Spectral analysis of reported crime data

Abstract

This paper applies spectral analysis to data on reported crimes for a medium-sized city. The objective is to examine certain aspects of the statistical properties of crime data, as well as to illustrate the utility of the spectral-analytic method for criminal justice research. A discussion of the theoretical underpinnings of the spectral analysis technique is provided through analogy with familiar least squares regression methods. Both spectral and cross-spectral analysis are then employed to test selected hypotheses concerning the data.