ASSESSING THE PREVENTION OF JUVENILE DELINQUENCY

ASSESSING THE PREVENTION OF JUVENILE DELINQUENCY:

AN EVALUATION OF A COURT-BASED DELINQUENCY

PREVENTION PROGRAM

Jacqueline Bergdahl, Sarah Twill, Michael Norris, and Matthew Ream

In 2002, the State of Ohio mandated juvenile courts to provide prevention for at-risk youth. This study examined official court records to evaluate the effectiveness of a prevention program administered by the Greene County Juvenile Court. A sample of 362 youth referred to the program for the years 2002 to 2009 by concerned caretakers, teachers, and police was analyzed. Con- sistent with intake goals, 81.7% of clients were referred for at-risk but not actually delinquent behaviors. Completion of the prevention program did not predict future court referrals, but neither did seriousness of referral behavior. Children with two biological parents were significantly more likely to complete the program, whereas referrals to Strengthening Families Program and substance abuse screening significantly predicted program noncompletion. Implications for policy and research are discussed.

Key Points for the Family Court Community:

� This article highlights efforts by county juvenile court to implement a secondary prevention program for at-risk but not officially court-referred youth.

� Delinquency prevention research depends on good juvenile court data and adequate comparison groups. � Evidence-based predelinquent interventions with external process and outcome evaluations should be the standard.

Keywords: At-Risk Juveniles; Caretaker and School Referrals; Delinquency Prevention; Juvenile Court; and State-

Mandated Prevention.

INTRODUCTION

In an effort to reduce juvenile delinquency, preventive programs began to emerge soon after the creation of the first juvenile court in Cook County, Illinois, in 1899 (Scott & Grisso, 1997). The Chicago Area Project emphasized cultivating conventional behavior in at-risk youth for sev- eral decades after its creation in 1932, and the Cambridge-Somerville Youth Study offered pre- ventive treatment from 1942 to 1976 (McCord, McCord, & Zola, 1959; Lundman, 2001). In 1961, the President’s Commission on Juvenile Delinquency and Youth Crime recognized the need for deinstitutionalization of status offenders and community-based programs, which led to the Juvenile Justice and Delinquency Prevention Act of 1974 and establishment of the Office of Juvenile Justice and Delinquency Prevention (OJJDP). Funding became available for implement- ing and evaluating prevention and diversion programs, and despite a regression toward “get tough” juvenile policies in the 1990s (Howitt, Moore, & Gaulier, 1998), delinquency prevention programs continue to be popular (Mihalic, Irwin, Fagan, Ballard, & Elliott, 2004; Bartollas & Schmalleger, 2013).

This article discusses the nature of state-mandated, predelinquent prevention and then describes and evaluates a prevention program provided by a county juvenile court in Ohio. We examined offi- cial court records for 362 youth referred to the program by not only law enforcement but also con- cerned caretakers and school officials between 2002 and 2009. We present sample demographics and

Correspondence: Jacqueline.Bergdahl@wright.edu; Sarah.Twill@wright.edu; Michael.Norris@wright.edu

FAMILY COURT REVIEW, Vol. 53 No. 4, October 2015 617–631 VC 2015 Association of Family and Conciliation Courts

 

 

discuss at-risk behaviors leading to program referrals, types of treatment specified by the court, case outcomes, study limitations, and suggestions for future research.

DELINQUENCY PREVENTION LITERATURE

Studies of prevention and diversion programs have significantly increased our understanding of juvenile delinquency, including the behavioral, demographic, and familial variables associated with it, as well as trends in juvenile offending over time. While the literature describes numerous preven- tion efforts, very few of these programs have been systematically evaluated to determine if they have been both implemented as designed (process evaluation) and effective (outcome evaluation; Mihalic et al., 2004).

The pioneering work of Sheldon and Elenor Glueck (1950) showed the critical influence of family on juvenile delinquency and the importance of early behavioral problems in predicting later delin- quency and crime. Several subsequent studies have examined risk factors for delinquency and ana- lyzed what works to prevent delinquency. Risk factors include both structural influences like low family income, high transience, and overcrowding (Glueck & Glueck, 1950; Sampson & Laub, 1993) and parental risk factors including poor supervision (Sampson & Laub, 1993; Loeber & Stouthamer-Loeber, 1986), parental rejection, harsh or inconsistent discipline (Sampson & Laub, 1993), negative responses by parents, and the absence of a healthy parent–child relationship (Loeber & Stouthamer-Loeber, 1986).

