생성형 AI의 유학생 학술 글쓰기 및 교과 과제에서의 지속적인 사용의도 연구 : -기대 확인, 지각된 유용성 및 만족도를 중심으로-

다국어 초록 (Multilingual Abstract)

Generative artificial intelligence （AI） has become an important learning support tool in higher education by assisting students with academic writing, content generation, language revision, and idea development. For international students studying in South Korea, who often face linguistic and academic challenges in non-native language environments, generative AI provides valuable support for completing academic writing and course assignments. However, limited research has examined the factors influencing international students’ continuance intention to use generative AI in academic contexts.
This study investigated the relationships among expectation confirmation, perceived usefulness, satisfaction, and continuance intention based on the Expectation-Confirmation Model （ECM）. Data were collected from 362 international students at Korean universities who had experience using generative AI for academic writing and course assignments. Structural equation modeling （SEM） was employed to analyze the proposed research model.
The results showed that expectation confirmation had significant positive effects on perceived usefulness and satisfaction. Perceived usefulness positively influenced both satisfaction and continuance intention, while satisfaction also had a significant positive effect on continuance intention. However, the direct effect of expectation confirmation on continuance intention was not statistically significant. In addition, satisfaction significantly mediated the relationships between expectation confirmation and continuance intention and between perceived usefulness and continuance intention. The sequential mediating effect of perceived usefulness and satisfaction between expectation confirmation and continuance intention was also statistically significant.
These findings suggest that international students’ continuance intention to use generative AI is formed through the interaction of cognitive evaluation and emotional experience. This study extends the application of the Expectation-Confirmation Model to the context of generative AI in higher education and provides practical implications for developing AI-supported learning environments and educational support strategies for international students.

목차 (Table of Contents)

• Ⅰ. 서론 1
• 1. 연구의 필요성 및 목적 1
• 2. 연구 문제 3
• 3. 용어의 정의 5
• Ⅱ. 이론적 배경 8
• 1. 유학생의 학술 글쓰기와 생성형 AI 활용 8
• 2. 기대확인모형의 구조와 적용 11
• 3. 주요 변수의 개념과 변수 간 관계 14
• 4. 연구모형 및 가설 도출 19
• Ⅲ. 연구방법 24
• 1. 연구 대상 24
• 2. 측정 도구 26
• 3. 연구 설계 및 절차 30
• 4. 자료 처리 32
• Ⅳ. 연구 결과 34
• 1. 예비분석 34
• 2. 기술통계 및 상관분석 36
• 3. 주요 변수 간의 구조적 관계 37
• 4. 만족도의 매개효과 41
• 5. 지각된 유용성과 만족도의 순차적 매개효과 42
• Ⅴ. 논의 45
• 1. 기대확인, 지각된 유용성, 만족도 및 지속적인 사용의도의 구조적 관계에 대한 논의 45
• 2. 만족도의 매개효과에 대한 논의 47
• 3. 기대확인의 지각된 유용성과 만족도를 통한 순차적 간접효과에 대한 논의 49
• 4. 종합 논의 51
• Ⅵ. 결론 및 제언 53
• 1. 요약 53
• 2. 결론 54
• 3. 제언 55
• 참 고 문 헌 57
• 부록 61
• Abstract 73