Improvement of Thematic Classification in Offender Profiling: Classifying Serbian Homicides Using Multiple Correspondence, Cluster, and Discriminant Function Analyses

Abstract

This paper investigates thematic classification of homicides for the purpose of behavioural investigative analysis (e.g. offender profiling). Previous research has predominantly used smallest space analysis (SSA) to conceptualise and classify offences into thematic groups based on crime scene behaviour data. This paper introduces a combined approach utilising multiple correspondence analysis (MCA), cluster analysis (CA), and discriminant function analysis (DFA) to define and differentiate crime scenes into expressive or instrumental and impersonal or personal crimes. MCA is used to derive the latent structural dimensions in the crime data and produce quantitative scores for each offence along these dimensions. Two-step CA was then utilised to classify offences. Offence dimensional scores were then used to predict cluster membership under DFA, producing cluster centroids corresponding to MCA dimensions. Centroids were plotted on the MCA correspondence map to simultaneously conceptualise crime classification and the latent structure of the Serbian crime data. Classification of offences based on MCA dimensional scores were 91.5% accurate. This MCA-CA-DFA approach may reduce some of the more subjective aspects of SSA methodology used in classification, whilst producing a product more amenable to objective and cumulative review. Implications for offender profiling research utilising SSA and this approach are discussed.

The patterns of homicide offence characteristics and their associations with offender psychopathology

“…Multiple correspondence analysis is analogous to principal component analysis; it is an exploratory multivariate technique that allows the pattern analysis of the associations between more than two multilevel categorical variables (Joyal, Côté, Meloche, & Hodgins, ). JCA produces a graph based on meaningful underlying latent dimensions that capture most of the inertia (Goodwill, Allen, & Kolarevic, ; Greenacre, ). Inertia indicates the amount of variance explained by the solution (Goodwill et al, ).…”

“…JCA produces a graph based on meaningful underlying latent dimensions that capture most of the inertia (Goodwill, Allen, & Kolarevic, ; Greenacre, ). Inertia indicates the amount of variance explained by the solution (Goodwill et al, ). Each point in the graph represents a category in the analysis (Husson & Josse, ).…”

“…Each point in the graph represents a category in the analysis (Husson & Josse, ). Points located in proximity to one another are considered more similar than distal points and, thus, may be indicative of theoretically meaningful associations (Abdi & Valentin, ; Goodwill et al, ). In general, the further a category is from the origin, the more distinct it is.…”

Characterising the Personality of the Public Safety Offender and Non‐offender using Decision Trees: The Case of Colombia

“…There are a range of personality theories that contribute to explaining the offender's personality (Jackson & Bekerian, ; Pinizzotto & Finkel, ) and many instruments for exploring the criminal's personality profile (Archer, Buffington‐Vollum, Stredny, & Handel, ; K. M. Davis & Archer, ; Kline, ; Mullen & Edens, ; Nikolova, Hendry, Douglas, Edens, & Lilienfeld, ; Rabin, Borgos, & Saykin, ; Wood, Nezworski, Lilienfeld, & Garb, ). A variety of statistical techniques are also available for analysing these personality tests (Bennell, Goodwill, & Chinneck, ; Goodwill, Allen, & Kolarevic, ; Goodwill et al ., ; Homant & Kennedy, ; Neuman & Wiegand, ). However, criminal personality profiling has occurred largely in the absence of a well‐defined profiling framework and an empirical knowledge base (Becker, ; Snook, Cullen, Bennell, Taylor, & Gendreau, ; Snook, Eastwood, Gendreau, Goggin, & Cullen, ).…”

Crime linkage of sex offences in Japan by multiple correspondence analysis

“…Although there have been a few studies using MCA for classifying incidents into several behavioral themes (Goodwill, Allen, & Kolarevic, 2014) or into criminal's hunting process scripts (Beauregard, Proulx, Rossmo, Leclerc, & Allaire, 2007), the authors could not find any peer-reviewed articles that examined MCA in the field of behavioral crime linkage, excepting Santtila, Fritzon, and Tamelander (2004). Santtila et al (2004) analyzed the data containing 45 dichotomous (categorical) variables by a principal component analysis (PCA) to link arson cases.…”