peace builders research-combo

Major PeaceBuilders Research and Program Articles The following publications are attached, which provide information and background on the large-scale, community-based randomized control trial evaluating a theoretically-driven violence-prevention program, PeaceBuilders®.

References Cited

Embry, D. D., Flannery, D. J., Vazsonyi, A. T., Powell, K. E., & Atha, H. (1996). PeaceBuilders: A theoretically driven, school-based model for early violence prevention. American Journal of Preventive Medicine, 12(5, Suppl), 91.

This paper describes the theory, experimental design and baseline findings of the randomized control group study.

Krug, E. G., Brener, N. D., Dahlberg, L. L., Ryan, G. W., & Powell, K. E. (1997). The impact of an elementary school-based violence prevention program on visits to the school nurse. American Journal of Preventive Medicine, 13(6), 459-463.

This paper reports on reductions in illnesses and injuries as measured by nurses’ office visits. This is possibly the first experimental proof of actual proximal reductions in violent injuries as a result of a violent-injury prevention effort.

Flannery, D. J., Vazsonyi, A. T., Liau, A. K., Guo, S., Powell, K. E., Atha, H., et al. (2003). Initial behavior outcomes for the PeaceBuilders universal school-based violence prevention program. Developmental Psychology, 39(2), 292-308.

The paper provides teacher and self-report outcomes using standardized measures related to social competence and aggression. Both measures have strong prediction for life-course aggression or resiliency.

Vazsonyi, A. T., Belliston, L. M., & Flannery, D. J. (2004). Evaluation of a School-Based, Universal Violence Prevention Program: Low-, Medium-, and High-Risk Children. Youth Violence and Juvenile Justice, 2(2), 185-206.

This paper reports on differential effects of the intervention for the most-high risk youth, based on baseline data.

The People Magazine article from April 5, 1999 tells the story of the use of PeaceBuilders in Salinas, CA and other communities.

Initial Behavior Outcomes for the PeaceBuilders Universal School-BasedViolence Prevention Program

Daniel J. FlanneryKent State University

Alexander T. VazsonyiAuburn University

Albert K. LiauKent State University

Shenyang GuoUniversity of North Carolina at Chapel Hill

Kenneth E. PowellCenters for Disease Control and Prevention

Henry AthaPima County (AZ) Community Services Department

Wendy VesterdalUniversity of Arizona

Dennis EmbryPAXIS Institute

PeaceBuilders is a universal, elementary-school-based violence prevention program that attempts to alterthe climate of a school by teaching students and staff simple rules and activities aimed at improving childsocial competence and reducing aggressive behavior. Eight matched schools (N � 4,000 students inGrades K–5) were randomly assigned to either immediate postbaseline intervention (PBI) or to a delayedintervention 1 year later (PBD). Hierarchical linear modeling was used to analyze results from assess-ments in the fall and spring of 2 consecutive school years. In Year 1, significant gains in teacher-ratedsocial competence for students in Grades K–2, in child self-reported peace-building behavior in GradesK–5, and reductions in aggressive behavior in Grades 3–5 were found for PBI but not PBD schools.Differential effects in Year 1 were also observed for aggression and prosocial behavior. Most effects weremaintained in Year 2 for PBI schools, including increases in child prosocial behavior in Grades K–2.Implications for early universal school-based prevention and challenges related to evaluating large-scaleprevention trials are discussed.

Despite recent downturns in national rates of violence perpetra-tion by juveniles, a significant number of young people remainboth perpetrators and victims of interpersonal violence (Dahlberg,1998; Mercy & Potter, 1996; Sickmund, Snyder, & Poe-Yamagata,1997; Snyder & Sickmund, 1999). For example, though the overallhomicide rate in the United States has declined, rates for homicideand nonfatal injuries among children and adolescents remain at

significantly high levels (Snyder & Sickmund, 1999). The propor-tion of young people who self-report having committed seriousacts of violence has also held steady since peaking in the early1990s (Snyder, 2000).

Violence occurs at home, in neighborhoods, and at school.Many recent studies illustrate the impact that exposure to violenceand victimization from violence have on mental health and behav-

Daniel J. Flannery and Albert K. Liau, Institute for the Study andPrevention of Violence, Kent State University; Alexander T. Vazsonyi,Department of Human Development and Family Studies, Auburn Univer-sity; Shenyang Guo, School of Social Work, University of North Carolinaat Chapel Hill; Kenneth E. Powell, Centers for Disease Control andPrevention, Atlanta, Georgia; Henry Atha, Pima County Community Ser-vices Department, Tucson, Arizona; Wendy Vesterdal, Department ofFamily and Consumer Resources, University of Arizona; Dennis Embry,PAXIS Institute, Tucson, Arizona.

Albert K. Liau is now at the Psychological Studies Group, NationalInstitute of Education, Singapore. Kenneth E. Powell is now at the GeorgiaDepartment of Human Resources, Division of Public Health, Atlanta,Georgia.

This project was supported in part by Cooperative Agreements U81-CCU010038–03 and U81-CCU513508–01 from the National Center for

Injury Prevention and Control, Centers for Disease Control and Prevention,Atlanta, Georgia.

We gratefully acknowledge the contributions of Laura Williams, KellyWester, Laurie Biebelhausen, and Lara Belliston to data management andanalysis. We also appreciate the comments of Thomas Simon on an earlierversion of the manuscript. We are grateful to the students, staff, and parentsin the Sunnyside and Tucson unified school districts for their ongoingsupport and participation.

PeaceBuilders is a registered trademark of Heartsprings, Inc. The use oftrade names is for identification only and does not constitute endorsementby the U.S. Public Health Service or the U.S. Department of Health andHuman Services.

Correspondence concerning this article should be addressed to Daniel J.Flannery, Institute for the Study and Prevention of Violence, Kent StateUniversity, 230 Auditorium Building, Kent, Ohio 44242. E-mail:[email protected]

Developmental Psychology Copyright 2003 by the American Psychological Association, Inc.2003, Vol. 39, No. 2, 292–308 0012-1649/03/$12.00 DOI: 10.1037/0012-1649.39.2.292

292

ior, including an increased risk for engaging in violent behavior(Elliott, Hamburg, & Williams, 1998; Flannery, 1997; Singer,Anglin, Song, & Lunghofer, 1995; Singer et al., 1999). Althoughthe risk of homicide victimization at school remains low (Kachuret al., 1996), levels of exposure to violence and victimization fromviolence at school remain high, particularly for elementary andmiddle school children (Kaufman et al., 2000; Singer et al., 1999).While recent data suggest a decline in the number of studentscarrying weapons to school (7% of high school students werefound to have done so within the previous 30 days; Kann et al.,2000), the use of firearms and other weapons has heightened thelethality of violence among young people (Rushforth & Flannery,1999) and has significantly increased the likelihood that specificconflicts will escalate into lethal exchanges (Fagan & Wilkinson,1997). In fact, despite recent declines in gun use and lethal formsof violence, the proportion of young people involved in nonfatalviolence has not declined (Snyder, 2000). Arrest rates for aggra-vated assaults remain almost 70% higher than they were in 1983,and this is the offense most frequently captured in self-reportsof violence (U.S. Department of Health and Human Services[USDHHS], 2001).

The data are clear. Violence among young people remains asignificant public health problem (USDHHS, 2001). Althoughmany school and community-based violence prevention programsexist, relatively few have been rigorously evaluated (Sherman etal., 1997; Thornton, Craft, Dahlberg, Lynch, & Baer, 2000). Ifpsychologists are to inform public policy and facilitate risk pre-vention for young people, it is imperative that we identify, throughapplied evaluation studies, programs that effectively prevent youthviolent behavior and its associated precursors (i.e., aggression) andrigorously evaluate the behavioral outcomes associated with theseinterventions (Powell & Hawkins, 1996; Satcher, Powell, Mercy,& Rosenberg, 1996; USDHHS, 2001). In the present study, weexamined the potential impact of a universal elementary-school-based violence prevention program on student aggression andsocial competence.

Preventive Interventions

The need to provide early prevention is illustrated by the mul-titude of studies that show that violent behavior occurs along adevelopmental continuum of behavioral severity (e.g., Flannery &Huff, 1999; Flannery & Williams, 1999; Tolan, Guerra, & Ken-dall, 1995; Tremblay, et al., 1992). The precursors to more seriousviolence perpetration in adolescence (e.g., homicide, assault) areyoung children’s aggressive behaviors such as hitting, kicking, andverbal insults and threats (Conduct Problems Prevention ResearchGroup [CPPRG], 1999; Dahlberg, 1998; Huesmann et al., 1996;Huesmann & Moise, 1999; Singer & Flannery, 2000; Stoolmiller,Eddy, & Reid, 2000; Tremblay, Pagani-Kurtz, Masse, Vitaro, &Pihl, 1995). These are the triggers that can escalate interpersonalconflict into violence and are the behaviors that need to be targetedin preventive interventions in elementary schools. Young peoplewithout the skills and competencies to resolve conflicts or solveproblems are at increased risk for violence victimization andperpetration (Lochman & Dodge, 1994). Longitudinal research hasconsistently demonstrated that aggressive, peer-rejected childrenin first grade are at increased risk for engaging in delinquent,violent behavior in adolescence (Hawkins et al., 2000; Loeber &

Farrington, 1998; Tolan & Gorman-Smith, 1998; Tremblay et al.,1992, 1995; Walker, Colvin, & Ramsey, 1995) and for becomingantisocial adults (Eron & Huesmann, 1990).

Promising studies exist showing that the developmental trajec-tory of youth violence may be altered (CPPRG, 1999; Dahlberg,1998; Englander-Golden, Jackson, Crane, Schwarzkopf, & Lyle,1989; Hawkins, 1995; Howard, Flora, & Griffin, 1999; Reid,Eddy, Fetrow, & Stoolmiller, 1999; Stoolmiller et al., 2000; Trem-blay et al., 1991). Several studies have now demonstrated thataggressive behavior can be reduced by altering the social environ-ments at school (Farrell & Meyer, 1997; Gottfredson, 1997;Greenberg, Kusche, Cook, & Quamma, 1995; Grossman etal.,1997; Reid et al., 1999; Stoolmiller et al., 2000), particularly byemphasizing rewards and praise for prosocial behavior (CPPRG,1999; Walker et al., 1995) and improving social competence(Hawkins, Catalano, Kosterman, Abbott, & Hill, 1999; O’Donnell,Hawkins, Catalano, Abbott, & Day, 1995) while reducing cues thatmight increase hostility (Lochman & Dodge, 1994).

The Good Behavior Game (GBG) is one type of school-basedprevention program that has established clear evidence of reducedaggressive behavior and other forms of child and adolescent prob-lem behavior such as tobacco use and poor academic achievement(Kellam & Anthony, 1998; Kellam, Ling, Merisca, Hendricks, &Ialongo, 1998; Salend, Reynolds, & Coyle, 1989). The GBG usesclassroom behavior management as the primary means of reducingaggression and problem behavior. Student teams are rewarded byteachers if no member of a team exhibits undesirable behaviorswhile engaged in game sessions. Teachers begin by rewardingteams with tangible reinforcers and then gradually move to lesstangible rewards.

The Linking the Interests of Families and Teachers (LIFT)program has also used the GBG as a core element of a 10-weekuniversal preventive intervention strategy to reduce aggression andincrease social competence (Reid et al., 1999; Stoolmiller et al.,2000). The overall intervention consisted of parent training, theGBG program, and systematic communication between teachersand parents. The intervention had immediate and significant ef-fects on physical aggression among students on the playground aswell as some impact on increased child social competence. Exam-ination of the LIFT program outcomes has also shown strongdifferential effects of treatment, with children highest on aggres-sion at baseline benefiting the most from the intervention (Reid etal., 1999; Stoolmiller et al., 2000). These short-term and differen-tial effects on aggression are important to consider given thepressure to demonstrate significant behavior change with relativelybrief school-based interventions.

Another program of research has been conducted by the Con-duct Problems Prevention Research Group (CPPRG, 1999). TheCPPRG has implemented and evaluated the Fast Track preventiontrial for conduct problem behavior for elementary school childrenat high risk for long-term antisocial behavior. Fast Track is adevelopmentally based, long-term multicomponent and multisiteintervention that has been evaluated using a randomized designwith a nonintervention control group. After 1 year of intervention(from kindergarten to Grade 1), the group found moderate positiveeffects on children’s social competence and conduct problems,including child aggressive behavior, for children in the interven-tion schools compared with children in the control group.

293SPECIAL ISSUE: VIOLENCE PREVENTION

Several other programs of research have demonstrated childbehavior change in the areas of improved social competence orreductions in aggressive behavior in the classroom or on theplayground (e.g., Grossman et al., 1997; Hawkins et al., 1999;Tremblay et al., 1995; also see Thornton et al., 2000; USDHHS,2001). In the current study, we sought to expand on these studiesby examining the effects of an early elementary-school-baseduniversal preventive intervention program called PeaceBuilders.This program attempts to alter individual child behavior—in par-ticular, to reduce aggressive behavior and increase social compe-tence—by changing the culture or climate of an entire school.There is some evidence that PeaceBuilders affects the incidence ofassault-related and violent injury. Specifically, Krug and col-leagues (Krug, Brener, Dahlberg, Ryan, & Powell, 1997) foundthat the frequency of injuries due to fighting for children in GradesK–5 whose schools were randomized to PeaceBuilders did notincrease over a 1-year period, although the incidence of injuriesdue to fighting for children in control schools increased 56% overthe same period. Although these are meaningful archival data, wereport here on teacher and child self-reports of social competenceand aggression, which have high predictive value for long-termprevention efforts (CPPRG, 1999; Tolan et al., 1995; Tremblay etal., 1995; Vazsonyi, Vesterdal, Flannery, & Belliston, 1999;Walker et al., 1995). School is a logical public health setting forchanging the cognitive, social, and imitative characteristics ofchildren at risk for violence. For example, schools can be thoughtof as large antecedent and reinforcement systems that can increaseor decrease antisocial and prosocial behavior (Mayer & Sulzer-Azaroff, 1990). We still lack consistent evidence of whether arelatively low-cost, widely implemented universal preventive in-tervention approach in the early elementary grades will lead tosignificant and sustainable behavior change.

