Correctional Service Canada
www.csc-scc.gc.ca
Common menu bar links
Research Reports
This Web page has been archived on the Web.
Offender Risk/Needs Assessment: A Study Of Conditional Releases
The points of view expressed in this research report are those of the authors and do not necessarily reflect the views of the Correctional Service of Canada. This report is also available in French. Ce rapport est égalment disponible en Français. It is available from the Communications Branch, Correctional Service of Canada, 340 Laurier Avenue West, Ottawa, Ontario, K1A 0P9.
Prepared By:
Laurence L. Motiuk
Frank J. Porporino
Research and Statistics Branch
Correctional Service of Canada
February 1989
ACKNOWLEDGEMENTS
This project was initiated with the full support of the Division of Community Release Programs and Support Services at National Headquarters. Dru Allen deserves acknowledgement for encouraging and facilitating this research effort. Data collection proceeded with the competent assistance of Elizabeth Fabiano as the very methodical project coordinator, and Louis Brunet and Tanya Barker as the very conscientious coders. Peter Maitland was exceptionally helpful in retrieving case files from operational units across the country. Daniel Stripinis and Po Yee Lee assisted with the data analysis. In the later stages of the project, Len Epp assisted in reviewing and commenting on drafts of the report. We wish to thank all of these individuals for their assistance.
TABLE OF CONTENTS
LIST OF TABLES
TABLE 1 Percentage Distribution of Offenders by Region
TABLE 2 Characteristics of Sample
TABLE 3 Percentage Distribution by Risk Level
TABLE 4 Percentage of Cases Failing on Conditional Release Within Risk Level Categories
TABLE 5 Percentage Distribution of Offenders by Needs Level
TABLE 6 Outcome on Conditional Release for Cases With Identified Needs
TABLE 7 Percentage Outcome for Cases With Identified Needs on Parole or Mandatory Supervision
TABLE 8 Percentage Distribution of Needs Levels Within Risk Level Categories
TABLE 11 Outcome on Conditional Release for Different Risk/Needs Levels
LIST OF FIGURES
FIGURE 2 Percentage of Cases With Identified Needs on Parole and Mandatory Supervision
FIGURE 3 Percentage of Cases Who Failed on Conditional Release Within Needs Level Categories
I. INTRODUCTION
In Canada, as in other jurisdictions, the assessment of risk (of reoffending) has become an increasingly important and focused concern for both the Correctional Service of Canada and the National Parole Board. This has been illustrated recently, for example, by the clearer articulation of release decision-making policies that are based principally on considerations of risk (National Parole Board, 1988). In order to assess risk systematically and consistently, the National Parole Board has officially adopted the release risk scoring system developed by Nuffield (1982).
Although the General Statistical Information on Recidivism Scale has been established as an accurate predictor of release outcome, the criteria utilized for assessing "risk" are static in nature and cannot be responsive to differential management strategies1. As with many other instruments used for this purpose, the criminal history criteria that form the basis of the "risk assessment" are essentially fixed and unchangeable. Consequently, if risk scores are to be used to determine levels of supervision, then the problem of changing levels of supervision over time becomes clear.
1. Hann and Harman (1988) have recently revalidated the Statistical Information for Recidivism Scale with respect to conditional release outcome and this statistically derived scoring system has withstood the test of time.
This dilemma has led various jurisdictions to develop assessment tools made up of need factors that are both empirically related to release outcome and responsive to intervention (Clements, 1982; Duffee & Clark, 1981; Duffee & Duffee, 1985). The rationale is that as need areas are addressed, the relative risk the offender represents will be reduced and less supervision and contact will be required to manage that offender effectively in the community.
In most jurisdictions that have adopted a differential approach to community case management, risk and needs are combined to determine required levels or intensity of supervision (Clear & Gallagher, 1983). For example, the State of Wisconsin's Assessment of Client Needs Scale was developed in order to assess the problem and deficit areas of offenders while under community-based supervision (Baird, Heinz, & Bemus, 1979). Probation and parole agents follow a structured procedure to identify the presence or absence of various needs, establish a treatment approach to deal with those needs, make referrals to professional service providers when necessary, and use other available community resources for dealing with the various need categories.
