统计与决策2025，Vol.41Issue(9)：35-41,7.DOI:10.13546/j.cnki.tjyjc.2025.09.006
样本轮换下事后分层的最短距离校准估计方法
Shortest Distance Calibration Estimation Method for Poststratification Under Sample Rotation
摘要
Abstract
In order to reflect the changes and development patterns of the surveyed population in a timely and accurate man-ner,the survey department usually applies the continuous sampling method to conduct the investigation of the surveyed popula-tion.For the sample rotation method of continuous sampling survey,this paper applies the post-stratification regression estimation form to construct the current population mean estimator.The estimator is calibrated with the previous survey value and the current auxiliary variable as constraints,and the calibration weights are obtained by using the shortest distance method to construct the calibration combined estimator.At the same time,the variance expression and the minimum variance of the calibration combina-tion estimator are offered and compared with the post-stratification regression combination estimator.Finally,Monte Carlo random simulations are used to verify the calibration efficiency of the combined estimators under several common distributions.In the two consecutive surveys of the data of small and micro enterprises in the"four below"selenium industry from 2022 to 2023,the sam-pling design and calibration estimation method are used to design and estimate the sampling survey scheme.The results of both the simulation experiments and the case analysis show that the post-stratification calibration estimation method has better applica-bility and estimation effect.关键词
样本轮换/事后分层/辅助变量/校准组合估计量Key words
sample rotation/poststratification/auxiliary variables/calibration combination estimator分类
社会科学引用本文复制引用
马金萍,雒晶晶..样本轮换下事后分层的最短距离校准估计方法[J].统计与决策,2025,41(9):35-41,7.基金项目
国家社会科学基金重大项目(21&ZD147) （21&ZD147）
国家社会科学基金资助项目(23BTJ024) （23BTJ024）
陕西省哲学社会科学研究专项(2025YB0215) （2025YB0215）
西安财经大学研究生创新基金资助项目(23YCZC08) （23YCZC08）
