Longitudinal study of metacognition’s role in self-efficacy and hope development

Higher education aims not only to impart knowledge but also to cultivate critical thinking, reflective capacity, and proactive coping with social challenges. Metacognition—awareness and regulation of one’s own thought processes—plays a key role in self-regulation and learning. This study investigates whether metacognitive self-awareness of one’s cognitive biases (MCS) predicts the development of general self-efficacy (GSE) and hope over time. Grounded in self-regulated learning theory, which posits that conscious control over cognition enhances regulation of emotions and actions, the authors hypothesised that higher MCS would be positively related to trajectories of self-efficacy and hope across college years.

Perceived self-efficacy is the belief in one’s ability to influence outcomes and regulate behavior, effort, persistence, and emotions, and is linked to better task performance and academic adjustment. Contemporary work emphasizes general self-efficacy (GSE) as a cross-situational belief supporting motivation and self-regulated learning. Hope, per Snyder’s theory, comprises agency (motivation to initiate/sustain action) and pathways (finding routes to goals). Higher hope is associated with better psychological adaptation, lower depression, better social relationships, health, and academic success; longitudinal and meta-analytic evidence shows hope predicts persistence and completion. Although correlated with GSE, hope is conceptually distinct, especially regarding emotional processes; factor-analytic work supports their separability. Metacognition involves knowledge, monitoring, and control of cognition, with applications across memory, decision-making, development, and critical thinking. The present work focuses on metacognitive self-awareness of biases (MCS)—acknowledging one’s susceptibility to common biases (e.g., positivity, confirmation, optimism/ planning fallacy). MCS reflects reflective, self-diagnostic motivation and correlates positively with adaptive metacognitive constructs and positive emotions, and negatively with maladaptive metacognitions (e.g., rumination). Prior educational research indicates metacognition and self-efficacy jointly predict achievement, with mediation between them; training metacognitive strategies reduces procrastination and bolsters efficacy. Developmental research links verbal ability to metacognitive knowledge and later comprehension. These literatures suggest interrelations among MCS, GSE, and hope in achieving academic goals, motivating a longitudinal examination of their developmental trajectories.

Design: Three-year longitudinal study with five waves of assessment spaced at approximately six-month intervals (end of each semester, before exams). Participants: Randomly recruited undergraduate students from the Departments of Humanities and Social Sciences at the University of Gdansk. Power analyses for repeated-measures designs targeted N≈198–220; a larger sample was recruited to ensure power for small effects. Exclusion criterion: age >26 years (2.6% of original sample) to reduce age-related confounds. Inclusion criterion: presence on campus on data-collection day. Wave samples: Wave 1 (Mar–Apr 2014): N=416 (380F/38M); Wave 2 (Nov 2015): N=395 (358F/37M; included new participants); Wave 3 (Apr 2015): N=352 (321F/31M); Wave 4 (Nov–Dec 2016): N=345 (313F/34M); Wave 5 (Apr–May 2016): N=363 (331F/32M). Age at Wave 1: M=20.1 years (SD=2.66); overall range 19–25 (M=20.9, SD=1.15). A subset of 257 completed all five waves, with no significant differences from intermittent participants. Measures: - Metacognitive Self Questionnaire (MCSQ-21): 21 items assessing awareness of common cognitive biases on a 6-point Likert scale. Reliability (Cronbach’s alpha, McDonald’s omega) across waves ranged ~0.69–0.78 (alpha) and 0.74–0.83 (omega); construct validity supported (bi-factor CFA; RMSEA satisfactory). Each item reflects a bias episodically; higher scores denote greater self-acknowledged bias awareness. - Generalized Self-Efficacy Scale (GSES; Polish version): 10 items, 4-point Likert responses; single-factor structure. Reliability consistently high (alpha/omega ≈0.85–0.91). CFA fit acceptable with allowed residual correlations between items 4 and 6. - Adult Dispositional Hope Scale (Polish version): 8 substantive items (4 agency, 4 pathways) plus 4 fillers; 8-point Likert scale. Reliability alpha ~0.75–0.83, omega ~0.80–0.86; CFA with residual covariances allowed among selected items; RMSEA within acceptable ranges. Procedure: Informed consent obtained; confidentiality assured; ethics approval (University Ethical Committee no. 17a/2013; Helsinki compliant). Sessions administered by trained staff individually or in groups up to 30, near end of semesters. Paper–pencil administration; questionnaires (MCSQ-21, GSES, Hope) presented in randomized order (randomization effect non-significant, F<1). No incentives. Full debriefing at final participation. Statistical analyses: Initial partial correlations between MCS and outcomes (GSES, Hope), controlling for time, with bootstrap confidence intervals. Longitudinal CFA with Satorra–Bentler corrections; model fit via RMSEA (<0.06), SRMR (<0.08), CFI (>0.95). Latent growth curve models (LGCMs) with intercept loadings fixed to 1 and slope loadings 0–4 for the five waves; random intercept and slope; linear time; MCS entered as fixed covariate across time. Nested model comparisons via −2 log-likelihood differences assessed MCS effects. Causal mediation analysis estimated indirect effects of MCS on GSE over time; moderation analysis tested MCS × Time effects on Hope. Analyses in R 4.0.4 (lavaan 0.6.8; lme4 1.1-26; mediation 4.5.0).

