In our new paper out today, autistic adults held a “get to know you” conversation with an unfamiliar autistic or typically-developing (TD) person. We were curious: would social interaction outcomes differ when their partner was also autistic? THREAD journals.sagepub.com/doi/full/10.11…
Most studies attempting to understand social disability in autism focus exclusively on individual characteristics, like social cognitive ability (e.g., theory of mind). This presumes that social interaction difficulties in autism are driven solely by the autistic person.
But social interaction by definition involves more than one person, and relational dynamics— in which each person influences and is influenced by the other— is key to understanding determinants of partner compatibility and social quality.
A relational perspective emphasizes that social ability for autistic people (or anyone) isn’t uniform but varies depending on social context and the characteristics of their interaction partner. Until recently, autism research has largely overlooked these extrinsic factors.
One of these factors may be communicative and cognitive differences between autistic and non-autistic people that can present relational barriers for mutual understanding and affiliation. @milton_damian calls this “the double empathy problem” (DEP).
The DEP now has considerable empirical support (see work by @cjcrompton @Brett_Heasman @DrGeoffBird & others). For instance, not only do autistic adults often struggle to infer the thoughts and motivations of TD adults, recent findings indicate that the reverse is true as well.
What might this mean for social interaction? A purely deficit-model of autistic sociability would predict *worse* social quality between two autistic people than between an autistic and a TD person. After all wouldn’t it include twice the number of socially-disabled people?
But the DEP would predict the opposite. Greater dispositional and neurologic similarity might facilitate better social quality between two autistic partners than between an autistic and a TD partner. Anecdotally, many autistic people often report this experience.
In this study we wanted to test this empirically. We examined whether real-world social interaction quality & first impression formation differed between unfamiliar autistic adults compared to between unfamiliar autistic and TD adults. This comparison is new w/in autism research.
Participants (67 ASD; 58 TD) were assigned to one of three partnership conditions (ASD-ASD; TD-TD; ASD-TD) in which they had a 5 minute unstructured conversation with an unfamiliar partner and then evaluated the quality of the interaction and their impressions of their partner.
We used the Actor Partner Interdependence Model to assess how each participant’s characteristics (e.g. their diagnostic status) and the interaction between them related to each participant’s evaluations of each other and the interaction. 
As we’ve found in our prior studies w/ video clips, autistic adults were rated as more awkward, less attractive, and less socially warm than TD adults by both autistic and TD partners. So some assessments of autistic adults seem relatively consistent even when formed in person.
However autistic adults weren’t rated as less intelligent or less trustworthy than TD adults, and unlike our video clip studies they weren’t rated as less likable. So direct interaction may work to lessen some negative evaluations of autism that occur during indirect observation.
This aligns with previous work suggesting that direct experience with (and knowledge about) autism can reduce stigma and improve TD first impressions of autistic people. Unfortunately such opportunities may be denied to them when negative judgments are made from afar.
TDs also rated conversations w/ autistic & TD partners to be of similar quality. Thus negative TD evaluations of autistic adults did *not* translate to lower social quality, which suggests that TD impressions are based on social presentation differences not conversational content
Importantly, although trait assessments of autistic adults did not differ for autistic and TD partners, interest in future social interaction did. 🚨 TD but *not* autistic participants preferred future interaction with TD partners over autistic partners. 🚨
Autistic adults actually trended towards the opposite, preferring future interaction w/ other autistic adults over TD adults. They also reported disclosing more about themselves to autistic partners than to TD partners, and felt closer to their partners than did TD participants.
These findings support the DEP and suggest that social interaction difficulties in autism are not an absolute or inherent characteristic of the individual. Rather, social quality is a relational characteristic and dependent on the fit between the person & the social environment.
In sum, social disability in autism is context-dependent and emerges to a greater degree within cross-diagnostic interactions. This may occur because of a mismatch in cognitive & communication styles that may improve w/ increased familiarity & acceptance. We’re testing this now.
We hope this paper shows how studying actual social interaction in ASD elicits a deeper understanding than lab studies alone. Big thanks to Kerrianne Morrison @kmdebrabander @DesiRJones @DanielFaso & Rob Ackerman. Please contact me if you can’t access it and I’ll send you a copy