The items selected for the Wisconsin Assessment of Client Needs Scale are typical of those included in other needs assessment instruments. A total of twelve need areas are covered as follows: academic/vocational skills, employment, financial management, marital/family relationships, companions, emotional stability, alcohol usage, other drug usage, mental ability, health, sexual behaviour, and an overall assessment of client needs.
Several studies have been conducted in order to evaluate the practical utility of instruments such as the Assessment of Client Needs Scale. Baird et al. (1979) reported on the community supervision of a large sample of probationers and parolees according to the needs assessment scale. Over a six month period, it was found that supervision levels could be reduced for a significant proportion of offenders who were initially assessed as requiring maximum levels of supervision. In another study, Baird et al. (1979) reported that the Assessment of Client Needs Scale was a valid indicator of the amount of time spent by case managers in supervising individual cases.
Despite the recent trend towards more systematic and structured assessment of offender needs, there is little research linking levels of need to outcome (Duffee & Duffee, 1985). Although previous work has looked at different approaches for the assessment of offender needs areas, there is little evidence to demonstrate how these assessments relate to outcomes. If a meaningful and significant relationship could be shown between identified needs and conditional release outcomes, then a combined risk/needs assessment procedure may be useful for differentiating offenders under community supervision.
In working towards the establishment of clear and comprehensive national standards for community supervision, the Correctional Service of Canada has recently endorsed the potential usefulness of risk/needs assessment (CSC, 1988). The present study takes a first step at examining the potential for an offender risk/needs assessment process in establishing guidelines or standards for varying levels of supervision on conditional release.
II. PRESENT STUDY
The present study was intended as a preliminary investigation of the relationships between risk and needs factors and particular outcomes under conditional release. As part of the study, we looked at both traditional risk factors (e.g., criminal history) and needs factors (e.g., alcohol abuse), and sought to determine whether independently, or in combination, these factors were related to release outcomes.
Major Aims of the Study
The present study had three related aims. First, to re-examine the predictive validity of a traditional risk assessment procedure, separately for Parole and Mandatory Supervision samples. A second purpose was to attempt to identify some of the major needs characteristics of those offenders who succeed or fail while under conditional release. Although criminal history factors have been found to be useful in the prediction of release outcome, there have been no specific attempts to compare these risk factors with identified needs. Finally, a third purpose was to examine how a combination of risk and needs assessment might yield improved predictions of release outcome.
Sample Selection
The sample of cases for the study was selected as follows. The automated Offender Information Service was used to identify all offenders on conditional release who either: 1) successfully completed their parole or mandatory supervision (i.e., reached their warrant expiry date) between January 1 and December 31, 1985, or 2) had their parole or mandatory supervision revoked during that year. This search yielded approximately 4,523 cases from all across Canada.
At this point, further criteria were applied to generate the final sample. We sought to give all five regions of the Correctional Service Canada equal representation. That is, the same number of subjects would be drawn from each region. Within each region, male offenders would be classified into one of four groups: 1) reached warrant expiry date while on parole; 2) revoked while on parole; 3) reached warrant expiry date while on mandatory supervision; or 4) revoked while on mandatory supervision. Within these four conditions, cases were selected randomly by region. This resulted in a randomly stratified sample of approximately 600 cases; 150 cases per conditional release grouping.
However, information on all 600 cases was not accessible. Some of the files were difficult to locate, some were being held on active status (i.e., on current use), and still others were eliminated because it was found that they did not, in fact, meet the sampling criteria. A total of 221 cases remained at this stage of the sampling process. These cases consisted of 103 (46.6%) offenders who were on parole and 118 (53.4%) who were on mandatory supervision. Of those who were on parole, 43 (41.8%) had been revoked, while for those who were on mandatory supervision, 56 (47.5%) had been revoked. In Table 1, we show the percentage distribution of offenders in each region for both types of conditional release outcomes.
| Region | Release Outcome | |
|---|---|---|
| Reached Warrant Expiry | Revoked | |
| Atlantic | 20.4% (20) | 18.9% (23) |
| Quebec | 20.4% (20) | 26.2% (32) |
| Ontario | 12.2% (12) | 11.5% (14) |
| Prairies | 30.6% (30) | 26.2% (32) |
| Pacific | 16.3% (16) | 17.2% (21) |
Data Gathering Process
Information was gathered from the community case management files of each case. A structured information coding schedule was developed in order to gather relevant data in 12 general areas: demographics, juvenile history, index offence(s), offence dynamics, socio-economic and psychological background, institutional adjustment, pre-release performance, release data, community supervision performance, final outcome data, original offence(s), and prior criminal history. Although considerable case specific information was collected, this report focuses only on some of the risk/needs information that was gathered.
