Использование понятия относительной погрешности оценивания для расчета выборки из генеральной совокупности с малой долей признака

Abstract

The paper elucidates how to determine a sample size from the general population if the analyzed variable’s value is small. The Ukrainian sociologist Mykola Churylov was the first who highlighted this problem in the national scientific literature. The author argues that it is necessary to calculate both absolute and relative error. He also points to the sample size from the general population (in case of small variable’s value) that should be larger than while using traditional formulas. A new enhanced approach presented in the paper includes using within a representative random sample only a priori characteristics.