It is not settled yet within the scientific community whether autistic children have a category-learning impairment. Studies have shown that, while children with ASD can form categories based on rules, they have trouble doing so based on prototypes. However, no experiment has been carried out on the abstraction of prototypes in people with autism. Research has found, though, that autistic children tend to focus on random rather than important, 'meaningful cues' this is called 'stimulus overselectivity'. I would like to design an experiment in order to see if participants with ASD hierarchize features during category-learning, and if they do, how so. The results could help shed light on both the phenomena of overselectivity and impaired prototype-abstraction in people with ASD. This project is a first step towards a revised approach to helping autistic children deal with novelty and generalize information, two problems which are common across the entire spectrum.