Корабльов, Віктор (orcid.org/0009-0000-2487-6259) (2025) Chapter VI. Integration of Diffusion AI Models into the Educational Process as a Driver of Creative Innovation (1). ІЦО НАПН України, м. Київ, Україна, pp. 80-96. ISBN 978-617-8330-53-8
|
Text
PОЗДІЛ VI.Монограф_ШІ_в_освіті.pdf Download (1MB) |
Abstract
The rapid integration of artificial intelligence into various spheres necessitates the modernization of educational content and methods, particularly in informatics education. This chapter addresses the development and experimental validation of methodological materials for teaching the "Graphic Editor" topic to 5th-grade students in the New Ukrainian School (NUS), utilizing AI diffusion models as a driver of creative innovation. Traditional approaches often limit students' creative potential and fail to leverage modern technological capabilities. The proposed methodology integrates the Krita graphic editor with the Krita AI plugin, based on Stable Diffusion models, to augment the creative process. This approach involves project-based learning, problem-solving tasks, and gamification elements, where AI acts as a creative partner, offering stylistic suggestions, color palettes, and compositional variations based on student prompts (prompt engineering). An 8-week pedagogical experiment involving an experimental group (N=22, using Krita+Krita AI) and a control group (N=24, traditional methods) was conducted. Pre- and post-experiment assessments measured theoretical knowledge, practical skills, creativity, and motivation. The results demonstrated a significantly higher increase in learning achievements in the experimental group (average level increased by approx. 35.34%) compared to the control group (approx. 16.74%). The share of students achieving a high level rose from 13.64% to 36.36% in the experimental group, while the share at the initial level decreased from 22.73% to 4.55%. Statistical analysis (Student's t-test, Welch's t-test for homogeneity) confirmed the statistically significant difference between the groups (p < 0.05). The findings validate the hypothesis that integrating AI diffusion models enhances learning effectiveness, fosters creativity, and increases student motivation in graphic editor training. Recommendations include integrating these materials into curricula and providing teacher training on AI tools and prompt engineering.
Downloads
Downloads per month over past year
Actions (login required)
![]() |
View Item |


