FORUM on Corrections Research

Risk classification for young offenders

and Edward P. Rawana¹

Lakehead Regional Family Centre and Lakehead University

and Byron Lod

Probation Services, Ministry of Community and Social Services

After a young offender becomes involved with the criminal justice system, a probation officer plays a significant role in the case management decisions for that young offender. Some decisions are usually based on some form of offender risk/needs classification which helps the probation officer make decisions on the treatment needs of the youth and to assess the youth's risk of reoffending.

Several tools have been developed to assess risk and need; however, many were initially designed for adult offenders and are not adjusted to focus on the risk and need factors specific to adolescent offenders. Previous research² has emphasized the importance of evaluating the validity of any risk screening instrument. This is because the use of any risk screening device makes the probation or correctional worker accountable for the manner in which he or she uses resources to deal with young clients. Most young offender classification research does not include an adequate sampling of minority groups such as Aboriginal youths. Also, there is little information on the risk and need factors which apply to female youths.

A collaborative effort among all the probation offices in northwestern Ontario (extending from White River to the Manitoba border), a mental health facility for children and a local university provided the necessary components to conduct a thorough evaluation of a risk and need instrument for young offenders. The instrument was developed and implemented in Ontario and is called the Ministry's Risk/Need Assessment Form. It was specifically designed to identify offending youths at risk of reoffending and to aid the probation officer in addressing treatment needs.

This article evaluates the validity of this relatively new risk and need tool through a focus on young offenders in northwestern Ontario where there is a large proportion of Aboriginal youths.

Risk/Need Assessment Form

The Risk/Need Assessment Form, otherwise known as the Youth Level of Service Inventory (YLSI), was preceded by the Level of Supervision Inventory (LSI) which is currently used with adult offenders. Both the Risk/Need Assessment Form and the LSI are based on Andrew, Bonta and Hoge's four principles of risk classification: risk principle, need principle, responsivity principle and professional discretion.³

The Risk/Need Assessment Form is designed for Phase I young offenders, or youths aged 12 to 15 years. It assesses 42 items which are grouped into eight individual risk and need factors: prior/current offences/disposition, family circumstances/parenting, education/ employment, peer relations, substance abuse, leisure/recreation, personality/behaviour and attitudes/orientation. Scores can range from 0 to 42. Youths with scores ranging from 0 to 8 are classified as low risk, scores ranging from 9 to 26 indicate moderate risk, scores of 27 to 34 are considered high risk, and youths scoring between 35 to 42 are considered very high risk.

Table

Risk/Needs Level by Outcome
Risk Level
	Low		Moderate		High		Very High
	frequency	%	frequency	%	frequency	%	frequency	%
Recidivist	17/76	22%	52/76	68	5/76	7%	2/76	3%
Non-recidivist	99/174	57%	71/174	41%	4/174	2%	0/174	0%

The Ministry of Community and Social Services of Ontario began to use the form in all probation offices and mandated that all young offenders be assessed on the eight factors, and then reassessed at every six months of their disposition.

There are some preliminary unpublished data on the tool suggesting it has adequate reliability and validity;⁴ however, the results were obtained from a region in which the form was developed and the norms set.

Northwestern Ontario

While the probation offices of northwestern Ontario cover a large area, little research has been conducted with young offenders in this region with respect to risk and need. Evaluating the instrument in this region was important because of the region's uniqueness in terms of its population diversity.

Northwestern Ontario has two major probation offices with access to numerous remote Indian reserves. There is, therefore, a gross representation of Aboriginal young offenders: almost 50% of the population of young offenders in northwestern Ontario is Aboriginal and many of these are female.⁵

Methodology

The sample of young offenders was drawn from the clientele of two regional probation offices over a nine-month period. Two hundred and fifty young offenders were assessed by probation officers. Several information sources were used to assess youths on the Risk/Need Assessment Form, including interviews with the youths and their parents, and a review of pertinent record and file information. The average age of the youths was 14.3 years (SD = 11.1) at the time of assessment. These were 166 (66.4%) males and 84 (33.6%) females. One hundred and twenty-six (50.4%) were Aboriginal youths and 124 (49.6%) were non-Aboriginal youths.

