Advancing equity through metacognitive instruction in higher education: a systematic review

Abstract

University students’ metacognition, including metacognitive knowledge, monitoring, and regulation, plays a crucial role in their academic success in higher education but students often have systematic self-monitoring problems, including the Dunning-Kruger effect, calibration errors and fluency illusions. This systematic review collates empirical evidence from around the world on these errors and examines the structural effectiveness of institutional and instructional practices and interventions for reducing errors and promoting equity in education. Following the PICOS framework, a thorough search was conducted across five electronic databases (PsycINFO, ERIC, Scopus, Web of Science and Google Scholar)for peer-reviewed empirical literature published from January 2007 onwards. The active search, screening, and selection workflows were executed from January 2025 to June 2026, with the final literature update completed on June 15, 2026. While the search window remained open to contemporary literature up to the final search date, the final pool of eligible included studies empirically spanned from 2007 to December 2024. The methodological quality and risk of bias were assessed using the Mixed Methods: Appraisal Tool and the certainty of evidence was assessed by an adapted GRADE approach. The entire process followed the PRISMA checklist and the review was registered on PROSPERO to enhance transparency. Of 1,288 identified, 37 were eligible primary empirical studies. The results were summarized using a narrative synthesis, which resulted in two categories: baseline errors( n  = 15) and interventions ( n  = 22). In 11 studies, there was evidence of severe overconfidence and miscalibration among students who are not performing well. Explicit metacognitive instruction ( n  = 10)was found to have high-certainty evidence of improving calibration discrepancies and narrowing demographic performance gaps among tested interventions. Scaffolded problem-solving frameworks ( n  = 5) and digital prompting tools ( n  = 7)demonstrated moderate-to-low certainty as there were inconsistencies in the baseline of each, student exposure to cognitive load, and an absence of proven demographic equity closures. University students are highly susceptible to metacognitive miscalibration. Self-regulation and equity are significantly improved by explicit classroom instruction, whereas technology-mediated interventions exhibit moderate-to-low certainty due to cognitive load constraints and fail to systematically narrow historical performance gaps. Sustained success will require structured training of instructors and integration of the curricula across the entire organization https://www.crd.york.ac.uk/PROSPERO/recorddashboard , PROSPERO CRD420261436943.