George Pallivathukal Renjith, Kyaw Soe Htoo Htoo, Donald Preethy Mary, Samson Renu Sarah, Hj Ismail Abdul Rashid
Oral Pathology and Oral Biology, Manipal University College Malaysia, Melaka, MYS.
Community Medicine, Manipal University College Malaysia, Melaka, MYS.
Cureus. 2024 Jan 27;16(1):e53032. doi: 10.7759/cureus.53032. eCollection 2024 Jan.
The impact of generative artificial intelligence-based Chatbots on medical education, particularly in Southeast Asia, is understudied regarding healthcare students' perceptions of its academic utility. Sociodemographic profiles and educational strategies influence prospective healthcare practitioners' attitudes toward AI tools.
This study aimed to assess healthcare university students' knowledge, attitude, and practice regarding ChatGPT for academic purposes. It explored chatbot usage frequency, purposes, satisfaction levels, and associations between age, gender, and ChatGPT variables.
Four hundred forty-three undergraduate students at a Malaysian tertiary healthcare institute participated, revealing varying awareness levels of ChatGPT's academic utility. Despite concerns about accuracy, ethics, and dependency, participants generally held positive attitudes toward ChatGPT in academics.
Multiple logistic regression highlighted associations between demographics, knowledge, attitude, and academic ChatGPT use. MBBS students were significantly more likely to use ChatGPT for academics than BDS and FIS students. Final-year students exhibited the highest likelihood of academic ChatGPT use. Higher knowledge and positive attitudes correlated with increased academic usage. Most users (45.8%) employed ChatGPT to aid specific assignment sections while completing most work independently. Some did not use it (41.1%), while others heavily relied on it (9.3%). Users also employed it for various purposes, from generating questions to understanding concepts. Thematic analysis of responses showed students' concerns about data accuracy, plagiarism, ethical issues, and dependency on ChatGPT for academic tasks.
This study aids in creating guidelines for implementing GAI chatbots in healthcare education, emphasizing benefits, and risks, and informing AI developers and educators about ChatGPT's potential in academia.
基于生成式人工智能的聊天机器人对医学教育的影响,尤其是在东南亚地区,在医学生对其学术效用的认知方面研究不足。社会人口统计学特征和教育策略会影响未来医疗从业者对人工智能工具的态度。
本研究旨在评估医科大学学生在学术用途上对ChatGPT的知识、态度和实践情况。它探讨了聊天机器人的使用频率、用途、满意度水平,以及年龄、性别与ChatGPT相关变量之间的关联。
马来西亚一所高等医疗保健机构的443名本科生参与了研究,结果显示他们对ChatGPT学术效用的认知水平各不相同。尽管对准确性、伦理和依赖性存在担忧,但参与者总体上对ChatGPT在学术方面持积极态度。
多元逻辑回归突出了人口统计学、知识、态度与学术使用ChatGPT之间的关联。医学学士学生比牙医学士和医学信息学学生更有可能在学术上使用ChatGPT。最后一年的学生使用ChatGPT进行学术活动的可能性最高。更高的知识水平和积极态度与学术使用的增加相关。大多数用户(45.8%)在独立完成大部分作业时使用ChatGPT来辅助特定的作业部分。一些人不使用它(41.1%),而另一些人则严重依赖它(9.3%)。用户还将其用于各种目的,从提出问题到理解概念。对回复的主题分析表明,学生们担心数据准确性、抄袭、伦理问题以及在学术任务中对ChatGPT的依赖。
本研究有助于制定在医疗教育中实施生成式人工智能聊天机器人的指导方针,强调其益处和风险,并让人工智能开发者和教育工作者了解ChatGPT在学术界的潜力。