A key aspect of multi-factorial theories is that they have been developed from analyses of empirical studies of sex offenders, their personal histories and their offence patterns. Multi-factorial explanations incorporate physiological, psychological and emotional dimensions relating to the offender, along with situational and historical (fixed) matters. Apart from these issues, the models are also mindful of time and change; thus, change in physiological, psychological and emotional states is important. Key contributions to this area of theory include: the confluence model of sexual aggression (Malamuth et al, 1996); integrated theory (Marshall and Barbaree, 1990); the quadripartite model of sexual aggression (Hall and Hirschman, 1991); the four preconditions model (Finkelhor, 1984); the pathways model (Ward and Siegert, 2002); the integrated theory of sexual offending (Ward and Beech, 2006), and a ‘new integrated theory’ of ‘child sexual abuse’ (Smallbone et al, 2008). Summarising the content of (multi-factorial) theories, Hanson and Morton-Bourgon (2005, pp 1154–5) comment:
Contemporary theories posit … that, apart from sexual deviancy and lifestyle instability, there may be three additional characteristics of persistent sexual offenders: (a) negative family background, (b) problems forming affectionate bonds with friends and lovers, and (c) attitudes tolerant of sexual assault.
However, while Hanson and Morton-Bourgon (2005) acknowledge the value of multi-factorial explanations of sex offending, they note that the evidence behind the identification of the factors and how they interrelate with reoffending is weak. They suggest that if the primary therapeutic objective of correctional work with sex offenders is to reduce recidivism (reconviction), then detailed attention must focus on what, statistically, can be shown to be directly related to reoffending. A statistical means of analysing large amounts of quantitative data from many studies is meta-analysis, and it is by using meta-analysis that issues directly related to sex offending have been identified. Table 2.2 outlines the key features of meta-analysis.
Table 2.2: What is meta-analysis?
Although meta-analysis is an efficient and effective way of collating and analysing large amounts of data, and it is clearly very influential, the analysis is susceptible to bias. A good meta-analysis carefully outlines the detail of the data included and excluded, and highlights the limitations of the study.
Karl Hanson and colleagues (eg Hanson and Bussiere, 1998; Harris and Hanson, 2004; Hanson and Morton-Bourgon, 2005) conducted a series of meta-analyses relating to sex offender recidivism. The early study (Hanson and Bussiere, 1998) was of major significance in highlighting the low recidivism rate of sex offenders. Moreover, meta-analyses raise questions about some elements common in treatment programmes that appeared to be unrelated to reoffending (the index offence, the nature of that offence, the denial of responsibility and the motivation for treatment). Hanson and Morton-Bourgon (2005, p 1154) suggest that the key factors linked to recidivism are ‘deviant sexual preferences and anti-social orientation’, while many well-established treatment targets – such as ‘psychological distress, denial of sex crime, victim empathy, stated motivation for treatment’ – ‘had little or no relationship with sexual or violent recidivism’.
The importance of developing a subtle (multi-factorial) understanding of sex offenders is not only to develop treatment programmes, but also to more accurately assess risk. There is an immense psychological literature on risk assessment and risk management in relation to sex offenders (for reviews of this literature, see Beech and Ward, 2004; Bengtson and Långström, 2007; see also Chapter Five). The literature distinguishes between actuarial and clinical approaches to assessing risk (Grubin, 1999). Actuarial approaches use risk factors that have been consistently identified in a wide range of research studies of convicted sex offenders. Typically, these factors are previous offences, relationship history and criminality (Beech and Ward, 2004, p 32), and are described as being static (ie they are not amenable to change). Clinical approaches rely on the judgement of the clinician dealing with the individual person who may pose a risk to others. The clinician’s assessment may or may not be based on research literature, but generally includes consideration of the dynamic factors (eg mood, attitudes, physical circumstances – including the availability of victims) affecting the individual under assessment. The actuarial tendency in assessing sex offender risk is strong (Beech and Ward, 2004, p 33) and the research literature continues to indicate that a pure actuarial approach is more accurate in predicting reoffending (Bengtson and Långström, 2007). However, Beech and Ward (2004, p 32) summarise a range of weaknesses with a pure actuarial approach (including its inability to deal with the unusual and contingent dynamic factors, and its dependence on official recidivism data that may underestimate recidivism rates and therefore levels of risk). Recently, Ward and Beech (2015) have critiqued the concept of ‘dynamic risk factors’ as being inadequately theorised and potentially leading clinical practice and research into a cul-de-sac. They argue that the concept needs clearer definition, and that descriptive and explanatory components need to be clearly distinguished. They propose a way forward of developing ‘exemplars’ of sex offending that first describe the phenomenon, and then seek to explain it. This work is in the early stage of development but looks promising as a way of further clarifying psychological approaches to assessing risk.
The third area of theory highlighted by Ward and Hudson (1998) concerns ‘micro-’issues. Two particular areas are given as examples of micro-issues: relapse prevention (RP) and ‘grooming’. Much has been written over many years about relapse prevention (eg Pithers, 1990; Ward, 2000; see also p 131 below). The notion of ‘relapse’ was derived and developed from addiction theory (Marlatt, 1985), which identified and named the stages of ‘relapse’. In relation to sex offending, the model was slightly adapted, but it was very popular from the mid-1980s through to the mid-2000s. RP modules featured in both community and prison programmes. The model has come in for serious criticism from Ward (2000), who proposed an alternative ‘self-regulation’ model of relapse. Understanding the minutiae of reoffending remains an important research target and treatment component, although the pure RP model is increasingly considered to be unsuitable for use with sex offenders (Laws, 2003; Yates et al, 2010). More recently, consideration of the ‘pathways’ to offending model has led to the development of more sophisticated conceptualising of the processes leading to offending behaviour (see pp 132–3 below). Kingston and Yates (2012) and Ward and colleagues (2006), among others, have developed the ‘self-regulation’ model of sexual offending, within which are identified four offence pathways, all of which have implications for intervention(s). The model is conceptually linked with the Good Lives therapeutic model, which is discussed later in the book.
The word ‘grooming’, referring to how a sex offender prepares both potential victims and the environment where s/he intends to offend, has entered the popular vocabulary through media reportage, and (in the UK) through the Sex Offenders Act 2003, which created a specific offence of ‘Grooming’. Olson et al (2007, p 241) describe grooming as ‘the subtle communication strategies that child sexual abusers use to prepare their potential victims to accept sexual contact’. This involves both controlling the victim by concealing the threat posed, and manipulating the environment to ensure that the offence will occur without disturbance. Campbell (2009) presents a clear account of the processes of grooming. She highlights that grooming involves ‘strategies to present a specific image in an interpersonal relationship through the use of language that indicates the sender’s “persona” to the receiver’ (Campbell, 2009, p 432). Campbell (2009, p 434) identifies five particular grooming strategies employed by offenders: supplication (presenting