Matching Services to Criminogenic Needs Following Comprehensive Risk Assessment Implementation in Juvenile Probation

Abstract

One emphasis of juvenile justice reform has been implementation of risk assessment instruments to improve case planning. This study examined the ability of juvenile probation departments to apply the risk-needs-responsivity (RNR) framework into case planning following a comprehensive implementation protocol. Data were collected on 385 adolescent offenders across three probation departments following implementation of the Structured Assessment of Violence Risk for Youth (SAVRY) and an RNR-related case planning policy. As expected, as risk levels of youth increased, probation departments assigned more services and addressed more criminogenic need areas in their case plans. Most case plans (86%) adhered to the policy to limit the number of needs addressed at one time. The quality of service-to-need matching varied by criminogenic need area, risk level, and site. Implications to juvenile courts’ and probation officers’ case planning and the challenges of research on service-to-need matching are discussed.

Keywords

juvenile probation criminogenic needs service-to-need matching SAVRY risk-needs-responsivity (RNR)

Over the last 20 years, the juvenile justice system in the United States has been involved in what has been referred to as the fourth wave of reform (Grisso, 2017). This wave acknowledges adolescents are different from adults and strongly pursues the integration of research, including research on adolescent development and immaturity, into juvenile justice practice (National Research Council, 2013). This is a significant shift from the prior wave of reform that emphasized punishment and treated youth much like adults. Although the concepts are not new (see the Juvenile Justice and Delinquency Prevention Act, 2002), one emphasis of the reform effort over the last decade has been the implementation of risk assessment instruments to improve case processing decisions and outcomes. The National Resource Council (2013) and others (Seigle, Walsh, & Weber, 2014; Vincent, Guy, & Grisso, 2012) strongly recommended structured risk and need assessment tools (also referred to as reduction-oriented instruments; Monahan & Skeem, 2014), be used to identify low-risk youth who could be handled less formally, to match youth to appropriate treatment, and to target high-risk youth for more intensive interventions.

These recommendations are based largely on the tenets of the risk-needs-responsivity (RNR) model for case management, which has considerable evidence as a method for reducing recidivism among adult offenders (Andrews & Bonta, 2010, 2017; Andrews, Bonta, & Hoge, 1990). The RNR framework suggests the highest risk offenders should receive the most intensive programming to reduce risk of reoffending (risk principle), and the programming should specifically target the individual’s criminogenic needs (the variable risk factors that appear to be driving their offending; need principle) while taking into account specific characteristics that may affect treatment response (responsivity principle). Factors that affect treatment response include both general (the treatment strategy used) and specific (individual characteristics such as strengths, motivation, etc.) responsivity factors.

Although the principles of the RNR model are conceptually very simple, the handful of studies to date, which primarily come from the adult system, have indicated that implementing RNR principles is much more difficult (Flores, Travis, & Latessa, 2004; Haas & DeTardo-Bora, 2009; Miller & Maloney, 2013; Viglione, Rudes, & Taxman, 2014). Surveys of practitioners in correctional settings (Flores et al., 2004) and adult probation officers (Miller & Maloney, 2013) report overall poor adherence to use of risk assessment to guide treatment goals. For example, in both interviews and observations of probation officers (PO), Viglione and colleagues (2014) found that although POs overwhelmingly administered the risk instrument, they rarely linked the scores to supervision and case management decisions. Studies in juvenile settings have not reported better results. Surveys of juvenile POs and judges indicated they tended to undervalue and, consequently, underuse results of risk instruments in case management, and many did not sustain the practice of using the tools at all (Shook & Saari, 2007). One study reported average adherence to the policy for simply administering a risk assessment was only 55% (Young, Moline, Farrell, & Biere, 2006). The goal of the current study was to examine whether integration of RNR principles, primarily the need principle, by juvenile POs (JPO) could be improved from past studies if probation offices underwent a comprehensive risk assessment implementation process.

Implementing the Need Principle

Very few studies have examined the need principle at the individual level. The majority of early RNR studies from the adult literature examined how groups of adult offenders fared after receiving services designed to address categories of criminogenic needs, in general, versus services that do not target any of these needs (e.g., Dowden & Andrews, 2000; Lowenkamp, Pealer, Smith, & Latessa, 2006). This approach differs from an individual approach whereby services are examined to determine whether they were a direct match to individual offenders’ specific criminogenic needs (or variable risk factors), commonly referred to as service-to-need match. The few service-to-need match studies conducted in juvenile settings indicate that service-to-need match has a strong association with later offending (Luong & Wormith, 2011; Peterson-Badali, Skilling, & Hoqanee, 2015; Vieira, Skilling, & Peterson-Badali, 2009). Vieira et al. (2009), for example, reported youth probationers who received services aligned with their criminogenic needs as identified by the Youth Level of Service/Case Management Inventory (YLS/CMI; Hoge & Andrews, 2006) reoffended at a rate of 25% versus 75% for youths who received services that did not match their needs, irrespective of youths’ risk levels.

Together, all studies of the need principle indicate it is an effective method for recidivism reduction. However, the need principle actually needs to be integrated into case management decisions for the benefit to be realized. Only a handful of studies can speak to the quality of implementation of service-to-need match in juvenile probation. In an archival study of youth probationers in two departments instructed to use the Saskatchewan Youth Edition of the Level of Service Inventory (LSI-SK; Andrews, Bonta, & Wormith, 2001), Luong and Wormith (2011) found youth workers (i.e., POs) were more likely to overrefer or underrefer youth within several need areas, rather than to match service referrals to need areas. For example, only 13% of youth scored moderate to high risk on the education scale, yet 68.8% were assigned an educational intervention. Vieira et al.’s (2009) youth probation sample had an average of only 35% of their criminogenic needs (as identified by the YLS/CMI) matched to an appropriate service. Peterson-Badali et al. (2015) reported that youth had an average of five needs identified on the YLS/CMI but an average of only 1.4 needs actually addressed during probation.

Studies have consistently demonstrated there is need for improvement in the implementation of the need principle. Some of the reasons identified for poor implementation of the need principle have included lack of service availability, the need to target high-impact needs first, and limitations in defining POs’ roles in case planning (Haqanee, Peterson-Badali, & Skilling, 2015). Other barriers, however, may stem from the quality of the implementation process for the risk assessment as a whole and lack of judge buy-in (Vincent, Guy, Perrault, & Gershenson, 2016).

