The effect of socioeconomic background on children’s educational attainment has long been investigated, but the extent to which high school mediates this effect is largely unknown. This study investigates the direct effect of family income on educational attainment not mediated by the rank of high school, using data from the longitudinal surveys of ninth-grade students and their mothers as a sample in Japan (n = 1,761). The results indicate that the total effects of family income on university enrollment and educational rank are substantial. Regression with residuals (RWR) analysis reveals that although high school type mediates the effects of family income on university enrollment and education rank to some extent (about 22% and 29%, respectively), the direct effects of family income are considerable. The result also suggests a positive interaction effect between family income and high school selectivity on the education rank. In other words, the higher the rank of high school, the greater the direct effect of family income on the education rank, implying a boosting advantage pattern rather than a compensatory advantage. I argue that the analytical framework used in this study can be applied to many methodological settings in sociological studies.