Citation
A gender-specific analysis of community-based juvenile justice reform

Material Information

Title:
A gender-specific analysis of community-based juvenile justice reform the effectiveness of family therapy programs for delinquent girls
Creator:
Aultman-Bettridge, Tonya
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
xiv, 157 leaves : ; 28 cm.

Subjects

Subjects / Keywords:
Juvenile justice, Administration of -- United States ( lcsh )
Community-based social services -- United States ( lcsh )
Family psychotherapy -- United States ( lcsh )
Female juvenile delinquents -- Services for -- United States ( lcsh )
Community-based social services ( fast )
Family psychotherapy ( fast )
Female juvenile delinquents -- Services for ( fast )
Juvenile justice, Administration of ( fast )
United States ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Colorado at Denver and Health Sciences Center, 2007. Public affairs
Bibliography:
Includes bibliographical references (leaves 150-157).
General Note:
School of Public Affairs
Statement of Responsibility:
by Tonya Aultman-Bettridge.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
166268817 ( OCLC )
ocn166268817

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

A GENDER-SPECIFIC ANALYSIS OF COMMUNITY -BASED JUVENILE JUSTICE REFORM: THE EFFECTIVENESS OF FAMILY THERAPY PROGRAMS FOR DELINQUENT GIRLS by Tonya Aultman-Bettridge B.A., University of Colorado at Boulder, 1995 B.S., University of Colorado at Boulder, 1995 M.C.J., University of Colorado at Denver, 1997 A thesis submitted to the University of Colorado at Denver and Health Sciences Center in partial fulfillment of the requirements for the degree of Doctor of Philosophy Public Affairs 2007 l /) .. t

PAGE 2

2007 by Tonya Aultman-Bettridge All rights reserved.

PAGE 3

This thesis for the Doctor of Philosophy degree by Tonya Aultman-Bettridge has been approved

PAGE 4

Aulbnan, Bettridge, Tonya (Ph.D. Public Affairs) A Gender-Specific Analysis of Community-Based Juvenile Justice Reform: The Effectiveness of Family Therapy Programs for Delinquent Girls Thesis directed by Professor Eric Poole ABSTRACT In an effort to improve outcomes and operation of its criminal and juvenile justice systems, the Washington State Legislature passed the 1997 Community Justice Accountability Act. A cornerstone of this Act was a significant push to implement "research-proven, cost effective intervention programs for juvenile offenders on a statewide basis" (Washington State Institute for Public Policy, 1999: I). Evaluations of programs serving both adult and youthful offenders were closely examined by the legislature's research center, the Washington State Institute of Public Policy (WSIPP). The WSIPP estimated the cost-savings of a variety of programs, based on their ability to prevent further criminal or delinquent offenses (Aos, Phipps, Bamoski, & Leib, 2001). Over the past seven years, the CJAA and the findings of the WSIPP have largely driven juvenile justice policy decisions throughout the state. Despite some positive indicators of changes in Washington State juvenile crime, some offense areas remain problematic. Among these is the increase in juvenile crimes committed by female offenders.

PAGE 5

This paper examines Washington State juvenile justice reform efforts, focusing on the effects of community-based family therapy programs for delinquent girls. The study looks at changes in risks for continuity of delinquent behavior and actual recidivism for a sample of 157 girls referred to Functional Family Therapy and Multi-systemic Therapy programs between January 2004 and June 2006. An equivalent comparison group pretest-posttest design is used to compare these effects to girls not receiving services from one of these two programs. Findings indicate some promise. Small improvements in risk factors were observed for girls completing the programs. This improvement was slightly greater than improvements for comparison group girls, but the differences were not significantly different. There were no significant differences in recidivism rates between girls participating in the programs and the comparison groups. More study is needed to determine how program implementation may be affecting outcome. Further study is needed in order to better understand how these programs might be effective in treating delinquent girls. This abstract accurately represents the content of the candidate's thesis. I recommend its publication. Signed

PAGE 6

DEDICATION I dedicate this thesis to my husband and to my son. Without their inspiration, love, and support, this would have not been possible. I also dedicate this to my parents and my wonderful extended family, for their unflagging love and support.

PAGE 7

ACKNOWLEDGEMENT My sincere thanks go to my advisor, Eric Poole, for his support during this journey. I also wish to extend my thanks to the wonderful members of my committee and to the entire GSP A staff for their assistance in this process. A heartfelt thank-you also goes to Peter Selby and Andrew Keller at TriWest Group, for their unfaltering encouragement and support. Without them this project would have never been possible. Finally, thanks to the hardworking and dedicated staff at the King County Juvenile Court for allowing me to experience the great things they are doing for youth and their families.

PAGE 8

CONTENTS Figures ............................................................................................................. xi Tables ............................................................................................................. xii CHAPTER 1. INTRODUCTION ....................................................................................... The Promise of Gender-Specific Programs ......................................... .4 Juvenile Justice Reform Efforts in Washington State ........................... 5 2. RESPONDING TO JUVENILE DELINQUENCY: THEORY, PRACTICE AND GENDER ISSUES ...................................... 11 Understanding Female Delinquency: Early Approaches .................... 12 Social, Environmental and Ecological Factors Female Delinquency ........................................................................... 17 Individual Personality, Temperament and Social Skills-Related Factors ............................................................. 20 Neighborhood, Recreation and Employment .......................... 24 School Attitude and Achievement Factors ............................. 25 Relationships with Delinquent Peers ...................................... 26 Family Issues and Parenting Practices ................................... .28 Alcohol and Drug Use and Mental Health Issues .................. 30 Community and Societal Factors ............................................ 31 Vll

PAGE 9

Using the Socio-Ecological Model in Assessing and Treating Delinquent Girls .................................................................. 33 3. SYSTEM REFORM: CHANGING APPROACHES TO THE TREATMENT OF JUVENILE OFFENDERS IN WASHINGTON STATE ............................................................................... 34 Functional Family Therapy ......................................................................... .43 Measuring Program Effects ............................................................... 49 Multi-Systemic Therapy .............................................................................. 50 Measuring Program Effects ................................................................ 56 Washington State's Experience with FFT and MST Programs ..................................................................................................... 56 A Conceptual Framework for Examining Washington State's Reforms ....................................................................... 57 Research Questions ............................................................................ 60 4. STUDY DESIGN AND METHODS ........................................................ 63 Research Design ................................................................................. 63 Sample ................................................................................................ 66 Data Collection ................................................................................... 68 JJIS Program Database Extract .............................................. 68 Court Filings .......................................................................... 69 Risk of Re-Offense at Program Entry ..................................... 70 Outcome Data: Changes in Risk Factors for Re-Offending and Actual Recidivism ................................................ 74 Data Analysis ......................................................................... 81 Vl11

PAGE 10

Data Limitations ................................................................................... 83 Equivalent Comparison Group Design .................................... 84 Implementation Quality ofPrograms ....................................... 86 5. DEMOGRP AHIC, RISK AND OFFENSE PROFILES OF DELINQUENT GIRLS REFERRED TO THE FFT AND MST PROGRAMS ............................................................................ 88 Referring Offenses .............................................................................. 89 Risk for Re-Offense ............................................................................ 93 Program Experiences of Girls Referred to FFT and MST ................ 102 Outcomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 06 Changes in Family Functioning ............................................. 06 Changes in Peer and Adult (non-family) Relationships ....... .1 08 Changes in Substance Abuse and Mental Health Risk .......... 1 09 Changes in Attitudes Towards Delinquency Risk Scores ...... 111 Changes in Attitudes Towards Aggression Risk Scores ........ 113 Changes in Social Skills Risk Scores .................................... .114 Summary of Changes in Risk for Re-Offending ................... .116 Recidivism Outcomes ...................................................................... 117 Recidivism Outcomes Summary ........................................... 121 Factors Associated with Improved Recidivism Outcomes ................ 121 ix

PAGE 11

6. DISSCUSSION AND CONCLUSIIONS .................................................. 125 Discussion: Findings for Each Research Question ............................ 125 Implications for Juvenile Justice Programming for Girls .................. 132 Implications for FFT and MST .............................................. 133 Directions for Future Research .......................................................... 133 APPENDIX A. WSJCA Pre-Screen ............................................................................ 135 B. WSJCA Full Assessment ................................................................... 139 REFERENCES .......................................................................................................... 150 X

PAGE 12

FIGURES Figure 3.1 Washington Juvenile Court Assessment Process ............................................. 3 8 4.1 Pretest-Posttest Design ..................................................................................... 63 4.3 Risk Level Definitions Using Criminal History and Social History Risk Scores ........................................................................ 72 5.1 Referral to Engagement Process .................................................................... 1 04 5.2 Program Engagement and Completion Process ............................................. } 06 Xl

PAGE 13

TABLES Table 3.1 WSJCA domains and descriptions ................................................................... 39 3.2 Risk levels based on Criminal and Social History scores .............................. .41 3.3 Recidivism rates for youth based on risk category ........................................ .41 3.4 JJIS program eligibility criteria ...................................................................... .44 4.1 Full Screen score ranges: 12 domains .............................................................. 73 4.2 Comparison of recidivism, based on data source ............................................. 77 4.3 Group equivalence: Comparison vs. program girls ......................................... 86 5.1 Demographic comparison of girls referred to FFT or MST versus all female juvenile court filings ............................................................ 89 5.2 Distribution of offense types for all juvenile court filings against girls, January 2004 to June 2006 ...................................................................... 90 5.3 Comparison of filing distributions, all female delinquency filings vs. MST and FFT referrals (January 2004 to June 2006) ...................................... 91 5.4 Specific offenses for girls referred to FFT and MST ....................................... 93 5.5 Offense types: FFT vs. MST referred youth .................................................... 94 5.6 Pre-Screen risk levels for youth referred to FFT and MST ............................. 95 5.7 Distribution across Pre-Screen risk levels ....................................................... 96 5.8 Criminal and Social History Pre-Screen scores ............................................... 97 5.9 WSJCA average domain scores for all girls referred ....................................... 98 xu

PAGE 14

5.10 Domain risk levels for all girls referred ......................................................... 1 00 5.11 Percentage of girls scoring in the "High" range, by program ........................ I 02 5.12 Reason for withdrawal of referral (MST only) ............................................. .103 5.13 Completion rates for girls engaged in FFT vs. MST .................................... .104 5.14 Program lengths of stay for girls completing FFT vs. MST .......................... } 05 5.15 Changes in family functioning: Program participants vs. comparison group ................................................... 108 5.16 Changes in family functioning, by type of participation ............................... 1 08 5.17 Changes in relationships: Program participants versus comparison .............. 1 09 5.18 Changes in relationship risk scores, by type of participation ........................ 11 0 5.19 Changes in substance abuse risk: Program participants versus ..................... 112 comparison group 5.20 Changes in substance abuse risk scores: By type ofparticipation ................ .ll2 5.21 Changes in attitudes toward delinquency risk scores: Program vs. comparison group ...................................................................... 113 5.22 Changes in attitudes toward delinquency risk scores, by type of participation ............................................................................................... 113 5.23 Changes in attitudes toward aggression risk scores: Program vs. comparison group ...................................................................... 115 5.24 Changes in attitudes toward delinquency risk scores, by type of participation ............................................................................................... 115 5.25 Changes in social skills risk scores: Program vs. comparison group ............ 116 5.26 Changes in social skills risk scores, by type of participation ........................ 116 xiii

PAGE 15

5.27 Six-month recidivism rates: Program vs. comparison group ........................ .l19 5.28 Six-month recidivism rates, by program completion ..................................... 119 5.29 Six-month recidivism rates, by risk level ...................................................... 120 5.30 Twelve-month recidivism rates: Program vs. comparison group .................. 120 5.31 Twelve-month recidivism rates, by program completion .............................. 121 5.32 Time to recidivism, by program group .......................................................... 121 5.33 Number of new filings, by program group .................................................... 121 5.34 Twelve-month recidivism rates, by risk level. .............................................. .l22 5.35 Twelve-month recidivism rates, FFT vs. MST .............................................. 122 5.36 Correlation matrix, means and standard deviations for variables in regression analysis (n=146) .................................................. 124 5.37 Logistic Regression results ............................................................................ 125 XIV

PAGE 16

CHAPTER 1 INTRODUCTION During the late 1980s and 1990s, crimes committed by young people drew a great deal of attention in America. Rising juvenile crime rates (particularly violent crime) increased awareness of youth gang presence throughout the country and a small but tragic number of school shootings combined to create a juvenile justice climate where "get tough" approaches and issues of accountability have been favored over a more treatment-oriented policy model; consequently, nearly all states have passed laws increasing sanctions for juvenile offenders, including a greater use of incarceration and waivers to criminal court (National Conference of State Legislators, 1994-2002; Sickmund, Synder, & PoeYamagata, 1997). In more recent years, while arrests for adolescent males have been declining in many offense categories, delinquency rates for girls (and violent crime in particular) either outpaced increases in rates for boys or decreased at a much smaller rate {American Bar Association &National Bar Association, 2001 ). This led to a recent Office of Juvenile Justice and Delinquency Prevention (OJJDP) report calling female offenders "the fastest growing segments of the juvenile and criminal justice systems" (Acoca, 1999: 2). 1

PAGE 17

Consistent with rising concern over juvenile crime and an increasingly public safety-oriented perspective regarding the issue, rates of youth incarceration in the 1990s grew dramatically (Sickmund et al., 1997). In addition, most states experienced dramatic increases in the use of secure short-term detention for both pre-adjudicated and sentenced youth, including those committing technical probation and parole violations. With this rise in use of detention came two significant challenges currently faced across the country. The first was overcrowding in detention facilities, leading to a number of lawsuits that have forced states to cap populations in their largest facilities (see American Civil Liberties Union, 2002; US Department of Justice, 2003). The second was a growing concern over minority overrepresentation in secure detention settings (Feldman, Males & Schiraldi, 2001). These concerns prompted many states to embark upon significant detention reform efforts (Fox & Lyons, 2003). One important focus often missing from these new reform efforts is attention to the growing number of delinquent girls being housed in detention facilities. While rates of detention have been increasing overall, the incarceration rate for young female offenders has outpaced growth in male detention rates. Between 1988 and 1997 juvenile male detention rates increased 30%, compared with a 65% increase for girls. In addition, girls are more likely than boys to be detained "for minor offenses that do not warrant detention according to the principle ofthe 'least restrictive alternative,' such as technical violations of probation and parole" (ABA & NBA, 2001: 27). 2

PAGE 18

While girls have overall lower rates of recidivism (as measured by new offenses) than boys, they tend to experience higher detention re-admission rates for issues stemming from their previous (rather than a new) criminal offense, such as contempt of court, technical violations of the terms of probation or parole, and failure to meet court-mandated program goals (Stanfield, 1995). This suggests that successful programs for girls need to focus as strongly on short-term treatment success, such as completing of probation or meeting court-mandated goals, as on general delinquency recidivism. One possible explanation for girls' trouble in meeting short-term treatment goals and maintaining positive behavior for the duration of their probation period is the discord between what girls need, based on the unique nature and causes of their offending, compared with what they actually get in programs designed for boys. Belknap and her colleagues (Belknap & Gaarder, 2002; Belknap & Holsinger, 1997) have examined the experiences of incarcerated girls in a variety of juvenile justice systems and settings, including those in juvenile detention facilities in Ohio and Colorado. Their research revealed problems identified by girls in receiving gender appropriate health care and nutrition information, including treatments for pregnancy, sexually-transmitted diseases, and eating disorders. Girls also had fewer educational and recreational opportunities than boys in the same facilities. Caseworkers interviewed in these studies confirmed many of the issues reported by the girls, including lack of treatment opportunities (compared with boys) and lack of gender3

PAGE 19

appropriate health care. Caseworkers also noted a difficulty in establishing a trusting relationship with girls that led to barriers in both the assessment and treatment processes. Finally, some staff exhibited negative attitudes towards working with girls, including one who stated: "Girls are more difficult [than boys]. I hate them" (as quoted in Belknap & Holsinger, 1997: 397). While seemingly extreme, similar and more subtle biases are evident across systems. For example, when the Baltimore Female Intervention Team (FIT) was originally established (a model requiring all-girl caseloads for FIT members), it was necessary to transfer cases among workers to facilitate the caseload balance in a male-dominated system. The ratio established for the "swaps" reflects the problem of negative staff perceptions, with one girl "swapped" from a caseload for each additional 10 boys (Daniel, 1999). The Promise of Gender-Specific Programs Despite current barriers to the effective treatment of delinquent girls, some information is emerging regarding promising gender-specific treatment practices. These findings center on female pathways to delinquency and how girls' needs differ from those of male delinquents. Delinquent girls' needs, often stemming from a past history of abuse, unhealthy relationships, and substance use, have been found to include: a need for physical, psychological and emotional safety; a need for female role models and relationships with adults characterized by trust and respect; a need to 4

PAGE 20

understand their own sexuality; and a need for self-efficacy and self-respect (Green, Peters & Associates, 1998). This emerging knowledge regarding effective treatment options for female offenders is a vital component of overall juvenile justice reform. In order to reduce rates of detention, the needs of all youth in these facilities must be addressed in a way that maintains the public safety, but also increases the chances that an individual youth will stay out of the juvenile justice system in the future. Until then, reform efforts may hope to change the "face" of the juvenile justice population but not to significantly reduce its numbers. Studying Juvenile Justice Reform Efforts in Washington State As states respond to growing detention populations and changes in the demographics of the youthful offenders being served, the State ofWashington is being recognized as a model by many other states. Approaches developed in the past decade of its reform efforts have since been adopted by jurisdictions in Florida, New York, Colorado and Illinois. These efforts are built upon the growing literature in effective juvenile justice strategies and are targeted towards moving to a greater use of community-based models of service delivery and more restricted use of incarceration. Within the State of Washington, King County is the largest juvenile court jurisdiction implementing these reform efforts. It includes the City of Seattle and a significant proportion of the state's juvenile offenses. 5

PAGE 21

In an effort to improve outcomes and operation of its criminal and juvenile justice systems, the Washington State Legislature passed the 1997 Community Justice Accountability Act (CJAA). A cornerstone of this Act was a significant push to implement "research-proven, cost-effective intervention programs for juvenile offenders on a statewide basis" (Washington State Institute for Public Policy, 1999: 1). Evaluations of programs serving both adult and youthful offenders were closely examined by the legislature's research center, the Washington State Institute of Public Policy (WSIPP). The WSIPP estimated the cost-savings of a variety of programs, based on their ability to prevent further criminal or delinquent offenses (Aos, Phipps, Barnoski, & Leib, 2001). Over the past seven years, the CJAA and the findings of the WSIPP have largely driven juvenile justice policy decisions throughout the state. In its response to the CJAA, King County developed in 2002 the Juvenile Justice Operational Master Plan (JJOMP), which included the utilization of a standardized risk assessment tool and the implementation of three programs identified by the WSIPP (1999) as effective in preventing recidivism in juveniles involved with the courts: Anger Replacement Training (ART), Functional Family Therapy (FFT), and Multi-Systemic Therapy (MST). These three programs are collectively referred to as Juvenile Justice Intervention Services (JJIS). The present study is located in this larger evaluation of the JJIS initiative. Trends in the Washington State juvenile justice system and preliminary WSIPP evaluation efforts provide early indications of the effectiveness of the state's 6

PAGE 22

comprehensive reform efforts, specifically involving these programs (WSIPP, 2003). Early research findings that support positive effects in King County have led to an expansion of JJIS services to all court-involved youth meeting eligibility criteria, beginning with youth who reside in areas targeted by the county as those having disproportionate representation in juvenile court. Despite some positive indicators of changes in Washington State juvenile crime, including decreases in all types of juvenile crime except drug and alcohol offenses between 1993 and 2002 (Governor's Juvenile Justice Advisory Committee [GJJAC], 2004), certain offense areas have been targeted as problematic. Among these is the increase in juvenile crimes committed by female offenders. In 2002, the juvenile arrest rate was at the lowest point since 1980; however, at a time when the proportion of juvenile males in the state arrested for crimes declined (down 4% from 1996), the proportion of juvenile females in the state arrested increased (up 11% from 1996). In 2002, girls accounted for nearly 30% of all juvenile arrests, representing a 17% increase during the previous decade (GJJAC, 2004). The King County Juvenile Court has directed much of its intervention efforts to those programs proven to intervene successfully with youthful offenders. All three of the interventions (ART, FFT and MST) have demonstrated success with a fairly diverse range of youth either involved in the juvenile justice system or at significant risk of court involvement. Although much of the research and development regarding the ART program has concentrated on male offenders, the court has recently explored 7

PAGE 23

the implementation of a gender-specific ART intervention. The MST and FFT programs, however, have some evidence to support positive program effects in samples including both boys and girls, though none has specifically looked at the programs' effect solely on girls. FFT is a structured family therapy program designed for at-risk and court involved adolescents (aged 11 to 18) with significant family risk factors. The focus of the intervention is to identify family risk factors that can be changed and protective factors that can be enhanced, while both engaging and motivating families to participate in the change process. Like other evidence-based practices, the role of fidelity in implementing the program model is crucial to its success (Alexander, Pugh, Parsons, & Sexton, 2000). All family members attend therapy sessions, which are designed to foster communication, problem-solving, and negotiating skills of the family as a whole. Ultimate aims of the program are to reduce family disorganization and modify chaotic family processes in order to effectively address an adolescent's acting-out delinquent behavior (Carr, 2000). Multiple studies ofFFT have demonstrated effects at reducing recidivism, often persisting as long as five years post-treatment. Overall effect sizes have varied from study to study, but most report statistically and substantively significant differences of sizable magnitude. Studies summarized by Alexander and his colleagues (2000) report reductions in recidivism rates from 25% to 55%, with an average of35%. 8

PAGE 24

MST is a family and community-based intervention program that works with seriously delinquent youth and their families. The program targets risk and protective factors in multiple domains of a youth's social envirorunent (tenned "nested family contexts"): family, school, peers, neighborhood, and support networks (Henggeler, Rone, Thomas, & TimmonsMitchell, 1998). A short, home-based model of intensive services is delivered to youth and their families, and is designed to foster family relationships and parenting practices that will help youth to improve function in their various social contexts (Burns, Shoenwald, Burchard, Faw, & Santos, 2000; Carr, 2000). MST programs have repeatedly demonstrated lower recidivism rates for youth participating in the program than control groups, with effects ranging from 25% to 70% reductions in recidivism (Henggeler et al., 1998). While these two interventions (along with the ART program, which is not included in this analysis) have demonstrated success in reducing juvenile recidivism, these programs have not been tested as gender-specific interventions for female offenders. A body of research is emerging that indicates that girls involved in the juvenile justice system differ from their male counterparts in multiple ways, including their patterns of delinquent activity and both risk and protective factors, i.e., their pathways to delinquency (see, for example, ChesneyLind, 1989; ChesneyLind & Shelden, 1992; ABA & NBA, 2001). In order to successfully intervene with youth who become involved with the King County Juvenile Court, programs must be developed and implemented that 9

PAGE 25

serve the needs of girls, who now make up close to one-third of all new juvenile arrests. Two programs already implemented in King County, FFT and MST, have both theoretical perspectives and clinical practices that are consistent with risk factors and needs of girls involved with the juvenile justice system. This research explores the effects of these two programs on the girls being served by the King County juvenile court. Understanding the types of programs that can effectively intervene with delinquent girls will contribute significantly to the body of knowledge on how to reduce or prevent juvenile delinquency among a population increasing its involvement with the juvenile justice system. 10

PAGE 26

CHAPTER2 RESPONDING TO JUVENILE DELINQUENCY: THEORY, PRACTICE AND GENDER ISSUES The problem of juvenile delinquency has been an important social issue for centuries in America. Theoretical and applied perspectives have fluctuated over the past 300 years, in concert with changes in the social sciences and with shifts in religious and societal perspectives. Traditional scientific and cultural orientations have generally taken a single-cause, single-solution approach to understanding delinquent behavior and, therefore, have largely failed to provide juvenile justice practitioners with many viable intervention strategies. Past theory development and practical research both have provided a great deal of insight into some forms of delinquent behavior, but have not produced a clear and effective course of action for policy makers. This lack of direction for juvenile justice programming culminated most famously in the mid 1970's general frustration that "nothing works" in rehabilitating delinquents (Martinson, 1974). In particular, the general view of juvenile delinquency as a problem typified by inner-city (often immigrant or minority) males has left serious gaps in the knowledge needed to adequately inform policy concerning juvenile delinquency. The resulting system of American juvenile justice is a male-centered one that seems 11

PAGE 27

increasingly ethnically and gender biased and is unable to respond to the needs of a fast growing female offender population. Understanding Female Delinquency: Early Approaches Early studies of juvenile delinquency either ignored the behavior of girls or viewed it in light of the female's role in reproducing a delinquent population. In some cases, girls' delinquency has been addressed tangentially, but none of the predominant social science theories was formulated to apply to females who commit delinquent acts. The significance of gender can be seen in the evolution of delinquency theory and juvenile justice practice over time. Formal scientific study of delinquency has long excluded an examination of girls, but social and cultural opinions on the types of behavior that constitute delinquency, the causes of "delinquent" behavior, and approaches to its prevention and treatment have very much been "gender-specific" (Schlossman & Wallach, 1998), with girls experiencing the American juvenile justice system in a way quite different from their male counterparts. Some early biological and social scientists did directly address the issue of female deviance, crime, and delinquency. As was the case with male delinquency, these explanations were couched in the societal biases, including racism and sexism, which were prevalent in their time. Lombroso' s ( 1911 I 1968) biological explanation of female delinquency, for example, was based on a belief that female personalities were more immature, vengeful, and cruel. These traits, however, were balanced by 12

PAGE 28

maternal feelings coupled with lower levels of passion and intelligence than those possessed by men (Shoemaker, 1990). In short, Lombroso' s theory implied that more "masculine" women and girls (those not having maternal feelings or who were highly passionate and intelligent) were those engaging in criminal behavior. Similarly, Freud ( 1927 /1949) posited that female criminals and delinquents were acting out as a result of their jealousy of male dominance and power. The theories ofLombroso and Freud were typical ofProgressive Era views of delinquency. Scientific studies of evolution and human physiology converged at that time with a growing social concern for an emerging population of urban youth viewed as beyond the control of family authority, requiring intervention of the state (Krisberg & Austin, 1993; Mennel, 1973). Most who study progressive area approaches to the identification and control of juvenile delinquency agree that emergent policies unfairly targeted lower-class children, particularly those from minority immigrant populations (Schlossman & Wallach, 1998). The patriarchal perspective of needing to protect (take control of) delinquent youth had a particularly harsh effect on girls. Stereotypes regarding women, and immigrant women in particular, meant that girls at the time were treated significantly harsher than their male counterparts, often receiving more severe punishments, despite infrequently being accused of violating any criminal statutes at all (Schlossman & Wallach, 1998). Early studies of physical delinquent "types" that dominated the Progressive Era did not include the physical attributes of female delinquents. Much of this is 13

