Аннотация
This article reports on two complementary multi-year case studies of integrating open-access PhET Interactive Simulations into the teaching of two notoriously demanding topics in the Kazakhstani school curriculum –stoichiometry in chemistry and probability theory in mathematics –and translates the resulting evidence into a unified setof design principles and curricular recommendations for higher education teacher preparation. Case 1 examines the use of “Reactants, Products and Leftovers” in 8th-grade chemistry across three Almaty schools; Case 2 examines the use of “Plinko Probability” in 10th-11th-grade mathematics across two Almaty institutions. Both interventions used a structured Predict-Observe-Explain design with parallel 15-item pre/post-tests, student questionnaires, and lesson-time logs. Across both subjects, simulation-supported instruction was associated with substantial gains in conceptual understanding (median Cohen’s d ≈ 1.0-1.1), increased intrinsic engagement, and a measurable reduction in the time required to reach class-level mastery. Crucially, the same four cross-cutting themes emerged in both subjects: visualisation of abstract sub-microscopic and stochastic phenomena, productive failure under low-stakes simulation, the limited appeal of competitive game elements, and the centrality of teacher coaching. We argue thatthis convergence across two very different STEM subjects strengthens the case for embedding digital game-based learning (DGBL) into Kazakhstani higher education through a four-component framework –conceptual orientation, simulation literacy, gamified lesson design and reflective practicum –applicable equally to chemistry and mathematics teacher preparation. Implications are discussed for university curriculum developers, methodologists and policymakers seeking to scale evidence-based gamification across the school-university continuum.