Present Study

The overarching goal of the current study was to determine whether the process of service-to-need match may be improved, relative to prior studies, following a comprehensive risk assessment and RNR implementation process in juvenile probation offices. Conclusions that can be drawn from prior work about implementation of RNR principles may be limited in that most research examined service-to-need matching in probation offices that had been using a risk assessment instrument for quite some time prior to the study being conducted. Thus, it is unclear to what extent findings represented a drift in protocol or a lack of office policies and procedures and training regarding how to use risk assessment in case planning decisions altogether. In addition, studies of service-to-need match are somewhat difficult to interpret because it is not always feasible to address all of a youth’s criminogenic needs. POs that have been trained in RNR are instructed to take risk level into account and give more intervention to higher risk youth and less intervention to lower risk youth. Thus, criminogenic needs must be prioritized in the context of other youth characteristics, which generally has not been taken into account.

The current study used data from the Risk/Needs Assessment in Juvenile Probation: Implementation Study (RNAJP; Vincent et al., 2016; Vincent, Paiva-Salisbury, Cook, Guy, & Perrault, 2012) to examine the quality of service-to-need match following a structured implementation process for a valid risk assessment tool using an RNR approach. RNAJP was a multisite, pre- (before use of any risk assessment)/post- (after implementation of a valid risk assessment and RNR-related case planning policy) quasiexperimental study of six juvenile probation offices in two states. None of the probation offices were using a risk assessment prior to the study. As described in previous publications (see Vincent, Paiva-Salisbury, et al., 2012; Vincent et al., 2016), the implementation protocol involved, among other procedures, developing policies and new case plan forms so that risk assessment would be used to make service referrals based on youths’ criminogenic needs and overall risk levels. The RNAJP study found most juvenile probation offices had good adherence to their risk assessment administration policy and experienced significant shifts in dispositions, rates of placement, and levels of supervision in a manner that was consistent with the risk principle (Vincent et al., 2016). Qualitative interviews with POs at 3 months and 10 months postimplementation indicated most POs were using the risk assessment in their decisions about services, and their most common response to how they selected service referrals was “the fit between services and needs” (Vincent, Paiva-Salisbury, et al., 2012).

This study utilized data from Louisiana, one of the RNAJP states, which implemented the Structured Assessment of Violence Risk for Youth (SAVRY; Borum, Bartel, & Forth, 2006), to examine the quality of service-to-need matching relative to their RNR-related case planning policies. Good adherence to their policies would be present if (a) youth were referred to services that matched their individual criminogenic needs as indicated by the SAVRY, (b) youth were not referred to services designed to address criminogenic need areas that they did not have, (c) the number of services received and criminogenic need areas addressed were positively correlated with youths’ risk level, and (d) case plans prioritized need areas so they did not burden the youth and families with too many services at one time (a maximum of three services).

Method

Participants

This study utilized the sample of newly adjudicated youth (N = 452) from the three probation offices (Sites 1, 2, and 3) in Louisiana that adopted the SAVRY and participated in the RNAJP study. The sample of 452 youth was generated over the period February 2009 to October 2010, immediately after each site completed its full implementation process for the SAVRY and case planning. In all, 25% of the sample was adjudicated for a violent offense, with the most common being assault or threat with a firearm (20.6% of the sample). Aside from assault, common most serious index offenses were theft or break and enter (26%), adjudicated status offenses (19%), and minor miscellaneous offenses such as disturbing the peace (17%).

Because the goal of the current study was to see how well service referrals matched need areas identified by the SAVRY, cases were excluded if they did not receive the SAVRY assessment before or immediately following their probation disposition (n = 45), were in an out-of-home placement for the entire follow-up period (n = 6), or reoffended prior to a SAVRY being conducted, which may have shifted the manner in which services were managed (n = 16). Thus, the final sample consisted of 385 adolescent offenders (Site 1: n = 190; Site 2: n = 76; and Site 3: n = 119) who were primarily male (73%) and African American (79.2%), with only 19% being White, non-Hispanic. The average age was 15.2 years (SD = 1.6, range: 7-18 years).

Measures

SAVRY

Each office adopted a policy that JPOs were to assess youth with the SAVRY predisposition (or immediately postdisposition when completion of a predisposition report was not possible) and to reassess youth every 6 months or following a major life-changing event (Borum et al., 2006). Each SAVRY was to be approved by a supervisor. The SAVRY was designed to assess violence risk in adolescents aged 12 to 18 years; however, it also is a valid predictor of nonviolent offending (Olver, Stockdale, & Wormith, 2009). It comprises 6 items defining Protective Factors (which may lower the likelihood of risk) and 24 items defining Risk Factors (which increase the likelihood of risk). Risk Factors are rated as low, moderate, or high. Items are rated as “critical” if the evaluator sees it as strongly related to the youth’s offending and in need of immediate intervention. The 10 risk factor items comprising the Historical domain are static, and the remaining 14 risk items on the Individual/Clinical and Social/Contextual domains are dynamic.

The SAVRY uses a structured professional judgment (SPJ) approach (as opposed to a formulaic actuarial approach), meaning the final determination of an examinee’s overall level of risk for violence or reoffending is the examiner’s (in this case, PO’s) Summary Risk Rating (SRR; low, moderate, high risk) based on his or her professional judgment as informed by a systematic appraisal of the most relevant risk factors, including idiosyncratic factors noted by the evaluator. This assures that examiners assess risk factors that are empirically associated with violence or reoffending, consider the applicability of these factors to the specific examinee’s risk for reoffending (criminogenic needs), and classify the severity to make their final SRR. Thus, a prime advantage of the SAVRY is that it not only structures the process by which a valid decision is made about the likelihood that one will recidivate, but it also includes dynamic risk factors related to need areas that may be effective targets for treatment.

Meta-analyses have shown the SAVRY to have good predictive validity in a variety of young offender populations (average Areas under the Curve [AUC] of 0.71; Guy, 2008; Singh, Grann, & Fazel, 2011) that is comparable for both violent and nonviolent reoffending (Olver et al., 2009). For the current study, level of risk was measured according to the SAVRY SRR, the overall risk level ratings by POs. The field interrater reliability among the POs in the study sites was good for the SAVRY SRR (intra-class correlation coefficients [ICC₁] = .71) with ICCs ranging from acceptable (ICC₁ = .62) to excellent (ICC₁ = .84) for the majority of items (Vincent, Guy, Fusco, & Gershenson, 2012). The SAVRY SRRs for the current sample significantly predicted both violent and nonviolent reoffending over an average 18-month period (Vincent et al., 2016).