The most relevant documents, for coding purposes, were the following: Penitentiary Placement Form, Finger Print Serial Number (FPS) Sheet, Pre-sentence Report, Psychological Assessment forms, police reports, and the Quarterly Supervision Reports. These documents were not available in every case. In these instances, numerical codes were used to indicate that information on a particular factor was either "missing", "not specified", or "not relevant". Most of the variables involved objective ratings (e.g., birthdate) on the part of the coder. A small number of variables involved some interpretation (e.g., emotional stability) as to the status of the subject on a particular factor.
Two coders were trained for an initial two week period on proper procedures for reviewing case files. Reliability checks were also made throughout the study in order to ensure that each coder was being consistent, and to minimize the subjectivity involved in rating a number of items.
A project coordinator was responsible for overseeing the review of files. The project coordinator and the two coders all had considerable experience in the field of criminology. In addition, one of the coders had worked for a number of years as a community case management officer in the Correctional Service Canada. Any discrepancies on the coding of information were discussed and resolved by consensus between the coders and project coordinator.
III. FINDINGS
Characteristics of Sample
The sample of cases studied had been released either under parole or mandatory supervision. In order to provide a general picture of the differences between these two groups of offenders, we looked at how they varied on a selected set of background characteristics. These findings are shown in Table 2.
| Category | Release Type | p | |||
|---|---|---|---|---|---|
| % | n | % | n | ||
| Previous History of: | |||||
| Substance Use/abuse | 46.6 | 48 | 77.1 | 91 | *** |
| Mental Health Problems | 1.9 | 12 | 5.9 | 7 | ns |
| Type of Index Offence: | |||||
| Sexual | 1.8 | 8 | 5.9 | 7 | ns |
| Assaultive | 53.4 | 55 | 42.4 | 50 | ns |
| At Time of Offence: | |||||
| Under Influence of Alcohol | 46.5 | 46 | 60.5 | 69 | * |
| Under Influence of Drugs | 17.2 | 17 | 22.1 | 25 | ns |
| Under Supervision | 29.1 | 30 | 70.0 | 79 | *** |
| History Prior to Release: | |||||
| Protective Custody | 10.7 | 11 | 23.7 | 28 | * |
| Administrative Segregation | 9.7 | 10 | 14.4 | 17 | ns |
| Admitted to Regional Treatment or Psychiatric Centre | 2.9 | 3 | 11.0 | 13 | * |
| Note: ns = nonsignificant; *p< .05; **p< .01; ***p< .001 | |||||
Parole releases were not differentiated from mandatory supervision releases on every dimension. There were no significant differences noted in terms of history of mental health problems, whether the index offence was sexual or assaultive in nature, whether the offender was under the influence of drugs at the time of the index offence, and whether there was an institutional record of administrative segregation.
On the other hand, some clear differences did emerge. First, it was noted that less than half of the offenders released on parole (46.6%) had a history of substance use/abuse, compared to 77.1% of those released under mandatory supervision. Furthermore, a high percentage of offenders released under mandatory supervision (60.5%) had been under the influence of alcohol at the time of their index offence.
Secondly, although most of the mandatory supervision cases (70%) had been under some form of community supervision at the time of their index offence, less than a third (29.1%) of those released on parole had a similar history.
Third, a comparison of characteristics related to institutional record indicated that those released under mandatory supervision were both more likely to have been placed in protective custody (23.7%), and more likely to have been admitted to regional treatment or psychiatric centres (11.0%).
Release Risk Assessment
In order to determine level of risk, the Statistical Information on Recidivism Scale was scored retrospectively through an assessment of case file information for each offender. The items were scored in the standard weighted format outlined by Nuffield (1982) and the total risk score was simply the summation of items corresponding to the following levels of risk: "Very Good" (-27 to -6), "Good" (-5 to -1), "Fair" (0 to +4), "Fair to Poor (+5 to +8), and "Poor" (+9 to +30). In order to facilitate the presentation of data in percentage form, these five risk levels were reassigned into three groupings ("Low", "Medium", and "High") as shown below. It should be noted that since the risk level groupings were collapsed, this narrows the range of differentiation that the scoring system can account for in release outcome.
