Novel Word Learning from Social Robots by Autistic and Non-Autistic Children
This publication is included in the Autistic Autism Scholarship Project. one author of this publication is identified as autistic in the project.
About the autistic author markerSommer, K., Russell, S., & Paynter, J. (2026). Novel Word Learning from Social Robots by Autistic and Non-Autistic Children [Preprint]. Center for Open Science. https://doi.org/10.31234/osf.io/nf6mh_v2
This publication is integrated into AutiHub through:
Authors
Abstract
This pre-registered study examined whether autistic preschool children could learn novel words from a social robot during a shared eBook reading session, how their engagement with the robot relates to word learning, and whether their performance differs from that of non-autistic peers. Thirty-one autistic preschoolers from an early intervention service completed a structured robot-mediated eBook session introducing four novel words, followed by a 12-trial two-alternative forced-choice receptive test, with engagement coded from video across five channels (Proximity, Eye Contact, Touch, Communication, and Emotion). Comparison data from 38 non-autistic children were collected using identical methodology in daycare settings, and all hypotheses were pre-registered (OSF: https://osf.io/sg79u/?view_only=55e46ce76b4c41c18eecade810277599). Word identification accuracy was significantly above chance in both groups, with no significant group difference. Pre-registered generalized linear mixed models found no significant association between any engagement channel and word learning in autistic children, although an exploratory group × touch interaction indicated that greater touch-based engagement was associated with lower accuracy specifically in autistic children. Non-autistic children showed higher eye contact and communication engagement, but the groups did not differ in overall engagement. Among autistic preschoolers able to participate in structured robot-mediated tasks, word learning appears comparable to that of non-autistic peers. The absence of engagement–learning associations suggests either that conventional behavioral engagement metrics do not capture what drives learning, or that structured story-based robot sessions support learning independently of engagement, both of which warrant mechanistic research before such interventions are scaled.
Bibliography cited by this reference
Cited references are imported from external metadata sources when they are available. The list may be partial.
No cited bibliography is currently integrated for this reference. This does not necessarily mean that no bibliography exists.