Downloads & Free Reading Options - Results

INTRODUCTION Metacognition is often defined as “cognition about cognition”, referring to individuals’ awareness and understanding of their own thought processes (Flavell, 1979). In educational science, metacognition usually encompasses three components: (a) metacognitive knowledge related to what learners know about cognition and their own learning; (b) metacognitive skills referring to the regulation of cognition and the use of planning, monitoring, control, and evaluation strategies; (c) metacognitive experiences, which are the feelings and judgments associated with cognitive activity (Zohar & Barzilai, 2013). Other models place greater emphasis on the distinction between metacognitive monitoring and control, as in cognitive (neuro)science. While the first refers to assessing or judging the state of one’s own cognitive activity, the second refers to using these assessments and judgments to guide behavior or strategy selection (Fleming & Lau, 2014; Nelson, 1996). Metacognition is also frequently studied within the self-regulated learning framework and is considered one of its core dimensions, alongside cognitive and motivational components (Panadero, 2017). Furthermore, other models associate metacognition with executive functions as a high-level cognitive process (Roebers, 2017) or even theories of mind as both involve reflecting on mental states (Kuhn, 2000). A considerable number of studies have examined the relation between metacognition and academic achievement using both correlational and intervention approaches. The numerous existing meta-analyses in the field consistently report positive and relatively strong effects throughout development (e.g., de Boer et al., 2018; Eberhart et al., 2025; Ohtani & Hisasaka, 2018). Moreover, it is often claimed that metacognitive interventions are particularly beneficial for vulnerable students, especially students from disadvantaged socioeconomic backgrounds (de Boer et al., 2018; Education Endowment Foundation, 2018). Yet, surprisingly, only a few studies have primarily focused on the effect of SES on metacognitive development (e.g., Callan et al., 2016; Pappas et al., 2003; Wang, 1993). This has led to the topic being considered under-researched (Muijs & Bokhove, 2020), unlike other cognitive domains such as executive function and language development, which have been the focus of numerous meta-analyses (e.g., Korous et al., 2020; Lawson et al., 2018; Rakesh et al., 2025). However, given the great interest in metacognition in education, and that SES is frequently measured and used as a control variable in this field, it is likely that more relevant data could be retrieved from studies that did not focus on this specific question. Considering the importance of thoroughly understanding the mechanisms behind educational inequalities — one of the most significant challenges facing education systems (OECD, 2023) — and early evidence suggesting that metacognition may partly explain them (Koğar, 2021; Maximino-Pinheiro et al., 2024), the effect of SES on metacognition warrants further investigation. Thus, we aim to conduct the first systematic review and meta-analysis synthesizing all relevant empirical studies to quantify the association between SES and metacognition from infancy to university, and to test the potential moderators of this association. We are particularly interested in whether this relation varies by demographics, the specific SES and metacognition components considered, the types of measurement used, whether the SES–metacognition association is the primary focus of the study, and various study and sample characteristics. RESEARCH QUESTIONS - Primary research questions We aim to address the following primary question: (Q1) What is the direction and overall strength of the association between SES and metacognition? - Secondary research questions In addition, we are interested in whether the relation between SES and metacognition varies with: (Q2) demographics, including age, gender and ethnicity; (Q3) the SES component considered (including but not limited to education, occupation, income and composite measures); (Q4) the metacognitive component assessed (including but not limited to metacognitive knowledge, skills, experiences and composite measures) and the type of measurement (including but not limited to behavioral tasks, observations, self- and hetero-report questionnaires); (Q5) whether the investigation of the SES–metacognition association is the primary focus of the study or not. (Q6) several study and sample characteristics to test for potential moderating roles, including but not limited to type of publication (i.e., peer-reviewed or grey literature), type of association (i.e., raw association or adjusted association controlling for other variables), year of publication, country, sample size, intended sample SES (including but not limited to low-SES or “at-risk” participants, middle-SES, high-SES or diverse sample), SES distribution, age distribution, gender ratio, ethnicity ratio, and study design (i.e., single time-point measure, cross-sectional or longitudinal approach).