PAGE 29

likely due to practical constraints of a puritan society's view of male scientists studying female body types (Mennel, 1973). However, another biological explanation for human criminal/deviant behavior, rooted in hereditary factors, did have significant implications for girls. Influenced by Mendel's studies of genetics, hereditary theories of delinquency (eugenic theories) were ultimately pessimistic in nature, concluding that reform was not possible (one can't change genetics). These genetic theories, when combined with widespread uses of intelligence tests showing delinquents to be less intelligent than non-delinquents, led the head of one large reform school to demand "permanent institutionalization for such [delinquent] girls, thus assuring ''that this race should end with them; they shall never become the mothers of children who are like themselves" (Charles Goddard, Vineland Training School, quoted in Mennel, 1973: 99). While this view did not completely permeate community and governmental institutions, the sterilization of"degenerate children," many of them girls, became common briefly during the early 20th century (Mennel, 1973). As is evidenced by most criminological theory from the late 1800s to the tum of the 20th century, social prejudice and bias drove scientific theory as much (if not more so) as empirical inquiry and evidence (Mennel, 1973). The dominant puritanical views of the time, however, certainly impacted the perception and treatment of juvenile girls more so than emerging scientific thought. Schlossman and Wallace ( 1998) in their review of juvenile justice and child welfare records of the time reported that girls were frequently labeled delinquent and subsequently 14

PAGE 30

institutionalized for "immoral" behavior. Accusations were often triggered by possible indicators of future promiscuity, regardless ofwhether a girl exhibited any actual problem behavior (Schlossman & Wallace, 1998). This was particularly true for immigrant girls, whose cultural norms and behavior were seen as inappropriate by upper-class Protestant activists and officials. Girls were commonly sent to reform schools for behavior different from their male counterparts, especially for sexual behavior (Mennel, 1973). Many early advocates for female delinquents, including the Chicago Juvenile Protective Association, strongly urged that wayward girls be trained in proper behavior, rather than being incapacitated in institutions (Menne}, 1973). These beliefs were translated into concrete programs in some reform schools, wherein female college students were used to provide education, cultural appreciation, and self-care information opportunities to institutionalized girls. The idea was to provide girls with the ability to leave the institution and to provide for themselves. '"fhis self-sufficiency was stressed based on a perspective that many ofthe problem behaviors were the result of a dependency on others who corrupted them (Mennel, 1973 ). However, gender-specific programming for delinquent girls has generally been the exception, rather than the rule, beginning with Progressive Era reform schools and continuing into the contemporary juvenile justice system of the 21st century. As juvenile justice theory moved from biological and genetics-based perspectives to anthropological and sociological approaches in explaining juvenile 15

PAGE 31

behavior, girls continued to be largely overlooked. Further, as theories became less dominated by the morality issues that had previously led to the incarceration of girls for any kind of behavior considered gender-inappropriate, the consideration of girls even belonging to the group of youth termed "delinquent" nearly disappeared entirely. Sociological theory from the 1920s through the 1960s contains few references to delinquent girls. More modem theories of female delinquency, developed during the 1970s, have focused either on a belief of the inherent sexual nature driving female delinquency (including gender roles and generation tensions regarding sexual morality beginning in the 1960s and 1970s) or on the greater participation of women in the workforce as a result of the successful women's rights movement (Chesney Lind & Pasko, 2004; Shoemaker, 1990). Feminist scholars tend to disagree strongly with these views of female delinquency, citing a complete lack of evidence that there is any relationship between women's emancipation and a growth in female crime and delinquency (Chesney-Lind & Belknap, 2002). These scholars also point out that the historical differential juvenile justice treatment based on gender has persisted. Girls continue to be detained for less serious offenses than do their male counter parts, with some still held for status offenses (ChesneyLind & Belknap, 2002). Recently, more attention has been paid to female delinquents, particularly as their numbers in juvenile detention and other residential institutions have increased. This focus has underscored the gender gap in traditional juvenile delinquency theory. 16

PAGE 32

While a new body of knowledge regarding girls is emerging, it tends to be rooted in the critical theory orientation of feminist scholarship. Much work points out where traditional theory has failed to explain female crime and delinquency, but there remains a great deal of work to be done into creating a unified theory of female delinquency. New approaches to the study of delinquency overall, often called Integrated Theory (Agnew, 2005b) or the Socio-Ecological model (Elliott, 2000) have taken a broader approach to understanding juvenile behavior than have traditional theories. These approaches are built upon research that emerged during the 1990s which focused on the prevention of youth crime and violence by understanding developmental pathways to delinquency and isolating the specific factors associated with delinquent behavior (see, for example, Thornberry, Huizinga, & Loeber, 1995; Hawkins, Herrenkohl, Farrington, Brewer, Catalano, Harachi, & Cothern, 2000). These new perspectives seem to offer a great deal of promise for the examination of female delinquency. Social, Environmental and Ecological Factors in Female Delinquency One of the biggest challenges facing the juvenile justice system is to develop a framework for understanding juvenile delinquency that successfully links the social world of delinquents (both boys and girls equally) with those factors specifically related to their patterns of behavior, and that can then respond with 17

PAGE 33

programs that will address those factors related tore-offending. Traditional approaches have failed to account for the underlying causes of female delinquency (and, arguably, male delinquency as well). Structural and macro-level explanations of crime and delinquency may be useful from a broad pubic policy perspective, but they lack utility in informing specific program decisions concerning the treatment of the individual juvenile offender. A body of delinquency research has accumulated regarding individual, social and ecological characteristics of juvenile delinquency. This work has built upon and integrated many aspects of existing juvenile theory, although it has not always had theory testing as its main purpose. Instead, recent research has focused on describing various aspects of the lives of delinquent youth, as well as on finding patterns and building predictive models that isolate factors which make it more likely that a youth will engage in delinquent behavior. Most recently, Agnew (2005b) has proposed a general theory of crime and delinquency that focuses on the different "domains" of a person's life that give rise to possible causative factors of offending1 These domains include individual personality and temperament factors, as well as social factors such as the family, peer groups, school, work, and the community/neighborhood where one lives. This approach may represent the most promising perspective for a comprehensive understanding of female delinquency that has been put forth thus far in the field of delinquency 1 Agnew's theory pertains to both criminal offending and delinquent behavior. Since this paper is focused on delinquency, that term will be used most often to discuss integrated theories. 18

PAGE 34

research. The focus on specific events and circumstances in the life of the individual is compatible with the feminist view of juvenile delinquency studies that has contended that an understanding of the "social lives" of women and girls has been missing in the study of crime and delinquency (ChesneyLind & Belknap, 2002). Also, this approach is more directly applicable to juvenile justice practice. It allows for an examination of specific factors that influence an individual's delinquent behavior, many of which can be targeted for change within an individualized treatment plan. In the socio-ecological approach, there are a number of general areas (domains) in the lives of a young person where he or she may be at greater or lesser risk for undertaking or continuing delinquent behavior. The key to the approach is to identify specific "risk factors," i.e., events or circumstances in an individual's life which create a greater risk for involvement in delinquency. As more of these factors are present, the risk for engaging in delinquent behavior increases (WSIPP, 2003). There is additional literature which points to "protective factors," i.e., events or circumstances in an individual's life that serve to insulate one from the risk of engaging in delinquent behavior. Different individuals may have differing patterns of risk and protective factors and, therefore, require different treatment approaches. While some of the factors presented here are static, meaning they cannot be changed, many can be addressed through specific intervention strategies. 19

PAGE 35

Individual Personality, Temperament and Social Skills-Related Factors Early work in the area of psychiatry saw many scholars finding a need to move beyond biological explanations of juvenile delinquency and study the "larger picture" regarding these youth-their communities, families and schools-the "total life situations" of youth (Mennel, 1973: 160). Toward this goal, the philanthropic Juvenile Protective Association funded Dr. William Healy to conduct a five-year (1913-1918) study of delinquent youth. Healy, while interested in eugenics, pursued diagnostic work with individual children. Ultimately, his conclusions were that each individual delinquent is unique, with different roots of delinquency, different patterns of behavior, and different needed courses of treatment. Healy's case-study approach documented the genetic, physical, psychological and social characteristics of 1 ,000 repeat juvenile offenders. His sample included both male and female delinquents of differing racial backgrounds. His research and work on individual cases within the emerging juvenile justice system served as the origin of the casework approach to intervention (Menne], 1973). Healy's work was initially intended to identify appropriate treatment interventions for individual delinquents, not general causal processes: [In] "arraying data for the purpose of generalization about relative values of causative factors we experience difficulty. Typing or classifying a delinquent may be an impossible task" (Healy, as quoted in Menne], 1973: 165). Such conclusions resulted in little mention 20

PAGE 36

of the role of gender in delinquency. For the most part, his observations regarding female offenders are limited to noting them as exceptions to a pattern observed with their male counterparts. Healy and his colleague, Augusta Bronner, continued to refine their studies and theories. They eventually concluded that delinquency was generally rooted in poor parenting and advocated the removal of some delinquent youth from their home environments; however, they were equally critical of juvenile institutions that neglected the individual needs of the children in their care (Mennel, 1973). In 1934, an extensive study of delinquent youth treated at Healy's Judge Baker Center found high rates of recidivism (Beard, 1934). The findings pointed more to a failure of the juvenile court (which often did not follow Healy's suggestions for therapy or removal from the home) than of Healy's methods. Nevertheless, he was disappointed in the results of the study and undertook a more careful examination of a small number of delinquent youth. He continued to advocate not only for individualizing treatment of delinquent youth, but also for addressing their social and economic conditions. However, psychiatric treatment became the method of choice for treating a delinquent offender. In agreement with Healy was the popular belief that "the key to understanding delinquency was psychiatric study of the individual delinquent's attitude toward his or her familial environment" (Mennel, 1973: 171 ). Even some ofthe more contemporary studies of individual personality factors as predictors of delinquency have encountered problems similar to those identified by 21

PAGE 37

Healy; that is, causal factors are extensive and sometimes contradictory (Shoemaker, 1990) and do not depict generalized delinquent "types." However, perhaps because individual personality factors are more likely to be affected by intervention strategies (as opposed to structural theories such as poverty or racism), there continues to be wide use of personality and other psycho-social measures in classifying and treating delinquents. Indeed, while sociologically-based theories have gained a great deal more popularity in recent decades, individual personality factors still play a dominant role in the study and treatment of delinquent youth (Shoemaker, 1990). Agnew (2005a) argues that youth who are highly irritable or lack self control are at higher risk for engaging in delinquent behavior. These youth have a low level of internal controls that may inhibit their tendencies toward delinquent behavior and are more sensitive to stressful situations (or social strains), reacting to them more strongly. Other researchers have pointed to a broader cluster of individual personality factors predictive of a higher rate of delinquent activity, including: internalizing disorders, hyperactivity, risk taking behavior, and favorable attitudes towards aggression and deviant behavior (Hawkins et al., 2000). Reviews of studies of these individual factors have found small, but significant, effects for these factors on future delinquent behavior (U.S. Department of Health and Human Services, 2001). Other researchers have pointed to similar antisocial values and beliefs that are linked to future delinquent behavior: minimization or denial of delinquent behavior, negative views of the law or conventional behavior, and lack of empathy (WSIPP, 22

PAGE 38

2003). A validation study of nearly 10,000 youth assessed using the Washington Juvenile Court Assessment (WSJCA) found a significant correlation2 between these individual attitudes and beliefs2 and felony recidivism in juvenile offenders (Barnoski, 2004), with youth showing negative attitudes and beliefs being more likely to engage in subsequent felony delinquency. While few studies of delinquent girls specifically adresses many of the personality factors measured by the WSJCA, there has been some study (Molidor, 1996) of aggressive behavior in girls, generally expressed as a reaction to past victimization. Another important area of individual and personality factors is a youth's collection of social skills, a broader view of what Agnew (2005a) refers to as internal social controls: consequential thinking and problem solving, dealing with conflict and interpersonal relations, monitoring/controlling internal and external triggers of problem behavior, and controlling impulsive behavior (WSIPP, 2003). The WJCRA validation study found a significant correlation between youth social skills and future felony offending, with youth with lower skills being more likely to commit new felonies. Studies of recidivism using the WSJCA tool have found that the instrument (which will be discussed in detail in Chapter 3) does predict recidivism equally well for both girls and boys. However, gender differences in the domains of the instrument 2 Pearson's r is the measure of correlation reported here. 23

PAGE 39

focusing on individual characteristics (Attitudes and Skills domains) have not been reported. Neighborhood, Recreation and Employment Merton's (1938) theory of anomie explains juvenile delinquency in terms of some individuals' lack of access to legitimate means of economic achievement. These individuals engage in illegitimate economic activities in reaction to their relative disadvantage. This theory is similar to social disorganization is that it places emphasis on economic disadvantage and urban poverty as a central social force behind crime and delinquency (Agnew, 2005b). Juvenile delinquents inhabit a social world where economic achievement is the fundamental cultural goal; however, due to uneven distribution of opportunities in education and employment, youth find these goals unattainable by traditional means. Delinquent behavior then becomes an alternate means for achieving economic goals. More recent studies of delinquency have not found clear links between employment and risk for further delinquency (Barmoski, 2004). There seems to be some evidence that employment may provide a youth with a variety of social skills and positive adult relationships and role models that can act as protective factors against delinquency. Also, it is important to note that social changes since Merton's original theory formulation have shifted focus from work to school for most adolescents, and anomie theory applies equally well to blocked educational opportunities which later translate into economic gains. 24

PAGE 40

School Attitudes and Achievement Factors Despite the fact that low educational achievement is highly correlated with female delinquency (Acoca, 2000), many state detention and corrections systems lack services that provide youth with strong links to school (Fox & Lyons, 2003). Youth attitudes towards school and educational achievement are important risk factors in juvenile delinquency, which thus raises issues for community-based intervention. The literature regarding school experiences points to poor school performance and low attachment to school as predictors of delinquent behavior. While the U.S. Department of Health and Human Services (2001) report documents only minor effects of school factors on violent behavior among adolescents, Hawkins and his colleagues (2000) point out that the relationship between academic achievement and serious delinquency is stronger for girls than for boys. Other reviews of the literature on female delinquency (ABA & NBA, 2001) have reached the same conclusion. In their validation study of the WSJCA (Bamoski, 2004), researchers found significant a correlation between both static (historical) school risk factors (r =.13) and dynamic (current) school factors and felony re-offending3 3 Correlations reported are bivariate (zero-order) correlations. The authors also report odds-ratios for each factor's independent contribution to the predictive model, controlling for the other factors. In addition, they report AUC (area under the receiver operator characteristic) to account for the use of a dichotomous dependent variable (recidivism "yes" or "no"). See Barnoski (2004) for a full description. 25

PAGE 41

Relationships and Association with Delinquent Peers In Delinquent Boys, Cohen (1955) linked the lack of economic opportunities to poor school performance and association with delinquent peer groups, particularly in the form of youth gangs. His theory combined parts of social learning theory (the theory that all human behavior is learned from others) and social disorganization explanations of economic strain placed on the lower class. Cohen contended that young males in poor urban areas experience failure in the school system, leading to poor self esteem and perceptions of rejection. Youth, in tum, tend to reject middle class value systems, dropping out of school and forming relationships with like minded (and possibly already delinquent) peers. These groups' resentment of middle class values and opportunities lead them to express their frustration through delinquent behavior. Similarly, Cloward and Ohlin ( 1960) formulated a theory of youth gang delinquency that forged a link between structural disadvantage and individual behavior. In this view, youth accept traditional "middle class" economic goals but reject the legitimate means of achieving them, often referred to as "Differential Opportunity Structure." Essentially, delinquency arises from a "blocked opportunity structure," whereas individual (or gangs of) youth react to the lack of opportunities by emulating values and norms observed in the community. Cloward and Ohlin ( 1960) add to the ideas of anomie and differential opportunity structures by focusing much of their efforts on associations among peer 26

PAGE 42

groups and the way in which individuals mirror their behavior after those closest to them (social learning). In this view of delinquency, peer associations emerge as a significant contributor to delinquent behavior and may be the link between macro level social structures and reasons why different individuals adapt their behavior in the face of strain in different ways. Based on Bamoski's (2004) validation study of the WSJCA, negative peer group relationships stood out as the dynamic factor most strongly correlated with felony recidivism. The only factor with a higher correlation was a youth's criminal history. Many researchers (see, for example, Colvin, 2000) point to differential association and social learning theories to explain how youth who spend most of their time in gangs or in peer groups share their delinquent values. The group serves to reinforce delinquent patterns of thinking and behavior. Much has been made, particularly in the media, about increasing violent crimes and gang membership among girls. Some feminist scholars dismiss this representation and argue that girls have always been a part of youth gang culture (ChesneyLind & Pasko, 2004). However, these scholars emphasize that the nature of the role that gangs play in delinquency differs by sex. Ethnographic studies have pointed to the gang as a means to escape neighborhood or family violence among girls, but as a refuge to deal with societal rejection or failure among boys. In addition, whereas gangs tend to be associated with drug-related crime and violence for its male members, qualitative studies of girls in gangs suggest that gang membership tends to 27

PAGE 43

be more associated with status offenses (such as running away) (ChesneyLind & Pasko, 2004). Family Issues and Poor Parenting Practices Studies of delinquency have always had some degree of focus on the family environment. Progressive Era reformers often advocated for the removal of delinquents from unfit homes, citing the family influence on delinquency (Menne!, 1973). Even structural theories of social disorganization point to the creation of a group of unsupervised and unguided children as a contributing factor to rates of delinquency in a given area (Shaw & McKay, 1942). Agnew's (2005b) integrated theory of delinquency focuses specifically on poor parenting practices as an important contributor to risk for future delinquent behavior. This includes poor parental supervision, negative interactions between children and parents, and child abuse. Where individual factors contribute to poor internal control, negative parenting practices contribute to poor external controls on delinquency. Further, delinquent youth learn anti-social behavior patterns from parents who are abusive or who may support criminal and values (Agnew, 2005b). Parental criminality is a factor that has also been identified in other research as contributing to risk of future delinquency, particularly violent behavior. A recent review of delinquency studies reported that boys having fathers with criminal histories were between two and four times more likely to engage in violent behavior than boys having fathers with no criminal history (Hawkins et al., 2000). Other 28

PAGE 44

studies (also cited in Hawkins et al., 2000) have found no correlation; moreover, no evidence exists for a link between parental criminality and delinquency among girls. Other parental practicesfamily management, discipline, and supervision have consistently been demonstrated to have an impact on future delinquent behavior (Hawkins et al., 2000; U.S. Department of Health and Human Services, 2001). For example, severe or inconsistent disciplinary practices and inadequate supervision put children at greater risk for delinquency. These parenting practices exert moderate or strong effects among younger youth, but only minor effects among adolescent youth, who are beginning to be more greatly influenced by their peers (U.S. Department of Health and Human Services, 2001 ). This is corroborated by the WSJCA validation study which found a significant correlation between family problems and felony recidivism in juvenile offenders (Bamoski, 2004). Family instability and poor parenting practices have been shown to be related to delinquent behavior, exerting stronger effects for girls than for boys (Kroupa, 1988). Other researchers (Kerpelman & Smith, 1999) report that high levels of mother-daughter conflict may lead younger girls to seek relationships with older males as a "safe haven" from their family dysfunction, often resulting in more problematic relations. Underlying family conflict and instability typically represent a stronger delinquency risk effect among females than males (ABA & NBA, 2001). Feminist scholars argue that the relationship between family dynamics and delinquency is different for girls and boys and that gender-specific models of 29

PAGE 45

delinquent behavior be developed (ChesneyLind & Pasko, 2004). Family functioning remains an important factor in delinquency causation; consequently, a consideration of family dynamics would seem to be vital in any program designed to intervene with juvenile justice-involved girls. Alcohol and Drug Use and Mental Health Issues While past victimization is a predictor of general delinquency in all youth (Lipsey & Derzon, 1998), delinquent girls are more likely than delinquent boys to have histories of physical and sexual victimization (Belknap & Holsinger, 1999; Belknap, Holsinger, & Dunn, 1997; Sherman, 1999). Other studies (McCabe, Lansing, Garland, & Hough, 2002) also have found significantly higher rates of psychopathology and familial risk factors among female versus male delinquents. Mental health needs of girls often go untreated due to lack of family means for accessing treatment, stigmatization or denial. In addition, adequate services are lacking within the juvenile justice system (Fox & Lyons, 2003). Many youth also face co-occurring issues, having substance abuse issues in addition to mental health needs. The links between these two factors and delinquency are complex (WSIPP, 2003). Most likely, other factors underlie these problems and are the more causative explanations for youthful offending. The issue of prostitution and its link to substance abuse (in girls who tum to prostitution as a means for supporting the addiction of themselves, partners or parents) critically distinguishes female and male delinquency (Sherman, 2000). Feminist scholars (see, for example, 30

PAGE 46

ChesneyLind, 1995; Gaardner & Belknap, 2002) contend these differences reflect the values and gender roles of a patriarchal and sexist society. Community and Societal Factors The role of society in the creation of delinquents has been the focus of macro level theorists since the emergence of sociology as a discipline. Societal explanations for individual behavior were the foundation of these theories, which contend that there is more to how an individual acts than just biological or psychological factors. In many ways, these theories were the earliest precursors to a socio-ecological perspective on delinquency. Early proponents of social disorganization and social anomie theories explained deviant and delinquent behavior as a function of individual reactions to the greater social world around them (Merton, 1957). Generally, social disorganization explains delinquent behavior in terms of a breakdown of the social order, and the associated informal social controls, of the area in which the delinquent lives (Lander, 1954). Pioneered by Shaw and McKay (1942) in Chicago, social disorganization research examined rates of delinquency, based upon the characteristics of the community where the delinquent youth lived, as opposed to where the offense occurred, and looked at the correlation between rates of delinquency and social structures (Chesney-Lind & Pasko, 2004). Socially disorganized neighborhoods are characterized by residents who have difficulty finding work and economic situations that limit parents' ability to supervise their children, which undermines the 31

PAGE 47

community's ability to control youth. These communities, largely due to economic disadvantage, have residents who neither form close ties in their neighborhood nor experience stable social institutions or community organizations (Shaw & McKay, 1942). This climate of low social control, then, fails to infuse juveniles with positive social values and to monitor them effectively. Shaw and McKay's work, coupled with the derivative efforts of others, has been used in many crime mapping studies that have looked at zones of delinquent behavior. This technique was the foundation of their longer-term intervention strategy, the Chicago Area Project; however, critics of this approach have pointed to an over-reliance on class and economic conditions and a neglect of other important social variables, most notably the effects of culture. Follow up studies have shown that neighborhoods that exhibit many of the characteristics identified by Shaw and McKay may be struggling economically but have strong cultural ties. Many researchers have found that these cultural bonds are sufficient to create an organized community that, essentially, protects itself from the negative effects of economic disadvantage (Shoemaker, 1990). Another important criticism of early social disorganization theory is that the perspective focused on rates of male delinquency, with scant attention paid to female offending; moreover, later case studies focused on delinquent boys, particularly in the context of gang delinquency (ChesneyLind & Pasko, 2004). By focusing their studies on males in economically stressed areas, social disorganization theorists 32

PAGE 48

framed a view of deviant behavior that proved too narrow to provide a comparative view of gender-specific delinquency. Using the Socio-Ecological Perspective in Assessing and Treating Delinquent Girls In some ways, traditional sociological approaches to delinquency have provided little assistance in formulating effective approaches for treating youth once they become involved in the juvenile justice system. For girls, in pa11icular, the lack of gender-specific research on the nature and causes of delinquency has lead to a dearth of juvenile justice programs for girls. A socio-ecological approach to delinquency, focusing on the underlying life factors that put youth at risk for delinquent behavior, may take more effort to adopt and implement, particularly for policymakers; however, it may offer the best hope for intervention programs to succeed. The "nothing works" doctrine, which dominated juvenile justice policy for decades, in part reflected a failure to recognize that delinquency has a complex system of causative factors and that effective intervention requires the targeting of multiple areas of a youth's life. Some emerging programs targeting youth problem behavior are based on this principle and are demonstrating success with both male and female delinquent or at-risk juvenile populations (WSIPP, 2003). This paper examines two programs based on the premise that effective intervention strategies must address those areas of a youth's life that put her at risk 33

PAGE 49

for future delinquent behavior: Multi-Systemic Therapy and Functional Family Therapy. While these programs use the family as the primary location for the delivery of intervention services, they both use models that are based on the belief that youth have needs in multiple domains of their lives. Because these programs have an individualized focus, where intervention strategies are directed at making changes in those areas putting the youth most at risk for delinquency, it seems that these programs may be quite promising gender-responsive interventions. This work explores the degree to which these two programs fit with what is currently known about the needs of female offenders and whether they are effective at reducing their levels of risk and subsequent offending. 34

PAGE 50

CHAPTER3 SYSTEM REFORM: CHANGING APPROACHES TO THE TREATMENT OF JUVENILE OFFENDERS IN WASHINGTON STATE By the end of the 1990s, juvenile incarceration rates were skyrocketing throughout the country. This presented states with the option of either investing in costly new detention and residential treatment facilities or exploring juvenile justice reform efforts that focused on community-based alternatives. At the same time, emerging research held promise that some programs targeting young offenders were successful. Washington State made the decision to use this new information on effective intervention strategies in its statewide reform efforts. The state became one of the first in the nation to embrace a research-based (or evidence-based) perspective in developing juvenile justice policy. In 1997, the legislature passed the Community Juvenile Accountability Act (CJAA), aimed at reducing juvenile crime and recidivism rates in the state through the implementation of strategies and programs with demonstrated efficacy. The CJAA resulted in two major changes to the juvenile justice system in Washington. Both changes were strongly grounded in socio-ecological research regarding risk and protective factors for delinquency, as well as in evaluation research showing efficacy of some intervention programs. The first major change was to develop, test, and implement a comprehensive, empirically-based assessment 35

PAGE 51

instrument that would identify youthful offenders at highest risk for re-offending, isolate the underlying factors directly tied to their offending behavior, and provide suggestions for intervention programs and strategies targeting those criminogenic factors. The second major change was the nature of funding for local court programs. Under the CJAA, the state began a policy of only funding programs "shown to reduce recidivism cost-effectively" (WSIPP, 1999:1 ). In its response to new statewide mandates under CJAA, King County Juvenile Court (which includes the City of Seattle) developed a Juvenile Justice Operational Master Plan (JJOMP). A key component of the JJOMP was the implementation of the Washington State Juvenile Court Assessment (WSCJA), and the use of three intervention programs identified by the WSIPP ( 1999) as effective in preventing recidivism in juveniles involved with the courts. These programs are Anger Replacement Training (ART), Functional Family Therapy (FFT), and Multi-Systemic Therapy (MST). These three programs are now collectively referred to within the court as Juvenile Justice Intervention Services (JJIS). Under the new plan, youth who are referred to the King County Juvenile Court for a delinquency petition are screened using the WSJCA. A preliminary "pre screen" is administered first in order to establish whether a youth appears to be at high, moderate, or low risk for re-offending. Low risk youth may be offered a diversion from the court or, if adjudicated, are placed on low-level probation supervision. Further services for these youth are not typically recommended or 36

