Exploring the Acceptance of Large Language Models as an Integrated Reading Tool: A UTAUT-Based Analysis

Abstract

Recent advances in large language models (LLM) have introduced new possibilities for computer-assisted language learning. However, empirical studies on integrating ChatGPT or other LLMs into language learning platforms remain limited. In response to this gap, the present study examines the acceptance of an LLM-assisted reading platform. In this platform, LLM is used to generate glossary, translations, assessment questions, and to provide instant assistance through an embedded Chatbot. A post-usage survey based on the extended Unified Theory of Acceptance and Use of Technology, with the additional constructs of perceived intelligence and task-technology fit, was administered to 175 undergraduates in China, following 1 month of platform use. PLS-SEM analysis indicated that usability-related constructs, specifically effort expectancy, and facilitating conditions, didn’t significantly influence undergraduates’ behavioral intention to use the platform. In contrast, given LLMs’ flexible alignment with diverse reading tasks, perceived intelligence and task-technology fit emerged as crucial drivers of sustained engagement, alongside other significant performance-oriented and affective factors, such as performance expectancy and hedonic motivation. Furthermore, it was observed that social influence also had a considerable effect on undergraduates’ behavioral intention of using that platform. These findings offer important implications for the design and application of LLM-assisted educational technologies, highlighting the importance of learners’ performance objectives, playful features, and social drivers.

Ähnliche Arbeiten