This study sought to clarify the nature of the relationship between conduct disorder (CD), early-onset alcohol abuse (EOAA), some other externalizing-related constructs and adult violent antisociality (VA). It addressed two key questions: (i) whether EOAA mediated the link between CD and VA; and (ii) whether the effects of EOAA on VA were, in turn, mediated by impulsiveness, ventro-medial prefrontal cortex (vm-PFC) dysfunction and social deviance as measured by the Psychopathy Checklist-Revised (PCL-R). It tested the hypothesis that in the context of early disinhibitory psychopathology, e.g. CD, EOAA disrupts the neural substrates of self-regulation in vm-PFC during a critical neurodevelopmental period (i.e. before age 20). Consequently, on entry into adulthood the vm-PFC is functionally impaired and personality suffers maladaptive development which would then take the form of increased impulsiveness and social deviance, placing the individual at high risk of violent antisocial behaviour. Using a cross sectional design, DSM-IV Axis I and II disorders, psychopathy, impulsiveness, vm-PFC functioning, history of drug and alcohol use, and both amount and severity of violence were assessed in 100 patients with personality disorders detained in secure hospital settings. Patients identified as having a history of EOAA, compared with those with no alcohol abuse history, were more impulsive, scored higher on the social deviance factor of psychopathy (PCL-R F2), were more conduct disordered, and showed a higher level of VA. Regression analysis showed that CD, EOAA, impulsiveness and PCL-R F2 significantly predicted VA, although PCL-R F2 rendered the effects of CD insignificant when used conjointly in regression analysis. A multiple mediation model explaining about 20% of the variance in VA showed that EOAA partially mediated the effects of CD on VA, after controlling for age, cannabis misuse and ADHD. A separate multiple mediation model explaining 50% of the variance in VA showed that PCL-R F2 and impulsiveness partially mediated the effect of EOAA on VA. However, contrary to the prediction arising from the hypothesis, the effects of vm-PFC functioning on VA were insignificant. Although the study suffered from some limitations, results suggest that both impulsiveness and social deviance contribute importantly to a pathway leading from CD through adolescent alcohol abuse to maladaptive personality development and adult VA.

Background: The research project aims to explore the mental representations therapists have about their patients and therapy between sessions (Intersession Experience, ISE) through exploratory mixed method sequential design. The study aims to gain an understanding of the way UK therapists make sense of their ISE; review the themes within an existing measure of ISE; explore the associations between aspects of therapist ISE and alliance for a series of UK therapists; and gain explanatory qualitative feedback to contextualise quantitative findings. Method: The research project employed a three stage mixed method design. Phase one comprised exploratory qualitative interviews of UK therapists about their ISE, analysed using a semantic level Thematic Analysis (Braun & Clarke, 2006). Phase two comprised a series of single case time series analysis designs where therapists completed a measure of their ISE and alliance weekly over the course of therapy. Phase three comprised explanatory deductive qualitative interviews about phase two findings. Results: ISE may be instrumental for therapy; ISE can be spontaneous or planned, may affect in-session processes and may impact on therapist wellbeing. Intersession thoughts were associated with problem solving, re-experiencing aspects of therapy and internal reflection. Positive emotional ISE preceded increases in alliance ratings. Negative emotional ISE and alliance were bi-directionally negatively associated within the sample. Explanatory phase three themes suggest ISE as an indicator of need for intervention (E.g. supervision), transference of patient difficulties and to help therapy progress. Conclusion: Therapist ISE may be conceptualised as ‘unfinished business’ and appears to relate to emotional processing of in-session content in the form of right brain activation (Kahneman’s system 1) or instrumental problem solving and reflection in the form of left brain processing (Kahneman’s system 2). ISE may hold implications for the therapeutic relationship, in-session processes and therapist wellbeing. ISE appears to be an important process variable linked to the therapeutic alliance and warrants further exploration in the form of patient/therapist dyads, exploration of ISE across therapy phases and the association to patient outcomes and therapist wellbeing.

