Ordinal Logistic Regression Analysis of Correlates of Crime Severity: The case of Tigray Region, Ethiopia
No Thumbnail Available
Date
2011-06
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Addis Abeba university
Abstract
The cumulative logit or the proportional odds regression model is one of the popular choices to
study covariate effects on ordinal responses. This research is aimed at modeling a categorical
response i.e., crime severity outcome in terms of some predictors, determines the goodness of fit
as well as validity of the assumptions and selecting an appropriate and more parsimonious model
there by proffer useful suggestions and recommendations. The proportional odds model was used
as a tool to model the two major factors viz. socio-demographic (sex, age, education status,
marital and employment status) and environmental (urban, rural) that affected the outcomes of
crime. The fit, of the model was illustrated with 2,753 crime records obtained from regional
police commission. This study provides some graphical and numerical methods for checking the
adequacy of the proportional odds regression model. The methods focus on evaluating crime
severity for specific covariate effects. The tested model showed good fit and performed
differently depending on categorization of outcome, adequacy in relation to assumptions and
goodness of fit. Findings of this study have shown that criminal’s age, educational background,
employment status, marital status, and area of crime committed are significantly affect the
outcome of crime. Criminals who are young, illiterate, employed, single household member, and
rural areas of crime committed increased the odds of being in either serious or medium crime
categories, than those criminals who are aged, educated, unemployed, married, and urban area of
crime committed. The finding of this study indicates that the rise of crime in the region is
generally as the result of direct effect of poverty. Policies and plans have to be put in place to
improve young age individual education and crime prevention agencies need to issue to
implement crime prevention strategies.
Keywords: crime severity, ordinal logistic regression, proportional odds model
Description
Keywords
Crime Severity, Ordinal Logistic Regression, Proportional Odds Model