A restricted randomization test for determining the statistical significance of the relative increase in population in the vicinity of urban facilities

Abstract

This paper proposes a restricted randomization test for determining the statistical significance of the relative increase of population in the vicinity of urban facilities. First, we develop a method for simulating realizations of the spatial distributions of change in population under the null hypothesis, which is that there is no relationship between the spatial distribution of change in population and the distribution of facilities. Second, based on the method, we propose a restricted randomization test for determining whether the relative increase in population around facilities is statistically significant. Third, we develop a method for correcting multiple testing problems when the proposed test is repeatedly executed using different parameters. Finally we conclude with suggestions for future study.