Implicit learning seems to come naturally for children with autism, but not for children with specific language impairment: Evidence from behavioral and ERP data.

The groups were small, especially the specific language impairment group, so real differences could have been missed. Several apparent differences, including slower deterministic responses in specific language impairment and better verbal reports in autism, did not survive correction for multiple comparisons. One child with autism and one child with specific language impairment contributed incomplete data, only to probabilistic analyses. The brain-response components were hard to identify in children, and the time windows and labels were uncertain. The behavioral blocks were large and may not have detected subtle learning differences, such as a slower learning rate. The task used plane directions and button presses, not speech or language.

Do not read the different brain responses as proof that children with specific language impairment cannot learn patterns. They responded as quickly as typical children on the task, and the study did not test speech or language therapy.