The PeaceBuilders Program

PeaceBuilders is a universal school-wide violence preventionprogram for elementary schools (Grades K–5) implemented by allstaff and students in a school (Embry, Flannery, Vazsonyi, Powell,& Atha, 1996). PeaceBuilders focuses on individual behaviorchange in proximal interpersonal and social settings (Tolan &Guerra, 1994). The program incorporates an ongoing, long-termstrategy to alter the climate and culture of the entire school (Embry& Flannery, 1999; Embry et al., 1996; Flannery, 1997). Theintervention is purposely woven into the school’s everyday routinerather than presented as a time- or subject-limited curriculum.Thus, PeaceBuilders is not offered as a set number of sessions orhours per week but includes activities that can be implemented ona daily basis in any classroom by any teacher or staff person.Specifically, PeaceBuilders attempts to change characteristics ofthe setting (antecedents) that trigger aggressive, hostile behavior,and it increases the daily frequency and salience of both live andsymbolic prosocial models. If there are more prosocial cues andmodels in a school and these behaviors are consistently reinforcedand rewarded, then over time, child social competence will in-crease and the frequency and intensity of aggressive behaviors willdecline. PeaceBuilders specifically rewards prosocial behaviorsand provides strategies to avoid the differential or accidentalreinforcement of negative behaviors and conflict that sometimeshappens with conflict mediation programs (Webster, 1993).

All children and staff in a school learn five simple rules via acommon language, which makes the intervention easy to learn andmaintain: (a) praise people, (b) avoid put-downs, (c) seek wisepeople as advisers and friends, (d) notice and correct hurts wecause, and (e) right wrongs. To help students learn these principles,PeaceBuilders includes (a) daily rituals related to its language andprinciples that are meant to foster a sense of belonging; (b) cuesand symbols that can be applied to diverse community settings; (c)specific prompts to “transfer” across people, behaviors, and time;and (d) new materials or strategies introduced for times andcircumstances when positive behavior might otherwise decay (Em-bry, 1980; Embry et al., 1996; Stokes & Baer, 1977).

For example, staff and students are encouraged to use “praisenotes” to pay attention to and reinforce positive, prosocial behaviorin the classroom, at school, and at home. “Peace feet” might beplaced by the drinking fountains to encourage children not to cutin line while waiting their turn, and students are sometimes sent tothe principal for kind acts or good deeds rather than just fordiscipline problems (principal “preferrals”). PeaceBuilder rulesand principles are prominently displayed throughout the school,and students complete activities from a specially designed comicbook in which they are the designated hero (see Embry et al.,1996). Adults more actively monitor “hot spots” in school such aslunchrooms and hallways in between activities, praising prosocialbehavior. All of these strategies and activities are geared towardcreating a positive climate and culture in the entire school, with anemphasis on reinforcement of positive behavior rather than simplythe reduction of negative behavior.

The training of teachers in the implementation of the presentintervention had several phases, including a preintervention orien-tation for all faculty and staff of the schools, a half-day trainingworkshop on the basic PeaceBuilders model, and extensive sitecoaching (on average, 2 hr per week) in the first 3 to 4 months ofthe intervention and then on an as-needed basis. All training andcoaching were conducted by the model developer (Embry et al.,1996) as a means of facilitating internal validity. Each participat-ing school also received specific in-service sessions on importantissues identified by staff (e.g., implementing activities with specialneeds children), periodic group forums to discuss successes andchallenges to implementation, and occasional 1-day institutes thatfocused on applying and creating new materials and interventions.Attendance was voluntary at the institutes and forums. Additionaldescription of program materials and training is available else-where (e.g., Embry et al., 1996).

Hypotheses

Our hypothesis was that youth aggressive behavior would bereduced by initiating prevention early in childhood and by increas-ing children’s resilience and social competence. A dual focus onreducing aggression and increasing social skills and competenciesis important because the prognosis for children with a combinationof low social competence, aggressiveness, and poor emotional andcognitive preparation is poor (CPPRG, 1999; Kellam, Mayer,Rebok, & Hawkins, 1998; Tolan et al., 1995; Weissberg & Bell,1997). We also examined the differential effectiveness of theintervention given evidence that treatment outcome effects mayvary depending on a child’s initial behavior status prior to partic-ipating in an intervention (Reid et al., 1999; Stoolmiller et al.,

294 FLANNERY ET AL.

2000). We examined both short-term change in aggression andcompetence (compared with controls) over the 1st year of inter-vention and longer-term change in Year 2, when all schoolsreceived intervention. Specifically, in the 1st year, we expectedthat children in the intervention-school group, compared withthose in the control-school group, would report greater improve-ments in social competence and greater reductions in aggressivebehavior. By the end of the 2 school years, when both groups werereceiving the intervention, we expected that, relative to baselinelevels, students in both conditions would exhibit significant in-creases in competence and prosocial behavior and decreases inaggressive behavior.

Method

The study protocol was approved by the Institutional Review Board forHuman Subjects at the University of Arizona in Tucson and by therespective schools’ research review committees. Parents were notified ofthe project through letters mailed to their homes and by school-distributednewsletters. Parents were given the opportunity to withdraw their childfrom any data collection. Students were also informed that their participa-tion was voluntary and were provided an opportunity for alternative class-room activities if they chose not to take part. If a student was engaged inanother activity (e.g., band class), we returned to attempt to gather infor-mation at a later date. At the time of survey administration, students wereasked to give oral assent, and questions were answered regarding theirparticipation. All students received rewards such as stickers or pencils forcompleting the surveys and interviews.

Eight elementary schools (Grades K–5) in Pima County, Arizona, wereselected from two large school districts to participate on the basis of havinghigh rates of juvenile arrests and histories of suspensions and expulsions.After we met with school administrators to discuss the purpose and scopeof the study, all schools that were initially contacted agreed to participate.Schools were located in all areas of town, including some in the central cityand others on the outskirts of town. One of the eight schools consisted ofa pair of schools in the same neighborhood, a school for Grades K–2 anda school for Grades 3–5 (approximately 1 block apart), and was treated asa single school for pairing, intervention, analysis, and discussion (School2A). All of the other schools were self-contained Grades K–5 schools. Oneschool that was randomly assigned to the delayed intervention condition(School 1B) did not gather initial baseline data but joined the study at

Time 2 in the spring of Year 1. All participating schools remained in thestudy through the first 2 intervention years.

Design and Procedure

Prior to baseline data collection, the eight project schools were matchedinto four pairs primarily on the basis of geographic proximity, but we alsoconsidered the percentage of ethnic students, the percentage of studentseligible for free or reduced-price lunch, and the percentage of students inEnglish as a Second Language (ESL) classrooms (see Table 1). School 2Acontained fewer Hispanic and more Native American students than itscomparison School 2B, but these schools were paired because of their closegeographic proximity. Four schools were then randomly assigned asPeaceBuilders immediate intervention (PBI) schools and began the pro-gram in the fall of 1994 immediately following baseline data collection.The remaining schools began the PeaceBuilders program in 1995 after 1year of baseline data collection and are hereafter referred to as PeaceBuild-ers delayed (PBD) schools (see Figure 1). PBD schools received compen-sation in Year 1 ($1,000) as an incentive for them not to engage in anyPeaceBuilders program-related activities.

We randomized at the school level because all students and staff in aschool were exposed to and participated in the intervention. Students in thefour PBI schools were exposed to PeaceBuilders for a total of 2 schoolyears, and PBD schools participated in the intervention for 1 school yearbetween the fall and spring semesters of Year 2. Owing to limited re-sources, we did not collect any child self-report data from new kindergartenstudents in Year 2, we collected Grades 1 and 2 child self-reports only forstudents who had participated in Year 1, and we did not follow Year 1 fifthgraders into sixth grade. Further, students new to PBI schools in Year 2 ofthe intervention were not included in these analyses.

Students in Grades 3–5 completed 100-item self-report surveys at eachdata collection point. Surveys were administered in classrooms of about 20students with at least two research assistants present to read the entiresurvey aloud and to answer questions. This procedure resulted in fewsurveys with missing or incomplete data. Surveys were pilot tested in twoelementary schools prior to data collection to assess the appropriateness ofthe items for young children. All child survey items were answered withthe anchors no, a little, or a lot.

For students in Grades K–2, self-report data were collected throughindividual 20-item, face-to-face interviews. The 20 items were pilot testedwith same-age children. Owing to time constraints (we were only able tointerview as many children as time permitted during a single class period),

Table 1School-Level Demographic Characteristics (%) of Matched Pairs

Matchedschools Caucasian

AfricanAmerican Hispanic

NativeAmerican

AsianAmerican

Freeluncha

ESLpairsb

1A (n � 704) 63.3 9.7 22.7 0.6 3.7 55 51B (n � 551) 62.5 14.6 18.5 1.9 2.5 58 8

2A (n � 817) 11.6 0.2 33.5 54.6 0.3 94 62B (n � 377) 29.4 5.2 62.2 1.7 1.4 60 29

3A (n � 550) 8.8 2.8 74.4 13.4 0.6 60 293B (n � 573) 4.8 0.8 91.8 2.5 0.3 94 68

4A (n � 327) 28.0 2.8 65.9 2.1 1.3 89 284B (n � 780) 36.0 3.5 58.5 1.0 1.0 73 21

Note. “A” schools are those randomly assigned to the PeaceBuilders immediate (PBI) intervention, whichoccurred immediately after baseline data collection. “B” schools were assigned to the PeaceBuilders delayed(PBD) condition. a Percentage eligible for federally funded free or reduced-price lunch programs. b Studentsfor whom English was their second language.


we randomly preselected 50% of students in each kindergarten, first-grade,and second-grade class to be interviewed. Individual interviews, whichtook about 5–8 min to complete, were conducted at a table outside of thechild’s classroom in a quiet area. In Grades K–2 in the participatingschools, the classes averaged about 20 students per classroom. By ran-domly preselecting half, we were attempting to target about 10 students perclass. Although there were no refusals of children in Grades K–2 toparticipate, on average we were able to complete 8 interviews per class, foran effective participation rate of 80%. Reasons for not interviewing all 10children included the following: Students were absent on the day of datacollection; students were engaged in an alternative school activity duringthe time interviews were conducted (e.g., band class), or we ran out of time.Limited time and resources precluded our being able to interview childrenat a later date. Self-report interviews in Year 2 were conducted only foravailable students who were interviewed in Year 1. Teachers continued toreport via surveys on all kindergarten, first-, and second-grade children intheir classrooms.

At the time of each data collection, teachers of children in Grades K– 5completed a 45-item instrument for each student in their classes. Teachersprovided written consent prior to participation. For Grades 3–5, bothstudents and teachers answered questions on bubble scan sheets thatcontained preassigned identification codes for data-tracking purposes. Nonames appeared on student data collection instruments. The preassigned IDcode allowed us to distribute numbered surveys to specific students on theday of data collection as well as to link student and teacher data over time.All student and teacher surveys were available in both English and Spanish.Schools received compensation for their general funds depending on thepercentage of teachers who completed surveys (e.g., $300 for 90% teacherparticipation).

Sample

On average, students across the 2 years examined were mostly Hispanic(51%), followed by Caucasians (28%), Native Americans (13%), AfricanAmericans (6%), and Asian Americans (1.5%). Seventy-one percent (n

� 1,101) of students reported that “Mom” took care of them the most, 15%reported “Dad,” 7% reported some other relative, and 2% each reported astepparent or some other adult. According to parent reports at Time 2 (n �809), 63% of children lived in homes with both parents present, 16% werefrom mother-only homes, and 12% lived with “one parent and otheradults.” Parent reports of household incomes, although based on a sub-sample of our families, were evenly distributed among the lower range ofsocioeconomic groups: 22% reported an annual household income of$7,000 or less; 19%, an income between $7,000 and $15,000; 24%, anincome between $15,000 and $25,000; 23%, an income between $25,000and $40,000; and 12%, an income greater than $40,000 per year. Themajority of our parents had completed the equivalent of high school or less:15% completed less than ninth grade; 12% completed less than highschool; 28% completed high school; 38% completed some college; and 7%completed 4 or more years of college. Compared with 1990 U.S. Censusdata, our sample was similar to the population of the metropolitan area(Pima County, AZ) on family composition, household income, and parentlevel of education. The only exception was for child ethnicity. In general,our sample comprised higher percentages of minority children (and thusfewer Caucasians) than were in the greater metropolitan area from whichthe sample was drawn.

Student and teacher sample sizes are reported in Figure 2. Studentresponse rates ranged from 86% to 93%, and teacher response rates from75% to 86%. Fewer than 1% of parents chose to withdraw their child fromany of the data collections. Similarly, fewer than 1% of children availableat each data collection time refused to complete a survey or interview,usually citing disinterest.

Variables and Instrumentation

Demographic variables. Demographic information gathered from stu-dents included age, gender, and grade in school. Teachers reported onchildren’s ethnicity by categorizing them into one of six groups: Hispanic,Caucasian, Native American, African American, Asian American, andother.

Figure 1. Overview of project design, data collection, and intervention schedule.

296 FLANNERY ET AL.

Aggressive behavior. Teachers reported on child aggressive behaviorusing items adapted from the Aggressive Behavior subscale of Achen-bach’s (1991) Teacher Report Form (TRF). The TRF has been usedextensively as both a clinical screening instrument and in large surveyresearch to assess child externalizing behavior problems (Achenbach,

1991; Grossman et al., 1997). The 25-item Aggressive Behavior subscaleasks teachers to rate child behavior on a 3-point scale in which 0 � nottrue, 1 � somewhat or sometimes true, and 2 � very true or often true. Theitems demonstrated high internal reliability (� � .95 at baseline) in oursample.

Figure 2. Student and teacher sample sizes at each data collection point. The unit of randomization was theschool. aOf the children selected to be sampled, only 50% of the students in Grades K–2 were targeted toparticipate in the child self-report portion of the study.


