For this assignment, students will be analyzing a major problem facing todays juvenile justice system.? Chosen Problem: Mental Health and Juvenile J

Factors Associated With Mental Health and Juvenile Justice Involvement Among Children With Severe Emotional Disturbance Kelly N. Graves Bennett College for Women

James M. Frabutt Terri L. Shelton University of North Carolina at Greensboro

Recent research has highlighted the fact that there is an overrepresentation of children with mental health problems in the juvenile justice system. Thus, this study uses a clinical sample of children receiving mental health services to examine demographic (e.g., age, ethnicity), person- level (e.g., anxious and/or depressed), family-level (e.g., number of transitions in living situa- tions), and school-level factors associated with being involved in the mental health and juvenile justice service systems (i.e., dual involvement). Analyses were conducted separately by gender to investigate differences in dual involvement and possible differences in the predictors of dual involvement. For boys and girls, older adolescents and a higher number of living transitions were associated with dual involvement. For girls only, depression and/or anxiety and social problems were associated with dual involvement. The findings highlight the need for greater collaboration among service systems given the strong overlap between mental health and juve- nile justice involvement for many children.

Keywords: juvenile justice; mental health; serious emotional disturbance; system of care

According to the Surgeon General’s report on mental health, more than four millionchildren suffer from a major mental illness (Office of the Surgeon General, 2001). Researchers have documented that between 40% to 90% of children and adolescents involved in the juvenile justice system also suffer from a mental illness compared to 18% to 22% of the general youth population (Cocozza, Stern, & Blau, 2005; Kazdin, 2000; Teplin, 2001; Teplin, Abram, McClelland, Dulcan, & Mericle, 2002). Thus, it is likely that children’s men- tal health problems play a major role in their offending behaviors. Despite these concerns, lit- tle is known about clinically related factors that might be associated with dual involvement in mental health and juvenile justice systems. The current study uses a subsample of children receiving community-based mental health services to examine factors that are associated with dual involvement. Furthermore, because relatively little attention has been given to differ- ences in clinically relevant factors based on gender, the current study examines these factors separately by gender to predict whether a child with mental health problems also will become involved in the juvenile justice system.

Youth Violence and Juvenile Justice

Volume 5 Number 2 April 2007 147-167

© 2007 Sage Publications 10.1177/1541204006292870

http://yvj.sagepub.com hosted at

http://online.sagepub.com

147

Many children and adolescents who access community-based mental health services are diagnosed with a serious emotional disturbance (SED). SED is defined as having a clinical diagnosis, a functional impairment, and disturbances in multiple domains within the child’s life (e.g., school, home, community, etc.; Pumariega & Winters, 2003). The SED population is estimated to encompass approximately 4.5 to 6.3 million children (6% to 8%) in the United States (Friedman, Katz-Leavy, Manderscheid, & Sondheimer, 1999). By definition, these children and adolescents experience problems across multiple domains and often require coordinated, multiple-systems intervention (Hansen, Litzelman, & Marsh, & Milspaw, 2004). Previous research using community samples has indicated that almost 46% to 88% of children involved with the juvenile justice system also were diagnosed with a SED (Lyons, Baerger, Quigley, Erlich, & Griffin, 2001).

Based on Bronfenbrenner’s (1979) theory of social ecology, these children exemplify the postulation that behavior can be multiply determined. The social ecology concept posits that behavior can stem from internal biological and psychological mechanisms as well as exter- nal interactions with others across family, peer, and school domains. Because development has numerous contextual influences, the current study examines factors across a variety of domains (in addition to demographic factors such as gender and ethnicity), including person-level, family-level, and school-level factors to identify comprehensive, clinically rel- evant factors related to dual involvement. In doing so, the current study also adheres to best practice guidelines, which have been recommended by the President’s New Freedom Commission (2002) and many other researchers (e.g., Greene, Peters, & Associates, 1998; Hansen et al., 2004).

