Beyond over- or under-sampling: autistic children's inflexibility in sampling costly information

Abstract

BackgroundEfficient information sampling is crucial for human inference and decision-making, even for young children. Information sampling is also closely associated with the core symptoms of autism spectrum disorder (ASD), since both the social interaction difficulties and repetitive behaviors suggest that autistic people may sample information from the environment distinctively. Previous research on information sampling in ASD focused mainly on adolescents and adults, and on whether they over- or under-sample. The specific ways in which autistic children sample information, especially when facing explicit costs and adapting to environmental changes, remain unclear. MethodsWe employed an adapted bead task to investigate the sampling behavior of 24 autistic and 41 neurotypical children, matched for age and IQ. In each trial of our experiment, children gathered information about an unknown target isle by drawing samples from it and then guessed the target between two isles based on their samples. In conditions where sampling was costly, children needed to weigh the benefits of information against the costs of acquiring additional samples. Through computational modeling and intricate behavioral measures, we revealed how the two groups of children differed in sampling decisions and underlying cognitive mechanisms. ResultsUnder conditions involving costs, autistic children showed less efficient sampling than their neurotypical peers. This inefficiency was due to their increased variability in the number of samples taken across trials rather than a systematic bias. Computational models indicated that while both groups shared a similar decision process, autistic childrens sampling decisions were less influenced by dynamic changes and more driven by recent evidence, thus leading to their increased sampling variation and reduced efficiency. LimitationsTo refine ASD subtyping and correlate symptom severity with behavioral characteristics and computational findings, future research may need larger participant groups and more comprehensive clinical assessments. ConclusionsThis study reveals an inefficiency of autistic children in information sampling and tracks down this inefficiency to their increased sampling variability, primarily due to their cognitive preference for more local and static information. These findings are consistent with several influential behavioral theories of ASD and highlight the needs of a multi-level understanding of cognitive flexibility in ASD.