This paper, based on the report from Tomita & Muta (2010), re-examined the influence of school factors and family-related factors on learning achievements in Malawi, which is one of the very low-income countries, and confirmed Heyneman and Loxley's finding (1983) : School factors are more influential and family-related factors are less influential on learning achievements in lowincome countries (called HL effect) in more comprehensive manner. Two different statistical methodologies were employed: 1)separate analyses of school- and pupil-level variables based on linear regression and structural equation modeling (SEM), and 2)simultaneous analyses of school- and pupil-level variables based on hierarchical linear modeling (HLM). Results from both methodologies confirmed HL effect in the case of Malawi. Finally, the policy implication of the result was discussed.