A comparison of Poisson-one-inflated power series distributions for modeling rural out-migration at the household level

Abstract

A class of Poisson-one-inflated power series distributions (the binomial, the Poisson, the negative binomial, the geometric, the log-series and the misrecorded Poisson) are proposed for modeling rural out-migration at the household level. The probability mass functions of the mixture distributions are derived and fitted to the data in Sharma (1985). Their performance in describing the total number of out- migrants from a household are subsequently compared. It was found that while the Poisson-one-inflated log-series distribution was superior to others for modeling out-migration from semi-urban areas and growth centres, the Poisson-one-inflated geometric distribution was better in the case of out-migration from remote areas.

Global Journal of Mathematical Sciences Vol.3(2) 2004: 113-122