Abstract

But instead of that, we propose in this paper that we think about it using a more of like a building-up type of model, in which you are building the connections between chat-based generative AI’s that exist out there right now and how they relate to people who have taken over directing their own education at universities. Contrary to the statistical mediation model, the conceptual model specifies several psychological pathways via which ChatGPT will impact upon learning. These are thought of as separate from being side-by-side with one another, but as conceptually distinct processes. According to social cognitive theory and self-determination theory, it discusses how ChatGPT is introduced into the learning situation and has an effect on its self-guided learning, because i influence people's way that they know themselves. If someone feels supported when it comes to instruction, and what kinds of ways a person would learn. And it is not like a multi- mediator model – it is giving us some explanation of how many different ways students can be autonomous in their own learning through this. In particular, the theoretical contribution of this study is to propose a theoretical bridging model. On the one hand, according to Social Cognitive Theory’s causal structure, we explain that ChatGPT is explained by it: a learning-environment factor affecting cognitive capability/perceived social support/technology belief, which would then impact learning behaviour. On the other hand, it uses SDT to illustrate that such psychological variables are converted into lasting, autonomous learning activity by internalizing competence, autonomy, and internally motivated reasons. As such, in terms of this, there can be considered digital literacy, as well as teachers and support for technology that aren’t mutually exclusive alternative mediations or replacements, but rather symbolic representations standing in for the more fundamental psychological pathways of thought and society and willpower. In other words, this research innovation isn’t a question of whether ChatGPT “affects” students’ self-directed study; instead, it reveals many varied psychological routes through which generative AI “affects” students’ study on their own, and how these can all be divided into a division of work. It’s like we’ve had some criticism of our previous studies about using AI to help students learn; folks think there hasn’t been much variety in what causes it.