Students’ Perspectives on the Use of Artificial Intelligence （AI） in EFL Reading and Writing Instruction = 영어 읽기 및 쓰기 교육에서의 인공지능（AI） 활용에 대한 학생들의 인식 연구

국문 초록 (Abstract)

인공지능（AI）의 빠른 발전은 외국어로서의 영어（EFL） 교육, 특히 읽기와 쓰기 교육에 큰 변화를 가져왔다. 대규모 언어 모델（LLMs）과 지능형 챗봇을 비롯한 AI 기반 도구는 학습자에게 즉각적인 피드백과 언어 지원을 제공하며 자기주도 학습을 촉진하고 있다. 그러나 이러한 도구의 활용이 고등교육에서 확대됨에 따라 AI 의존, 학문적 정직성, 그리고 교육 현장에서의 적절한 활용 방식에 대한 관심도 함께 높아지고 있다. 이에 본 연구는 대구대학교 학부생을 대상으로, 특히 인문대학 학생들을 중심으로 영어 읽기 및 쓰기 교육에서 AI 활용 실태와 이에 대한 인식을 조사하였다. 본 연구는 양적 연구 방법을 적용하였으며, 구조화된 설문지를 통해 학부생 65명（N = 65）의 자료를 수집하였다. 수집된 자료는 기술 통계와 추론 통계를 활용하여 분석하였다. 연구 결과, 학생들은 AI가 시간 관리, 읽기 이해, 그리고 쓰기 활동을 지원하는 유용한 학습 도구라고 인식하는 경향을 보였다. 반면, AI에 대한 과도한 의존, 자신의 쓰기 능력에 대한 자신감 저하, 그리고 학문적 정직성 유지에 대한 우려도 함께 나타났다. 또한 응답자의 56.92%는 대학의 AI 활용 정책을 잘 알지 못한다고 응답하여, 관련 정책에 대한 안내와 소통이 충분하지 않은 것으로 나타났다. 이러한 결과는 AI가 언어 학습을 효과적으로 지원하기 위해서는 명확한 제도적 가이드라인과 적절한 교수·학습 방안이 함께 마련되어야 함을 시사한다. 따라서 본 연구는 AI의 책임 있는 활용을 위한 정책의 투명성을 높이고, 학생들에게 AI 활용에 대한 실질적인 지침을 제공하며, 독립적인 사고를 유지하면서 AI를 학습 지원 도구로 활용할 수 있도록 과정 중심의 평가 방식을 확대할 것을 제안한다. 아울러 디지털 근접발달영역（d-ZPD） 개념을 바탕으로 AI를 활용한 학습 지원 방안을 제시하며, EFL 교육에서 AI의 효과적이고 책임 있는 활용을 위한 교육과정 개발과 정책 수립에 실질적인 시사점을 제공하고자 한다.

다국어 초록 (Multilingual Abstract)

The rapid development of Artificial Intelligence （AI） has brought significant changes to English as a Foreign Language （EFL） teaching and learning, particularly in reading and writing instruction. AI-based tools, including Large Language Models （LLMs） and intelligent chatbots, provide learners with immediate feedback, language support, and opportunities for independent learning. At the same time, their growing use in higher education has raised questions about students' dependence on AI, academic integrity, and appropriate classroom use. This study examined undergraduate EFL students' use of AI and their perceptions of its roles in reading and writing instruction at a Humanities college in Southern part of Korea. A quantitative research design was employed. Data were collected from 65 undergraduate students （N = 65） using a structured questionnaire and analyzed through descriptive and inferential statistical methods. The findings indicate that students generally viewed AI as a useful learning tool that improved time management, reading comprehension, and writing support. However, many participants also expressed concerns about becoming overly dependent on AI, losing confidence in their own writing ability, and maintaining academic integrity. In addition, more than half of the participants （56.92%） reported that they were unaware of the university's policies regarding AI use. The findings suggest that AI is perceived to be effectively supporting language learning when it is integrated within clear institutional guidelines and appropriate teaching practices. Based on the students’ confessions revealed in the data, this paper argues for greater policy transparency, increased guidance on responsible AI use, and process-oriented assessment that encourages students to use AI as a learning support rather than as a substitute for independent thinking. These findings provide practical implications for curriculum design and AI integration in EFL instruction and support the use of an AI in the framework of Zone of Proximal Development to promote meaningful and responsible AI-assisted learning.

