İki özellik arasındaki ilişki hesaplamalarında yanlış tahminleri ve yanlılık tahminlerinin simülasyon tekniği ile araştırılması

Abstract

In this study, applicable to variety of economic data that non-normal gamma distribution is considered. For this reason related to gamma distribution with parameter a and p, five hundered income average and five hundered house-hold are calculated simultaneously by use of a computer. The relationship between this two variables colud be measured by the use of, least square methods, regression and correlation coefficient techniques are compared by Jackknife methods. This comparision results strongly suggest that, results obtained by Jackknife estimation technique much more dependable than, results that obtained by least square methods. Key Words : Jackknife, bias, simulation, unbiased, gamma distribution, least suqare, regression coeffcient, correlation coefficinet.