層化集落抽出에서의 不偏比抽定量에 關한 硏究

Abstract

母數의 推定量으로서 標本資料와 補助資料를 함께 利用하는 比推定量은 標本統計量보다 推定量의 精度를 높일 수 있으나 一般的으로 不偏推定量이 되지 못한다.

本 硏究는 層化集落抽出인 경우의 不偏比推定量을 얻기 위해서 單純任意抽出에서의 「Hartely-Ross」의 不偏比推定量을 擴張하여 不偏分離比推定量과 不偏結合比推定量을 導出하였고 또한 各各의 分散을 구하여 有效性을 比較 分析하였다.

Unbiased ratio estimators in stratified clusler sampling are presented. The two unbiased ratio estimators, separate unbiased ratio estimator and combined unbiased ratio estimator, are obtained by extension of the Hartely-Ross unbiased ratio estimator and their variances are developed by using the moments of bivariate sample cumulants. the relative efficiencies of these estimators with respect to the biased ratio estimators are investigated.

Unbiased ratio estimators in stratified clusler sampling are presented. The two unbiased ratio estimators, separate unbiased ratio estimator and combined unbiased ratio estimator, are obtained by extension of the Hartely-Ross unbiased ratio estimator and their variances are developed by using the moments of bivariate sample cumulants. the relative efficiencies of these estimators with respect to the biased ratio estimators are investigated.