Interaction in Synchronous Computer-Mediated Communication: The Effects of Interlocutor, Task, and State Anxiety

Abstract

There is an increasingly large body of research that has addressed how interaction via Synchronous Computer-Mediated Communication (SCMC) may support second language (L2) development (see Ziegler, 2016 for a review). Various empirical studies in SCMC have demonstrated that type of interlocutor (e.g., Liu, 2017) and type of task (e.g., Blake, 2000; Yilmaz & Granena, 2010) can significantly impact the amount of negotiation that occurs during interaction. Nevertheless, the number of empirical studies is limited and they vary widely in their methodology. In addition, other researchers have suggested that one of the possible advantages of SCMC over other modes of communication is its potential to reduce L2 learner anxiety (e.g., Abrams, 2003).

The present study aims to contribute to instructed second language acquisition (ISLA) research by employing two types of tasks along with three types of interlocutors in SCMC to address whether 1) production of language-related episodes (LREs) on task, 2) amount of talk on task, and 3) L2 lexical recognition and production scores over time are related to type of interlocutor, type of task, and learners’ state anxiety. It included 82 adult intermediate L2 learners of Spanish who collaborated with either a peer, professor, or native speaker via Zoom instant messaging to complete 1) an information gap task and 2) a decision-making task. After completing each experimental task, they completed a state anxiety questionnaire adapted from Baralt and Gurzynski-Weiss (2011).

Statistical analyses revealed that a combination of Type of interlocutor and Type of task differentially affected the production of LREs and the amount of talk on task. In terms of participants’ lexical recognition and production performance over time, the analyses indicated that while Type of interlocutor played a limited role, participants were significantly more accurate in their recognition and written production of the decision-making task target items. It was also found that interlocutor type and task type were not predictors of participants’ overall state anxiety. Ultimately, these findings provide a better understanding of the roles of interlocutor status and the type of task in text-based SCMC and their impact on subsequent recognition and written production of L2 lexical forms.