How Can Large Language Models Improve Health Research Skills: A Perspective Study

Area of Science:

• Health Research
• Medical Education
• Artificial Intelligence

Background:

• Health researchers face challenges in text translation, information interpretation, and statistical analysis.
• Artificial Intelligence (AI) models offer potential solutions by simplifying access to information and automating complex tasks.

Purpose of the Study:

• To examine university students' perspectives on the impact of AI models in addressing health research challenges.
• To assess student familiarity, usage, and perceived benefits and risks of AI in health research.

Main Methods:

• A survey study was conducted in 2023 involving 445 students from Kerman University of Medical Sciences.
• A researcher-developed questionnaire was used, with data analyzed using descriptive statistics and linear regression.
• Open-ended responses were coded and analyzed using MAXQDA software.

Main Results:

• 60% of students reported moderate to high familiarity with AI, but only 52% reported moderate to high usage.
• AI models were rated most favorably for article translation and rephrasing (Mean=3.33).
• Significant risks identified include AI providing false information (n=34) and failing to adhere to ethical standards (n=10).

Conclusions:

• AI models demonstrate potential for enhancing health research efficiency and quality.
• Further efforts are needed to mitigate identified risks and promote wider AI adoption in health research.