Partial and semi-partial statistics of spatial associations for multivariate areal data

Abstract

The analysis of correlation structures among multivariate spatially aggregated data has become increasingly important and poses substantial challenges. This article concerns the development of partial and semi‐partial statistics of spatial associations in the context of multivariate spatial areal data extending Moran's I and Geary's C. The proposed statistical tools describe global or local associations among spatially aggregated measurements for pairs of different components conditional on all remaining components. The new statistics are tested through a simulation study and illustrated using aggregated data on the COVID‐19 pandemic recorded at the Community level in Spain.