Recent research has found that examining factors across multiple domains may be more insightful when trying to understand antisocial behavior, particularly because there may be gender differences in the predictors of risk (Gorman-Smith & Loeber, 2005). In that study, antisocial behavior among girls was more influenced by parenting variables such as parental monitoring than among boys, whereas antisocial behavior among boys was more influenced by deviant peers than among girls. Those findings hint at the possibility that particular eco- logical systems (e.g., family vs. school) might have unique, gender-based influences on development. Thus, the current study expands on this possibility by examining whether there also might be different predictors based on gender for dual involvement with an SED sam- ple using factors that represent the multiple ecological systems within a child’s life.

Demographic Factors

Although the violent crime index declined in 2003 for the ninth consecutive year (falling 48% from its 1994 peak), there continues to be areas of concern with respect to arrest rates. For example, between 1980 and 2003, arrest rates for simple assault increased 269% for females and 102% for males (Snyder, 2005). Although boys repeatedly are reported to be more violent than girls (Kashani, Jones, Bumby, & Thomas, 1999), that trend is changing, with rates of violence among girls quickly approaching the rates of violence among boys (Snyder, 2005). Even given these statistics, some research indicates the existence of chival- rous treatment of female offenders in the initial stages of criminal processing (Visher, 1983), with girls receiving “lighter” punishments for illegal behaviors.

148 Youth Violence and Juvenile Justice

Those findings bring to the forefront the significant controversy regarding dispropor- tionate minority contact with the juvenile justice system. In Visher’s (1983) study, older, European American girls were less likely to be arrested than were younger, African American girls. However, the published statistics indicate that antisocial behavior and juvenile justice involvement disproportionately involve minorities, with African American children and ado- lescents reported to engage in more violent behaviors compared to European American or Hispanic children and adolescents (Blum, Ireland, & Blum, 2003; Earls, 1994; Kashani et al., 1999; Snyder, 2005). Of the estimated 1,400 murder arrests in 1999 and 2003, 49% and 48% were African American adolescents, respectively (Snyder, 2005; U.S. Department of Justice, 2000). In fact, some have reported that ethnicity has a very strong effect size in terms of pre- dicting violent behaviors across age ranges (d = .17, p < .01; Lipsey & Derzon, 1998). However, the federal Office of Juvenile Justice and Delinquency Prevention (OJJDP) repeat- edly has raised the concern that the percentage of minority youth involved in the juvenile jus- tice system is disproportionate to their representation in the general population. Specifically, recent estimates report that minority youth represent 34% of the juvenile population in the United States, but 62% of the nation’s detained youth (Hsia, Bridges, & McHale, 2004). National and state data (Ekpunobi, Wilson, Chunn, Huang, & Davis, 2002; Frazier & Bishop, 1995; Leiber, 2002; Snyder, 2005) consistently report finding systemwide disproportionate minority contact. Thus, it is unclear whether ethnicity is a true risk indicator, is associated with some other risk indicator (e.g., poverty, access to prevention services), and/or whether systemwide problems exist in terms of the overidentification of African American youth for juvenile justice involvement.

Person-Level Factors

Although there is a myriad of person-level factors that might be associated with dual involvement, the current study selected three factors that previous research has identified as strongly and consistently linked with involvement in the juvenile justice system. These three factors are anxious and/or depressed, depressed and/or withdrawn, and inattention and/or hyperactivity symptoms.

Research has indicated that involvement with the juvenile justice system often co-occurs with internalizing symptoms such as depression and anxiety, particularly among girls (Simmons, 2002; Teplin et al., 2002). It is possible that high levels of internalizing symp- toms increase the likelihood of “lashing out” behaviors that might increase the risk for juve- nile justice contact. For example, girls are at a higher risk for suicidal ideation compared to boys, and suicidal ideation and suicide attempts are associated with an increased likelihood of antisocial behaviors (Chandy, Blum, & Resnick, 1996). From a socialization perspective, it is possible that because girls are socialized away from aggression throughout their lives, the outlet for internalized anger often is unrefined, resulting in overt, impulsive, and some- times aggressive acts rather than healthy methods of exposing internalized feelings such as assertiveness or problem-solving strategies (Simmons, 2002). Furthermore, when females do attempt to address intense feelings through physical aggression, they often are punished more than males for doing so (e.g., Stueve, O’Donnell, & Link, 2001) without being taught alternate forms of conflict resolution. Consistent with the frustration–aggression hypothesis,

