Computers in Human Behavior Reports (Mar 2025)
Uncovering personalized L2 motivation and self-regulation in ChatGPT-assisted language learning: A hybrid PLS-SEM-ANN approach
Abstract
Currently, chatbots powered by artificial intelligence (AI) have gained considerable attention due to their ability to provide personalized language learning (PLL) for learners. In this regard, recent studies have extensively explored learners' emotional aspects, such as their attitudes and acceptance of personalized language learning in chatbots. It is, however, unclear what factors might determine their cognitive behaviors in such a personalized language learning environment, particularly their self-regulation. To fill the gap, the researchers collected data from 133 Iranian EFL learners who had personalized language learning through ChatGPT in their language learning institute and answered our questionnaire that tapped on their personalized L2 motivational self-system (PEL2MSS) and their personalized self-regulation (PESRL). The researchers analyzed the empirical data using a hybrid SEM-artificial neural network (SEM-ANN), in contrast to previous literature that primarily relied on structural equation modeling (SEM). The results showed that ChatGPT significantly responded to language learners' current L2-self and their ought to L2-self to pass their obligation, and metrics to reach their goals resulted in seeking more assistance from ChatGPT and evaluating their language learning progress with it. Moreover, the sign of digital self-authenticity was also discovered by the researchers, where learners dedicated more motivation to learn language with ChatGPT in comparison with their previous language learning environments, which culminated in having more self-evaluation, goal-setting, and daily academic schedule to learn language with ChatGPT. Additionally, the ANN analysis supported the linear findings of the PLS-SEM by showing that language learners' current L2-self, digital self-authenticity, and ought to L2-self were the most significant motivational factors affecting their PESRL. Based on these findings, a new conceptual framework for the PLL was developed in the literature, and the research view was shifted from covering language learners' emotional aspects to their cognitive aspects in this environment. Thus we recommend that language teachers should avoid seeing ChatGPT as a tool that learners use for cheating; rather, it can be used as a co-teacher outside of the classroom to help students cover their present language learning needs, which might not be covered in the classroom due to the time restriction.