목차 (Table of Contents)

• CHAPTER 1: INTRODUCTION 1
• 1.1 Background of the Study 1
• 1.2 Statement of the Problem 2
• 1.3 Purpose of the Study 3
• 1.4 Research Questions 4
• 1.5 Significance of the Study 5
• 1.6 Definition of Terms 5
• 1.7. Overview of the Thesis 6
• CHAPTER 2: LITERATURE REVIEW 9
• 2.1 Theoretical Frameworks for AI in EFL 9
• 2.1.1 The Digital Zone of Proximal Development (ZPD) and the AI-MKO 9
• 2.2 AI in Reading and Writing: Pedagogical Affordances 10
• 2.2.1 Reading Instruction: Adaptive Comprehension and Differentiation 10
• 2.2.2 Writing Instruction: From Automated Evaluation to Collaborative Generation 12
• 2.3 Student Perspectives: Empowerment and Anxiety 12
• 2.3.1 Motivation, Self-Efficacy, and the Face-Saving Factor 13
• 2.3.2 The Risk of Over-Reliance and Cognitive Atrophy 13
• 2.4 The Paradox of Invisible Labor and De-skilling Anxiety 14
• 2.5 The Burden of Academic Integrity and Assessment Redesign 15
• 2.6 Pedagogical Strategies for AI-Integrated EFL Instruction 15
• 2.6.1 The Evolution of Prompt Engineering as a Core Literacy 15
• 2.6.2 The Human-in-the-Loop (HITL) and Evaluative Post-Editing 16
• 2.6.3 Critical AI Literacy and the Validation Phase 17
• 2.6.4 AI-Scaffolded Differentiated Learning and the ZPD 17
• 2.7 The South Korean Higher Education Context: A Synthesis 18
• 2.7.1 Institutional Drivers and Policy Implementation 18
• 2.7.2 The Convergence of Tradition and Innovation 19
• CHAPTER 3: METHOD 20
• 3.1 Introduction 20
• 3.2 Research Design 20
• 3.3 Research Setting and Participants 21
• 3.4 Research Instrument 21
• 3.5 Data Collection Procedures 23
• 3.6 Data Analysis 23
• 3.7 Reliability and Validity 24
• 3.8 Ethical Considerations 25
• 3.9 Scope and Delimitations 25
• CHAPTER 4: RESULTS 26
• 4.1 Overview 26
• 4.2. Demographic Information and AI Usage Patterns 26
• 4.2.1. Participant Demographics 26
• 4.2.2. Analysis of AI Tool Selection and Usage Purposes 26
• 4.3. Perceived Benefits of AI in Literacy Activities 29
• 4.4. Perceived Concerns and Risks 30
• 4.5. Institutional Guidelines and Educational Needs for AI Integration 32
• 4.6. Evaluation of Institutional Support and Future Intent (Q32–Q33) 34
• 4.6.1. The Awareness Gap (Q32) 34
• 4.6.2. Latent Demand for Specialized Training (Q33) 35
• 4.6.3. Synthesis: Moving from Passive to Active Integration 35
• 4.7. Qualitative Analysis of Student Open-Ended Responses (Q34–Q36) 35
• 4.7.1. Operational Friction, Semantic Misalignment, and Ethical Anxiety (Q34) 36
• 4.7.2. Explicit Demand for Metacognitive and Strategy-Based Scaffolding (Q35) 37
• 4.7.3. Intellectual Sovereignty and the Institutional Accountability Gap (Q36) 38
• 4.8. Summary of Findings 39
• CHAPTER 5: DISCUSSION 41
• 5.1 Overview of Key Findings 41
• 5.2. The Shift from Translation to Generative Scaffolding 41
• 5.3. The Perception Paradox: Receptive vs. Productive Skills 42
• 5.4. Polarization and the Wait-and-See Majority 42
• 5.5. Addressing the Transparency Gap 44
• 5.6. Pedagogical Recommendations and Implications 44
• 5.6.1. Establishing a Clear-Box AI Policy 45
• 5.6.2. Shifting from Output to Process-Oriented Assessment 45
• 5.7. Conclusion of the Discussion 46
• References 48
• 요약/초록 54
• Appendix 56
• Appendix A: A Survey Questionnaire for Korean students’ English Classes 56
• Appendix B: Transcribed Participant Open-Ended Survey Responses (Q34–Q36) 60
• Acknowledgement 74