Nature, extent and correlates of bullying and assault in penal populations
The present thesis is a detailed examination of bullying behaviour in Young Offender Institutions and assaultive behaviour in adult prisons and Young Offender Institutions. An amalgam of methods of enquiry were used throughout, which comprised i) questionnaires, ii) structured interviews. Hi) focus groups, iv) standardised measures of personality, intelli^ce and social background, v) analysis of official discipline report records, vi) analysis ff official assault incident report records and vii) analysis of computerised prisoner records. This mixture of methods, known as ‘triangulation’, was adopted in an attempt to achieve a more reliable and valid representation of bullying and assaultive behaviour occurring within penal establishments. Information was obtained from a variety of subject groups, including convicted young offenders, young offenders on remand, convicted adult prisoners, adult prisoners on remand, prison officers, prisons management and specialist staff working in the prisons. Data were analysed by means of parametric and non-parametric statistical techniques. Seven cross-sectional studies were designed and conducted, the results of which are reported herein. The first five studies examined only Young Offender Institutions. Concerning young offenders, the levels of bullying ivere comparatively high when compared with studies done on analogous populations. Among young offenders, the most common types of bullying were similar to those shown in previous studies, such as taxing, threats and name calling. More staff in Young Offender Institutions perceived bullying as a problem both nationally and in their own establishment than did young offenders. While staff and young offenders had discordant opinions as to the levels and types of bullying taking place within Young Offender Institutions, they had concordant views as to the characteristics of ‘bullies’ and ‘victims’. The types of bullying (i.e. covert and overt) varied considerably according to type of Young Offender Institution and type of young offender under study. The introduction of an anti-bullying initiative in one Young Offender Institution appeared to change the way bullying was manifested, as noted in the prison records, by reducing overt bullying behaviour and increasing more subversive and covert bullying. The remaining two studies in the thesis examined assault in both adult prisons and Young Offender Institutions. Results revealed that the typology of assaults in adult prisons and young offender establishments were dissimilar in important respects. In Young Offender Institutions the assaults on prisoners were more likely to be ‘spontaneous’ and result in less severe injury to the victim, whereas in adult prisons the assaults were more likely to be ‘planned’ and result in more severe injury to the victim. When looking at sub-groups within Scottish prisons and Young Offender Institutions using discriminant function analysis, victims of assaults on prisoners were distinguishable from both perpetrators of assaults on staff and perpetrators of assaults on other prisoners, using a range of social background factors. Victims of assault in adult prisons were more accurately identified (91% correctly identified, compared with 43% at ‘chance’) than victims in Young Offender Institutions (73% correctly identified, compared with 47% at ‘chance’). The introduction of anti-bullying initiatives into young offender establishments, and in particular, how they might effect overt and covert bullying in contrasting^ ways, is reflected upon. Moreover, the importance of obtaining information from a variety of subject groups and an amalgam of data gathering techniques ts highlighted. The utility of using factors relating to an inmate’s social background, personality and intelligence to predict involvement in bullying is discussed. Finally, the main findings from the thesis are discussed in relation to the relevant literature, practical implications for intervention and areas where future research may be necessary.