Conversational topic maintenance and related cognitive abilities in autistic versus neurotypical children

Keeping a conversation going is the social glue of friendships. The DSM criteria for autism list difficulties with back-and-forth conversation but does not necessitate that all autistic children will be equally impacted. We carried out three studies (two pre-registered) with verbally fluent school children (age 5-9 years) to investigate how autistic and neurotypical children maintain a conversation topic. We also investigated within-group relationships between conversational ability and cognitive and sociocognitive predictors. Study 1 found autistic children were more likely than neurotypical controls to give off-topic and generic minimal responses (e.g. 'mm', 'oh') and were less likely to give non-verbal responses (e.g. nodding or use of facial affect to respond). Nonetheless, the autistic group provided topic-supporting responses 62% of the time, indicating some aspects of conversation topic maintenance are a relative strength. Studies 2 and 3 found large individual differences in topic-supporting conversational responding among both neurotypical and autistic children. These were positively related to theory of mind ability and age in both groups. Conversational skills lie on a continuum for the general population and differences by diagnostic group are a matter of degree. Given the importance for peer relationships, we suggest a whole classroom approach to supporting conversation skills in all children.Lay abstractChildren who struggle to maintain conversation with peers often have fewer friends and lower popularity ratings, which can affect wellbeing. Verbal social communication more broadly is linked to both behavioural difficulties and emotional problems. We carried out three studies to examine children's ability to provide responses which keep a back and forth conversation going. The first study found that while autistic children had on average greater difficulties than their neurotypical peers with certain aspects of conversation topic maintenance, for other aspects the autistic group showed considerable strengths. Both studies 2 (neurotypical children) and 3 (autistic children) found relationships between, on the one hand, conversational ability, and on the other, the ability to consider another's viewpoint and the ability to maintain and update information in short term memory. We suggest support for social conversation skills should be part of mainstream classroom curricula for autistic and neurotypical children alike.