Service Referrals

Research assistants (RA; two master’s level graduate students and one PO) within each probation office recorded all service referrals in every youth’s case plans for a period up to 10 months following each youth’s adjudication. Services were defined as community-based services aimed at treatment or rehabilitation (e.g., mentoring programs, functional family therapy, and counseling). Interventions such as community service and electronic monitoring were not counted because these are sanctions rather than treatment-oriented approaches that meet criminogenic needs. Data recorded from youth’s files and case plans included the type of service referred (e.g., anger management), the start and end dates of participation, and the youth’s treatment status at the end of the tracking period (i.e., active, completed, and terminated unsuccessfully). By the end of the data tracking period, all but 1% of the services to which youth were referred had actually been attended at least once. These service referrals were coded as “pending,” meaning the youth had been assigned the referral, but had not yet attended the service.

Implementation Procedures

Researchers guided each probation office through a structured implementation process for the SAVRY and RNR-based case management practices, which had to be completed before any data collection was initiated. These procedures, which have been described in detail elsewhere (see Vincent, Guy, & Grisso, 2012; Vincent, Paiva-Salisbury, et al., 2012), included a detailed risk assessment policy including how to make disposition recommendations, assign supervision level, and match needs to services in case planning. Administrators revised the office’s case plan template to be organized according to criminogenic need areas identified by the SAVRY (e.g., substance abuse, family circumstances). Each office completed a service matrix that listed each of the services available in their communities according to the criminogenic need areas addressed and the intensity of the service.

POs completed a 2-day workshop on the SAVRY, a minimum of three additional practice cases, and a half-day training on RNR. The half-day training included (a) RNR principles and research evidence; (b) the new office policy regarding how to use the SAVRY in case processing recommendations that follow RNR principles, including service-to-need matching; (c) how to identify and prioritize need areas from the SAVRY; and (d) use of the service matrix and case plan form. POs were instructed to (a) follow the risk principle, such that the number and intensity of services should be positively related to risk level, (b) prioritize criminogenic needs so youth are not expected to receive more than three services at any given time (translated into three needs or less), and (c) give few to no services to low-risk youth. Needs were to be prioritized according to those with the most risk factors rated high or critical. Offices also received booster training on the SAVRY and service-to-need matching every 6 months.

Criminogenic Need Areas

Unlike most risk assessments, the SAVRY is not constructed according to scales that define homogeneous need areas. Thus, the researchers developed six need areas to be as consistent as possible with those described in the RNR model. The POs were instructed to fill out a need areas worksheet with each SAVRY. Consistent with the SPJ approach, rather than “score” need areas, POs were trained to consider youths’ ratings on the items within each of these need areas to identify the areas of greatest concern. The need areas were disruptive behavior problems (a combination of personality problems/behaviors and antisocial attitudes), substance abuse, family/parenting problems, education, negative peer relations, and emotional stability concerns. Need areas comprised mainly dynamic risk factors, but some static risk factors were included when marked as critical, which would mean the POs thought the factor continued to be strongly related to the youths’ offending. Table 1 provides the need areas and Cronbach’s alphas derived from a database of 4,705 SAVRYs conducted by POs in Louisiana (α’s ranging from .50 for scales with few items up to .82).

Table 1:

SAVRY Criminogenic Need Areas

Need area and SAVRY items
Disruptive behavior problems 17. Negative attitudes 18. Risk taking/impulsivity 20. Anger management problems 21. Low empathy/remorse 23. Poor compliance
Emotional stability concerns 5. Self-harm or suicide attempts (if critical) 13. Stress and poor coping 20. Anger management 22. Attention deficit/hyperactivity disorder
Substance abuse 19. Substance abuse difficulties
Family/parenting problems 6. Exposure to violence in the home (if critical) 7. Childhood history of maltreatment (if critical) 8. Parental/caregiver criminality (if critical) 14. Poor parental management
Education 10. Poor school achievement 24. Low interest/commitment to school
Negative peer relations 11. Peer delinquency 12. Peer rejection 15. Lack of personal/social support

Note. SAVRY = Structured Assessment of Violence Risk for Youth.

For the purpose of this study, researchers summed items into a “need score” within each need area by assigning a value of 0, 1, or 2 for low, moderate, or high ratings on each SAVRY risk item. The scores separating the tertiles within each range of scores from the development sample of 4,705 cases were considered the cut-off scores for high, moderate, or low needs. The researchers coded criminogenic need areas across SAVRY reassessments and calculated an overall criminogenic need present/absent variable for each youth. A need area was considered “present” if youth had moderate or high scores in the need area domain for any of their SAVRY assessments (ranging from 1 to 3 assessments per youth) and it was “absent” if youth scored low across all assessments.

Service-to-Need Match

The authors created a codebook to categorize services according to which need area(s) the service addressed (e.g., family counseling met the family/parenting problems need area), based on the service matrix created for each probation office. The codebook also indicated how to code matches for services that addressed more than one need area (e.g., Multisystemic Therapy). Three master’s level RA’s trained on the SAVRY and codebook conducted the service-to-need match coding.

Service-to-need match was coded for each SAVRY need area for each youth. A “good match” was when youth had a need area identified as present by the SAVRY (i.e., high or moderate need score) and at least one service referral assigned to address that need area, or when youth did not receive services for need areas that were not present. A “bad match” was defined as (a) “under-prescription” of service referrals when a youth did not receive a service aligned with a need area scored as present, and (b) “over-prescription” when POs assigned at least one service to a youth targeting a need area that was scored as absent. There was high interrater agreement on the rating scale, with ICC₁ values for each need area falling in the excellent range (.92 or higher) based on 30 randomly selected cases.

Data Analyses

Analyses were performed to examine adherence to the need principle, or service-to-need matching approach. Quality of service-to-need match was operationally defined as a good match or a bad match (due to over- or underprescription) for each need area for each youth. The authors conducted ANOVA or chi-square analyses to examine differences in the quality of service-to-need match by risk level and criminogenic need areas. Risk level was categorized as Low versus Moderate to High risk to be consistent with past research (e.g., Luong & Wormith, 2011; Vitopoulos, Peterson-Badali, & Skilling, 2012).

Results

Risk, Criminogenic Needs, and Service Referrals

Risk Level and Criminogenic Need Areas

The SAVRY SRR identified 40.3% (n = 153) of youth as low risk, 46.3% (n = 176) as moderate risk, and only 13.4% (n = 51) as high risk. Combining moderate to high risk youth into one category resulted in a group of 227 youth (59.7%). Five youth (1.3%) were missing the SAVRY SRR. With respect to need areas, youth ranged from having 0 to 6 SAVRY need areas calculated as present (i.e., moderate or high rating) with an average of 3.6 needs (SD = 1.8). The most common need areas overall were negative peer relations (78.7%), disruptive behavior problems (65.7%), education (66%), and emotional stability concerns (59.5%; see Table 2).