Effective interventions for delinquency prevention include training programs for parents, interven- tions by system professionals, and school- and community-based prevention. Successful training pro- grams for parents typically include parental skill enhancement and behavioral modeling, broad-based content, at least 20 contact hours, clearly identified risk factors, clinical consideration of development, identification of socioeconomic and cultural factors, a skilled staff, and follow-up (Webster-Stratton & Taylor, 2001). Skilled staff was found important for stabilizing the educational experiences of children, identifying and treating emotional and behavioral disorders, and combating abuse and neglect (Web- ster-Stratton & Taylor, 2001; Mann & Reynolds, 2006). School- and community-based prevention pro- grams that achieved the most success included preschool or early-education programs, mentoring, after-school programs, church programs (Welsh & Farrington 2007), and home- and school-based programs (Shoenfelt & Huddleston 2006; Manuel & Jørgenson 2013).

Whether reviewing individual studies, reviews of studies with common research questions, or meta-analyses of similar investigations with weighted average effects, the vast majority of research comes from the United States, and the most effective and promising prevention programs tend to be home- and school-based. Additionally, many studies have methodological issues and/or insignificant differences between treatment and comparison group outcomes. For example, Manuel and Jørgenson (2013) in their international review found only one secondary prevention study with both high- quality methodology and significant improvement for the treatment group.

STATE-MANDATED DELINQUENCY PREVENTION

In January 2002, the Ohio General Assembly approved an amendment to the Ohio Revised Code that allowed for the development of programs for youth deemed at risk of formal court involvement (Ohio State House of Representatives, 2002). This act expanded the responsibilities of Ohio juvenile courts, while enabling development of prevention programs through allocation of federal funds and other resources.

The law mandates counties “. . . to develop a service coordination process to deal with children alleged to be or at risk of becoming unruly children. . .” (Ohio State House of Representatives, 2002, p. 1). The law requires Ohio counties to place children in preventive programs. Interestingly, a sur- vey we did of U.S. juvenile justice statutes revealed only four other states—Florida, New York,

618 FAMILY COURT REVIEW

 

 

Illinois, and New Jersey—in which delinquency prevention programs are mandated by law. Finally, the scholarly literature suggests predelinquent interventions definitely showed a preference for fam- ily- and school-based programs over court- and corrections-based interventions, so the program eval- uated here may be disadvantaged. One the other hand, the Greene County Juvenile Court’s (GCJC) prevention is consistent with the literature’s emphasis on early prevention, because children as young as 6 years old are eligible for the prevention program and may be referred by parents, guardians, teachers, and counselors, as well as law enforcement (Gidley, 2002).

From a public health perspective, the type of intervention specified by Ohio is secondary preven- tion. Primary prevention programs such as Drug Abuse Resistance Education (DARE) are aimed at the entire juvenile population, secondary prevention tries to identify at-risk youth and target them for interventions, and tertiary prevention focuses on delinquency that already exists and tries to keep it from getting worse (Cohen & Chehimi, 2007). Secondary prevention has the potential to create sec- ondary deviance (Lemert, 1951; Dickerson, Collins-Camargo, & Martin-Galijatovic, 2012) by expos- ing children who have committed deviant but not necessarily delinquent acts to the trauma and stigma of juvenile court, often called net widening (Schur, 1971; Shdaimah, Bryant, Sander, & Cornelius, 2011). From the labeling perspective, early screening and intervention may have an aggra- vating effect, despite noble intentions, in which children accept labels such as “at risk,” “unruly,” or “predelinquent” in a self-fulfilling prophecy of future delinquency and crime (Merton, 1957; Macallair, 2001). For example, the corrections-based Scared Straight program is widely considered to do more harm than good (Lundman, 2001; Petrosino et al., 2013). Although the labeling perspec- tive has had a great influence on juvenile courts, a competing theory is deterrence, through which children learn their lesson and avoid future court involvement.

THE GCJC PREVENTION PROGRAM

In response to the Ohio State House of Representatives’ legislation, the GCJC created the preven- tion program within its Diversion Department in 2002. The principal form of diversion for all but fel- ony and drug-related delinquency in Greene County is Teen Court, a tertiary prevention program (Norris, Twill, & Kim, 2011). The focus of this research is the GCJC’s prevention program, initially funded by a grant from Job and Family Services, Family and Children First, and Children’s Services, with 1-year extensions after 2004 and a funding increase in 2008.

