Testing Longitudinal Change in Spatial Distributions and Patterns: An Extension of the Spatial Point Pattern Test

The spatial point pattern test (SPPT) provides a nonparametric framework for comparing spatial distributions of events across areal units, but its application to longitudinal data has been limited to sequential pairwise comparisons or a fixed-baseline extension with a thin trajectory taxonomy. This article develops a fully longitudinal SPPT that bootstraps all time periods symmetrically, introduces simulation-based p values, expands trajectory classification to include linear and quadratic trend detection, and classifies baseline crime levels using Jenks natural breaks. All classifications are strictly local, preserving the unit-level specificity and direct mapability that distinguish the SPPT from cluster-based trajectory methods. The method is demonstrated using twenty-three years of residential burglary data across 1,019 dissemination areas in Vancouver, British Columbia. Results reveal widespread distributional oscillation without directional trend, spatially structured decelerating changes consistent with prior research on the changing geography of residential burglary, and substantial crime reduction (drop to zero) in nearly one-fifth of the study area. The method complements existing trajectory approaches while offering a formal inferential framework unavailable in prior longitudinal SPPT extensions.