On the Level of Performance of Selected Link Functions in the Identification of Poverty Correlates in Nigeria

Abstract

Effective poverty reduction requires proper identification of correlates of poverty. This study therefore used the generalized linear modeling approach to identify poverty correlates in Nigeria based on the 2003/2004 National Living Standard Survey (NLSS) data. This approach is adequate because it respects the underlying distribution of poverty indicators as the response variable in models to facilitate inference on poverty. The results of the study show that sex of the household head, age in years of the household head, father’s education level, father’s work, mother’s work, household size, occupation group of the household head, and educational group for highest level attained by the household head are strong correlates of poverty while mother’s education level and literacy of the household head are weak correlates of poverty. Having fewer numbers of children and ensuring education of children are recommended as measures to reduce the burden of poverty.

Keywords: Link functions, Poverty correlates, Human dignity, Lack of understanding, Regression