ChatGPT-Assisted Hypertension Knowledge in Nursing Students: Assessment of Accuracy, AI Anxiety, and Cognitive Load
试验速览
- 阶段
- 不适用
- 状态
- 尚未招募
- 入组人数
- 96
- 主要终点
- Hypertension Prevention Attitudes Scale
研究概览
简要总结
ChatGPT provides quick access to information, research support, and study materials, but concerns remain regarding its reliability, accuracy, and inability to offer personalized care principles essential in nursing. Although previous studies show its high accuracy in clinical responses, over-reliance on AI-generated medical information necessitates cautious use. The study will explore both the benefits and limitations of ChatGPT in nursing education, particularly in hypertension learning.
详细描述
The use of artificial intelligence (AI) tools in education is increasing rapidly. ChatGPT, developed by OpenAI, is an AI-based chatbot that provides an interactive learning environment. It generates fluent and knowledge-based responses based on books, online sources, and articles published until 2021. The widespread adoption of ChatGPT across various fields has sparked debates about its role and limitations. In nursing education, students frequently use ChatGPT for quick access to information, research support, and exam preparation. However, concerns regarding its reliability arise due to the unknown sources of its responses and the potential for misinformation. ChatGPT also has limitations in interpreting complex, context-dependent answers and lacks the ability to apply the principle of individualized care, which is fundamental in nursing practice.
Studies have demonstrated ChatGPT's varying performance in the healthcare field. Hypertension, a chronic disease affecting over a billion people worldwide, is a critical topic for nursing students, as their understanding of the condition can positively impact patient care. Previous research has shown that ChatGPT provides clinically appropriate answers to hypertension-related questions with a high accuracy rate of 92.5%. Additionally, the GPT-4 version of ChatGPT correctly answered over 86% of the questions in the United States Medical Licensing Examination (USMLE).
This randomized controlled study aims to assess the effectiveness of ChatGPT in teaching hypertension to nursing students while also evaluating their levels of AI-related anxiety and cognitive load. Given the increasing presence of AI tools in education, understanding both their advantages and limitations is crucial for their optimal integration into nursing education.
The study population consists of students enrolled in the nursing program at a private university. The study aims to reach the entire population, specifically 96 students who have completed the Internal Medicine Nursing course. Students meeting the inclusion criteria will be informed about the study and invited to participate. Volunteers will complete an Introductory Information Form and be randomly assigned to intervention (ChatGPT) or control groups in a 1:1 ratio using computer-based randomization (48 students per group). The intervention group will answer questions from the Hypertension Prevention Attitude Scale using ChatGPT, while the control group will use traditional methods. Afterward, both groups will complete the Artificial Intelligence Anxiety Scale and Cognitive Load Scale, concluding data collection. All collected data will be analyzed using the SPSS for Windows 22.0 statistical software package.
研究设计
- 研究类型
- Interventional
- 分配方式
- Randomized
- 干预模型
- Parallel
- 主要目的
- Other
- 盲法
- Single (Outcomes Assessor)
盲法说明
Outcome assessor blind
入排标准
- 性别
- All
- 接受健康志愿者
- 是
入选标准
- •To be enrolled in the nursing program at a private university, during the 2024-2025 academic year.
- •To have taken the Internal Medicine Nursing course (In this course, students receive 4 hours of theoretical lessons on nursing care for hypertension patients).
- •To be willing to volunteer for participation in the study.
排除标准
- •Students who wish to withdraw from the research at any stage will not be included in the study.
结局指标
主要结局
Hypertension Prevention Attitudes Scale
时间窗: At the beginning of the study first admission.
The scale consists of 26 items and subdimensions, including protection and control, habits and lifestyle, nutrition attitudes, mental state and physical activity, and disease and risk knowledge. The items are rated on a five-point Likert scale, ranging from "Strongly Disagree" to "Strongly Agree." The scale scores can range from 26 to 130. There is a positive relationship between the scale scores and attitudes toward hypertension prevention. The Cronbach's alpha value of the scale is 0.91.
Artificial Intelligence Anxiety Scale
时间窗: Immediately after the intervention (answering the scale questions)
The Artificial Intelligence Anxiety Scale (AIAS) was developed by Wang and Wang (2019) and adapted into Turkish by Akkaya et al. (2021). The scale is a 5-point Likert type, consisting of 21 items and 4 factors. These factors are: Learning, Job Change, Socio-technical Blindness, and Artificial Intelligence Structuring. The minimum score that can be obtained from the scale is 21, and the maximum score is 105. A higher score indicates a higher level of AI anxiety. The Cronbach's alpha coefficient of the scale is reported to be 0.95.
Cognitive Load Scale
时间窗: Immediately after the intervention (answering the scale questions)
The scale developed by Paas and Van Merriënboer (1993) aims to measure the cognitive load of students during individual study processes. It was adapted into Turkish by Kılıç and Karadeniz (2004). The scale is a symmetric, Likert-type scale with scores ranging from 1 to 9. It allows the assessment of the effort a student exerts during their individual learning process. According to the scale, cognitive load increases from 1 to 9. Scores between 1-4 are considered low cognitive load, while scores between 5-9 are considered high cognitive load. Paas and Van Merriënboer (1993) reported an internal consistency coefficient of 0.82 for the scale, while Kılıç and Karadeniz (2004) calculated an internal consistency coefficient of 0.90 for the Turkish version.
次要结局
未报告次要终点
研究者
Nursemin ÜNAL
Associate Professor
Ankara University