PROGRAM DESCRIPTION

As described in its brochure, the GCJC prevention program is “. . . designed to be a partnership between families and the community to assist youth who are at risk of court involvement through expanding community support, empowering families and educating children to decrease the number of youth placed on diversion or probation” (Gidley, 2002, p. 2). Services are offered to children between the ages of 6 and 17, and the program’s principal objective is to intervene before unruly or at-risk behavior escalates and requires formal juvenile court involvement; therefore, the prevention program does not accept any youth who have had prior involvement with the juvenile court. Children can be referred not only by law enforcement officials, but also by parents and guardians, school offi- cials, and children’s protective services.

The prevention program staff screens referrals to determine if program eligibility requirements have been met. During the intake process, staff members consider the child’s reported behavioral problems and family situation and make recommendations for treatment. Once children are admitted into the program, they must comply with the conditions of a prevention contract that outlines specific recommendations and rules. Recommended treatment includes case management, community service and/or restitution, Family Fun Night, mental health assessment, mentoring and tutoring, Substance Abuse Subtle Screening Inventory (SASSI), anger management classes, and Strengthening Families

Bergdahl et al/ASSESSING THE PREVENTION OF JUVENILE DELINQUENCY 619

 

 

Program (SFP). Outcomes for these program recommendations are examined later in Table 5 and descriptions of each type of treatment follows.

EXPLANATION OF PROGRAM RECOMMENDATIONS

This section describes the types of treatment that could be recommended for children. Multiple program recommendations could be made for any particular child.

Case management was a common recommendation in case files. After consulting GCJC staff, we determined that this was a catch-all category for prevention candidates who were not specifically referred to other program interventions. Community service may be recommended to promote pro- social behaviors or to make restitution, in which case the number of hours is determined by the amount of restitution owed, at $6.00 for each hour of service. Family Fun Night involves a Wednes- day dinner and game night in the GCJC building and is designed to improve pro-social behavioral interactions among family members. Mentoring and tutoring programs are offered by the court as well as by schools, churches and Big Brothers Big Sisters of America. SASSI uses a one-page sub- stance abuse screening and, based on the score, this can lead to recommendations for treatment from county mental health agencies.

The SFP is identified as “promising” by OJJDP and sequentially counsels children and their care- takers separately, together, and in groups of multiple families (Mihalic et al., 2004; Kumpfer, Xie, & Driscoll, 2012). GCJC uses two forms of SFP based on the ages of the youth involved: Program I is for children aged 5–10 and Program II is for children aged 10–14. Each program meets weekly for 7 consecutive weeks, for 2 hours per week. Parents and youth meet separately with their facilitators to learn communication skills for the first hour and then together as a family for the second hour to practice the skills they just learned. Contracts are voided if the children and families fail to complete the specified recommendations or are otherwise uncooperative, and dropouts may be referred to the formal juvenile court if their behavior warrants it.

This article examines the relationships between behaviors and characteristics leading to program referrals, treatment types, and case outcomes for youth referred to the program in order to determine what predicts future court contact.

METHOD

RESEARCH SAMPLE

All prevention program cases available in juvenile court records from the program’s inception on March 1, 2002, through January 1, 2009, were coded and entered into a data set using SPSS statisti- cal software. Variables of interest included demographic information (sex, race, age, family situation, number of siblings), source of referral, behaviors that led to the referral, court recommendations for clients, and case outcomes. Each case was assigned a three-digit code for confidentiality require- ments and reference purposes.

Table 1 describes the referral source, family situation, and case outcome and what is included in each category of these variables. There are four sources of referrals to the program (family, school, parent and school jointly, and community), four family situations of the children referred (two bio- logical parents, single biological parent, biological parent and partner, and other), and four possible client outcomes (completion with no further court referral, completion with further court referral, fail- ure to complete with no further court referral, failure to complete with further court referral).

The majority of subjects were male (61.9%), White (87.2%), and between the ages of 12 and 14 (52.8%). The ages of referred youth ranged from 5 to 19 years, and the median and modal age for referrals was 12 years. Roughly a third (34.2%) were 11 or younger and 13.1% were 15 or older.

620 FAMILY COURT REVIEW

 

 

Missing data were a problem. Case files missing both staff recommendations and case outcome ranged from 12.1% for 2002 data to 44.4% for 2008, with no consistent trend. The percentage of cases with crucial missing data varied from year to year, but on average, almost a quarter of cases had to be omitted for each year of the program’s operation. Some of the missing data can be attrib- uted to several revisions to the intake form during the 7-year period that produced the data analyzed here. Additionally, referrals to the program came from different sources—parents, teachers, school counselors, and law enforcement—in itself a source of variability. Another reason for incomplete records might be court staff prioritizing paperwork for formally adjudicated delinquency cases and lack of oversight by administrators for the data collection process.