PAGE 52

required by the court as a condition of probation. This practice is based on evidence that programs targeting low-risk offenders frequently do not work (Aos, Phipps, Bamoski, & Leib, 200 I) and a belief that further engaging these youth in the juvenile justice system will likely draw those youth more deeply into the system and increase their chances of future problem behavior (WSIPP, 2003 ). Youth who are assessed at a moderate or high risk for re-offending undergo the full assessment process. Drawing information from as many sources as possible, youth probation officers complete items that cover 12 domains of the offender's life (see Table 2.1 for a description of these domains). The domains represent factors which have consistently been demonstrated to predict future youth offending behavior. The purpose of the full assessment is three-fold. It is intended (I) to isolate and target the specific problem areas that put a youth at heightened risk for re offending; (2) to use the identified risk areas, as well as strengths and resiliencies identified, to create a case plan for treatment of the youth, including referrals to appropriate treatment programs; and (3) to monitor the youth's progress in treatment in order to determine whether court interventions are working to reduce a youth's risk ofre-offending (WSIPP, 2003). Results may be used to refer a youth for more in-depth mental health or substance abuse assessments. Also, judges may use the results to recommend specific treatment options or specific program completion as a condition of the youth's probation. For all youth, the probation officer uses the assessment to work with the 37

PAGE 53

youth to establish goals and tasks to be met during the probation period, to develop a comprehensive case plan for the youth, and to refer youth to appropriate programs. Re-assessments may be administered at intervals to monitor a youth's progress towards probation goals. A final assessment is administered after completion of treatment and/or at probation termination to assess changes in risk for re-offending. ( Delinquency Petition Filed in the Court Pre-screen is administered I ( Pre-screen Risk Is Low J Pre-screen Risk Is Moderate or High I Diversion Low Level Probation Supervision Full Assessment Is administered Further Mental Health or Substance Abuse Assessment if Indicated Program Referral Court may offer special courts or programs as terms of probation Complete Reassessment at end of program/services and/or on probation termination FIGURE 3.1: WASHINGTON JUVENILE COURT ASSESSMENT PROCESS 38

PAGE 54

Table 3.1 WSJCA domains and descriptions Domain Description 1. Criminal Official record of youth's past offending behavior. History 2. Demographics Males have slightly higher risk for re-offending than do females. 3. School Includes truancy, school dropout, a lack of commitment to and a lack of commitment to or interest in school. Relationships with positive peers and teachers can be protective factors. 4. Use ofFree Youth involved in both structured and unstructured pro-social activities Time in the community have a lower risk for re-offending. Youth with no interest in pro-social activities have a greater risk. 5. Employment Employment provides youth with important job and social skills that are protective factors against re-offending. Youth who have been employed and who have had good relationships with their employers have a lower risk of re-offending. 6. Relationships A youth's relationship with antisocial peers increases a youth's continued involvement in delinquent behavior. Negative adult role models in the community also increase risk. Positive peer and adult role models reduce a youth's risk for re-offending. 7. Family Parent criminal behavior (including substance use and abuse), family problems, as well as lack of supervision and poor parenting skills represent an environment where a youth is at a higher risk for reoffending. 8.Alcoholand Substance abuse in itself represents a degree of delinquency. However, Drug Use alcohol and drug use seem to be related to other forms of delinquency only to the extent that the use interferes with the youth's social functioning (e.g., performance in school, family and peer relationships, etc.). 9. Mental Health Mental health issues seem to be tangentially related to offending. To the degree that mental health problems result in problems functioning socially, they_ contribute to risk ofre-offending. 10. Attitudes and How a youth perceives his or her behavior contributes to either risk of Behaviors or protection against re-offending. Youth that have anti-social attitudes are generally at higher risk for continuing delinquent behavior. 11. Aggression The extent to which a youth employs aggression, generally or situationally, creates a greater risk for re-offending. 12. Skills The degree to which a youth brings to bear appropriate social skills for dealing with difficult situation affects continued delinquency. Youth with poor impulse control, problem solving and other cognitive skills are at greater risk for future delinquent behavior. Note: Adapted from the WSJCA Manual (WSIPP, 2003). 39

PAGE 55

In March 2004, the WSIPP released a report detailing its validation of the WSJCA (Bamoski, 2004). The validation efforts included more than 10,000 preand full-screen assessments conducted between 1999 and 2001. The study examined the degree to which the levels of risk predicted by the pre-screen instrument accurately projected youths' risk for re-offending, as well as the associations between the full assessment risk domain (scale scores) and subsequent offending. The pre-screen is a shorter version (27 items) of the full assessment and consists of a criminal history and a social history section. The criminal history includes items such as the youth's age at first offense and number of previous felony and misdemeanor filings, along with history of escapes and failure to appear in court. The social history domain includes items on the youth's family history, peer relationships current school attendance and performance, substance use, and mental health issues. Each of the two pre-screen sections is scored separately. During an evaluation of the Washington State Early Intervention Program (Matson & Barnoski, 2003), a scoring matrix was empirically derived that assigns a specific level of recidivism risk based on the two scores. This matrix is shown in Table 3.2. Findings from the validation study indicate that risk level (high, moderate or low), as determined by criminal history risk and social history risk scores generated by the pre-screen, was associated with a youth's subsequent offending. 40

PAGE 56

Table 3.2 Risk levels based on criminal and social history scores Social History Score Criminal History Score OtoS 6to 9 10 to 18 0 to 2 Low Risk Low Risk Moderate Risk 3 to 4 Low Risk Moderate Risk High Risk 5 to 7 Low Risk Moderate Risk High Risk 8 to 31 Moderate Risk High Risk High Risk Note: Adapted from the WSJCA Manual (WSIPP, 2003). The study compared pre-screen scores on the two dimensions to rates of recidivism in the 18 months following the assessment. It found that youth with the lowest scores on both domains had the lowest recidivism rates (7.8%), and those with the highest scores on the two domains had the highest risk for recidivism (37.9%). Table 3.3 Recidivism rates for youth based on risk category Social History Score Criminal History Score OtoS 6to9 10 to 18 0 to 2 (Low) 7.8% (Low) 10.9% (Mod) 11.2% 3 to4 (Low) 8.9% (Mod) 16.2% (High) 20.0% 5 to 7 (Low) 13.2% (Mod) 21.4% (High) 27.8% 8 to 31 (Mod) 22.4% (High) 29.7% (High) 37.9% Note: Adapted from the WSJCA Manual (WSIPP, 2003). 41

PAGE 57

Predictive validity of the assessment was also examined for different ethnic groups and by gender. The findings held across all of the sub-groups included. For both males and females, youth assessed at high risk for re-offending had felony recidivism rates more than three times higher than youth assessed as low risk. However, females had consistently lower recidivism rates than did males, with girls assessed as high-risk actually having a lower felony recidivism rate (18%) than males assessed as moderate risk (23%). So, practically, a "high" risk girl may actually be at a lower risk for recidivism than a "moderate" risk boy. Indeed, girls assessed to be at "moderate" risk for re-offending had the same actual recidivism rates as did boys assessed to be at low risk (13%). Continuing along the logical path, then, there seems to be a risk that fewer girls may be diverted into the lower risk category of youth than would occur if the assessment was more gender-specific. So, in a sense, the court may be over-serving low risk girls and drawing them further in the juvenile justice system than is necessary, as informed by research findings in the area. At this time, the WSIPP has chosen to account for the gender differences by increasing the risk weight of being male on the pre-screen to try to account for the difference, rather than re-norming the tool on the two different populations and using different score cutoffs for risk levels for boys and girls. Another important component of the validation study was to examine the individual domains of the full assessment and their contribution to the prediction of 42

PAGE 58

recidivism. While the correlations were not particularly strong, each of the domains tested did show a significant association with felony recidivism. The study examined the association between specific domain scores. Separate scores are calculated within each domain for static (historical, unchangeable) risk, dynamic (current, changeable) risk, as well as static and dynamic protective factors. The strongest correlation to felony recidivism was with the Criminal History domain score (r =.18), followed by the risk score on the School History risk (r =.13) and on the Current Relationships (dynamic) risk domain (r =.13). Correlations ofr =.10 or higher were also observed on Family History, Current Family, Current School, and Youth Attitudes and Skills risk scores (attitudes and skills are both dynamic factors). Protective factor scores were significantly negatively correlated with recidivism; however, these correlations were weaker (less than .1 0). In addition, the weakest risk score associations were observed in the Alcohol and Drug Use and the Mental Health domains (Barnoski, 2004). While the WSJCA has been vetted as a tool for accurately describing a youth's risk for re-offending (as high, moderate, or low), it has been less thoroughly tested for its second intended purpose, i.e., to provide insight into the type of treatment program most likely to be effective for the youth. Currently, youth are referred to programs based on a match between the risk factors most elevated on the WSCJA and those targeted by the intervention, along with general program referral requirements, such as age and offense restrictions. While program recommendations 43

PAGE 59

continue to be the responsibility of the youth's Juvenile Probation Counselor (JPC), the WSJCA assessment report is an important tool used in making a program referral. Multiple treatment options are available to probation youth. For those with significantly elevated risks in the Alcohol and Drug Use or Mental Health domains, more thorough assessments are conducted and the youth may be referred to outpatient or even to short-term inpatient substance abuse or mental health treatment, if warranted. For other youth who are not in need of mental health or substance abuse inpatient or outpatient treatment (the majority of youth), there are three program options available through the JJIS. The number of programs has been limited because only a small number of programs meet the requirement of "proven practice" that has been set forth. Throughout the state, more programs are being examined and implemented as more research becomes available; however, at this time most youth are referred to the three JJIS programs. Table 3.4 shows the referral criteria for each program. T bl 3 4 JJIS a e r 'bTt t program e 1g11 1 tty en ena Program Criteria Aggression Replacement Training Moderate to High Risk (ART) Aggression Score of at least 1 point Functional Family Therapy Moderate to High Risk (FFT) Family risk factor score of at least 6 Multisystemic Therapy Moderate to High Risk (MST) Family risk factor score of at least 6 Two of these programs, Multisystemic Therapy and Functional Family Therapy have been targeted for study here. Each of the two interventions has a 44

PAGE 60

different set of constructs underlying the program theory and relies on distinct program elements to achieve positive youth outcomes; however they share two important common elements. First, each program is built upon individually-tailored treatment perspectives that thoroughly evaluate and assess a young person's risk/protective factors and family needs and strengths. This perspective is compatible with recommendations for the treatment of delinquent girls, who have often been served using programs that are designed for and tailored to boys. Interventions that target specific needs of the individual can account for the different histories, risk factors and needs of girls involved in juvenile justice. These can then deliver the appropriate treatment and support, rather than a one-size-fits-all generic treatment model likely designed for males and wholly inappropriate for females. The second important commonality is that each program moves beyond a traditional model of intervention or treatment of juvenile delinquent behavior that focuses almost exclusively on the individual offender and individual life and social skills, such as impulse control and anger management. Both programs are embedded in the larger socio-ecological frameworks of individual youth and the "nested social contexts" of young people and their families. The social ecological model, when applied to juvenile justice reform efforts, centers on moving youth out of secure detention and residential commitment settings and providing effective rehabilitation and treatment services in the home and the community. Youth operate in a variety of distinct social domains: individual 45

PAGE 61

(temperament and skills), family, peer groups, school, and the neighborhood/community. Problem behavior is the result of difficulties (risk factors) present within any of these systems or problems in interactions among these systems, e.g., family-school relations (Henggler et al., 1998). This philosophy is at the core of the MST and FFT programs being implemented by King County under the umbrella of the JJIS project. These ecological approaches to intervention diverge somewhat from more traditional rehabilitative models for treating offenders. They were developed largely in reaction to failures of individual-focused programs to effectively prevent the re occurrence of problem behavior in youth. From a social ecological standpoint, individual-based programs are ineffective in that do not consider the entire social and environmental context in which an adolescent functions (including gender-specific histories, risks and needs). Skill-building programs (focusing on anger management, for example) may be able to change an individual's attitude or level of pro-social skills; however, once the individual returns to a delinquent peer group or dysfunctional family environment, the negative influence of that environment quickly erodes any positive benefits of individual treatment because other prominent problems in the young person's life have not been addressed. In addition, both the MST and FFT programs have been thoroughly evaluated and have demonstrated success with general populations of juvenile offenders populations that included both boys and girls (Alexander, et al., 2000; Henngeler et 46

PAGE 62

al., 1998). This empirical validation, combined with the programs' orientations toward individual assessment and treatment, make them promising as effective gender-specific intervention strategies. Functional Family Therapy FFT is designed as a short-term intervention program. Depending on youth level of risk and severity of family problems, the intensity of the program ranges from 8 tol2 hours for lower risk/need youth and families to 26 to 30 direct service hours for more intensive cases. Sessions are usually spread over 3 months. The FFT model relies heavily on evaluation research for its theoretical foundation, clinical practice, and program refinement (Alexander et al., 2000). Finally, supervision and monitoring of program delivery are critical in adherence to the principles and assumptions of the intervention model. The program relies on three specific phases delivered sequentially: Engagement and Motivation, Behavior Change, and Generalization. Each phase has its own distinct goals, therapist role characteristics, and techniques. Each phase also contains an assessment and intervention component, so that the intervention is matched to an individual youth's and her family's needs, while remaining a structured intervention program (Alexander et al., 2000). Specific family and individual youth risk factors for re-offending are targeted within each of the phases. Therapist competence in delivering the FFT model is the central factor in ensuring fidelity to the design. Therapists, in order to properly deliver service, need to 47

PAGE 63

learn specified FFT techniques, accurately assess individuals in the therapy settings, and use the appropriate technique at the appropriate time and situation (Alexander et al., 2000). An important factor in promoting competent service delivery has been the continued involvement ofFFT, Inc., in therapist training and supervision. The program creators specify the following characteristics and qualifications for successful therapists (Alexander et al., 2000: 48): a thorough understanding of the FFT model willingness to maintain therapeutic focus "non-judgmentalness" and "non-blamingness" reasonable intelligence, humor, warmth, and acceptance the ability to structure Master's degree in psychology, counseling, marriage and family therapy, social work or related area (although research demonstrates non-Master's level therapists can provide effective services under close supervision) licensed according to state laws maintain at least a half-time active caseload (5 to 8 families) full-time caseloads should be between 12 to 16 "active" families The program utilizes self-reports, structured progress notes, standardized supervision, and family reports to assure that treatment principles are adhered to within the program. Therapist competence is critical in achieving expected program outcomes. A recent preliminary analysis released by WSIPP (2000b) found that just over half of the FFT therapists delivering the service in Washington State courts were doing so either "competently" or "highly competently." The study found differences in program outcomes, with more competent therapists achieving better outcomes (as 48

PAGE 64

measured by youth recidivism rates); however, significant variations in outcomes within competency categories led the authors to express concerns regarding whether the existing rating system, employed for the study, was sufficiently and correctly distinguishing between competent and incompetent therapists. The authors recommend the development and implementation of a refined rating system for monitoring in the future. Other studies have also demonstrated links between specific therapist behavior in delivering FFT and short-term client outcomes, namely completion versus program drop-out; and in immediate negative reactions during therapy sessions. Robbins and his colleagues (2003) found that differences in individual family members' "alliance" with the therapists were positively associated with likelihood of program mortality. That is to say, if some family members formed an alliance relationship with the therapists while others did not, the family was more likely to drop out of the program. The strength of alliance between therapists and family members was measured using the Vanderbilt Therapeutic Alliance Scale-Revised (completed by observers of sessions). Interestingly, however, overall family alliance with the therapist (that is, all family members forming an alliance relationship with the therapists) was not significantly associated with program attrition. The authors point out the significant applications for these findings during the initial engagement phase of the program, where family members often engage in blaming behaviors and try to get the therapist to side with his/her position. The authors point out that the 49

PAGE 65

therapist's role in reframing blaming behavior and not appearing to take sides during this initial phase can be important in family retention in the program. In an earlier study, Robbins and his colleagues (1996) underscored the use of reframing techniques in overcoming adolescent resistance and negativity during initial therapy sessions. The authors report that only reframing techniques (contrasted with framing, acknowledgement, pacer, or prompts) used following negative statements of the adolescent produced a non-negative response. Measuring Program Effects As reported previously, at least one study has associated FFT processes (namely, therapist delivery ofthe services) with retention of program participants (Robbins, Alexander, Newall, & Turner, 2003). Other evaluations reported participation completion rates for the program. Because the Engagement and Motivation (first) phase is the foundation ofFFT service delivery, monitoring client engagement (via program completion rates) becomes central to the evaluation of the FFT program. A study ofFFT implemented with rural, low-income white families reported an 82% completion rate (Gordon, Graves, & Aburthnot, 1995). A summary of research provided by program developers report 80% to 90% retention rates for FFT families\ even in those families having higher risk youth (Alexander et al., 2000). Other short-term effects reported by a wide variety of evaluations of the FFT program implemented throughout the country (and one site in Sweden) include 4 Not all studies have FFT reviewed reported the actual retention rates for clients. 50

PAGE 66

reductions in the overall units of service provided to families, reductions in the number of foster care placement referrals, fewer out-of-home placements, and higher program completion rates (Alexander et al., 2000). In addition, evaluations of FFT programs have shown positive changes in critical risk areas for juvenile delinquency, including improvements in family dynamics and communication; greater parent involvement; and reduced maternal depression, somatization, and anxiety (Alexander et al., 2000). Multiple studies of FFT have demonstrated effects at reducing recidivism, often persisting as long as five years post-treatment. Overall effect sizes have varied from study to study, but most report statistically significant differences of substantive magnitude. Studies summarized by Alexander and his colleagues (2000) report reductions in recidivism rates of25% to 55%, with an average of35% across the studies. A recent evaluation of FFT programs operating in Washington State juvenile courts (WSIPP, 2002b) found significant reductions only in felony recidivism; however, the issue of identifying competent FFT therapists in order to match outcomes with faithful delivery of services may have confounded the study results. Other positive long-term effects revealed in previous evaluations of the program include fewer subsequent out-of-home placements (post-program) and lower recidivism rates for siblings participating in the program 51

PAGE 67

Multisystemic Therapy (MST) Multisystemic therapy is a familyand community-based intervention program that works with seriously delinquent youth and their families. The program targets risk and protective factors in multiple domains of a youth's social environment (termed "nested family contexts"): family, school, peers, neighborhood, and support networks (Henggeler, Schoenwald, Liao, Letourneau & Edwards, 2002). A short, home-based model of intensive services is delivered to youth and their families and is designed to foster family relationships and parenting practices that will help youth to improve function in their various social contexts (Bums et al., 2000; Carr 2000). MST targets serious, chronic, and violent juvenile offenders characterized as having antisocial, emotional, and behavioral problems. The program is designed for youth ages 12 to 17 at imminent risk for out-of-home placement (Bums et al., 2000). Long-term program effects have been demonstrated for serious youthful offenders and those with significant substance abuse problems (Henggeler et al., 2002), and more recent research has indicated promising effects of MST for adolescent sexual offenders (Borduin, 1999). In major clinical trials ofMST, males have represented the majority of participants; however, in one study, female offenders comprised over 40% of the sample. The program has been demonstrated effective with both African American and Caucasian offenders and with youth from both two-parent and single-parent homes. Youth must have at least one parent figure in the home in order to be eligible 52

PAGE 68

for the program and meet at least one of the following criteria: (1) be considered a violent or chronic offender, (2) be at imminent risk for out-of-home placement, and (3) be diagnosed as substance-abusing or dependent (Henggeler et al., 1998). Due to the intensive nature and associated cost of the program, evaluations interested in efficient service delivery must consider whether the program is indeed reserved for serious offending youth who possess one or more of these characteristics. Treatment fidelity to the MST model is measured based on therapist adherence to the nine core principles of MST (Henggeler et al., 1998: 17): 1) The primary purpose of assessment is to understand the fit between identified programs and their broader systemic contexts. 2) Therapeutic contacts emphasize the positive and use systemic strengths as levers for change. 3) Interventions are designed to promote responsible behavior and decrease irresponsible behavior among family members. 4) Interventions are present-focused and action-oriented, targeting specific and well-defined problems. 5) Interventions target sequences of behavior within and between multiple systems that maintain identified problems. 6) Interventions are developmentally appropriate and fit the developmental needs of the youth. 7) Interventions are designed to require daily or weekly effort by family members. 8) Intervention effectiveness is evaluated continually from multiple perspectives, with providers assuming accountability for overcoming barriers to successful outcomes. 9) Interventions are designed to promote treatment generalization and long-term maintenance of therapeutic change by empowering caregivers to address family members' needs across multiple systemic contexts. Because the therapist is the central figure in the faithful delivery ofMST services, careful staffing, training, and supervision are critical program elements. The 53

PAGE 69

following characteristics are outlined for MST therapists (Henggeler et al., 1998: 2728): Master's-level therapists are desired, although services can be delivered by highly competent Bachelor's-level professionals. High levels of motivation, flexibility, common sense, and "street smarts" are requisites. Caseloads should be between four and six families at any given time. MST staff should be distinct, that is, working to deliver MST services to youth and families and not also providing other program services. Fidelity to the MST model is enhanced through a standardized and intensive training for prospective therapists, along with a set of supervisory practices that include monitoring therapist adherence to the model and therapist skill in each aspect of the assessment and intervention process (Schoenwald, Brown, & Henngeler, 2000). Except in special circumstances, MST supervision occurs in a small-group format: MST "teams" generally consist of three or four therapists. The overall nature, including duration and frequency, is likely to vary depending on the needs of the MST teams. New teams may need more frequent supervisory sessions, while more experienced teams may meet less frequently. Supervisory sessions usually last between one and two hours (Schoenwald et al., 2000). Weekly consultation with a MST expert is an additional important feature of implementing the model with fidelity. Consultation also occurs in a small-group format and focuses on monitoring and support of MST therapists and supervisors, providing guidelines for the implementation of specific treatment modalities, 54

PAGE 70

updating information regarding MST research findings, and identifying organizational or system barriers to the implementation of MST (Schoenwald et al., 2000). Research on the relationship between therapist adherence to the MST model and client characteristics and outcomes (Schoenwald et al., 2003) found links between participant characteristics and immediate outcomes with degree of therapist adherence. Three models were run to test the influence of participant characteristics on adherence: family demographics, youth referral characteristics, and pre-treatment functioning. In the family demographics model, gender and the number of parents in the home were unrelated to therapist adherence. The analysis revealed that the greater the caregiver educational and economic disadvantage, the greater the therapist adherence to the model. It was hypothesized that therapists' perceptions of clients' level of need for assistance (via lack of personal or other resources) may lead to greater focus on the MST model. In addition, caregiver ethnicity was associated with higher therapist adherence, with Caucasian families reporting higher therapist adherence than non-Caucasian families; however, this difference was a reflection of greater adherence being reported when there was a match between therapist and caregiver ethnicity. Youth referral characteristics included two measures: referral source (juvenile justice or other) and reason for referral. The source of the youth's referral to MST was not related to therapist adherence. The study found higher adherence when youth 55

PAGE 71

were referred only for substance use than when youth were referred for both substance use and a criminal offense. There was also higher adherence with youth referred for a status offense than for a criminal offense and substance abuse. In effect, therapist adherence decreased with increased seriousness of referral problems. However, in cases where youth were referred for more serious problems, it was noted that the clinicians were relatively new to the programs. The relationship should be tested to see if it still holds true for more experienced therapists. Youth pre-treatment functioning was assessed using four items (parentreported), with a 5-point Likert scale ranging from 1 =never to 5=almost always (previous two months): youth displayed depressed affect nervousness substance use temper outbursts Parents also reported lifetime number of youth arrests and number of out-of-home placements and completed items from the Child Health Questionnaire (Landraf, Abetz, & Ware, cited in Henggeler et al., 2002) to measure pre-treatment functioning. Youth school functioning was assessed through caregiver reports of the number of school days missed during the previous month and whether youth had been suspended during previous six months. Prior placement variables, youth pre-treatment emotional/behavioral functioning, and school absenteeism were not related to therapist adherence. 56

PAGE 72

Adherence was negatively related to arrests and school suspensions. To some degree, adherence was lower for families of youth with more severe antisocial behavior. Measuring Program Effects MST program retention rates have varied considerably across the programs that have been evaluated, ranging from 63% tolOO% (Henngeler, et al., 2000; Randall & Cunningham, 2003). Demonstrated short-term effects for MST services include improvements in those family and youth risk factors targeted by the intervention, including decreased aggressive behavior and increased family cohesion reported in MST trials for substance-abusing juvenile offenders (Randall & Cunningham, 2003). A review of short-term outcomes reported in the literature include higher retention/program completion rates; lower aggressive behavior and self-reported alcohol and substance use; greater family cohesion; and lower incidence of out-of home placement (Henggeler et al., 1998). MST programs have repeatedly demonstrated lower recidivism rates for youth participating in the program than for control groups, with effects ranging from a 25% reduction in recidivism to a 70% reduction (Henggeler et al., 1998). Other long-term outcomes measured in MST studies include reductions in subsequent arrests for serious or violent offenses, reductions in subsequent arrests for substance-related offenses, and reductions in subsequent arrests for sexual offenses (see Bourdin, 1999) 57

PAGE 73

A Conceptual Framework for Examining Washington State Reforms The general multi-systemic, individual family needs/strengths-based approaches ofFFT and MST implemented in Washington State make them particularly appealing as potentially effective programs for young female offenders. In addition, both programs are nested within a larger system reform effort that assesses an individual offender's treatment needs, works to match those youth to appropriate programs based on those needs, and provides treatment strategies that are relevant within the socio-ecological context of a youth's life. The following general conceptual framework for behavior change within the FFT and MST programs will guide the present evaluation. First, delinquent youth and their families undergo a comprehensive assessment that is designed to identify their specific strengths and needs for change. This assessment drives an individualized treatment plan, where the entire family is engaged in therapy sessions (including only family members or those invited in by the family) aimed at changing the skills and interactions of individual family members to improve overall family functioning. Individual support or therapy augments the family therapy when needed (MST). Families are engaged in the process by helping to create the treatment plan and are motivated to change through a structured engagement process. Therapists deliver services (therapy and support sessions) in a natural environment (in the home, in the community, or at the youth's 58

PAGE 74

school). This includes structural family therapy, behavioral training, and cognitive based sessions attended by the entire family. Progress towards treatment goals is monitored by the therapist and reviewed with the families. Any barriers to progress are identified by the therapist and addressed in sessions with the family. Towards the end of treatment, families learn to generalize their learning from therapy sessions and apply new skills and behaviors to their everyday family processes. As a result of achieving individualized goals of improved family functioning, parents better supervise, monitor, and support the youth, thereby better managing the youth's problem behavior. Because of this improved family supervision and support, youth experience reductions in a number of risk factors for continued delinquency. Based on previous empirical findings and programming recommendations advanced in the literature (see ABA & NBA, 2001; Chesney-Lind & Pasko, 2004; US Department of Health and Human Services, 2001), the additional factors (beyond family functioning) of pro social peer and adult relationships, substance abuse, and mental health issues have been targeted for study here. Better family relationships and parental management, then, should lead to improved social relationships, less drug and alcohol use, and fewer mental health symptoms. When these risk factors are reduced, the chances of the girl continuing to engage in delinquent behavior should also be reduced. Based on this conceptual mode, the present study will assess the efficacy of FFT and MST programs in the treatment of young women involved in the juvenile 59