10.1177/1541204003262224ARTICLEYouth Violence and Juvenile JusticeVazsonyi et al. / EVALUATION OF A VIOLENCE PREVENTION PROGRAM

EVALUATION OF ASCHOOL-BASED, UNIVERSALVIOLENCE PREVENTION PROGRAM:

Low-, Medium-, and High-Risk Children

Alexander T. VazsonyiLara M. BellistonAuburn University

Daniel J. FlanneryKent State University

The current investigation examined the differential effectiveness of PeaceBuilders, alarge-scale, universal violence prevention program, on male and female youth identi-fied as low, medium, or high risk for future violence. It included eight urban schools ran-domly assigned to intensive intervention and wait-list control conditions. The currentsample included N = 2,380 predominantly minority children in kindergarten throughfifth grade. Results indicated differential effectiveness of the intervention, by level ofrisk; high-risk children reported more decreases in aggression and more increases insocial competence in comparison to children at medium and low levels of risk. Findingsadd to a growing number of promising science-based prevention efforts that seek toreduce aggression and increase social competence; they provide encouraging evidencethat relatively low-cost, schoolwide efforts have the potential to save society millions invictim, adjudication, and incarceration costs.

Keywords: aggression; social competence; violence prevention; ethnicity

Young people are the primary perpetrators, victims, and often witnesses of interper-sonal violence in our society (Snyder & Sickmund, 1999). Children who live in a climate ofviolence learn to suppress empathy and learn that violence is an acceptable means to achiev-ing their goals (Beland, 1996). This growing problem is evident in national crime statistics.

185

Authors’ Note: A previous version of this article was presented at the 9th Biennial Meetings of the Society forResearch on Adolescence in New Orleans (April 2002). This project was supported in part by cooperative agree-ments from the National Center for Injury Prevention and Control, Centers for Disease Control and Prevention(#U81-CCU010038-03 and #U81-CCU513508-01), Atlanta, GA. We would like to thank Dennis Embry as wellas students, staff, and parents in the Sunnyside and Tucson Unified School Districts for their participation.PeaceBuilders is a registered trademark of Heartsprings, Inc. The use of trade names is for identification only anddoes not constitute endorsement by the Public Health Service or the U.S. Department of Health and Human Ser-vices. Please address correspondence related to this article to Alexander T. Vazsonyi, Ph.D., Dept. of HumanDevelopment and Family Studies, 284 Spidle Hall, Auburn, AL 36849; e-mail: [email protected].

Youth Violence and Juvenile Justice, Vol. 2 No. 2, April 2004 185-206DOI: 10.1177/1541204003262224© 2004 Sage Publications

Even though recent publications of the National Crime and Victimization Survey (NCVS)and Uniform Crime Reports (UCR) note continued decreases in violence over the past sev-eral years (U.S. Department of Justice [USDOJ], 2001a, 2001b), juvenile violence remainshigh. Although firearm-related homicides have decreased, the youth homicide rate hasremained fairly stable; moreover, the overall youth violence index, assault with injury, androbbery with a weapon have increased (U.S. Department of Health and Human Services[USDHHS], 2001). In addition, a cross-national comparison shows that the rate of adoles-cent homicides involving a firearm is over 15 times higher in the United States than in 12European countries combined (Centers for Disease Control and Prevention [CDC], 1997).

Prevention and intervention efforts designed to ameliorate violence have identified anumber of individual, family, school, peer, and community risk factors that contribute todelinquency and future violence (Andrews & Trawick-Smith, 1996; Consortium on theSchool-Based Promotion of Social Competence, 1994). Although many of these factors canhelp identify individuals at risk for problem behaviors, good prevention efforts need to tar-get risk factors most amenable to change, such as skills training, behavior monitoring andreinforcement, behavioral techniques for classroom management, and building schoolcapacity (USDHHS, 2001). A number of individual-level risk factors can be targeted byviolence prevention programs. Such factors include general offenses, substance use,aggression, problem behaviors, and antisocial attitudes (Gottfredson, 2001; USDHHS,2001). Several of these risk factors are highly confounded with rates of deviance; however,the most salient behavioral predictor of later violence and delinquency is early aggressionbetween ages 8 and 10 years (Farrington, 1987; Gottfredson, 2001; Hawkins et al., 1998;Lipsey & Derzon, 1998; Loeber & Dishion, 1983; O’Donnell, Hawkins, & Abbott, 1995;Viemerö, 1996; USDHHS, 2001). Furthermore, aggression in the school context is highlyproblematic during grade school as it violates peer group and social norms (Bierman &Montminy, 1993; Coie & Dodge, 1998). Cross-sectional research has demonstrated thatchildhood aggression can foretell official delinquency status (Vazsonyi, Vesterdal,Flannery, & Belliston, 1999). Longitudinal investigations have also demonstrated thataggressive behavior is relatively stable over time and part of a general pattern of antisocialbehavior that is associated with later self-reported violence, arrests, and convictions for vio-lent offenses (Farrington, 1987; Lipsey & Derzon, 1998; Loeber, Farrington, Stouthamer-Loeber, Moffitt, & Caspi, 1998; Viemerö, 1996).

Most violence and delinquency prevention research has focused on reducing aggres-sion; however, researchers have also emphasized protective factors that may interact withrisk factors to buffer or reduce risk of future violence (Bierman, Miller, & Stabb, 1987; Coie& Koeppl, 1990; USDHHS, 2001). Individual-level factors that protect against delin-quency include a positive social orientation and an intolerant attitude toward interpersonalviolence and deviance (USDHHS, 2001). Recent prevention efforts have targeted behav-ioral measures of social competence and prosocial skills (e.g., Blechman, 1996; O’Donnellet al., 1995). Children who lack these skills are more likely to rely on their negative pat-terns of interaction and demonstrate more negative behavioral outcomes (Ollendick, Weist,Borden, & Greene, 1992; Quinn, Mathur, & Rutherford, 1995; Walker & McConnell,1988). However, few rigorous evaluation studies have been completed examining risk andprotective factors of juvenile violence and deviance.

A next important step for researchers is to identify how risk and protective factorswork together to influence problem behaviors. Kupersmidt, Coie, and Dodge (1990) foundthat aggressiveness and social competence predicted delinquency in elementary school.Similarly, Hämäläinen and Pulkkinen (1995) found that rates of recidivism were greater for

186 Youth Violence and Juvenile Justice

criminals who had been more aggressive and less prosocial when they were young. Con-versely, in a review of intervention programs, Coie and Koeppl (1990) observed that a num-ber of programs targeting aggressive, disruptive, and rejected children focused their pre-vention efforts on increasing prosocial behaviors and paid insufficient attention to reducingaggressive behaviors (e.g., Frey, Hirschstein, & Guzzo, 2000; Gottfredson, Gottfredson, &Skroban, 1998; Prinz, Blechman, & Dumas, 1994). Both behaviors must be changed toachieve intervention effectiveness (Blechman, 1996; Coie & Koeppl, 1990; Gresham &Elliott, 1987; USDHHS, 2001; Wasserman & Miller, 1998). Therefore, rigorous evaluationstudies of programmatic efforts should focus on risk and protective factors, and they shouldevaluate how both of these behaviors change following an intervention.

School-Based Violence Prevention

Recent violence prevention efforts have shifted to large-scale, universal program-matic efforts (Powell et al., 1996). Although prevention efforts have occurred in multiplecontexts, school-based interventions have several advantages (Catalano, Arthur, Hawkins,Berglund, & Olson, 1998; Gottfredson, 2001; USDHHS, 2001). For example, schools arean optimal setting for preventions and interventions; children spend a great deal of time atschool with teachers and peers, and large groups of at-risk children can be easily targeted(Beland, 1996; Blechman, 1996). Effective strategies for universal school implementationinclude behavioral monitoring and reinforcement, classroom management, and skills train-ing; students receive direction from their primary teacher and support from other schoolstaff members. This approach recognizes that behavior change takes time; it also recognizesthat the total school atmosphere needs to change as reinforcements are implemented acrossschool experiences (e.g., Farrell, Meyer, Kung, & Sullivan, 2001; Gottfredson, 2001;USDHHS, 2001).

Several large-scale, school-based violence prevention programs targeting elementaryschool students have documented promising findings of program effectiveness (cf. the Stu-dents for Peace Project, Kelder et al., 1996; Orpinas et al., 2000). For example, the Resolv-ing Conflicts Creatively Program (RCCP) (Aber, Jones, Brown, Chaudry, & Samples,1998) found that the program did not reverse negative or positive behavior patterns but sig-nificantly slowed the trajectories for increasing aggressiveness and decreasing social com-petence, particularly for students who were exposed to most of the programmatic compo-nents. Similarly, findings from the Fast Track prevention trial by the Conduct ProblemsPrevention Research Group (CPPRG) indicated that the program has decreased rates ofconduct problems in children identified as being at high risk for behavior problems in kin-dergarten (baseline; 27% children with conduct problems in the intervention group vs. 37%in the control group; CPPRG, 2002). In another effort evaluating the effects of Peacemak-ers, Shapiro, Burgoon, Welker, and Clough (2002) found decreases in self-reported andteacher-reported aggressive behaviors as well as decreases in the number of disciplinaryincidents and suspensions following program implementation. The study also indicatedstronger program effects for boys than for girls and for younger children than older ones.Finally, teacher-reported data showed more consistent and stronger program effects thanstudent data, although self-reported student data corroborated findings based on teacherreports.

Additional programs require some discussion. Again focusing on a high-risk sampleof children, the Metropolitan Area Child Study (MACS) (Eron, Huesmann, Spindler,

Vazsonyi et al. / EVALUATION OF A VIOLENCE PREVENTION PROGRAM 187

Guerra, & Henry, 2002; Guerra, Eron, Huesmann, Tolan, & Van Acker, 1997) providedevidence of program effectiveness. Findings indicated that the program was most beneficialwhen it was administered during the early school years and where it was supported by a 2-year follow-up intervention. They also indicated that the intervention was equally effectivefor boys and girls; in fact, although median levels of aggression increased over time in inter-vention and control conditions, a significant number of children moved from clinical tononclinical status for externalizing behavior problems following the intervention.

The Responding in Peaceful and Positive Ways (RIPP) (Farrell & Meyer, 1997;Farrell, Meyer, & White, 2001) program and evaluation study provided evidence of areduction in violent behaviors and less in-school suspensions following the intervention.The reduction in violent behaviors was most evident in students who had high levels of vio-lent behaviors at pretest, which indicated a differential programmatic effect.

Finally, two studies evaluating Linking the Interests of Families and Teachers (LIFT)(Reid, Eddy, Fetrow, & Stoolmiller, 1999; Stoolmiller, Eddy, & Reid, 2000) found supportfor reducing young children’s physical playground aggression and increasing teacher rat-ings of peer-preferred behaviors. Differential effectiveness for reducing children’s aggres-sion were found over time, namely, that children with the highest levels of aggression atpretest showed more changes than children with lower pretest scores. To assess this differ-ential effectiveness, Stoolmiller et al. (2000) measured the effect sizes at four levels ofaggression and found medium to high effect sizes for children with the highest levels ofaggression at pretest.

These findings are encouraging and are consistent with Durlak and Wells’ (1997)meta-analysis that showed that programs targeting reducing negative behaviors and pro-moting social competency show promise. These programs addressed a specific recommen-dation by Durlak and Wells (1997) and Weissberg and Bell (1997) to evaluate program suc-cess for at-risk populations. In particular, these studies started to address the differentialeffectiveness of programs, how well the programs work for children at risk for future vio-lence, rather than addressing main effects between intervention and control groups. Indeed,Stoolmiller et al. (2000) identified differential effectiveness as a key issue for universal pro-grams. However, researchers disagree on how to best determine risk for future delinquency.Because of low base rates, only a small number of children become classified as officiallydelinquent, approximately 5% to 6% of boys (Vazsonyi et al., 1999). RIPP, LIFT, andPeaceBuilders have addressed differential effectiveness utilizing regression methodol-ogy (Farrell, Meyer, & White, 2001; Flannery et al., 2003; Stoolmiller et al., 2000). In par-ticular, previous research on PeaceBuilders found differential effectiveness for teacher-reported aggression, self-reported aggression, and prosocial behavior.

Based on these studies, the purpose of this article is to test the differential effective-ness hypothesis, namely, that programs have greater effects on children with high rates ofproblem behaviors as opposed to children with very low rates. Rather than utilizing aregression procedure, an alternative method for assessing differential effectiveness is toassign children to risk levels. In addition, instead of classifying children at risk by officialdelinquency status, risk determination should include more children by identifying vari-ables that predict delinquency that are not confounded with measures of delinquency(Loeber & Dishion, 1983). LeBlanc (1998) advocated using a “multiple-gating” proceduredeveloped first by Loeber and Dishion (1983) that uses several assessments or predictors asscreening gates. The first step is to apply the first predictor to the full sample, temporarilyclassifying children into risk and nonrisk samples by the primary factor. Subsequently, inthe second step, children are maintained or dropped from the risk sample based on the sec-


ond predictor. The result is that a larger number of children are classified at risk for a partic-ular outcome, which may be beneficial for determining how effective programs are forchildren most at risk for future problems.

Thus, the current investigation examined the differential effectiveness of Peace-Builders on children identified as low, medium, or high risk for future problems. Childrenwere classified by the multiple-gating procedure into three risk groups (low, medium, andhigh risk) based on teachers’ assessments of aggression and social competence.

PeaceBuilders Violence Prevention Program

PeaceBuilders is a schoolwide, universal violence prevention program that is theoret-ically based (Embry, Flannery, Vazsonyi, Powell, & Atha, 1996). The program attempts tochange antecedents that trigger aggressive behavior, reward prosocial behavior, and pro-vide strategies to avoid reinforcing negative behavior. PeaceBuilders is organized aroundfive main principles: (a) PeaceBuilders praise people, (b) PeaceBuilders avoid put-downs,(c) PeaceBuilders seek wise people, (d) PeaceBuilders notice hurts they have caused, and(e) PeaceBuilders right wrongs. The intervention structure uses several behavior techniquesto promote change: symbolic and live models, role-plays and rehearsals, and group andindividual rewards.

