Prediction Leads to Increased and Sustained Priming: A Self‐Replication and Extension of Grüter et al. (2021)

Abstract Active engagement in prediction has been shown to enhance learning across domains. In a small‐scale study, Grüter et al. (2021) found preliminary evidence that engaging second language (L2) learners in a “guessing game” task that required predicting a partner's utterances increased their own production of a grammatical construction (double‐object datives) that was consistently used by that partner. In a preregistered close replication and extension of Grüter et al. (2021), we fully replicate this finding with first language (L1) Korean, Japanese, and Chinese L2 learners of English and additionally demonstrate that the critical effects are maintained in a delayed posttest. Yet, contrary to observations in the original study, proficiency was not a robust modulating variable. These findings contribute to the theoretical understanding of prediction error as a mechanism for L2 learning and provide an empirical foundation for potential implementations of the guessing game paradigm in evidence‐based pedagogical practice.