New MEDIS:ON study shows that relying on GenAI over understanding — not frequency of use — predicts lower achievement in Czech seventh-graders.
Colleagues from the MEDIS:ON project, led by SΛTIS member Tomáš Lintner, have published a new article in Education and Information Technologies titled Generative AI use and academic achievement: a cross-sectional study of Czech seventh-grade students. The study is co-authored by SΛTIS member Markéta Žáková.
Drawing on the deep versus surface approaches to learning framework, the study distinguishes between two separate aspects of student engagement with GenAI: how frequently students use these tools, and the extent to which they endorse reliance on GenAI over understanding — defined as the belief that knowing how to prompt AI effectively matters more than developing one's own conceptual grasp of the material.
Using cross-sectional survey data from a nationally stratified sample of 2,307 seventh-grade students, the authors estimated structural equation models predicting IRT-scaled achievement in mathematics and Czech language. The frequency of GenAI use was not positively associated with achievement in either subject. In contrast, a stronger tendency to rely on GenAI instead of building understanding was consistently and negatively associated with achievement, with a more pronounced effect in Czech language. Neither frequency of use nor reliance significantly moderated the associations between learning-related dispositions and achievement.
The findings suggest that the educational implications of GenAI depend less on how often students use AI tools and more on whether AI is approached as a support for understanding or as a substitute for cognitive engagement. The results contribute to ongoing debates on epistemic offloading and the role of AI in shaping students' orientations toward learning.
Read the full article here.
SΛTIS member Tomáš Lintner appeared on Czech Television to present findings from the ai:dolescence research collection. His study shows that three-quarters of Czech high school students use AI tools daily, roughly half in ways that break school rules, and a smaller share even to create deepfake content targeting classmates.
From March 9–20, the SATIS Lab hosted Mauritz Cartier van Dissel, a PhD candidate at the Complexity Science Hub in Vienna. Mauritz worked with Tomáš Lintner on a joint paper on large-scale social network analysis. The visit also included a shared dinner and a walk through Brno with CSH colleague Giacomo Zelbi.
