Metacognition refers to students’ self-regulation abilities in their learning process, involving planning, monitoring, and assessing one’s understanding and performance. For instance, the degree to which students monitor their progress during preparing for exams and recognize necessary adjustment for academic success. Previous studies indicated early metacognitive predicts subsequent academic achievement. Though meta-analyses examined the relationships between metacognition and academic outcomes, relying primarily on cross-sectional studies would risk bias from common method variance. To address this, He and colleagues conducted a three-level meta-analysis utilizing only longitudinal studies.
The meta-analysis included 28 studies, yielding 66 effect sizes, published between 1990 and 2023. These studies featured time-lag between measures of metacognition and academic outcomes ranging from 3 to 24 months, with an average of 10.3 months. Results revealed a small but significant positive correlation between initial metacognition and subsequent academic achievement (r = +0.22). Moderator analyses found on-line measures (r = +0.32), which utilized behavioral measure (e.g., on-going task performance), were stronger predictors than off-line measures (r=+0.17), which mainly used self-report questionnaire, after controlling other moderators. Effects also varied by metacognition components with largest effect for metacognitive skill (r = +0.34), followed by metacognitive awareness (r=+0.22), metacognitive strategy study (r = +0.21), and metacognitive knowledge (r = +0.17).
Additionally, examining the reversal link, results found a small positive correlation between initial academic achievement and subsequent metacognition (r=+0.12). Furthermore, autoregressive cross-lagged meta-analytic structural equation modelling revealed significant small bidirectional associations: students’ metacognition predicted their subsequent academic performance (beta = +0.038), and their academic outcomes also predicted future metacognition (beta = +0.079).
While longitudinal data shed light on dynamic relationships, the authors caution that causal inferences cannot be definitively drawn from fitting hypothesized model. Overall, the meta-analysis provides evidence that metacognition and achievement reciprocally and weakly predict each other over time.
Source: He, G., Chen, S., Lin, H., & Su, A. (2024). The association between initial metacognition and subsequent academic achievement: A meta-analysis of longitudinal studies. Educational Psychology Review, 36(3), 81. https://doi.org/10.1007/s10648-024-09922-w