Another look at across-crime similarity coefficients for use in behavioural linkage analysis: an attempt to replicate Woodhams, Grant, and Price (2007)

Abstract

In the absence of physical evidence, investigators must often rely on offence behaviours when determining whether several crimes are linked to a common offender. A variety of factors can potentially influence the degree to which accurate linking is possible, including the similarity coefficient used to assess across-crime similarity. The current study examines the performance of two similarity coefficients that have recently been compared to one another, Jaccard's coefficient (J) and the taxonomic similarity index (D s ), using samples of two crime types, serial homicide (N 0237) and serial burglary (N 0210). In contrast to previous research, the results indicate that D s does not significantly outperform J with respect to linking accuracy. In addition, both coefficients lead to higher levels of linking accuracy in cases of serial homicide compared to serial burglary. Potential explanations for these findings are presented and their implications are discussed.

The impact of data degradation and sample size on the performance of two similarity coefficients used in behavioural linkage analysis

“…For example, as Melnyk et al [14] argue, J only accounts for across-crime similarity at the most discrete behavioural level and, therefore, it is very sensitive to even slight variations in behaviours exhibited across crimes, as well as variations in the way that crime scene information is reported and recorded by police officers. Issues such as these have led researchers to recommend that other similarity coefficients be examined to determine if they might be more suitable for BLA than J [12,27,28].…”

“…As discussed by Melnyk et al [14], part of the appeal of J is undoubtedly its simplicity. For a pair of crimes, A and B, J is:…”

“…In other words, if a specific behaviour is absent from two crimes, J will not increase. To some, it is thought that ignoring joint non-occurrences in this way makes sense ''because the absence of a behaviour in any given crime, or crime report, may be due to factors other than its actual non-occurrence'' [14]. For example, a victim may not have witnessed a behaviour or remembered its occurrence, or the behaviour may not have been accurately reported to or recorded by the police [29].…”

“…Primarily for the purpose of determining the extent to which it is possible to link crimes, the degree of behavioural stability and distinctiveness exhibited by serial offenders has been examined in cases of sexual crime [1,11,12], homicide [13][14][15], arson [16], burglary [14,[17][18][19][20], vehicle theft [21,22], and robbery [23]. Like the research conducted by personality psychologists, the results of these studies indicate that offenders do exhibit behavioural distinctiveness in a somewhat stable fashion across their crimes, but that situational factors (e.g., victim resistance) also have a significant influence on offender behaviour, thus limiting the degree to which BLA will be possible [2].…”

Crime concentration theory

“…Two of the more prominent measures are Jaccard's similarity coefficient and the taxonomic similarity index. Thus behavioural linkage analysis has been used to determine whether, in the absence of physical evidence, unsolved crimes were committed by the same offender, based on across-crime similarity coefficients (Melnyk et al 2010;Bouhana et al 2014).…”