Overview This thesis reviewed forensic aspects of Intellectual Disabilities (ID) and Autism Spectrum Disorder (ASD). Chapter two was a case study where an individual with ID and ASD who exhibited forensic/Challenging Behaviour (CB) was assessed and intervention offered. Chapter three then focussed on the assessment of people with ID and ASD by critiquing the Wechsler Adult Intelligence Scale (WAIS), version 3 and 4 (WAIS-III & WAIS-IV) (Wechsler, 1997; Wechsler, 2008a, 2008b, 2008c) and discussed application issues for people with ID and ASD. Chapter 4 and 5 examined intervention in more depth. Chapter 4 applied Weiner’s (1985, 1993) theory of achievement, motivation and emotion as a way of understanding contextual intervention and Chapter 5 examined the effect of psychotherapeutic interventions on CB. Chapter Two The case study discussed in chapter two examined explanations of violence and diagnoses. ASD and an ID were identified as useful neurodevelopmental diagnoses that could be used as a framework to understand the individual’s needs. Some explanations of violence were related to features of these diagnoses. Social factors were also identified as increasing risk. Violence was modelled by a key male figure during childhood development which appeared to have been retained as a template of acceptable behaviour, supporting the work of Bandura on social learning (1973). This, and the experience of being a victim of this violence, also supported the work of Douglas, Hart, Webster and Belfrage (2013) who indicated that social factors can increase the risk of violence. In turn this affected attachment supporting the work of Bowlby (1980). The behavioural approach (Cipani & Schock, 2010) was also applied to identify functions of violence. A psychotherapeutic intervention was offered to improve one aspect of emotional recognition, face perception, with the aim of enabling him to comprehend visual feedback from the environment, make interpretations about the emotions experienced by others, and modify his behaviour accordingly. There was little difference to his level of skills following the intervention. It was concluded that particular skills development may be difficult for someone with a neurodevelopmental difficulty if the biological structures are not there to support that skill, or skills may take a long time to develop. In this case it was concluded that the best intervention was to consider a placement for him in an appropriate environment that would cater for his ASD and ID related needs. Comments were made about the quality of assessments and for improvements to be considered to formulation and intervention. Chapter Three Chapter three was a critique of the WAIS, a tool widely used as part of the process to identify whether someone has an ID. This may determine if an individual may be unfit to plead in court or to determine the services appropriate for that individual’s needs. The review focussed on the most recent version, WAIS-IV (Wechsler, 2008a, 2008b, 2008c) but also referred to other versions in particular the WAIS-III (Wechsler, 1997). The conclusion was there were some flaws to the assessments that required further research and also some omissions in the epistemology. The critique examined different aspects of validity and reliability. The examination of convergent validity yielded mixed results with some positive correlations with other tests and some weak correlations. This difference could be caused by some tests measuring different aspects of ‘g’. The consequences of differences can be significant, for example the scores of the Wechsler Intelligence Scale for Children, version 4 (WISC-IV) (Wechsler, 2003) and the WAIS-IV scores could differ by up to 10 points and could mean individuals moving from a children’s to an adult service could become ineligible for the same levels of support. The results for test-retest reliability were good in the short term but more research was recommended to determine if the latest version of the WAIS would be reliable in the long term. The ‘Flynn effect’ (Flynn, 2007) indicated that levels of intelligence were rising so the WAIS-IV compensated for this and currently measures intelligence at around 3 points lower than the WAIS-III. The stability of this effect was unclear. The results for internal reliability were good although more