Do individuals with autism spectrum disorder not experience the uncanny valley? A psychological experiment and feature analysis using human and robot faces
Recently, there have been growing expectations and attempts to use robots in the therapeutic treatment of individuals with autism spectrum disorder (ASD). However, the uncanny valley effect, which refers to the discomfort or eeriness people feel when robots appear almost human, but not quite realistic, poses a challenge when designing such robots. Previous studies have been inconsistent regarding how individuals with ASD perceive the uncanny valley in response to agent faces. It is important to use images of robots to determine whether the uncanny valley occurs for individuals with ASD and whether robots can be used for their therapy. Therefore, we used robot and human facial images as experimental stimuli to investigate the uncanny valley effect in typically developing individuals and those with ASD. Thirty-two individuals with ASD (30 men, 2 women; 32 analyzed) and 47 typically developing individuals (31 men, 14 women, 2 unspecified; 31 analyzed) participated in the study. The results show that the typically developing individuals exhibited the uncanny valley effect, whereas the individuals with ASD showed a less distinct effect. Those with ASD were less likely to perceive eeriness in face images that were ambiguous in classification as human or robot, leading to a reduced sense of eeriness. Furthermore, an analysis of face image features suggested that individuals with ASD placed more emphasis on local information rather than global information to determine their likability of the face images, which explains their results well. Finally, the therapeutic application of robots for individuals with ASD is discussed. • Uncanny valley (UV) is a negative reaction to an almost humanlike object in shape. (82) • There is no consensus on the UV of people with autism spectrum disorder (ASD). (79) • We used human and robot faces to show that UV was not distinct in persons with ASD. (83) • Image feature analysis revealed that persons with ASD focused on local information. (83) • Those with ASD have difficulty integrating local information thereby preventing UV. (83)
Authors
- Lu Li (ORCID: https://orcid.org/0009-0006-8839-8659)
- K. UEDA
- Naoki Nishikawa
- Hirokazu Kumazaki
- Taku Imaizumi
Institutions
- Tokyo University of Information Sciences (JP)
- Tokyo University of the Arts (JP)
- Nagasaki University (JP)
- The University of Tokyo (JP)
Publication Details
- Journal
- Computers in Human Behavior Reports
- Published
- 2026-01-05
- DOI
- https://doi.org/10.1016/j.chbr.2026.100927
- Primary Topic
- Autism Spectrum Disorder Research
- Type
- article
- Field-Weighted Citation Impact
- 0.00