A systematic review of statistical learning in autism spectrum disorder.

Most included samples were in the average cognitive range, so the review says little about autistic people with below-average cognition. No included study covered minimally verbal autistic people, although the paper notes many autistic people remain minimally verbal. Samples were male-biased, and few studies reported race, ethnicity, or socioeconomic status. Studies used different tasks, matching rules, instructions, stimuli, and age ranges, so differences between studies are hard to compare. Most studies did not report effect sizes, so the review cannot say how large any differences were. Neuroimaging findings were inconsistent and did not identify a stable brain mechanism. The studies were cross-sectional, so they cannot show whether learning differences cause language or social difficulties.

The review was funded by the National Institutes of Health, United States. The supplied text did not report other conflicts of interest.

Do not conclude that autistic people have a general pattern-learning deficit that should guide therapy. Behavioral studies mostly showed typical learning, and this review did not test whether changing pattern learning improves communication or social outcomes.