A total of 510 case files were initially coded, after which 96 cases missing both intervention rec- ommendations and client outcomes were dropped from the analysis, as were 51 cases missing client outcomes, resulting in a sample of 362 cases.

DATA ANALYSIS

Chi-square statistics were used to analyze the bivariate relationships between case outcome (the dependent variable) and all independent variables. Cases were removed or variable categories were collapsed to ensure that the assumptions about cell size were met. The dependent variable of case outcome was dichotomized into successfully completed versus not successfully completed for multi- variate analysis. Logistic regression analysis was run separately for demographic variables,

Table 1 Coding for Referral Source, Family Situation and Case Outcomes

Variable Name and categories Categories include

Referral Source: Family Father, mother, father & mother, adoptive parent(s) and grandparent School School official Parent & School Parent and school official Community Prevention staff, diversion director, child & family counselor, probation

officer, Children Service Board, police and school & police Family Situation:

Two biological parents Natural mother & father or parents with joint custody Single biological parent Mother or father Biological parent & partner Natural mother & step parent/boyfriend or natural father & step parent/

girlfriend Other Grandparent(s), aunt/uncle, foster care, adoptive parents or guardian

Case Outcome: Successful completion without later court

involvement Successful completion of diversion program with no later court

involvement indicated in record Successful completion with later court

involvement Successful completion of diversion program with: -later return to diversion program with new charge -later traditional juvenile court -later Teen court -re-enrollment into Prevention -later court involvement not otherwise specified

Unsuccessful completion with no later court involvement

Unsuccessful completion with no later court involvement indicated in record:

-did not complete Strengthening Families -did not complete mental health assessment

Unsuccessful completion with later court involvement

Unsuccessful completion with formal charge: -sent to diversion program -sent to traditional juvenile court -sent to Teen court -later formal charge not specified

Bergdahl et al/ASSESSING THE PREVENTION OF JUVENILE DELINQUENCY 621

 

 

instigating behaviors, and recommendations. Forward logistic regression was used to select which of these independent variables best predicted the dependent variable. And finally, the enter method was used to build the simplest model to predict the dependent variable.

Of the youth referred to the prevention program, 98.3% agreed to participate, 0.3% declined, and 1.4% had missing data for this variable. As it is an elective program, those who declined to partici- pate had no further court involvement unless arrested for official delinquency. The three children and their families who volunteered for the prevention program a second time with no formal court referral were an exception.

Table 2 is a cross-tabulation of source of referral by program outcome. School officials referred the majority of clients in the sample, 33.1% of accepted cases. Referrals from concerned caretakers (parents, grandparents, or other guardians) were 19.1% of cases, and 5.5% were referred jointly by school officials and parents. Other actors in the community (police, police and schools jointly, and juvenile court staff) referred 6.9% of clients. These percentages should be interpreted with caution because the source of referral was missing for 35.4% of the 362 cases.

Chi-square statistics were used to analyze the bivariate relationships between case outcome as a dependent variable and family background type as an independent variable in Table 3. Cases were removed or variable categories were collapsed to ensure that the assumptions about cell size were met. In Table 2, source of referral was not a statistically significant predictor of program completion or later court involvement.

In Table 3, the largest proportion of clients came from families with only a single biological par- ent (35.1%), while 19.1% came from a home with two biological parents, and 15.7% were from a home with one biological parent and a partner. Two- biological-parent families had the highest per- centage of program completion (60.9%); followed by other nonfamily, which included foster care (n 5 1), adoptive parents (n 5 11), and guardians (n 5 2) with 56.2% completion; followed by those with an unknown family situation (53.3% completion) and one biological parent (45.6% completion).

Ten at-risk behaviors were coded and entered as potential independent variables into Table 4: poor school performance, disrespect of parents, shirking chores, school detention, school disruption, lying, stealing, temper tantrums or outbursts, fighting with siblings, and destructive behavior. Most problem behaviors likely occurred at home, such as disrespect of parents (63.4%), shirking chores (14.1%), and fighting with siblings (11.4%), or at school, such as poor performance (45.9%), disrup- tion (26.6%), and detention (17.7%).