Table 2:

Appropriateness of Service Referral by SAVRY Identified Need Area (N = 385)

Need area	No need,n (%)	Need present,n (%)	Good service match, n (%)	Overprescription,n (%)	Underprescription, n (%)
Disruptive behavior problems	132 (34.3)	253 (65.7)	179 (46)	39 (29)	167 (66)
Emotional stability concerns	156 (40.5)	229 (59.5)	193 (50.1)	29 (19)	163 (71.1)
Substance abuse	228 (59.2)	157 (40.8)	286 (74.2)	14 (6.1)	85 (54.1)
Family/parenting problems	208 (54)	177 (46)	241 (63)	37 (16)	107 (60.4)
Education	131 (34)	254 (66)	159 (41.2)	30 (23)	196 (77.1)
Negative peer relations	82 (21.3)	303 (78.7)	148 (38.4)	37 (45)	200 (66)

Note. Percentages for columns were calculated using different denominators. A criminogenic need area being present or absent, and the rates of good service matching are reflected in the percentage of youth out of the entire sample (N = 385). The percentage of youth with overprescription includes only youth who had the need area absent, whereas underprescription is the percent of youth who had the need area present. SAVRY = Structured Assessment of Violence Risk for Youth.

Service Referrals

The number of services for each youth ranged from 0 to 11 (M = 2.1, SD = 1.6). JPOs made a total of 758 service referrals for the 385 youth in the sample; however, 55 service referrals (1%) could not be coded due to missing information about the type of service. Of the remaining 703 services, referrals addressing the emotional stability need area were most frequently assigned (n = 144, 19%). In general, service referrals were fairly evenly distributed across most other need areas: substance abuse (n = 126, 16.6%), negative peer relations (n = 114, 15%), family/parenting problems (n = 108, 14.2%), and disruptive behavior problems (n = 93, 12.3%). Education-related services (which also included employment) was the only need area addressed fairly infrequently (5.5%). The final 10% of service referrals would be categorized as responsivity factors (e.g., independent living) or were services that met multiple need areas.

Was There Adherence to Risk and Need Principles in Case Planning?

Risk Principle

In accordance with the risk principle, probation departments were instructed to align the number of services provided with youth’s risk level. There was a significant difference in the number of service referrals given by risk level, F(2, 377) = 2.89, p = .05, [1.88, 2.2] η_p² = .02, such that high-risk youth received significantly more services (M = 2.5, SD = 1.7) than moderate- (M = 2.0, SD = 1.5) and low-risk youth (M = 1.9, SD = 1.7). Low-risk youth did not receive significantly less services than moderate-risk youth.

Number of Need Areas Addressed

POs were instructed to prioritize a maximum of three criminogenic need areas for each youth, commensurate with the youth’s risk level (e.g., lower risk youth get less). To investigate adherence to this policy, researchers examined the total number of need areas addressed based on the first SAVRY assessment. On average, youth had 1.7 (SD = 1.4) need areas addressed with at least one service, and the majority of the sample had three or less need areas addressed at any one time (n = 330, 86%) following their initial SAVRY. However, as expected, there was a significant difference by risk level such that low-risk youth had fewer need areas addressed (M = 1.5, SD = 1.4), on average, than high/moderate-risk youth (M = 1.8, SD = 1.5; t(378) = −2.03, [−2.51, −1.87] p = .04, d = .20). The majority of low-risk youth had one need or less addressed by a service (63%).

Service-to-Need Matching

Table 2 provides the percentages of youth identified as having each need area present along with the percentage receiving a good match (need area present and service received, or need area absent and service not received) versus an over- or underprescription of services. Youth identified as having a substance use (n = 157) or family/parenting problems (n = 177) need were the most likely to receive a good match (see Table 2). The need area with the most overprescription of services was negative peer relations (e.g., boys/girls groups, social skills training) with 45% of youth identified as low in this area receiving a service. When services were overprescribed, on average across all need areas, youth were referred to one or two more services than needed. Overprescription was particularly high for the disruptive behavior problems, emotional stability, and family/parenting need areas such that youth who did not appear to have these needs received up to three related services.

In general, there was considerably more underprescription than overprescription, with 50% or more youth needing a service they did not receive. The most commonly underprescribed need areas were education (77.1%) and emotional stability (71.1%). A total of 28 youth did not have a single need area present, but all of them were still given an average of one service (SD = 0.8, range: 1 to 3), which were most commonly related to negative peers (n = 11, 39.3%) or disruptive behavior (n = 9, 32.1%).

Service-to-Need Matching and Risk Level

One limitation in the aforementioned analyses of quality of service-to-need match as a whole is that the analyses do not consider (a) differences in youths’ risk levels, and (b) severity of the needs when moderate and high scores within each need area are combined into one “present” category. Rates of underprescription would be inflated in these analyses if some youth were low risk and had only moderate criminogenic needs that may not require an intervention to decrease the likelihood of reoffending. Thus, we would be less concerned about under- than overprescription for low-risk youth, but would be concerned about both under- and overprescription for higher risk youth.

To deconstruct the service-to-need match findings further, the researchers conducted chi square tests separately by risk level (see Table 3). Table 3 indicates low-risk youth had higher rates of underprescription than overprescription. Table 4 displays the severity of needs (according to need area scores) by risk level. As Table 4 indicates, among low-risk youth who had a need, most scored in the moderate level with the exception of the disruptive behavior and emotional stability need areas. The moderate- to high-risk youth also had significantly more underprescription than overprescription in most need areas (see Table 3); however, unlike the low-risk youth, Table 4 indicates a majority of these youth were scoring high in several of these need areas for which they did not receive services.

