Ashkenazi, P. 1 , Ashkenazi, P. 1 & Sasson, A. 1, 2
1 Bar Ilan University
2 Tel Aviv University
The study examines the effect of stimulus diversity on artificial grammar learning (AGL). Although research in language acquisition and cognitive learning has demonstrated the contribution of stimulus diversity to learning, no experiment has examined it in a classic visual AGL task. In this study we explore whether learning and transfer ability are affected by the diversity level of training sequences. Fifty-nine adults were divided randomly into two experimental conditions: high stimulus diversity, in which participants were presented with sixty different letter sequences, and low stimulus diversity, in which participants were presented with twenty letter sequences, each appearing three times. Both groups classified test strings for their grammaticality in a baseline learning test and a transfer test. Results of the d' measures in the learning task indicated that the high stimulus diversity group performed significantly better than the low stimulus diversity group. The transfer data indicated that performance exceeded chance only in the high stimulus diversity group. Individual-level data, combined with the attribution analysis indicated that those exceeding chance level attributed more grammaticality classifications to familiarity, while those performing below chance reported relying more on intuition. Taken together, our results highlight the central role of stimulus diversity in AGL phenomena.