AI-Enhanced Speech Recognition for Advancing Pronunciation Skills

Abstract

This research explores the effectiveness of Artificial Intelligence-powered Speech Recognition Technology (AI-SRT) in supporting the development of English pronunciation and oral communication skills among English as a Foreign Language (EFL) learners in Saudi Context. It also examines learners’ perceptions and attitudes toward integrating AI-SRT into their educational routines. Adopting a quasi-experimental design with pre- and post-assessments, complemented by a mixed-methods survey that included Likert-scale items and open-ended prompts, the study engaged a group of EFL learners who systematically used AI-SRT tools during a defined instructional phase. Quantitative outcomes were derived from standardized measurements before and after the intervention, capturing objective gains in pronunciation precision and speaking fluency. Descriptive statistics assessed central tendency, variability, and score distributions, while inferential tests—specifically paired-sample t-tests—determined the significance of observed improvements. Meanwhile, qualitative data from learners’ open-ended responses underwent thematic analysis to extract recurring themes and key perspectives on their experiences with AI-SRT. The results highlight AI-SRT’s considerable potential in EFL contexts, revealing marked improvements in learners’ pronunciation accuracy and oral proficiency. Additionally, participants generally reported positive impressions regarding the technology’s ease of use and instructional benefits. These findings offer valuable insights for language educators, curriculum planners, and educational technology designers striving to create AI-informed interventions tailored to EFL learners’ needs in spoken language development.