Acknowledging the Influence of Geographic Outliers: An Analytical Caution in Examining Chronic Repeat Victimization

Abstract

Evidence suggests that repeat victimization at specific addresses is higher in areas identified by computer algorithms as being ‘hot spots’. Recently developed explanations for the existence of repeat victimization2 within high-crime areas extend the tenets of crime pattern and routine activities theories. The emerging hypotheses account for the apparent overlap by suggesting that offenders in the area are more prolific (returning to known or ‘quickly replenishing’ opportunity more frequently) or that there is a concentration of nearby offenders whose routine activities expose them to the same victims.3 The current demonstration presents a methodological caution to those engaged in this stream of research. Some of the apparent overlap between chronic repeat victimization at precise addresses and hot spots can be attributed to analytic factors associated with the mapping software used to identify geographic concentrations of crime. This implies the need for careful consideration of the appropriateness of analytic strategies that are significantly influenced by the presence of extreme values or spatial outliers.