PAGE 75

justice system within a framework driven by a socio-ecological approach to intervention. Because both FFT and MST seem to be consistent with current knowledge regarding promising gender-specific programming (ABA & NBA, 2001; U.S. Department of Health and Human Services, 2001), it is hypothesized that these two interventions will ( 1) address the unique needs of delinquent girls, (2) be effective at reducing the risk factors for delinquency that seem most salient for young women (i.e., poor family relationships, negative peer groups, unhealthy peer or adult relationships, drug and alcohol use, and mental health problems), and (3) reduce recidivism of delinquent girls by reducing these risks. Research Questions Research Question # 1 a: Do girls referred to FFT and MST programs have significant risk factors for continued delinquency that are targeted by the two programs? Are these factors consistent with what has been put forth in the literature regarding the causes of delinquency and treatment needs for girls? It has been suggested that females have a different set of underlying risk factors for delinquency and, therefore, require different treatment modalities. The degree to which the girls in this study exhibit factors that are consistent with this literature has serious implications as to whether findings from this evaluation may inform practice with delinquent girls, in general. Washington State juvenile justice reform efforts are based on the assumptions that (I) intervention is more likely to be effective when it is directed towards those at highest risk for continued juvenile 60

PAGE 76

justice involvement, and (2) in order to be effective, intervention strategies must target those factors directly related to a youth's offending behavior. Research Question #lb: Do the riskfactors align with current knowledge regarding the issues that put girls at risk for juvenile justice involvement? Research Question #lc: How do girls enrolled in FFT differ from those in MST? H1: Girls enrolled in MST will have been assessed with moderate to severe substance abuse and mental health problems, levels higher than girls enrolled in FFT. H2: Girls referred to the two programs will have at least moderate risk factors in the area of interpersonal relationships (including peer and intimate relationships). Research Question #2: To what extent does delivery of the two programs affect girls' improvement in the above identified risk factors after participating in either FFT or MST? H3 : Girls (participating in either FFT or MST) whose families are successfully engaged in treatment and whose treatment plans are based on their own needs/strengths and consistently monitor progress towards goals will demonstrate improvement in family functioning processes (specifically, parent monitoring, positive support, and intervention with problem behavior). H4 : Girls participating in MST will demonstrate significant improvement (over girls participating in FFT) in risk factors related to peer groups/relationship, positive adult relationships, substance abuse, and mental health issues. 61

PAGE 77

Hs: Girls demonstrating significant change in family functioning will be more likely to show improvements in peer relationships and social functioning. Research Question #3: Are subsequent rates of re-offending related to the observed changes in targeted risk and protective factors? H1: Rates of subsequent re-offending are inversely related to reductions in the targeted risk factors. H8 : Rates of subsequent re-offending are inversely related to improvements in the targeted protective factors (family functioning, peer and social relationships, and substance use/mental health issues). Research Question #4: Do girls participating in FFT or MST have lower rates of recidivism than matched girls who do not participate? H9 : Girls participating in FFT or MST will have lower rates of recidivism than girls who do not participate in one of the two programs. 62

PAGE 78

CHAPTER4 STUDY DESIGN AND METHODS Research Design The research project employs a descriptive analysis of the demographic characteristics, offense profiles, and risks for re-offending for girls being referred to the FFT and MST programs. A pretest-posttest quasi-experimental design, using multiple group comparisons, is used to assess the effects of these programs as they are applied within Washington State juvenile justice reform efforts. In the first stage of the analysis, the two groups are girls referred to the FFT program and the MST program (see Figure 4.1). Xr Orrr x2 XJ Participation in FFT Xr x2 XJ Equivalent group of girls not participating in FFT Xr OMST x2 XJ Participation in MST. Xr Equivalent group of x2 XJ girls not participating in FFT. Note: X1 = Risk factors for re-offendmg at Tlille 1; X2 = Risk factors for re-offendmg at Tlille 2; XJ Subsequent recidivism. FIGURE 4.1 PRETEST POSTTEST DESIGN 63

PAGE 79

The initial descriptive analysis examines the offenses and risk profiles of girls being referred to the two groups. In addition, it looks at their involvement with the court within the context of the larger group of girls involved in the juvenile justice system in King County. It also compares their risk factors to those targeted by the program in order to analyze the degree to which girls are being referred to programs designed to meet their individual needs. Looking at the two groups separately allows for an examination of the degree to which the right youth is referred to a program and the degree to which effects differ across the two programs. In addition, comparisons of program effects are made between those who successfully complete the program and those who are referred but who never enroll in program services or who drop out of the program prior to completion. In order to answer research question #4 (posed in Chapter 3) regarding the attribution of outcomes to the programs, an equivalent group design will be employed, adding comparison groups to the above model. This will allow for a comparison between outcomes for youth participating in the two programs with a group of similar girls who are not referred to either program. The use of an equivalent comparison group is preferred to relying on girls who are referred but do not participate or those who drop out as the only basis for comparison because of underlying biases that may exist between those two groups. 64

PAGE 80

The degree to which the programs are able to engage youth and their families, by getting them enrolled in services and keeping them through successful completion, is important to program outcomes, so comparisons are made in both the shortand long-term outcomes between those youth who are engaged versus those who are not. However, youth and families who refuse to participate in the program or who drop out early may be more resistant to treatment in general; therefore, this group would be expected to have worse outcomes, regardless of the intervention strategy. This necessitates the use of a comparison group of youth not referred to the program in order to better isolate the program's effects, using a group of youth matched on specific criteria to the treatment group. It is important to note that the primary purpose of this design is not to compare the two programs to one another in terms of greater/lesser effect, but to illustrate that the two programs are both hypothesized to have positive effects, but on two slightly different populations. This research design, along with its associated methods, is part of a larger overall evaluation of the JJIS expansion in the King County Juvenile Court being conducted by TriWest Group (who employs this author). The data collected in the larger evaluation will be analyzed here as a separate inquiry regarding program effects specific to girls. Descriptions of the methods below include the structure of the overall data collection process, but refer only to the sample of youth selected for the present study. 65

PAGE 81

Sample All girls in King County, WA, referred to either FFT or MST between January 2004 and June 2006 were identified through the JJIS Referrals Database maintained by King County Juvenile Court staff. Records for referrals made to the two programs in the time period were extracted based on a gender search criteria of "female" (n=186). Further analysis of these files showed that in 13 cases, there were duplicate girls referred to the programs (meaning the same youth was referred more than once during the time period). In 11 of these 13 cases, the most recent referral was the one associated with youth completion of the program and was, thus, the referral used for the study. In two cases, girls had two separate referrals to the programs, both of which were withdrawn before either of the girls engaged in services. The most recent referral was used in the analysis for these two cases. Once referrals for those 173 girls were extracted, the files were analyzed for missing data, particularly for JCN (Juvenile Control Number) and JUVIS number, both of which are important county court and state system identifiers for links to other data sources. Sixteen girls who were referred to the programs were missing a JCN and could not be matched to the court filings based on a search by name and date of birth. These girls were excluded from the analysis, leaving a total referral sample of 157 girls. In order to describe these girls in contrast with others who were not referred to the two programs, as well as to compare the outcomes for girls participating in the 66

PAGE 82

programs to those not referred, risk assessment and offense data for all girls with court filings during the time period were examined. A "filing" is defined as the formal action of the prosecutor's office to pursue a delinquency case in juvenile court. Data on delinquency petitions filed in King County Juvenile Court between January 2004 and June 2006 were obtained through an extract of JJWEB, the primary King County Juvenile Court administrative database. Assessment data were pulled to match the first filing occurring in the time period for each girl. A comparison group of girls was also pulled from these court records. Girls who had filings in the same time period were initially eligible for the comparison group. From this pool, a comparison group (n=203) was chosen as an equivalent comparison group to the treatment groups outlined above. Inclusion in the comparison group was based upon basic demographics (gender, race, and age) and risk profiles similar to those of the girls in the two programs. Risk profiles (risk for re-offending, as determined by the WSCJA) were used rather than the initial offense for inclusion in the comparison because the dependent variables in the study are related to risk for re-offending and actual recidivism. Group differences on initial risk for re-offending would cause significant threats to validity within this research design. This initial selection process over-sampled girls for the comparison group so that exclusions could be made if differences were detected based on seriousness of offense or on demographic variables (race and age). An analysis of the equivalence of the treatment and comparison groups is presented later in this chapter. 67

PAGE 83

Data Collection Data for this study come entirely from existing court records. The King County Juvenile Court maintains three major databases. pertaining to the girls included in this research. The JJWEB database is the primary administrative court database. This electronic system maintains records on all court referrals from law enforcement, filings by the prosecutor's office, case dispositions, and detention admissions and releases. The "Assessments.com"5 database is the electronic system used to administer the WSJCRA. A separate database is maintained by the JJIS program staff to record program referrals, engagements, and terminations. JJIS Program Database Extract The JJIS database is maintained by court staff. This database includes information regarding the youth referred (including state and county court identifiers), as well as date ofbirth, gender, race, and ethnicity. The program also includes information regarding the youth's probation officer (who made the referral), the date the referral was made, dates of program participation and the type of discharge (generally, either successful program completion or drop-out). The JJIS data extract also contains demographic data regarding youth served, including gender, race/ethnicity, and date of birth. Gender is recorded as male or female. Race or ethnicity is categorized using the following categories (as recorded s Assessments.com is a private, for-profit company specializing in web-based data collection and case management systems for juvenile justice (and other) systems. The State of Washington contracts with Assessments.com to maintain the electronic system for administering, scoring, and storing the WSJCA. 68

PAGE 84

by the court: African American, Asian/Pacific Islander, Caucasian, Hispanic, and Native American (including Alaskan Native). Court Filings (Initial/Referring Offenses) Court information for the program sample of girls was extracted from the JJWEB database. This includes offense information, court filing information, and disposition of the filing. Offense information includes date of offense, seriousness of offense (based on a court numerical ranking system), and text description of the offense. The filing information includes the Case ID (used to link tables), the date of the filing and the type of court action. Finally, information on the date and nature of the disposition of the case (e.g., adjudication, deferment, diversion, or dismissal) is recorded. Initial offenses. Initial offenses are defined as the most serious offense for which a filing occurred on the closest date prior to referral to the program (or first filing within the period for comparison group youth). A "filing" represents a formal charge in juvenile court filed by the prosecutor's office (referred to as a delinquency petition in some jurisdictions). In some cases, an initial offense could not be linked specifically to the FFT or MST program referral. In these cases, a filing occurring within two weeks of the program referral was used to define the youth's initial offense. Additionally, offense and filing information for 10 girls either could not be located at all in court records or happened well before the program referral (between 69

PAGE 85

2 and 3 years). The average number of days between the program referral and the court filing closest to the referral was 81 days (median= 78 days). Risk for Re-Offense at Program Entry Baseline data on the risk factors for future delinquent behavior for girls enrolled in FFT and MST were extracted from the Washington State Juvenile Court Assessment (WSCJA), maintained by Assessments.com. According to statewide policy, juvenile probation counselors (JPCs) administer the WSJCA pre-screen to all youth who have a filing in juvenile court. Policy dictates that youth who score a moderate or high risk on the pre-screen then receive a full assessment. [See the discussion of the administration of the WSJCA in Chapter 3 for a complete description of the assessment process. Copies of the pre-screen and the full assessment are included as Appendices A and B, respectively.] The pre-screen assessment contains items related to a youth's past involvement with the juvenile justice system (criminal history) and a selection of items from the full assessment that address the youth's social environment (social history). A numeric score is computed within the pre-screen for criminal history (ranging from 0 to 31) and for social history (ranging from 0 to 18), where a higher score is indicative of a greater risk for subsequent offending. The two scores together are used to determine whether a youth's history and current environment put her at a high, moderate, or low risk for re-offending, based on the matrix reproduced in Figure 4.3. 70

PAGE 86

Criminal History Score 0 to 2 3 to4 5 to 7 8 to 31 Social History Risk Score 0 to 5 6 to 9 10 to 18 Low Low Low Moderate Low Moderate Moderate Hi h Note: This figure is reprinted from the WSCJA Pre-Screen a public domain document produced by the Washington State Institute for Public Policy (2003). FIGURE 4.3 RISK LEVEL DEFINITIONS USING CRIMINAL HISTORY AND SOCIAL HISTORY RISK SCORES The full assessment covers 12 domains of risk and protective factors for youth offending. Up to four numeric scores are computed for each domain, representing Static Risk, Dynamic Risk, Static Protective, and Dynamic Protective factors (WSIPP, 2003). The static and dynamic risk scores represent the degree to which a youth is at risk for future offending (with a higher score indicating greater risk). Static scores are generated from historical items and, thus, are not subject to change by an intervention (e.g., age at first offense and number of previous adjudications). Dynamic scores are generated from assessment items based on the youth's current environment. These scores, when elevated, represent areas that may be targeted for intervention, as they can be changed (e.g., current school enrollment and association with delinquent peers). Protective factors are items identified by the assessment that may help to insulate a youth from the risk of re-offending .. As with risk factors, these factors can be static (graduating from high school) or dynamic (having positive adult 71

PAGE 87

role models and relationships). Table 4.1 shows the score ranges for the 12 WSJCA domains, by type of factor. Table 4.1 Full screen score ranges: 12 domains Possible Score Ranges Domain Static Dynamic Risk Protective Risk Protective 1. Criminal History 0-31 none none none 2. Demographics (gender) 0-1 none none none 3. School 0-5 0-2 0-21 017 4. Use of Free Time none 0-4 0 -1 0-6 5. Employment 0-2 0-4 0-1 0-7 6. Relationships 0-5 0-4 0 -10 0-10 7. Family 016 0-4 0-37 0-22 8. Drugs and Alcohol 0-22 0-8 0-25 0-5 9.Mental Health 0-9 0-5 0-6 0-8 10. Attitudes and Behaviors 0 1 0 I 0-22 017 11 Aggression none none 011 0-8 12. Skills none none 018 0-28 Because the WSJCA data are housed within Assessments.com, a statewide database, records are filed under the state JUVIS control number rather than under the King County Juvenile Court JCN number; consequently, all assessment data matches had to be linked based on the JUVIS number. Unfortunately, the availability and verifiability ofthis number are less than that of the JCN, making matches somewhat difficult. After an initial attempt to link files using the JUVIS number, 42 girls in the program sample (n=157) had no record of assessments. The Assessments.com extract included the youths' first, middle and last names as well as dates ofbirth, so efforts to 72

PAGE 88

match on other identifiers could be made in the event of missing (n=7) or inaccurate JUVIS numbers in the county records. Based on the youth's name and date ofbirth, assessment data were located for an additional 10 girls6 For the remaining 12 girls (8%) in the sample, court and assessment records could not be linked. These cases were dropped from the final sample. Once all of the assessments for each girl in the sample were extracted, the respective pre-screen, initial, and reassessment/final assessments were matched with program participation dates. This was done to ensure that the closest assessments to the program start and discharge, respectively were used. While 96 of the girls had pre-screens, an additional 27 girls who did not have pre-screens available did have full initial assessments that were used to calculate the initial risk levee. Either an initial or pre-screen assessment was available for 130 (93%) ofthe girls in the study sample. Ofthese 130 girls, 102 had both an initial assessment and a post assessment ( 65% of the original sample). In order to determine whether this much missing data would bias the sample, test for significant differences (chi-square) were run for youth who had complete data and those who did not. There were no significant differences between the two groups on race, program status 6 In all 10 cases, both the names and dates of birth matched exactly. ln addition, no cases occurred where there were two youth in the Assessments.com data file with the same name. 7 Because the pre-screen risk score is based on items that are contained within the full assessment, the low, moderate or high risk levels can be calculated from a full WSJCA. 73

PAGE 89

(completion vs. tennination), specific program (FFT vs. MST), the site where the program was delivered, or the initial risk level of the youth (high, moderate, or low)8 Outcome Data: Changes in Risk Factors for Re-Offending and Actual Recidivism Data regarding changes in a youth's risk factors for re-offending are calculated based on the change from initial assessment to post assessment. Postassessment data are taken from reassessments or final assessments completed by the youth's probation counselor either upon completion of the program or completion of the youth's probation. The original data collection plan for the present study was to receive files quarterly from the state legislature-supported research organization, the WSIPP; however, a statewide project to convert data systems resulted in problems in obtaining 2006 recidivism files. After discussions with court staff, the possibility of using county records from JJWEB was considered. The benefit ofusing WSIPP as the source of recidivism data is that it maintains offenses statewide, not just those occurring in King County. Since many youth may commit delinquent acts in surrounding communities, it is advantageous to look at statewide offenses, rather than just those occurring within the county. Still, there are some disadvantages to this data source. Foremost is the need to rely on the 8 Results of the chi square tests are as follows: race, x2=3.52, p>.OS, df=4, n=327; program status, x2=6.02, p>.OS, df=l, n=327; specific program, x2=2.05, p>.OS, df=l, n=327; site, X2=4.80, p>.OS, df=l, n=327; risk level, x2=2.10, p>.OS, df=2, n=327. 74

PAGE 90

statewide JUVIS number to link youth records. Because King County Juvenile Court relies primarily on the JCN for its electronic records, the JUVIS number is more often missing or incorrect for youth than is the JCN. As a result, some recidivism events will be missed by the initial WSIPP data merge because the youths' JUVIS numbers do not match or because of other data errors occurring in the process of transferring the county data to WSIPP, as well as its subsequent cleaning and formatting. In order to determine the degree of differences between the two sources, King County Juvenile Court records and WSIPP recidivism records were examined for a sample of327 youth participating in the MST program between 2001 and 20059 For each youth, all subsequent offenses, as recorded in the King County Juvenile Court database (JJWEB), were compared with all offenses recorded in the WSIPP recidivism files. In each file, the first recidivism event was located (defined as the first offense occurring following referral to the MST program). Using the JJWEB database, recidivism events were found for 165 ofthe youth (50%). Using the WSIPP database, recidivism events were found for 146 youth (45%). In most cases, the groups overlapped; however, as can be seen in Table 4.2, there were some youth missed by each source. A total of 35 youth had recidivism events that were identified in the WSIPP data set but were missed using the JJWEB database. A greater number 9 This sample was taken from a previous analysis ofMST program outcomes, conducted by TriWest Group in 2006 (King County Juvenile Court, 2006). 75

PAGE 91

of youth (53) had recidivism events that were identified in the JJWEB database, but not in the WSIPP database. Table 4.2 Comparison of recidivism, by data source Recidivism Event Recidivism Event found in found in JJWEB WSIPPdata No Yes Total No 127 35 162 Yes 54 111 165 Total 181 146 327 There were no statistically significant differences in race, gender, type of offense, or initial risk level (high, moderate, or low) between the offenses detected in the WSIPP data set but not in JJWEB or between those detected in JJWEB but not in the WSIPP data file10. As a result of these analyses, it was determined that using the King County court filings provides an adequate estimation of juvenile recidivism for youth involved in this study. Recidivism here is defined as any new filing in King County Juvenile Court occurring after engagement in the FFT or MST program and within 18 months of engagement (referral dates are used for youth who are referred to the program but don't receive program services). Studies of juvenile justice programs often use different frames of reference for defining a recidivism event. While there are costs and benefits to each, this analysis 1 Chi square statistics for the tests are as follows: race, X2=3.38, p>.05, df=4, n=327; gender, X2=3.79 p>.05 df=4, n=327; type of offense, X2=4.56, p>.05, df=4, n=327; risk level, x2=2.10, p>.05, df=4, n=327. 76

PAGE 92

attempts to balance more conservative and liberal definitions of recidivism. The typical practice used by WSIPP in evaluating Washington State Juvenile Justice programs is to include only filings resulting in a formal adjudication. A longer period of study is needed when relying on adjudications because ofthe often lengthy court and administrative processes. Also, using adjudications generally means that deferred sentences or court diversions are not included. The use of official filings in juvenile court as a recidivism measure is not as conservative as one relying on adjudications. The advantage of using court filings is that they include cases that end in deferred judgments, court diversions, etc. In this sense, they provide a better picture of further involvement in delinquent behavior than adjudications alone; moreover, a shorter follow-up period of study can be used. Many studies of FFT (Alexander et al., 2000) and MST (Hennegler et al., 1998) have used re-arrest or self-reported delinquency to measure recidivism. Both of these measures have the advantage of casting a wider net and thus having greater sensitivity. However, arrest records are more difficult to match up to court-involved youth and are more difficult to interpret than official court data. Self-report measures are useful but are hard to verify and costly to obtain. The court filing measure here represents a compromise resulting in data that are reasonably obtained and verifiable, as well as useful in describing how the programs impact girls in relation to their experiences in juvenile court. 77

PAGE 93

To create a more robust picture ofre-offending, three separate measures of the re-offense variable were used. First is a dichotomous (yes/no) measure of whether any re-offense occurred. Second is the number of new filings reported (in 6 months, in 12 months, and in 18 months, where possible). The third is a calculation of the number of days between program start and the first new offense for which there was a formal juvenile court filing. Time periods for recidivism are from program engagement (or assessment date for the comparison group) to 12-months and 18-months from the starting date. Again, different approaches to follow-up times for recidivism have been used in juvenile justice program evaluations. The WSIPP recommends an 18-month follow up period to allow enough time for offenders to repeat delinquent behavior (Bamoski, 2004). Evaluations cited by program designers tend to use a 12-month follow-up period (Alexander et al., 2000; Henngeler et al., 1998) or, in some cases, a 6-month follow-up period. The 12-month period has also been used by juvenile justice agencies in other jurisdictions. Statewide recidivism analyses of youth committed to the Division ofYouth Corrections in Colorado, for example, uses a standardized definition of recidivism that includes 12 months post-discharge. This standard was developed following studies showing that the vast majority of youth who do re offend, do so within the first year following release (Division ofYouth Corrections, 2006). 78

PAGE 94

For this study, length of the recidivism period was an important element in gathering a larger sample size by extending data collection to referrals happening through June 2006)\. This means that for some youth only 6 months of recidivism data are available. For youth entering the programs earlier (2004 and early 2005), 18 months of data are available. This study reports rates for each sub-group of youth (based on the time period in which they entered), using the longest time possible. Because the recidivism data file is current through December 2006, 6-month rates are reported for all youth, 12-month rates are reported for youth entering the program prior to December 2005, and 18-month rates are reported for youth entering the program prior to July 2005. The primary dependent variables for the study are: (1) intermediate changes in risk factors related to family processes, relationships, substance abuse, and mental health symptoms, attitudes, and skills, as they are measured by the standardized WSJCA; and (2) the incidence of recidivism. Appendix B contains the full WSCJA assessment and shows all items used to capture these concepts. Broadly, these variables are defined as follows. 1. Family Processes (related to delinquency). These are the mechanisms by which parents monitor and manage a youth's behavior. These include parental supervision, youth compliance with parental authority, parent support networks, and appropriate positive rewards for youth behavior. 79

PAGE 95

2. Positive Social Relationships. These relationships include positive adult relationships and ties with members of the community that discourage youth from getting into trouble. In addition, it includes whether a youth primarily associates with pro-social versus negative peer groups and the nature of romantic relationships. The youth's opinion of the anti-social behavior of negative peers also plays a large role in the area of social relationships. 3. Substance Abuse. This measures the youth's current {past month) use of drugs or alcohol. It includes specific social functioning problems caused by drug or alcohol use and any current (past month) drug and alcohol treatment experiences. 4. Mental Health Symptoms. This measures the youth's current (past month) mental health problems or treatment experiences, specifically as they relate to a youth's social functioning. 5. Attitudes towards Delinquency. Attitudes towards delinquency is defined as how a youth views her delinquent behavior. This includes accepting responsibility for behavior, victim empathy, and respect for authority. 6. Social skills. This measure reflects the probation counselor's assessments regarding the degree to which a youth can recognize and control behavior triggers, deal with conflict situations, cope with life stressors, and engage in realistic goal setting. 80

PAGE 96

Each of the above constructs represents a separate domain on the WSCJA. Each domain has a Static Risk, Static Protective, Dynamic Risk, and Dynamic Protective numeric score. Table 4.1 (page 73) shows the score ranges for each domain of the assessment. Data Analysis Research Question(s) #la: Are the appropriate girls being referred to the two programs? Do girls referred for FFT and MST programs have significant risk factors for continued delinquency that are targeted by the two programs? Are these factors consistent with what has been put forth in the literature regarding the causes of delinquency and treatment needs for girls? How do these girls compare to the overall population of girls involved injuvenile court? Frequency distributions (race) and means (age) of the demographic characteristics of girls referred to the two programs are presented. When testing for mean or proportion differences between the sample and the overall population, a one sample test for mean (or proportion) differences is used (Healy, 2002). The confidence level for all statistical tests is 95% (a= .05; Z critical= .96). Frequency distributions of the type of offenses committed by girls referred to the programs are reported. General descriptive statistics (mean, median, standard deviations) are used to illustrate the severity of risk factors in the four areas operationalized above. These scores are compared to norms for the risk assessment 81

PAGE 97

and are categorized using score cut-offs for low, moderate, and high levels of risk established during the instrument's validation (Bamoski, 2004). These categories are reflective of score distributions across the domains. The low category represents the bottom third of scores; the moderate category represents the middle third of scores; and the high risk category is the top third of scores. Research Question # 1 b: How do girls enrolled in FFT differ from MST? Tests for group differences (independent samples t-test for means; chi-square for frequency distributions) are used to examine any differences in risk factors for girls enrolled in the FFT versus the MST program. Research Question #2: To what extent does delivery of the two programs affect girls improvement in the above identified risk factors after participating in either FFT or MST? Improvements in risk factors are analyzed using paired sample t-tests comparing risk scores prior to program enrollment with those measured six months after completing the program. Difference scores will be correlated with scores on tools measuring therapists' delivery ofthe services according to the program model. Research Question #3: Are subsequent rates of re-o !fending related to the observed changes in targeted risk and protective factors? Change scores (from pre to post) in each risk factor are used to determine the degree to which they can predict the incidence (yes or no) of re-offending using a multivariate regression model (independent variables= change score on each risk 82

PAGE 98

factor; dependent variables= did the girl re-offend (yes/no); the amount of time elapsing between program engagement and offense; and number of new filings in the time frame). Finally, beta weights in the regression model are used to examine the relative effects associated with each of the risk factors. Bivariate correlations between each risk factor change score and the number of repeat incidents are also computed. Research Question #4: Do girls participating in FFT or MST have lower rates of recidivism than similar girls who do not participate? Chi-square analyses are used to compare differences in re-offending among girls in FFT, MST, and the comparison group. Data Limitations In working with existing data sets, cleaning issues and missing data are always problematic. In this case, because the initial sample size was so small, linking up the data sets by hand was possible, so a careful matching of data was acheived even in some cases where identifiers were missing or wrong, thus minimizing some missing data issues. The greatest concern involving missing data lies with post assessments. While court policy dictates that reassessments be conducted with youth at specific intervals (generally no longer than six months) and that final assessments be completed at program discharge, this policy has not been translated completely into practice. Improvements in this area were made over the study period; unfortunately, this means that the later the youth came into the program, the more 83

