Search

Conventional methods used to identify crime hotspots at the small-area scale are frequentist and employ data for one time period. Methodologically, these approaches are limited by an inability to overcome the small number problem, which occurs in spatiotemporal analysis at the small-area level when crime and population counts for areas are low. The small number problem may lead to unstable risk estimates and unreliable results. Also, conventional approaches use only one data observation per area, providing limited information about the temporal processes influencing hotspots and how law enforcement resources should be allocated to manage crime change. Examining violent crime in the Regional Municipality of York, Ontario, for 2006 and 2007, this research illustrates a Bayesian spatiotemporal modeling approach that analyzes crime trend and identifies hotspots while addressing the small number problem and overcoming limitations of conventional frequentist methods. Specifically, this research tests for an overall trend of violent crime for the study region, determines area-specific violent crime trends for small-area units, and identifies hotspots based on crime trend from 2006 to 2007. Overall violent crime trend was found to be insignificant despite increasing area-specific trends in the north and decreasing area-specific trends in the southeast. Posterior probabilities of area-specific trends greater than zero were mapped to identify hotspots, highlighting hotspots in the north of the study region. We discuss the conceptual differences between this Bayesian spatiotemporal method and conventional frequentist approaches as well as the effectiveness of this Bayesian spatiotemporal approach for identifying hotspots from a law enforcement perspective. [ABSTRACT FROM AUTHOR]

Objectives: Explore Bayesian spatio-temporal methods to analyse local patterns of crime change over time at the small-area level through an application to property crime data in the Regional Municipality of York, Ontario, Canada. Methods: This research represents the first application of Bayesian spatio-temporal modeling to crime trend analysis at a large map scale. The Bayesian model, fitted by Markov chain Monte Carlo simulation using WinBUGS, stabilized risk estimates in small (census dissemination) areas and controlled for spatial autocorrelation (through spatial random effects modeling), deprivation, and scarce data. It estimated (1) (linear) mean trend; (2) area-specific differential trends; and (3) (posterior) probabilities of area-specific differential trends differing from zero (i.e. away from the mean trend) for revealing locations of hot and cold spots. Results: Property crime exhibited a declining mean trend across the study region from 2006 to 2007. Variation of area-specific trends was statistically significant, which was apparent from the map of (95 % credible interval) differential trends. Hot spots in the north and south west, and cold spots in the middle and east of the region were identified. Conclusions: Bayesian spatio-temporal analysis contributes to a detailed understanding of small-area crime trends and risks. It estimates crime trend for each area as well as an overall mean trend. The new approach of identifying hot/cold spots through analysing and mapping probabilities of area-specific crime trends differing from the mean trend highlights specific locations where crime situation is deteriorating or improving over time. Future research should analyse trends over three or more periods (allowing for non-linear time trends) and associated (changing) local risk factors. [ABSTRACT FROM AUTHOR]

Spatial cluster detection is an exploratory spatial data analysis technique that identifies areas or groups of areas with disproportionately high risk. Several local cluster detection methods have been developed; yet no research has critiqued these methods as they contribute to spatial studies of crime. This study aims to identify the locations of drug offence hotspots in Toronto and compare the clusters detected through four methods: (1) spatial scan statistic - Euclidean distance, (2) spatial scan statistic - non-Euclidean contiguity, (3) flexibly shaped scan statistic, and (4) local Moran's I. It was found that all methods detected clusters in the downtown, with fewer methods detecting clusters in the west and east of Toronto. It was observed that the spatial scan statistic detected the largest and most circular clusters, making it a suitable tool to inform general policing initiatives and highlight possible variables to be included in confirmatory research. The local Moran's I method, in contrast, found the smallest and most compact clusters, indicating that it is an appropriate test for identifying areas where resource intensive crime prevention and policing efforts should be targeted. [ABSTRACT FROM AUTHOR]