Family background and school performance: a non-parametric approach

Abstract

A large theoretical and empirical literature suggests that there is an intergenerational transfer of socioeconomic characteristics of the family, such as income and education. In this sense, the objective of this work is to investigate a relation between the familiar characteristics and a school proficiency. For this, a non-parametric model, the generalized additive model (GAM), was estimated through a two stages non-parametric estimator. The micro data of Saeb 2011 and the School Census were used, corresponding to the grades of the apprenticeship examination of students in the 5th and 9th grade of public and private schools. The results suggested as variables linked to the family background exert a great influence on the students’ school performance in a variable with variables related to the school, the teachers and the principals, especially for the students of the public school.