Determinants of Behavioral Intention to Use Large Language Models for Learning Among Educational Technology Students in Nigerian Universities

Abstract

Large language models (LLMs) such as ChatGPT and Claude are Artificial Intelligence systems trained on massive datasets to understand and generate human-like responses to user prompts. Yet most studies on LLM adoption in higher education focus on usage patterns, with limited attention given to how factors such as trust, academic integrity concerns, performance expectancy, and facilitating conditions influence students' use of these tools for learning. This study, therefore, determined the influence of these factors on Educational Technology students' intention to use LLMs for learning. A descriptive survey design of the correlational type was adopted, with 250 undergraduates sampled through simple random sampling from two Nigerian universities. Data were collected using a validated questionnaire and analyzed using multiple regression and paired-samples t-test at 0.05 level of significance. Results showed that all four predictors significantly influenced behavioral intention, F(4, 245) = 21.35, p = .000, R² = .258, with Performance Expectancy as the strongest positive predictor (β = .371) and Academic Integrity Concern as the only significant negative predictor (β = -.193). Students also reported significantly higher intention to use ChatGPT (M = 3.11) than Claude (M = 2.75), t(249) = 6.61, p = .000. The study concludes that Educational Technology students' intention to use LLMs is formed by their beliefs about the tools' benefits, the conditions enabling use, their trust in the output, and their ethical concerns about using it, and recommends raising LLM awareness, closing access gaps, clarifying integrity guidelines, and increasing Claude exposure among Nigerian students.