ChatGPT in Dental Education: A Dual-Method Analysis of Technology Acceptance and Predictive Modeling

Abstract

Introduction: artificial intelligence (AI) is increasingly integrated into dental education, yet little is known about the adoption of AI-powered tools such as ChatGPT. Understanding the determinants of students’ behavioral intention to use these tools is crucial for effective integration into curricula. This study extends the Technology Acceptance Model (TAM) by incorporating Perceived Value (VAL) and Task-Technology Fit (TTF) and employs a hybrid analytical approach.Method: a structured questionnaire comprising 50 items across ten constructs was distributed to 318 dental students in Indonesia, with 263 valid responses collected. Data were analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM) to test reliability, validity, and hypothesized relationships. To enhance predictive validation, six machine learning (ML) classifiers (AdaBoostM1, J48, BayesNet, Logistic Regression, OneR, and LWL) were applied.Results: PLS-SEM results revealed that Perceived Value (β = 0.347, p < 0.001) and Perceived Usefulness (β = 0.321, p < 0.001) were the strongest predictors of intention to use ChatGPT. Additional significant effects were found for Perceived Enjoyment, Trust, and Perceived Accuracy. Conversely, Perceived Ease of Use, Social Influence, and Facilitating Conditions were not significant. Mediation analysis confirmed that Perceived Usefulness mediated the effects of TTF, Trust, and Accuracy on adoption intention. ML analysis corroborated these findings, with AdaBoostM1 achieving the highest predictive accuracy (87.3%).Conclusions: adoption of ChatGPT in dental education is predominantly driven by perceived academic value, usefulness, and engaging learning experiences rather than ease of use or social factors. The validated framework integrating TAM, Perceived Value, and TTF provides both theoretical advancement and practical guidance for integrating AI into dental education. The hybrid use of PLS-SEM and ML enhances model robustness and offers a replicable methodology for future educational technology research.

Ähnliche Arbeiten