This research was authorised as part of the overall current Youth Unit review with FReMO requirements being satisfied by those within Policy Development tasked with carrying out this review.


The participants were 69 youth offenders residing in the Waikeria Bay of Plenty Regional prison, Hawkes Bay, and Christchurch prison Young Offender Units (YOU). Those in remand were excluded from the study as they had not yet been sentenced, and thus may not become part of the target population (sentenced in the youth unit rather than community sentence, or no sentence). It was decided that using these three Youth Units would produce the most representative sample of young offenders within New Zealand. The Rimutaka Youth Unit was excluded from the study due to time and resource constraints. Only four youth offenders refused to take part in the study.

Psychometric Instruments

Four psychometric instruments were used for the purposes of this study: The Risk Screen Youth Offending Scale (RSYO); The Psychopathic Checklist: Youth Version (PCL: YV); the Youth Level of Service/Case Management Inventory (YLS/CMI); and Risk of Conviction X Risk of Imprisonment (RoC*RoI). Details of these instruments is contained in Appendix A.

A structured interview was designed by the researcher based on the interview and scoring guidelines that would enable the PCL: YV, RSYO, and YLS/CMI to be appropriately scored from one interview. This approach removed duplication of data enabled all the data to be scored from a single one and a half hour interview. RoC*RoI information was obtained from the Corrections Department computerised offender management system at the time of interview.

The PCL: YV, RSYO, and YLS/CMI were all hand scored based on the scoring guidelines from their respective manuals. Ratings on the YLS/CMI and RSYO were made blind to the PCL: YV ratings. While the RSYO and YLS/CMI were designed to be administered by a wide range of corrections staff with a limited amount of training, the PCL: YV required specialist training. The principal researcher in this study, Dr Nick Wilson is very experienced in the use of the

PCL-R and PCL: SV and has also attended advanced training with one of the PCL: YV developers, Dr Robert Hare.


All potential participants were approached first by the Young Offender Unit staff and provided with a participant information sheet that explained the purpose of the study, confidentiality and the limitations thereof, and that involvement in the study would in no way impact on their management within the Youth Unit for good or for bad (see Appendix B).

All young offenders who signed the consent form were then scheduled for an interview with either Dr Nick Wilson or another experienced Psychological Service Clinical Psychologist, Rehina Rolleston. Study files on each youth offender were prepared and reviewed prior to the interview. These included: electronic reports contained on IOMS; their current criminal history. In addition, prison files were reviewed by the researchers who appended relevant notes and copies of important documents to the study file. Interviews took place in November and early December 2003 at the three YOU’s. At the end of the interviews, participants were asked for their permission and details to ring and interview a significant other of their choice for the purpose of scoring the RSYO. A small koha 1 was provided to each participant with the permission of the three YOU unit managers in recognition of the time involved.

The RSYO’s were scored from the information obtained from the structured interview and the interview with the significant other. Structured phone interviews with the significant others were carried out following the guidelines from the user manual for the RSYO. These were conducted on a Monday and Tuesday night between 5pm and 7pm in December 2003. Within the RSYO there are four scales. The first scale is a reliability scale that is rated from a set of questions that are asked of both the offender and the significant other. Where a significant other was unable to be contacted (approximately 40%) the reliability scale was unable to be calculated. Items in the other scales are also scored from the questions that are asked of both the offender and the significant other. In order to retain a degree of consistency these items were rated based on the offender’s answers rather than that of the significant other. This is so that these items are scored the same across all of the tests regardless of whether or not the significant other could be contacted. Within the second scale there is a set of items that are rated based on questions that are only asked of the significant other. Where there was no significant other, these items were rated based on the interview and file information.

The PCL: YV and the YLS/CMI were both scored from the structured interview and file notes. The data from all three instruments, the RoC*RoI, and demographic and offence details were then entered into an excel spreadsheet. This data was then analysed using the Statistica 5.0 computerised data analysis software.

1 Maori term for gift showing respect. In the present study this consisted of a can of coke and a small chocolate bar.