Effectiveness of Deep Learning–Based Reading Platforms on EFL Reading Comprehension: A Quasi-Experimental Study

Abstract

This study aimed to examine the effectiveness of a deep learning–based reading platform in improving EFL students’ reading comprehension achievement. A quantitative approach was employed using a quasi-experimental design, specifically the non-equivalent control group design. The sample consisted of two classes of university students (N = 60), divided into an experimental group and a control group. The experimental group used a deep learning–based reading platform equipped with adaptive reading features, difficult vocabulary analysis, AI-generated questions, and automated feedback, while the control group received conventional teaching. A 30-item reading comprehension test, validated and tested for reliability, served as the research instrument. Data were analyzed using paired sample t-tests, independent sample t-tests, and effect size calculation. The results revealed a significant improvement in the experimental group’s reading comprehension scores (p < 0.05). Furthermore, a significant difference was found between experimental and control groups in the post-test results, with an effect size of 1.30, indicating a large effect. These findings demonstrated that the deep learning–based reading platform was more effective than traditional instructional methods in enhancing EFL students’ reading comprehension.