research independent of the WAIS was recommended. The application of the WAIS for individuals with an ASD and ID was considered with the WAIS being found to be a limited tool for discriminating between categories of ASD, although what can be expected is an uneven profile. It was also unclear how useful the WAIS was to discriminate between individuals with a substance misuse/abuse disorder and normal controls. There were also difficulties with the application of the WAIS to individuals with ID. Difficulties measuring intelligence at lower levels were identified as scores can be unstable. Research was also recommended to provide a comprehensive view of the effect upon results of variables such as nationality, gender and age. Chapter Four Chapter four was an empirical research project investigating Weiner’s (1985, 1993) theory of achievement, motivation and emotion as a means of understanding the responses of unqualified nursing staff to clients with ASD exhibiting CB. One hundred and twenty seven unqualified nursing staff responded to an Attributional Style Questionnaire (ASQ) (Dagnan & Cairns, 2004) and questions based on a vignette. The quantitative results found weak support for Weiner’s model. There was weak evidence for the effect of demographic variables and the type of client (ASD or ID) and the type of service (ASD or Mental Health/ID) did not affect the results. Qualitative data was themed and coded according to Braun and Clarke (2006) and Boyatzis (1998) and provided an insight into the psychological processes influencing staff responses to CB. The results suggested that intervention was influenced by staff, patient and contextual variables. A recommendation was made for a contemporary exploration of the topic and to improve the validity and reliability of the ASQ. Chapter Five Chapter five was a systematic review of psychotherapeutic interventions aimed at reducing the CB exhibited by people with ID only due to a paucity of research using people with ASD as participants. A systematic literature search was conducted using 5 databases, a search of relevant journals and contact made with experts. From this 8 papers met the inclusion/exclusion criteria and these were analysed qualitatively. Three types of psychotherapeutic intervention were used within the ten papers, Cognitive Behavioural Therapy (CBT), “Soles of the Feet” meditation and Solution-Focussed Brief Therapy. All types of psychotherapeutic intervention reduced CB although the results of studies using CBT were mixed. There were shortcomings to these conclusions; these were methodological issues such as a need to ensure the consistency of outcome measures and the types of CB used to measure change. The effect of intervention may have been modified by variables such as length of treatment, the context and intrapersonal variables. Conclusions The chapters of this research thesis indicated that neurodevelopmental factors influenced forensic/CB, although social role models, attachment and trauma were also significant. Assessments were important as they may affect fitness to plead and services/treatment offered. Assessments such as the WAIS are supported by research although further research has been recommended. The use of the WAIS with people with ASD and ID was limited as people with ASD appear to exhibit uneven profiles when completing the WAIS and there were inaccuracies when using the WAIS with people with lower levels of intelligence. The difficulties people with ASD have with social communication skills can also affect performance and support is needed when administering assessments. Intervention for people with a neurodevelopmental difficulty can also be successful. Consideration is needed regarding appropriate intervention for each individual, how long before the intervention is effective, the context including the type of support offered and how that affects outcomes, intrapersonal variables such as intelligence, and whether the treatment includes one to one support. More research of an improved quality is required to determine the effectiveness of both assessment and treatment including contextual based treatment such as the interaction between staff and patients within an institutional setting. Considering these further would improve the service for this client group and contribute to a reduction of risk.

