<![CDATA[Value-laden religious subjects such as Qur'an–Hadith combine cognitive mastery, recitation skill, textual interpretation, and the behavioral internalization of values, a multidimensional target that conventional, cognition-focused achievement measures capture only partially, and whose underlying motivational and social mechanisms remain empirically underexamined. Drawing on achievement-goal theory and Bandura's social cognitive theory, this study tests whether social learning mediates the effect of students' mastery goal orientation on Qur'an–Hadith learning performance, with performance operationalized as a five-dimensional construct rather than as cognitive test scores alone. Using an explanatory cross-sectional survey, 280 of 935 grade-XI students across three Madrasah Aliyah in Padang City, Indonesia, were selected by proportional random sampling and completed a Likert-scale questionnaire, which was analyzed with partial least squares structural equation modeling (PLS-SEM). Mastery goal orientation significantly predicted learning performance (β = 0.416) and social learning (β = 0.728); social learning significantly predicted learning performance (β = 0.473) and complementarily mediated the orientation–performance relationship (indirect β = 0.345), with the model explaining 68.3% of the variance in performance and 53.1% in social learning (all p < 0.05). These estimates carry one caveat: discriminant validity is supported only marginally (maximum HTMT = 0.892 against the 0.90 benchmark; the Fornell–Larcker criterion is not met), so the social-learning/performance boundary is empirically thin, and single-source self-report data raise common-method-bias concerns. The findings tentatively support pairing individual mastery-oriented strategies with structured peer- and teacher-modeling in Islamic secondary education, and point to multi-source or longitudinal designs to confirm the construct boundary]]>
