The (in)appropriateness of aggregating across crime types

Researchers in the spatial analysis literature commonly aggregate crime types: property, violent, and total crime, for example. This practice is done for a number of reasons that range from low crime counts to confidentiality to the form of data provided to researchers. But is this practice appropri...

Full description

Saved in:
Bibliographic Details
Published in:Applied geography (Sevenoaks) 2012-11, Vol.35 (1-2), p.275-282
Main Authors: Andresen, Martin A., Linning, Shannon J.
Format: Article
Language:English
Subjects:
Aggregation
Crime at places
Crime types
Spatial point pattern test
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Researchers in the spatial analysis literature commonly aggregate crime types: property, violent, and total crime, for example. This practice is done for a number of reasons that range from low crime counts to confidentiality to the form of data provided to researchers. But is this practice appropriate, particularly in the context of studying spatial patterns? In this paper, we analyze the appropriateness of such crime type aggregations. We find that in almost all cases, aggregating crime types is inappropriate for polygon-based analyses. Only in the cases of micro-spatial units of analysis (street segments) may these aggregations be appropriate. Therefore, aggregating crime types in a study that implicitly or explicitly considers the spatial patterns of crime is inappropriate. ► We examine the appropriateness of aggregating across crime types. ► We use an area-based spatial point pattern test. ► Spatial crime patterns differ significantly across crime types. ► Most often aggregating crime types is inappropriate. ► Such aggregations may only be appropriate at micro-spatial scales.
ISSN:0143-6228
1873-7730
DOI:10.1016/j.apgeog.2012.07.007