Abstract

In view of the overall spatial heterogeneity and simple and easy-to-obtain geographic coordinate information in social surveys, Our study discusses the "spatial effect" and "sample representativeness", systematically studies the role of geographic auxiliary information in view of spatial heterogeneity, introduces geographic coordinates into the spatial sampling survey, and discusses the statistical properties of the new method and its applicability in practice by using the stochastic simulation method and the empirical research method. First, the hierarchical sampling is improved by using geographic coordinates, and through the simulation of spatial linear model and spatial random field, it is found that the spatial stratified sampling with the participation of geographic coordinates can improve the estimation accuracy, especially when the sample size is large. Secondly, for the population with strong spatial heterogeneity, the sampling unit sorting problem is transformed into the single-source shortest path problem, and the Dijkstra algorithm is used to carry out ordered cube sampling, which effectively solves the spatial autocorrelation problem. Thirdly, taking the Chinese Household Survey as an example, the empirical study shows the accuracy and efficiency of the spatial stratification sampling and ordinal cube sampling methods with the participation of geographical coordinates in estimating the per capital disposable income and the number of permanent residents in urban areas.