Table 3:

Appropriateness of Service Referral by SAVRY Identified Need Area by Risk Level (n = 380)

Need area	No need, n (%)	Need present, n (%)	Good service match, n (%)	Overprescription, n (%)	Underprescription, n (%)
Low Risk^a
Disruptive behavior problems	102 (66.7)	51 (33.3)	90 (58.8)	28 (27.5)	35 (68.6)
Emotional stability concerns	108 (70.6)	45 (29.4)	98 (64)	21 (19.4)	34 (75.6)
Substance abuse	107 (70)	46 (30)	125 (81.6)	7 (6.5)	21 (45.6)
Family/parenting problems	105 (68.6)	48 (31.4)	101 (66)	20 (19)	32 (66.7)
Education	89 (58.2)	64 (41.8)	79 (51.6)	20 (22.4)	54 (84.3)
Negative peer relations	66 (43)	87 (57)	62 (40.5)	29 (44)	62 (71.2)
Mod/High Risk^b
Disruptive behavior problems	30 (13.2)	197 (86.8)	87 (38.3)	11 (36.7)	129 (65.4)
Emotional stability concerns	47 (20.7)	180 (79.3)	92 (40.5)	8 (17)	127 (70.5)
Substance abuse	120 (52.9)	107 (47.1)	159 (70)	7 (5.8)	61 (57)
Family/parenting problems	102 (45)	125 (55)	136 (60)	17 (16.6)	74 (59.2)
Education	42 (18.5)	185 (81.5)	79 (34.8)	10 (23.8)	138 (74.5)
Negative peer relations	16 (7)	211 (93)	83 (36.5)	8 (50)	136 (64.4)

Note. SAVRY = Structured Assessment of Violence Risk for Youth.

n = 153. ^b n = 227.

Table 4:

Severity of Criminogenic Need Area by Risk Level (N = 380)

Need area	No need,n (%)	Need present, n (%)
Need area	No need,n (%)	Moderate level	High level
Low Risk^a
Disruptive behavior problems	102 (66.7)	25 (16.3)	26 (17)
Emotional stability concerns	108 (70.6)	19 (12.4)	26 (17)
Substance abuse	107 (70)	28 (18.3)	18 (11.8)
Family/parenting problems	105 (68.6)	33 (21.6)	15 (9.8)
Education	89 (58.2)	35 (23)	29 (19)
Negative peer relations	66 (43.1)	70 (45.8)	17 (11.1)
Mod/High Risk^b
Disruptive behavior problems	30 (13.2)	39 (17.2)	158 (70)
Emotional stability concerns	47 (20.7)	34 (15)	146 (64.3)
Substance abuse	120 (53)	54 (24)	53 (24)
Family/parenting problems	104 (45.8)	63 (27.8)	60 (26.4)
Education	42 (18.5)	83 (36.6)	102 (45)
Negative peer relations	16 (7)	123 (54.2)	88 (38.8)

n = 153. ^bn = 227.

When Did the Quality of Service-to-Need Matching Differ?

The results indicate that the service-to-need match process was not well implemented, with less than 50% of needs being addressed by a service within each criminogenic need category. Thus, we conducted ad hoc analyses to examine whether the ostensibly poor results might be explained by site-level differences in implementation quality, such that poor implementation of service-to-need matching in one site might be watering down good implementation in the other sites. Examination of site-level differences indicated the quality of service-to-need matching was best in Site 1, which had the highest rates of good matches across all need areas (see Table 5). However, Site 1 also was the most likely site to overprescribe services in disruptive behavior problems, negative peer relations, and education. Site 3 had a lower quality service-to-need matching with the highest rates of underprescription in most areas.

Table 5:

Probation Office Differences in Quality of Match for Each Need Area by Site (N = 385)

Need area	Good service match,n (%)	Bad service match, n (%)		χ ²	V
Need area	Good service match,n (%)	Overprescription	Underprescription	χ ²	V
Disruptive behavior problems
Site 1	92 (48.4)	31 (16.3)	67 (35.3)	1.38	0.06
Site 2	37 (48.7)	5 (6.6)	34 (44.7)
Site 3	50 (42)	3 (2.5)	66 (55.5)
Emotional stability concerns
Site 1	111 (58.4)	10 (5.3)	69 (36.3)	15.89*	0.20
Site 2	40 (52.6)	17 (22.4)	19 (25)
Site 3	42 (35.3)	2 (1.7)	75 (63)
Substance abuse
Site 1	150 (79)	8 (4.2)	32 (16.8)	4.71	0.11
Site 2	55 (72.4)	1 (1.3)	20 (26.3)
Site 3	81 (68.1)	5 (4.2)	33 (27.7)
Family/parenting problems
Site 1	128 (67.4)	13 (6.8)	49 (25.8)	15.97*	0.20
Site 2	37 (48.7)	11 (14.5)	28 (36.8)
Site 3	76 (64)	13 (11)	30 (25.2)
Education
Site 1	95 (50)	20 (10.5)	75 (39.5)	16.14*	0.21
Site 2	32 (42.1)	7 (9.2)	37 (48.7)
Site 3	32 (27)	3 (2.5)	84 (70.6)
Negative peer relations
Site 1	102 (53.7)	30 (15.8)	58 (30.5)	36.93*	0.31
Site 2	19 (25)	3 (4)	54 (71.1)
Site 3	27 (22.7)	4 (3.4)	88 (74)

Note. Sample sizes vary by site; Site 1 (n = 190), Site 2 (n = 76), Site 3 (n = 119).

p ⩽ .001.

Site 2 had the highest percentage of youth identified by the SAVRY SRR as low risk (60%) compared with Sites 1 (38.1%) and 3 (31.3%; χ²(2) = 18.13, p ⩽ .001, V = .15); however, their sample did not have significantly fewer need areas, on average (M = 3.3, SD = 1.9), than youth in Site 1 (M = 3.1, SD = 1.9). Despite the high rate of low-risk youth, Site 2 gave all their youth significantly more services: M = 2.5, SD = 2.4, range: 0-11; than the other sites: Site 1 (M = 1.9, SD = 1.0, range: 0-5) and Site 3: (M = 2.0, SD = 1.7, range: 0-9); F(2, 382) = 4.65, p = .01, η_p² = .02. Moreover, Site 2 was the only site to not have a significant association between the number of services assigned and youth’s risk level. POs, on average, assigned roughly the same number of services to low-risk youth (M = 2.5, SD = 2.7), moderate-risk youth (M = 2.8, SD = 2.4), and high-risk youth (M = 2.4, SD = 1.2). A total of 20% of the low-risk youth from Site 2 received four or more services; whereas the other sites gave low-risk youth: Site 1 (M = 1.8, SD = 1.0) and Site 3 (M = 1.6, SD = 1.0) and moderate-risk youth: Site 1 (M = 1.9, SD = 1.0) and Site 3 (M = 1.9, SD = 1.7) less services than high-risk youth: Site 1 (M = 2.3, SD = 1.0) and Site 3 (M = 3.1, SD = 2.5).