PAGE 99

likely she is to have a post assessment available. She also will have less follow-up time available in measuring recidivism. Equivalent Comparison Group Design Random assignment to comparison and program groups can eliminate some of the bias issues that arise in cases where external forces influence group membership. While this represents a potential threat to the external validity of the study, it can be managed within an equivalent group design by ensuring the groups are matched based on the independent variables in the study. In this case, a group of girls with similar demographic characteristics, offenses, and risk levels were selected to serve as a comparison group for girls referred to the two programs. While an equivalent group of girls was selected for comparison, outcome analyses on pre-post risk scores and recidivism only include those girls engaging in the programs for whom all the data elements were available. This reduces the sizes of the two samples considerably. As shown in Table 4.3, the comparison group did not significantly differ from the program group girls based on race, age, or risk for re-offending (X2=3.62, df=4, p>.05 for race and x2=1.72, df= 3, n=IOO, p>.05 for risk). 84

PAGE 100

Table 4.3 Group equivalence: Comparison vs. program girls matched for final outcome analysis FFTandMST Comparison Group Group Race/Etbnicity Number Percent Number Percent African American 32 30% 39 39% Asian/Pacific Islander 5 5% 6 6% Caucasian 59 54% 41 41% Hispanic 0 --7 7% Native American 0 --2 2% Other 3 2% 0 --Total 108 100% 157 100% Average Age at Assessment 16.0 years 15.7 years Risk Level High 59 60% 71 66% Moderate 34 32% 34 32% Low 6 6% 3 3% The primary issue with this design is that, because of the nature of the study itself, it is a comparison rather than control group design. In other words, the two programs are being compared with other court interventions (all of the girls in both groups are under probation supervision and receive some combination of court intervention). Girls in the comparison group, for example, may receive short-term drug and alcohol or mental health treatment. In addition, they may have a mentor appointed by the court or may have participated in Aggression Replacement Training (ART). Each girl is under the supervision of a Juvenile Probation Counselor (JPC) who provides individualized case management services -this includes girls in the FFT and MST programs, as well as those in the comparison group. All interpretations of program outcomes must be made within the context of the specific FFT and MST interventions versus traditional court services received by other youth. This study 85

PAGE 101

examines the effects ofthe two programs compared to more traditional (as well as less intensive and less costly) juvenile justice system services. It cannot, however, compare program effects with no intervention at all. There is not sufficient data within the court records to account for what combination of traditional court services each girl receives. Implementation Quality of the Programs As mentioned previously, the extent to which the FFT and MST programs are implemented with fidelity to their models can impact their effects considerably. Regretfully, while all of the teams in King County work closely with the national FFT and MST providers to train staff and monitor their implementation, no fidelity data were made available for this study for two reasons. First, the MST data were not maintained by the program in such a way as to associate a particular therapist's program adherence in his or her work with an individual client (which means it could not be linked with a particular client's outcomes). Some MST teams did report general therapist adherence ratings; however, these data were so inconsistent they were determined not to be useful for this study. The FFT teams declined to provide any data on individual therapist adherence, citing confidentiality concerns. Program adherence can be estimated, qualitatively, based on evolution of the program over time. The time period for this study was chosen to capture the most recent time period possible. The court began implementing FFT and MST programs in 1999 and 2000. In 2003, the WSIPP released findings from a preliminary program 86

PAGE 102

evaluation and reported differential effects related to therapist's adherence to the program model (WSIPP, 2003b ). The WSIPP recommended a set of quality control standards to ensure better deliver of the programs across the entire state. The court has since worked more closely with the program staff who have in tum collaborated with staff from the national centers that certify the local programs, to increase adherence. Informal discussions with court personnel confirm anecdotally that the programs have improved adherence since 2003, but that some room for improvement remains. There are also some broad indicators of program implementation quality that may prove helpful, such as engagement and retention rates and length of service delivery. These data, while not definitive, can provide insight into the fidelity of implementation in the two programs. 87

PAGE 103

CHAPTER5 RECIDIVISM RISK, PROGRAM EXPERIENCES AND OUTCOMES FOR GIRLS REFERRED TO FFT AND MST During the 30 months ofthe study, 157 girls who had filings in King County Juvenile Court were referred to either the FFT {n= 70) or MST (n=77) program. The race and ethnic backgrounds of the girls were not entirely consistent with the general distribution of girls with filings in the King County Juvenile Court. A higher proportion of Caucasian girls were referred to the two programs than is generally observed in the overall population. This difference was statistically significant (Z=2.32, p<.05). The girls' average age at the time of program referral was 15.7 years. This was not significantly different from the overall female population with filings in the court (!l = 15.9 years; Z= 1.80, p>.05). Table 5.1 Demographic comparison of girls referred to FFT or MST versus all female juvenile court filings All Girls Referrals to FFTorMST Race/Ethnicity Number Percent Number Percent African American 1044 39% 52 33% Asian/Pacific Islander 195 7% 6 4% Caucasian 1I58 43% 82 52% 156 6% II 7% Native American I35 5% 6 4% Other/Unknown 7
PAGE 104

Referring Offenses During the 30 months of the study (January 2004 through June 2006) there were 10,159 delinquency petitions filed in King County Juvenile Court. Girls made up 27% of these filings (n=2,756). A total of 1,789 individual girls had a delinquency petition filed during the period, with some girls having multiple filings within the period and others having only one. Just over half (52%) of all filings11 were for misdemeanors, with the remaining 48% for felony offenses. Offenses were divided evenly across person (43%) and property (47%) offenses. The remaining filings were for drug and/or alcohol-related offenses (8%) or other offenses (3%). Table 5.2 below shows the distribution of offenses for filings in King County Juvenile Court involving girls between January 2004 and June 2006. Table 5.2: Distribution of offense types for all juvenile court filings against girls, January 2004 to June 2006 Type of Offense Number of Filinzs Percent Person Felony 593 22% Property Felony 608 23% Drug/ Alcohol Felony 207 8% Other Felony 36 1% Person Misdemeanor 557 21% Property Misdemeanor 648 24% Other Misdemeanor 46 2% Total 2,695 100% While the MST and FFT programs are designed to serve girls at moderate and high risk for re-offending, they are not appropriate for some of the most severe offenders. Girls committing serious offenses are likely to be incarcerated within the 11 In cases where multiple offenses were included in a single filing, the most serious offense is counted. 89

PAGE 105

jurisdiction of the Juvenile Rehabilitative Authority (JRA) because of their greater risk to public safety. In addition, the programs generally exclude youth adjudicated for serious violent felony offenses. As expected, the distribution of offenses for girls referred to the MST and FFT program is different from that of the overall filings for delinquent girls. Just over one-third (35%) of girls referred to the programs had filings for a felony offense directly preceding their referral, compared with nearly half ( 48%) of all female delinquency filings. This difference was statistically significant (Z = -3.26, p<.05). A lower proportion of referred girls had filings for person offenses (38% vs. 43%), but this difference was not significant. Table 5.3 shows a comparison of the distributions of all female delinquency filings with those of girls referred to the MST and FFT programs. Table 5.3 Comparison of filing distributions, all female delinquency filings vs. MST and FFT referrals (January 2004 to June 2006) AU Girls' Filines Pro2ram Referrals Type of Offense Number of Percent Number of Percent Filine:s Filin2S Person Felony 593 22% 10 7% Property Felony 608 23% 30 20% Drug/ Alcohol Felony 207 8% 10 7% Other Felony 36 <1% I <1% Person Misdemeanor 557 21% 46 31% -Property Misdemeanor 648 24% 39 27% Other or Drug Misdemeanor 46 2% II 7% Total 2695 100% 147 100% .. Note: Court Filing data for 10 grrls m the sample could not be located. The "other category mcludes one case classified as a "non-offense." 90

PAGE 106

Girls referred to the MST and FFT programs are not completely representative of the overall population of girls referred to King County Juvenile Court, which is to be expected, given that there are specific criteria for program eligibility. The fact that fewer non-white girls are being referred to the program may be an issue warranting further attention. Because the assessment tool that is used to generate a program recommendations has been validated across different racial and ethnic groups, it is possible that JPC discretion is contributing to the racial differences observed in program referrals. This may be because of a perceived bias about the programs themselves (e.g., a belief that the programs are not culturally appropriate for some groups). However, given the data available, it is impossible to control for the external influences on JPC decision making that may be accounting for this ethnic difference. A much lower proportion of girls in the two programs' samples had delinquency filings for a violent (person) offense than the female delinquent population as a whole". This is consistent with program criteria that tend to exclude most serious violent offenders from eligibility. An analysis of all filings indicated that a higher proportion of African American and Native American youth (30% and 40%, respectively) had filings related to a person felony offense, compared to Caucasian and Hispanic youth (16% and 18%, respectively) who in tum had higher rates of property-related felonies (X2 = 34.4, p<.05, df =6, n=2,695). Differences in program referrals may be due to the nature of the charges that result in court filings. Whether 91

PAGE 107

or not this represents a systemic bias within the court, however, is beyond the scope of this research. As shown in Table 5.4, girls referred to the two programs had filings for a variety of delinquent offenses. By far the largest proportion of girls was involved with the court on assault charges (31 %), followed by theft charges (25%). Smaller proportions of girls (6% in each category) were referred to the court for burglary, criminal trespassing, malicious mischief (generally vandalism/property destruction), and motor vehicle theft. A few girls were referred for either drug possession (3%) or distribution (4%). Table 5.4 Specific offenses for girls referred to FFT and MST Offense Number of Filings Percent Assault 45 31% Burglary 9 6% Criminal Trespass 2 1% Driving While Intoxicated 1 <1% Escape/Obstruction 4 3% Forgery 2 1% Harassment 6 4% Malicious Mischief (prop) 9 6% Motor Vehicle Theft 9 6% Narcotics Distribution 6 4% Narcotics Possession 5 3% Possession Stolen Property 4 3% Prostitution 1 <1% Robbery 3 2% Sex Offense 1 <1% .. Status 4 3% Theft 36 25% Total 147 100% Four girls (3%) were in court for status offense charges; three of those were for alcohol offenses and one for a charge described in court records as "at-risk youth, 92

PAGE 108

contemptnon-offense." One girl was referred on a prostitution charge, while one was referred for a sex offense (child molestation). Regarding general types of offenses, there were no statistically significant differences between girls referred to the FFT program vs. the MST program. However, more MST girls were referred for drugand alcohol-related felony offenses (11% vs. 1 %), while more FFT girls were referred for person felonies (11% vs. 4%). Table 5.5 shows the offense distributions for girls referred to each of the programs. Table 5.5 Offense types: FFT vs. MST referred youth FFT MST Type of Offense Number of Percent Number of Percent Filings Filings Person Felo!!Y_ 8 11% 3 4% Property Felony 14 20% 16 20% Drug/Alcohol Felon_y_ 1 1% 9 11% Other Felony 0 --1 1% Person Misdemeanor 24 34% 23 29% Property Misdemeanor 18 25% 23 29% Other or Dru_g_ Misdemeanor 3 3% 2 3% Total 70 100% 77 100% Risk for Re-Offense As discussed in Chapter 3, the FFT and MST programs have much in common in their treatment approaches, specifically their focus on delivery of communitybased family therapy interventions. There are important differences, however. The MST program is a more intensive program, costing around $4,000 to $4,500 per youth (Henngeler et al., 1998; WSIPP, 2006). The FFT program is shorter and less intensive than MST. It also costs less, averaging between $2,000 and $2,500 per 93

PAGE 109

youth (Alexander et al., 2000; WSIPP, 2006). Washington State reform efforts have been heavily concentrated around the idea of diverting low risk youth from the system onto low-level probation supervision, then matching moderate and high risk youth with the appropriate type and intensity of service. As shown in Table 5.6, the referrals to these two programs seem very much in line with these principles. Table 5.6 Pre-Screen risk levels for youth referred to FFT and MST FFT MST All Youth Risk Number of Percent Number of Percent Number of Percent Level Youth Youth Youth High 27 44% 53 80% 80 62% Moderate 31 50% 14 21% 45 35% Low 4 7% 1 2% 5 4% Total 62 100% 68 100% 130 100% Note: Pre-screen assessments are mtssmg for 17 youth. Overall, the majority of girls referred to the two programs were assessed at high risk tore-offend, based on the pre-screen (see Table 5.7). However, a significantly higher proportion of girls referred to the more intensive MST program was assessed at high risk than for the FFT program (y} = 16.4, p<.05, df=5, n=l30). These moderate and high risk girls were fairly evenly distributed across the higher end scores on both the criminal history and the social history dimensions of the tool. 94

PAGE 110

Table 5. 7 Distribution across pre-screen risk levels (N= 130) Criminal History Score Social History Risk Score 0 to S 6 to 9 l 0 to l R 0 to 2 1% 0% 0% (n=1) L L 3 to 4 0% 10% 5% L (n=13) M (n=S) 5 to 7 2% 15% 19% (n=3) L (n=18) M (n=25) 8 to 31 12% 27% 11% (n=l6) M (n=35) H (n=l4) Note: L=Low, M=Moderate, H=High Only 14 girls (12% of the sample) were scored in the moderate or high category based on a high criminal justice score combined with a low social history score. Another 19 girls (15% of the sample) were scored in the moderate to high category based on a relatively low criminal history score (4 or lower) and a higher social history score (6 or more). Three percent of the girls were low risk. The remaining girls (70%) were moderate or high risk because of higher criminal history M H H H and social history scores. The average scores and standard deviations were similar, as shown in Table 5.8. Both averages fell into the highest risk range. 95

PAGE 111

Table 5.8 Criminal and Social History Pre-Screen scores Std. Upper Score N Mean Deviation Rane:e Criminal History Pre Screen 130 7.72 3.13 8-31 Social History Pre Screen 130 8.33 3.00 1018 Std. N Mean Deviation Sit!. FFT Criminal History 62 6.61 2.70 Program Social History 62 7.44 2.73 MST Criminal History 68 8.74 3.17 t=4.09 p<.OS Program Social History 68 9.15 3.02 t=3.38 p<.05 Girls referred to the FFT program had somewhat lower scores on both the criminal history and the social history domains of the pre-screen. On average, girls referred to the less intensive FFT program had criminal and social history prescreen scores that fell into the moderate risk range, whereas, on average, these scores for girls referred to the MST program fell into the high-risk range for the criminal history domain and just outside the moderate risk range for social history. Because both of these programs are family therapy-based, dynamic family risk factors for re-offending are part of the general referral guidelines for the program. Typically, a family risk score of 6 or higher warrants referral to one of the two programs. In addition, the programs work with youth and families to address other risk areas that may be present within the social ecology of the youth's life. This may include problems in school, relationships with peers and positive adult role modes, substance use, mental health symptoms, and the individual youth's attitudes (including views and use of aggression) and social skills. 96

PAGE 112

Table 5.9 displays the average static risk and dynamic risk scores for gi referred to the programs for 11 risk domains of the WSCJA. Since the second c is gender, this score is constant for all youth in the sample. Average scores are in relation to the score maximums on each domain. In addition, the range of considered high risk, based on population norms, is displayed. Table 5.9 WSCJ average domain scores for all girls referred High Std. Domain N Range Mean Deviation Criminal llistory Static Risk 123 8-31 8.34 3.00 School Static Risk 123 5-5 3.45 1.57 Dynamic Risk 123 7-22 7.37 5.75 Use of Free Time Static Risk 123 n/a .00 .00 Dynamic Risk 123 n/a .38 .49 Employment Static Risk 123 n/a .06 .27 Dynamic Risk 123 n/a .02 .13 Relationships Static Risk 123 2-3 1.94 .62 Dynamic Risk 123 3-8 3.78 1.81 Family Static Risk 123 3-13 4.43 2.37 Dynamic Risk 123 9-34 11.72 5.22 Substance Use Static Risk 123 6-22 7.02 6.54 Dynamic Risk 123 6-24 5.15 6.28 Mental Health Static Risk 123 2-5 2.36 2.06 Dynamic Risk 123 2-4 .33 .78 Attitudes Dynamic Risk 123 4-23 8.52 4.75 Aggression Dynamic Risk 123 3-13 5.67 2.79 Social Skills Dynamic Risk 123 3-18 8.28 5.58 Based on the standard deviations presented in Table 5.9, it is clear that there is significant amount of variation across the domain scores for the girls in the san however the average scores tend to fall into the high risk score range. 97

PAGE 113

Raw scores, even when considered in relation to domain maximums, are difficult to interpret; moreover, comparisons across the domains are impossible because they have different score ranges. In order to make the analysis of specific domains more meaningful, risk scores for each domain area were converted from a numeric value to a "high," "moderate," or "low" risk classifications, based on cutoff scores created by WSIPP after the initial validation of the tool. These normative categories were established based on the more than 20,000 juvenile justice-involved youth in the state who were administered the full assessment between January 1999 and January 2000. It is important to note that these norms are based on a delinquent population, not on the juvenile population as a whole. Based on actual scores for these youth, score cutoffs for "high" (top third), "moderate" (middle third), and "low" (bottom third) risk categories were established. These score norms were not available for either the "Use of Free Time" or the "Employment" domains. Table 5.1 0 shows the percentage of girls in the sample falling into each risk level, for those domains where norms are available. 98

PAGE 114

Table 5.10 Domain risk levels for all girls referred Domain Percent of Youth and 10) Low Risk Moderate Hie:h Risk Criminal History Static Risk 9% (II) 34% (42) 57% (70) School Static Risk 41% (50) 32% (39) 27% (34) Dynamic Risk 30% (37) 12% (15) 58% (7I) Relationships Static Risk 5% (6) 7% (9) 88% (108) Dvnamic Risk 5% (6) 2I% (26) 74% (9I) Family Static Risk 2%(2) I9% (23) 80% (98) Dynamic Risk 7% (9) 2I% (26) 72% (88) Drugs & Alcohol Static Risk 35% (43) I9% (23) 46% (57) Dynamic Risk 47% (58) 16% (20) 36% (45) Mental Health Static Risk 25% (3I) I5% (19) 59% (73) Dynamic Risk 80% (98) I2% (15) 8% (10) Attitudes Dynamic Risk 6%(7) II%(14) 83% (102) Aggression Dynamic Risk I% (I) I4% (17) 85% (105) Skills Dynamic Risk I5% (18) 9% (II) 76% (94) More than three-quarters of the girls referred to the two programs fell into the high risk level in five of the seven domains highlighted in this study: relationships, family, attitudes, aggression, and interpersonal skills. Two other areas highlighted as being of particular importance for female delinquents are substance abuse and mental health issues, with 36% of girls scoring in the high range on the dynamic (current) substance abuse domain and 8% scoring high on the dynamic (current) mental health domain. These findings fit with the two programs' target areas and program criteria; they are not designed to treat severe levels of substance use or mental health problems. Girls presenting such problems would likely need at least some level of inpatient treatment followed by intensive outpatient treatment that is beyond the scope of the programs. 99

PAGE 115

In initial validation studies of the WSCJA, girls assessed at a moderate risk level had recidivism outcomes similar to boys categorized as low risk. Even though the new version of the tool (used for all the girls in the sample) accounts for the gender differences by slightly weighting the risk for male youth, there remained concerns that the moderate risk girls being referred to the program may actually be lower risk than the general populations these programs target. Given the full assessment results presented here, however, this concern does not seem warranted. Girls being referred to the program do show high levels of risk for re offending. This risk manifests in multiple domains in the social-ecology of these girls' lives-their families, their friends, their relationships with other positive adult role models, and their own interpersonal characteristics. School, while not targeted for study, was also a significant risk area for nearly 60% of girls in the sample. The results of the risk assessment indicate that girls follow multiple pathways to their delinquency and that solutions designed to intervene must be multi-faceted in order to adequately address their behavior. 100

PAGE 116

The only significant differences between the two programs in the proportion of girls referred scoring in the high risk range across the domains was in the criminal history domain (x2= 12.43, p<.05, df=1, n=123) and the static school risk domain (X2 = 6.65, p<.05, df=1, n=123). Girls referred to the MST program were more likely to score in the high risk range, than were girls referred to FFT. Somewhat small sample sizes could account for a lack of statistical differences between the two programs, but there is clearly an underlying pattern. Across all but one domain, a higher proportion of girls referred to the MST program fell into the high risk range than did girls referred to FFT, (see Table 5.11). Table 5.11 Percentage of girls scoring in the "High" range, by program Percent of Youth in the High Domain Score Rane:e FFT MST n=59 o=64 Criminal History Static Risk 41%(24) 72% (46) School Static Risk 22%(13) 33% (2If Dynamic Risk 56% (33) 59% (38) Relationships Static Risk 88% (52) 88% (56) Dynamic Risk 64% (38) 83% (53) Family Static Risk 73% (43) 86% (55) Dynamic Risk 75% (44) 69% (44) Drugs & Alcohol Static Risk 42% (25) 50% (32) Dynamic Risk 29% (17) 44% (28) Mental Health Static Risk 58% (34) 61% (39) Dynamic Risk 7%(4) 9%(6) Attitudes Dynamic Risk 82% (48) 84% (54) Aggression Dynamic Risk 82% (48) 89% (57) Skills Dynamic Risk 75o/o (44) 78% (50) On the dynamic family risk domain, a slightly higher (but not significant) proportion of the FFT group scored in the high risk range. Given that FFT is more 101

PAGE 117

centered around issues within the family, while MST broadly addresses multiple issues, with the family as the primary lever for change, this finding seems to support the notion that girls being referred to the two programs are those who match their treatment modalities. Program Experiences of Girls Referred to FFT and MST Out of the 157 program referrals during the period, 21 were withdrawn prior to engagement in program services (defined as a first meeting between therapist and family) (see Table 5.12). All of these were referrals to the MST program. Table 5.12 Reason for withdrawal of referral (MST only) Reason for Referral Withdrawal Number Percent Youth and/or Family Refusal to Participate 5 28% Youth Goin_g_ to Inpatient Treatment 4 22% Youth on the Run (warrant status) 3 16% Family_ Moved 1 6% JPC Doesn't Support 1 6% No room in program 1 6% Other 3 16% Total 18 100% Missing Reason 3 All but one of the youth whose referrals were withdrawn were categorized as high risk for recidivism in the pre-screen. This finding is not surprising given that the majority of reasons for withdrawal were due to youth (or family) refusing to participate in treatment, a youth being unavailable (on the run), or the youth's needs were significant enough to warrant an inpatient-level of treatment. For the remaining 136 girls, the average time from program referral to engagement was just over one month (39 days). On average the time between referral 102

PAGE 118

and engagement was longer for girls in the FFT program than the MST program. The difference was not significant. Figure 5.1 summarizes the referral process. Girls Referred to MST: Referrals Withdrawn 21(27%) / Girls Engaged in MST: 58 (73%) Mean = 33 dayt. Girls Referred to FFT: """'" Referrals Withdrawn = 0 7 Girls Engaged in FFT: 70 (100%) Mean "'44 daYI FIGURE 5.1 REFERRAL TO ENGAGEMENT PROCESS Of the youth who did engage in treatment, nearly three-quarters (74%) ofthe girls completed one ofthe two programs. Table 5.13 shows the difference in completion rates between the two programs. Overall, the differences aren't statistically significant but are large enough that they may be meaningful, progranunatically. Table 5.13 Completion rates for girls engaged in FFT vs. MST FFT MST Type of Offense Number Percent Number Percent Completed Program 52 69% 49 80% Dropped Out 23 31% 12 20% Other evaluations of the two programs have typically reported program completion rates higher than 80% (Alexander et al., 2000; Henngeler et al., 1998), so these rates are somewhat lower than those for programs shown to have been successful in prior evaluation studies. 103

PAGE 119

There were no significant differences in program completion rates across the different levels of risk for re-offense determined in the pre-screen assessment, although all five of the low risk youth referred to the programs completed (statistics not presented). For youth assessed at high risk, 71% completed one of the programs, compared with 75% of moderate risk youth. Average length oftreatment was consistent with the program models (see Table 5.14). On average, FFT youth received services for 4.4 months, just slightly longer than the prescribed length of treatment of3 to 4 months (Alexander et al., 2000). MST youth also had an average length of stay (5.6 months) that was slightly longer than is recommended in the program mode (3 to 5 months). The standard deviation for both groups was 2 months, meaning that many youth had treatment considerably shorter than what is recommended and many went much longer (a few youth had lengths of stay of one year). In examining the length of treatment for girls completing their respective programs, just over half of the girls received treatment for a length of time consistent with their program's model. Table 5.14 Program lengths of stay for girls completing FFT vs. MST FFT (n=Sl) MST (n=49) Mean Std. Mean Std. (in Deviation (in months) Deviation of Treatment 4.4 1.9 5.6 2.3 Percentage of Treatment stays within program recommended rages FFT MST Recommended Length 3 to 4 months 3 to 5 months Pen:entage Percenta e Below/Within/ Above Below Within Above Below Within Above Recommended Ranees 14% 37% 50% 5% 73% 23% 104

PAGE 120

Only 37% of the girls completing the FFT program were served within the target range, compared with 73% of girls completing MST. This difference is statistically significant (X2 = 12.5, df=2, n= 101, p<.05). Across both programs, the majority of girls (and their families) who were not served within the recommended time frames took longer to complete treatment than is typically recommended. Figure 5.2 provides a summary of the program engagement and completion process. These treatment "dosages" are an important part of both programs' theoretical frameworks, which call for intensive treatment provided over a relatively short period of time. Deviations from recommended lengths of treatment are one indication that programs are not being delivered with high fidelity to their respective models. Given that an association between fidelity and outcomes has been demonstrated in previous evaluation studies (Alexander et al, 1998; Schoenwald et al., 2000), program outcomes for girls in King County may not be as good as those that have been shown for other juvenile justice populations served. Girls Refemd to FFT: """"'" Referrals Withdrawn = 0 7' Girls Engeged in FFT: 70 (100%) Meandays Girls Completing FFT: 52(77%) Mean Umonthl Glrls Referred to MST: Referrals Withdrawn """" 21 (27%) / Girls Enpged in MST: 56 (73%) Mean 33 daYII """ Program Dropouts / """ 12 (21 %) Gins Completing MST: 49(79%) Mean 5.6montha FIGURE 5.2 PROGRAM ENGAGEMENT AND COMPLETION PROCESS 105

PAGE 121

Outcomes Changes in Family Functioning Overall, girls participating in FFT and MST did demonstrate a small, but statistically significant, decrease in family function risk scores, from an average score of 11.6 to 10.2 (t=2.86, p<.05) (see Table 5.15). In addition, a slightly larger percentage of girls participating in the programs did show a score improvement (decrease in risk score), although this difference was not significant. While scores decreased both for girls completing the two programs and for those who dropped out prior to completion, the difference for program completers was larger and statistically significant. There were no significant differences in the proportion of girls showing improvement across the groups -either completers versus non-completers or for FFT vs. MST. However, the youth completing the programs overall, and the MST group specifically, had a much higher percentage of girls who did show improvement in their family function (see Table 5.16). 106

