- Enhancing EFL Speaking Skills through AI-Supported Literary Works: A Quantitative Study on Drama, Poetry, and Short Stories

Abstract

Recent advancements in artificial intelligence (AI) have opened new avenues for enhancing English?as?a?foreign?language (EFL) instruction. Integrating AI?driven literary tasks promises to enrich speaking practice by providing personalized feedback and authentic language exposure. However, empirical evidence on the efficacy of such interventions remains limited. This study examined the impact of a 12?week AI?assisted literary intervention on EFL speaking proficiency among university learners.  Twenty?eight undergraduate participants (mean age?=?20.4?years) completed a pre?test/post?test design. Speaking performance was assessed using a standardized rubric covering pronunciation, grammar, vocabulary, fluency, and interaction. Scores were analyzed with paired?sample t tests and Cohen’s d for effect size. Additionally, semi?structured interviews captured learners’ qualitative perceptions of the intervention.  Mean speaking scores increased from 12.86?(SD?=?2.14) in the pre?test to 24.07?(SD?=?1.03) in the post?test, reflecting an 11.21?point gain (t(27)?=?42.47, p?<?.001) and a large overall effect size (d?=?1.55). Sub?scale analyses revealed significant improvements across all dimensions (p?<?.001), with the greatest effects on vocabulary (d?=?1.72) and interaction (d?=?1.64). Qualitative feedback highlighted enhanced vocabulary acquisition, increased confidence, immediate corrective feedback, heightened motivation, and appreciation for the personalized nature of the AI support. The modest sample size and absence of a control group constrain the generalizability of findings and preclude definitive attribution of gains to the AI component. Uniformly large effect sizes raise concerns regarding potential ceiling effects or measurement bias. The AI?assisted literary approach demonstrates promising potential for boosting EFL speaking proficiency, yet future research should employ larger, randomized controlled trials to validate efficacy, explore long?term retention, and assess scalability across diverse learner populations.