O pewnym problemie analizy wzdłużnej. Ocena zdarzeń demograficznych metodą kohort

Abstract

A longitudinal analysis requires utilization of statistics of demographic events located in cohorts history. And so, too often, and in practice always, these statistics are prepared by age (in completed years) of units or by antiquity (in completed duration periods) of cohort's elements. That should be then separated what refers to age given in numbers of events regarding each of both cohorts contributing to event's appearance at that age. The article is devoted just to a method allowing to do such a separation, explaining it on the ground of statistics referring to first marriages (of single people) of French women in 1962. Such a re-grouping of data by generations delivers to satisfactory results.