Project CEMAPRE internal
|Title||Analysis of school failure through multilevel models|
|Participants||Amélia Bastos (Principal Investigator), Maria Eugénia Ferrão|
|Summary||Multilevel models offer social science scientists a unique approach to the understanding of|
individual, contextual and interaction factors implications on a given outcome variable. The
association between education and poverty or social exclusion is well documented in the literature.
It is also well known that there are substantial school variations in educational outcomes and that
compositional school effect influences the student performance, and therefore educational success.
This project investigates school failure with microdata in Portugal. Drawing on multilevel models we
analyse the association of the number of student's failures in compulsory education - our endogenous
variable - with factors such as: age, gender, household sociodemographic and economic background and
children's living conditions - our covariates. Furthermore, we also attempt to shed light on the
impact of school in the endogenous variable.
The factors analyzed are grouped in two levels: schools (and compositional characteristics of
population served), and individual sociodemographic and economic attributes of the students in these
schools. Disentangling the effects of these two sources of variation in school outcomes is important
for the debate of education public policy as well as for the design of poverty reduction policies.