whether reduced social orienting is related to effortful social seeking; (2) consider whether it is useful to measure traits of anxiety when task designs are rewarding in nature, instead follow up as to whether state anxiety influences social responsivity and predictive processing; (3) conduct predictive processing based tasks under social and non-social contexts to understand the effects of completing a task with another person.