Implementing metacognitive strategies in STEM classrooms has been shown to improve test scores across various demographics and to strengthen long-term learning. So why don’t more science and engineering instructors use them?

In a junior-level fluid mechanics class at a major U.S. university, students aren’t only solving for pressure drops and Reynolds numbers. Each week, they also write short reflections: How did I plan today? Am I monitoring my understanding? How will I evaluate my performance later? It’s an exercise more often associated with an English seminar than with engineering. But here it serves a different purpose: to help students practice metacognition, or thinking about how they think.

The practice stems from a 2024 paper in the International Journal of Mechanical Engineering Education, led by Dr. Renee M. Clark of the University of Pittsburgh with colleagues at the University of South Florida. Their multi-year study found that when metacognitive strategies are woven deliberately into STEM instruction, students’ performance improves “statistically or practically significantly for various demographic strata.” Yet despite the documented gains, such strategies remain far from standard practice.

At the center of Clark’s work is a straightforward premise: equip students with tools to plan, monitor and evaluate their own learning. These self-regulatory skills are central to metacognition. As Clark and her colleagues write, “metacognitive strategies of planning, monitoring, and evaluating can be promoted through systematic reflection to drive self-directed, lifelong learning.”

Over three years, students in a fluid-mechanics course received structured in-class instruction on strategies for planning and tracking their problem-solving. They were also prompted to reflect weekly on their study habits, exam preparation, and evolving understanding of course concepts.

To test the effect, Clark’s team ran a comparison cohort that completed the same course without metacognitive scaffolding. Both groups took the same final exam, and both completed the Metacognitive Activities Inventory at the end of the term. Students in the metacognition-supported group showed a notably greater positive shift in self-regulatory behaviors. Their written reflections also revealed themes such as “carefulness, organization, and diligence” that correlated positively with stronger exam scores.

“Throughout the three years of the study, the weekly reflections were viewed foremost by students as a means of academic support,” Clark et al. write, noting benefits for “course performance, content understanding, exam preparation, accountability, and improvement opportunities.” They also observed that “more female versus male students may engage with and/or value reflection, suggesting a potential means of supporting gender equity in STEM.”

Taken together, the findings suggest that encouraging students not just to solve problems but to reflect on how they solve them can meaningfully deepen learning, and potentially help close performance gaps.

Clark’s team emphasizes that the benefits extend beyond grades. By reinforcing habits of reflection and self-monitoring, instructors can help students build intellectual skills that endure beyond a single course. Metacognition, they argue, is a “general skill that supports more-specific cognitive skills” and is “indispensable to learning and academic success.” Over time, many students came to see the value in the exercise. Clark’s group notes “a positive shift in students’ perspectives regarding the value of the reflection questions,” suggesting that what begins as an imposed task can evolve into an internalized practice.

Still, adoption of metacognitive practices across STEM courses remains limited. Time is one barrier: a typical 15-week semester leaves little room for additions to already dense technical syllabi, particularly in courses that must prepare students for standardized exams. Some instructors may be unfamiliar with metacognitive pedagogy, and some students may initially view reflective writing as peripheral to mastering technical content. Clark’s team observed that student opinions grew more positive only gradually, as the semester progressed. “Focus groups… showed a trend towards greater valuation of the weekly reflections over time,” they write, while noting that some improvements may reflect refinements to the reflection prompts themselves.

Other research suggests the picture is even more complex. A 2025 study led by Dr. Min Zhong and colleagues at UT Austin found that while metacognitive strategies are “widely encouraged,” improvements in students’ metacognitive awareness are not consistent across classrooms. Based on data from multiple biology courses, the team reported that “natural growth of overall metacognitive awareness is not significant in all students,” pointing to the need for more targeted interventions. Differences appeared not only across disciplines but across academic levels: the researchers found “a substantial variance of metacognition between entry-level and upper-level students, primarily centered around metacognitive knowledge.” For Zhong’s group, these patterns underscored “the critical necessity to enhance entry-level students’ cognition-related knowledge early on in their academic journey.”

Even with these caveats, Clark and her colleagues argue that integrating reflection into existing coursework can provide meaningful benefits without requiring wholesale redesign. Their findings, they conclude, “provide evidence for the potential enhancement of course performance with metacognition support,” including for students who begin the semester with little experience in self-directed learning.

At a moment when scientific and technical fields are rapidly evolving, teaching students to think about their own learning may be as important as teaching the content itself. Metacognitive habits could help shape scientists and engineers who are not only technically proficient but intellectually adaptable, into becoming proactive and self-aware learners for life.