Discussion

Several studies of juvenile probation samples have been conducted to examine the quality of service-to-need matching (e.g., Luong & Wormith, 2011; Peterson-Badali et al., 2015; Vieira et al., 2009). These studies have reported some significant limitations, indicating that less than half of youth appear to be getting the services they need. The current study adds to the literature by examining the quality of service-to-need matching following a rigorous implementation process of a risk assessment and RNR-related policies in case planning. The analyses in this study also examined adherence to the need principle by attempting to account for the fact that not every criminogenic need can feasibly be addressed in a case plan. Real-world factors, such as prioritizing needs based on their severity, limiting the number of concurrently received services, and creating plans that are realistic and attainable for the youth and family are all important to consider in our examinations of the quality of the matching process. Of equal importance is the need to investigate whether service referrals were allocated consistent with the risk principle as it is important to limit the amount of intervention given to lower risk youth.

Did Probation Departments Adhere to the Case Planning Protocol?

This study found some support for adherence to RNR principles among probation departments when assigning treatment-related services to young offenders. Probation departments generally followed the risk principle by assigning more interventions to higher risk youth than to lower risk youth, albeit the effect sizes were small so there is room for improvement. They also followed their case planning protocol with respect to limiting the number of criminogenic need areas that were targeted for intervention at any one time. Less than 15% of youth received interventions designed to address more than three criminogenic need areas in tandem, and low-risk youth, on average, were expected to address fewer needs than higher risk youth. The impetus for limiting the number of criminogenic need areas to target comes from the notion that behavior change is difficult. Moreover, any required attendance in interventions must consider that youth also are expected to fulfill responsibilities in other areas, such as satisfying their educational requirements, maintaining any employment commitments, or meeting family obligations such as caring for siblings. Furthermore, youth are often dependent on their parents or public transportation (if available) to attend appointments. This can create a significant burden for families and working parents, especially single parents. In consideration of these issues, a maximum of three criminogenic needs (which translated to three services) was the rule of thumb selected by these offices and the researchers assisting the risk assessment implementation effort. However, the ideal number of needs and services to target in a supervision plan is an area for further study. The goal is to target as many criminogenic needs as possible while still setting realistic expectations.

The case planning protocol was not followed well for low-risk youth. Overall, low-risk youth were expected to attend an average of just under two service interventions and averaged 1.6 criminogenic need areas addressed. These averages were even higher in one probation office (Site 2) that served an unusually high rate of low-risk youth and clearly did not differentiate the number of service referrals by risk level. The problem of overservicing low-risk youth was demonstrated in a rigorous analysis of data from multiple juvenile probation offices in Texas where low-risk youth were still attending one or two programs, often with higher risk youth, and were more likely to be rearrested than low-risk youth not attending programs (Fabelo, Arrigona, Thompson, Clemens, & Marchbanks, 2015).

Did Probation Departments Refer Youth to Services That Matched Their Needs?

Probation departments performed better in some aspects of service-to-need matching than others. Examination of specific criminogenic need areas indicated relatively good matching occurred in the substance abuse area; almost 75% of youth had this need matched appropriately. Similarly, the family/parenting problem area had good rates of matching, where 63% of youth were appropriately matched to a service. The better match in these criminogenic need areas is somewhat consistent with Luong and Wormith (2011) and is not terribly surprising because these need areas are some of the most tangible. Generally, probation offices were better at the matching process when needs were absent (meaning it not be appropriate to give a service), as demonstrated by significantly higher rates of underprescription compared with overprescription. Unfortunately, 50% or less of youth having criminogenic need areas other than substance use or family/parenting were underprescribed services.

The most disconcerting of these was disruptive behavioral problems (similar to Antisocial Personality within the RNR framework), where less than half of youth had the need appropriately addressed, 30% received a service they should not have, and more than 65% did not receive a service they needed. This area has been reported as having one of the worst matches in other studies (Luong & Wormith, 2011; Peterson-Badali et al., 2015) and, unfortunately, appears to have the strongest association with later recidivism (Perrault, Vincent, & Guy, 2017; Peterson-Badali et al., 2015). Interpretation of this result is complicated in all of the studies with probation samples to date. The current and prior studies examined only community-based services, which may be lacking in the ability to address this particularly recalcitrant need area. It is likely probation offices expect personality/behavior problems to be handled in residential treatment or correctional facilities, which have not been included in most service-to-need matches. Future research should code services received both within and outside of out-of-home placements.

On the plus side, at least 50% of the underprescription of services identified in this study stemmed from low-risk youth not receiving services. This is a positive result because most of these needs were only moderate in level, and probation offices were trained to give less intervention to low-risk youth. Despite the larger percentages of underprescription for low-risk youth, relative to higher risk youth, low-risk youth were still overserviced across the board. The most striking service-to-need mismatch for low-risk youth was the overuse of peer-oriented services. This was likely a result of probation offices wanting to provide something rather than nothing; however, these services may be particularly problematic as they increase the likelihood of exposure to higher risk youth and, therefore, may increase peer contagion effects that are positively correlated with later delinquency (Dishion & Tipsord, 2011).

There are a couple potential explanations for the overservicing of low-risk youth. One is that the system is simply uncomfortable supervising youth without giving them an intervention, and they continue to believe that something is better than nothing, despite research indicating the opposite (Fabelo et al., 2015). Another reason may be that a risk assessment does not tell us about the mental health or other noncriminogenic needs of youth, some of which may have been common among low-risk youth given their most prevalent need area was emotional stability concerns. When low-risk youth in this study were seemingly being overprescribed mental health-related services, they were receiving multiple services in this area. It may be that other assessments indicated these interventions were necessary. Finally, it may be the case that low risk is not equal to “no risk” (Taxman & Caudy, 2015); however, research is needed to provide guidance to juvenile justice agencies regarding the line for what is harmful versus helpful for these youth.

Implications

A primary interest of this study was to examine whether undergoing a comprehensive implementation procedure for a valid risk assessment and RNR-related case planning improved service-to-need matching. Clearly, because the research design involved the implementation protocol and the SAVRY being implemented at the same time, there was not a comparison group that enabled us to determine whether the training on RNR and the new case planning policies improved service-to-need matching beyond training on the SAVRY alone. It is also difficult to make any direct comparisons between the results and those from other studies in light of differences in the specific risk assessment instrument used and in the operational definitions of “good match” across studies. However, compared with Peterson-Badali et al. (2015), who reported youth had an average of five needs with only 1.4 needs addressed (about 25%), the current results appear to be an improvement, with youth having an average of 3.6 needs and 1.7 addressed (almost half).