The PeaceBuilders program was implemented in the school setting by teachers, prin-cipals, and other support staff. members Teachers use a variety of materials to help teachand encourage students to be PeaceBuilders: “I Help Build Peace” story/workbooks, medi-ation essays, Praise Boards (written records of positive events), games (The Peace ScoutGame where anonymous scouts send secret notes), home notes, posters made by children,PeaceCards and secret notes. Teachers received an hour-long preintervention orientation, 3to 4 hr of training workshops, and 2 hr of site coaching per week that occurred during thefirst 8 to 12 weeks of program implementation. Additional help sessions were offered whenschools had specific questions regarding PeaceBuilders (Embry et al., 1996).

The study design included nine project schools with children in kindergarten throughfifth grades. One Grade K-2 school and one Grade 3-5 school were combined to form a sin-gle K-5 unit. These eight school units were then grouped into four matched pairs. Within thepairs, schools were randomly assigned as intervention (Wave 1) or wait-list control (Wave2) schools. Baseline data (Time 1) for all schools were conducted in the fall of 1994. Wave 1schools received the intervention following baseline, in the fall of 1994, and Wave 2schools received the intervention in the fall of 1995. Data collection occurred every fall andspring for 2 years; Time 2 data collection occurred spring 1995, Time 3 data collectionoccurred fall 1995, and Time 4 data collection occurred spring 1996.

A preliminary analysis of the effectiveness of PeaceBuilders compared children in aninitial treatment versus delayed treatment condition. Children in both conditions had similarbaseline levels of aggression and social competence. Hierarchical linear modeling deter-mined that children who received the intervention in the initial treatment condition showedsignificant increases in teacher-rated social competence and child self-reports of peacebuilding as compared to the delayed treatment condition, over a 2-year period. Similarly,after 12 months, children in the delayed treatment condition showed significantly higherrates of aggression than children in the continuous treatment condition (Flannery et al.,2003). Whereas preliminary longitudinal analyses show some changes in children’s behav-


iors because of PeaceBuilders interventions, these results need to be more carefully exam-ined, especially for children who vary in degree of risk for future violence at baseline.

Risk status established through a multiple-gating procedure has been used to examinehow level of risk subsequently predicts delinquency or externalizing behaviors in previousstudies (Lipsey & Derzon, 1998; Lochman & The Conduct Problems Prevention ResearchGroup, 1995; Loeber & Dishion, 1983; Patterson, Capaldi, & Bank, 1991); however, it hasnot been utilized in evaluation research. One exception is the MACS project, where Guerraet al. (1997) reported splitting children into two risk categories. The most recent reportsof MACS program effects (Eron et al., 2002) focused on program effects in the high-riskgroup of children. Across violence prevention programs, program effectiveness has notbeen evaluated at multiple levels of risk. Therefore, the current investigation cannot makespecific predictions about program effectiveness by level of risk based on previous work; atthe same time, we did expect the greatest amount of change in high-risk youth and thesmallest amount of change in low-risk youth.

More specifically, we expected differential program effects on behavioral outcomemeasures by levels of risk. For example, we expected that children identified as high riskwould show the greatest program effects in teacher-reported aggression and social compe-tence, and self-reported aggression and prosocial behavior. We also expected more modestprogram effects for children identified at medium risk for future violence and delinquency,and we expected very few program effects for children identified at low risk for future vio-lence and delinquency. Finally, we expected that we would find similar programmaticeffects by levels of risk for boys and girls.

Method

Sample

The sample for the current study is based on the PeaceBuilders violence preventionevaluation project conducted in the Tucson metropolitan area (Embry et al., 1996; Flanneryet al., 2003; Vazsonyi et al., 1999). The targeted region had experienced an increase in vio-lent offenses from 1990 to 1993—increases in juvenile arrests for violent crimes and homi-cides, vandalism, and weapons violations. Juvenile arrests for total, property, and violentcrimes continued to increase and peaked in 1995. Since 1995, juvenile arrests have beendecreasing. However, property crimes have decreased at a higher rate than violent crimes,which are still high at similar levels as reported in 1990-1991 (Geospatial and StatisticalData Center [Geostat], 2003). In addition to community-wide efforts to increase social andcognitive competencies related to preventing violence, a comprehensive program, Peace-Builders, was implemented within two city school districts.

Two school districts were chosen based on police crime maps; these maps identifiedareas with high levels of violent crimes and high neighborhood stress (e.g., domestic vio-lence, transition and mobility, poverty levels). Nine schools (one K-2 and 3-5 were com-bined to form one school unit) were invited for participation based on these data (Embryet al., 1996). Schools were matched into four pairs based on geographic proximity, studentethnicity, percentage of students eligible for free or reduced lunch, and percentage of stu-dents with English as their second language (Embry et al., 1996; see Table 1). It is importantto note that some of the matched school pairs differed on key variables (e.g., student ethnic-


ity). Schools were then randomly assigned to the initial intervention condition or the wait-list control condition.

The sample included approximately 4,600 children from kindergarten through fifthgrade. Of the total sample, 50% were Hispanic, 29% White, 15% Native American, 5%African American, and 1% Asian/Pacific Islander. Children were roughly evenly dividedby sex between the intervention and wait-list control conditions. Figure 1 graphically pres-ents program intervention and data collection periods. One half of the schools received thePeaceBuilders intervention in the fall of the first year (Wave 1), while the other one halfreceived the intervention during the following fall (Wave 2). For the current investigation,we focused on children in third through fifth grades who had teacher ratings of aggressionand social competence and self-reports of aggression and prosocial behaviors at the base-line (Time 1). Data from children in both treatment conditions (Wave 1 and 2) were aggre-gated to permit a comparison of children’s pretest scores and posttest scores. Data at pretestwere Time 1 data for Wave 1 children and Time 3 data for Wave 2 children. Data at posttestwere Time 2 data for Wave 1 children and Time 4 data for Wave 2 children. ANOVAs com-paring pretest scores by Wave showed only one significant difference for female prosocialbehavior; Wave 2 girls reported slightly higher prosocial behaviors at pretest than Wave 1girls (F = 5.96, p < .05, d = .19).

The total number of students enrolled in project schools in kindergarten throughGrade 5 at initial data collection was N = 4,679. Some students were excluded from the sam-ple because of incomplete data. Teacher data were available on children in kindergartenthrough Grade 5. Complete teacher data (K-5) were collected from n = 2,380 children (Mage = 8.5 years), a response rate of 50.8% (see Table 2). Because of cognitive and languageability, child self-report data were only available for Grades 3 through 5. Complete childself-report data (3-5) were obtained for n = 1,170 children (M age = 9.8 years), a responserate of 52.2% (see Table 2). The low response rates (as compared to Flannery et al., 2003)are due largely to the construction of this sample. First, to classify the students into risk cat-egories by teacher reports, the sample was limited to students with baseline teacher-reported data. Second, to compare pretest and posttest data, Wave 1 children had to have


TABLE 1School Level Demographic Characteristics:

Percentage of Ethnicity and Socioeconomic Variables

School School School School School School School School1A 1B 2A 2B 3A 3B 4A 4B

(n = 704) (n = 551) (n = 817) (n = 377) (n = 550) (n = 573) (n = 327) (n = 780)

African American 9.7 14.6 0.2 5.2 2.8 0.8 2.8 3.5Asian/Pacific Islander 3.7 2.5 0.3 1.4 0.6 0.3 1.3 1.0Hispanic 22.7 18.5 33.5 62.2 74.4 91.8 65.9 58.5Native American 0.6 1.9 54.6 1.7 13.4 2.5 2.1 1.0White 63.3 62.5 11.3 29.4 8.8 4.8 28.0 36.0Free lunch 55 55 94 60 60 94 89 73English as Second

Language 5 8 6 29 29 68 28 21

NOTE: Schools were matched in pairs based on geographic proximity, student ethnicity, percentage of students eli-gible for free or reduced lunch, and percentage of students with English as their second language. Schools were ran-domly assigned to the initial intervention condition (A schools) or the wait-list control condition (B schools).

Time 1 and Time 2 data, whereas Wave 2 children had to have Time 3 and Time 4 data.Thus, students without child and/or teacher data at Time 1 and Time 3, for example, weredropped from the sample. This selection process did not appear to vary by sex or race.Approximately 50% of boys and girls were dropped, and race percentages ranged from20% to 25% for all groups except Native Americans (41%). In addition to sample construc-tion issues, subject attrition rates were related to relatively high residential mobility withinschool districts.

Procedures

Data were collected by trained project staff members from teachers in Grades Kthrough 5, and children in Grades 3 through 5. During regular school hours, children com-pleted in-class surveys administered by project staff members who read all questions aloud.The survey took approximately 1 hr to complete, and students received small incentives fortheir participation, such as stickers or pencils. Teachers completed surveys for each child intheir classroom. They received data collection packets at the time of the student survey datacollection. Each teacher received $20 for participation. In addition, the schools were eligi-ble for schoolwide incentives based on the number of teacher surveys returned ($300 to$500). During the initial phase of the project, data were collected at four points in time (twofall and two spring, in consecutive school years) for schools assigned to two interventionconditions. Schools in Wave 1 started the intervention immediately following baseline datacollection. Schools in Wave 2 began the intervention about 1 year later following Time 3data collection.


Figure 1. Overview of Project Design, Data Collection, and Intervention Schedule

Measures

Social competence (teacher report). The 19-item short-form version of the Walker-McConnell Scale of Social Competence (Walker & McConnell, 1995) measured socialskills and school adjustment as rated by teachers. The instrument has been used in long-term follow-up studies and has predictive value, particularly for children with seriousbehavior problems (Fifeld, 1987; Hops, 1987). The scale includes three subscales: SchoolAdjustment (e.g., “student attends to assigned tasks” and “produces work of acceptablequality given his or her skills”); Peer Preferred Behavior (e.g., “invites peers to play” and“shares laughter with peers”); and Teacher Preferred Behavior (e.g., “can accept not gettinghis or her way” and “compromises with peers when a situation calls for it”). Teachers ratedeach item on a 5-point scale from 1 = never to 5 = frequently (α = .95). The three subscaleswere summed to produce an overall Social Competence score (Flannery et al., 2003;Vazsonyi et al., 1999).


TABLE 2Sample Description and Response Rates

Low Medium HighNot Risk Risk Risk Total

Classified n n n n N %

Teacher report (K-5)with baseline dataa (1,132) 1,147 1,099 1,301 3,547 4,679 75.8

Teacher report (K-5) withpretest/posttest scoresb 840 746 794 2,380 4,679 50.8

Males 455 341 376 1,172 2,380 49.2Females 385 405 418 1,208 2,380 50.8

Child self-report (3-5) withpretest/posttest scoresb 443 332 395 1,170 2,243 52.2

Males 221 146 197 564 1,170 48.2Females 222 186 198 606 1,170 51.8

Low Medium HighNot Risk Risk Risk

Classified % % % n Nd %

Ethnicityc

African American 20.1 21.8 18.4 39.7 174 3,086 5.6Asian/Pacific Islander 20.4 40.8 22.4 16.3 49 3,086 1.6Hispanic 25.6 25.8 24.7 23.8 1,540 3,086 49.9Native American 40.8 18.7 18.1 22.3 497 3,086 16.1White 24.0 28.2 21.7 26.2 826 3,086 26.8

NOTE: a. Baseline data for all children were collected at Time 1.b. Data at pretest were Time 1 data for Wave 1 children and Time 3 data for Wave 2 children. Data at posttest wereTime 2 data for Wave 1 children and Time 4 data for Wave 2 children.c. Ethnicity statistics reported here were obtained from archival school records and are reported for students withbaseline data.d. Ethnicity was available for 3,086 children out of the 4,679 children eligible with baseline data, 34.0% of the sam-ple was missing data on ethnicity.

Aggressive behavior (TRF). Physical and nonphysical aggressive behavior was mea-sured by the 25-item Achenbach’s Child Behavior Checklist Teacher Report Form(Achenbach, 1991; Flannery et al., 2003; Vazsonyi et al., 1999). Teachers were asked torecall children’s behavior over the past 2 months; examples include “The child argues a lot,”“The child gets in many fights,” and “The child threatens people.” Responses were given ona 3-point Likert-type scale, 0 = not true, 1 = somewhat true, or 2 = very true (α = .95).

Prosocial behavior (child report). This 16-item scale was developed by the researchteam to measure how much children engaged in prosocial acts over the past 2 weeks(Flannery et al., 2003; Vazsonyi et al., 1999). Children responded to questions such as “Idid things to help other kids,” “I smiled at others,” and “I apologized to a grown-up atschool.” Responses were given on a 3-point scale, 1 = no, 2 = a little, and 3 = a lot (α = .92).

Aggressive behavior (YSR). This scale consisted of nine items from the Delinquencyand Aggression subscales of the Child Behavior Checklist-Youth Self Report (Achenbach,1991; Flannery et al., 2003; Vazsonyi et al., 1999). Questions asked about physical andnonphysical aggression over the past 2 weeks, for example, “I teased other kids at school,”“I hit someone,” and “I tried to get other students to fight.” Responses were given on a 3-point Likert-type scale, 0 = no, 1 = a little, to 2 = a lot (α = .95).

Plan of Analysis

Initial descriptive statistics on teacher reports and child self-reports were computedfor all children with baseline (Time 1) data. These data were used to classify children intothree risk groups: low, medium, and high risk.

Analyses for the current study were computed using general linear modeling (GLM).GLM covers a variety of linear models of analyses of variance and covariance, regression,and repeated measures models (Howell, 1992); it also adjusts for unequal cell sizes and pro-vides estimated marginal means (predicted estimates of the population marginal meanbased on regression; Searle, Speed, & Milliken, 1980). GLM repeated-measures proce-dures account for variation in the pretest and posttest scores by computing a pooled valuefor multivariate tests and subsequently determines change over time using estimatedmarginal means (SPSS, 1999).

