Profiles of learners based on their cognitive and metacognitive learning strategy use: occurrence and relations with gender, intrinsic motivation, and perceived autonomy support

The study addresses persistent concerns about secondary school students’ poor achievement in science, especially Physics, and gender disparities favoring boys. Grounded in self-regulated learning theory, the authors examine how cognitive and metacognitive learning strategies (rehearsal, organization, elaboration, critical thinking, metacognitive self-regulation) relate to motivation and autonomy support in lower secondary Physics. The research aims to identify distinct learner profiles based on combinations of strategy use and to test differences across gender, intrinsic motivation, and perceived teacher autonomy support. The authors hypothesized more than two latent profiles, including a less self-regulatory metacognitive group; that profiles heavy in organization would include more girls; and that profiles characterized by higher-order strategies would show higher intrinsic motivation and perceived autonomy support.

Within the self-regulated learning framework, cognition (strategy use), metacognition (planning, monitoring, evaluation), and motivation (e.g., intrinsic motivation) interact to influence learning. Deep-level strategies (elaboration, critical thinking) support deeper understanding, whereas surface-level strategies (rehearsal, organization) emphasize memorization. Autonomy support fosters intrinsic motivation and adaptive strategy use. Prior person-centered research has revealed multiple learner profiles across educational levels (e.g., four profiles among elementary, secondary, and university students) with varying links to achievement. Evidence on gender differences in strategy use is mixed: some studies report girls’ greater strategy use or preference for memorization, others show boys’ higher use of rote learning or critical thinking, with variations by culture and subject. Little is known about Physics-specific strategy profiles in lower secondary school in developing countries, warranting a person-centered exploration and examination of gender, intrinsic motivation, and perceived autonomy support across profiles.

Design: Cross-sectional, person-centered study using latent profile analysis (LPA) with a three-step approach. Participants: 579 Grade 9 students were sampled from six schools in Central Uganda; data from 573 students were analyzed after excluding six without consent. Females comprised 56% (n=321). Mean age was 14.3 years (SD=1.51); 50.9% were day scholars. Ethics and procedure: Ethical approval was obtained from Mbarara University of Science and Technology REC (no. 86/2019) and Universität Potsdam REC. Written informed consent was collected. An anonymized questionnaire was administered during Physics classes (~30 minutes) with researchers present. Measures: - Cognitive and metacognitive learning strategies: MSLQ (Pintrich et al., 1991) subscales for rehearsal, organization, elaboration, critical thinking, and metacognitive self-regulation, adapted to “Physics class.” 7-point Likert scale (1–7). CFA indicated good fit; internal reliabilities acceptable (alphas ~0.82–0.90 reported in Table 3). - Perceived teacher autonomy support: Adapted Learning Climate Questionnaire (Williams & Deci, 1996); 15 items on a 7-point scale; one item with loading <0.40 was removed; CFA fit acceptable; α=0.90. - Intrinsic motivation for Physics: 5-item subscale from adapted Science Motivation Questionnaire II (Glynn et al., 2011); 5-point Likert (1–5); CFA fit acceptable; α=0.73. Preliminary analyses: Missing data (0.5%) handled with FIML. Data were approximately normal (Shapiro–Wilk p=0.78). KMO=0.93; Bartlett’s χ²=2789.65, p<0.05. CFAs met fit criteria (e.g., CFI/TLI ≥0.90; RMSEA ≤0.06; SRMR ≤0.08). Measurement invariance across gender for the learning strategy scale was supported at configural, metric, scalar, and strict levels (ΔCFI≤0.01; ΔRMSEA≤0.015). Variable-centered gender comparisons: Independent t-tests compared males and females on strategy use. Latent Profile Analysis: Series of LPA models (1–6 profiles) estimated with information criteria (AIC, BIC, ABIC), likelihood ratio tests (LMR, adjusted LMR, BLRT), and entropy. The 5-profile model was rejected due to a class <5%. A 4-profile solution was selected based on significant BLRT, satisfactory entropy (~0.80+), classification probabilities (>0.85), class sizes, and theoretical interpretability, despite mixed LMR results. Posterior classification probabilities supported adequate class assignment. Predictors of profile membership: In step 3, multinomial logistic regression included gender, perceived autonomy support, and intrinsic motivation as covariates (Mplus 8 auxiliary procedure), reporting coefficients and odds ratios while preserving class assignments. Additional analyses used SPSS v20.

