Abstract :
This study explored the effectiveness of an Artificial Intelligence (AI)-powered program for fostering Cognitive Academic English Language Proficiency (CALP) among STEM (Science, Technology, Engineering, and Mathematics) college students. The study utilized quasi-experimental research method (pretest-posttest control group design). Sixty participants enrolled in STEM program at the Faculty of Education, Minia University were randomly assigned to two intact groups: a treatment group (n=30) and a non-treatment group (n=30). Both groups participated in pre- and post-tests to gauge their CALP levels. The participants in the treatment group were trained and instructed using an AI-powered program designed by the researchers whereas their counterparts in the non-treatment group did not receive such intervention as they received regular instruction. Instruments of the study included a test of English academic language proficiency and wh-communication prompts in interviews, a test of English formal academic learning, and a test of English- speaking skills with a scoring checklist of specific speaking subskills. The findings revealed that the participants in the treatment group significantly surpassed their counterparts in the non-treatment group in the post-performance of the test of English formal academic learning, and the test of English-speaking skills. Valuable insights into the potential of AI-powered interventions were contributed. Additionally, the development of effective and innovative language learning tools tailored to the specific needs of STEM college students for promoting their academic English language proficiency were introduced. Suggestions for further research and recommendations were also presented.