Trust Predicts Actual Use of AI Chatbot as a Virtual Nutrition Assistant Among Dietetic Students in Taiwan: A Path Analysis
- PMID: 41220106
- DOI: 10.1111/jhn.70156
Trust Predicts Actual Use of AI Chatbot as a Virtual Nutrition Assistant Among Dietetic Students in Taiwan: A Path Analysis
Abstract
Background: Although studies have demonstrated ChatGPT's potential as an AI nutritionist, no research has yet investigated the factors influencing its adoption and actual use of ChatGPT as a virtual nutrition assistant among dietetic students.
Objective: In this study, we examined factors influencing ChatGPT's actual usage among dietetic students, focusing on the role of trust in its adoption for academic and clinical nutrition tasks.
Methods: A web-based survey was conducted among dietetic students in Taiwan with prior experience using ChatGPT. The survey was designed based on an extended Technology Acceptance Model. Partial least squares structural equation modeling was applied to assess the structural model and test hypotheses.
Results: In total, 150 dietetic students (82% female, mean age = 20.66 ± 3.96 years) completed the survey. Regarding specific nutrition-related tasks, ChatGPT's actual use scores were highest for "understanding nutrition knowledge" and "analyzing the calorie and nutrient content of a dietary record", while lower scores were observed for "preparing for the registered dietitian exam" and "analyzing patients' nutritional status and providing dietary recommendations". The PLS-SEM model respectively explained 28.2% and 44.6% of the variance in actual use (AU) for academic and clinical nutrition tasks. Trust (TR) and behavioral intention to use (BIU) independently predicted AU for academic tasks, including understanding nutrition knowledge (BIU: β = 0.336 and TR: β = 0.341) and preparing for the registered dietitian exam (BIU: β = 0.310 and TR: β = 0.311). For clinical tasks, such as analyzing the calorie and nutrient contents of a dietary record (β = 0.523) and evaluating patients' nutritional status (β = 0.381), TR was the sole significant predictor.
Conclusions: Trust is a key factor driving the adoption and actual use of ChatGPT as an AI nutrition assistant, particularly in clinical nutrition tasks.
Keywords: AI nutritionis; AI‐enabled dietitian; ChatGPT; Technology Acceptance Model; artificial intelligence; educational technology; trust.
© 2025 The British Dietetic Association Ltd.
References
-
- World Health O. Ethics and governance of artificial intelligence for health: WHO guidance. World Health Organization; 2021.
-
- Organization WH. Ethics and governance of artificial intelligence for health: large multi‐modal models. WHO guidance. World Health Organization; 2024.
-
- G. Eysenbach, “The Role of ChatGPT, Generative Language Models, and Artificial Intelligence in Medical Education: A Conversation With ChatGPT and a Call for Papers,” JMIR Med Educ 9 (March 2023): e46885, https://doi.org/10.2196/46885.
-
- F. Dzuali, K. Seiger, R. Novoa, et al., “ChatGPT May Improve Access to Language‐Concordant Care for Patients With Non–English language Preferences,” JMIR Med Educ 10 (2024): e51435, https://doi.org/10.2196/51435.
-
- H. Naveed, A. U. Khan, S. Qiu, et al., “A Comprehensive Overview of Large Language Models,” ACM Trans Intell Syst Technol 16, no. 5 (2025): 1–72, https://doi.org/10.1145/3744746.
MeSH terms
Grants and funding
LinkOut - more resources
Full Text Sources
