AI-Driven Gamified and Non-Gamified Platforms in EFL Listening: Comparative Effects of Duolingo and Replika on Comprehension, Engagement, and Motivation

Abstract Listening comprehension is not only a vital pillar of English language acquisition but also one of the most underexplored skills in adolescent EFL education. This study offers a novel, head-to-head comparison of two AI-powered platforms with contrasting pedagogical designs—Duolingo, a gamified system built around rewards and progress tracking, and Replika, a non-gamified, conversation-based AI. Addressing a clear gap in the literature, we examine how these divergent approaches influence high school learners’ listening comprehension, engagement, and motivation. Using a concurrent mixed-methods design, we worked with 53 Iranian high school students (aged 14–15), randomly assigned to a Duolingo group (n = 27) or a Replika group (n = 26) in a 12-week, 24-session program. Quantitative data came from researcher-developed pre- and post-tests and perception questionnaires; qualitative insights were drawn from classroom observations and semi-structured interviews (n = 15). Results revealed that while both groups achieved significant gains in listening comprehension, Duolingo’s adaptive gamification consistently drove higher engagement and motivation. Replika, in contrast, nurtured more sustained, reflective listening through personalized dialogue. These findings show that design architecture—whether gamified or conversational—matters for how AI can address long-standing gaps in EFL instruction. Beyond improving listening comprehension, the study illustrates how targeted integration of AI-driven strategies can help educators, curriculum designers, and EdTech developers create more engaging, context-sensitive, and sustainable language learning experiences.