Research shows that excessive use of ChatGPT by students leads to lower grades and lower motivation
While generative AI like ChatGPT is expected to be a tool to improve learning efficiency, it can also be used to hinder learning, such as by dumping homework on others or cheating during tests. A paper published in the academic journal 'Education and Information Technologies' by
Personality correlates of academic use of generative artificial intelligence and its outcomes: does fairness matter? | Education and Information Technologies
https://link.springer.com/article/10.1007/s10639-025-13489-6
Too much ChatGPT? Study ties AI reliance to lower grades and motivation
https://www.psypost.org/too-much-chatgpt-study-ties-ai-reliance-to-lower-grades-and-motivation/
AI is widespread among students, with one survey finding that nearly all UK university students use generative AI, but how they use it is not uniform: Past research has shown that students who are responsible, punctual and participate in class tend to be less reliant on AI tools.
Inspired by these findings, a Pakistani research group conducted three surveys of 326 students enrolled at three Pakistani universities to explore the relationship between generative AI usage patterns and student personality and academic performance.
The first survey focused on three of
The results showed that of the three personality traits surveyed, only conscientiousness was significantly associated with AI use, and that students with higher scores on conscientiousness were less likely to use generative AI for their studies. This suggests that 'conscientious people tend to prefer relying on their own efforts and dislike shortcuts,' supporting previous research showing a link between conscientiousness and independent learning.
On the other hand, contrary to expectations, openness, which refers to a tendency to actively try out new technologies, or neuroticism, which refers to a tendency to feel anxious or resistant to issues regarding the accuracy or ethics of such tools, did not show a significant relationship with AI usage.
'Our study finds that conscientious students are less likely to rely on generative AI for their academic tasks, likely due to their greater self-control and ethical standards,' said Sundas Azeem, corresponding author of the paper and assistant professor at the Shaheed Zulfikar Ali Bhutto Institute of Science and Technology in Pakistan. 'They may prefer more cognitively engaging learning activities, such as exploring multiple sources of information, research and discussion.'
The research group also analyzed how the perception of whether students' grades were evaluated fairly affected the relationship between personality traits and AI use. As a result, only openness and the perception of fairness of grading were marginally significant. Specifically, when students with high openness perceived the grading system as fair, their use of AI was slightly reduced. In contrast, conscientiousness and neuroticism were not significantly related.
Regarding this, Dr. Azeem said, 'We were surprised to find that fairness of grading only slightly affected the use of generative AI, and that this only affected the personality trait of openness. This suggests that generative AI is becoming more widespread regardless of the fairness of grading. This result is suggestive, as we had expected that students who perceived their grading to be unfair would turn to generative AI tools to get higher grades. In addition, people with high openness are generally quicker adopters of technology, but this was not reported in this study.'
Further analysis found that increased AI use was associated with several negative outcomes. For example, students who relied more on AI reported lower academic self-efficacy, meaning they felt it was harder for them to succeed on their own. They also found that reliance on AI was associated with learned helplessness, the tendency to believe that effort is futile and that one has no control over outcomes. Additionally, increased AI use was associated with a slight decline in academic achievement as measured by GPA.
These trends suggest that the overuse of generative AI may undermine student agency and reduce their motivation to engage deeply in class.
Looking back on the study as a whole, Azeem said, 'While our research sounds the alarm about the potential negative effects of generative AI on students, there is also literature supporting the positive outcomes of generative AI. As AI tools become increasingly prevalent in education, policymakers, educators, and EdTech developers will need to move away from viewing generative AI in a binary way: good or bad. We believe that exploring how to use generative AI responsibly while mitigating risks will be key to improving the quality of learning.'
Related Posts:
