New Study Challenges Assumptions About Emotion Recognition in Autistic Individuals

A recent study involving 235 participants suggests that individuals on the autism spectrum may not have a deficit in recognizing emotions, and in some cases, may even be more accurate than neurotypical individuals when interacting with others on the spectrum. The research, published in the journal Molecular Autism on September 12, 2026, was led by Dr. Yonat Rum from the Hebrew University of Jerusalem and the University of Cambridge's Autism Research Centre.

Unlike traditional self-report questionnaires, this study employed a novel method. It involved participants watching videos of ten adults, five autistic and five neurotypical, recounting personal stories. During their storytelling, these individuals used a slider to indicate their emotional state every half-second, which was recorded. The participants then rated 12 specific emotions on a scale of 0 to 8. The 235 study participants, 141 of whom were on the autism spectrum and 94 neurotypical, were tasked with replicating these emotional ratings while watching videos of both autistic and neurotypical storytellers, without knowing their diagnostic status.

The findings revealed no significant overall difference between the groups in general emotion recognition. However, a key discovery was that participants on the autism spectrum demonstrated significantly greater accuracy in identifying the emotions of the autistic storytellers compared to the neurotypical participants. Conversely, both groups performed similarly when assessing the emotions of neurotypical storytellers. Interestingly, neurotypical participants reported feeling more empathetic and interested in the stories, but this subjective feeling did not translate into higher accuracy.

Researchers suggest these findings partially support the "double empathy theory," which posits that communication difficulties between autistic and neurotypical individuals are bidirectional, stemming from different ways of expressing and interpreting emotions. Dr. Rum stated that this perspective could shift focus from "correcting" autistic individuals to fostering mutual understanding between all parties. This theory was first proposed in 2012, suggesting that challenges in interaction arise from the interface between different neurotypes, not solely from one individual.

The study's implications are particularly relevant to Israel, where 1.8% of children are diagnosed with autism, and the number of children receiving disability benefits for autism has nearly quintupled in a decade, reaching approximately 60,000 in 2024. This highlights a large service system built on the assumption of one-sided difficulty. If the challenge is indeed bidirectional, resources might need to be reallocated from training autistic individuals to training their environment, including educators, employers, and professionals.

However, the researchers caution that the study has limitations. The assessments were based on recorded videos, not live interactions, and the autistic sample consisted of volunteers, potentially not representing the full spectrum, especially those requiring higher support. The effect size of the observed advantage was also noted as small to moderate.