Graves et al. / Dual Involvement 149

or that frustration can trigger aggression (Berkowitz, 1989), it is at that point that aggression might manifest itself among boys and girls. However, for some girls, increased consequences and social ridicule may lead to increased shame and guilt. Such isolation stifles their con- tinued development and places them at further risk for developing psychological sequale such as clinical depression and/or anxiety (Orbach-Isreal, 2003). Because of the different socialization influences on development (Maccoby, 1990), and because of the more intense accumulation of multiple risks (Gorman-Smith, Tolan, Loeber, & Henry, 1998; McCord, 1982), it is hypothesized that the links between internalizing symptoms (i.e., anxious and/or depressed, depressed and/or withdrawn) and dual involvement will be stronger for girls than for boys.

Antisocial behavior also has been linked with higher rates of attention-deficit/hyperactiv- ity disorder (ADHD; Zoccolillo, 1993) and more general attention problems (e.g., Loeber, Green, Keenan, & Lahey, 1995). The combination of antisocial behavior, inattention, and hyperactivity–impulsivity sets into motion a pattern of person–environment interactions between the child and others, which often fosters and maintains individual differences among hyperactive and impulsive children compared to children who do not display such character- istics (Moffitt, Caspi, Rutter, & Silva, 2001). Although each of these factors has been linked to juvenile justice involvement individually, when combined, children can exhibit a general personality profile of disinhibition that can increase the risk for police contact (i.e., calls to the police, arrest decisions, court intake decisions). That possibility has been empirically val- idated, indicating that children who have hyperactivity, impulsivity, and attention problems (manifested as a general syndrome of disinhibition), and a history of conduct problems, are at the greatest risk for perpetuating antisocial behavior (Carlson, Tamm, & Gaub, 1997; Lynam, 1996, 1999).

Family-Level and School Factors

Children involved in the juvenile justice system often come from families with overex- tended resources. For example, high levels of caregiver strain have been linked with comor- bid diagnostic profiles and greater psychological distress (Brannan & Heflinger, 1997; Garland, Aarons, Brown, Wood, & Hough, 2003); however, different patterns based on child gender have not been empirically investigated. Furthermore, research has indicated that as the number of living transitions increases, child functioning decreases (particularly in the school environment; Simmons, Burgeson, Carlton-Ford, & Blyth, 1988). If levels of functioning and school success decrease, there may be an increased likelihood of antisocial or delinquent behavior, raising the risk of becoming involved with the juvenile justice system. More gener- ally, statistics indicate that family stressors such as a high number of living transitions and limited resources increases the risk of juvenile justice involvement (U.S. Department of Justice [USDoJ], 1995).

School failures characterized by high absenteeism and poor academic performances have been identified as risk factors for juvenile justice involvement (USDoJ, 1995). Some research indicates that this relationship might be stronger for girls compared to boys (e.g., Rankin, 1980; Thornton, Craft, Dahlberg, Lynch, & Baer, 2002). For example, some studies have documented that educational failure is an almost-universal correlate of delinquent girls,

150 Youth Violence and Juvenile Justice

whereas that is not necessarily the case among delinquent boys (Thornton, Craft, Dahlberg, Lynch, & Baer, 2002). Explanations for this difference are not immediately clear. However, some have speculated that girls who experience school failure resort to adopting a “bad girl” image to gain status because school success status appears unattainable (Koroki & Chesney- Lind, 1985). Thus, school failure might set into motion a pattern of peer rejection and con- frontation with teachers and parents, resulting in the increased likelihood that those children will gravitate toward deviant peer groups to achieve a sense of acceptance. Because girls tend to emphasize social relationships to a greater degree than boys, it is likely that school failure might be a stronger factor for dual involvement for girls compared to boys.