To maximize sample size, GLM analyses were conducted separately by sex and byteacher and self-report data. Risk status was entered in the model as a between-subjectsvariable with three levels of risk. Age and race were also included in the model ascovariates. Age was a continuous covariate, and race was a categorical covariate, namely,White versus non-White. Because of the inclusion of covariates in the model, GLM analy-ses were conducted in two steps (Winer, 1971). The first step ran the model with thecovariates and reported the between-subjects portion of the model. The second step ran themodel without the covariates and reported the within-subjects portion of the model. Subse-quently, pairwise comparisons were conducted based on the estimated marginal means.Because differences over time were hypothesized a priori, significant pairwise comparisonswere reported regardless of the significance of the omnibus F statistic (Girden, 1992;Tabachnick & Fidell, 1989).

One assumption of ANCOVA is that the regression coefficients are equal (Howell,1992). A significant covariate would violate that assumption, usually invalidating the use of


ANCOVA for modeling data (Howell, 1992; G. Hudson, personal communication, March8, 2002; SPSS, 2002). However, in the current study, the covariates included in the modelwere not treatment effects but naturally occurring variations in the population. Therefore,we were interested in the percentage of variance in the model explained by each of thecovariates (Howell, 1992). For GLM repeated measures, this was done in two procedures:first by analyzing the slope and variance of the covariates for the pooled dependent variable(consistent with GLM multivariate tests) and subsequently by examining whether thecovariate influenced change in the dependent variables over the course of the program byusing change scores (Howell, 1992; G. Hudson, personal communication, March 8, 2002;SPSS, 2002; Tabachnick & Fidell, 1989). The second step of the analyses determined theeffect of the program over time for each of the three risk groups after controlling for theeffects of covariates. Pairwise comparisons of the estimated marginal means determinedprogrammatic effects on children’s behaviors at different levels of risk.

Results

At-Risk Status

Children’s at-risk status was determined by teacher-reported aggression andsocial competence scores collected at baseline. Low, medium, and high risk was defined bya multiple-gating procedure that utilized a median split of the two risk variables separatelyby sex (see Table 3).

Boys and girls were classified as high risk if they had scores above the median inaggression and below the median for social competence; in other words, they exhibited highnegative behaviors and few positive ones. Individuals were classified as low risk if they hadscores below the median score for aggression; these children also reported high social com-petence scores. The medium-risk group was characterized by scores above the median inaggression and social competence or by scores below the median in aggression and socialcompetence. In other words, these children exhibited a mixture of positive and negativebehaviors (see Figure 2 for the multiple-gating procedure; see Table 3 for the number ofchildren in each of the risk groups by sex; there were at least 125 children in each riskcategory for boys and girls).


TABLE 3Descriptive Statistics of Teacher Reports by Sex (N = 3,554)

Males Females(n = 1,765) (n = 1,779)

M SD Median M SD Median

Social competence 3.49 .82 3.53 3.86 .76 3.89Aggression (TRF) 1.45 .52 1.24 1.23 .38 1.04

NOTE: Social competence and aggression (TRF) are teacher-reports. TRF = Teacher Report Form of theAchenbach Child Behavior Checklist (Achenbach, 1991).

Winer Model Step 1:

Inclusion of Covariates and Between-Subjects Variable

The first step of the Winer model includes both covariates, age and race, and the inde-pendent variable, risk status. The results of the analyses are presented in Table 4. The firstthree columns report the multivariate F statistics for each dependent variable. The next fourcolumns report the slope and percentage of variance explained by each of the covariates inthe pooled dependent variable. The last four columns report the slope and percentage ofvariance explained by the covariates on the amount of change over time in each dependentvariable.

Significant effects for age were found for female social competence, male teacher-rated aggression, male and female prosocial behaviors, and male self-reported aggression.Significant effects for race were found only for female prosocial behavior as well as maleand female self-reported aggression. Significant effects for risk as a covariate were foundfor male and female social competence, teacher-reported aggression, and self-reportedaggression. No significant effects for risk were found for prosocial behavior.

When analyzing the effects of the covariates on the pooled dependent variables, ageaccounted for a large percentage of the variance in prosocial behavior (8.9% girls, 12.8%boys). For both of these variables the slopes were negative; as age increased, prosocialbehavior decreased. Age also accounted for 4.4% of the variance in male teacher-ratedaggression and 3.2% of the variance in female social competence. Race accounted for verylittle variance in the dependent variables; the highest percentage of variance attributed torace was for male self-reported aggression (1.4%). Even though age and race accounted forsome proportion of the variance in the pooled dependent variables, additional analysesneeded to determine the percentage of variance these covariates explained in pre/postchange scores. In these scores, age accounted for a small proportion of the variance in male(2.5%) and female (1.5%) changes of social competence and male self-reported aggression(1.5%). Race accounted for very little variability in change scores, namely, 1.7% of the


Figure 2. Multiple-Gating Procedure

TA

BL

E 4

Win

er M

odel

(St

ep 1

): B

etw

een-

Subj

ects

FV

alue

s, a

nd C

ovar

iate

Ana

lyse

s Sl

ope,

and

R2

Val

ues

for

Gen

eral

Lin

ear

Mod

elin

g (G

LM

) m

odel

Poo

led

DV

Cha

nge

Scor

es

Slop

e%

Slop

e%

Slop

e%

Slop

e%

Age

Rac

eR

isk

of A

geV

aria

nce

of R

ace

Var

ianc

eof

Age

Var

ianc

eof

Rac

eV

aria

nce

Fd

Fd

Fdβ

Age

βR

ace

βA

geβ

Rac

e

Mal

es Soci

al c

ompe

tenc

ea2.

451.

3538

1.40

*–.

001

1.82

–.00

40.

07–.

063

2.54

.014

0.00

Agg

ress

ion

(TR

F)a,

c6.

21*

1.09

342.

29*

–.00

14.

43–.

002

0.04

.001

0.11

.003

1.69

Pros

ocia

l beh

avio

rb10

8.11

*0.

241.

33–.

1912

.78

–.00

20.

04–.

023

0.19

.037

0.11

Agg

ress

ion

(YSR

)b,c

6.95

*7.

64*

18.7

2*–.

005

1.36

–.10

61.

42.0

561.

53.0

470.

24Fe

mal

esSo

cial

com

pete

ncea

6.39

*1.

6731

8.00

*–.

002

3.23

–.00

40.

05–.

044

1.45

.001

0.00

Agg

ress

ion

(TR

F)a,

c0.

290.

8220

8.51

*.0

000.

00.0

020.

04.0

160.

78.0

470.

54Pr

osoc

ial b

ehav

iorb

65.6

0*4.

40*

0.95

–.15

28.

91–.

008

0.57

–.00

40.

00–.

007

0.00

Agg

ress

ion

(YSR

)b,c

2.04

5.89

*11

.83*

–.00

20.

55–.

005

0.83

–.00

30.

00.0

370.

32

a. T

each

er r

epor

ts o

f so

cial

com

pete

nce

and

aggr

essi

on a

re li

sted

fir

st,n

rang

es f

rom

323

to 4

11.

b. C

hild

sel

f-re

port

s of

pro

soci

al b

ehav

ior

and

aggr

essi

on,n

rang

es f

rom

125

to 1

94.

c. T

RF

= T

each

er R

epor

t For

m o

f A

chen

bach

’s C

hild

Beh

avio

r C

heck

list;

YSR

= Y

outh

Sel

f-R

epor

t For

m o

f A

chen

bach

’s C

hild

Beh

avio

r C

heck

list (

Ach

enba

ch, 1

991)

.d.

Mul

tivar

iate

Fst

atis

tic is

sig

nifi

cant

at *

p<

.05.

197

variance in male teacher-reported aggression changes. The second step of the analysesdemonstrated significant changes over time in each of the three levels of risk. Therefore, aseries of analyses was completed in the second step of the Winer model; results are reportedin Table 5.

Winer Model Step 2:

Changes Over Time by Level of Risk

High risk. Significant changes over time for children classified as high risk werefound for male and female teacher-reported social competence and aggression scores; nosignificant changes over time were found for self-reported prosocial behavior or aggres-sion. These significant differences were in the hypothesized direction for social competenceand aggression. Social competence scores for high-risk children increased significantly forboys (d = .36) and girls (d = .44). Teacher-rated aggression scores decreased significantlyfor boys (d = –.13) and girls (d = –.24).

Medium risk. For children classified at medium-risk status, significant changes overtime were found for male and female teacher-reported social competence. No significantchanges over time were found for teacher-rated aggression or self-reported prosocial be-havior and aggression. As hypothesized, medium-risk teacher-rated social competencescores increased for boys (d = .34) and girls (d = .31).

Low risk. For children classified at low risk, significant changes were found for maleand female teacher-rated aggression. No significant changes were found for teacher-ratedsocial competence or self-reported prosocial behavior and aggression. Contrary to hypothe-ses, teacher-reported aggression increased for boys (d = .31) and girls (d = .15).

Discussion

Based on criteria established by the surgeon general and the Centers for Disease Con-trol and Prevention, the PeaceBuilders (Embry et al., 1996) violence prevention programtargets decreasing risk factors and increasing protective factors in a universal school-basedprogram utilizing effective strategies for behavior change (USDHHS, 2001). The currentinvestigation examined whether the PeaceBuilders violence prevention program had a dif-ferential effect on children’s behavioral outcomes by levels of risk (low, medium, andhigh); more specifically, we were interested in four outcomes, namely, teacher-reportedaggression and social competence and self-reported aggression and prosocial behavior. Inaddition, we were interested in determining the effects of sex, age, and race on programeffectiveness. Although researchers have classified children at risk for future problems inprevious work (Lochman & The Conduct Problems Prevention Research Group, 1995;Patterson et al., 1991), most of these comparisons have considered children’s behaviordifferences at one point in time and not in the context of an intervention.

Findings indicated that the effects of PeaceBuilders were not universal across riskcategories. Significant behavior changes were found for children classified at high risk forfuture violence at baseline. Consistent with expectations and previous research on differen-tial effectiveness (Farrell, Meyer, & White, 2001; Flannery et al., 2003; Stoolmiller et al.,


TABLE 5Winer Model (Step 2):

Pretest-Posttest Scores by Risk, Within-Subjects F Value,and Significant Pairwise Comparisons

Low Risk Medium Risk High Risk

Pretest Posttest Pretest Posttest Pretest Posttest

Time Sig.Risk Pairwise

M SD M SD M SD M SD M SD M SD Fd Comparisonsd

MalesSocial competencea 4.13 .56 4.18 .67 3.55 .63 3.78 .74 2.93 .57 3.16 .71 10.96* b,cAggression (TRF)a,c 1.08 .18 1.15 .27 1.34 .39 1.36 .44 1.82 .51 1.75 .55 12.58* a,cProsocial behaviorb 1.86 .54 1.82 .55 1.84 .55 1.72 .57 1.82 .59 1.80 .56 1.65Aggression (YSR)b,c 1.26 .35 1.32 .39 1.35 .43 1.39 .46 1.53 .53 1.52 .51 1.55

FemalesSocial competencea 4.44 .49 4.44 .60 3.93 .63 4.13 .65 3.29 .59 3.58 .72 25.54* b,cAggression (TRF)a,c 1.02 .07 1.06 .17 1.12 .22 1.14 .24 1.48 .48 1.37 .43 28.14* a,cProsocial behaviorb 2.17 .51 2.07 .53 2.10 .49 2.05 .50 2.13 .47 2.02 .55 0.58Aggression (YSR)b,c 1.10 .24 1.12 .21 1.15 .29 1.17 .23 1.20 .32 1.23 .34 0.34

NOTE: a-low risk; b-medium risk; c-high risk.a. Teacher reports of social competence and aggression are listed first, n ranges from 341 to 455 for teacher-reports (Grades K-5) and from 146 to 222 for child self-reports (Grades 3-5).b. Child self-reports of prosocial behavior and aggression, n ranges from 125 to 194.c. TRF = Teacher Report Form of Achenbach’s Child Behavior Checklist; YSR = Youth Self-Report Form of Achenbach’s Child Behavior Checklist (Achenbach, 1991).d. F statistic and pairwise comparisons are significant at *p < .05.

199

2000), high-risk children showed the most significant changes over time; teacher-reportedaggression decreased, whereas teacher-rated social competence increased. Moreover, theseeffects were found for boys and girls. However, no positive program effects were found foreither of the self-reported variables, prosocial behavior and aggression. For medium-riskchildren, only teacher-rated social competence increased, whereas no effects were foundfor teacher- or self-reported aggression. Findings for children classified at low risk showedunexpected changes over time, namely, increases in teacher-reported aggression. At thesame time, these children maintained their relatively high levels of social competence. Inaddition, though aggression increased, the levels of aggression still remained substantiallybelow medium- and high-risk groups, scores that would continue to result in low-riskclassification.

In conclusion, the findings from the current evaluation effort are encouraging.Together with other recent efforts (e.g., CPPRG, 2002; Eron et al., 2002; Farrell, Meyer, &White, 2001; Shapiro et al., 2002), large-scale universal violence prevention programs suchas PeaceBuilders show promise for changing children’s behaviors, in particular for chang-ing risk and protective factors for future violence (cf., CPPRG, 2002). Our findings add to agrowing number of investigations that have provided evidence on differential program effi-cacy for high-risk children and youth (e.g., CPPRG, 2002; Eron et al., 2002; Farrell, Meyer,Kung, & Sullivan, 2001); they suggest that students classified at high risk for future behav-ior problems significantly decreased on measures of aggression and increased on measuresof social competence. In addition to differential program effectiveness being examined viaregression methodology as shown by previous work, the current study demonstrated differ-ential effectiveness through the multiple-gating procedure that utilized two variables forrisk classification, presence of negative and lack of positive behaviors at baseline. Futureevaluation research should continue to evaluate the effectiveness of violence preventionprograms in children who are most at risk.