For other problem behaviors, the location was not suggested by the category of behavior: destruc- tive behavior caused 18.3% of referrals, temper tantrums/outbursts accounted for 19.9% of referrals, while lying accounted for 12.7%. Stealing led to 13.0% of referrals, and other rarer and more serious norm violations tended to accrue in an “Other, please specify” category not shown in Table 4: drug

Table 2 Source of Referral by Program Outcomes

Program Outcome Family School Parent & school Community Unknown v2

Completed: no 27.5% 30.0% 30.3% 28.0% 29.7% 10.020 court involvement n519 n536 n56 n57 n538 Completed: later 23.2% 20.8% 25.0% 24.0% 26.6% court involvement n516 n525 n55 n56 n534 Incomplete: no 30.4% 23.3% 40.0% 28.0% 18.8% court involvement n521 n528 n58 n57 n524 Incomplete: later 18.8% 25.8% 5.0% 20.0% 25.0% court involvement n513 n531 n51 n55 n532 Totals 100.0% 100.0% 100.0% 100.0% 100.0%

Note: * p < .05, ** p < .01, ***p < .001

622 FAMILY COURT REVIEW

 

 

use (3.3%), runaway (1.1%), mental health issues (0.6%), physical violence (0.3%), hitting parents (0.3%), and curfew violation (0.3%). Although Table 4 is a cross-tabulation of referral behaviors by program outcomes, none of the instigating behaviors had a significant effect on the four possible out- come combinations of program completion and later court involvement.

Table 5 shows case recommendations by program outcomes. In Table 5, eight recommendations were coded as present or absent: case management, community service, Family Fun Night, mental health assessment, mentor/tutor program, SASSI, anger management, and SFP. Overall, only 53.0% of participants completed the prevention program; this is disappointing, but most referrals were for predelinquent behavior and participation was entirely voluntary. Three recommendations signifi- cantly reduced program completion: only 28% of children referred for SASSI completed the program (p 5 .023); anger management participants had a 47% completion rate (p 5 .042), and only 46% of children referred to SFP were completers (p 5 .000). Referrals for substance abuse have had signifi- cant attrition effects in previous research (Bartollas & Schmalleger, 2013). SFP is considered an effective intervention for high-risk families (Kumpfer et al., 2012), so it is not known from this result whether the families were too dysfunctional to accept the treatment, SFP was not effectively adminis- tered by the juvenile court, or if other factors were present. Anger management was statistically sig- nificant in bivariate analysis, but was not in multivariate analysis and so was dropped in the final model.

More than half (55.2%) of those who completed a prevention program had no later court involve- ment compared to 51.8% of those who did not complete their program, but this difference did not achieve statistical significance.

For those who attended the intake meeting and elected to participate in the program, staff recom- mendations included case management (55.7%), SFP (55.4%), anger management classes (22.9%), mental health assessment (15.2%), Family Fun Night (13.9%), mentoring and/or tutoring (6.9%), drug and/or alcohol screening (6.9%), and community service (4.7%).

Forward logistic regression was conducted to determine which independent variables were predic- tors of successful completion of the diversion program. The independent variables entered into the model included: demographic variables (person of color, gender, age, number of siblings, referral source, and family situation), instigating behaviors (poor school performance, disrespect of parents, shirking chores, school detention, school disruption, lying, stealing, temper tantrums or outbursts, fighting with siblings, and destructive behavior) and recommendations (case management, commu- nity service, Family Fun Night, mental health assessment, mentor/tutor program, SASSI, anger man- agement and SFP). The logistic regression model is shown in Table 6.

Prevention program participant’s family situation was statistically significant when tested by chi- square (see Table 3) but logistic regression analyzes each category separately for this type of

Table 3 Family Situation of Referral by Program Outcomes

Program Outcome

Two biological parents

Single biological parent

Biological parent & partner

Other non -family Unknown v2

Completed: no 34.8% 27.6% 29.8% 28.1% 27.3% 21.51* court involvement n524 n535 n517 n59 n521 Completed: later 26.1% 23.6% 15.8% 28.1% 26.0% court involvement n518 n530 n59 n59 n520 Incomplete: no 29.0% 22.8% 33.3% 31.2% 13.0% court involvement n520 n529 n519 n510 n510 Incomplete: later 10.1% 26.0% 21.1% 12.5% 33.8% court involvement n57 n533 n512 n54 n526 Totals 100.0% 100.0% 100.0% 100.0% 100.0%