Some of the strengths of the implementation protocol that should be examined in other studies and considered by other probation offices were the RNR-related policies and the use of a service matrix to standardize probation officers’ selection of services to address specific need areas. There were clear site-level differences in the quality of service-to-need matching, some of which were due to service availability, a known barrier to service-to-need matching (Haqanee et al., 2015). Site 1 had the highest rates of overprescription largely because they had the most services available in their area; Site 3 had the exact opposite experience due to a lack of available interventions. Site 2 did not appear to follow their policies well at all. Evidently, implementation of best practices is challenging and requires ongoing supervisory oversight of case plans, continual retraining of probation officers, and staff incentives for good case planning,

Limitations and Future Directions

There are a few limitations that may have impacted the findings of this study. One limitation is that the data were gathered for this study only a few months after the SAVRY had been implemented. It can take 3 years for a new risk assessment procedure to be implemented well enough to show results (Flores, Lowenkamp, Holsinger, & Latessa, 2006). Therefore, it is possible that POs became more adept at assigning services to needs over time. A second issue is that this was the first attempt at examination of service-to-need matching using a structured professional judgment tool such as the SAVRY, which does not have a clear method for calculating need areas. POs were trained in a structured method to identify needs. However, it may have simply been too difficult, or led to decisions that were too idiosyncratic. Without a comparison group using an actuarial tool, it is unknown if, and to what extent, this may have interfered with their ability to follow the need principle. Moreover, a better method for quantifying whether need areas were present when using an SPJ tool would have been to have POs document the need areas they identified using a combination of item ratings and their judgment, as they were trained to do. The researchers were limited to using a scoring system to identify the moderate- or high-need areas for each youth. Although the scoring approach should have corresponded to POs’ selected need areas, this is not how the tool would be used in practice. A quantitative method to identification of needs has some limitations. For example, if one is rating the SAVRY correctly, a youth who had a substance use problem in the past but is currently sober and does not have a need for treatment would still be scored as moderate, and, therefore, identified as having a substance abuse need. Future research examining implementation of the SAVRY should have POs record the need areas they identified and test whether there are better methods for combining items.

Another limitation of the study was the manner in which we defined “services” and the dependency on probation records. The services tracked in this study were formal interventions received in the community. We did not include programming received by youth when they were sent to out-of-home placements or correctional facilities where, conceivably, disruptive behavioral problems were most likely to be addressed. The high rates of underprescription reported for education services may be due to limited documentation because educational services often are not paid for by the probation departments. Our rates of matching in the education area were in direct contrast to the overservicing reported by Luong and Wormith (2011). Other service-to-need matching studies have included interventions performed by POs, such as cognitive tasks to change youths’ criminal thinking (Peterson-Badali et al., 2015). We did not include PO-related interventions in this study aside from in-house classes led by POs (e.g., victim awareness). POs in the study sites were not trained to provide other interventions, and it was not part of the probation culture to do so. Last, it is important to recognize that POs do not always have complete control over what services youth are expected to attend. In some jurisdictions, these determinations are made in the court’s probation orders, which the probation officers must follow. Fortunately, court orders made with the benefit of a predisposition report and recommendations based on the SAVRY occurred in at least half of the cases in this sample, and most of the judges in these jurisdictions let POs make the service decisions. However, it is possible some of the services received in this study were ordered by the judge based on their own impressions of the case. Any attempts at adopting an RNA and RNR-related case planning should make a concerted effort to train judges and attorneys on these concepts and obtain their buy-in.

Conclusion

Effective and appropriate treatment is critical to efforts for rehabilitating youthful offenders in hopes of reducing their likelihood of future offending. Youth with fewer criminogenic needs addressed with appropriate service referrals have been shown to reoffend at fast rates with a greater number of convictions (Vieira et al., 2009). Our results generally align with previous findings of the mixed results on the quality of service-to-need matching; some criminogenic need areas appear easier to address with services (e.g., substance abuse), while others seem to result in the overprescription (e.g., negative peer relations) and underprescription (e.g., disruptive behavior problems) of treatment services. The present research suggests the need for further efforts to ensure better matching of services to youths’ criminogenic needs. Continued focus on the need principle within the RNR framework and the advancement of service-to-need matching through empirical research is imperative to improve the policies and practices within the juvenile justice system.

Footnotes

Authors’ Note:

The authors wish to acknowledge the assistance of Scarlett Woods, MA, and Geno Salomone, MA, for their dedicated assistance with data coding. This research was funded by the John D. and Catherine T. MacArthur Foundation as part of the Models for Change Research Network (Principal Investigator: G.M.V.).

ORCID iD

Rebecca J. Nelson

Rebecca J. Nelson, MA, is a PhD student in the Clinical Psychology Program at the University of Rhode Island. Her clinical and research interests center on better understanding unique aspects of the criminal justice system, such as treatment services to reduce reoffending, multicultural considerations, substance use problems, aggression and intimate partner violence, and motivation for and compliance in mental health treatment, for both adult and juvenile offenders, with the aim of strengthening system policies and procedures.

Gina M. Vincent, PhD, is an associate professor, codirector of the Law & Psychiatry Program, and director of Translational Law and Psychiatry Research in the Department of Psychiatry at the University of Massachusetts Medical School. She has received funding from the National Institutes of Health (NIMH), National Institute on Drug Abuse (NIDA), the MacArthur Foundation, and Office of Juvenile Justice and Delinquency Prevention (OJJDP) for studies relevant to risk for reoffending, callous-unemotional traits, and mental health and substance use problems among youth involved in the juvenile justice system.

References

Andrews

D. A.

Bonta

(2010). Rehabilitating criminal justice policy and practice. Psychology, Public Policy, and Law, 16, 39-55.

Andrews

D. A.

Bonta

(2017). The psychology of criminal conduct (6th ed.). Abingdon, UK: Taylor & Francis.

Andrews

D. A.

Bonta

Hoge

(1990). Classification for effective rehabilitation: Rediscovering psychology. Criminal Justice and Behavior, 17, 19-52. doi:10.1177/0093854890017001004

Andrews

D. A.

Bonta

Wormith

J. S.

(2001). Level of Service Inventory–Saskatchewan youth edition. Toronto, Ontario, Canada: Multi-Health Systems.

Borum

Bartel

Forth

(2006). Structured Assessment of Violence Risk in Youth (SAVRY). Odessa, FL: Psychological Assessment Resources.

Dishion

T. J.

Tipsord

J. M.

(2011). Peer contagion in child and adolescent social and emotional development. Annual Review of Psychology, 62, 189-214. doi:10.1146/annurev.psych.093008.100412

Dowden

Andrews

D. A.

(2000). Effective correctional treatment and violent reoffending: A meta-analysis. Canadian Journal of Criminology, 42, 449-467.

Fabelo

Arrigona

Thompson

M. D.

Clemens

Marchbanks

M. P.

(2015). Closer to home: An analysis of the state and local impact of the Texas juvenile justice reforms. New York, NY: Council of State Governments Justice Center.