Limitations of the Current Study

A number of limitations require some discussion. One important consideration is theinsider versus outsider perspective. In the current study, teachers reported more significantbehavior changes than did children’s self-reports similar to Shapiro et al. (2002). However,in the current study, no significant results were found by risk level in the differential effec-tiveness of PeaceBuilders for aggression or prosocial behavior. Findings by Stanger andLewis (1993) based on comparisons of behavior ratings between teacher, child, and parentreports on the Child Behavior Checklist have some important implications for the currentstudy. They found that children generally report more problems than do teachers; they sug-gested that one possible reason for this is that teachers rate behaviors only during schoolhours, whereas children rate their behaviors across contexts. In addition, they suggestedthat teachers attend to externalizing behaviors, such as aggression, differently than do chil-dren, because these behaviors cause management problems and may be more salient for theteachers than children. It may be that teachers attended to children’s changed behaviorwithin the school environment, whereas children attended to their behaviors in school andin other contexts outside of schools with siblings or peers. Thus, the differential effective-ness of PeaceBuilders would be limited in generalizability to the school environment.

Another issue requiring discussion is the one of quasi-experimental design. The cur-rent study did not contain a true control condition. Due to Institutional Review Board (IRB)


requirements and practical considerations, all children received the intervention at somepoint, thus creating a wait-list control condition, where one half of the students received theintervention 1 year later. Children in this latter condition did complete data collectionsprior to the intervention and may have been aware of PeaceBuilders prior to actual inter-vention at their school. With quasi-experimental designs, the study is compromisedbecause program effects cannot be clearly determined; however, quasi-experimentaldesign emphasizes the ecological context and optimizes generalizability because programsare evaluated as they are implemented (Henrich, Brown, & Aber, 1999). Third, withoutcomparing change over time among students classified in risk categories in intervention andcontrol conditions, we could not determine whether changes in reported behaviors were dueto regression to the mean. Therefore, one plausible explanation for our findings couldinclude regression to the mean.

Fourth, the current study was limited by participant attrition. The PeaceBuilders eval-uation study was conducted in high-risk neighborhoods that experienced very high residen-tial mobility that limited the number of students with longitudinal data (for a discussion, seeFlannery et al., 2003). A final limitation is that results may be attributable to teacher bias.Additional analyses conducted by Belliston (2000) and Flannery et al. (2003) have docu-mented varied fidelity of implementation. However, analyses have indicated that fidelity ofimplementation did not affect the differential effectiveness of the program by risk category.

Conclusions/Implications

Universal, school-based programs such as PeaceBuilders show promise for reducingaggression and increasing social competence (Flannery et al., 2003; Shapiro et al., 2002).Such relatively low-cost programs that attempt to blanket the school population haveimportant policy implications in that spending only a few hundred dollars per child duringelementary school might save the criminal justice system millions later on, when individu-als enter it during adolescence and adulthood (Cohen, 1998). Specifically, Cohen (1998)estimated the costs of a criminal on society based on calculations such as mean number ofoffenses, victim cost of crime, cost of investigation and adjudication, incarceration, fore-gone earnings, and opportunity cost of time. He noted that the benefits of programs thatreduce crime might exceed the cost estimates computed, in terms of affecting large socialproblems, reducing fear of crime, reducing private security measures, or changing lifestyledue to decreased risk of victimization (e.g., walking vs. taking a cab). Cohen estimated that,for a juvenile career, the present lifetime costs range between U.S.$80,000 and $325,000;for an adult offender, $1.2 million, total costs ranging from $1.3 to $1.5 million for juvenileand adult career offenses. When combining comorbid problems of criminality, drug use,and high school dropout, costs to society range from $1.7 to $2.3 million (Cohen, 1998). Incontrast, the entire PeaceBuilders project budget for project administration, project devel-opment, project implementation, training, follow-ups, evaluation design, and data collec-tion and analysis cost less than $200 per child over the project’s 3-year period. Thus, thecost of the program is minimal compared to potential costs due to a life of crime and vio-lence. Universal programs such as PeaceBuilders seem effective and cost-efficient becausethey can reach an entire population of children, not only children at risk. By reaching agreater number of children, such programs change the school climate, reduce the number ofclassroom disruptions, and ultimately reduce the total number of children at risk for futureviolence.


REFERENCES

Aber, J. L., Jones, S. M., Brown, J. M., Chaudry, N., & Samples, F. (1998). Resolving conflict cre-atively: Evaluating the developmental effects of a school-based violence prevention programin a neighborhood and classroom context. Development and Psychopathology, 10, 187-213.

Achenbach, T. M. (1991). Manual for the Child Behavior Checklist/4-18 and 1991 Profile.Burlington: University of Vermont Department of Psychiatry.

Andrews, L., & Trawick-Smith, J. (1996). An ecological model for early childhood violence preven-tion. In R. L. Hampton & P. Jenkins (Eds.), Prevention violence in America. Issues in chil-dren’s and families’ lives (pp. 233-261). Thousand Oaks, CA: Sage.

Beland, K. R. (1996). A school-wide approach to violence prevention. In R. L. Hampton, P. Jenkins,& T. P. Gullotta (Eds.), Preventing violence in America (pp. 209-231). Thousand Oaks, CA:Sage.

Belliston, L. M. (2000). The impact of the PeaceBuilders school-based violence prevention programon low, medium, and high-risk children. Unpublished master’s thesis, Auburn University,Alabama.

Bierman, K. L., Miller, C. L., & Stabb, S. D. (1987). Improving the social behavior and peer accep-tance of rejected boys: Effects of social skill training with instructions and prohibitions. Jour-nal of Consulting and Clinical Psychology, 55, 194-200.

Bierman, K. L., & Montminy, H. P. (1993). Developmental issues in social-skills assessment andintervention with children and adolescents. Behavior Modification, 17, 229-254.

Blechman, E. A. (1996). Coping, competence, and aggression prevention: II. Universal school-basedprevention. Applied and Preventive Psychology, 5(1), 19-35.

Catalano, R. F., Arthur, M. W., Hawkins, J. D., Berglund, L., & Olson, J. J. (1998). Comprehensivecommunity- and school-based interventions to prevent antisocial behavior. In R. Loeber &D. P. Farrington (Eds.), Serious and violent juvenile offenders: Risk factors and successfulinterventions (pp. 248-283). Thousand Oaks, CA: Sage.

Centers for Disease Control and Prevention. (1997). Rates of homicide, suicide, and firearm-relateddeath among children—26 industrialized countries. Morbidity and Mortality Weekly Report,46(5), 101-105.

Cohen, M. A. (1998). The monetary value of saving a high-risk youth. Journal of Quantitative Crimi-nology, 14(1), 5-33.

Coie, J. D., & Dodge, K. A. (1998). Aggression and antisocial behavior. In W. Damon (Ed.), Hand-book of child psychology (pp. 780-845). New York: John Wiley.

Coie, J. D., & Koeppl, G. K. (1990). Adapting intervention to the problems of aggressive and disrup-tive rejected children. In S. R. Asher & J. D. Coie (Eds.), Peer rejection in childhood (pp. 309-337). Cambridge, UK: Cambridge University Press.

Conduct Problems Prevention Research Group. (2002). Evaluation of the first 3 years of the FastTrack prevention trial with children at risk for adolescent conduct problems. Journal of Abnor-mal Child Psychology, 30(1), 19-35.

Consortium on the School-Based Promotion of Social Competence. (1994). The school-based promo-tion of social competence: Theory, research, practice, and policy. In R. J. Haggerty, L. R.Sherrod, N. Garmezy, & M. Rutter (Eds.), Stress, risk and resilience in children and adoles-cents (pp. 268-316). Cambridge, UK: Cambridge University Press.

Durlak, J. A., & Wells, A. M. (1997). Primary prevention mental health programs for children andadolescents: A meta-analytic review. American Journal of Community Psychology, 25(2), 115-152.


Embry, D. D., Flannery, D. J., Vazsonyi, A. T., Powell, K. E., & Atha, H. (1996). PeaceBuilders: Atheoretically driven, school-based model for early violence prevention. American Journal ofPreventive Medicine, 12(5), 91-100.

Eron, L., Huesmann, R., Spindler, A., Guerra, N., & Henry, D. (2002). A cognitive-ecologicalapproach to preventing aggression in urban settings: Initial outcomes for high-risk children.Journal of Consulting and Clinical Psychology, 70(1), 179-194.

Farrell, A. D., & Meyer, A. L. (1997). The effectiveness of a school-based curriculum for reducingviolence among urban sixth-grade students. American Journal of Public Health, 87(6), 979-984.

Farrell, A. D., Meyer, A. L., Kung, E. M., & Sullivan, T. N. (2001). Development and evaluation ofschool-based violence prevention programs. Journal of Clinical Child Psychology, 30(1), 207-220.

Farrell, A. D., Meyer, A. L., & White, K. S. (2001). Evaluation of Responding in Peaceful and PositiveWays (RIPP): A school-based prevention program for reducing violence among urban adoles-cents. Journal of Clinical Child Psychology, 30(4), 451-463.

Farrington, D. P. (1987). Early precursors of frequent offending. In J. Q. Wilson & G. C. Loury (Eds.),From children to citizens: Family, schools, and delinquency prevention (pp. 27-50). NewYork: Springer-Verlag.

Fifeld, B. (1987). A concurrent validation study of the SSRS-T and the Walker-McConnell Scale ofSocial Competence and School Adjustment [Technical Report]. Eugene, OR: Eugene SchoolDistrict.

Flannery, D. J., Vazsonyi, A. T., Liau, A. K., Guo, S., Powell, K. E., Atha, H. A., et al. (2003). Initialbehavior outcomes for PeaceBuilders universal school-based violence prevention program.Developmental Psychology, 39(2), 292-308.

Frey, K. S., Hirschstein, M. K., & Guzzo, B. A. (2000). Second step: Preventing aggression by pro-moting social competence. Journal of Emotional and Behavioral Disorders, 8(2), 102-111.

Geospatial and Statistical Data Center. (2003). Uniform crime reports county data: Juvenile arrests.Available at http://fisher.lib.virginia.edu/crime/

Girden, E. R. (1992). ANOVA: Repeated measures. Newbury Park, CA: Sage.Gottfredson, D. C. (2001). Schools and delinquency. New York: Cambridge University Press.Gottfredson, D. C., Gottfredson, G. D., & Skroban, S. (1998). Can prevention work where it is needed

most? Evaluation Review, 22(3), 315-340.Gresham, F., & Elliott, S. (1987). The relationship between adaptive behavior and social skills: Issues

in definition and assessment. Journal of Special Education, 21(1), 167-181.Guerra, N. G., Eron, L. D., Huesmann, L. R., Tolan, P., & Van Acker, R. (1997). A cognitive-

ecological approach to the prevention and mitigation of violence and aggression in inner-cityyouth. In D. P. Fry & K. Bjoerkqvist (Eds.), Cultural variation in conflict resolution: Alterna-tives to violence (pp. 199-213). Mahwah, NJ: Lawrence Erlbaum.

Hämäläinen, M., & Pulkkinen, L. (1995). Aggressive and non-prosocial behavior as precursors ofcriminality. Studies on Crime and Crime Prevention, 4(1), 6-21.

Hawkins, D. F., Herrenkohl, T., Farrington. D. P., Brewer, D., Catalano, R. F., & Harachi, T. W.(1998). A review of predictors of youth violence. In R. Loeber & D. P. Farrington (Eds.), Seri-ous and violent juvenile offenders: Risk factors and successful interventions (pp. 106-146).Thousand Oaks, CA: Sage.

Henrich, C. C., Brown, J. L., & Aber, J. L. (1999). Evaluating the effectiveness of school-based vio-lence prevention: Developmental approaches. Social Policy Report, Society of Research inChild Development, 13(3), 1-17.


Hops, H. (1987). Behavior correlates of positive and negative sociometric status among same-sexchildren. (Available from Oregon Research Institute, 149 W. 12th Avenue, Eugene, OR97401)

Howell, D. C. (1992). Statistical methods for psychology (3rd ed.). Belmont, CA: Duxbury.Kelder, S. H., Orpinas, P., McAlister, A., Frankowski, R., Parcel, G. S., & Friday, J. (1996). The

Students for Peace Project: A comprehensive violence-prevention program for middle schoolstudents. American Journal of Preventive Medicine, 12(5), 22-30.

Kupersmidt, J. B., Coie, J. D., & Dodge, K. A. (1990). The role of poor peer relations in the develop-ment of disorder. In S. R. Asher & J. D. Coie (Eds.), Peer rejection in childhood (pp. 309-337).Cambridge, UK: Cambridge University Press.

LeBlanc, M. (1998). Screening of serious and violent juvenile offenders: Identification, classification,and prediction. In R. Loeber & D. P. Farrington (Eds.), Serious and violent juvenile offenders:Risk factors and successful interventions (pp. 167-193). Thousand Oaks, CA: Sage.

Lipsey, M. W., & Derzon, J. H. (1998). Predictors of violent or serious delinquency in adolescenceand early adulthood: A synthesis of longitudinal research. In R. Loeber & D. P. Farrington(Eds.), Serious and violent juvenile offenders: Risk factors and successful interventions(pp. 248-283). Thousand Oaks, CA: Sage.

Lochman, J. E., & The Conduct Problems Prevention Research Group. (1995). Screening of childbehavior problems for prevention programs at school entry. Journal of Consulting and ClinicalPsychology, 63(4), 549-559.

Loeber, R., & Dishion, T. (1983). Early predictors of male delinquency: A review. Psychological Bul-letin, 94(1), 68-99.

Loeber, R., Farrington, D. P., Stouthamer-Loeber, M., Moffitt, T. E., & Caspi, A. (1998). The devel-opment of male offending: Key findings from the first decade of the Pittsburgh Youth Study.Studies on Crime and Crime Prevention, 7(2), 141-171.

O’Donnell, J., Hawkins, J. D., & Abbott, R. D. (1995). Predicting serious delinquency and substanceuse among aggressive boys. Journal of Consulting and Clinical Psychology, 63(4), 529-537.

Ollendick, T. H., Weist, M. D., Borden, C., & Greene, R. W. (1992). Sociometric status and academic,behavioral, and psychological adjustment: A five-year longitudinal study. Journal of Consult-ing and Clinical Psychology, 60(1), 80-87.