| Risk Level | S.I.R. Score |
|---|---|
| Low | (-27 to - 1) |
| Medium | (0 to + 4) |
| High | (+5 to +30) |
Table 3 shows how cases were distributed by risk level, for both the parole and mandatory supervision samples. As expected, the majority of those released on parole were classified as low risk (59.2%). In contrast, most of those released under mandatory supervision were classified as high risk (62.7%).
| Risk Level | Release Type | |
|---|---|---|
| Parole (n = 103) | Mandatory Supervision (n = 118) | |
| Low | 59.2% (61) | 14.4% (17) |
| Medium | 15.5% (16) | 22.9% (27) |
| High | 25.2% (26) | 62.7% (74) |
In Table 4, we show the overall percentage of cases who failed within the three risk level groupings. Again, as expected, those offenders who were classified as high risk had the highest failure rate (50%) on conditional release, with those cases classified as medium or low risk showing proportionately lower rates of failure.
| Risk Levels | ||
|---|---|---|
| Low | Medium | High |
| 38.5% (n = 78) | 44.2% (n = 43) | 50.0% (n = 100) |
Figure 1 shows the relationship between risk level and failure rates, separately for parole and mandatory supervision cases. As Figure 1 shows, the Statistical Information on Recidivism Scale is quite able to identify cases who will fail on parole. In this case, the probability of failure is as expected, increasing proportionately with level of risk. On the other hand, differentiation by risk level seems to break down when we look at mandatory supervision cases as a separate group. We now see that the low risk mandatory supervision group had the highest failure rate (52.9%) relative to both the medium risk group (40.7%) and the high risk group (48.6%). This finding can perhaps be accounted for by the small sample sizes we examined. Nevertheless, the results do raise questions about the usefulness of the Statistical Information on Recidivism Scale with mandatory supervision cases.
FIGURE 1
Percentage of Cases Who Failed on Parole or Mandatory Supervision Within Risk Level Categories
Release Needs Assessment
The degree of uncertainty in classifying offenders into differential risk levels serves to illustrate the kinds of practical challenges experienced by correctional managers. Although we re-examined the usefulness of the Statistical Information on Recidivism Scale as a tool for differentiating offenders by level of risk, our main focus in this report was to explore the relationship between identified needs and conditional release outcomes.
In reviewing case files, we extracted a range of information on need dimensions covered by needs assessment instruments. In particular, we paralleled our coding procedures against Wisconsin's Assessment of Client Needs Scale. We excluded several need areas for which it was decided there would be inconsistencies in reporting (e.g., sexual behaviour). Information was collected on eight need factors: Employment, Financial Management, Marital/Family Relationships, Living Arrangement, Companions, Alcohol/Drug Usage, Emotional Stability, and Attitude. Weighted ratings (see Appendix) on each need dimension were arrived at for each subject from reviews of documentation for the first quarter of community supervision, with particular emphasis on the first Quarterly Supervision Report. These weighted ratings were rescored into a binary format, indicating simply whether a particular need was identified by the supervising case management officer. We then added the number of needs that were identified into a total Needs Score for each case. Arbitrarily, we divided these total scores into three needs level groupings as shown below.
| Need Level | # of Identified Needs |
|---|---|
| Low | (1 to 3) |
| Medium | (4 to 6) |
| High | (7 to 8) |
Table 5 shows how offenders were distributed by needs level groupings, separately for both parole and mandatory supervision cases. We note that many of those released on parole had no identified needs (40%). In contrast, only 9.5% of cases released under mandatory supervision had no identified needs, while over half had at least one or as many as three identified needs (54.3%). In addition, only the former group had as many as seven or eight identified needs (8.6%).
| Needs Level | Release Type | |
|---|---|---|
| Parole (n = 100) | Mandatory Supervision (n = 105) | |
| None | 40.0% (40) | 9.5% (10) |
| Low (1 to 3) | 47.0% (47) | 54.3% (57) |
| Medium (4 to 6) | 13.0% (13) | 27.6% (29) |
| High (7 to 8) | -- | 8.6% (9) |
An important question that remains is the relationship between identified needs and subsequent release outcome. Table 6 shows the number of cases in the sample where a particular need area was identified. For each need area, the Table shows the percentage distribution of failures and successes associated with that need. Cross tabulations revealed that four of the eight need factors studied did not significantly differentiate between failure and success on conditional release. The four need factors found not to be significant were Employment; Financial; Marital/ Family; and Emotional Stability.
