AI-Generated Visual Representations In Science Education: A Didactic Transposition Perspective in Genetics Across Educational Level
17th International Baskent Congress On Socıal, Humanities, Administrative, and Educational Sciences, Ankara, Turkey, 12 - 14 February 2026, pp.1341-1358, (Full Text)
- Publication Type: Conference Paper / Full Text
- City: Ankara
- Country: Turkey
- Page Numbers: pp.1341-1358
- Ondokuz Mayıs University Affiliated: Yes
Abstract
Artificial intelligence plays a role across diverse areas, including healthcare and security, and the education sector is no exception. Genetics is a contemporary and significant subject in the science curriculum. However, it presents challenges for student learning and for teacher instruction. In particular, difficulties arise in comprehending concepts presented at the micro level. To improve the quality of science education, it is necessary first to ensure that abstract concepts are comprehensible to students and presented visually. Visual representations are powerful tools for making subject matter more straightforward to understand and for eliminating the tendency toward superficial learning. The concept of didactic transposition addresses changes in the process of transforming scientific knowledge into learned knowledge. AI-supported systems can be effective tools for supporting the transposition of scientific knowledge, as well as for concretizing abstract concepts. The purpose of this study is to examine AI-generated visuals created for different levels of education in genetics within the theoretical framework of didactic transposition. This study adopts a qualitative research approach. Visual representations generated using artificial intelligence tools, including ChatGPT, Copilot, and Leonardo, were systematically analyzed. A total of 27 images were generated by AI for middle school, high school, and university levels. The visuals were analyzed within the framework of the didactic transposition theory, focusing on scientific appropriateness and visual theme. The findings reveal that images generated by AI tools across grade levels in visual content and scientific accuracy, and that some analyzed images lack scientific terminology and are visually inaccurate.
Keywords: Artificial intelligence in science education, Didactic transposition, Genetics, AI-generated visual representations