The purpose of this study was to investigate the way middle school students recognize the effectiveness of three learning methods in mathematics― classroom instruction, self-study, and private tutoring. Specially, how do students’ recognition for effectiveness of each learning method change during their transition from middle school to high school? Study data were secondary data derived from the Korea Youth Panel Survey for 2003–2007. Findings revealed that private tutoring, self-study, and classroom instruction, in that order, were the effective learning modes recognized by middle school juniors; patterns of recognition did not change across grades. Application of a piecewise latent growth model uncovered a gap in effectiveness levels between middle school and high school, characterized by a drop in average levels for all three methods. This result implies a discontinuity in terms of vertical articulation in the curriculum. Using a multivariate latent growth model, this study investigated the relationship between the initial level of effectiveness and the slope of change for these three learning methods. Study findings revealed that a strong correlations among initial levels of effectiveness; slopes in the same direction indicated that students believed different methods effective simultaneously. This implies that students utilized private tutoring as an enrichment strategy, not a remedial strategy. A higher initial level of effective classroom instruction was linked to a steeper decrease in the slopes for the effectiveness of self-study and private tutoring. Indeed, a higher initial level of effective self-study related to a steeper decrease in the slopes for the effectiveness of classroom instruction and private tutoring. However, the initial level of effectiveness for private tutoring was not related to reduced slopes for other learning methods These results suggest that to decrease dependency on private tutoring in high school, students must experience effective classroom instruction and self-study during middle school.