Valid and reliable measures of Acceptance and Commitment Therapy's (ACT) purported mechanism(s) of change are paramount in establishing its incremental validity above other extant psychological therapies. However, the most widely used general measure of psychological flexibility and the six ACT processes of which it is composed of, the Acceptance and Action Questionnaire (AAQ-II), has been significantly criticised, principally in terms of being a conflated distress measure due to having items with poor face and content validity. Theoretically, there has also been debate about whether ACT's 'hexaflex' model, which conceptually underpins the AAQ-II and other ACT process measures, is comprised of fewer than six distinct processes as per current ACT theory. The overall aim of our study was therefore to: 1) develop a new general measure of ACT processes; 2) test the measure's validity and reliability; and 3) establish whether the measure's factor structure accords with Hayes et al.'s (2011) clustering of ACT's six processes into three dyadic processes. The measure was named the Comprehensive assessment of ACT processes ('CompACT'). Using a Delphi-consensus methodology, ACT experts (N= 10) rated the face and content validity of 106 items. Items were taken from existing ACT process measures and two mindfulness measures, as well as generated by ourselves and the ACT experts within our study. Items which failed to reach pre-defined ratings thresholds were removed and remaining items were re-rated until a pre-defined consensus threshold was reached. After three rounds, an initial 37-item version of the CompACT was produced. The CompACT was then disseminated to a non-clinical sample of participants (N = 377) mainly via social media. Participants also completed other measures to assess different aspects of the CompACT's validity. A three-factor structure was found to be the most stable for a 30-itemed version of the CompACT. An examination of the distribution and contents of item loadings also confirmed that the CompACT's three factors could be defined in accordance with Hayes et al's. (2011) three dyadic ACT processes. The CompACT performed in theoretically predicted ways against measures of other related and distinct constructs, and was mostly comparable to the AAQ-II in terms of its magnitude of associations with these other measures, with the exception of a distress measure. The CompACT's associations with these measures also supported ACT's central hypothesis: higher levels of psychological inflexibility are associated with higher levels of distress and lower levels of health and wellbeing. Overall, the CompACT shows initial promise as a comprehensive measure of ACT processes, and more generally as a measure of psychological flexibility. The psychometrically viability of Hayes et al.'s (2011) three dyadic ACT processes as a conceptual model for the CompACT further strengthens previous suggestions that ACT's clinical model may be composed of fewer than six distinct processes. The CompACT was adjudged by ACT experts to possess items with better face and content validity than the AAQ-II. However, in order to be considered as a viable alternative measure to the AAQ-II, further testing of the CompACT's psychometric properties is needed.

Introduction: The concept of pornography addiction, although widely accepted, continues to be debated, and as such no diagnostic criteria exist. Opponents and proponents of pornography continue to cite evidence that supports the claim that pornography is either harmful or beneficial, respectively. However, a review of the literature pertaining to pornography use revealed conceptual and methodological shortcomings which limit the conclusions made in the existing literature base. Without adequate research to further our understanding of the complex relationship individuals have with pornography, we risk pathologising or condoning behaviours that are legal and consensual, which may cause distress for some individuals, or even divert attention from deeper issues, such as an individual’s thinking styles and feelings of shame. Researchers acknowledge that people’s relationship with pornography is complex (Hardy, 1998) and individuals experience it in different ways, for example the time spent viewing pornography, the environment in which they view it, who they view it with, and the genre of pornography they watch (Attwood, 2005; Hald & Malamuth, 2008; Malamuth, Addison, & Koss, 2000; Poulsen, Busby, & Galovan, 2013; Reid, Li, Gilliland, Stein, & Fong, 2011). Given that additional variables may be involved when people pathologise their pornography use, particularly rigidity of thinking (Reid et al., 2009), it is worth investigating such variables further to help discern differences between those who pathologise their pornography use and those who do not. Aims: The overall aim of this study was to explore whether thinking styles impact on the meaning given to pornography use. The primary objective was to compare participants who perceived their pornography use to be problematic (porn addict) with individuals who did not (non-addict) on the following dependent variables: thinking styles, shame, levels of desirability responding, levels of religiosity, and perceived effects of pornography use. As this was an exploratory approach a two-tailed hypothesis was employed. The secondary objective was to capture in-depth qualitative experiences participants had with pornography, and pornography addiction. Design: This study used a mixed-method sequential explanatory design (MMSE), using questionnaires to collect quantitative data and interviews to collect qualitative data. Method: Participants (n=265) were recruited from both UK National Health Service (NHS) and non-NHS sites. For non-NHS participants, the questionnaires were made available online and advertised through social media. NHS recruitment occurred at a specialist sexual health clinic and was advertised using posters displayed in relevant clinical areas. The study predominantly employed a maximum-variation sampling technique. This is a purposive sampling technique used for this research to ensure that an eclectic sample of diverse demographics was captured. Quantitative data was collected using a demographic questionnaire and four validated measures; the Inventory of Cognitive Distortions (Yurica & DiTomasso, 2001), the Test of Self-Conscious Affect-3 (Tangney, Dearing, Wagner, & Gramzow, 2000), the Balanced Inventory of Desirable Responding (Paulhus, 1991; 1998), and the Pornography