- Descriptive patterns: Students reported highest use of elaboration and metacognitive self-regulation; critical thinking had the lowest mean. - Gender differences (variable-centered): Girls scored higher in organization (p=0.03) and elaboration (p=0.02). No significant gender differences in rehearsal, critical thinking, or metacognition. - Latent profiles (n=573): Four profiles were identified: 1) Competent strategy users (n=152; 26.5%): Highest means on elaboration, organization, critical thinking, and elevated metacognitive self-regulation. 2) Struggling strategy users (n=62; 10.9%): Lowest usage across all strategies (all z-scores negative). 3) Surface-level learners (n=204; 35.6%): High rehearsal, lower organization, elaboration, and critical thinking; slightly above-average metacognition. 4) Deep-level learners (n=155; 27.0%): Emphasis on higher-order strategies (critical thinking, metacognition) relative to rehearsal and elaboration. - Predictors of membership: - Gender: Girls were 2.4–2.7 times more likely than boys to belong to the competent strategy user and surface-level learner profiles relative to struggling and deep-level profiles. - Perceived autonomy support: Higher perceived teacher autonomy support increased the likelihood of membership in the competent user profile (and, relative to other profiles, also favored adaptive profiles); lower autonomy support predicted the struggling profile. - Intrinsic motivation: Higher intrinsic motivation increased the likelihood of membership in the deep-level learner profile (and the competent profile relative to others); lower intrinsic motivation predicted the struggling profile.

The study demonstrates meaningful heterogeneity in Physics learning strategies among lower secondary students, aligning with self-regulated learning theory. Identifying four distinct profiles clarifies that many students rely on surface approaches, while smaller groups either struggle broadly with strategy use or engage in deep-level strategies emphasizing critical thinking and metacognition. The findings address the research questions by showing that gender, intrinsic motivation, and perceived autonomy support systematically differentiate profile membership: girls were more prevalent among competent and surface-level profiles; higher intrinsic motivation aligned with deep-level learning; and perceived teacher autonomy support was strongly associated with competent strategy use. These results underscore the pedagogical importance of autonomy-supportive teaching and motivational scaffolding to cultivate deeper strategy repertoires, particularly fostering critical thinking and metacognitive regulation. The presence of a notable struggling group highlights an urgent need for explicit instruction and support in strategy development early in secondary school to prevent widening gaps as academic demands increase.

This study contributes a person-centered understanding of lower secondary students’ Physics learning strategies, revealing four profiles distinguished by cognitive and metacognitive strategy combinations. It shows that perceived teacher autonomy support and intrinsic motivation are higher in adaptive profiles (competent and deep-level), while girls are more likely to be competent and surface-level users and show higher organization and elaboration. Implications include promoting autonomy-supportive instruction and explicitly teaching deep-learning strategies, especially to girls, to address gendered patterns and enhance higher-order strategy use. Future research should examine longitudinal stability and evolution of profiles, incorporate additional data sources (e.g., learning diaries, classroom observations), extend to other science domains, and link profiles to Physics achievement to evaluate performance implications.

- Self-report measures may introduce bias; complementary qualitative or behavioral data (e.g., learning diaries) are recommended. - Cross-sectional design precludes causal inference and does not capture profile stability over time. - Focus on Physics limits generalizability to other science subjects. - Lack of Physics achievement data prevented examining relations between profiles and performance.