Higher-order thinking in STEM and science education: A systematic review of constructs, frameworks, and applications

This systematic review synthesizes how Higher-Order Thinking (HOT) is defined, theoretically framed, implemented, and assessed in STEM and science education. A Scopus search conducted on January 13, 2025 identified empirical qualitative, quantitative, and mixed-methods studies published between 2014 and 2024. Following PRISMA 2020 procedures, 24 studies met the predefined eligibility criteria and were included in the narrative and thematic synthesis. Methodological quality was appraised using the Mixed Methods Appraisal Tool, with appraisal findings used to calibrate the strength of interpretation rather than exclude eligible studies. The synthesis identified four recurring and overlapping HOT construct families: critical thinking and argumentation, problem-solving and scientific reasoning, creativity and innovation, and metacognition and self-regulation. Bloom’s revised taxonomy was the most visible cognitive taxonomy, primarily guiding learning objectives and assessment design, while constructivist, inquiry-based, problem-based, discovery, and metacognitive models structured learning processes. No included study operationalized the SOLO Taxonomy as its primary empirical framework. Active and technology-supported approaches—including digital games, Internet of Things–supported investigation, educational robotics, flipped learning, online metacognitive instruction, and digital assessment—were generally associated with favorable HOT outcomes, although evidence strength varied because many studies used non-randomized, short-term, or context-specific designs. The review contributes an integrated analytical map linking HOT constructs, framework functions, learning activities, technologies, and assessment practices. It highlights the need for clearer construct–instrument alignment, stronger longitudinal and transfer designs, and sustained teacher professional development.