This document discusses using a weighted dissimilarity function with the CLARANS clustering algorithm for spatial data. It defines spatial and non-spatial attributes of data and different relationship types in spatial data. The weighted dissimilarity function calculates distances between polygon spatial and non-spatial attributes. CLARANS is described as an iterative algorithm that considers random neighbors to find local minima cluster costs. Experimental results on student/class data show the weighted dissimilarity function improves silhouette index and runtime over varying weights and numbers of clusters. Increasing clusters increases computational time linearly.