PAGE 122

Table 5.15 Changes in family functioning: Program participants versus comparison group Pretest Posttest Sig. Percent Sig. test Risk Score Risk Score Improving for% Diff All program youth 11.6 10.2 t=2.86 40% x2=1.78 (n=93) p< .05 p>.05 Comparison Group 9.2 8.3 t=l.96 37% df=2 (n=91) p> .05 .. Note: Includes all youth engaged (began recetvmg servtces), regardless of whether they completed or dropped out of the program. Excludes youth who were referred to the program, but never began receiving services. Table 5.16 Changes in family functioning, by type of participation Pretest Posttest Sig. Percent Sig. test for Risk Risk Score Improving %Diff Score All program youth 11.6 10.2 t=2.86 40% (n=93) p< .05 x2=t.79 Completers (n=64) 11.3 9.7 t=2.55 42% p>.05 j)< .05 df=2 Non-completers t= 1.28 n=lOl (n=29) 12.1 11.2 p> .05 36% FFT ( completers 11.5 10.4 t=l.3 32% x2=2.92 only) (n=34) p>.05 p>.05 MST (completers t=2.3 53% df=2 only) (n=30) 11.2 8.9 p<.05 The small sample size may contribute to the lack of statistical significance; however, a 21% difference in the number of girls who completed MST improving versus the number of girls who completed FFT showing improvement would likely be a meaningful difference from a programmatic perspective. 107

PAGE 123

Changes in Peer and Adult (non-family) Relationships Girls participating in the two programs demonstrated a small (statistically significant) decrease in dynamic relationship risk scores (see Table 5.17); however, a slightly larger change was observed for the comparison group. A higher percentage of girls participating in the programs showed improvement, but the difference between the two groups was not significant. For girls who completed the program, the change was much larger than for the participant group as a whole (see Table 5 .18). Relationship risk scores were virtually unchanged for girls who dropped out of the programs. In addition, scores decreased for a larger proportion of completing youth. While this difference was not statistically significant, it appears to be of substantive importance given its sheer size (22%). Table 5.17 Changes in relationship risk scores: Program participants versus comparison group Pretest Posttest Sig. Percent Sig. test Risk Score Risk Score for% Diff All program youth 3.6 3.2 t=2.29 32% x2=3.5o {n=93) p<.05 p>.05 Comparison Group t=3.53 df=2 (n=91) 3.4 2.8 p<.05 28% 108

PAGE 124

Table 5.18 Changes in relationship risk scores, by type of participation Pretest Posttest Sig. Percent Sig. test for Risk Risk Score Improving %Diff Score All program youth 3.6 3.2 t=2.29 32% (n=93) p<.05 x2=5.54 Completers (n=64) 3.4 2.8 t=2.75 39% p>.05 p<.05 df=2 Non-completers 3.9 4.0 t= -.. 28 17% =101 (n=29) p>.05 FFT ( completers 3.3 3.1 t=l.8 30% x2=2.92 only) (n=34) p>.05 p>.05 MST ( completers 4.1 3.5 t=2.34 50% df=2 only) (n=30) p<.05 While girls completing the MST program displayed a significant decrease in dynamic relationship risk scores, girls completing FFT did not. In addition, a significantly higher proportion ofMST girls had a decrease in the dynamic relationship risk score. Changes in Substance Abuse and Mental Health Risk Scores While nearly half of the girls participating in the two programs had some level of static risk factors in both the substance abuse and mental health domains, very few had dynamic risk scores in those areas to measure for change. Across the group, baseline dynamic risk scores for the mental health domain were sr low (.38) that little room exists for further reduction. On the substance abuse domain, however, there were some improvements for girls participating in FFT and MST (see Table 5.19). While on average these decreases in risk score from pretest to posttest were of 109

PAGE 125

comparable magnitude to others observed previously, they were not statistically significant, largely due to the extremely wide distribution of both pretest and posttest scores. Girls in the comparison group had significantly larger improvements in this domain. However, it should be noted that neither the MST nor the FFT program targets substance abuse, specifically12 Many of the girls in the comparison group may have received interventions that focused exclusively (or nearly exclusively) on their substance abuse issues, which may explain large improvements in this area, with slightly smaller improvements in other, related, domains. Overall, youth completing their programs showed significant improvement on the substance abuse risk domain compared with girls not completing their program. This is true across the average improvement for the groups as well as the higher proportion of girls in the completion group showing improvement. Between the two programs, significant improvement was seen for the MST girls only. Again, the FFT program is not designed, specifically, to address this issue. 12 Some trials ofMST, however, have shown promise for intervening with substance abusing youth and in reducing levels of substance abuse (Henggeler et al., 2002). 110

PAGE 126

Table 5.19 Changes in substance abuse risk: Program participants versus comparison group Pretest Posttest Sig. Percent Sig. test Risk Score Risk Score Improvin for% Diff All program youth 4.6 3.3 t=1.90 32% (n=93) p>.05 x2=3.31 Comparison Group 4.3 2.4 t=3.16 27% p>.05 (n=91) p<.05 Table 5.20 Changes in substance abuse risk scores: By type of participation Pretest Posttest Sig. Percent Sig. test for Risk Risk Score Improving 0/o Diff Score All program youth 4.6 3.3 t=1.90 32% (n=93) p>.05 x2=7.39 Completers (n=64) 4.8 2.8 t=2.l7 38% p<.05 p<.05 df-=2 Non-completers 4.2 4.5 t=-0.26 21% (n=29) p>.05 FFT ( completers 3.3 3.09 t=l.8 35% x2=3.54 only) (n=34) p> .05 p>.OS MST (completers 4.1 3.5 t=2.34 40% df-=2 only) (n=30) p<.OS Changes in Attitudes Towards Delinquency Girls participating in the two programs demonstrated a decrease in attitudes towards delinquency risk scores, from an average score of8.5 to 7.3 (see Table 5.21). A similar change was observed in the comparison group. Additionally, the percentage of girls in the two groups who showed some improvement in this domain is the same. For girls who completed the program the change was much larger than for girls who dropped out of the program. In addition, scores decreased for a larger proportion of 111

PAGE 127

completing youth than for non-completing youth (41% and 24%, respectively). This difference was not statistically significant, but was of considerable magnitude (17%). Girls completing either FFT or MST had small but significant decreases in attitudes towards delinquency risk scores. While not significant, a substantively higher percentage of girls completing MST showed improvement over girls completing the FFT program, 47% vs. 35%, respectively. Table 5.21 Changes in attitudes toward delinquency risk scores: Program vs. comparison group Pretest Posttest Sig. Percent Sig. test Risk Score Risk Score Improving for 0/o Diff All program youth 8.5 7.3 t=3.01 36% (n=93) p<.05 no diff Comparison Group 7.6 6.1 t=3.89 36% (n=91) p< .05 Table 5.22 Changes in attitudes toward delinquency risk scores, by type of participation Pretest Posttest Sig. Percent Sig. test for Risk Risk Score Improving % Diff Score All program youth 8.5 7.3 t=3.01 36% (n=93) p<.05 x2=2.47 Completers (n=64) 8.1 6.7 t=3.12 41% p>.05 p<:: .05 df=2 Non-completers 9.4 8.9 t=.75 24% (n=29) p>.05 FFT ( completers 8.15 7.19 t=2.13 35% x2=5.52 only) (n=34) p<.05 p>.05 MST (completers 8.98 7.41 t=2.58 47% df=2 only) (n-30) p<.05 112

PAGE 128

Changes in Attitudes Toward Aggression Girls participating in the two programs demonstrated a decrease in attitudes toward aggression risk scores, from an average score of 5.8 to 5.0 (see Table 5.23). However, a similar change was also observed in the comparison group (from 4.8 to 4.0). Additionally, the percentage of girls in the program and comparison groups who showed some improvement in this domain is similar. Girls who completed the programs had changes that were significant. Girls dropping out had smaller changes that were not statistically significant. In addition, there was observed improvement for a slightly larger proportion of completing youth. This difference was not statistically significant. Girls completing either FFT or MST had significant decreases in attitudes towards aggression risk scores. While MST realized a larger average change in risk scores, FFT had a higher percentage of girls showing improvement on the aggression risk domain; however, the difference was not statistically significant. This finding may indicate that FFT, as a lower intensity program, can affect change on the aggression domain, however, the magnitude of the effects are greater for the higher intensity MST program. 113

PAGE 129

Table 5.23 Changes in attitudes toward aggression risk scores: Program vs. comparison group Pretest Posttest Sig. Percent Sig. test Risk Score Risk Score lmprovin2 for 0/o Diff All program youth 5.8 5.0 t=3.01 31% (n=93) p<.05 x2=2.47 Comparison Group t=3.35 p>.05 (n=91) 4.8 4.0 p<.05 28% df=2 Table 5.24 Changes in attitudes toward delinquency risk scores, by type of participation Pretest Posttest Sig. Percent Sig. test for Risk Risk Score Improving %Diff Score All program youth 5.8 5.0 t=3.01 31% (n=93) p<.05 Cornpleters (n=64) t=3.12 x2=2.47 5.6 4.8 p<.05 33% p>.05 Non-cornpleters t=l.67 df=2 (n=29) 6.3 5.8 p>.05 28% FFT ( cornpleters 5.6 5.0 t=2.24 35% only) (n=34) p<.05 x2=.32 MST (cornpleters t=3.79 p>.05 only) (n=30) 9.6 6.5 p<.05 30% df=2 Changes in Social Skills The single largest average change for girls participating in the two programs was the improvement in their social skills (i.e., a decrease in social skills risk scores). Girls participating in the programs improved, on average, from 8.1 to 5.6 (see Table 5.25). While comparison group girls also realized a statistically significant change, the decrease was smaller. Further, 42% of the program girls improved their social 114

PAGE 130

skills, compared with 30% of comparison girls. While this difference is not statistically significant, it is of considerable magnitude. Girls who completed the programs had changes in social skills that were significant. Girls dropping out had smaller changes that were not statistically significant. In addition, there was observed improvement in social skills for a slightly larger proportion of completing youth, though this difference was not statistically significant. Table 5.25 Changes in social skills risk scores: Program vs. comparison group Pretest Posttest Sig. Percent Sig.test Risk Score Risk Score lm__I!rovin_g for% Diff All program youth 8.1 5.6 t=4.78 42% (n=93) p<.05 x2=3.19 Comparison Group 5.7 4.5 t=2.66 30% p>.05 (n=91) p<.05 Table 5.26 Changes in social skills risk scores, by type of participation Pretest Posttest Sig. Percent Sig. test for Risk Risk Score Improving 0/o Diff Score All program youth 8.1 5.6 t=4.78 42% (n=93) p<.05 Completers (n=64) 7.6 5.1 t=4.18 45% x2=2.47 p<.05 p>.05 Non-completers 9.2 6.7 t=2.40 35% (n-29) p>.05 FFT ( completers 7.3 5.3 t=3.43 47% only) (n=34) p<.05 x2=t.32 MST (completers 6.5 t=3.79 43% p>.05 only) (n-30) 9.6 p<.05 115

PAGE 131

Girls completing either FFT or MST had statistically significant decreases in social skills risk scores, although the MST group did have a larger change. Again, even though the average change was small, a slightly higher proportion ofFFT girls did have an improvement. Summary of Changes in Risk for Re-Offending Girls participating in the programs had improvements in risk scores across all ofthe domains measured here, with the exception of substance abuse (although MST completers showed statistically significant improvement); however, all of the changes observed were relatively small. Comparison group girls also showed improvements in all of the domains except for family functioning. This finding is expected given that the two interventions being tested are family therapy-based and the family risk domain is the one most consistently addressed by both programs. In addition to the improvements in family risk scores, the improvements in attitudes towards aggression and in social skills for girls participating in the program stand out. While comparison group girls also had significant changes in these areas, the magnitude of the changes was much larger for the participating girls. Other patterns emerged within the more specific program groups. While many program changes may not have been statistically significant (which may be due to small sample sizes), the repetition of the same patterns across all ofthe risk domains do provide some important programmatic information. First, youth completing the 116

PAGE 132

program consistently showed greater improvements over those youth who dropped out prior to completion. This finding is likely a spurious one. Youth who complete the program are likely to be more compliant with treatment demands and, overall, more amenable to positive change than are dropouts. The comparisons are interesting to make, given that these two programs have relatively low rates of engagement (when compared to other FFT and MST studies). More intensive efforts to engage youth could produce fewer dropouts and better outcomes. On the other hand, it is unrealistic to expect that every youth who engages in the program will finish successfully. Finally, there was a consistent pattern that changes of a greater magnitude were observed for the MST girls (completing the program) than for their FFT counterparts. Again, the MST program is a more intensive one, targeting higher risk youth. For those scales where it seems the programs are having the greatest effects (family functioning, attitudes towards aggression, and social skills), similar proportions of FFT and MST girls showed improvements. The changes for the MST girls, however, were larger. Recidivism Outcomes The rates of youth with new filings in King County Juvenile Court within six months of their program start (or study start date for comparison group youth) were similar for both program and comparison group youth (see Table 5.27). The percentage of youth with a new filing was slightly higher for program participants 117

PAGE 133

(33%) than for comparison group youth (27%); however, this difference was not statistically significant. Table 5.27 Six-month recidivism rates: Program vs. comparison group New Filings (6 months) Program Group Number of Percent Sig. Youth Program Youth 36 33% x2=.9I Comparison Group 28 27% p>.05 df=l Rates for six-month recidivism, particularly, are affected by program dropouts. Youth dropouts had program recidivism rates that were considerably higher than those youth who completed the program (see Table 5.28). Again differences were not significant, likely to due to the small size of the dropout group; however, the difference between dropouts and completers is certainly substantial. As mentioned previously, the youth dropping out of the program are likely predisposed tore-offending. Excluding them from the analysis artificially lowers the program "failure" rate by removing those who fail sooner. Table 5.28 Six-month recidivism rates, by program completion New Filings (6 months) Program Group Number of Percent Sig. Youth Program Dropouts 15 46% x2=3.14 Program Completers 21 28% p>.05 Comparison Group 28 27% df=2 118

PAGE 134

Recidivism rates were consistent with the initial overall risk level, with low risk youth having the lowest rates of six-month recidivism and high risk youth having the highest rate (see Table 5.29). The differences were significant (X= 18.6, p<.05). Table 5.29 Six-month recidivism rates, by risk level New Filings (6 months) Risk Level Number of Percent Sig. Youth High Risk Youth 53 36% x2=18.6 Moderate Risk Youth 16 23% p<.OS Low Risk Youth 1 11% df=2 Of the youth in this sample, all but one recidivism event (when counting the first event) occurred within 12 months of the program start date; therefore, 12-month recidivism rates are used for the remainder of this analysis. Twelve-month recidivism rates were very similar to those for the six-month time period (see Table 5.30). Girls participating in the programs had recidivism rates that were slightly higher than the comparison group, although the difference isn't significant. Table 5.30 Twelve-month recidivism rates: Program vs. comparison group New Filings (12 months}_ Program Group Number of Percent Sig. Youth Program Youth 29 37% x2=.91 Comparison Group 32 31% p>.05 df=1 119

PAGE 135

Youth dropping out of the program had much higher rates of recidivism than did youth who completed the program. Table 5.31 Twelve-month recidivism rates, by program completion New Filings (6 months) Program Group Number of Percent Sig. Youth Program Drop Outs 10 53% x2=s.o2 Program Completers 19 32% Comparison Group 32 31% p<.05 Program participants tended to have a shorter length of time to their first recidivism event (3.9 months) than did youth in the comparison group (6.7 months) (see Table 5.32). The program group's average time to recidivism is actually slightly lower than the average program length of stay ( 4.4 months). The program group, on average, did have a slightly lower number of filings occurring in the 12-month recidivism follow up-period (see Table 5.33). This difference was not statistically significant. Table 5.32 Time to recidivism, by program group Std. Std. Error Group N Mean Deviation Mean Program Group 32 3.9083 4.45149 .78692 Comparison Group 46 6.6722 5.17697 .76330 Table 5.33 Number of new filings, by program group Std. Std. Error N Mean Deviation Mean Program Group 78 .83 1.362 .154 Comparison Group 105 1.oo i 1.468 .143 120

PAGE 136

High risk youth had higher rates of recidivism than did moderate or low risk youth. Moderate risk youth actually had lower recidivism rates than did low risk youth, but this is based on a small number of low risk youth whore-offended (see Table 5.34). More importantly, 12-month recidivism rates for the two programs were identical (see Table 5.35). Table 5.34 Twelve-month recidivism rates, by risk level New Filings (6 months) Risk Level Number of Percent Sig. Youth High 31 50% x2=3.53 Moderate 10 31% p>.05 Low 3 38% df=2 Table 5.35 Twelve-month recidivism rates, FFT vs. MST New Filings (6 months) Program Number of Percent Sig. Youth FFT 16 37% x2=.65 MST 13 37% p>.05 df=l Recidivism Outcomes Summary Overall, the small improvements on risk domains for the youth did not translate to success in terms of new filings in King County Juvenile Court. The program group youth did not show improvement on any of the recidivism measures over the comparison group of youth. As expected, youth who dropped out of the program had worse outcomes than did youth who completed the program. There were few differences in longer-term recidivism rates for moderate or high risk youth. In 121

PAGE 137

addition, there were no significant differences between the two programs on 12month recidivism rates. The following analyses examine the relationships between those changes in risk factors and recidivism outcomes. Factors Associated with Improved Recidivism Outcomes In order to reduce the number of variables used in regression equations (due to sample size limitations), the variables used for the analysis are restricted to those where the largest improvements were observed for youth participating in the programs and those in the comparison group: family functioning (program group only), attitudes towards delinquency, attitudes towards aggression, and social skills. Initially, bivariate correlations between these factors and 12-month recidivism were run in order to examine the relationship for each variable independently. As shown in Table 3.36, the variables representing change scores on each of the risk factors were all highly correlated with one another; however, the only risk change that was associated with the number of recidivism events was the change in attitudes towards aggression. 122

PAGE 138

Table 5.36 Correlation matrix, means and standard deviations for variables in regression analysis (n= 146) x. x2 x3 Xs Xa Family Functioning 1.00 .573 .463 .355 .017 X2 Attitudes Towards Delinquency 1.00 .702 .575 .113 X3 Attitudes Towards Aggression 1.00 .410 .220 Social Skills 1.00 .124 Xs 12-Month Recidivism 1.00 Mean .80 .63 .52 .55 .34 Standard Deviation .75 .70 .68 .65 --Note: r 2:.17 significant at p<.05. Because recidivism is a dichotomous variable (l=yes), the mean represents the proportion of youth re-offending. The variables were entered into a stepwise logistic regression analysis to examine the degree to which these change scores could predict recidivism outcomes. Taken together, changes in these factors did not significantly predict 12-month recidivism rates. The lack of statistical significance, however, is in part attributable to the ratio of variables to sample size. Within the model, only change on the attitudes towards aggression domain was a significant predictor. There was no additional predictive power in the model when the other variables were introduced; therefore the stepwise regression (condition) ended after the first iteration. The model did not change when program versus comparison group membership was added. 123

PAGE 139

Table 5.37 Logistic Regression results Exp 8 S.E. Wald df Sia. (B) Step Attitudes 1.27 1(a) Towards .246 .088 7.762 1 .005 9 Agrression Constant -.403 .167 5.834 1 .016 .668 Score df Sia. Step 1 Variables Family Functioning .013 1 .908 Attitudes Towards Delinquency .299 1 .584 Social Skills .821 1 .365 Overall Statistics 2.105 3 .551 The same pattern held true for the relationship between changes on these risk domains and the number of recidivism events within the 12 month period and for the relationship between risk factor change and time to first recidivism event. The overall regression model using changes in risk factors to predict number of events was not significant (F=2.25, df= 1, n= 101, p> .05). Within the model, the only individual risk change that was significant was attitudes towards aggression. This effect remained stable when program versus comparison group membership was added to the model. 124

PAGE 140

CHAPTER6 DISCUSSION AND CONCLUSIONS Washington State has recently been at the national forefront of community based juvenile justice reform. Its efforts in implementing research-based strategies for identifying and intervening with youth most at risk for re-offending have produced, generally, many positive results; however, the research base and resulting juvenile justice policies have not specifically included an examination of how to best intervene with a growing population of delinquent girls. This study examined the King County Juvenile Court's use ofFFT and MST programs with this specific population, within the larger framework of the overall Washington State model. Of particular interest was the assessment of the girls' risk for re-offending, the observed changes. in those risk factors following program participation, and the association between changes on those risk factors and re-offending. Discussion: Findings for Each Research Question Research Question(s) #]a: Are the appropriate girls being referred the two programs? More specifically, how do these girls compare to the overall population of girls involved in juvenile court? Do girls referred for FFT and MST programs have significant risk factors for continued delinquency that are targeted by the two 125

PAGE 141

programs? Are these factors consistent with what has been put forth in the literature regarding the causes of delinquency and treatment needs for girls? An important part of Washington State community-based juvenile justice reform is matching the right risk level of youth to the best intervention, depending on needs. This includes diverting low risk youth from the system then matching moderate risk and high risk youth to the right levels and types of treatment. In terms of the girls referred to these two programs between January 2004 and June 2006, this seems to be working for the court. With few exceptions, girls referred to the programs were charged with fairly serious acts of delinquency (generally felony and misdemeanor offenses). Nearly all of the girls referred to the program were initially assessed at either a moderate or high risk tore-offend. In addition, girls were scored in the high risk area tore-offend across multiple socio-ecological domains, illustrating that their pathways to delinquency are multiple and complex. During the initial validation of the WSCJA, girls assessed at a moderate risk for re-offending had recidivism rates that were the same as those for low risk boys. An adjustment was made to the pre-screen to more heavily weight the risk score for male youth. A sizeable proportion of girls Uust over one-third) referred to the FFT and MST programs were assessed as moderate risk for re-offending. This led to a hypothesis that these girls' risk levels may be somewhat overstated, making them less appropriate for the programs. In looking at full risk assessment score distributions, however, it appears that, as a group, the youth referred to the programs showed high 126

PAGE 142

levels of risk across multiple domains of the full assessment, particularly on those factors that tend to be targeted by the programs: family, relationships, attitudes towards delinquency and aggression, and social skills. This is an encouraging finding given that juvenile justice systems often have targeted girls for supervision and services who had committed minor delinquent acts or status offenses. The King County Juvenile Court does seem to be delivering its more intensive services to youth with higher needs who seem to have a genuine risk for further delinquent behavior. The score distributions across the risk domains measured in the WSJCA track with the overall risk levels assigned by the pre-screen. These findings indicate that program referrals represent a good match between girls' needs and the programs' intervention goals. In addition, the differences between FFT and MST program referrals show that girls at higher risk are those being referred to the more intensive MST program. Family risk factors were more concentrated for girls being referred to FFT, whereas a larger proportion of girls referred to MST had scores falling into the high risk range across the remaining WSCJA domains. Risk levels across the WSJCA domains were mostly consistent with many of the factors that have been cited in the theoretical literature regarding female delinquency (ABA & NBA, 2001; ChesneyLind & Pasko, 2004). Girls had high risk scores on the family and relationships domains, as expected; however, the proportion of girls scoring in the highest risk range (more than three quarters of all of the girls referred to the two programs) on the attitudes towards delinquency, attitudes towards 127

PAGE 143

aggression, and social skills was expected, although the magnitude of these issues was somewhat greater than anticipated. Most of the studies of female delinquency, particular those from a feminist perspective, have emphasized the external social reality of girls' lives (past victimization, intimate relationship issues, and gender biased societal expectations and pressures regarding appropriate behavior) and frame delinquent behaviors as reactions to those external forces. While this may explain the development of these attitudes and skills, more research needs to be done on the best strategies for changing these attitudes and helping to develop important social skills for female delinquents, specifically. The current FFT and MST models focus on strategies for doing this on an individual youth basis, but do not address whether or not there is a general need for gender-specific therapy techniques that may yield stronger results for girls. Research Question #2: To what extent does delivery of the two programs affect girls' improvement in the above identified risk factors after participating in either FFT or MST? On average, girls participating in the FFT and MST programs showed improvement on the family, relationships, substance abuse, attitudes towards delinquency, attitudes towards aggression, and social skills risk domains. Baseline scores on the mental health domain were too low to test for changes. While girls did show improvement in these areas, most of the improvements were small (although 128

PAGE 144

some were statistically significant). Moreover, less than half of the participating girls showed measurable improvement on any of the risk domains. Girls in the two programs showed significant improvement in the dynamic family risk domain, whereas comparison group girls did not have a significant change. A larger proportion of girls in the program groups had improvements in this area as well. While most of the percentage differences observed were not statistically significant, this could be due to relatively small sample sizes within the groups. The consistency in the direction of the changes does suggest the group of differences warrant further investigation. Larger proportions of girls in the program groups than those in the comparison group showed improvements in the dynamic risk domains for relationships, attitudes towards aggression, and social skills. Improvements on the attitudes towards delinquency domain were about the same for both groups. Again, there is evidence for small program effects. Still, it seems that the programs (as they were implemented during this period) were not able to effect any large changes. These results may be explained in two ways. First, the program conceptual models address some of the girls' needs in this area, but additional programmatic components are needed to produce larger effects. Second, the program models do accurately address these factors, but the programs are not being implemented with fidelity to the level that is needed to produce larger and more dramatic results. A lack of program data in the area of fidelity to the model and a lack of specific information 129

PAGE 145

on the exact treatment strategies or goals are significant barriers to definitively addressing this question. In the examination of program processes, two findings emerged that indicate that these programs have not been implemented in exactly the same way as other implementation sites that have shown more promising results. The first is that both programs had a relatively high rate of treatment "failures" (program drop-outs), particularly in comparison to other treatment sites. The second is that, on the whole, lengths of service delivery for the programs were longer than is recommended in the program models. Because the two programs emphasize the need for high levels of family engagement and brief, but intensive, intervention, these two pieces of data indicate that program implementation is not at the level of fidelity that is needed to produce more pronounced positive results. Research Question #3: Are subsequent rates of re-offending related to the observed changes in targeted risk and protective factors? The only change in dynamic risk factors that was positively associated with lower rates of recidivism was improvement on the attitudes towards aggression domain. There was a significant correlation between improvement in this area and lower rates of recidivism; however, two data issues may contribute to an inability to associate these changes with re-offending. First, the sample sizes are relatively small, particularly in relation to the number of variables used in the regression equations. Second, and more importantly, there was a lack of variation in the change variables 130

PAGE 146

being tested. The proportion of girls showing improvement in risk domain scores was low on every domain (less than 50%); moreover, a few girls in the "not improving" category actually exhibited a negative numerical change (i.e., their risk increased). The overall lack of variance on these change variables makes the correlations difficult to interpret. Research Question #4: Do girls participating in FFT or MST have lower rates of recidivism than similar girls who do not participate? There were no significant differences between girls participating in the program and those comprising the comparison group on rates of re-offending or on the number of subsequent offenses. While program girls, on average, re-offended sooner than did girls in the comparison group, it is important to note that the comparison group girls received court services outside of the FFT and MST programs. Still, it was anticipated that FFT and MST program participants would have lower recidivism rates, based on previous evaluations of these program. Given the empirically established link between program fidelity and outcomes, the failure to reduce re-offending could be due to program implementation issues, rather than the appropriateness of the programs for delinquent girls. Implications for Juvenile Justice Programming for Girls The research literature reviewed points to a greater need for the study of delinquent girls and for the development ofbetter gender-specific programming in 131

