150 likes | 239 Views
Something Fishy? Uncovering heterogeneity in the distribution of crime victimisation in general populations. Tim Hope and Paul Norris SCCJR (CJ-QUEST) University of Edinburgh December 2008. The Distribution of Property Crime in the BCS. Understanding and Modelling the Distribution of Crime.
E N D
Something Fishy? Uncovering heterogeneity in the distribution of crime victimisation in general populations Tim Hope and Paul Norris SCCJR (CJ-QUEST) University of Edinburgh December 2008
Understanding and Modelling the Distribution of Crime • The distribution shown on the previous slide poses two questions :- - Substantive question: What is the data generation process that underpins the distribution? - Statistical question: What kind of dependent variable is best employed to model victimisation?
Theoretical Models of Victimisation • Simple Exposure (pure heterogeneity) • Simple RV (pure state-dependency) • Mixture Model - Large proportion of the population experience no victimisation - Small proportion of the population experience chronic victimisation - One or more groups for low-level victimisation T. Hope and A. Trickett (2004). ‘La distribution de la victimationdans la population’, Déviance et Société, 28 (3), 385-404.
Dependent Variable for Victimisation Research • Type of crime victimisation • Type of incident • One type verses more generalist victim • Frequency of crime victimisation • Nominal (0,1), Ordinal (0, 1, 2+), Count (0-n) • Distribution of count variables – Poisson verses Negative Binomial
Data • British Crime Survey - England and Wales (BCS) • 1992, 1996, 2001, 2003/04, 2006/07 • Crime types • Household Property Crime (6 questions) • Count data (victim screeners, capped at 6) • Scottish Crime Victimisation Survey (SCVS) • 1993, 1996, 2000, 2003, 2006
Latent Class Models • Latent Class Analysis (LCA) is analogous to cluster analysis but: • -Can handle missing data -Can handle non-normal data • -Can be used with longitudinal data
Victim Type A Simple LCA Model Age Household Income Neighbourhood Type + Vandalism Victimisation + Forced Entry Victimisation = Total Victimisation Household Theft Victimisation LCA indicator considers both level of victimisation and type of crime
Accuracy Verses Parsimony • How many groups are required? -range of statistical indicators -substantive interpretation is crucial • Within group variation?
Distribution of Victimisation Indicators • Count data often modelled using Poisson distribution • Victimisation appears to follow Negative Binomial distribution • What about zero-inflation?
ABIC for BCS Data • Lower ABIC figures represent better fit between model and data Results based on Negative Binomial Distribution. Results using zero-inflated Negative Binomial reveal an identical pattern but exhibit a slightly worse fit to the data • ABIC suggests six groups should be used
Results for Scottish Data • Distribution of property crime in Scottish data is very similar to BCS Results based on Negative Binomial Distribution. Results using zero-inflated Negative Binomial reveal an identical pattern but exhibit a slightly worse fit to the data • ABIC statistic suggests 4 class solution is optimal
Summary • Overall distribution obscures heterogeneity • Heterogeneity of both substantive and statistical interest • Most “uncertainty” occurs around the middle of the distribution • Key issues around how solution is affected by sample design, prevalence of incidents and how useful apparent classes are for analysis