When prototypes are not best: judgments made by children with autism.

The current study used a factorial comparison experimental design to investigate conflicting findings on prototype effects shown by children with autism (Klinger and Dawson, Dev Psychopathol 13:111-124, 2001; Molesworth et al., J Child Psychol Psychiatry 46:661-672, 2005). The aim was to see whether children with high-functioning autism could demonstrate prototype effects via categorization responses and whether failure to do so was related to difficulty understanding ambiguous task demands. Two thirds of the autism group did show an effect. The remainder, a sub-group defined by performance on a control task, did not. The discussion focuses on the influence of heterogeneity within the autism group and the ability to resolve ambiguity on task performance. Finally, an alternative experimental design is recommended for further research into these issues.