Consumption Effects Scale (Hald & Malamuth, 2008). All interviews were conducted either through the audio function on Skype© or over the telephone. Results: Participants reported themselves as belonging to one of three groups; addicts, somewhat addicts, or non-addicts. MANOVA analyses revealed that groups significantly differed in their propensity for cognitive distortions, reported effects of their pornography use, the impact of their religious beliefs, and time spent viewing pornography. Significant differences were not found for the shame scales, or for social desirability. Multinomial logistical regression revealed that negative impact of pornography on participants’ lives in general, on their sex life, dysfunctional thinking styles (overall and externalisation of self-worth, magnification and fortune telling, minimisation and arbitrary inferences, and perfectionism), and impact of religious beliefs significantly predicted group membership. Furthermore, regression analysis supported the hypothesis that thinking styles mediated the relationship between time spent viewing pornography and the overall negative impact of pornography. Qualitative results supported these findings, and thinking styles were revealed to influence the discourses participants had regarding pornography. Primary themes identified were participants’ relationship with pornography and perceived causation of pornography addiction, significance of social norms, and impact of expert opinion. Furthermore, although absent in the quantitative findings, the concept of shame was raised as an influential factor in the pathologising of pornography use, thus supporting the notion that a conflict in values, paired with a relatively inflexible cognitive style, may lead to pathologising, and shame will be a product of that process. Discussion: This study demonstrates the role thinking styles play in how people evaluate their pornography use. Thinking styles both predict whether an individual perceives their pornography use to be problematic or not, and are evident in the discourses people use when discussing their pornography use and the concept of pornography addiction. Specifically, individuals with a propensity for rigid thinking styles are more likely to negatively evaluate their pornography use. Furthermore, the similarities and differences apparent in group comparisons can be understood within a values theoretical framework; it may be that individuals with rigid thinking styles are more likely to endorse particular values that are incongruent with their pornography use behaviours. Conversely, individuals with more flexible thinking styles may be more likely to endorse values which are not incongruent with their pornography use behaviours. This is important from a research and treatment perspective, as it may not be the behaviour itself (pornography use) that is problematic and the target for intervention, but the cognitive framework individuals use in relation to the behaviour. Current treatment offered to those who self-report as pornography addicts often ignore the role of thinking styles and values. In light of this study’s findings, thinking styles should be a focus in future research and treatment as it may help to reduce cognitive dissonance and engender agency.

The aim of the current thesis was to explore sexual knowledge and risk in the assessment and treatment of adolescent males with intellectual developmental disorders (IDD), who display harmful sexual behaviour (HSB). A variety of methods were used in this undertaking, including a systematic review, a single case study, a critique of a psychometric measure and two empirical studies. The findings of the systematic review indicated that although the identified/assessed risk variables related to recidivism were found to be similar in cases and controls, some were found to be more prevalent for sexual offenders with IDD. However, it was difficult to draw any firm conclusions due to inconsistencies in findings and limitations in the methodological quality of included studies. The case study detailed the assessment and treatment of an adolescent male with limited cognitive ability and a history of harmful sexual behaviour. The assessment highlighted the utility of the integrated theories of sexual offending, and how his adverse early life experiences and developmental trauma disorder may have negatively impacted on his attachments, social and self-regulation skills, low-self-esteem and the development of his harmful sexual behaviour. The post-intervention assessment highlighted the client had made some positive shifts within his identified treatment targets, although there was still room for improvement with regards to his level of socio-sexual knowledge, his ability to self-regulate and his tendency to engage in aggressive and rule-breaking behaviours. In the critique of the Knowledge Test and Quick Quiz components of the Assessment of Sexual Knowledge –ASK (Butler, Leighton & Galea, 2003) the Knowledge Test measure demonstrated some good psychometric properties and had undergone extensive research and robust testing with individuals with IDD during its development. However, weaknesses of the measure considered were its lack of normative data, the ambiguous wording of some items, and limited empirical research regarding the tool’s effectiveness in measuring sexual knowledge across different populations of individuals with IDD. The empirical research consisted of two studies. In the first study a questionnaire adapted from the Knowledge Test of the Assessment of Sexual Knowledge – ASK, was tested for its ability to accurately measure sexual knowledge in adolescent males with and without IDD and their counterparts who display harmful sexual behaviour. Tests examining both the reliability and validity of the adapted measure suggested its psychometric properties were promising. The second study sought to explore whether there were differences in sexual knowledge in adolescents with and without IDD and their counterparts who display harmful sexual behaviour (HSB). Significant differences were observed between groups. In the HSB groups, adolescents with IDD experienced significantly higher rates of sexual victimisation than No IDD adolescents. For those with IDD the HSB group were 12 times more likely to view sexually explicit material than the Non HSB group. The findings suggested that a lack of sexual knowledge might not be a contributory factor in why some young people with IDD go on to display harmful sexual behaviour.

