AI- Assisted Academic Writing in ESL Context: A Corpus- Based Linguistic Analysis

Abstract

This paper analyses the use of artificial intelligence in language education in the field of ESL academic writing based on theory of Second Language Acquisition and Corpus Linguistics. It seeks to research the effectiveness of AI-assisted writing aids in the academic writing performance of the learners and particularly how AI affects lexical diversity, syntactic complexity, grammatical accuracy, cohesion and writing quality. The study adopted the qualitative case study approach, analyzed undergraduates at Government Graduate College, Zafarwal who were chosen through purposive sampling. Data were gathered using AI-assisted and non-assisted writing samples, unstructured interviews and think-aloud protocol. The linguistic analysis, using corpus methods, evaluated lexical density, syntactic richness, cohesiveness and accuracy whereas thematic analysis was used to evaluate perception, confidence, and metalinguistic knowledge of the learners. The results have shown that AI-aided writing was more efficient concerning lexical diversity, sentence complexity, sentence structure, and cohesion and the students had more self-belief, motivation and perceived awareness of the language usage. According to think-aloud procedures, AI facilitate decision-making, but excessive dependence will result in paraphrasing and low creativity rates. These findings indicate that AI inclusion in strategies and educational methods can be effective to achieve learning outcomes, growth in autonomy, and improvement of creativity and analytical thinking. The study concludes that AI-supported technologies may be an effective method to improve ESL academic writing and curriculum building and further studies in AI-enhanced language learning.

Keywords: AI in writing, writing academic English, Corpus analysis, perception of learners, cognitive scaffolding, complex syntax, lexical variety.