PAGE 147

juvenile justice. Previous studies (see, for example, ABA & NBA, 2001) have shown that delinquent girls often have different life circumstances than delinquent boys and, thus, have different treatment needs. The findings here, regarding the general areas of risk for re-offending, show that girls have a variety of pathways to delinquency. Critics have pointed out that the traditional "one-size-fits-all" approach to juvenile justice has excluded the unique needs of delinquent girls; however, the variations of risk areas found here suggest that a female-specific approach must recognize that even within the group of delinquent girls there is still significant variation, and each girl may have different needs to be addressed. Based on previous research, it was expected that delinquent girls would have elevated risks for delinquency on the family (Sherman, 1999), relationships (Chesney-Lind & Pasko, 2004), and school (Acoca, 200) domains, specifically. In contrast, the areas where girls in this study exhibited the highest risk levels were in their attitudes towards delinquency, attitudes towards aggression, and social skills risk domains. While female aggression has been discussed to some extent in the literature, this has generally been done in the context offemale gang members. Molidor (1996), for example, has characterized some female aggression in terms of reactions to past victimization and abuse. Certainly, the findings here show that girls referred to these programs have significant aggression issues. Given that attitudes towards aggression was the only domain that was associated with recidivism reductions for girls in this study, this is an area that needs to be an important focus for treatment of these girls. 132

PAGE 148

Given the high correlations between attitudes towards aggression and family issues, attitudes towards delinquency, and social skills, programs must address the issue of covariation in risk domains that may provide insight as to why some changes in risk factors matter more than others .. Implications for FFT and MST Programs Overall, present findings support the potential of these programs as salient gender-specific strategies for girls, but more study is needed, particularly in relation to the implementation of the two programs. Future research should focus on gathering data on the actual treatment plans and goals established for girls and linking those directly to pretest and post-test risk scores. Both program models should be studied in relation to their ability to deliver treatment that addresses the attitude and skills issues discussed here, alongside their intervention with youth, their families, schools and communities. In addition, controlling for therapist adherence to the program treatment principles will be vital in establishing that these programs can be delivered in a way that meets the needs of girls and in demonstrating their effectiveness in reducing their multiple risks for delinquency. Directions for Future Research While the links between the risk factors contained in the WSCJA andreoffending are clear, the question remains whether achieving substantial changes in these domains is associated with less re-offending. Because restricted variance on the change scores poses a significant challenge in program evaluation, future studies in 133

PAGE 149

the area will need either to have significantly larger sample sizes or to transform the change variable to maximize its variability. What is apparent from this research is that the girls participating in either FFT or MST exhibit significant risks for re-offending at the time of their referrals to these programs. Their risk profiles indicate multi-faceted problems stretching across all aspects of their lives. In particular, these girls had substantial attitudes and skills related risk issues that should become an important part of their treatment plans. Both the FFT and MST treatment models are flexible enough to accommodate these treatment needs and to tailor plans and goals to focus on these areas. This study does not demonstrate that these programs are effective interventions for girls, but it does reveal their potential for addressing their risks for subsequent delinquent involvement. Further evaluations should explore more carefully the interactions between the domains of risk experienced by girls in the program and the treatment received by the girls and their families in order to better understand how these specific treatment modalities can be delivered in order to maximize the positive results for juvenile justice-involved girls. 134

PAGE 150

APPENDIX A WSJCA Pre-Screen Risk Assessment Name __________________________________ __ Initiated ___ 1 __ 1 __ I __ I __ LI_I __ LI DOMAIN 1 Record of Referrals Resulting 1n DtvcrsJOn, Deferred Adjudtcat/On, AdjudtcatJOn, Commttment to the OJ VISIOn of Youth Corrections, or Conv1ctton Referrals, not offenses, are used to assess the persistence of re-offendina by the Circle the aooropriate score Age at first offense: The age at the time of the offense for which the youth was referred to juvenile Over 16 0 court for the first time on a non-traffic misdemeanor or felony that resulted in a Diversion, Deferred 16 1 Adjudication, Adjudication, Commitment to the Division of Youth Corrections, or Conviction. 15 2 13 to 14 3 Under 13 4 Felony and misdemeanor referrals: Items 2 & 3 are mutually exclusive and should add to the total number of referrals that resulted in a Diversion, Deferred Adjudication, Adjudication, Commitment to the Division of Youth Corrections, or Conviction. 2. Misdemeanor referrals: Total referrals in which the most serious offense was a non-traffic None or one 0 misdemeanor. Two 1 Three or four 2 Five or more 3 3. Felony referrals: Total referrals for a felony offense that resulted in a Diversion, Deferred None 0 Adjudication, Adjudication, Commitment to the Division of Youth Corrections, or Conviction. One 2 (regardless of whether successfully completed). Two 4 Three or more 6 Against-person or weapon referrals: Items 4, 5, and 6 are mutually exclusive and should add to the total number of referrals that involve an against-person or weapon offense, including sex offenses, that resulted in a Diversion, Deferred Adjudication, Adjudication, Commitment to the Division of Youth Corrections, or Conviction (regardless of whether successfully completed). 4. Weapon referrals: Total referrals for which the most serious offense was a firearm/weapon None 0 charge or a weapon enhancement finding. One or more 1 5. Against-person misdemeanor referrals: Total referrals for which the most serious offense was an None 0 against-person misdemeanor, including sexual misconduct. An against-person misdemeanor involves One 1 threats, force, or physical harm to another person. Two or more 2 6. Against-person felony referrals: Total referrals for an against-person felony, including sex None 0 offenses. An against-person felony involves force or physical harm to another person. One or two 2 Three or more 4 Su offense referrals: Items 7 and 8 are mutually exclusive and should add to the total number of referrals that involve unlawful sexual behavior or another offense, the underlying factual basis of which involves unlawful sexual behavior that resulted in a Diversion, Deferred Adjudication, Adjudication, Commitment to the Division of Youth Corrections. or Conviction. 7. Misdemeanor sex offense referrals: Total misdemeanor sex offenses or misdemeanors where None the underlying factual basis involves unlawful sexual behavior. One Two or more B. Felony sex offense referrals: Total felony sex offenses or felonies where the underlying factual None basis involves unlawful sexual behavior. One Two or more 9. Court orders where youth served at least one day confined in detention: Total court and None 0 modification orders for which the youth served at least one day physically confined in a detention One 1 facility. A day served includes credit for time served. Two 2 Three or more 3 10. Court orders where youth served at least one day confined under DYC: Total court and None 0 modification orders for which the youth served at least one day confined under the authority of the Division One 2 or Youth Corrections (DYC). Two or more 4 11. Escapes: Total number of attempted or actual escape filings. None 0 One 1 Two or more 2 12. Failure-to-appear in court warrants: Total number of failures-to-appear in ccurt thai resulted in None 0 a warrant being issued. Exclude failure-to-appear warrants for non-criminal matters. One 1 Two or more 2 Criminal History Score: (Maximum of 31 points) 135

PAGE 151

WSJCA Pre-Screen Risk Assessment 2a. Youth's current school enrollment status, regardless of attendance: If the youth is in home school as a result of being expelled or dropping out, check the expelled or dropped out box, otherwise check enrolled. Sum of 2a to 2d: -----Sum of 4a and 4b: -----5. History of court-ordered or DSS out-of-home and shelter care placements exceeding 30 days: Exclude DYC commitments. 6. History of runaways or times kicked out of home: Include times the youth did not voluntarily return within 24 hours, and include incidents not reported by or to law enforcement 7. History of jaiUimprlsonment of persons who were ever involved In the household for at least 3 months: who are currently involved with the Mother and father refer to current parent or legal guardian. 8. Sum of jail/Imprisonment history: ___ 136 Mother/female caretaker Father/male caretaker Older sibling Younger sibling

PAGE 152

WSJCA Pre-Screen Risk Assessment 9. 10. Current parental authority and control: 11 c. Alcohol use within the previous 11d. Drug use within the previous 4 any history that is suspected, whether or not substantiated; exclude reports of abuse or incidents of abuse, whether or not substantiated, but exclude reports proven to be Sum of 12a and 12b: ----13. History of being a victim of neglect: Include suspected incidents of neglect, whether or not substantiated, but exclude reports proven to be 14. Mental health problems: Such as schizophrenia, bi-polar, mood, thought, personality and adjustment disorders. Exclude substance abuse and specialeducetion since those issues are considered elsewhere. Conftrm by a licensed mental health Not a victim of physical 0 Physically abused by family member 0 Physically abused by someone outside the family 0 Sexually abused by family member 0 Sexually abused by someone outside the family Maximum Score of 1 point: 0 Not victim of neglect 0 Victim of neglect 0 No history of mental health problem(s) 0 Diagnosed with mental health problem(s) 0 Only mental health medication prescribed 0 Only mental health treatment prescribed 0 Mental health medication and treatment prescribed 137

PAGE 153

WSJCA Pre-Screen Risk Assessment Pre-Screen Attitude/Behavior lndicato111 15. Reports/evidence of violence not Included In 0 No reports of violence that are not included criminal hisiOI)' criminal history: Includes displaying a weapon, 0 Reports of violence that are not included in criminal history deliberately hurting someone, violent outbursts, violent temper, fire starting, animal cruelly, destructiveness, volatility, and intense reactions. 16. Problem with sexual aggression not included In 0 No reports of sexual aggression that are not included in criminal history: Reports of aggressive sex, sex for criminal history power, young sex partners, voyeurism, exposure, etc .. 0 Reports of sexual aggression that are not included in criminal history 17. Accepts responsibility for anti-social behavior: 0 Accepts responsibility for anti-social behavior 0 Minimizes, denies, justifies, excuses, or blames others 0 Accepts anto-social behavior as okay 0 Proud of anti-social behavior 18. Attitude toward responsible law abiding behavior: 0 Abides by conventions/values 0 Believes conventions/values sometime apply to him or her 0 Does not believe conventions/values apply to him or her 0 Resents or is hostile toward responsible behavior 19. Belief In yelling and verbal aggression to resolve a 0 Believes verbal aggression is rarely appropriate disagreement or conflict: 0 Believes verbal aggression is sometimes appropriate 0 Believes verbal aooression is often appropriate 20. Belief in fighting and physical aggression to resolve 0 Believes physical aggression is never appropriate a disagreement or conflict: 0 Believes physical aggression is rarely appropriate 0 Believes physical aggression is somelimes appropriate 0 Believes pt1ysical aggression is often aQQroPriate Risk Level Definitions Using Criminal History and Social History Risk Scants Score .. .- . __ .. 'Oio5 ._-_ . . . ": '6to9 .. '10lo18 _, Low Low Low Low Moderate Risk Level:----138

PAGE 154

APPENDIXB Washington State Juvenile Court Assessment Name ______ L.aet First Initiated I I Month CaY Year JUVIS Control Number LLI __ I __ I_j_j Gender: 0 Male 0 Female --DOMAIN 1 Record of Referrals Rcsullong on Convictoon, Doversoon, or Deferred Adjudocat10n/Oosposotoon I Referrals, rather than offenses, are used to assess the persistence of re-offending by the youth. Include only refeffals that resulted in a conviction, diversion, deferred adjudication, or deferred disposition (regardless of whether successfully completed. 1. Age at first offense: The age at the time of the offense for which the youth was referred to juvenile court for the first time on a non-traffic misdemeanor or felony that resulted in a conviction, diversion, deferred -----adjudication, or deferred disposition. Felony and misdemeanor nefeml/s: Items 2 and 3 are mutually exclusive and should add to the total number of referrals that resulted in a conviction diversion, deferred adjudication, or deferred disf}Osition. 2. Misdemeanor referrals: Total number of referrals for which the most serious offense was a non-traffoc misdemeanor that resulted in a conviction, diversion, deferred adjudication, or deferred disposition -----(regardless of whether successfully completed). 3. Felony referrals: Total number of referrals for a felony offense that resulted in a conviction, diversion, -----deferred adjudication, or deferred disposition (regardless of whether successfully completed). Against-person or weapon referrals: Items 4, 5, and 6 are mutually exclusive and should add to the total number of referrals that involve an against-person or weapon offense, including sex offenses, that resulted in a conviction, diversion, deferred adjudication, or defeffed disposition (regardless of whether successfully completed). 4. Weapon referrals: Total number of referrals for which the most serious offense was a firearm/weapon -----charge or a weapon enhancement finding. 5. Against-person misdemeanor referrals: Total number of referrals for which the most serious offense was an against-person misdemeanor, including sexual misconduct. An against-person misdemeanor involves -----threats, force, or physical harm to another person (e.g., assault, coercion, harassment, intimidation, etc.). 6. Against-person felony referrals: Total number of referrals for an against-person felony, including sex offenses. An against-person felony involves force or physical harm to another person (e.g., homicide, murder, manslaughter, assault, robbery, kidnapping, domestic violence, harassment, criminal mistreatment, ----intimidation, coercion, etc.). Sax offense referrals: Items 7 and 8 are mutually exclusive and should add to the total number of referrals that involve a sex offense or sexual misconduct that resulted in a conviction, diversion, deferred adjudication, or deferred disposition. 7. Sexual misconduct misdemeanor referrals: Total number of referrals for which the most serious offense was a sexual misconduct misdemeanor or misdemeanor with sexual motivation. A sexual misdemeanor -----includes crimes such as obscene phone calls, indecent exposure, obscenity, pornography, or public indecency. 8. Felony sex offense referrals: Total number of referrals for a felony sex offense or felony involving sexual motivation. Felony sex offenses include carnal knowledge, child molestation, communication with minor for Immoral purpose, Incest, indecent exposure, indecent liberties, promoting pornography, rape, rape of child, -----sexual exploitation of child, sexual misconduct with minor, statutory rape, or voyeurism. 9. Disposition orders where youth served lit leaat one day confined In detention: Total number of disposition orders and modifiCation orders for which the youth served at least one day physically confined in a -----county detention facility. A day served includes credit for time served. 10. Dleposltlon orders where youth served at least one day confined under JRA: Total number of disposition orders and modification orders for which the youth served at least one day confined under the -----authority of the Juvenile Rehabilitation Administration (JRA). A day served includes credit for time served. 11. Escapes: Total number of attempted or actual escapes that resulted in a conviction. ---12. Failure-to-appear In court warrants: Total number of failures-to-appear in court that resulted in a warrant -----being issued. Exclude failure-to-appear warrants for non-criminal matters. Domain 1: Criminal History January 2003 139

PAGE 155

DOMAIN 2A School H1story 1. Youth Is a special education student or has a formal diagnosis of a special education need: 2. History of expulsions and suspensions since the first grade: 3. Age at first expulsion or suspension: 4. Youth has been enrolled In a community school during the laat 6 months, regardless of snendance: Behavioral AOHDIAOO Learning Mental retardation 0 No expel/suspend 0 1 expel/suspend 1. Youth's current school enrollment status, regardless of attendance: 0 Graduated, GEO If the youth Is in home school as a result of being expelled or dropping 0 Enrolled fuii-Ume out, check the expelled or dropped out box; otherwise check enrolled, If in 0 Enrolled part-time 2. Type of school in which youth Is enrolled: 3. Youth believes there Is value In getting an education: 4. Youth believes school provides an encouraging environment for him or hlr: 5. Teachers, staff, or coachas the youth likes or feels comfortable talking with: 6. Youth's involvement In school activities during most recant term: School leadership; social service clubs; music, dance, drama, art; athletics; other extracurricular activities. 7. Youth's conduct In the most recant term: Fighting or threatening students; threatening teachers/staff; overly disruptive behavior; drug/alcohol use; crimes (e.g., theft, vandalism); lying, cheating, dishonasty. 8. Number of expulsions and suspensions In the most recant term: 9. Youth's attendance in the most recent term: Partial-day absence means a/lending majority of classes and missing minority. Full-day absence means missing majority of classes. A truancy petition is equal to 7 unexcused absences In a month or 10 in a year. 10. Youth's academic performance In the most recant school term: 11. Interviewer's assessment of likelihood the youth will stay in and graduate from high school or an equivalent vocational school: Domain 2: School 143 140 0 Public academic 0 Vocational 0 Alternative 0 No ONo 0 No 0 4or5 0 6or7 0 14to 15 years old 0 16 to 18 years old 0 Suspended 0 Dropped out 0 Expelled 0 Private academic 0 Home school 0 College January 2003

PAGE 156

DOMAIN 3A Hostonc Usc of Free Tome 1 History of structuntd recreational activities within the put 5 yea1'11: Youth has participated in structured and supervised pro-social community activities such as religious group/church, community group, cultural group, club athletics, or other communit activities. 2 History of unstructured pro-soelal recreational activities within the past 5 yea1'11: Youth has engaged in activities that positively occupy the youth's time such as reading, hobbies. ate 0 Involved in 2 or more structured activities 0 Involved in 1 structured activity 0 Never involved in structured activities 0 Involved in 2 or more pro-Social unstructured activities 0 Involved in 1 pro-social unstructured activity 0 Never involved in pro-social unstructured activities DOMAIN 38 Current Usc of Free Tome 1. Current lntereat and Involvement In structured recreational 0 CorrenUy involved in 2 or more stnJctured acti vities actlvltlu: Youth participates in structured and supervised pro0 Currently involved in 1 structured activity social community activities such as religious groJp/church, 0 Currently interested but not i nvo l ved community group, cultural group club, athletics. or other 0 Currently not interested in any structured activit ies community activity. 2 Types of structured recreational activities In which youth Community/cultural group 0 No 0 Yes currently participates: Hobby group or club 0 No 0 Yes Athletics 0 No 0 Yes Religious group/church 0 No 0 Yes Volunteer or_ganization 0 No 0 Yes 3 Current Interest and Involvement In unstructured 0 Currently involved in 2 or more unstructured activities recreational activities: Youth engages in activities that 0 Currently involved in 1 unstructured activity positively occupy his or her lime, such as reading, hobbies, etc 0 Currently interested but not involved 0 Currently not interested in any unstructured act i vities DOMAIN 4A Employment Hostory 1 1 History of employment: 2 History of successful employment : 3 History of problems while employed: 4 History of positive personal ralatlonshlp(s) with past employer( a) or adult coworker(s) : 1 Understanding of what Is required to maintain a job: 2 Current interest In employment: 3 Current employment status : Current positive personal ralationshlp(s) with employer(s) or adult coworker(s): Domain 3 : Usa of Free Time Doma i n 4 : Employment 141 0 Too young for employment consideration 0 Never been employed 0 Never fired or quit because of problems 0 Fired or quit because of poor performance 0 Fired or quit because he or she could not get along 0 Never had any positive relationships 0 Had 1 positive relationship 0 Had2 0 Lacks knowledge of w1hat It takes to mainta i n a job 0 Has knowledge of abilities to maintain a job 0 Has 0 Currently employed 0 Not employed but highly interested in employment 0 Not employed but somew1hat interested 0 Not employed and not interested in employment 0 Not c urrently employed 0 Employed but no pos1ti ve relationsh i ps 0 At leas t 1 January 2003

PAGE 157

DOMAIN SA H1story of Relationships 1. History of positive adult non-family relationships not connected to school or employment: Adults, who are not teachers and not part of the youth's family, who can provide support and model pro-social behavior, such as religious leader, club member, community_person, etc. 2. History of anti-social friends/companions: Anti-social pears are youths hostile to or disruptive of the legal social order; youths who violate the law and the rights of others. 3. History of gang membership/association: 0 No positive adull relationships 0 1 positive adult relationship 0 2 positive adult relationships 0 3 or more positive adults relationships 0 Never had consistent friends or companions 0 Only had pro-social friends 0 Had pro-social friends and anti-social friends 0 Only had anti-social friends 0 Never been a gang member/associate OBee n e be/ s iat .. DOMAIN 58 Current RelatiOnships (within last four weeks) 1. Current positive aduH non-family relationships not 0 No positive adult relationships connected to school or employment: Adults, who are 0 1 positive adult relationship not teachers and not part of the youth's family, who can 0 2 positive adult relationships provide support and model pro-social behavior, such as 0 3 or more positive adults relationships religious leader, club member, community person, etc. 2. Current pro-social community ties: Youth feels there 0 No pro-social community ties are people in his or her community who discourage him or 0 Some pro-social community ties her from getting into trouble or are willing to help the youth. 0 Has strong pro-social community ties 3. Current friends/companions youth actually spends 0 No consistent friends or companions time with: 0 Only pro-social friends 0 Pro-social friends and anti-social friends 0 Only_ anti-social friends 4. Currently a gang member/associate: 0 Not a gang member/associate 0 Ga11g member/associate 5. Currently In a "romantic," Intimate, or sexual 0 Not romantically involved with anyone relationship: 0 Romantically involved with a pro-social person 0 Romantically involved with an anti-social person/criminal 6. Currently admires/emulates anti-social peers: 0 Does not admire, emulate anti-social peers 0 Somewhat admires, emulates anti-social peers 0 Admires. emulates anti-social peers 7. Current resistance to anti-social peer Influence: 0 Does not associate with anti-social friends 0 Usually resists going along with anti-social peers 0 Rarely resists goes along with anti-social peers 0 leads anti-social peers Domain 5: Relationships January 2003 142 I

PAGE 158

DOMAIN 6A Famtly Htslory 1. History of court-ordered or DSHS voluntary out-of-home and 0 No out-of-home placements exceeding 30 days shelter care placements e:aceedlng 30 days: Exclude JRA 0 1 out-of-home placement commitments 0 2 out-of-home placements 0 3 or more out-of-home placements 2. History of running away or gattlng kicked out of home: Include 0 No history of running away or being kicked out times the youth did not voluntarily return within 24 hours, and include 0 1 instance of running away/kicked out incidents not reported by or to law enforcement. 0 2 to 3 instances of running away/kicked out 0 4 to 5 instances of running away/kicked out 0 Over 5 instances of running away/kicked out 3. History of petitions filed: Include all petitions regardless of whether Youth-at-risk 0 No 0 Yes the petition was grantad; ARP ONo OYes CHINS 0 No 0 Yes Dependency ONo 0 Yes 4. History of jail/imprisonment of persons who were ever Involved In Mother/female caretaker ONo 0 Yes the household for at least 3 months: Father/male caretaker 0 No 0 Yes Older sibling ONo 0 Yes Younger sibling ONo 0 Yes Other member 0 No 0 Yes 5. Youth has been living under any "adult supervision" during the 0 No, living with peers without adult supervision, last 4 weeks. Adult supervision must be someone who is responsible do not complete Domain 6B for the youth's welfare, either legally or with parental consent. 0 No, living alone without adult supervision, do not complete Domain 6B 0 No, transient without adult supervision, do not I complete Domain 6B 0 Yes, must complete Domain 6B DOMAIN 68 Current Ltvtng Arrangements I Complete this section if the vouth hes been living under any adult suiJSrvision during the last 4 weeks. 1. All persons with whom you1h Is currently living (within the last 4 BioiOciical mother 0 No 0 Yes weeks): Biolocical father ONo 0 Yes Non-bioloaical mother 0 No 0 Yes ONo OYes Older 0 No 0 Yes Youncer sibli,:;Q("S) 0 No 0 Yes Grandparent( Sf ONo 0 Yes Other relativelsl 0 No 0 Yes Lone-term oarental oartnerls) ONo 0 Yes Short-term oarenta!Dartnerls) 0 No 0 Yes Youth's romantic oartner 0 No 0 Yes You1h's child ONo OYes Foster/oroUP home 0 No 0 Yes Youth's friends 0 No o Yes Transient (street. moving around) 0 No 0 Yes 2. Annual combined Income of youth and family: 0 Under $15,000 0 $15,000 to $34,999 0 $35,000 to $49,999 0 $50,000 and over 3. Jail/Imprisonment history of persons who are currently Involved Mother/female caretaker 0 No 0 Yes with the household: Father/male caretaker 0 No OYes Older sibling 0 No 0 Yes Younger sibling 0 No 0 Yes Other member 0 No 0 Yes 4. Problem history of parents who are currently Involved with the Alcohol 0 No 0 Yes I household: Drugs 0 No 0 Yes Mental health 0 No OYes Physical health ONo 0 Yes Emolovment 0 No 0 Yes Domain 6: Family January 2003 143

PAGE 159

5. Problem history of siblings who are currently Involved with the Alcohol 0 No 0 Yes household: Drugs 0 No 0 Yes Mental health ONo 0 Yes Physical health ONo 0 Yes EmPIO}'IIlent 0 No 0 Yes 6. Support network for family: Extended family and/or family friends 0 No support network who can provide additional support to the family. 0 Some support network 0 Strona support network 7. Family willingness to help support youth: 0 Consistently willing to support youth 0 Inconsistently willing to support youth 0 Little or no willingness to support youth 0 Hostile, berating, and/or belittling of youth 8. Family provides opportunities for youth to participate in family 0 No opportunities for involvement provided activities and decisions affecting the youth: 0 Some opportunities for involvement provided 0 Opportunities for involvement provided 9. Youth hae run away or been kicked out of home within the last 4 0 Has not run away/kicked out of home weeks: Include limes youth did not voluntarily return within 24 hours, 0 Has run away/kicked out within last 4 weeks and include incidents not reported by or to law enforcement. 0 Is currently kicked out of home or is a runaway 10. Family member(s) youth feels close to or has good relationship Mother/female caretaker 0 No 0 Yes with: Father/male caretaker ONo OYes Male sibling 0 No OYes Female sibling 0 No 0 Yes EK!ended family ONo 0 Yes 11. Level of conflict between parents, between youth and parents, 0 Some conflict that is well managed among siblings: 0 Verbal intimidation, yelling, healed arguments 0 Threats of physical abuse 0 Domestic violence: physicaVsexual abuse 12. Parental supervision: Parents know whom youth is with, when 0 Consistent good supervision youth will r&turn, where youth is going, and what youth is doing. 0 Sporadic supervision 0 supervision 13. Parental authority and control: 0 Youth usually obeys and follows rules 0 Youth sometimes obeys or obeys some rules 0 Youth consistently disobeys and/or is hostile 14. Consistent appropriate punishment for bad behavior: 0 Consistently appropriate punishment Appropriate means clear communication, timely response, and 0 Consistently overly severe punishment response proportionate to conduct. 0 Consistently insufficient punishment 0 Inconsistent or erratic punishment 15. Consistent appropriate rewards for good behavior: Appropriate 0 Consistently appropriate rewards means clear communication, timely response, and response 0 Consistently overly indulgenVoverty protective proportionate to conduct; rewards mean affection, praise, etc. 0 Consistently insufficient rewards 0 Inconsistent or erratic rewards 16. Parental characterization of youth's anti-social behavior: 0 Disapproves of youth's anti-social behavior 0 Minimizes, denies, justifies, excuses behavior, or blames others/circumstances 0 Accepts youth's anti-social behavior as okay 0 Proud of youth's anti-social behavior Domain 6: Family January 2003 144