Orpinas, P., Kelder, S., Frankowski, R., Murray, N., Zhang, Q., & McAlister, A. (2000). Outcomeevaluation of a multi-component violence prevention program for middle schools: The Stu-dents for Peace Project. Health Education Research, 15, 45-58.

Patterson, G. R., Capaldi, D., & Bank, L. (1991). An early starter model for predicting delinquency. InD. J. Pepler & K. H. Rubin (Eds.), The development and treatment of childhood aggression(pp. 139-168). Hillsdale, NJ: Lawrence Erlbaum.

Powell, K. E., Dahlberg, L. L., Friday, J., Mercy, J. A., Thornton, T., & Crawford, S. (1996). Preven-tion of youth violence: Rationale and characteristics of 15 evaluation projects. American Jour-nal of Preventive Medicine, 12(5), 1-2.

Prinz, R. J., Blechman, E. A., & Dumas, J. E. (1994). An evaluation of peer coping–skills training forchildhood aggression. Journal of Clinical Child Psychology, 23(2), 193-203.

Quinn, M. M., Mathur, S. R., & Rutherford, R. B. (1995). Early identification of antisocial boys: Amulti-method approach. Education and Treatment of Children, 18(3), 272-281.

Reid, J. B., Eddy, J. M., Fetrow, R. A., & Stoolmiller, M. (1999). Description and immediate impactsof a preventive intervention for conduct problems. American Journal of Community Psychol-ogy, 27(4), 483-517.

Searle, S. R., Speed, F. M., & Milliken, G. A. (1980). Population marginal means in the linear model:An alternative to least squares means. American Statistician, 34(4), 216-221.


Shapiro, J. P., Burgoon, J. D., Welker, C. J., & Clough, J. B. (2002). Evaluation of the PeacemakersProgram: Prevention for students in grades four through eight. Psychology in the Schools,39(1), 87-100.

Snyder, H. N., & Sickmund, M. (1999). Juvenile offenders and victims: 1999 national report. Wash-ington, DC: Office of Juvenile Justice and Delinquency Prevention.

SPSS. (1999). SPSS 10.0 syntax reference guide. Chicago: Author.SPSS. (2002). AnswerNet solution #100010254. Available at www.spss.comStanger, C., & Lewis, M. (1993). Agreement among parents, teachers, and children on internalizing

and externalizing behavior problems. Journal of Clinical Child Psychology, 22(1), 107-115.Stoolmiller, M., Eddy, J. M., & Reid, J. B. (2000). Detecting and describing preventive intervention

effects in a universal school-based randomized trial targeting delinquent and violent behavior.Journal of Consulting and Clinical Psychology, 68(2), 296-306.

Tabachnik, B. G., & Fidell, L. S. (1989). Using multivariate statistics. New York: Harper & Row.U.S. Department of Health and Human Services. (2001). Youth violence: A report of the surgeon gen-

eral. Rockville, MD: U.S. Department of Health and Human Services, Centers for DiseaseControl and Prevention, National Center for Injury Prevention and Control; Substance Abuseand Mental Health Services Administration, Center for Mental Health Services; and NationalInstitutes of Health, National Institute of Mental Health.

U.S. Department of Justice. (2001a). Crime index trends, 2000 preliminary figures. Available atwww.fbi.gov/pressrel/pressrel01/ucrprelim2000.htm

U. S. Department of Justice. (2001b). Four measures of serious violent crime. Available at www.ojp.usdoj.gov/bjs/glance/tables/4meastab.htm

Vazsonyi, A. T., Vesterdal, W. J., Flannery, D. J., & Belliston, L. M. (1999). The utility of child self-reports and teacher ratings in classifying children’s official delinquency status. Studies ofCrime and Crime Prevention, 8(2), 1-20.

Viemerö, V. (1996). Factors in childhood that predict later criminal behavior. Aggressive Behavior,22(2), 87-97.

Walker, H. M., & McConnell, S. R. (1988). The Walker-McConnell Scale of Social Competence andSchool Adjustment: A social skills rating scale for teachers. Austin, TX: PRO-ED.

Walker, H. M., & McConnell, S. R. (1995). The Walker-McConnell Scale of Social Competence andSchool Adjustment (SSCSA). Florence, KY: Thomson Learning.

Wasserman, G. A., & Miller, L. S. (1998). The prevention of serious and violent juvenile offending. InR. Loeber & D. P. Farrington (Eds.), Serious and violent juvenile offenders: Risk factors andsuccessful interventions (pp. 197-247). Thousand Oaks, CA: Sage.

Weissberg, R. P., & Bell, D. N. (1997). A meta-analytic review of primary prevention programs forchildren and adolescents: Contributions and caveats. American Journal of Community Psy-chology, 25(2), 207-214.

Winer, B. J. (1971). Statistical principles in experimental design (2nd ed.). New York: McGraw-Hill.

Alexander T. Vazsonyi is an associate professor of human development and familystudies at Auburn University. He received his Ph.D. from the University of Arizona. Hisresearch interests include etiological risk factors in adolescent problem behaviors,deviance, and violence. Recent publications have appeared in the Journal of Research inCrime and Delinquency, the Journal of Quantitative Criminology, and Criminal Justiceand Behavior.


Lara M. Belliston is currently a doctoral candidate in the Department of Human Devel-opment and Family Studies at Auburn University. Her research interests include adoles-cent development, family relationships, and program evaluation.

Daniel J. Flannery, Ph.D., is a professor of justice studies and director of the Institutefor the Study and Prevention of Violence at Kent State University. He received his Ph.D.in 1991 in clinical psychology from The Ohio State University. He was coeditor (withC. R. Huff) of Youth Violence: Prevention, Intervention and Social Policy (1999). Hisprimary areas of interest are in youth violence prevention, the link between violence andmental health, and program evaluation.


Child self-report of aggressive behavior in Grades 3–5 was assessedusing items generated specifically for this study. The 9-item scale con-tained items such as “I hit someone” or “I put down other kids” that wererated on a 3-point scale ranging from no (1) to a lot (3). The scaledemonstrated adequate internal consistency (� � .86 at baseline). Childrenin Grades K–2 answered yes or no to five items assessing whether they gotinto trouble at school, if they ever got into fights, and if they ever cut in line(� � .66 at baseline).

Social competence. Teachers rated child social competence using theelementary school version (Grades K-6) 19-item short form of the Walker–McConnell (W-M) Scale of Social Competence and School Adjustment(Walker, Irvin, Noell, & Singer, 1992; Walker & McConnell, 1995). TheW-M scale has three subscales: School Adjustment (7 items), Peer-Preferred Behaviors (7 items), and Teacher-Preferred Behaviors (5 items).The School Adjustment subscale assesses adaptive social–behavioral com-petencies highly valued by teachers within classroom instructional con-texts. Peer-Preferred Behaviors reflect peer values concerning forms ofsocial behavior that govern peer dynamics and social relations withinfree-play settings. Teacher-Preferred Behaviors reflect teacher ratings ofsensitivity, empathy, cooperation, self-control, and socially mature formsof behavior in peer relations. Teachers responded to such items as “appro-priately copes with aggression from others” on a 5-point Likert scaleranging from never (1) to frequently (5). The W-M scale has demonstratedhigh internal consistency and test–retest reliability and correlates withother teacher and child self-report measures of social competence (Walker& McConnell, 1995). In the present sample, the internal consistency of theW-M scale was high (� � .95 at baseline). The W-M scale has been usedin other preventive intervention studies with elementary-school-age chil-dren to differentiate behavior outcomes between treatment groups (e.g.,Reid et al., 1999).

Prosocial behavior. Prosocial behavior for children in Grades 3–5 wasmeasured with a 16-item instrument designed for this study. The itemsassessed child self-reported empathy, caring, helpfulness, and support ofothers. Sample items include “I helped adults at school without beingasked” and “I helped other kids.” Children responded to each item using a3-point scale that included no (1), a little (2), and a lot (3). The scale itemsloaded on a single factor (eigenvalue � 6.34) and displayed high internalconsistency (� � .92 at baseline). Children in Grades K–2 answered yes,sometimes or no, not really to six questions assessing sharing, helpfulness,saying “thank you,” and saying “I’m sorry” (� � .51 at baseline).

Peace-building behavior. Child self-report of peace-building behaviorin Grades 3–5 was assessed with three items: “I helped build peace atschool,” “I told other kids they were peace builders,” and “I earned rewardsfor peace building.” Responses on the 3-point scale ranged from no to a lot.The three items loaded on a single factor (eigenvalue � 1.86) and dem-onstrated adequate internal consistency (� � .72 at baseline). Children inGrades K–2 responded yes or no to four items about building peace suchas “I helped build peace at school” and “I earned rewards for peacebuilding.” This yes/no scale demonstrated marginal internal consistency(� � .58 at baseline).

Teacher training. Immediately after teachers participated in an in-service training session, workshop or institute, they completed a 10-itemsurvey designed to assess the clarity and effectiveness of the training andtheir impressions of whether the materials and program would be easy ordifficult to implement. Sample items, rated on a 5-point scale ranging fromstrongly agree to strongly disagree, included “The basic philosophy behindPeaceBuilders is easy to understand”; “The training provided for theprogram was clear, effective and easy to follow”; and “As an interventionprogram, PeaceBuilders will be difficult to implement.”

Implementation and fidelity. In the spring of Year 2 (Time 4), teacherscompleted an 8-item survey that assessed their use and implementation ofprogram materials. Some items assessed frequency of use, such as “I usethe PeaceBuilders curriculum in my classroom” answered on a 5-pointscale including daily (1), occasionally (3), and not at all (5). Other items

assessed degree of satisfaction or effectiveness of the program, such as“PeaceBuilders is easy to use” or “Overall, my school has implemented thePeaceBuilders curriculum,” answered on a 4-point scale ranging fromstrongly agree (1) to strongly disagree (4). Teachers were also asked toindicate the total number of core PeaceBuilders materials they used in theirclassrooms. These included the Action Guide, reproducible binders (sep-arate lessons on PeaceBuilders rules), the “I Help Build Peace” storybook,praise notes in class, praise notes sent home, “First Aid for Anger,” thePlayground Guide, and the Intensive PeaceBuilders Guide.

Analysis Plan

After presenting correlation data on the relationship between teacher-and child-reported outcomes, we provide some descriptive data on the levelof program implementation and teacher training. We then present data onsample attrition within and between school years and its relation to internalvalidity (differential attrition by intervention group) and external validity(characteristics lost to the sample). Then we turn to our main analyticquestions of year-to-year differences in the immediate and delayed inter-ventions’ effects on our outcomes of interest. We conducted two maintypes of analyses to address our specific hypotheses regarding school-yearchanges in behavior outcomes relative to baseline.

First, we constructed a three-level hierarchical linear model (Version 5,Bryk & Raudenbush, 1992) to examine change in behavior assessed at fourpoints in time over 2 school years. The three levels of the model reflectchange over time (Level 1), individual effects (Level 2), and school effects(Level 3). The model was constructed to examine both short-term (Year 1)and longer term (Year 2) change in outcomes after controlling for baselinelevels and student gender. Because there was not continuous interventionover the summer months, we decided to model our effects by creating aseries of dummy variables for each data collection time point (Neter,Wasserman, & Kutner, 1983), with baseline as the reference (spring ofYear 1 � Time 2, fall of Year 2 � Time 3, spring of Year 4 � Time 4).Specifically, we first examined change from baseline to the spring semester(Time 2) in Year 1 for PBI schools and PBD schools. We also examineddifferences in Year 2 between schools with 2 years of intervention (PBIschools) and schools with 1 year of intervention (PBD schools). In PBIschools, we expected the most significant changes to occur at Time 2,after 1 year of intervention.

Hierarchical linear modeling (HLM) has several advantages for theanalysis of longitudinal data. First, responses on any outcome variablefrom the same individual over time will be correlated, thus violating theassumption about independent sample observations embedded in moststatistical models dealing with cross-sectional data, and HLM takes thiscorrelation into account. This intraclass correlation also needs to be takeninto account when school is used as the unit of assignment to condition(Koepke & Flay, 1989; Murray & Wolfinger, 1994; Piper, Moberg, &King, 2000; Rooney & Murray, 1996). Second, when applying conven-tional linear models to analyzing longitudinal data, one generally under-estimates the standard errors of the impacts and therefore may erroneouslyassume statistical significance. HLM effectively handles this problem aswell as others inherent in longitudinal data, such as varying times betweenobservations, unequal groups at each data point over time, and the need tocontrol for the effects of potentially confounding independent variables(Bryk & Raudenbush, 1992; Diggle, Liang, & Zeger, 1994; Lindsey,1993). These advantages make HLM more appropriate than the moreconventional repeated measures analyses used in longitudinal studies.

The second main analytic approach was a differential analysis on Year 1baseline to Time 2 data for all outcome variables. Because of our delayedintervention model, Year 1 was the only period in which we had interven-tion schools compared with nonintervention schools. Our analytic proce-dure followed the protocol developed by Stoolmiller et al. (2000) andexamined the extent to which intervention effectiveness depended on anindividual’s initial (baseline) status on an outcome of interest.

298 FLANNERY ET AL.

Although we expected some gender differences between students atbaseline (e.g., boys being more aggressive, girls being more sociallycompetent), we did not expect differences on outcomes between schools atbaseline. In general, we expected the PBI and PBD schools to be signifi-cantly different at Time 2 (spring of Year 1) and perhaps at Time 3 (fall ofYear 2) because PBD schools would have just begun their interventions.We expected that the PBI and PBD schools might be significantly differentfrom each other at Time 4 (spring of Year 2), although we expected allscores at Time 4 to reflect improvement (e.g., in social competence) ordecline (e.g., in aggressive behavior) relative to baseline.

