AI-Supported Inside Reading for Academic Word List Development: Effects on Vocabulary Learning and Sentence Writing in an Azerbaijan, EFL University Classroom

Abstract

The current study analyzed the impact of using AI-supported inside reading on the acquisition of the words of the Academic Word List (AWL) and sentence writing (SW) performance of EFL university students in Azerbaijan. Sixty undergraduate students were randomly assigned to an experimental and a control group, following a quasi-experimental pre-test/post-test control group design. The study drew upon the theory of AI-mediated feedback and adopted a combination of structured academic reading and the Involvement Load Hypothesis from the field of vocational education to ensure that both receptive and productive lexical development were encouraged using AI-mediated feedback. The experimental group significantly outperformed the control group in gain scores (in AWL) (F(1,57) = 358.72, p < .001, η²p = .863), showing an average gain score of 10.83 points (43.4%) while the control group gained an average score of 2.87 points (11.6%). Sentence writing outcomes followed the same pattern (F(1, 57) = 452.20, p < .001, η²p = .888), with experimental gains of 12.50 points (39.9%) substantially exceeding control gains of 2.30 points (7.6%). A high correlation between AWL–SW (r = .74, p < .001) in the experimental group was obtained, which indicated that receptive and productive vocabulary development occurred simultaneously. The results of the research support the use of the inside reading pedagogical approach with the assistance of artificial intelligence as an effective method in academic vocabulary teaching in the post-Soviet university context. The limitations are a single-site sample and the lack of delayed retention measurement.