Abstract
AbstractMuch previous research on behavioural case linkage has used binary logistic regression to build predictive models that can discriminate between linked and unlinked offences. However, classification tree analysis has recently been proposed as a potential alternative owing to its ability to build user‐friendly and transparent predictive models. Building on previous research, the current study compares the relative ability of logistic regression analysis and classification tree analysis to construct predictive models for the purposes of case linkage. Two samples are utilised in this study: a sample of 376 serial car thefts committed in the UK and a sample of 160 serial residential burglaries committed in Finland. In both datasets, logistic regression and classification tree models achieve comparable levels of discrimination accuracy, but the classification tree models demonstrate problems in terms of reliability or usability that the logistic regression models do not. These findings suggest that future research is needed before classification tree analysis can be considered a viable alternative to logistic regression in behavioural case linkage. Copyright © 2012 John Wiley & Sons, Ltd.
Highlights
Much previous research on behavioural case linkage has used binary logistic regression to build predictive models that can discriminate between linked and unlinked offences
Separate Receiver Operating Characteristic (ROC) curves were constructed for each logistic regression model and the classification tree model for the burglary and car theft datasets. These analyses provided an insight into the relative ability of logistic regression and classification tree analysis to construct predictive models for the purposes of case linkage
Classification tree analysis was conducted on the training sample and subsequently applied to the test sample
Summary
Much previous research on behavioural case linkage has used binary logistic regression to build predictive models that can discriminate between linked and unlinked offences. Using this and other methodologies, a number of studies have demonstrated that certain types of offender behaviour can be used to distinguish between linked and unlinked offences to a statistically significant extent This evidence spans a variety of different crime types, including burglary, robbery, car theft, sexual assault, homicide, and arson (e.g., Bennell, Jones, & Melnyk, 2009; Melnyk, Bennell, Gauthier, & Gauthier, 2010; Santtila, Fritzon, & Tamelander, 2004; Tonkin et al, 2008; Woodhams & Toye, 2007). Woodhams and Toye (2007) showed that a logistic regression model combining three types of offender behaviour (control, planning, and intercrime distance2) was able to distinguish between linked and unlinked commercial robberies with a high degree of accuracy (AUC = 0.95; Swets, 1988) This level of accuracy suggests that behavioural case linkage may be a viable procedure for the police to use. One recent methodological issue that has been explored is the use of classification tree analysis instead of logistic regression to produce statistical models that can discriminate between linked and unlinked offences (Bennell, Woodhams, & Beauregard, in preparation)
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: Journal of Investigative Psychology and Offender Profiling
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
