A systematic review of group-based trajectory analyses of spatial patterns of crime

Read the full article See related articles

Listed in

This article is not in any list yet, why not save it to one of your lists.
Log in to save this article

Abstract

Originally developed in criminology to examine criminal careers, trajectory modelling, and particularly group-based trajectory modelling, is now widely applied to analyse a variety of other phenomena, including spatial crime dynamics. However, despite its use in this area of research for over two decades, no systematic review has synthesized its application to spatial-temporal analyses of crime, including in terms of the typical model specifications and findings. This article addresses that gap through a comprehensive literature search that identified 1,397 non-duplicate publications, of which 28 were relevant and collectively reported on 47 distinct group-based trajectory modelling analyses. Most of the analyses identified between three and eight latent trajectories, generally corresponding to groups of places with stable levels of negligible, low-to-moderate, or chronic crime. While some analyses uncovered more unique longitudinal trends, increasing the number of groups often produced only relatively subtle variations of these broader patterns, raising issues about their substantive value. Methodologically, while the analyses often compared and ultimately selected in their final models particular numbers of groups, data distributions, trajectory shapes, and diagnostic metrics, there was considerable variation in how thoroughly the modelling process was conducted and reported. Additionally, unique issues relatively specific to this area of research were also identified, including in terms of the treatment of missing data and zero-crime units, and the diversity of spatial resolutions. Overall, the review highlights the need for more systematic modelling practices and clearer reporting standards. The article concludes with recommendations for future applications of group-based trajectory modelling to spatial crime dynamics in order to advance our substantive understanding of how crime at different places evolves over time.

Article activity feed