고등학교 영어 교과서와 ChatGPT 생성 텍스트의 어휘 및 구문 특성 비교: 코퍼스 기반 분석

• Published in: 영어어문교육 Vol. 31; no. 2; pp. 155 - 172
• Main Author: 이영희(Younghee Cheri Lee), 강준수(Junsoo Kang)
• Format: Journal Article
• Language: Korean
• Published: 한국영어어문교육학회 30.06.2025; The English Teathers Association in Korea
• Subjects: 영어와문학; 고등학교 영어 교과서; generative AI; ChatGPT; 생성형 인공지능; lexical diversity; 어휘 다양성; 코퍼스; corpus; high school English textbook
• ISSN: 1226-2889

This study empirically examines the strengths and limitations of generative AI texts as supplementary educational resources by comparing the lexical and syntactic features of high school English textbooks with those of texts generated by ChatGPT. A parallel corpus comprising 47 passages from ten government-authorized textbooks and their AI-generated counterparts was compiled for analysis. Four linguistic metrics were analyzed: number of word types, standardized type-token ratio (STTR), mean word length (MWL), and mean sentence length (MSL). WordSmith Tools 7.0 was used for corpus analysis, and statistical significance was assessed through independent t-tests. The results showed that while textbook passages contained a greater number of word types, ChatGPT texts demonstrated higher lexical diversity and longer mean word length. No significant difference was observed in sentence length. These findings suggest that AI-generated texts can offer enriched lexical input while maintaining structural clarity, thus serving as effective supplementary materials in classroom instruction. The study offers both theoretical insights and practical implications for integrating generative AI into secondary English education, especially for enhancing vocabulary exposure and expanding learner-centered reading resources. KCI Citation Count: 0