This thesis aims to provide a broad overview of contemporary challenges in relation to sexual offending, with specific focus being placed on the need to balance the public’s perceptions of those convicted of sexual offences with the reality of their offence patterns, to determine what risk society really faces from these individuals. In doing this, it incorporates diverse methods, including a systematic review, an empirical research study, an individual case study and a critique of a psychometric measure. Following a general introduction in Chapter 1, Chapter 2 presents a systematic evaluation of 13 studies investigating public perceptions of sexual offenders, with specific focus on comparing public demographic variables as predictors of these attitudes. It was found that overall; the public’s perceptions were negative, with beliefs about high recidivism rates amongst other misperceptions. Demographic characteristics were shown to be inconsistent in predicting punitive judgments. Level of educational attainment was the only demographic variable investigated that was consistently associated, with higher levels of education resulting in more positive perceptions. In Chapter 3, the offence patterns of repeat sexual offenders are investigated, including an analysis of whether this population escalate, de-escalate or remain stable over time. The results indicate an overall pattern of stability, indicating that sexual recidivists commit the same category of offence from one offence to the next. However, escalation was also common, a finding that warrants further exploration. Chapter 4 explores the factors that contribute to the onset and escalation of the offence behaviours of a repeat sexual offender, through the use of psychological formulation. Results indicated that a different set of risk factors contributed to the onset of the client’s offending compared to those that contributed to escalation. Chapter 5 evaluates the Community Attitudes Toward Sex Offenders (CATSO) scale (Church, Wakeman, Miller, Clements, & Sun, 2008), an assessment measure used in Chapter 2. Finally, Chapter 6 provides a discussion and conclusion to the thesis, drawing together the implications of the research.

The role of violent video games (VVG) in increasing levels of aggression in players is a topic that has been debated for decades. While a simple cause and effect relationship is not present, it is apparent from the literature that VVGs do have a relationship with elevated levels of aggression within some players. The first chapter of this Thesis presents a systematic review of the literature focusing on additional factors included in the research to explain this finding. However, a lack of standardisation in both measurement and reporting of results making interpretation difficult. The second chapter provides a critique of the non-experimental methods used within video game (VG) research. The third chapter focuses on a non-experimental study which highlighted the differences in aggression, motivations for playing, empathy, and personality in a group of VVG players. The final chapter presents the results of a service evaluation of the management of VGs within a high secure psychiatric hospital, which highlighted the real-world implication of an unclear literature base. The conclusions of this thesis highlight the variation in the quality of the literature available and stresses the need to conduct research with targeted populations, both VG players and clinically populations, to fully understand the relationship between engaging in VVG and increased aggression.

