An investigation of inductive learning in autism

Fei Xu and her colleagues conduct research on early cognitive and language development in typically developing infants and children. Using a novel approach of conceptualizing autism as a learning disorder, the researchers aim to investigate the capacity of children with autism in making inductive generalizations.

They plan to use tasks developed with typically developing children to test children who have autism. Given that children with autism have difficulties generalizing learned strategies to novel situations, the researchers’ main goal is to see if these children have a deficit in inductive generalizations.

The first series of studies focuses on the children’s ability to make inductive generalizations (for example, the ability to use small amounts of data to draw general conclusions and make inferences at a higher level) in the context of learning new words. The second series focuses on their ability to engage in probabilistic reasoning (for example, to test their intuitions on basic probability tasks).

The proposed project has major implications for theories of autism. If autism is a learning disorder and children with autism are impaired in inductive generalization and probabilistic reasoning across domains, this would provide an explanation for how these basic learning mechanisms may be responsible for the well‐known difficulties these children have in language learning and social interaction. This line of research also has major implications for clinical psychology and the design of intervention programs. Training children with autism to better use data to make generalizations could improve their reasoning in multiple domains.