- Descriptive grand means across waves: MCSQ-21 M=4.33 (SD=0.45), GSES M=3.06 (SD=0.45), Hope M=4.66 (SD=0.70). - Partial correlations (controlling time): MCS with GSES r=.18 (95% CI [0.08, 0.27]); MCS with Hope r=.26 (95% CI [0.18, 0.35]). - GSES growth: LGCM fit good (χ2(34)=42.535, CFI=0.991, SRMR=0.064, RMSEA=0.032 [0, 0.060]). Intercept mean=2.986; slope mean=0.024 per half-year (p<.001). MCS significantly improved model constant (χ2(1)=25.59, p<.001) and predicted change: approximately +0.024 (SE=0.01) GSES points per six months per one-point higher MCS. Slightly faster GSES increases for higher MCS vs lower MCS (χ2(1)=3.85, p=.0498). - Mediation (GSES): Total time effect β=0.028 (95% CI [0.014, 0.040]); indirect effect via MCS β=0.004 (95% CI [0.002, 0.010]); proportion mediated ≈20.4% (95% CI [8.3, 53.0]). - Hope growth: LGCM fit acceptable (χ2(34)=56.795, CFI=0.976, SRMR=0.086, RMSEA=0.055 [0.028, 0.080]). Intercept mean=4.917; slope mean=0.027 per half-year. Larger slope variance for Hope (0.012) than GSES (0.003). - Moderation (Hope): Significant MCS × Time interaction (β=0.099, SE=0.030, p<.001). Individuals with above-median MCS showed faster increases in hope; those below median showed an irregular decrease over time.

Results support that metacognitive self-awareness of biases (MCS) is a psychological resource facilitating development of both general self-efficacy and hope during college years. For GSE, MCS exerts a linear, positive association with growth: higher MCS corresponds to slightly faster gains in coping-related efficacy. For hope, MCS functions as a moderator, suggesting a threshold-like process whereby adequate metacognitive skills help translate real-world academic challenges into growing hopeful expectations, while insufficient MCS may be associated with declines in hope, potentially due to miscalibrated goals and repeated setbacks. These patterns align with theories of self-regulated learning and with distinctions between GSE (cognitive expectations about action outcomes) and hope (integration of cognitive and emotional processes, including pathway thinking). The stronger moderating role of MCS for hope may reflect pathway-generation capacities central to hope but not to self-efficacy. The discussion links MCS with positive metacognitive emotions and with neural evidence implicating higher-order cortical systems in metacognitive processing related to hedonic evaluation, suggesting a broader cognitive–affective basis for MCS’s adaptive role. Practically, enhancing metacognitive awareness of biases may support students’ self-knowledge, improve calibration of effort and difficulty, and bolster trajectories of efficacy and hope, especially under academic stressors.

This longitudinal study demonstrates that metacognitive self-awareness of cognitive biases (MCS) contributes to the development of general self-efficacy and moderates the growth of hope during university education. Individuals higher in MCS show slightly faster increases in self-efficacy, and MCS buffers and enhances hope growth in the face of real-life academic challenges. These findings extend self-regulated learning theory by highlighting MCS as a key resource shaping motivational and affective expectancies over time. Future research should incorporate behavioral indicators of self-regulation and academic performance, examine mechanisms linking MCS to agency versus pathways components of hope, test targeted interventions (e.g., bias-awareness and metacognitive strategy training) to causally enhance MCS and downstream outcomes, and evaluate boundary conditions (e.g., task difficulty, feedback environments) and potential confounders across diverse student populations.

- Self-report measures only; absence of behavioral/observational indicators may limit inferences about enacted self-regulation. - Longitudinal attrition and intermittent participation could introduce bias, though those completing all waves (N=257) did not differ significantly from others. - Some uncontrolled confounders (e.g., external stressors, academic context differences) may have influenced outcomes. - Sample drawn from humanities and social sciences with gender imbalance, potentially limiting generalizability. - Measurement timing and calendar irregularities could affect trajectories; reliance on campus presence for inclusion may select for availability.