Integration of Spatial Data from Two Independent Surveys: A Model-Based Approach Using Geographically Weighted Regression

In large-scale surveys, many practical challenges arise, including increased expenses for data collection, a growing need for statistics at a small-area level, decreasing response rates, and the need for timely estimates. In recent years, integrating data from multiple surveys has emerged as one of the most popular approaches for making inferences about finite population. The integration of data offers solution to these challenges and gives precise estimates of the population parameters. For spatial data, the association between the study variable and covariates differs across various locations. It is called spatial non-stationarity. This article proposes a novel spatially integrated estimator for finite population total using geographically weighted regression model. This proposed approach combines data from two independent surveys, harnessing the power of spatial information. A simulation study was then carried out to evaluate the statistical properties of the proposed spatially integrated estimator. Additionally, a spatial proportionate bootstrap method for estimating the variance of the proposed integrated estimator has been introduced.