However, some have argued (e.g., Hawley, 1999; Vaughn, Vollenweider, Bost, Azria- Evans, & Snider, 2003) that aggression can be adaptive in certain social contexts because it creates “dominance status.” This dominance status sometimes functions to increase cohesion among the social group (Strayer & Trudel, 1984). Thus, if girls place more emphasis on social relationships and can achieve cohesion in a social group through aggression, it may be more likely that girls will resort to aggressive behaviors, raising the likelihood of juvenile justice contact. In contrast, some studies have indicated that school success is related to increased aggression among boys (but not among girls; Heimer & Matsueda, 1994). The hypothesized process is that a general increase in self-esteem, social acceptance, and admiration results in a decreased perceived likelihood of being punished for antisocial or risk-taking behaviors (Heimer & Matsueda, 1994). Thus, whereas school failure was related to increased likelihood of antisocial behavior among girls, school success was related to an increased likelihood of antisocial behavior among boys in some cases. It is hypothesized that school functioning will be differentially related to juvenile justice involvement for boys and for girls among clinically referred samples as well.

Hypotheses

In summary, the current study uses a clinical sample of youth identified with SED to determine what factors are related to dual involvement (i.e., mental health and juvenile jus- tice system involvement) separately among boys and girls. The use of a clinical sample is important in the current investigation because it allows for the examination of factors across multiple ecological systems within the context of a clinically referred population. Because it is important to remove any variance accounted for by general delinquency behaviors when predicting dual involvement, delinquent behaviors are controlled in all analyses. It is hypoth- esized that a larger proportion of boys would be dually involved than girls (Hypothesis 1). Based on the previous research reviewed above, each of the proposed factors (i.e., anxious and/or depressed, depressed and/or withdrawn, social problems, ADHD-type symptoms, caregiver strain, high number of living transitions, and low school functioning) is hypothe- sized to be positively associated with dual involvement (Hypothesis 2). In addition, the family- level factors (caregiver strain and number of living transitions) are hypothesized to be equally important among boys and for girls in terms of their relationships to dual involvement (Hypothesis 3). However, based on previous research (e.g., Charles, Abram, McClelland, & Teplin, 2003; Simmons, 2002; Snyder, 2005), it is hypothesized that the person-level factors of anxious and/or depressed, depressed and/or withdrawn, and social

Graves et al. / Dual Involvement 151

problems would be stronger predictors of dual involvement for girls than for boys (Hypothesis 4). Although previous findings have been somewhat inconsistent (Gorman- Smith & Loeber, 2005; Loeber & Farrington, 2000; Rankin, 1980; Thornton et al., 2002), it is hypothesized that school functioning would be a stronger predictor of dual involvement for girls than for boys (Hypothesis 5).

Method

Data Source

The current study uses a nationwide, representative subsample of children and adolescents receiving community-based mental health services through the Comprehensive Mental Health Services for Children and Their Families Program (funded by the federal Substance Abuse and Mental Health Services Administration, Center for Mental Health Services [CMHS]). More than 50,000 children and adolescents have entered this national program and received mental health services since 1993. The goal of that nationwide program is to provide services that are child-centered and family-focused, strengths-based, community- based, and culturally competent. The data used in the current study represents participants in this program between 1993 and 2002. Eligibility criteria for enrollment previously was determined by CMHS for purposes of the demonstration site grants and included: (a) being between age 5 and 18 years at intake (although only those at least age 11 years are included in the current study), (b) being a local county resident, (c) having a clinical diagnosis, (d) being separated or at risk of being removed from the home because of extreme behavioral or emotional difficulties, and (e) having multiple agency needs. The program also included an evaluation component that assessed system development and individual outcomes for children and families. All data collection protocols were established nationally. A full description of the national evaluation protocol and data-collection procedures is provided elsewhere (see Holden, Friedman, & Santiago, 2001).

Sample Selection

The current cross-sectional study focuses on European American and African American clinically referred children age 11 to 17 years who participated in the outcome study (N = 1,168). All children had at least one clinical diagnosis, with the most common diagnosis being a mood disorder (31%) followed by ADHD (22%). See Table 1 for percentages of children listed across all diagnostic categories. More than 68% of children had multiple diagnoses, with the average number of diagnoses being 1.86. In terms of psychotropic med- ication, 83% of children reported taking psychotropic medication on entry into the service system. The specific type of psychotropic medication was not identified in data collection and thus could not be reported here.

