ABSTRACT
Objectives
This study aims to explore the correlation of psychological domains and behavioural engagement with ChatGPT among undergraduate dental students.
Material and Methods
A cross-sectional study was conducted using a self-administered questionnaire (supporting material) distributed to 130 undergraduate dental students of a dental school in Karachi, Pakistan. The questionnaire assessed the frequency and purpose of ChatGPT usage across academic domains and specific learning contexts. It also evaluated students’ attitudes, perceptions, and behaviors using a Likert scale. Descriptive statistics were used for analysis, while Spearman’s correlation tests were applied to examine subgroup differences and relationships between attitudes and behaviors.
Results
Most participants reported learning about ChatGPT through peers or friends (n = 53) and social media (n = 38). The mean composite scores for the psychological domains [Perception(3.27), behaviour (3.21) and Attitude (3.19)] indicate a generally neutral-to-moderately positive inclination toward ChatGPT. The primary area of usage was basic dental sciences (n = 73). A strong positive correlation was observed between trust in ChatGPT for dental explanations and behavioral engagement (rs = 0.86, 95% CI [0.88, 0.95], p = 0.003, Holm–Bonferroni adjusted p = 0.006), while doubts regarding its accuracy were negatively correlated with engagement (rs = −0.91, 95% CI [−0.97, −0.76], p = 0.002, adjusted p = 0.006). Students who used ChatGPT for staying updated with dental knowledge (rs = 0.90, 95% CI [0.68, 0.91], p = 0.004, adjusted p = 0.004) and those willing to recommend it to peers (rs = 0.95, 95% CI [0.88, 0.98], p = 0.001, adjusted p = 0.004) also demonstrated significantly higher behavioural engagement.
Conclusion
Undergraduate dental students showed cautious but receptive attitudes toward ChatGPT, with psychological factors such as trust and perceived usefulness significantly shaping behavioral engagement. While the tool was mainly used for foundational learning, these findings support the need for structured, faculty-guided integration of AI in dental education.