AI-Enhanced Multisensory Learning Intervention to Improve Pronunciation in EFL Students

Abstract

Despite early exposure to English as a foreign language (EFL), Spanish learners often struggle with its pronunciation due to fundamental differences between the transparent orthography in Spanish and the deep orthography in English. This study examines the effectiveness of a multisensory learning (MSL) methodology enhanced by artificial intelligence (AI) tools and based on the dual-route model of reading, which distinguishes between sublexical decoding and lexical word recognition, in improving English pronunciation accuracy. 94 university students were divided into a target group and a control group. The intervention involved a pre-test and a post-test and a set of reading-aloud tasks, all conducted via the Wooflash platform and focused on five difficult English vowel phonemes: /ə/, /uː/, /ʌɪ/, /eɪ/, and /ɔː/. The objective was to compare the effectiveness of MSL and traditional training (TT), and to assess participants’ experiences of the MSL approach. Quantitative results showed that MSL was as effective as TT in improving pronunciation accuracy, while qualitative feedback from the post-activity questionnaire revealed emotional, cognitive and metacognitive benefits of the MSL intervention, including increased confidence and awareness of pronunciation challenges. These findings suggest the potential of MSL as a learner-centered and engaging tool for improving pronunciation in EFL contexts.