The four need factors which significantly differentiated between failure and success on conditional release were Living Arrangement; Companions; Alcohol/Drug Usage; and Attitude. These four need factors were related to outcome as follows: (1) the less stability in the offender's living situation, the greater the likelihood of that offender being revoked; (2) the more negative an offender's associations (i.e., pro-criminal) in the community, the higher the probability of being revoked; (3) the more disfunctioning an offender's alcohol/drug usage, the more likely that offender was to be revoked; and (4) the less responsive the offender was to assistance, the greater were the chances of being revoked.
| Dimension | # of Cases With Need | Failure % | Success % | p |
|---|---|---|---|---|
| Employment | 123 | 50.4 | 49.6 | ns |
| Financial | 59 | 50.8 | 49.2 | ns |
| Marital/Family | 58 | 48.3 | 51.7 | ns |
| Living Arrangement | 44 | 70.5 | 29.5 | *** |
| Companions | 39 | 69.2 | 30.8 | *** |
| Alcohol/Drug Usage | 50 | 68.0 | 32.0 | *** |
| Emotional Stability | 43 | 53.5 | 46.5 | ns |
| Attitude | 82 | 65.9 | 34.1 | *** |
| Note: ns = nonsignificant; *p< .05; **p< .01; ***p< .001 | ||||
Considering that four out of eight need factors significantly differentiated failures from successes on conditional release, it remains important to establish whether any significant differences exist with respect to release type. Figure 2 shows the percentage of cases with identified needs who were released on parole or mandatory supervision. As Figure 2 shows, all of the identified need factors significantly differentiated parole releases from mandatory supervision releases. Notably, of those with an identified Employment need, 32.0% were released on parole compared to 76.3% released on mandatory supervision. It would appear that release decisions are being made in a fashion quite consistent with the assessment of need factors.
FIGURE 2
Percentage of Cases With Identified Needs on Parole and Mandatory Supervision
In order to establish predictive efficiency with respect to both release type and release outcome, we again examined separately those cases released on parole and those released on mandatory supervision. As Table 7 indicates, only three need factors significantly differentiated failure from success on parole. Those factors were Companions, Alcohol/Drug Usage, and Attitude. All of the need factors that were related to outcome on parole were also related to outcome under mandatory supervision. In addition, those offenders who had instability in their living arrangement while under mandatory supervision were also more likely to be revoked.
| Dimension | Parole | Mandatory Supervision | ||||
|---|---|---|---|---|---|---|
| Failure | Success | p | Failure | Success | p | |
| Employment | 51.5% (17) | 48.5% (16) | ns | 50.0% (45) | 50.0% (45) | ns |
| Financial | 38.9% (7) | 61.1% (11) | ns | 56.1% (23) | 43.9% (18) | ns |
| Marital / Family | 45.0% (9) | 55.0% (11) | ns | 50.0% (19) | 50.0% (19) | ns |
| Living Arrangement | 71.4% (5) | 28.6% (2) | ns | 70.3% (26) | 29.7% (11) | *** |
| Companions | 72.7% (8) | 27.3% (3) | * | 67.9% (19) | 32.1% (9) | * |
| Alcohol / Drug Usage | 83.3% (10) | 16.7% (2) | *** | 63.2% (24) | 36.8% (14) | ** |
| Emotional Stability | 38.5% (5) | 61.5% (8) | ns | 60.0% (18) | 40.0% (12) | ns |
| Attitude | 77.8% (14) | 22.2% (14) | *** | 62.5% (40) | 37.5% (24) | *** |
| Note: ns = nonsignificant; *p , .05; **p < .01; ***p < .001 | ||||||
Although identified needs were found to significantly differentiate release types as well as release outcomes, we were also interested in whether a Needs Level grouping (i.e., which collapses the number of identified needs into needs levels) could improve the differentiation between failure and success on conditional release. Inspection of Figure 3 reveals that as Needs Level scores increase, so does the overall percentage of cases who failed on conditional release.
FIGURE 3
Percentage of Cases Who Failed on Conditional Release Within Needs Level Categories
In Figure 4, we show the relationship between level of need and failure rate, separately for the parole and mandatory supervision cases. A consistent trend is maintained for both the parole and mandatory supervision cases; as needs level increases so does the failure rate.
