The moderating effects of gender and need satisfaction on self-regulated learning through Artificial Intelligence (AI)

Artificial intelligence (AI) has the potential to support self-regulated learning (SRL) because of its strong anthropomorphic characteristics. However, most studies of AI in education have focused on cognitive outcomes in higher education, and little research has examined how psychological needs affect SRL with AI in the K–12 setting. SRL is a self-directed process driven by psychological factors that can be explained by the three basic needs of self-determination theory (SDT), i.e., autonomy, competence, and relatedness. This study fills a research gap by examining the moderating effects of need satisfaction and gender in predicting SRL among Grade 9 students. The results indicate that girls perceive more need support than boys. In predicting SRL, satisfaction of the need for autonomy and competence is moderated by both gender and AI knowledge, whereas satisfaction of the need for relatedness is moderated by gender only. Particularly among girls, the effects of autonomy and competence more strongly predict SRL when AI knowledge is low. These findings confirm the gender differences in need satisfaction when predicting SRL with a chatbot. The findings have implications for both teacher instruction and the design and development of intelligent learning environments.