Methods for the Analysis of Biographical Data

Abstract

This article suggests three approaches for analyzing biographical data. The non-parametric approach deals with the interaction between various demographic phenomena and leads to various types of dependence between them. The parametric approaches attempts to incorporate the heterogeneity observed in the population into the analysis, by linking temporary behavior with various characteristics that are either dependent on or independent from time. The semiparametric approach permits the simultaneous exploration of interactions and heterogeneity. The issue of non-observed heterogeneity in data is also explored. These methods are illustrated using examples of the application of data taken from surveys.