The impact of a chatbot-assisted flipped approach on EFL learner interaction

Abstract

Flipped learning is gaining prominence in EFL classrooms, where it has been shown to provide various advantages. However, the literature has not adequately addressed the integration of L2 interaction outside the classroom into flipped learning design. Accordingly, given the ubiquitous and interactive nature of chatbot technology, we designed a chatbot-assisted flipped approach to increase learners’ opportunities for L2 practice. To examine the impact of this approach, we investigated how the use of chatbots as conversational partners in the out-of-class phase of flipped learning affected students’ interactive performance and perceptions of interaction. We randomly assigned 87 elementary school EFL learners to experimental (n = 42) and control (n = 45) groups. For six weeks, both groups were taught using the same flipped approach except that after viewing a video, the experimental students practiced a pre-class dialogue with chatbots while the control group practiced the dialogue using a worksheet. Employing a sequential, explanatory mixed-methods design, we collected two sets of quantitative data (pre- and post-interactive speaking tests and an interaction questionnaire) and one set of qualitative data (semi-structured interviews). The quantitative data analyses showed that the experimental group outperformed the control group in terms of interactive speaking performance and the degree to which they perceived they had participated in learner-learner and learner-instructor interactions. Thematic analysis of the qualitative data supplemented the quantitative results by elucidating how the chatbots facilitated the learners’ interactive speaking performance and perceptions of interaction. This study highlights the potential of chatbot technology to bridge the out-of-class and in-class phases of flipped learning.

Keywords

• flipped learning
• artificial intelligence
• chatbot
• customized chatbot
• interaction