COMPSTAT 2022: Start Registration
View Submission - COMPSTAT2022
Title: Spbsampling: An R package for spatially balanced sampling Authors:  Francesco Pantalone - University of Southampton (United Kingdom) [presenting]
Roberto Benedetti - University of Chieti - Pescara (Italy)
Federica Piersimoni - ISTAT (Italy)
Abstract: In environmental, geological, biological, and agricultural surveys, among many others, usually, the main feature of the population of interest is to be geo-referenced. In these situations, we can expect that units closer to each other provide less information about a target of inference than units farther apart, as outlined in Tobler's first law of Geography. Therefore, it would be beneficial to the efficiency of the final estimates to consider the spatial dependence. Since traditional sampling designs generally do not take into account the spatial features of the population, several spatially balanced sampling designs have been introduced in the literature, which select samples well spread over the population of interest, or spatially balanced samples. We introduce the R package Spbsampling, which implements some of the designs recently introduced. We focus on sampling designs that achieve spatially balanced samples by means of an MCMC algorithm and the use of a summary index of a distance matrix. This allows for wider applicability, as a distance matrix can be defined for units according to variables different from geographical coordinates.