Document Type : Original Research Paper-English Issue

Authors

1 English Department, Shahid Rajaee Teacher Training University, Tehran, Iran

2 M.A. Student, Department of English Language Teaching, Faculty of Humanities, Shahid Rajaee Teacher Training University (SRTTU), Tehran, Iran

10.22061/tej.2025.12480.3295

Abstract

Background and Objectives: Flipped instruction (FI) inverts the traditional lecture-homework model, engaging learners with content before class through a variety of technologies. AI in this regard can bring adaptability and interactivity to both the pre-class and in-class phases, especially in understanding scientific concepts in higher education courses. While there is growing research on the value of AI-assisted FI in subjects such as science and engineering, the impact of this revolutionary instructional practice in teacher education courses remains open to further research. This study employs a quasi-experimental design to investigate the effects of AI-assisted FI on the understanding of technical and scientific concepts in the course Psychology of Language Learning.

Materials and Methods: The participants included three groups of BA students who enrolled in the course Psychology of Language Learning (n=73). Group 1 (n=26) received AI-assisted FI, where pre-class instructional content was prepared by NotebookLM, an AI-powered research and writing tool. Group 2 (n=25) received conventional FI, where pre-class instructional content included the instructor’s PowerPoints with voiceovers. Group 3 (n=22) received conventional instruction utilizing a lecture-based instructional approach. In-class phase activities included quizzes, group/pair work, completion of task sheets, question-and-answer activities, and oral discussions. Post-class reinforcement included summary writing, transcribing, and generating concept maps. The participants’ achievement in the course and understanding of the technical concepts were assessed by the researcher-made midterm and final exams.

Findings: A two-way Multivariate analysis of Variance (MANOVA) was used to compare the participants’ achievement in the course and understanding of the technical terms. The results illustrated a significant difference between the three groups in general achievement and in both the midterm and final exams of the course with a strong effect size. Tukey’s HSD test showed that Group 1, who experienced AI-assisted FI, outperformed both Groups 2 and 3 in midterm and final exams. It was also found that Group 2, who learned the technical concepts using conventional FI, outperformed Group 3, who participated in a traditional and lecture-based course, in both exams. No difference was observed between male and female students.

Conclusions: The significant improvement in conceptual understanding among students who experience AI-assisted FI suggests that integrating AI tools, such as NotebookLM, can meaningfully enhance learning experiences by providing personalized, adaptive, and interactive pre-class content. This implies a shift in pedagogical design, from traditional, instructor-centered delivery models to learner-centered environments where students actively engage with content before class. Moreover, the use of AI in pre-class instruction supports differentiated learning by accommodating individual pacing and comprehension levels, thus promoting educational equality. For instructors, AI tools reduce the need for repetitive content delivery, enabling them to focus on facilitating higher-order learning, critical thinking, and collaborative in-class activities. From an institutional perspective, the successful application of AI-assisted FI in this study can reform curriculum development, faculty training, and the integration of AI tools into learning. Notably, while this study focused on educational psychology in ELT, the model has broad potential for transferability to other fields, particularly those requiring mastery of complex or technical concepts, such as engineering education, STEM curricula, or health education.

Keywords

Main Subjects

CAPTCHA Image