Autistic people outperform neurotypicals in a cartoon version of the Reading the Mind in the Eyes. Liam Cross, Andrea Piovesan, Gray Atherton. Autism Research, July 20 2022. https://doi.org/10.1002/aur.2782
Abstract: Prior research suggests that while autistic people may demonstrate poorer facial emotion recognition when stimuli are human, these differences lessen when stimuli are anthropomorphic. To investigate this further, this work explores emotion recognition in autistic and neurotypical adults (n = 196). Groups were compared on a standard and a cartoon version of the Reading the Mind in the Eyes test. Results indicated that autistic individuals were not significantly different from neurotypicals on the standard version. However, autistic people outperformed neurotypicals on the cartoon version. The implications for these findings regarding emotion recognition deficits and the social motivation account of autism are discussed and support the view of socio-cognitive differences rather than deficits in this population.
Lay Summary: The Reading the Mind in the Eyes test and a cartoon version were tested on autistic and neurotypical adults. Autistic adults were not significantly different on the original test compared to neurotypicals, but they outperformed neurotypical adults on the cartoon version.
DISCUSSION
Results showed that autistic people did not perform significantly worse on the RME than NTs. While this was not as hypothesized, it may be that the high proportion of female participants boosted ASC performance. Previous work, for instance, has shown that autistic females do not perform differently to NT females (Holt et al., 2014) and that there is a female advantage in the RME (Kirkland et al., 2013). Additionally, this may be explained by adding definitions to the RME to remove variance associated with vocabulary differences. In line with our predictions, individuals with ASC outperformed NTs on the cartoon version of the task. These results suggest that autistic individuals may lack a human-specific specialization in this domain seen in NTs. For instance, NTs performed better on the RME than the C-RME, while this was not the case for autistic individuals, whose performance on the two versions did not significantly differ. This is the first study to find that autistic people outperform NTs on the RME.1 Our results suggest that NT people have more difficulty than autistic people when the agent being evaluated is anthropomorphic. Our findings also suggest that autistic people are perhaps not experiencing the same processing deficits as NTs when taking the C-RME. Why might this be?
We speculated that this could be driven by either a reduced motivation for actual human agents or an increased motivation to evaluate cartoon agents. The descriptive statistics in the present study suggest that the interaction seen in the present study was likely driven not by autistic individuals over performing on the C-RME but by NTs underperforming on the C-RME. Additionally, descriptive statistics relating to the measure of difficulty suggested that the autistic individuals found both tasks more difficult than NTs on average. However, this difference was not significant at the 0.05 level.
If NTs were to be used as a ‘benchmark’ for FER development, it would appear that there is a specialization for the human in typical development. As a result, anthropomorphic FER is more difficult for NTs. Considering that autistic people do not see a reduced performance on such measures, they may not experience the same specialized interest or aptitude for human FER. Interestingly, this does not lead to the deterioration of anthropomorphic FER. Specifically, autistic performance was not lower across both conditions, and it did not follow the same pattern as NTs. As such, it may be that autistic people also have an enhanced ability to perform anthropomorphic FER. This ability may develop through protracted engagement with anthropomorphic agents, as is suggested through research on restricted interests. The enhanced development of anthropomorphic FER in autistic people would contrast with research suggesting that FER deficits only increase as autistic people age (Lozier et al., 2014). Instead, perhaps autistic people, and those with high autistic traits, continue to develop anthropomorphic FER and ToM which allows them to eventually surpass NTs. This continued development may explain why studies on very young autistic children show impaired performance in both human and nonhuman face recognition (Chawarska & Volkmar, 2007), while studies on older autistic children, adolescents and adults show an intact or even relatively enhanced ability for nonhuman performance (for a review, see Atherton & Cross, 2018).
Autism research is rife with studies showing that autistic people enjoy engaging with the nonhuman and may be doing so increasingly throughout development, whether it be through animation (Holmgaard et al., 2013), contact with pets (Atherton et al., 2022), animal-assisted therapy (O'Haire, 2013), or even embodying the nonhuman during online game-play (Stendal & Balandin, 2015) (for a review, see Atherton & Cross, 2018). This type of engagement may allow autistic people to develop social expertise and derive social pleasure in ways that do not rely on human specialization, which may function as an extension of how autistic people begin to see themselves as more than human (Davidson & Smith, 2009). Future research should look to better understand autistic people's motivations for interacting with anthropomorphic agents. Cross et al. (2019) and Carter et al. (2016) suggest that using anthropomorphic agents in therapeutic contexts may also improve social understanding and connection. Including such agents in virtual settings and observing changes in responsiveness would be a valuable avenue for future research.