Procedures

Children were referred to their local community mental health program from a variety of sources, including caregivers, child-serving agencies (e.g., Department of Social Services,

152 Youth Violence and Juvenile Justice

Department of Juvenile Justice, Department of Public Health), and schools. Consent forms for treatment and for participation in the evaluation process were signed by the primary care- giver (or legal guardian if different from the caregiver) and the child. Families were informed that an interviewer would be contacting them within a few days to schedule an interview. Interviews were scheduled as soon as possible, but no later than 30 days after the initiation of services.

Trained evaluators conducted in-home interviews lasting approximately 2 hrs for care- givers and 2 hr for children. All instruments were read to children and their caregivers to minimize possible error due to differential reading abilities. Caregivers received U.S. $25 for their participation; children received gift certificates donated from local fast food restaurants.

Measures

Demographic Information Questionnaire (DIQ; Center for Mental Health Services [CMHS], 1997). The DIQ is a 37-item caregiver-reported questionnaire that measures child

Graves et al. / Dual Involvement 153

Table 1 Descriptive Statistics for the Sample (N == 1,168)

Indicator % M SD Range

Age 13.86 1.78 11.00 – 17.00 Male 63 African American 22 Custody status

Two-parent family 25 Single-parent family 53 Grandparents 6 Adoptive and/or foster parents 5 State custody 9 Other relatives 4

Caregiver education level Not a high school graduate 24 High school graduate 35 Attended some college 41

Family income Less than U.S. $15,000 46 Above $15,000 54

Clinical diagnoses Mood disorder 31 Attention-deficit/hyperactivity disorder 22 Oppositional defiant disorder 15 Conduct disorder 7 Adjustment disorder 7 Anxiety disorder 2 Substance use disorder 3

and family characteristics such as age, race, ethnicity, risk factors, family structure, physi- cal custody, referral source, presenting problems, family income living arrangements, edu- cation, household composition, physical health, and medications.

Person-Level Factors (Attention Problems, Social Problems, Anxious and/or Depressed, Depressed and/or Withdrawn)

All caregivers and children reported on the level of attention problems, social problems, anxious and/or depressed problems, and depressed and/or withdrawn problems using the Child Behavior Checklist (CBCL; Achenbach, 1991a) for caregivers, and the Youth Self Report (YSR: Achenbach, 1991b) for children. Caregiver and child reports of each construct were correlated highly with one another (at least r = .20, p < .001). Therefore, based on pre- vious research using this procedure (Loeber et al., 2000), and the author’s recommendation to combine reports (Achenbach, 1991a), composite scores were created by averaging the T-scores across caregiver and child reports for each construct separately. This composite score was used as the indicator of each person-level factor. Internal reliability (> .82), test– retest reliability (> .87 for all scales), and validity have been demonstrated in previous stud- ies (Achenbach, 1991a).

Attention and hyperactivity and/or impulsivity. To assess levels of attention and hyper- activity, the current study utilized the Attention Problems subscale on the CBCL and YSR, which includes 20 items related to inattention, hyperactivity, and impulsivity. Sample items include, “Now or within the last six months, my child can’t sit still, is restless, or is hyperactive” and “Now or within the last six months, I often act without thinking.” It is important to note that the measure used taps not only attention problems but also hyper- activity and/or impulsivity. However, because the author of the measure titled the subscale “Attention Problems” in the analyses, the label of that scale is used but refers to attention and hyperactivity and/or impulsivity problems. Caregivers and children responded to each item on a 3-point scale, from 0 = not true through 2 = very true or often true. Caregiver and adolescent reports were correlated .32 (p < .001). The current study used the T-score composite from the Attention Problems subscale, with higher T-scores indicating higher levels of attention problems.