All the youths were followed up six months after the assessment date, except for young offenders in custody who were followed up⁶ six months after their release. At follow-up, it was determined whether the youth was a recidivist. Recidivism was defined, for the purpose of this study, as any conviction for an offence committed up to six months after release. Conviction information was obtained from police reports and probation databases.

Results

The average total score on the Risk/Need Assessment Form was 11.2. Of these young offenders, 116 were classified as low risk, 123 as moderate risk, 9 as high risk and 2 as very high risk. At the end of the six-month follow-up, it was found that 30.4% of the young offender sample were recidivists and 69.6% were not.

The young offenders' performance on the Risk/Need Assessment was strongly related to their actual reoffending as demonstrated by the trends seen in examining the distribution of recidivists and non-recidivists on each of the four risk levels (see table). A large proportion of the recidivists were classified at moderate risk for reoffending (52/76) and a large proportion of the non-recidivists were classified at low risk (99/174, see table).

Figure

An examination of the total risk/needs scores of recidivating and non-recidivating young offenders provided additional support for the utility of the instrument. Despite the cultural diversity of the study's sample, recidivists (M = 15.74; SD = 8.01) were assessed with a significantly higher (p < .001) overall risk score than their non-recidivist counterparts (M = 9.22; SD = 7.46).

To test the validity of the tool further, individual risk and need factors were investigated for discrimination between recidivists and non-recidivists. Analyses indicated that all eight risk/needs factors were important predictors of recidivism, with recidivists scoring significantly higher on each factor than non-recidivists (see figure). This finding provides strong support for the usefulness of the form in northwestern Ontario, especially when one considers that it was developed in southern Ontario and that northwestern Ontario is overrepresented by Aboriginal young offenders. It is interesting to note that the strongest factor capable of discriminating between the two groups of young offenders was the attitude of the young offender. However, it was surprising to find that previous and current offences and disposition provided only a moderate discriminator of recidivism, since much of the literature has emphasized the predictive ability of past behaviour.

Discussion

The results strongly support the validity of the Risk/Need Assessment Form as an instrument for assessing a young offender's risk of reoffending. The eight risk/needs factors appear to be relevant and important in differentiating recidivists from non-recidivist youths. Therefore, one can conclude that recidivism can be predicted based on the youth's score on the form.

Past research⁷ has demonstrated the importance of validating instruments in jurisdictions other than where they were originally developed and the norms established. This is especially relevant since the place where it was developed does not have the diverse group of young offenders seen in other regions. The collaboration between probation and local resources provided the necessary efforts to address these issues. This study also included a sample of female offenders which was more than adequate when compared to other young offender studies. This investigation has demonstrated that the Risk/Need Assessment Form is not only valid in predicting risk, but also robust with respect to jurisdiction, ethnicity and sex.

---

1. 283 Lisgar Street, Thunder Bay, Ontario, P7B 6G6.

2. J.S. Wormith and C.S. Gladstone, "The Clinical and Statistical Prediction of Recidivism," Criminal Justice and Behavior, 11 (1984): 3-34.

3. D.A. Andrews, J. Bonta and R.D. Hoge, "Classification for Effective Rehabilitation: Rediscovering Psychology," Criminal Justice and Behavior, 17 (1990): 19-52.

4. R.D. Hoge, personal communication.

5. Statistical data on young offender population was obtained through Probation Services via their Young Offender Strategic Information System (YOSIS).

6. Risk/Needs Factors in figure: OFF - Prior and current offences/dispositions; FAM - Family circumstances/parenting; EDUC - Education/employment; PEER - Peer relations; SUB - Substance abuse; LEIS - Leisure/recreation; PERS - Personality/behaviour; ATT - Attitudes/orientation.

7. J.B. Ashford and C.W. LeCroy, "Predicting Recidivism: An Evaluation of the Wisconsin Juvenile Probation and Aftercare Risk Instrument," Criminal Justice and Behavior, 15 (1988): 141-151.