AI Speaking Practice and English Literature Students’ Fluency, Interaction, and Anxiety: A Qualitative Study
Speaking is the skill that many learners of English as a foreign language (EFL) find hardest to practice, and voice-based AI speaking applications now offer a private space for doing so. This qualitative descriptive study explored how English Literature students experience such applications with regard to speaking fluency, interactional competence, and speaking anxiety. Ten purposively selected students who had used voice-based AI speaking applications took part: four used a pronunciation-feedback application, two a conversation-practice application, and four a general-purpose application with a voice feature. Data came mainly from semi-structured interviews, supplemented by observation notes and voluntary reflective journals, and were analyzed through reflexive thematic analysis. Three themes were developed around one central idea, private practice with friction: a private space that lowers the cost of speaking; rehearsal of repair strategies, with initiative left largely to the AI; and friction from word-finding demands, recognition errors, heavy feedback, and demanding topics. The findings suggest that participants valued the applications as rehearsal spaces, while their perceived progress remains self-reported. Proposals for teachers and program leaders are offered, and directions for future research are discussed.
Authors
- Latifah
- Widhya Putriana
- M.Pd Nurul Anam (ORCID: https://orcid.org/0009-0001-2288-7584)
Publication Details
- Journal
- Academic Journal Perspective Education Language and Literature
- Published
- 2026-11-14
- DOI
- https://doi.org/10.33603/perspective.v14i2.13002
- Primary Topic
- EFL/ESL Teaching and Learning
- Type
- article
- Field-Weighted Citation Impact
- 0.00