Previous research concerning violent offending has been fragmented considering different elements of violent offending separately. The aim of this thesis is to consolidate the different areas of previous research into one cohesive model of assault offenders and offences. This model will consider the developmental, cognitive, behavioural and environmental constituents of assault offenders and offences in a temporal framework. Interviews were conducted with 35 male and 13 female offenders who had a conviction for assault. Grounded Theory analysis was used to categorise the data and construct a model of assault including developmental factors, the time preceding the offence, the offence and the time after the offence producing the Pathways Model of Assault (PMA). Initial construction of the PMA was conducted using 25 of the male participant interview transcripts. The PMA was then validated across gender through inclusion of the 13 female participant transcripts. The PMA was also subject to an inter-rater reliability test, which provided high consistency between the coding of two researchers using the final 10 male participant transcripts and 10 randomly chosen female participant transcripts. The PMA consisted of 10 stages where the individual differences of the participants could be mapped, thus providing “pathways” through the model. Five major pathways were found. Further exploration of the PMA through quantitative analyses provided validation of four of the pathways, with significant associations found between two of the offender types and two of the offence types. The offender types reported were under-controlled, representing persistent repeat violent offenders and over-controlled, representing onetime violent offenders. (For complete abstract open document)

abstract: The link between victimization and offending is well established in the literature, yet an unexplored causal pathway within this relationship is concerned with why some individuals engage in maladaptive coping in response to victimization. In particular, those with low self-control may be attracted to problematic yet immediately gratifying forms of coping post-victimization (e.g., substance use), which may increase their likelihood of violent offending in the future. Using three waves of adolescent panel data from the Gang Resistance Education and Training (GREAT) program, this research examines: (1) whether individuals with low-self control are more likely to engage in substance use coping following violent victimization, and (2) whether victims with low self-control who engage in substance use coping are more likely to commit violent offenses in the future. The results from negative binomial regressions support these hypotheses, even after controlling for prior offending, peer influences, prior substance abuse, and other forms of offending. The implications for integrating general strain and self-control theories, as well as for our understanding of the victimization-offending overlap, are discussed. / Dissertation/Thesis / M.S. Criminology and Criminal Justice 2011

The relationship between interpartner abuse (IPA) and criminal offending has received little scholarly attention, despite its important theoretical and practical implications. Two key questions about this relationship require attention. First, to what extent do IPA, violent offending, and property offending represent empirically distinct behavioural domains? Second, to what extent do these offence types share common predictors? The current study addressed these issues, and several additional issues, in a birth cohort of 950 New Zealand adults. Cohort members were questioned at ages 21, 25, and 30 years about the extent which they had engaged in IPA and criminal offending during the previous year. Information was also obtained from birth to late adolescence on a number of potential predictors of IPA and criminal offending, including socio-economic disadvantage, family dysfunction, childhood abuse, conduct disordered behaviours, deviant peer affiliations, substance abuse, academic ability, the obtainment of a high-school qualification, identification with an ethnic or racial minority, and gender. Confirmatory Factor Analysis results indicated that IPA, violent offending, and property offending represent three empirically distinct, albeit related, behavioural domains. Consistent with this finding were those obtained using Structural Equation Modelling techniques, which indicated that these offence types share many common childhood, adolescent, and demographic predictors. In addition, many predictors, but not all, were found to exert similar effects across these offence types. Analyses also indicated that shared predictors accounted for considerable proportions of the relationships between IPA, violent offending, and property offending. Finally, the vast majority of predictors were found to exert similar effects for males and females on each offence type. The current findings are discussed in relation to previous research and theory, and with respect to their implications for prevention-focused interventions for IPA and criminal offending.