Schedule of Events | Symposia

Sketchpad Series

Statistical Learning with Inner Speech Suppression – Behavioral and ERP Evidence from an Artificial Grammar Learning Task in Children

Poster Session E - Monday, March 31, 2025, 2:30 – 4:30 pm EDT, Back Bay Ballroom/Republic Ballroom
Also presenting in Data Blitz Session 1 - Saturday, March 29, 2025, 10:30 am – 12:00 pm EDT, Grand Ballroom.

Ziyi Cao¹ (zcao@mghihp.edu), Asiya Gul¹, Lauren Baron¹, Shivani Patel¹, Annika Schafer¹, Kelsey Black¹, Yael Arbel¹; ¹MGH Institute of Health Professions

Statistical learning is characterized as an implicit, automatic, and unconscious process. However, there are arguments that explicit processing may not be entirely excluded and might play a supplementary or distractive role in commonly-used statistical learning tasks. This study aimed to evaluate implicit learning in school-age children (8:0 – 12:0) by employing two strategies to suppress explicit processing in a visual Artificial Grammar Learning (AGL) task. First, stimuli were replaced with non-namable, abstract shapes to minimize verbalization that could aid memorization. Second, participants were randomly assigned to an inner-speech-suppression group (S) or a non-suppression group (NS) for comparison. The suppression group was required to repeatedly vocalize an irrelevant word (e.g., "tea") during the learning phase to suppress inner speech and working memory. During the learning phase, participants were exposed to sequential patterns governed by an underlying finite structure. In the subsequent testing phase, they judged whether new patterns were grammatical. A follow-up test one week later assessed knowledge consolidation. Preliminary behavioral results revealed that S and NS groups performed with similar accuracy, significantly above chance (p < .05). Accuracy levels were maintained across the one-week interval, with the S group showing a trend of improvement. ERP analyses, time-locked to both pattern completion and violation points, showed a P600 effect in response to grammatical violations at completion points in both groups, but at violation points only in the S group . Additionally, the S group exhibited an FN400 effect at both time points, which was absent in the NS group.

Topic Area: LANGUAGE: Development & aging