This study examined the impacts of an AI chatbot intervention for improving arithmetic word-problem-solving among Chinese primary students with dyslexia and explored how different types of motivation moderated these effects. Eighty-three students diagnosed with dyslexia were recruited and randomly assigned to either an AI chatbot-assisted learning condition or a traditional instruction control group. Over 6 weeks, the experimental AI chatbot group received adaptive, multisensory, and individually tailored support for arithmetic word problems. Results revealed that the Chatbot group witnessed significantly greater gains in arithmetic word-problem-solving performance than the control group. Moderation analyses showed that intrinsic motivation, but not extrinsic motivation or amotivation, significantly enhanced the intervention's impacts. Intrinsic motivation emerged as the key factor that amplifies the impact of the intervention. The experimental group exhibited substantial increases in intrinsic motivation and reductions in amotivation, whereas the control group's motivation remained unchanged. These findings indicate the importance of intrinsic motivation in maximizing the educational impact of AI-based interventions to support students with dyslexia.
Wang et al. (Sun,) studied this question.