Some New Demographic Equations in Survival Analysis under Generalized Population Model: Applications to Swedish and Indian Census Age-Data for Estimating Adult Mortality

Subrata Lahiri, International Institute for Population Sciences (IIPS)

This paper presents development of some new formulas in estimating "10-year conventional and cumulative life table survival ratios", defined by the following ratios--L(x+10, x+15)/L(x, x+5) and T(x+10)/T(x) in life table terminology respectively, from two enumerations (not necessarily multiple of 5 years apart) of any closed population that follows a generalized population model. Furthermore, it is assumed that the age-specific growth curve should resemble closely to a second-degree polynomial. Attempts have also been made to establish algebraic relationships between census survival ratios (conventional and cumulative) and the corresponding life table survival ratios under GPM. The formulas, developed here, have been applied to sufficiently accurate age-data of Sweden and that of India subject to response biases in age reporting. The proposed technique, which works quit well in assessing adult mortality even when the age-data are distorted due to age misreporting, may be extended for population projection and other demographic estimations.