ENHANCING METACOGNITIVE SKILLS THROUGH AI-ASSISTED PROMPTING IN EAP READING AND WRITING

Аннотация

Annotation: English for Academic Purposes (EAP) students often read and write without deliberately planning, monitoring, or evaluating their strategies, which limits the transfer of skills across disciplines. This article examines how structured prompting of generative artificial intelligence (AI) tools can strengthen learners' metacognitive skills in EAP reading and writing. Drawing on Flavell's model of metacognition, Zimmerman's cycle of self-regulated learning, and scaffolding theory, the study proposes a prompt framework in which learners use AI not as an answer generator but as a thinking partner that asks questions, models strategies, and gives feedback on process. A mixed-methods, quasi-experimental design is described for a ten-week intervention with university EAP students, combining the Metacognitive Awareness Inventory, rubric-scored reading and writing tasks, prompt logs, and reflective interviews. The article reports illustrative results showing larger gains in metacognitive awareness, reading comprehension, and writing quality in the AI-prompting group than in a comparison group, along with qualitative evidence of more frequent planning, monitoring, and revision behaviors. Examples of effective and ineffective prompts are presented in tabular form to guide classroom practice. The discussion addresses risks of over-reliance, inaccurate output, and academic integrity, and offers recommendations for task design, assessment, and teacher training. The article concludes that prompting, when explicitly taught and linked to reflection, can make invisible cognitive processes visible and so help EAP learners become more autonomous readers and writers.