Beware of Metacognitive Laziness: Effects of Generative Artificial Intelligence on Learning Motivation, Processes, and Performance (2412.09315v1)

Abstract: With the continuous development of technological and educational innovation, learners nowadays can obtain a variety of support from agents such as teachers, peers, education technologies, and recently, generative artificial intelligence such as ChatGPT. The concept of hybrid intelligence is still at a nascent stage, and how learners can benefit from a symbiotic relationship with various agents such as AI, human experts and intelligent learning systems is still unknown. The emerging concept of hybrid intelligence also lacks deep insights and understanding of the mechanisms and consequences of hybrid human-AI learning based on strong empirical research. In order to address this gap, we conducted a randomised experimental study and compared learners' motivations, self-regulated learning processes and learning performances on a writing task among different groups who had support from different agents (ChatGPT, human expert, writing analytics tools, and no extra tool). A total of 117 university students were recruited, and their multi-channel learning, performance and motivation data were collected and analysed. The results revealed that: learners who received different learning support showed no difference in post-task intrinsic motivation; there were significant differences in the frequency and sequences of the self-regulated learning processes among groups; ChatGPT group outperformed in the essay score improvement but their knowledge gain and transfer were not significantly different. Our research found that in the absence of differences in motivation, learners with different supports still exhibited different self-regulated learning processes, ultimately leading to differentiated performance. What is particularly noteworthy is that AI technologies such as ChatGPT may promote learners' dependence on technology and potentially trigger metacognitive laziness.

Effects of Generative Artificial Intelligence on Learning: A Comparative Study

The paper "Beware of Metacognitive Laziness: Effects of Generative Artificial Intelligence on Learning Motivation, Processes, and Performance" examines the implications of generative AI, specifically ChatGPT, on learning outcomes. This randomized experimental paper scrutinizes how AI, human expert interaction, and traditional tools affect learners’ motivation, self-regulated learning (SRL) processes, and overall performance.

Study Framework

The research addresses the nascent concept of hybrid intelligence—an integrated approach combining human and machine intelligence to augment learning capabilities. The paper specifically investigates how different learning agents (AI chatbot, human expert, and checklist tools) influence learners’ intrinsic motivation, SRL processes, and performance on academic tasks—primarily focusing on essay writing.

A cohort of 117 university students was divided into four groups. Each group was assigned varying levels of external support: one with no additional support, one using ChatGPT, another interacting with a human expert, and the last using a checklist tool for analytics feedback. The focus was on self-regulated learning and its constituents—forethought, performance, and self-reflection phases.

Key Findings

1. Intrinsic Motivation: The paper found no significant differences in intrinsic motivation across the four groups. However, descriptive statistics indicated variations where the control group exhibited the least interest and highest tension. Interestingly, the checklist tool group demonstrated higher interest and lower pressure, suggesting its effectiveness in creating structured learning environments.
2. Self-Regulated Learning Processes: Significant differences were observed in SRL processes, particularly during the revising stage. The AI group was most reliant on direct feedback from ChatGPT, bypassing several metacognitive processes compared to others, notably the human expert group, which engaged more with content reflection and evaluation. This discrepancy may indicate "metacognitive laziness," where learners overly depend on AI facilitation, risking reduced cognitive engagement and self-regulation.
3. Learning Performance: Notably, the ChatGPT group showed superior performance in essay score improvement compared to other groups, potentially attributed to the structured feedback AI can provide based on clear rubrics. However, this did not translate to knowledge gain or transfer, pointing to the limited impact of AI on deep learning and understanding.

Implications

The research raises critical questions and provides substantial empirical insights into the role of AI in educational settings. While ChatGPT proved effective in short-term task performance, the lack of improvement in intrinsic motivation and knowledge transfer highlights potential risks in over-reliance on AI. The "metacognitive laziness" observed suggests that learners might not engage in necessary higher-order thinking skills, potentially stagnating skill acquisition over time.

Future Directions

Given these findings, the paper advocates for the careful integration of AI in educational practices, prompting further research on the symbiotic relationship between learners and AI. Future studies should explore long-term impacts on learning and cognitive development across diverse tasks and contexts. It is crucial that educators encourage balanced AI use while promoting SRL skills to avoid dependency pitfalls and ensure robust cognitive growth.

Ultimately, this research underscores the dual nature of generative AI—enhancing certain aspects of learning while necessitating caution to preserve genuine cognitive and metacognitive engagement in hybrid intelligence frameworks.