Evaluating the Impact of the DIPEC–STEM Teaching Model on Secondary Students’ Computational Thinking: A Quasi-Experimental Study in Colombia

This study examines changes in computational thinking (CT) associated with the implementation of the DIPEC–STEM teaching model in technical upper-secondary education. Using a quantitative, one-group pretest–posttest quasi-experimental design, the intervention was implemented over one academic period in three public institutions in Medellín, Colombia. Participants (N = 130) were selected through non-probabilistic multistage sampling (purposive selection of institutions and convenience sampling of intact classes). CT was measured before and after the intervention using an adapted version of the Computational Thinking Test, assessing four CT dimensions: decomposition, pattern recognition, abstraction, and algorithmic thinking. Pre–post differences were evaluated using paired-samples inferential analysis and effect sizes (Cohen’s d). Results indicated a statistically significant improvement in overall CT performance after the intervention, with a large effect (d = 0.8957). Dimension-level effects were strongest for algorithmic thinking (d = 1.22) and pattern recognition (d = 0.76), modest for abstraction (d = 0.28), and slightly negative for decomposition (d = –0.20), evidencing differentiated patterns across CT skills. Additionally, a student perception survey administered to a subsample (n = 31) yielded consistently positive ratings regarding the clarity of the DIPEC–STEM sequence and the usefulness of the activities. Overall, the findings suggest that DIPEC–STEM is associated with meaningful gains in CT—particularly in skills related to algorithmic structuring and pattern analysis—within the constraints of authentic public-school settings.