Related Experiment Videos
Changes in artificial intelligence literacy, attitudes, and acceptance following ChatGPT-based training among
Ebru Bulut1, Hacer Yalniz Dilcen2, Beyza Aksu Dünya3
1Department of Midwifery, Health Sciences Faculty of Bartın University, Bartın, Türkiye, Turkey. esert@bartin.edu.tr.
Background:
The increasing role of artificial intelligence technologies in health education and clinical practice makes understanding the knowledge, attitudes, and readiness levels of future healthcare professionals in this field increasingly important. This study aimed to examine changes in midwifery students' attitudes, literacy, and acceptance levels regarding artificial intelligence (AI) following a brief ChatGPT-based artificial intelligence training program. Additionally, the study sought to explore how students perceive and integrate AI within professional contexts, supported by qualitative data.
Methods:
The study was conducted using an explanatory sequential mixed-methods design. In the quantitative phase, a single-group pretest-posttest quasi-experimental design involving a brief ChatGPT-based artificial intelligence training intervention was applied (n = 36 pre-test; n = 32 post-test). In the qualitative phase, semi-structured in-depth interviews were conducted with purposefully selected students (n = 7). Quantitative data were analyzed using descriptive statistics, paired-samples t-tests, and correlation analysis. Thematic analysis was used for the qualitative data.
Results:
Statistically significant improvements were observed in students' attitudes toward AI (p<0.001, Cohen's d=- 0.96, large effect) and AI acceptance (p=0.008, Cohen's d=- 0.51, moderate effect) following the training. In contrast, changes in AI literacy scores were positive but not statistically significant (p=0.351, Cohen's d =-0.17, small effect). These findings indicate that the magnitude of change varied across outcome domains, with stronger effect size values observed in attitudinal and acceptance-related measures compared to AI literacy. Qualitative findings revealed that students perceived AI as a supportive tool, reported reduced ethical concerns after the training, and expressed increased self-confidence in using AI tools. The training enhanced students' engagement with AI; however, the development of deep AI literacy was found to require repeated and practice-based learning experiences.
Conclusion:
Short-term AI training was associated with more positive attitudes and greater acceptance of AI among midwifery students; however, its contribution to AI literacy development appears more limited and may require sustained educational exposure. Therefore, integrating AI into health education curricula through pedagogically structured, practice-oriented, and ethically grounded educational modules may support students' gradual development of AI-related competencies while maintaining the principles of human-centered care.
Clinical Trial Number:
Not applicable.