Note: * p < .05, ** p < .01, ***p < .001

Bergdahl et al/ASSESSING THE PREVENTION OF JUVENILE DELINQUENCY 623

 

 

T a

b le

4 In

st ig

at in

g B

eh av

io rs

b y

P ro

g ra

m O

u tc

o m

es

% Y

es P

o o r

S ch

o o l

P er

fo rm

a n ce

D is

re sp

ec t

o f

P a re

n ts

S h ir

ki n g

C h o re

s S ch

o o l

D et

en ti

o n

S ch

o o l

D is

ru p ti

o n

L yi

n g

S te

a li

n g

T em

p er

T a n tr

u m

o r

O u tb

u rs

t

F ig

h ti

n g

w it

h S ib

li n g s

D es

tr u ct

iv e

B eh

a vi

o r

C o m

p le

te d :

N o

C o u rt

In v o lv

em en

t

2 8 .9

% 2 6 .6

% 2 5 .5

% 2 1 .9

% 2 6 .0

% 3 4 .8

% 2 7 .7

% 2 6 .4

% 3 1 .7

% 2 5 .8

%

C o m

p le

te d :

L at

er C

o u rt

In v o lv

em en

t

2 1 .1

% 2 2 .3

% 2 1 .6

% 2 9 .7

% 2 4 .0

% 1 7 .4

% 2 1 .3

% 2 7 .8

% 1 7 .1

% 2 4 .2

%

In co

m p le

te :

N o

C o u rt

In v o lv

em en

t

2 7 .1

% 2 8 .4

% 2 1 .6

% 2 1 .9

% 3 0 .2

% 3 2 .6

% 2 9 .8

% 2 7 .8

% 3 1 .7

% 2 8 .8

%

In co

m p le

te :

L at

er C

o u rt

In v o lv

em en

t

2 2 .9

% 2 2 .7

% 3 1 .4

% 2 6 .6

% 1 9 .8

% 1 5 .2

% 2 1 .3

% 1 8 .1

% 1 9 .5

% 2 1 .2

%

N o te

: N

o n e

w er

e st

at is

ti ca

ll y

si g n ifi

ca n t

624 FAMILY COURT REVIEW

 

 

T a

b le

5 R

ec o

m m

en d

at io

n s

b y

P ro

g ra

m O

u tc

o m

e

P ro

g ra

m O

u tc

o m

es C

a se

M a n a g e-

m en

t C

o m

m u n it

y S er

vi ce

F a m

il y

F u n

N ig

h t

M en

ta l

H ea

lt h

A ss

es sm

en t

M en

to r/

T u to

r P

ro g ra

m S A

S S I

A n g er

M a n a g em

en t

S tr

en g th

en in

g F

a m

il ie

s

C o m

p le

te :

N o

C o u rt

In v o lv

em en

t

2 7 .4

% 2 3 .5

% 2 6 .0

% 3 4 .5

% 2 0 .0

% 4 .0

% 3 0 .1

% 2 9 .0

%

C o m

p le

te :

L at

er C

o u rt

In v o lv

em en

t

2 3 .4

% 4 1 .2

% 2 4 .0

% 2 7 .3

% 3 6 .0

% 2 4 .0

% 1 6 .9

% 1 7 .0

%

In co

m p le

te :

N o

C o u rt

In v o lv

em en

t

2 9 .4

% 1 7 .6

% 3 2 .0

% 2 5 .5

% 3 2 .0

% 4 0 .0

% 3 4 .9

% 3 2 .0

%

In co

m p le

te :

L at

er C

o u rt

In v o lv

em en

t

1 9 .9

% 1 7 .6

% 1 8 .0

% 1 2 .7

% 1 2 .0

% 3 2 .0

% 1 8 .1

% 2 2 .0

%

T o ta

ls 5 5 .5

% 4 .7

% 1 3 .8

% 1 5 .2

% 6 .9

% 6 .9

% 2 2 .9

% 5 5 .2

% C

h i-

sq u ar

e 6 .6

9 2 .9

8 2 .1

4 3 .8

7 4 .4

3 9 .5

2 *

8 .1

8 *

1 9 .5

6 * * *

N o te

: T

o ta

ls re

fl ec

t p er

ce n ta

g es

o f

o v er

al l

re co

m m

en d at

io n s.

D o es

n o t

to ta

l 1 0 0 %

b ec

au se

m u lt

ip le

re co

m m

en d at

io n s

co u ld

b e

m ad

e.