Spatial autocorrelation of smuggling crimes in Iran provinces using weight matrices

Spatial autocorrelation of smuggling crimes in Iran provinces using weight matrices

Statistical autocorrelation describes spatial connections between observations/regions. As an application of the theory, this paper examines the smuggling of goods and foreign exchange in 31 provinces of Iran between 2011 and 2017. Coordinates are specified between spatial units using spatial weights matrices, sometimes referred to as $ m{W}$. The local structure and influences of the spatial relationships are described by these matrices. The choice of a weight function for the spatial weights matrix plays a significant role in the results of the spatial analysis. Here, we compare the statistical performances of different types of spatial weighting matrices based on Queen contiguity (first-order and second-order neighbor weights). The first-order contiguous Queen with the neighbors of 148 shows a significant correlation in goods smuggling and foreign exchange.