Dr. Li Yi：Biased mental face representations of autistic children in the general population

Abstract

Background

Autism stigma significantly impacts the quality of life for autistic individuals and their families. Methods to measure stigma, such as self-report scales, may be influenced by social desirability biases and fail to uncover the psychological mechanisms underlying stigma formation. The present study used the reverse-correlation image classification (RCIC) technique to investigate whether the mental faces of autistic children were represented more negatively than non-autistic children, and to examine potential associations between these biases and behavioral characteristics or diagnostic labels of autism.

Methods

In Study 1, participants selected between two faces (with random visual noise) that best matched their mental image of an autistic child. The selected and unselected faces were averaged to generate mental face representations, which were then rated on attractiveness, felt disgust, and behavioral avoidance. Studies 2 and 3 further examined the impact of behavioral characteristics and diagnostic labels on these mental representations by presenting participants with target persons paired with autistic behaviors or autism labels versus typical behaviors or non-autism labels.

Results

Study 1 found that, compared to mental representations of non-autistic children, participants rated mental representations of autistic children as less attractive, reported increased felt disgust toward them, and indicated a greater likelihood of avoiding them. Parents of autistic children similarly exhibited negative representations. Studies 2 & 3 revealed that arbitrary pairing of faces with autistic behaviors (vs. typical) or autism labels (vs. non-autism) led to less attractive mental images.

Limitations

The stimuli were limited to male children’s faces with neutral expressions, restricting generalizability across gender, age, and emotion. The design could not determine whether negative representations reflect autism‑specific stigma or a broader social bias triggered by perceived difference. The study did not explore individual differences or incorporate autistic perspectives in stimulus design.

Conclusions

Negative mental face representations of autism exist independently of physical traits, shaped by behavioral cues and diagnostic labels. These findings highlight a possible mechanism underlying stigma: the formation of negative mental face representations contributes to biased evaluations of autistic children. Future research could explore whether modifying such representations might offer a new pathway for supporting strategies.

Original Link：https://doi.org/10.1186/s13229-026-00709-0