Flores

A. W.

Lowenkamp

C. T.

Holsinger

A. M.

Latessa

E. J.

(2006). Predicting outcome with the Level of Service Inventory-Revised: The importance of implementation integrity. Journal of Criminal Justice, 34, 523-529. doi:10.1016/j.jcrimjus.2006.09.007

10.

Flores

A. W.

Travis

L. F.

Latessa

E. J.

(2004). Case classification for juvenile corrections: An assessment of the Youth Level of Service/Case Management Inventory (YLS/CMI), final report. Washington, DC: National Institute of Justice.

11.

Grisso

(2017). Policy brief: Assuring the future of developmental reform in juvenile justice: Recommendations of the fourth wave forecasting project. Retrieved from http://www.nysap.us/Products.html#FourthWave

12.

Guy

L. S.

(2008). Performance indicators of the structured professional judgment approach to assessing risk for violence to others: A meta-analytic survey (Unpublished doctoral thesis). Simon Fraser University, Vancouver, British Columbia, Canada.

13.

Haas

S. M.

DeTardo-Bora

K. A.

(2009). Inmate reentry and the utility of the LSI-R in case planning. Corrections Compendium, 34, 11-16, 49-52.

14.

Haqanee

Peterson-Badali

Skilling

(2015). Making “what works” work: Examining probation officers’ experiences addressing the criminogenic needs of juvenile offenders. Journal of Offender Rehabilitation, 54(1), 37-59. doi:10.1080/10509674.2014.980485

15.

Hoge

R. D.

Andrews

D. A.

(2006). Youth Level of Service/Case Management Inventory: User’s manual. North Tonawanda, NY: Multi-Health Systems.

16.

Juvenile Justice and Delinquency Prevention Act of 2002, the 21st Century Department of Justice Appropriation Authorization Act, Pub. L. No. 107-273, 116 Stat. 1758.

17.

Lowenkamp

C. T.

Pealer

Smith

Latessa

E. J.

(2006). Adhering to the risk and need principles: Does it matter for supervision based programs? Federal Probation, 70(3), 3-8.

18.

Luong

Wormith

(2011). Applying risk/need assessment to probation practice and its impact on the recidivism of young offenders. Criminal Justice and Behavior, 38, 1177-1199. doi:10.1177/0093854811421596

19.

Miller

Maloney

(2013). Practitioner compliance with risk/needs assessment tools: A theoretical and empirical assessment. Criminal Justice and Behavior, 40, 716-736. doi:10.1177/0093854812468883

20.

Monahan

Skeem

(2014). The evolution of violence risk assessment. CNS Spectrums, 19, 419-424. doi:10.1017/S1092852914000145

21.

National Resource Council. (2013). Reforming juvenile justice: A developmental approach. Washington, DC: The National Academics Press. doi:10.17226/14685

22.

Olver

Stockdale

Wormith

(2009). Risk assessment with young offenders: A meta-analysis of three assessment measures. Criminal Justice and Behavior, 36, 329-353. doi:10.1177/0093854809331457

23.

Perrault

R. T.

Vincent

G. M.

Guy

L. S.

(2017). Are risk assessments racially biased? Field study of the SAVRY and YLS/CMI in probation. Psychological Assessment, 29, 664-678.

24.

Peterson-Badali

Skilling

Hoqanee

(2015). Implementation of risk assessment in case management for youth in the justice system. Criminal Justice and Behavior, 42, 304-320. doi:10.1177/0093854814549595

25.

Seigle

Walsh

Weber

(2014). Core principles for reducing recidivism and improving other outcomes for youth in the juvenile justice system. New York, NY: Council of State Governments Justice Center.

26.

Shook

J. L.

Saari

R. C.

(2007). Structured decision making in juvenile justice: Judges’ and probation officers’ perceptions and use. Children and Youth Services Review, 29, 1335-1351. doi:10.1016/j.childyouth.2007.05.008

27.

Singh

J. P.

Grann

Fazel

(2011). A comparative study of violence risk assessment tools: A systematic review and metaregression analysis of 68 studies involving 25,980 participants. Clinical Psychology Review, 31, 499-513. doi:10.1016/j.cpr.2010.11.009

28.

Taxman

F. S.

Caudy

M. S.

(2015). Risk tells us who, but not what or how: Empirical assessment of the complexity of criminogenic needs to inform correctional programming. Criminology & Public Policy, 14, 71-103. doi:10.1111/1745-9133.12116

29.

Vieira

Skilling

Peterson-Badali

(2009). Matching court-ordered services with youths’ treatment needs: Predicting treatment success with young offenders. Criminal Justice and Behavior, 36, 385-401. doi:10.1177/0093854808331249

30.

Viglione

Rudes

D. S.

Taxman

F. S.

(2014). Misalignment in supervision: Implementing risk/needs assessment instrument in probation. Criminal Justice and Behavior, 42, 263-285. doi:10.1177/0093854814548447

31.

Vincent

G. M.

Guy

L. S.

Fusco

S. L.

Gershenson

B. G.

(2012). Field reliability of the SAVRY with probation officers: Implications for training. Law and Human Behavior, 36, 225-236. doi:10.1007/s10979-011-9284-2

32.

Vincent

G. M.

Guy

L. S.

Grisso

(2012). Risk assessment in juvenile justice: A guidebook for implementation. Retrieved from http://www.nysap.us/Risk%20Guidebook.pdf

33.

Vincent

G. M.

Guy

L. S.

Perrault

R. T.

Gershenson

(2016). Risk assessment matters, but only when implemented well: A multisite study in juvenile probation. Law and Human Behavior, 40, 683-696. doi:10.1037/lhb0000214

34.

Vincent

G. M.

Paiva-Salisbury

Cook

N. E.

Guy

L. S.

Perrault

R. T.

(2012). Impact of risk/needs assessment on juvenile probation officers’ decision-making: Importance of implementation. Psychology, Public Policy, and Law, 18, 549-576. doi:10.1037/a0027186

35.

Vitopoulos

N. A.

Peterson-Badali

Skilling

T. A.

(2012). The relationship between matching service to criminogenic need and recidivism in male and female youth: Examining the RNR principles in practice. Criminal Justice and Behavior, 39, 1025-1041. doi:10.1177/0093854812442895

36.

Young

Moline

Farrell

Biere

(2006). Best implementation practices: Disseminating new assessment technologies in a juvenile justice agency. Crime & Delinquency, 52, 135-158. doi:10.1177/0011128705281752