Autistic and Neurotypical Children’s Social Impressions of Off-Topic and Delayed Responding

Background:

On average, autistic groups are more likely than non-autistic groups to respond in an off-topic manner (Sng et al., 2020), and/or to leave significantly longer turn-taking gaps (Ochi et al., 2019). Exhibiting unconventional conversational behaviours could result in negative social impressions of autistic children from their neurotypical peers (e.g., Place & Becker, 1991). However, it is possible that autistic individuals do not make the same social judgments due to different communicative preferences (Granieri et al., 2020).

Objectives:

This study aimed to determine whether autistic children pattern similarly or differently to their neurotypical peers in rating off-topic or delayed responding as a deterrent to friendship or interaction, and if not, whether they are aware of the societal dispreference for these behaviours.

Methods:

Current results are based on twenty-eight participants (N=14 per group), but the final (pre-registered) sample will include thirty-six neurotypical and thirty-six autistic children. All participants are aged between 9;0 and 13;11-years, and groups are matched on chronological age, gender ratio, core language ability, and non-verbal reasoning skills. Participants listened to several short audio clips of conversations containing six conversational turns between two children. For each conversation, participants rated four statements on a 0-100 scale: two capturing their personal preferences, such as “I would enjoy chatting to the [target speaker]”, and two regarding their perceptions of “Most other people’s” impressions. A conversation production measure was also used to observe participants’ tendency to provide off-topic or delayed responses themselves.

Study 1 manipulated the typicality of the target speaker’s responses via the content of their utterances (On-topic vs. Off-topic). One example of an off-topic response to the statement “I went to that new restaurant in town last night” is “Oh no, I think my library books are due in today”. Study 2 manipulated the timing of the target speaker’s conversational turns. That is, the responses were all on-topic, but either came 200-msec after the offset of the preceding utterance (Typical) or 3000-msec after (Delayed).

Results:

Linear mixed effects models were conducted to compare participants’ ratings for On-topic vs. Off-topic conversations, and for Typical vs. Delayed conversations. In Study 1, participants rated On-topic conversations significantly more favourably (M = 73.67) than Off-topic conversations (M = 46.37) (p<.001). Similarly, in Study 2, there was a significant main effect of Timing (Non-delayed M = 73.67, Delayed M = 64.68, p=.014). For both studies, there was no interaction, and no effect of Diagnostic Group (Autistic M = 59.97, Neurotypical M = 63.17), with a small effect size for both the Topic (d=.02) and Timing (d=.19) analyses.

Conclusions:

The absence of interaction effects and main effects of Diagnostic Group (with very small effect sizes) suggests that autistic children are just as likely as their neurotypical peers to dis-prefer speakers who provide ‘atypical’ conversational responses. While it is possible that some autistic adults may develop a preference for atypical communication styles, it appears that the preferences of autistic children are not only indistinguishable from that of neurotypical children, but also align with their perceptions of societal norms.