A Study of Some Demographic Phenomena using Statistical Techniques

Abstract

Demography is a discipline that is particularly suitable for the use of models newlinebecause its events and entities-such as age, number of persons, number of births newlineand the number of deaths are unambiguously numerical. A population is a newlinecollectivity that increases or decreases by the entry or exit of members. This entry newlineor exit of members is known as fertility and mortality analysis in general. Further, newlineat a given moment each member of a population has a specifiable age, which newlineincreases by one unit with the passage one unit of time. The event or risks that newlinedetermine entry or exit vary with age, defined as the time since an event that fixes newlineage zero in specified population. For estimating fertility and mortality patterns newlineseveral parametric and non-parametric techniques have been proposed. The newlineparametric ones are non linear models that represent the mortality pattern as a newlinefunction of age and a number of parameters. The nonparametric approach does not newlineinvolve functional forms or parameters of such forms. In general non-parametric newlinemethods apply to very wide families of distributions rather than only to families newlinespecified by a particular form. This thesis attempts to analyze these two important newlinedemographic phenomena using different statistical techniques. newline