Using a novel AI-powered observation approach, a groundbreaking study reveals that eye contact avoidance may not be unique to autism

Children with autism are believed to face difficulties in social interactions, besides also lacking the ability to be attentive while interacting with others. In fact, eye contact avoidance is a key behavioral marker in the clinical diagnosis of autism. However, most developing children also seldom make direct eye contact during everyday play interactions, calling into question the applicability of this behavioral marker in diagnoses.

Conducting further investigations to decode the behavioral indicators of autism, a new study led by Lu Qu and Qiaoyun Liu at East China Normal University’s Shanghai Institute of Artificial Intelligence for Education revealed that autistic children exhibit social attention patterns similar to their typically developing peers during play, with a primary focus on toys rather than faces. Published online in the ECNU Review of Education on March 17, 2025, their findings challenge the longstanding assumption that reduced eye contact is a definitive marker of autism.

According to the researchers, most conventional studies use an artificial clinical setting to measure children’s joint attention skills. These tests involve presenting stimuli, such as toys, to children and observing their gaze patterns to assess their ability to follow and initiate joint attention. However, these tests are usually conducted in standardized lab environments and may not fully reflect children’s performance in natural settings.

Tiding over these conventional methods, the authors in this study utilized a novel, non-intrusive AI-powered observation lab to analyze gaze, vocalizations, and movements in natural settings—addressing limitations of traditional lab-based assessments like the Autism Diagnostic Observation Schedule (ADOS). This approach captured authentic behaviors, revealing that standardized tests may overlook critical social communication strategies. The research involved multimodal behavioral analysis of children across three groups: typically developing children, autistic children, and children with developmental delays. Ethical approval was obtained from East China Normal University, with funding from the China Postdoctoral Science Foundation and National Key Research and Development Program.

The results of this study were quite surprising. Both autistic and typically developing children spent 60%-80% of their playtime focused on toys and only 6%-14% looking at adults’ faces, suggesting that eye contact avoidance may not be unique to autism only. Nonverbal communicative cues, such as caregivers’ hand movements, were also found to play a significant role in joint attention during interactions. These results align with recent studies in Current Biology (2022), which found that children rarely look at faces during natural play, regardless of autism diagnosis.

These findings may have important implications for understanding attention spans in children with autism. According to the authors, the long-held belief that autistic children avoid eye contact may be exaggerated. While autistic children are known to look at their parents’ faces less often than typically developing children, this difference may not be that significant. Additionally, these findings suggest that children may use other communication cues to participate in social settings, especially during play-related activities.

Moreover, these findings could have significant implications for clinical practice as well. Many clinical interventions for autistic children focus on improving attention and encouraging eye contact. However, the results of this study suggest that these interventions may not be suitable for all autistic children, especially for play-related activities. Instead, the authors suggest that using alternative interventions, such as parents’ hand positions and communicative gestures, could be more effective.

“Our findings emphasize the need to rethink interventions focused solely on eye contact,” say lead researchers Qu and Liu. Adding further, they claim, “Targeting broader cues, such as gestures, could better support autistic children’s communication development.” The study calls for updated diagnostic frameworks and personalized intervention strategies that reflect the diverse ways in which children engage socially.

According to Qu and Liu, “In this era of artificial intelligence, our understanding of the core symptoms of autism needs continuous updating. Classical research paradigms must be re-examined, and continuous exploration and research must be conducted to understand children’s natural behaviors in social communication, especially in a natural environment.” Going ahead, the researchers hope that studies like this help improve and provide appropriate and effective support for autistic children.

Reference
Title of original paper: Is a Child Who Doesn’t Look at People Always Autistic? - A Closer Look at Joint Attention
Journal: ECNU Review of Education
DOI: https://doi.org/10.1177/20965311251319050