Validation of the Ohio Youth Assessment System Dispositional Tool (OYAS-DIS): An Examination of Race and Gender Differences

• DOI: 10.1177/1541204019859938
• Author: Christopher D’Amato,Christina A. Campbell,Jordan Papp
• Published date: 01 April 2020
• Date: 01 April 2020
• Subject Matter: Articles

Article
Youth Violence and Juvenile Justice
2020, Vol. 18(2) 196-211
Validation of the Ohio Youth
ª The Author(s) 2019
Article reuse guidelines:
Assessment System
sagepub.com/journals-permissions
DOI: 10.1177/1541204019859938
journals.sagepub.com/home/yvj
Dispositional Tool (OYAS-DIS):
An Examination of Race
and Gender Differences
Christina A. Campbell1 , Christopher D’Amato1, and Jordan Papp2
Abstract
The Ohio Youth Assessment System-Disposition Tool (OYAS-DIS) is a juvenile risk assessment that
is used in numerous states and jurisdictions to assess criminogenic risk of juvenile offenders. Still,
there is little published research on the predictive validity of the tool. The purpose of the current
study was to examine the predictive validity of OYAS-DIS, with a specific focus on understanding
prediction of recidivism across racial and gender subgroups. The sample consisted of 4,383 youth
that received a court petition in a single large Midwestern county juvenile court. The findings
indicated that the OYAS-DIS was a statistically significant predictor of recidivism across all racial and
gender subgroups. However, there was statistically significant variation in predictive validity across
subgroups. For instance, the tool was a statistically significantly better predictor of recidivism for
White males as compared to Black male youth. There was also statistically significant variation in the
predictive validity of certain domains (e.g., juvenile justice history) on the OYAS-DIS across racial
and gender subgroups. Implications of research favor the use of the OYAS-DIS to predict recidivism
for adjudicated juveniles.
Keywords
risk assessment, predictive validity, Ohio Youth Assessment System, racial disparities, gender
disparity
Juvenile offending has significantly decreased in last two decades; yet, there remain disparities
across gender and racial groups in the arrest rates and processing practices of the juvenile justice
system (Cochran & Mears, 2015; Office of Juvenile Justice and Delinquency Prevention [OJJDP],
2019). In fact, current reports suggest that the number of delinquency cases has decreased by 53%
yet issues of racial disproportionality remains (OJJDP, 2016). For instance, Black individuals
1 University of Cincinnati, Cincinnati, OH, USA
2 University of Michigan, Ann Arbor, MI, USA
Corresponding Author:
Christina A. Campbell, University of Cincinnati, 660 MC TDC, PO Box 210389, Cincinnati, OH 45221, USA.
Email: campb2ci@ucmail.uc.edu

