Pre‐Task Explicit Instruction, Input Modality, and Working Memory in L2 Oral Self‐Repair

ABSTRACT Despite the central role of tasks in language education and ensuing research documenting how task‐related variables might affect language performance and learning, it remains unclear whether pre‐task explicit instruction, input modality, and working memory (WM) influence how learners monitor and repair grammatical structures in real‐time speech. The present study addresses this gap by examining the effects of pre‐task explicit instruction, task input modalities, and WM on the self‐repair behavior of second language (L2) learners during oral task performance. The participants (89 learners) were randomly assigned to four groups based on explicit instruction on English passive voice (with or without) and input modality: a story retelling task with visual and auditory input or a dictogloss task with textual and auditory input. Immediately after completing the oral tasks, participants took part in individual stimulated recall interviews, during which their global form repairs (F G ‐repairs), local form repairs (F L ‐repairs), and content repairs (C‐repairs) were identified. The results of a two‐way MANCOVA, followed by several one‐way ANCOVAs, showed a significant main effect of pre‐task explicit instruction on learners’ F L ‐repairs, whereby explicit instruction heightened attention to the target linguistic form. There was also a significant interaction between explicit instruction and WM for F L ‐repairs, showing that higher WM was advantageous in benefiting from explicit instruction. Additionally, the input modality of tasks had a significant main effect on C‐repairs, revealing that the story retelling task led to greater engagement with the communicative message of tasks. The study holds theoretical and pedagogical implications for task designers and L2 teachers.