Formative Assessment Protocols Versus Summative Benchmarking in K–12 STEM Pedagogy: A Criterion-Referenced Evaluation of Learning Outcome Trajectories and Metacognitive Skill Acquisition

Abstract

This study systematically contrasts formative assessment protocols with summative benchmarking frameworks across K–12 STEM instructional contexts, examining their differential effects on learning outcome trajectories and metacognitive skill acquisition. Drawing on a mixed-methods design involving 1,847 students across fourteen secondary institutions in three national education systems, the investigation employed hierarchical linear modelling, think-aloud protocols, and validated self-regulation inventories. Findings indicate that iterative formative feedback cycles yield statistically significant gains in metacognitive monitoring (d = 0.61) relative to terminal summative evaluations, whereas summative benchmarking demonstrates superior predictive validity for standardised longitudinal performance indices. The paper concludes that a hybridised assessment architecture, calibrated to disciplinary epistemology and learner developmental stage, most effectively optimises both competency attainment and autonomous learning disposition in contemporary STEM education environments.