An Analysis of Factors Influencing Chatbots on College Students’ Programming Self-learning Ability

Abstract

Against the backdrop of the deep integration of chatbots into education, this study, based on Self-Determination Theory (SDT) and the UTAUT model, constructed a model of factors influencing college students’ self-directed learning ability in programming. Through a review of existing literature, six key determinants were identified: learning expectancy, effort expectancy, facilitation conditions, teaching interaction, learning motivation, and learning self-efficacy. The main goal of this study was to develop a model that outlines the interaction of these factors when students use chatbots to learn programming. To this end, a questionnaire survey was conducted among 296 college students who had used chatbots for programming self-study. The questionnaire consisted of a set of rigorously validated items. The collected data were carefully analysed, including the construction and validation of a structural equation model using AMOS software. The results show that learning motivation plays the most important role in the positive influence of chatbots on programming self-learning ability. Its effect was significantly greater than that of other factors, including teaching interaction, learning self-efficacy, learning expectancy, and effort expectancy. In addition, learning motivation was found to mediate the effect of learning self-efficacy on programming self-learning ability. Finally, based on the findings, the study provides a series of recommendations aimed at offering scientific and practical strategies to enhance college students’ ability to learn programming independently.