IMMEDIATE EFFECTS OF AN AI-BASED SPEAKING PLATFORM (SPEECHACE) ON THAI EFL STUDENTS’ SPEAKING PERFORMANCE

Sasithorn  Noppraphai; Yun  Luo; Napasri  Suwanajote

Sasithorn Noppraphai Faculty of Humanities and Social Sciences, Suan Sunandha Rajabhat University, Thailand
Yun Luo Faculty of Humanities and Social Sciences, Suan Sunandha Rajabhat University, Thailand
Napasri Suwanajote Faculty of Humanities and Social Sciences, Suan Sunandha Rajabhat University, Thailand

Keywords: Artificial intelligence, EFL speaking, Speechace

Abstract

Abstract— This study examined the immediate effects of an AI - based speaking platform, Speechace, on English speaking performance among Thai EFL university students . A quasi - experimental one - group pretest -posttest design was employed with 60 undergraduate English major students at a public university in Thailand . Participants completed a speaking task via the Speechace we bsite before and immediately after receiving automated AI - generated feedback focusing primarily on pronunciation and fluency . Speaking scores generated by the system were analyzed using descriptive statistics and paired - sample t- tests. In addition, a quest ionnaire was administered to explore students ’ perceptions of the platform in terms of usability, perceived impact, and satisfaction. The findings indicated a statistically significant increase in students’ post-test speaking scores compared to their pre - test performance, suggesting immediate performance gains following AI - assisted feedback. Questionnaire results revealed positive student perceptions toward Speechace, particularly regarding its usefulness for speaking practice and ease of use . While this study does not claim long-term improvement in overall speaking proficiency, the results highlight the potential of AI- supported platforms to facilitate short - term speaking performance adjustment and learner engagement in EFL classroom contexts . The study contributes classroom- based evidence from a Southeast Asian EFL setting and suggests that AI - based speaking tools can serve as effective supplementary resources for speaking practice rather than as standalone assessment instruments.