Comparação entre metodologias de idade-período-coorte para o estudo de probabilidades de progressão por série selecionadas no Brasil

The aim of this article is to realize a comparative methodological exercise of two APC estimators: the conventional restrict estimator obtained by the generalized linear models (MLGR) and the so-called intrinsec estimator (EI). Both instruments are contrasted by means of the parameter estimates and of the model’s efficiency. The object of our interest are the contributions of age, period and cohort effects to temporal changes in grande progression probabilities in Brazil. The APC modelling of grade progression probabilities is justified as age, period and cohort effects may signficantly affect school transitions: age effects reflect both the mininal age of school entry as the trade-off between study and work, which becames strong along the educacional carrier; period effects are associated with different economical and political conjuctures, as well as with the state of the educational policies; finally, cohort effects reflect social attributes proper of some group of individuals. This article shows the potentiality of the solution to the ageperiod-cohort model based on intrinsec estimator, which presents excelent statistical properties, namely: best adjust to predicted values, small variance, and significance of the parameters. Therefore, projections of grade progression probabilities based on this estimator may be very promissing.