Search

Intelligence-led policing and the introduction of the National Intelligence Model have emphasised the need to gather information around the crimes and characteristics of prolific offenders. The police and crime analysts have been tasked to find more reliable, efficient ways to carry out methods such as suspect prioritisation („offender profiling‟) and case linkage. Three tenets underpinning such tasks are homology (Alison, Bennell, Mokros, & Ormerod, 2002), offender consistency (for example, Canter, 1995) and inter-offender variation (for example, Goodwill & Alison, 2007). Research considering homology has drawn varied conclusions, with some studies providing support for this (for example, Canter & Fritzon, 1998), and others finding that, offenders who behave in similar ways within their crimes, do not readily share similar characteristics (for example, Mokros & Alison, 2002). Research investigating the consistency hypothesis, however, has been more promising, with support for the consistent exhibition of particular behaviours over crime series (for example, Bennell & Canter, 2002). There has been a consensus that elements of the offenders‟ spatial behaviour, notably inter-crime proximity, is consistent across crimes and provides particular accuracy at predicting whether two crimes are linked (for example, Markson, Woodhams, & Bond, 2010). Ideas drawn from environmental criminology theories, such as Routine Activity Theory (Cohen & Felson, 1979) help to explain why offenders usually commit offences in locations that are close together. Moreover, ideas from personality theory highlight the particular consistency of behaviours that are within the offenders‟ control (Funder & Colvin, 1991) which may explain why spatial behaviours are especially useful (Bennell & Jones, 2005). In addition, researchers have called for a more detailed examination of the impact context (such as the location of the offence) has on the exhibition of behaviours within crimes (Mokros & Alison, 2002). A recently developed model highlights the dynamic nature of the rape event, and how geographic and offence behaviours may interact (Beauregard, Proulx, Rossmo, & Leclerc, 2007). However, this model is limited in its application to the investigation of rapes perpetrated by an unknown offender, it uses analysis techniques that may not fully explore the qualitative inter-relationships between geographical and behavioural variables and it is based on a model (Rossmo, 1997) that has not been empirically tested (van der Kemp & van Koppen, 2001). Thus, the aim of this thesis was to develop a model of the spatial mobility of stranger rapists within their offences and examine how this behaviour is related to offence behaviour. Moreover, the ability of the model to predict offender background characteristics („offender profiling‟) and link crimes together is considered. The records of 112 detected stranger rape cases occurring between 2004 and 2006, reported to the Metropolitan Police Service were used as the main data set for this thesis. Thematic analysis resulted in four Geo-mobility styles being established; Intruded, Ambushed, Abducted and Followed. These were found to be related to three behavioural themes; Intruded rapes were related to those which reflected a broad Criminal style, Ambushed and Followed rapes were associated to those which reflected a broad Violent style, whilst Abducted rapes were associated to those which reflected a broad Sexual style. The Geo-mobility styles were not useful in predicting offender background characteristics, but neither were other spatial or offence behaviours. The Geo-mobility styles were also not exhibited consistently over a linked series (a sub-sample of the original data set). However, inter-initial approach and inter-attack proximity were found to be both consistent and accurate at predicting case linkage. Findings are discussed in terms of theories of rape behaviour, theories of spatial behaviour and ideas about homology, consistency and inter-offender variation. The implications for offender profiling and case linkage are considered, as well as the limitations of the present study and future research ideas.