FIGURE 4
Percentage of Cases Who Failed on Parole or Mandatory Supervison Within Needs Level Categories
Release Risk/Needs Assessment
A final aim of the study was to determine whether there were combinations of risk/needs levels which would yield improved release predictions. Tables 8 through 10 present the percentage distributions of needs levels within risk level categories.
Beginning with Table 8, we note considerable variation in needs levels within each risk level category. For example, there were 27 (35.5%) cases identified with no needs, 39 (51.3%) cases identified with low needs, and 10 (13.2%) cases identified with medium needs who were assessed as low risk for recidivism.
| Risk Level | Needs Levels | |||
|---|---|---|---|---|
| None | Low | Medium | High | |
| Low | 35.5% (27) | 51.3% (39) | 13.2% (10) | -- |
| Medium | 23.8% (10) | 50.0% (21) | 21.4% (9) | 4.8% (2) |
| High | 14.9% (13) | 50.6% (44) | 26.4% (23) | 8.1% (7) |
Table 9 presents the percentage distribution of needs levels within risk level categories specifically for those offenders released on parole. Again there is considerable variation in needs level within each risk level category.
| Risk Level | Needs Level | |||
|---|---|---|---|---|
| None | Low | Medium | High | |
| Low | 41.7% (25) | 46.7% (28) | 11.7% (7) | -- |
| Medium | 50.0% (8) | 25.0% (4) | 25.0% (4) | -- |
| High | 29.2% (7) | 62.5% (15) | 8.3% (12) | -- |
With respect to offenders released on mandatory supervision, Table 10 presents the percentage distribution of needs levels within risk level categories. Similarly, there is variation with respect to needs level within each risk level category. Noteworthy are the six cases within the high risk level category who were all identified as having no needs.
| Risk Level | Needs Level | |||
|---|---|---|---|---|
| None | Low | Medium | High | |
| Low | 12.5% (2) | 68.8% (11) | 18.7% (3) | -- |
| Medium | 7.7% (2) | 65.4% (17) | 19.2% (5) | 7.7% (2) |
| High | 9.5% (6) | 46.0% (29) | 33.3% (21) | 11.1% (7) |
While needs level appears to show ample variation within risk level categories, the overwhelming consideration would be whether there are combinations of risk/needs levels which yield improved release predictions. Table 11 presents the relationship between combined risk/needs levels and conditional release outcome. The data presented include the number of cases that fall within a particular risk/needs level and the percentage distribution of failures associated with that risk/needs level.
The risk/needs level combinations are rank ordered in Table 11 according to percentage failure from the highest to the lowest. It can be seen from the results that when risk/needs levels are combined, a pattern of needs level groupings emerges. That is, we see a natural grouping of needs levels so that, for example, regardless of risk level, medium needs cases have higher failure rates than low need cases. On the other hand, this kind of grouping phenomenon does not occur with risk levels. We can conclude that it is principally the level of need that determines failure rates on conditional release and not the level of risk.
Of special interest was the finding that the majority of offenders assessed as high risk/high need (71.4%) and high risk/medium need (52.2%) failed on conditional release. In contrast, only 30.8% of the offenders assessed as high risk/no need failed on conditional release. These findings suggest that a risk/needs instrument may be particularly useful for the differential allocation of controls and services to higher risk cases on conditional release.
| Level | # of Cases | % of Failure |
|---|---|---|
| High Risk/ High Need | 7 | 71.4 |
| Medium Risk/ Medium Need | 9 | 66.7 |
| Low Risk/ Medium Need | 10 | 60.0 |
| High Risk/ Medium Need | 23 | 52.2 |
| High Risk/ Low Need | 44 | 45.5 |
| Low Risk/ Low Need | 39 | 39.5 |
| Medium Risk/ Low Need | 21 | 39.1 |
| High Risk/ No Need | 13 | 30.8 |
| Low Risk/ No Need | 27 | 26.0 |
| Medium Risk/ No Need | 10 | 20.0 |
| Note: Cells that had expected counts less than 5 were not included, as tests of significance may not be valid. | ||
IV. DISCUSSION
In general, the results of the study indicate that offender risk/needs assessments may have some practical utility for community case managers. The following summarizes the major findings:
- the study reconfirmed the Statistical Information on Recidivism Scale's ability to predict outcome on parole. However, the scale's ability to predict mandatory supervision outcome is questioned.