Anxious and/or depressed. Caregivers and adolescents reported levels of adolescent anx- ious and/or depressed using the Anxious/Depressed subscale from the CBCL for caregivers and the Anxious/Depressed subscale from the YSR for adolescents. The CBCL contains 14 items (e.g., “Now or within the past six months, my child has been unhappy, sad, or depressed,” or “Now or within the past six months, my child has felt worthless or inferior”), and the YSR contains 16 items (e.g., “Now or within the past six months, I cry a lot,” or “Now or within the past six months, I feel lonely”). Caregivers and adolescents responded on a 3-point scale, from 0 = not true through 2 = very true or often true. Caregiver and ado- lescent reports were correlated .28 (p < .001). The current study used the T-score compos- ite from the Anxious/Depressed subscale, with higher T-scores indicating higher levels of anxiety and depression.

154 Youth Violence and Juvenile Justice

Social problems. Caregivers and adolescents reported levels of adolescent social problems using the Social Problems subscale from the CBCL for caregivers, and the Social Problems subscale from the YSR for adolescents. The CBCL and YSR contain 8 items, including “Now or within the past six months, my child gets teased a lot,” and “Now or within the past six months, my child is not liked by other kids.” Caregivers and adolescents responded on a 3-point scale, from 0 = not true through 2 = very true or often true. Caregiver and adoles- cent reports were correlated .42 (p < .001). The current study used the T-score composite from the Social Problems subscale, with higher T-scores indicating higher levels of social problems.

Depressed and/or withdrawn. Caregivers and adolescents reported levels of adolescent depression and withdrawal using the Depressed/Withdrawn subscale from the CBCL for caregivers and the Depressed/Withdrawn subscale from the YSR for adolescents. The CBCL contains 8 items (e.g., “Now or within the past six months, complains of loneliness,” and the YSR contains 6 items (e.g., “Now or within the past six months, I cry a lot,” or “Now or within the past six months, feels that nobody loves me”). Caregivers and adolescents responded on a 3-point scale, from 0 = not true through 2 = very true or often true. Caregiver and ado- lescent reports were correlated .20 (p < .001). The current study used the T-score composite from the Depressed/Withdrawn subscale, with higher T-scores indicating higher levels of depression and withdrawal.

Family-Level Factors

Caregiver strain. Caregivers reported on their levels of strain on the Caregiver Strain Questionnaire (CGSQ; Brannan & Heflinger, 1997). The CGSQ has 21 items that assess the degree to which a caregiver feels strained related to caring for a child with mental health needs. Items include, “interruption of personal time,” “financial strain,” and “feeling socially isolated.” Caregivers respond on a 4-point scale, with higher scores indicating more strain. The current study utilizes a composite strain score that is an average of the 21 items.

Living transitions. On the DIQ mentioned above, caregivers responded to the question, “How many times has the child changed living residences in the past six months?” Caregivers’ responses to that question were used as the indicator of the number of living transitions for each child.

School-Level Factors

To assess school functioning, caregivers completed the Child and Adolescent Functional Assessment Scale (CAFAS; Hodges, 1994). The current study utilized the School Role sub- scale, which assesses the degree of impairment in school functioning. Items included, “non- compliant behavior which results in persistent or repeated disruption,” “frequently truant,” and “disruptive behavior.” The CAFAS is rated on a 30-point scale (0 = no impairment, 10 = mild impairment, 20 = moderate impairment, 30 = severe impairment). Thus, higher scores indicate greater impairment in school functioning. Interrater reliability and validity

Graves et al. / Dual Involvement 155

have been demonstrated in previous studies (Hodges & Wong, 1996), and mental health professionals were trained to achieve high interrater correlations (> .80) between their rat- ings and criterion ratings established by the author (Hodges, 1994).

Dual Involvement

To assess dual involvement, one dichotomized item from the Delinquency Survey (DS; CMHS, 1994) was utilized. That item was

Collepals.com Plagiarism Free Papers

Are you looking for custom essay writing service or even dissertation writing services? Just request for our write my paper service, and we'll match you with the best essay writer in your subject! With an exceptional team of professional academic experts in a wide range of subjects, we can guarantee you an unrivaled quality of custom-written papers.

Get ZERO PLAGIARISM, HUMAN WRITTEN ESSAYS

Why Hire Collepals.com writers to do your paper?

Quality- We are experienced and have access to ample research materials.

We write plagiarism Free Content

Confidential- We never share or sell your personal information to third parties.

Support-Chat with us today! We are always waiting to answer all your questions.