Effects of Blended AI-Assisted and Teacher-Led Instruction on EFL Students’ Academic Writing Development and Writing Self-Efficacy at Salale University, Ethiopia
Generative artificial intelligence has entered Ethiopian higher education faster than pedagogical frameworks for its use. This study examined the effects of a blended model combining AI-assisted feedback with teacher-led instruction on EFL students’ academic writing development and writing self-efficacy. A one-group pre-test–post-test quasi-experimental design with an embedded qualitative strand was used. Thirty-seven third-year students of the Department of English Language and Literature at Salale University, enrolled in the Intermediate Writing course, received ten weeks of blended instruction in which AI tools supplied immediate draft-level feedback and the teacher supplied genre, content and revision instruction. Teaching materials were modified from the existing course textbook rather than newly prepared. Writing was assessed with an adapted analytic rubric and self-efficacy with the Second Language Writer Self-Efficacy Scale. Paired-samples t-tests showed significant gains in overall writing, from M = 51.38 to M = 64.22, t (36) = 10.00, p < .001, d = 1.64, and in writing self-efficacy, from M = 2.84 to M = 3.73, t (36) = 9.33, p < .001, d = 1.53. Gains were largest for content and organization and smallest for mechanics. Writing and self-efficacy gains correlated moderately, r = .46, p = .004.
Authors
- Yonas Tamiru Gebre (ORCID: https://orcid.org/0009-0006-2132-7928)
Institutions
- Salale University (ET)
Publication Details
- Journal
- Research Square
- Published
- 2026-11-20
- DOI
- https://doi.org/10.21203/rs.3.rs-11367514/v1
- Primary Topic
- EFL/ESL Teaching and Learning
- Type
- preprint