Prediction of Life Expectancy

Abstract

Life expectancy refers to the number of years a person is expected to live. The life expectancy for a particular person or population group depends on several variables such as their lifestyle, access to healthcare, diet, economical status and the relevant mortality and morbidity data. To make predictions on lifetime of a person, four independent variables were collected from death certificates and they are sex, cause of death, profession and race. During the study, impact of the four factors on life expectancy is measured. Independent samples t test and Kruskal wallis test were used to examine the independent variables. Except for the race, Kruskal wallis test indicated that age is not same across different categories of sex, cause of death and profession. But independent Sample’s t test indicated that there was no significant difference between males and females and also it resulted in the fact that, there was no evidence to say that there is a difference between the lifetimes of cancer patients and languishing people. However, Kruskal wallis test and independent sample’s t test gave contradictory results. After analyzing the variables, to make predictions, General Linear Models (GLM) and Kaplan Meier estimates were used. According to the results obtained from General Linear Models, sex and cause of death were statistically significant in the model. Being a non parametric test Kruskal wallis test always aligned with the model fitting results.