STEM-PjBL Deep Learning model with product differentiation: a model for enhancing pre-service teachers’ critical thinking

In recent years, teacher education has placed increasing emphasis on critical thinking as a key 21st-century competency. However, the integration of STEM, project-based learning (PjBL), deep learning and product differentiation is often fragmented, leaving a gap in supporting pre-service teachers’ critical thinking development. This study examines the effectiveness of a STEM-PjBL Deep Learning Model integrated with product differentiation. A sequential explanatory mixed-method design was applied to 70 undergraduate students in Indonesia, divided into experimental and control groups. The experimental group experienced the integrated model, while the control group followed conventional PjBL. Data were analyzed using ANCOVA, t-tests, N-Gain and thematic analysis. The results show that the experimental group achieved higher post-test scores (M = 84.62, SD = 6.57) than the control group (M = 72.08, SD = 6.32), with a large effect size (η2 = 0.372, p < 0.001). All critical thinking sub-skills improved at moderate levels, with consistently higher gains observed in the experimental group. Qualitative findings indicate that STEM integration supports analytical reasoning, product differentiation enhances ownership and reflection strengthens metacognitive awareness. Practically, this model can guide teacher education programs in integrating STEM-PjBL with differentiated outputs and structured reflection to foster meaningful learning experiences.