---

Campbell et al.
197
comprise roughly 16% of the youth population but 29% of delinquency caseloads in juvenile justice
(Onifade, Davidson, & Campbell, 2009). Further, justice-involved female youth represent the fastest
growing group of juvenile offenders (Chesney-Lind & Sheldon, 2004). Female youth accounted for
roughly 20% of juvenile arrests in 1992, while recent research estimates this number is currently
around 29% (Pusch & Holtfreter, 2018). Finally, there is also empirical evidence that suggests that
there are disparities in the allocation of programs and services. Cochran & Mears (2015) found that
justice-involved White male youth are more likely to receive rehabilitative services than female and
minority youth. These findings point to a need to examine evidence-based strategies that address
race and gender disparities that exist throughout juvenile justice system.
A common strategy juvenile justice agencies use to address disproportionality is incorporating
modern juvenile risk assessment tools into their practices. In fact, some researchers have gone as
far as saying that agencies who do not use risk assessments are unprofessional, unethical, or biased
(Grove & Meehl, 1996; Quincey, Harris, Rice, & Cormier, 1998). This is because risk assessment
tools serve many different functions for juvenile justice agencies and practitioners. For instance,
risk assessments help replace the use of professional/clinical judgments by standardizing decision-
making, informing agencies of potential risks and needs of youth, and assisting juvenile justice
actors with making evidence-based program recommendations for the youth offenders (Bonta &
Andrews, 2017). Most importantly, risk assessments have increased the ability for court actors to
predict which youth offenders have the highest probability of recidivating. As a result, practi-
tioners are better able to focus court resources on juveniles who are highest risk of future crime and
exhibit the greatest level of intervention needs (Bonta & Andrews, 2017; Olver, Stockdale, &
Wormith, 2009).
A current debate that continues is the extent to which risk assessments improve disparities that
remain in the justice system. Overall, research is promising and suggests that the adoption of
decision-making tools, like juvenile risk assessments, have potential to reduce disparities present
in (or that occur as the result of) juvenile justice decision-making processes (Schwalbe, Fraser, Day,
& Cooley, 2006). Modern day risk assessments that capture dynamic risks and needs have shifted
practitioners’ focus from crime type to eight central factors or treatment areas (i.e., antisocial
behavior, antisocial personality; cognition, peers, family, education, leisure recreation, and sub-
stance abuse) that reduce the likelihood of future delinquency. If risk assessments are fair and valid,
risk assessment tools should reduce racial, ethnic, and gender biases that may occur in juvenile
justice by increasing the reliability and consistency of assessing the risk of offenders through a
standardized process (Schwalbe et al., 2006). The hope is that these instruments will correctly
classify offenders by risk across all subgroups. However, evidence suggests that ethnic minorities
and girls consistently experience disproportionality as the result of several policies and practices that
may increase criminogenic risk (Liska, 1992; Schwalbe et al., 2006). Thus, it is through validation
studies where theories can be expanded concerning how social factors, treatment decisions, and
policies may threaten the validity of risk assessment tools.
Prior research has suggested that differences in social experiences of diverse populations may
lead to differences in risk assessment score and predictive validity. As result, static factors related to
criminal history or certain family characteristics may be both associated with certain subgroups and,
in some cases, arbitrarily increase certain juveniles risk level (Moore & Padavic, 2011). Research
has suggested that many juvenile risk assessment instruments do not take into account items that
may uniquely impact girls or specific ethnic groups. In fact, studies have questioned the effect of
trauma or abuse histories, poverty, potential neighborhood bias, school policies, education, policing
practices, and depression on youth’s risk level (see Shepherd, Luebbers, & Dolan, 2013; Van
Voorhis, Peiler, Presser, Spiropoulis, & Sutherland, 2001). Given there are several social factors
that may impact recidivism beyond items measured by these tools, more research is needed to
examine predictive validity across subgroups.

---

198
Youth Violence and Juvenile Justice 18(2)
Predictive Validity of Risk Assessment by Race and Gender
Proponents of standardized risk tools believe that risk assessments should work in a race neutral
manner (Bonta & Andrews, 2017; Latessa, Listwan, & Koetzle, 2014). However, there remain
concerns about the extent to which these tools are a measure of racial inequities. For example,
researchers have questioned the impact of surveillance and differential treatment on risk factors
measured on assessment tools (e.g., criminal history). Studies have hypothesized that racial dispa-
rities, like disproportionate minority contact, are the result of racial minorities being targeted in ways
that lead to increased reporting, arrests, referrals to the court, and biased case management practices
(C. Campbell, Papp, Barnes, Onifade, & Anderson, 2018; Harcourt, 2015; Ryan, Williams, &
Courtney, 2013). These mechanisms of differential treatment can exacerbate rates of recidivism
and increase risk score, which can consequently reduce the validity of risk assessment tools.
A study conducted by C. Campbell, Papp, Barnes, Onifade, and Anderson (2018) found that the
Youth Level of Service/Case Management Inventory (YLS/CMI) was slightly stronger predictor of
recidivism for White youth probationers as compared to Black youth probationers, although there
were no differences in criminogenic risk scores of White and Black probationers. Although the YLS/
CMI was a significant predictor of recidivism over and above chance across both racial groups, some
differences remained concerning the tools ability to differentiate across low-, moderate-, and high-
risk Black offenders (C. Campbell et al., 2018). Furthermore, Onifade, Davidson, and Campbell
(2009) found that while the YLS/CMI moderately predicted recidivism across a diverse sample of
juvenile probationers (Black and White male and females), the tool demonstrated better predictive
validity for White juveniles than Black juveniles. According to Schwalbe, Fraser, Day, and Cooley
(2006), risk assessments have the potential to underestimate intervention needs as well across race.
Schwalbe et al. (2006) found that when using the North Carolina Assessment of Risk (NCAR),
Black youth scored and recidivated at significantly higher rates than White youth. Further, while
NCAR predicted recidivism (measured as new adjudication) moderately well for both Black and
White youth, it...