Spontaneous use of generative artificial intelligence in preparing for college English test: GenAI adoption factors, learners' motivation, and self-efficacy

Abstract

This study investigates the adoption and impact of Generative Artificial Intelligence (GenAI) tools on English as a Foreign Language (EFL) learners, focusing on self-efficacy, motivation, and predicted exam performance. Using a mixed-methods approach, the research combines survey data (N = 595) and interviews (N = 8) from Chinese university students preparing for the College English Test 4. Quantitative analysis shows that positive attitudes, perceived usefulness, and ease of use predict GenAI usage frequency, while intrinsic motivation negatively correlates with use. GenAI use frequency did not significantly affect predicted test scores or self-efficacy levels. Younger students used GenAI more frequently. Qualitative data reveals that students primarily use GenAI as supplementary learning resources, which, while beneficial, do not substantially enhance self-efficacy or predicted exam performance. The study founds a knowledge gap in the effectiveness of these tools in EFL contexts. It suggests strategies for optimizing GenAI in EFL learning, including blending with traditional methods, providing user training, balancing technology with traditional approaches, ensuring equal access, and customizing GenAI for education. The findings imply a need for educators to supervise GenAI use and for institutions to facilitate effective integration. Future research should examine long-term and cultural effects of GenAI use in education.