Growing evidence suggests that the Salience Network, anchored on the bilateral anterior insula and dorsal anterior cingulate cortex plays a key role in the pathophysiology of schizophrenia. However, several questions regarding the exact nature of these abnormalities in schizophrenia remain unanswered. This thesis examines the neurobiological basis of Salience Network dysfunction in patients with schizophrenia. Specifically, in this work, we use multi-modal neuroimaging techniques to investigate abnormalities in the functioning, connectivity and neurochemistry of the Salience Network in schizophrenia. Chapter 1 introduces the concept of schizophrenia as a disorder of salience and the role of the Salience Network in this context. Chapter 2 describes participant recruitment, data collection and the neuroimaging techniques used in this work. Chapters 3 to 5 present the results from three research studies. In the first study, using functional Magnetic Resonance Imaging (fMRI) and Magnetoencephalography (MEG), we investigate regional brain activity in response to a task designed to engage the Salience Network. We report evidence demonstrating inefficient cerebral recruitment in patients with schizophrenia i.e. aberrant activity in task positive and task negative brain regions; along with reduced beta response to relevant stimuli in the Salience Network. These findings are reported in chapter 3. In chapter 4, we use high resolution 7T Magnetic Resonance Spectroscopy (MRS) and demonstrate that neurochemical abnormalities in schizophrenia vary between patients depending on their clinical status. Specifically, we show that glutathione and glutamate concentrations in the Anterior Cingulate Cortex are prominently reduced in patients with residual schizophrenia. We also provide evidence linking two popular neurochemical theories of schizophrenia – the NMDA receptor hypofunction hypothesis and the theory of oxidative stress. Chapter 5 explores the question of disrupted effective connectivity within the Salience Network in schizophrenia using resting-state fMRI. In this chapter, we report abnormalities in effective connectivity within the Salience Network in patients with schizophrenia. We also demonstrate that these aberrant causal interactions are related to the neurochemical abnormalities reported in chapter 4. Finally, in Chapter 6, we discuss the significance and translational potential of these findings and offer recommendations for future work. Overall, using a multimodal imaging study design and a wide range of measures, this thesis reports novel findings which further our understanding of Salience Network dysfunction in schizophrenia and provides a good foundation for future research studies in this area.

This thesis focussed on online sexual offending, particularly the access of indecent images of children (IIOC), and whether certain risk factors associated with contact sexual offending, namely offence-supportive cognitions and deviant sexual interest (DSI), were relevant for individuals who access IIOC. In references to offence-supportive cognitions for IIOC offenders, the issue is unclear. Research indicates that IIOC offenders do not hold these distortions, whereas the narratives from IIOC offenders indicates that distortions do exist for them. An examination of legal and illegal pornography, and its relationship to offence-supportive attitudes and sexual offence-related behaviour via a systematic review indicated that studies cited a relationship between pornography and sexual offence-related behaviours more often than a relationship with offence-supportive attitudes. However, the review found no statistically significant differences, with either outcome, regarding the number of studies citing a relationship compared to those that did not; so no consistent relationship with pornography was found. The systematic review also highlighted the difficultly in ascertaining the impact of IIOC on individuals due to the limited research comparing IIOC offenders to community controls. However, there did appear to be certain individuals for when pornography access (including IIOC) could be problematic. The case study highlighted the difference between global distortions and those specific to the offence in question, and how this might have relevance to IIOC offenders. Regarding the second risk factor of interest, DSI with IIOC offenders is not well understood, with research indicating these offenders to be more deviant than some contact sexual offenders, whilst the narratives of IIOC offenders suggest the opposite. Within the research project, DSI was found to be salient, with IIOC offenders differing from controls on multiple measures of DSI. DSI was also highlighted within the case study, regarding its role within the subsequent crossover offending of an IIOC offender; the Internet and accessing IIOC appeared influential in the development of the client’s sexual interest in children, and in turn, their contact offending. Discussions viewed the importance of measuring DSI within risk assessment, and its inclusion within interventions with IIOC offenders. Consideration was giving to the prevention of IIOC access online to prevent the development of offence-supportive beliefs and behaviour, and to increase the protection of children.