[PS-2.29] Commonly used statistical learning tasks do not capture stable individual differences in children: examining reliability across modalities

Arnon, I.

Psychology Department, Hebrew University

Infants, children and adults are capable of using statistical learning (SL) to extract information about their environment. While most studies look at group-level performance, a growing literature examines individual differences in SL and their predicted positive relation to learning outcomes. Several studies have found positive correlations between SL performance and language outcomes in children and adults. However, these studies use tasks designed to explore group-level performance without modification, resulting in psychometric shortcomings that impact reliability in adults. Even though similar measures are used to assess individual differences in children, no study to date has examined their reliability in development, limiting our ability to interpret their reported correlations with other measures. Here, we examine the reliability of three commonly used SL measures (two auditory and one visual) in children (mean-age 8;2) and adults. While the tasks showed moderate reliability in adults, they did not capture stable individual variation in children. None of the tasks were reliable across sessions, and all showed internal consistency measures well below psychometric standards. These findings raise significant concerns about the use of current SL measures to predict and explain individual differences in development. We discuss possible explanations for the difference in reliability between children and adults