- the findings of the study indicated a consistent relationship between identified need areas and both type of release and release outcome.
- a combination of risk/needs assessment can result in improved predictive efficiency.
With only a very simple scoring of the number of need areas that were identified by supervising case managers, we were able to achieve prediction of release outcome that surpassed that provided by a sophisticated statistical risk prediction instrument.
It would seem that further development and refinement of a risk/needs assessment tool would be a worthwhile pursuit, especially in light of its potential relevance to conditional release outcomes. Such an assessment instrument could be used to focus supervision resources. Moreover, it would provide a means of monitoring changes in the offender's behaviour, attitudes and circumstances which are clearly related to release outcome.
REFERENCES
Baird, S.C., Heinz, R.C., & Bemus, B.J.(1979). Project report# 14: A two Year follow-up. Wisconsin: Department of Health and Social Services, Case Classification/Staff Deployment Project, Bureau of Community Corrections.
Clear, T.R., & Gallagher, K.W.(1985). Probation and parole supervision: A review of current classification practices. Crime and Delinquency, 31, 423-443.
Clements, C.B.(1982). Offender needs assessment. College Park, MD: American Correctional Association.
Correctional Service of Canada. (1988). Standards for conditional release. Ottawa: Correctional Service of Canada.
Duffee, D.E., & Clark, D. (1985). The frequency and classification of the needs of offenders in community settings. Journal of Criminal Justice, 13, 243-268.
Duffee, D.E., & Duffee, B.W. (1981). Studying the needs of offenders in pre-release centres. Journal of Research in Crime and Delinquency, 18, 232-253.
Hann, R.G., & Harman, W.G. (1988).Release risk prediction: A test of the Nuffield Scoring System. Ottawa: Ministry of the Solicitor General of Canada.
National Parole Board of Canada. (1988). Policy and procedures manual. revised. Ottawa: National Parole Board of Canada.
Nuffield, J. (1982). Parole decision-making in Canada: Research towards decision guidelines. Ottawa: Communication Division.
Appendix A |
||||
Need Dimension Definitions and Scoring |
||||
Note: |
Need dimensions are scored in a binary format |
|||
("0" if not identified and "1" if identified) |
||||
Employment |
||||
(Percentage of time employed) |
||||
0. |
60% or more; |
|||
40-59%; |
||||
Disabled/students |
||||
1. |
less than 40% |
|||
Financial |
||||
0. |
Long-standing pattern of self-sufficiency: |
|||
e.g. good credit rating; |
||||
No current difficulties |
||||
1. |
Situational or minor difficulties; |
|||
Severe difficulties; may include |
||||
garnishment, bad checks or bankruptcy |
||||
Marital/Family |
||||
0. |
Relationships and support |
|||
exceptionally strong; |
||||
Relatively stable relationships |
||||
1. |
Some disorganization or stress but |
|||
potential for improvement; |
||||
Major disorganization or stress |
||||
Living Arrangement |
||||
(Stability of offender's living arrangements) |
||||
0. |
Above average; |
|||
Average |
||||
1. |
Below average |
|||
Companions |
||||
0. |
Good support and influence; |
|||
No adverse relationships |
||||
1. |
Associations with occasional |
|||
negative results; |
||||
Associations almost completely |
||||
negative |
||||
Alcohol/Drug Usage |
||||
0. |
No interference with functioning; |
|||
1. |
Occasional abuse, with some |
|||
disfunctioning; |
||||
Frequent abuse, serious |
||||
disfunctioning, needs treatment |
||||
Emotional Stability |
||||
0. |
Exceptionally well-adjusted; accepts |
|||
responsibility for actions; |
||||
No symptoms of emotional |
||||
instability; appropriate emotional |
||||
responses |
||||
1. |
Symptoms limit but do not prohibit |
|||
adequate functioning: e.g. excessive |
||||
anxiety |
||||
Symptoms prohibit adequate |
||||
functioning e.g., lashes out or |
||||
retreats into self |
||||
Attitude |
||||
0. |
Motivated to change, receptive to |
|||
assistance |
||||
1. |
Dependent, unwilling to accept |
|||
responsibility; |
||||
Rationalizes behaviour, negative, |
||||
not motivated to change |
||||