PAGE 160

DOMAIN 7A Alcohol and Drug H1story Disrupted functioning involves having a problem in any of these five life 818as: education, family conflicl, peer relationships, crime, or health, and usually indicates ttutment is wBtTented. Use that contributes to criminal behavior typically precipitates the commission of e crime; there is evidence or reason to believe the youth's criminal activity is filiated to alcohoVdrug use 1. Hletory of alcohol uae: 2. Hletory of drug un: 3. History of referrale for alcohol/drug assessment: 4 Hletory of attending alcohol/drug edu!;!tion clasee! for a alcohol/drug problem: 5. 1 Alcohol use within the previous 4 weeks: 2 Drug use within the previous 4 weeks: 3. Alcohol/drug treatment program participation within the previous 4 weeks: 4. Type of drugs used within the previous 4 weeks: Domain 7: Alcohol and Drugs Past use of alcohol Alcohol disrupted education Alcohol caused family conflict Alcohol interfered with keeping pro-social friends caused health problems contributed to criminal behavior Current alcohol use not dtsrupting functioning Alcohol disrupts education Alcohol causes family connie! Alcohol interferes with keeping pro-social friends Alcohol causes health problems Alcohol contributes to criminal behavior Current drug use not disrupting functioning Drugs disrupt education Drugs cause family conflicl Drugs interfere with keeping pro-social friends Drugs cause health problems Drugs contribute to criminal behavior 0 Alcohol/drug treatment not warranted 0 No ONo 0 No 0 No 0 No 0 No 0 No 0 No ONo 0 No 0 0 No 0 No 0 No 0 No 0 No 0 No 0 No 0 No 0 No ONo 0 Yes OYes 0 Yes 0 Yes 0 Yes 0 Yes 0 Yes 0 Yes 0 Yes ')Yes 0 0 Yes 0 Yes 0 Yes 0 Yes 0 Yes 0 0 Yes 0 Yes 0 Yes 0 Yes OYes Yes 0 Not currently attending needed alcohol/drug treatment program 0 CurrenUy attending alcohol/drug treatment program January 2003 145

PAGE 161

DOMAIN BA Mental Health Hstory 1. History of sulcldalldeatlon: 0 Has never thought about suicide 0 Has had serious thoughts about suicide 0 Has made a plan to commit suicide 0 Has attem_pted to commit suicide 2. History of physical abuse: Include suspected incidents 0 Not a victim of physical abuse of abuse, whether or not substantiated, but exclude 0 Physically abused by family member reports proven to be false. 0 Physically abused by someone outside the family 3. History of sexual abuse: Include suspected incidents 0 Not a victim of sexual abuse of abuse, whether or not substantleted, but exclude 0 Sexually abused by family member reports proven to be false. 0 Sexually abused by someone outside the family 4. History of being a victim of neglect: Include 0 Not a victim of neglect suspected incidents of neglect, whether or not 0 Victim of neglect substantiated, but exclude reports _!){oven to be false. 5. History of ADD/ADHD: Confirmed by a professional in 0 No history of ADD/ADHD the social service/healthcare field. 0 Diagnosed with ADD/ADHD 0 Only ADD/ADHD medication prescribed 0 Only ADD/ADHD treatment prescribed 0 ADD/ADHD medication and treatment prescribed 6. History of mental health problems: Such as 0 No history of mental health problem(s) schizophrenia, bi-polar, mood, thought, personality, and 0 Diagnosed with mental health problem(s) adjustment disorders. Exclude conduct disorder, 0 Only mental health medication prescribed oppositional defiant disorder, substance abuse, and 0 Only mental health treatment prescribed ADDIADHD. Confirmed by a professional in the social 0 Mental health medication and treatment prescribed servicelhealthcere field. 7. Currently has health Insurance: 0 No health insurance 0 Public insurance (Medicaid) 0 Private Insurance 8. Current mental health problem status: 0 No mental health problem(s), do not complete Domain 88 0 Mental health problem(s must complete Domain 88 DOMAIN BB Current Mental Health (W1thm the last 4 weeks) 1. Current suicidal Ideation within the previous 4 0 Does not have thoughts about suicide weeks: 0 Has had serious thoughts about suicide 0 Has recently made a plan to commit suicide 0 Has recently attempted to commit suicide 2. Currently diagnosed with ADD/ADHD: Conftrmed by a 0 No ADD/ADHD diagnosis professions/In the social servicelhealthcare field. 0 No ADD/ADHD medication currently prescribed Type of medication: 0 Currently taking ADD/ADHD medication 0 ADD/ADHD medication currently prescribed, but not taking 3. Mental health treatment currently prescribed 0 No current mental health problem excluding ADD/ADHD treatment: 0 No mental health treatment currently prescribed 0 Attending mental health treatment 0 Treatment currently prescribed, but not attending 4. Mental health medication currently prescribed 0 No current mental health problem excluding ADD/ADHD medication: 0 No mental health medication currently prescribed Type of medication: 0 Currently taking mental health medication 0 Mental heaHh medication currently prescribed, but not takirlg_ 5. Mental health problems currently Interfere with 0 No current mental health problem working with the youth: 0 Mental health problem(s) do not interfere in work with youth 0 Mental health problem(s) interfere in work with youth Domain 8: Mental Health January 2003 146

PAGE 162

DOMAIN 9A Att1tudes/Behav10rs 1. Primary emotion when committing crlme(s) 0 Nervous, afraid, worried, ambivalent, uncertain, or indecisive within the last 6-months: 0 Hyper, excited, or stimulated 0 Unconcerned or indifferent 0 Confident or brag_s about not oeHina caught 2. Primary purpose for committing crlme(s) 0 Anger within the last 6-months: 0 Revenge 0 Impulse 0 Sexual desire 0 Money or material gain, including drugs 0 Excitement, amusement, or fun 0 Peer status, acceptance. or attention 3. Optimism: Youth talks about future in positive 0 High aspirations: sense of purpose, commitment to beHer life way with plans or aspirations of a better life that 0 Normal aspirations: some sense of purpose could include: employment, education, raising a 0 Low aspirations: little sense of purpose or plans for better life family, travel, or other pro-social life goals. 0 Believes nothing_ matters; he or she will be dead before long 4. Impulsive; acts before thinking: 0 Uses self-control; usually thinks before acting 0 Some self-control; sometimes thinks before acting 0 Impulsive; often acts before thinking 0 Highly Impulsive; usually acts before thinking 5. Belief In control over anti-social behavior: 0 Believes he or she can avoid/stop anti-social behavior 0 Somewhat believes anti-social behavior is controllable 0 Believes his or her anti-social behavior is out of his or her control 6. Empathy, remorse, sympathy, or feelings for 0 Has empathy for his or her victim(s) the vlctlm(s) of criminal behavior: 0 Has some empathy for his or her victim(s) 0 Does not have empathy for his or her victim(s) 7. Respect for propeny of others: 0 Respects property of others 0 Respects personal property but not publicly accessible property: "It's not hurting anybody." 0 Conditional respect for personal property: "If they are stupid enough to leave it out. they deserve losing 0 No respect for property: "If I want something, it should be mine." 8. Respect for authority figures: 0 Respects most authority figures 0 Does not respect authority figures, and may resent some 0 Resents most authority figures 0 or 1s hostile toward most authority figures 9. Attitude toward responsible law abiding 0 Abides by conventions/values behavior: 0 Believes conventions/values sometime apply to him or her 0 Does not believe conventions/values apply to him or her 0 Resents or is hostile toward responsible behavior 10. Accepts responsibility for anti-social 0 Accepts responsibility for anti-social behavior behavior: 0 Minimizes, denies, justifies, excuses, or blames others 0 Accepts anti-social behavior as okay 0 Proud of anti-social behavior 11. Youth's belief In successfully meeting 0 Believes he or she will be successful conditions of coun supervision: 0 Unsure if he or she will be successful 0 Does not believe he or she will be successful Domain 9: Attitudes and Behaviors January 2003 147

PAGE 163

1. Tolerance for frustration: 2. Hostile Interpretation of actions and Intentions of otherw In a common non confrontational setting: 3. Belief In yelling and verbal aggression to resolve a disagreement or conflict: DOMAIN 98 Aggression 0 Rarely gets upset over small things or has temper tantrums 0 Sometimes gets upset over small things or has temper tantrums 0 Often gets upset over or has temper tantrums 0 Primarily positive view of intentions of others 0 Primarily negative view of intentions of others 0 Primarily hostile view of intentions of others 0 Believes verbal aggression is rarely appropriate 0 Believes verbal aggression is sometimes appropriate 0 Believes verbal is often 4. Belief In fighting and physical aggression 0 Believes physical aggression is never appropriate to resolve a dlsavreement or conflict: 0 Believes physical aggression is rarely appropriate 0 Believes physical aggression is sometimes appropriate 0 is often 5. Reports/evidence of violence not Included In criminal history: Violent destruction of property 0 No Report Violent outbursts, displays of temper, uncontrolled anger indicating potential for harm 0 No Report Deliberately inflicting physical pain 0 No RePOrt Using/threatening with a weapon 0 No Report Fire starting 0 No Report Animal cruelty 0 No Report 6. Reports of problem with sexual aggression not Included In criminal hlatory: 0 Yes 0 Yes 0 Yes 0 Yes OYes 0 Yes ...................................................................................... i\99iessive.sex al,io Report ... a'i'as .......... .. .... .. .... .. .. ...... ...... ........... .. ...... .. .......... .. .. ... seiac.;: pc;v.;e; a-liio ReiiOrt .. a vas .. ...... Child sex 0 No Report 0 Yes Domain 9: Attitudes and Behaviors January 2003 148

PAGE 164

DOMAIN 10 Skills 1. Consequential thlnklllfl: 0 Does not understand there are consequences to actions 0 Understands there are consequences to actions 0 Identifies consequences of actions 0 Acts to obtain desired consequences-good consequential thinking 2. Goal setting: 0 Does not set goals 0 Sets unrealistic goals 0 Sets somewhat realistic goals 0 Sets realistic goals 3. Problem-solving: 0 Cannot identify problem behaviors 0 Identifies problem behaviors 0 Thinks of solutions for problem behaviors 0 Applies appropriate solutions to problem behaviors 4. Situational perception: Ability to analyze 'he 0 Cannot analyze the situation for use of a pro-social skill situation, choose the best pro-social skill, end 0 Can analyze but not choose the best pro-social skill select the best time and place to use the pro0 Can choose the best skill but cannot select the best time and place social skill. 0 Can select the best time and place to use the best pro-social skill 5. Dealing with others: Basic social skills 0 Lacks basic social skills in dealing with others include listening, starting a conversation, having 0 Has basic social skills, lacks advanced skills in dealing with others e conversation, asking a question, saying thank 0 Sometimes uses advanced social skills in dealing with others you, introducing yourSflff, introducing other 0 Often uses advanced social skills in dealing with others people, and giving a compliment. Advanced social skills include asking for help, joining in, giving instructions, following instructions, apologizing, and convincing others. 6. Dealing with difficult situations: Includes 0 Lacks skills in dealing with difficult situations making a complaint, answering a complaint, 0 Rarely uses skills in dealing with difficult situations dealing with embarrassment, dealing with being 0 Sometimes uses skills in dealing with difficult situations left out, standing up for a friend, responding to 0 Often uses skills in dealing with difficult situations frustration, rasponding to failure, dealing with contradictory massages, dealing with accusation, getting ready for a difficult conversation, and dealing with group pressure. 7. Dealing with feelings/emotions: Includes 0 Lacks skills in dealing with feelings/emotions knowing his or her feelings, expressing feelings, 0 Rarely uses skills in dealing with feelings/emotions understanding the feelings of others, dealing 0 Sometimes uses skills in dealing with feelings/emotions with someone a/sa's anger, expressing 0 Often uses skills in dealing with feelings/emotions affection, dealing with fear, and rewarding one sa/f. 8. Monitoring of Internal triggers, distorted 0 Cannot identify internal triggers thoughts, that can lead to trouble: 0 Identifies internal triggers 0 Actively monitors/controls internal triggers 9. Monitoring of external triggers, events or 0 Cannot identify external triggers situations, that can lead to trouble: 0 Identifies external triggers 0 Actively monitors/controls extemaltriooers 10. Control of impulsive behaviors that get 0 Never had a problem with impulsive behavior youth Into trouble: Rafrsming, replacing anti0 Does not know techniques to control impulsive behavior social thoughts with pro-social thoughts, 0 Knows techniques to control impulsive behavior diversion, relaxation, problem solving, 0 Uses techniques to control impulsive behavior negotiation, ralapsa prevention. 11. Control of aggression: Includes asking 0 Never had a problem with aggression permission, sharing thoughts, helping others, 0 Lacks alternatives to aggression negotiating, using self control, standing up for 0 Rarely uses alternatives to aggression one's rights, responding to teasing, avoiding 0 Sometimes uses alternatives to aggression trouble with others, end keeping out of fights. 0 Often uses alternatives to aggression Domain 10: Skills January 2003 149

PAGE 165

REFERENCES A coca, L. ( 1999). Investing in girls: A 21st century strategy. Juvenile Justice, 6( 1) 3-13. Washington, DC: Office of Juvenile Justice and Delinquency Prevention. Acoca, L. (2000). Educate or incarcerate: Girls in the Florida and Duvall County juvenile justice systems. San Fransisco: CA: National Council on Crime and Delinquency. Agnew, R. (1998). Foundation for a general strain theory of crime and delinquency. In P.M. Sharp & B. W. Hancock (Eds). Juvenile Delinquency: Historical, theoretical and societal reactions to youth (pp. 185-211). Newark, NJ: Prentice Hall. Agnew, R. (2005a). Juvenile delinquency: Causes and control. (2nd ed.) Los Angeles: Roxbury Publishing Company. Agnew, R. (2005b ). Why do criminals offend? A general theory of crime and delinquency. Los Angeles, CA: Roxbury Publishing Company. Akers, R. L. (1985): Deviant behavior: A social/earning approach. Belmont, CA: Wadsworth Publishing. Alexander, J., Pugh, C., Parsons, B. & Sexton, T. (2000). Blueprints for violence prevention: Functional Family Therapy. Boulder, CO: University of Colorado Boulder, Center for the Study and Prevention of Violence. American Bar Association and National Bar Association. (2001). Justice by Gender: The Lack of Appropriate Prevention, Diversion and Treatment Alternatives for Girls in the Justice System. Retrieved October 2005 at http://www.abanet.org. American Civil Liberties Union (2002). ACLU and Cook County agree on plan to improve chronic conditions at IL juvenile detention center. Retrieved December 2003 at http://www.aclu.org. Angenent, H. & Man, A. (1996). Background/actors ofjuvenile delinquency. American University Studies. New York: Peter Lang. 150

PAGE 166

Aos, S., Miller, M., & Drake, E. (2006). Evidence-based public policy options to reduce future prison construction, crimina/justice costs, and crime rates. Olympia, WA: Washington Institute for Public Policy. Aos, S., Phipps, P., Barnoski, R. & Lieb, R. (2001). The comparative costs and benefits of programs to reduce rime. Olympia, WA: Washington State Institute for Public Policy. Barnoski, R. (2004). Assessing risk of re-offense: Validating the Washington State juvenile court assessment. Olympia, WA: Washington State Institute for Public Policy. Beard, B. B. (1934). Juvenile probation: An analysis of the case records of five hundred children studied at the Judge Baker Guidance Clinic and placed on probation in the juvenile court of Boston. New York, NY: American Book Company. Belknap, J. & Gaarder (2002). Tenuous borders: Girls transferred to adult court. Criminology, 40(3): 481-517. Belknap, J. & Holsinger, K. ( 1998). An overview of delinquent girls: How theory and practice have failed and the need for innovative changes. In R. Zaplin (Ed.) The female offender: What does the future hold? (pp. 31-64). Washington, DC: St. Mary's Press. Belknap, J., Holsinger, K. & Dunn, M. (1997). Understanding incarcerated girls: The results of a focus group study. The Prison Journal, 77(4), 381-404. Borduin, C. M. (1999). Multisystemic treatment of criminality and violence in adolescents. Journal of the American Academy of Child and Adolescent Psychiatry, 38 (3), 242-250. Bums, B. J., Schoenwald, S. K., Burchard, J.D., Faw, L., & Santos, A. B. (2000). Comprehensive community-based interventions for youth with severe emotional disorders: Multisystemic therapy and the Wraparound process. Journal of Child and Family Studies, 9 (3), 283-314. Carr, A. (2000). Evidence-based practice in family therapy and systemic consultation: Child-focused problems. Journal of Family Therapy, 22 (1), 29-61. 151

PAGE 167

ChesneyLind, M. (1989). Girls, crime and a woman's place: Toward a feminist Model of female delinquency. Crime and Delinquency, 35 (5), 5-29. ChesneyLind, M. & Belknap, J. (2002, November). Gender, delinquency and juvenile justice: What about girls? Paper presented at the annual meeting of the American Society of Criminology, Washington, DC. Chesney-Lind, M. & Pasko, L. (2004), The female offender: Girls, women and crime. Thousand Oaks, CA: Sage Publications. Chesney-Lind, M. & Shelden, R. G. (1992). Girls: Delinquency and juvenile justice. Pacific Grove, CA: Brooks/Cole Publishing Company. Cloward, R. A. & Ohlin, L. E. (1960). Delinquency and opportunity. New York: Free Press. Cohen, A. K. (1955). Delinquent boys: The culture of the gang. New York: Free Press. Cohen, L. E., & Felson, M. (1974). Social change and crime rate trends: A routine activity approach. American Sociological Review, 44 (4), 588-608. Colvin, M. (2000). Crime and coercion. New York: St. Martin's Press. Daniel, M. (1999). The Female Intervention Team. Juvenile Justice Bulletin, 6 (1 ), 1-5. Division ofYouth Corrections (2006). 2006 Recidivism Report. Denver, CO: Colorado Department of Health and Human Services. Eddy, J. M. & Gribskov, L. S. (1998). Juvenile justice and delinquency prevention in the United States: The influence of theories and traditions on policies and practices. In T.P. Gullotta, G.R. Adams, & R. Montemayor (Eds.) Delinquent violent youth: Theory and interventions (pp. 27-28). Thousand Oaks, CA, et al: Sage Publications. 152

PAGE 168

Elliott, D. S. (2000). Violent Offending Over the Life Course: A Sociological Perspective. InN. Krasnegor, N. Anderson, & D. Bynum (Eds.), Health and Behavior. Behavioral and Social Science Seminar Series, Vol. 1, Washington, DC: National Institutes of Health. Feldman, L., Males, M., & Schiraldi, V. (2001). A tale oftwojurisdictions: Youth crime and detention rates in Maryland and the District of Columbia. Retrieved October 2003 at http://www.buildingblocksforyouth.org. Ferdinand, T. N. (1998). History overtakes the juvenile justice system. In P.M. Sharp & B. W. Hancock (Eds). Juvenile delinquency: Historical, theoretical and societal reactions to youth (pp. 64-76). Newark, NJ: Prentice Hall. Freud, S. (1949). The Ego and the Id, (J. Riviere, Trans.) London: Hogarth. (Original work published in 1927). Fox, K. & Lyons, D. (2003) Juvenile confinement conditions and services. Legisbrief, 11(5). Denver, CO: National Conference ofState Legislatures. Gordon, D. A., Graves, K., & Arbuthnot, J. (1995). The effect of Functional Family Therapy for delinquents on adult criminal behavior. Criminal Justice and Behavior, 22 (1), 60-73. Governor's Juvenile Justice Advisory Committee. (2004). Annual Report: 2003. Olympia, W A: Author. Green, Peters & Associates ( 1998) Guiding principles for female programming: An inventory of best practices. Washington, DC: Office of Juvenile Justice and Delinquency Prevention. Hawkins, J. D., Herrenkohl, T., Farrington, D., Brewer, D., Catalano, R., Harachi, T., & Cothern, L. (2000). Predictors of youth violence. Juvenile Justice Bulletin, April2000. Washington, DC: Office of Juvenile Justice and Delinquency Prevention. Retreived October 2004 at http://www.ncjrs.gov/html/oiidp/jjbul2000 04 5/contents.html. Henggeler, S. W., Clingempeel, W. G., Brondino, M. J., & Pickrel, S. G. (2002). Four-year follow-up ofMultisystemic therapy with substance-abusing and substance-dependent juvenile offenders. Journal of the American Academy of Child and Adolescent Psychiatry, 41 (7), 868-874. 153

PAGE 169

Henggeler, S. W., Mihalic, S. F., Rone, L., Thomas, C., & Timmons-Mitchell, J. (1998). Blueprints for violence prevention: Multisystemic therapy. Boulder, CO: University of Colorado Boulder, Center for the Study and Prevention of Violence. Henggeler, S. W., Schoenwald, S. K., Liao, J. G., Letourneau, E. J., & Edwards, D. L. (2002). Transporting efficacious treatments to field settings: The link between supervisory practices and therapist fidelity in MST programs. Journal of Clinical and Child Psychology, 31 (2), 155-167. Hodge, R. D. (2001). Thejuvenile offender: Theory, research and applications (2nd ed.) Boston: Kluwer Academic Publishers. Hirschi, T. (1998). A control theory of delinquency. In P.M. Sharp & B. W. Hancock (eds). Juvenile delinquency: Historical, theoretical and societal reactions to youth (pp. 98-104). Newark, NJ: Prentice Hall. King County Juvenile Court (2006). Preliminary draft: Juvenile Justice Intervention Services program evaluation (unpublished). Seattle, W A: King County Juvenile Court with TriWest Group. Krisberg, B., & Austin, J. F. (1993). Reinventing juvenile justice. Newbury Park, Sage Publications. Lander, B. (1954). Towards an understanding ofjuvenile delinquency. New York: University of Columbia Press. Lipsey, M. W., & Derzon, J. H. (1998). Predictors ofviolent and serious delinquency in adolescence and early adulthood: A synthesis oflongitudinal research. In R. Loeber & D.P. Farringon (Eds.), Serious and Violent Juvenile Offenders: Risk Factors and Successful Interventions (pp. 86-105). Thousand Oaks, CA: Sage Publications. Lombroso, C. (1968). Crime. (H. P. Horton, Trans.) Montclair, NJ: Patterson Smith (Original work published in 1911 ). Martinson, R. (1974). What works? Questions and answers about prison reform. Public Interest, 36, 22-45. 154

PAGE 170

Matson, S., & Bamoski, R. (1997). Assessing risk: Washington State Juvenile Court Early Intervention Program. Olympia, WA: Washington State Institute for Public Policy. McCabe, K. M., Lansing, A. E., Garland, A., & Hough, R. (2002). Gender differences in psychopathology, functional impairment, and familial risk factors among adjudicated delinquents. Journal of the Academy of Child & Adolescent Psychiatry, 41 (7), 860-867. Menne], Robert M. (1973). Thorns and thistles: Juvenile delinquents in the United States, 1825-1940. Hanover, NH: University Press ofNew England. Merton, R. (1938). Social structure and anomie. American Sociological Review, 3, 672-682. Merton, R. (1957). Social theory and social structure. (Rev. ed.) London: The Free Press. Molidor, C. (1996). Female gang members: A profile of aggression and victimization. Social Work, 41,251-251. National Conference of State Legislatures. (1994-2002). Review of selected juvenile justice legislation. Retrieved November 2004 at http://www/ncsl.org Patterson, E. B. ( 1998). Poverty, income inequality, and community crime rates. In P. M. Sharp & B. W. Hancock (Eds). Juvenile delinquency: Historical, theoretical and societal reactions to youth (pp. 135-150). Newark, NJ: Prentice Hall. Regoli, R. M., & Hewitt, J.D. (2000). Delinquency in society (41h ed.). Boston, MA: McGraw Hill. Robbins, M.S., Alexander, J. F., Newell, R. M., & Turner, C. W. (1996). The immediate effect ofreframing of client attitude in family therapy. Journal of Family Psychology, 10 (1), pp. 28-34. Robbins, M. S., Alexander, J. F., Turner, C. W., & Perez, G. A. (2003). Alliance and dropout in family therapy for adolescents with behavior problems: Individual and systemic effects. Journal of Family Psychology, 17 (4), 534-544. Rubington, E., & Weibert, M.S. (1996). Deviance: The Interactionist Perspective, Boston: Allyn and Bacon. 155

PAGE 171

Schoenwald, S. K., Brown, T. L., & Henngeler, S. W. (2000). Inside Multisystemic therapy: Therapist, supervisory, and program practices. Journal of Emotional and Behavioral Disorders, 8 (2), 113-128. Schoenwald, S. K., Halliday-Boykins, C. A., & Henngeler, S. W. (2003). Client-level predictors of adherence to MST in community service settings. Family Process, 42 (3), 345-359. Schlossman, S., & Wallach, S. (1998). The crime of precocious sexuality: Female juvenile delinquency in the Progressive Era. In P.M. Sharp & B. W. Hancock (Eds). Juvenile delinquency: Historical, theoretical and societal reactions to youth (pp. 98-104): Newark, NJ: Prentice Hall. Shaw, C. R. & McKay, H. D. (1942). Juvenile delinquency and urban areas. Chicago: University of Chicago Press. Sherman, F. (1999). The Juvenile Rights Advocacy Project: Representing girls in context. Juvenile Justice Journal, 6 (1). Sherman, F. (2000). Prostitution and teenage girls. Women, Girls and Criminal Justice Newsletter, 6 (1), 83-84. Shoemaker, D. J. ( 1990). Theories of delinquency: An examination of explanations of delinquent behavior. New York: Oxford University Press. Sickmund, M., Snyder, H. N., & PoeYamagata, E. (1997). Juvenile offenders and victims: 1997 update on violence. Washington, DC: Office of Juvenile Justice and Delinquency Prevention. Stanfield, R. ( 1995). Pathways to juvenile justice reform. Seattle, W A: Annie E. Casey Foundation. Sykes, G. M. & Matza, D. (1998). Techniques of neutralization: A theory of delinquency. In P.M. Sharp & B. W. Hancock (eds). Juvenile Delinquency: Historical, theoretical and societal reactions to youth (pp. 105-110). Newark, NJ: Prentice Hall. Thornberry, T. P. ( 1998). Toward an interactional theory of delinquency. In P. M. Sharp & B. W. Hancock (eds). Juvenile delinquency: Historical, theoretical and societal reactions to youth (pp. 151-167). Newark, NJ: Prentice Hall. 156

PAGE 172

Thornberry, T. P., Huizinga, D. & Loeber, R. (1995). The prevention of serious delinquency and violence: Implicaions from the Program of Research on the Causes and Correlates of Delinquency. In J.C. Howell, B. Krisberg, J.D. Hawkins, & J. J. Wilson (Eds.) A sourcebook: Serious, violent, and chronic juvenile offenders (pp. 213-237). Thousand Oaks, CA: Sage Publications. U.S. Department of Health and Human Services (2001). Youth violence: A report of the Surgeon General. Washington, DC: author. U.S. Department of Justice (2003). Juvenile court statistics 1999: Celebrating 100 years of the juvenile court. Washington, DC: National Juvenile Court Data Archives. Washington State Institute for Public Policy (1999). The Community Juvenile Accountability Act: Research-proven intervention for the juvenile courts. Olympia, W A: Author. Washington State Institute for Public Policy (2002a). Washington State's implementation of Aggression Replacement Training for juvenile offenders: Preliminary findings. Olympia, WA: Author. Washington State Institute for Public Policy (2002b ). Washington State's implementation of Functional Family Therapy for juvenile offenders: Preliminary findings. Olympia, WA: Author. Washington State Institute for Public Policy (2003). Washington State Juvenile Court Assessment Manual V. 2.1. Olympia, WA: Author. 157