<![CDATA[Meta-analytic evidence shows that character education programmes yield only small average effects and that boarding in itself is not consistently associated with better student development; the quality of the arrangements operating inside the residential setting is therefore a more plausible focus of inquiry. Which of those arrangements carries greater predictive weight remains unknown, because the santri care system and extracurricular programme management have so far been examined separately rather than in a single model. This study estimates the relative predictive weight of both when they are modelled simultaneously. A cross-sectional survey analysed with multiple linear regression was administered to the entire population of 112 santri at Annakhil Darunnajah 6 Islamic Boarding School, Mukomuko, Bengkulu (census). Three scales care system (16 items, α = .897), extracurricular management (10 items, α = .851), and discipline (10 items, α = .827) met the criteria for content validity and internal consistency; inter-construct correlations corrected for attenuation (.61–.76) remained below the .85 discriminant threshold. Both predictors were positively and significantly associated with discipline: care system β = .515, 95% CI [.35, .68], p < .001, f² = .37; extracurricular management β = .270, 95% CI [.11, .43], p = .001, f² = .10. Together they explained 48.5% of the variance in discipline scores (R² = .485; adjusted R² = .476; F(2, 109) = 51.38, p < .001), with effective contributions of 33.9% and 14.7%, respectively. Descriptively, the care system is therefore the dominant predictor and extracurricular management a complementary predictor; the difference between the two coefficients was not formally tested. Because the design is correlational and both the predictors and the outcome were reported by the same respondents, the findings describe statistical association and explained variance within this population, not causal effects. The weakest indicators consistency of rule enforcement and habituation to the daily schedule mark the point at which the explanatory power of the model diminishes.]]>