Results

Zero-Order Correlations Among Outcome Variables

In preliminary analyses, we examined the zero-order correla-tions among outcome variables.1 The two main outcomes of in-terest, child social competence and aggression, were significantlyrelated; teacher-rated aggression was negatively related to teacher-rated social competence, r(1613) � �.56, p � .001, at baseline.This relationship was largely unchanged over the four data collec-tion points, and the correlation ranged from �.55 to �.66. Forchildren in Grades K–2, there were small to moderate relationshipsat baseline between child self-reported prosocial behavior andaggression, r(650) � �.03, ns; between aggression and peace-building behaviors, r(650) � �.08, p � .05; and between proso-cial and peace-building behaviors, r(650) � .25, p � .001. Forself-reports of children in Grades 3–5, the strongest relationshipwas between peace building and prosocial behavior, r(1879) �.69, p � .001. Relationships between aggressive behavior andprosocial behavior, r(1886) � �.23, p � .001, and between peacebuilding and aggressive behavior, r(1879) � �.13, p � .001, werenot as strong. Teacher reports of aggression were related to childself-reports of aggression at baseline, r(1316) � .34, p � .001, butrather modestly given the large sample size. The correlationsbetween age and most outcome variables were statistically signif-icant but weak, ranging from r(674) � .01, ns for child self-reportsof peace-building behavior to r(1878) � �.25, p � .001 for childself-reports of prosocial behavior in Grades 3–5. Correlations atbaseline between age and social competence and between age andaggression were significant but low, averaging .08 ( p � .01).

Teacher Satisfaction With Training

All regular and special education teachers in participatingschools participated in the half-day workshops (n � 194). Overthe 2 years of intervention, training questionnaires were gatheredfrom a total of 134 teachers (69%), 57 of whom were from PBIschools (43%) and the remainder of whom (n � 77) were fromPBD schools. Overall, 93% of teachers indicated they “stronglyagreed” or “agreed” that the basic philosophy behind the Peace-Builders intervention was easy to understand. Seventy-seven per-cent agreed or strongly agreed that the training provided was clear,effective, and easy to follow, and 83% agreed or strongly agreedthat the ideas would be easy to use in the classroom. Three of fourteachers who completed surveys believed that “PeaceBuilders willbe very successful as an intervention” and strongly agreed oragreed that “The school administration stands behind this inter-vention effort 100 percent.”

Level of Implementation and Fidelity

A total of 190 teachers (98%) completed a spring 1996 (Time 4)self-assessment of their use of intervention materials in theirclassrooms. Teachers completing the survey were distributedacross all participating schools and grades and represented allparticipating teachers of Grades K–5 in each school. Teacherswere equally divided between immediate- and delayed-intervention schools. The majority of teachers surveyed indicatedthat they used the PeaceBuilders curriculum in their classrooms ona daily (48%) or weekly (32%) basis. Nearly all teachers (98%)strongly agreed or agreed that “Overall, my school has imple-mented the PeaceBuilders curriculum,” 53% rated implementationas “extensive,” and 43% rated implementation as “moderate.”Teachers were also consistent in their agreement that the interven-tion “has decreased the level of violence in our school” (94%) and,conversely, that “PeaceBuilders has increased prosocial interac-tions in my class and in our school” (94%). Regarding the totalnumber of program materials used, teachers reported, on average,that they used at least four of the eight core sets of materials intheir classrooms. Teachers in the PBD schools reported, more thandid teachers in the PBI schools, that during Year 2 they were morelikely to use program materials daily (compared with weekly),�2(4, N � 190) � 14.64, p � .01.

Attrition

We first calculated attrition within each intervention year (frombaseline to Time 2 in Year 1 and from Time 3 to Time 4 in Year2) and between Years 1 and 2 to determine rates of attrition and todetermine whether there was differential attrition by interventiongroup. We also examined differences in outcomes between stu-dents with baseline-only data and those with baseline data plus atleast one additional data point over the 2-year period. In a secondset of analyses, we examined demographic characteristics relatedto attrition between PBI and PBD schools. Finally, we examinedour two main outcomes of interest, teacher-rated social compe-tence and aggression, to determine whether children lost from thesample after baseline were different from those children whoremained part of the sample. All attrition analyses on outcomeswere conducted separately for the Grades K–2 and Grades 3–5samples.2

Within each intervention year, the average rate of attrition (falldata but no spring data) was 12% in Year 1 and 17% in Year 2.Between-years attrition was 32% for students in Grades K–2 (331of 1,037 students) and 28% for students in Grades 3–5 (231 of

1 These zero-order correlations do not take into consideration intragroupcorrelation among students within classes and therefore serve only adescriptive and exploratory purpose.

2 To corroborate our attrition analyses, we also conducted logistic re-gressions with attrition status as the outcome variable. We ran regressionswith grade, gender, intervention-group membership, Grades K–2 teacher-rated social competence and aggression, as well as Grades 3–5 teacher-rated social competence and aggression as independent variables. Theresults of these regression analyses were consistent with the analysis ofvariance and chi-square results reported here. To control for possiblevariation due to grade or gender, we also ran regressions controlling forthose variables, and the results remained the same.


814). We did not include new kindergarten students in Year 2, nordid we track Year 1 fifth-grade students into Year 2. For allstudents assessed at baseline, 169 (10%) of 1,615 students (GradesK–2) and 120 (9.5%) of 1,140 students (Grades 3–5) had no otherdata over the 2-year period.

Students in Grades K–2 with baseline-only data (those lost toattrition) were rated by teachers at baseline as more aggressive,F(1, 1612) � 11.05, p � .01, and less socially competent, F(1,1611) � 7.09, p � .01, than were students who remained part ofthe study sample. Students in Grades 3–5 with baseline-only datawere also rated by their teachers as more aggressive, F(1,1258) � 14.70, p � .01, and less socially competent, F(1,1258) � 13.60, p � .01, than were students who remained partof the study sample. Rates of attrition from baseline were notsignificantly different by gender, grade, or intervention-groupmembership for either children in Grades K–2, �2(1, N� 1,615) � 0.804, p � .05, or children in Grades 3–5, �2(1,N � 1,260) � 0.389, p � .05.

Behavior Outcomes

HLM was our main analytic approach to examining school-leveleffects. We used a three-level hierarchical linear model, with thefirst level representing change over time, the second level repre-senting individual student differences (gender), and the third levelrepresenting differences between schools. The model examineddifferences between schools after controlling for baseline levels ofbehavior (�0ij) and gender. The Level 1 model was specified as

Yhij � �0ij � �1ijT2 � �2ijT3 � �3ijT4 � ehij.

The Level 2 model was specified as

�0ij � �00j � �01jMALE � r0ij

�1ij � �10j

�2ij � �20j � r2ij

�3ij � �30j � r3ij.

The Level 3 model was specified as

�00j � �000 � �001PBI � u00j

�01j � �010

�10j � �100 � �101PBI

�20j � �200 � �201PBI

�30j � �300 � �301PBI.

�0ij represents the intercept or baseline. T2 represents datacollected in the spring of Year 1 (Time 2), T3 represents datacollected in the fall of Year 2 (Time 3), and T4 represents datacollected in the spring of Year 2 (Time 4). The Level 1 error term,ehij, is assumed to be normally distributed with a zero mean and aconstant variance. At Level 2, MALE is the dichotomous gendervariable equal to 1 if the child was a boy and 0 if the child was agirl. The Level 2 random effects r0ij, r2ij, and r3ij are assumed to benormally distributed with a zero mean and a constant variance. Thevariable PBI represents the PeaceBuilders immediate intervention

(as opposed to the delayed intervention, or PBD). Adding the errorterm (u00j) to the intercept equation at the school level (Level 3)takes into account the autocorrelation within schools—namely, thenonindependence of students within a school.

In determining the specification of our model, we followed therecommendation of Snijder and Bosker (1999) by first testing thesignificance of random effects in our models. The models withsignificant random effects were then compared using Akaike’sinformation criterion (AIC) and Schwarz’s Bayesian criterion(SBC). These criteria measure whether specifying additional ran-dom effects improves fit if the models under comparison have thesame structure of fixed effects. A larger value of AIC or SBC is anindication of better fit (Littell, Milliken, Stroup, & Wolfinger,1996; see also Guo & Hussey, 1999). The model with the randomspecification above emerged most consistently as the model withthe largest likelihood function, and the best fit to the data, com-pared with all the other models that we explored.3

The fixed effects presented in Tables 2 and 3 illustrate semestereffects that reflect differences (not taking into account other fac-tors such as intervention) on outcomes over time. Individual ef-fects reflect Level 2 gender differences at baseline, and the schooleffects reflect differences between PBI and PBD schools at base-line. As shown in Tables 2 and 3, the random effects are statisti-cally significant, which indicates that specifying such extra hetero-geneity to control for intragroup correlation is necessary. Schooleffects reflect baseline differences in outcomes between PBI andPBD schools. The Level 3 effects are Semester � School interac-tion effects. These results show, after controlling for baseline(Level 1) and gender (Level 2), how immediate-interventionschools compared with delayed-intervention schools on the out-comes of interest at each subsequent point in time: spring of Year 1(Time 2), fall of Year 2 (Time 3), and spring of Year 2 (Time 4).

To address the issue of floor effects in the aggression scales, welog-transformed aggression scale scores for both teacher and childreports (Cohen & Cohen, 1983; cf. Stoolmiller et al., 2000). Forexample, at baseline, for Grades K–2 and Grades 3–5 teacher-reported aggression, teachers identified 33% and 37% of the chil-dren, respectively, as not engaging in any aggressive behavior. Atbaseline, for Grades K–2 and Grades 3–5 child self-reported ag-

3 To determine the best Level 2 random effects specification for ourmodel, we ran all possible permutations of random effects for all 10 of theoutcome variables. Following the recommendation of Snijder and Bosker(1999), we started by testing the significance of random effects in ourmodels. We found two models had significant random effects for most ofthe outcome variables. Model A (the model we used) had significantrandom effects for nine of the outcome variables, and Model B (with onlyone random effect at Level 2 for the intercept) had significant randomeffects for all 10 of the outcome variables. To determine which modelprovided a better fit to the data, we then compared the models using theAIC and the SBC. For example, for Grades K–2 teacher-rated competence,for Model A, AIC � �20,936 and SBC � �20,943, and for Model B,AIC � �21,077 and SBC � �21,072. Model A had consistently largerAIC and SBC values than Model B for all 10 outcome variables. Thus,Model A emerged most consistently as the model with the largest likeli-hood function, and the best fit to the data, compared with all the othermodels that we explored. However, for one of the outcome variables,Grades K–2 child-reported peace-building behavior, we used Model B,because two of the random effects at Level 2 for Model A were notsignificant.

300 FLANNERY ET AL.

Table 2Hierarchical Linear Modeling Results: Teacher Ratings of Child Social Competence andAggressive Behaviors

Fixed effects

Coefficients (standard errors)

Kindergarten–2nd gradeteacher ratings

3rd–5th gradeteacher ratings

Socialcompetence

Logaggression

Socialcompetence

Logaggression

Semester effectsBaseline 72.12*** (1.25) 1.49*** (.008) 73.28*** (0.94) 1.46*** (.007)Spring Year 1 2.52*** (0.51) �0.013** (.004) 1.39** (0.49) 0.020*** (.004)Fall Year 2 0.31 (0.75) �0.017** (.006) �0.08 (0.69) 0.005 (.005)Spring Year 2 2.24** (0.78) �0.011† (.006) �0.12 (0.66) 0.009† (.005)

Individual effectsGender: boy (Reference:

girl)�7.12*** (0.57) 0.061*** (.005) �7.22*** (0.53) 0.070*** (.004)

School effects (baseline)PBI (Reference: PBD) 2.27 (1.77) �0.034* (.010) 0.94 (1.18) 0.006 (.010)

Semester � School interactioneffects

Spring Year 1 � PBI 3.05*** (0.66) 0.006 (.005) 1.06† (0.62) 0.017** (.005)Fall Year 2 � PBI 8.20*** (1.08) �0.014† (.008) 4.98*** (0.95) �0.026*** (.007)Spring Year 2 � PBI 7.17*** (1.10) �0.009 (.008) 6.24*** (0.88) �0.019** (.007)

Random effects varianceLevel 2, r0ij 157*** .011*** 162*** .012***Level 2, r2ij 134*** .009*** 146*** .004***Level 2, r3ij 138*** .008*** 84.6*** .004***Level 3, u00j 4.66*** .000*** 1.71*** .000***

Note. PBI � PeaceBuilders immediate-intervention schools; PBD � PeaceBuilders delayed-intervention schools.† p � .10. * p � .05. ** p � .01. *** p � .001.

Table 3Hierarchical Linear Modeling Results: Child Self-Report of Aggressive, Prosocial, and PeaceBuilding Behaviors

Fixed effects

Coefficients (standard errors)

Kindergarten–2nd grade self-report 3rd–5th grade self-report

Log aggression Prosocial PeaceBuilding Log aggression Prosocial PeaceBuilding

Semester effectsBaseline 0.27*** (.01) 5.71*** (.05) 3.43*** (.06) 1.00*** (.01) 33.60*** (.58) 5.50*** (.13)Spring Year 1 �0.01 (.01) �0.05 (.05) 0.01 (.06) 0.01 (.00) �0.18 (.30) �0.32*** (.07)Fall Year 2 �0.01 (�.02) �0.05 (.06) 0.23** (.07) �0.01** (.00) 1.10** (.35) 0.71*** (.09)Spring Year 2 0.03† (�.02) �0.02 (.06) 0.17* (.10) �0.000 (.01) �1.00** (.37) 0.40*** (.09)

Individual effectsGender: boy (Reference: girl) 0.09*** (.01) �0.11** (.03) �0.11** (.04) 0.06*** (.00) �4.00*** (.26) �0.63*** (.06)

School effects (baseline)PBI (Reference: PBD) �0.02 (.02) �0.02 (.06) �0.03 (.07) �0.02 (.01) 1.40 (.78) 0.03 (.17)

Semester � School interaction effectsSpring Year 1 � PBI 0.02 (.02) 0.10 (.07) 0.19* (.08) �0.003 (.01) �0.77* (.39) 0.71*** (.10)Fall Year 2 � PBI �0.02 (.03) 0.19* (.09) 0.02 (.09) �0.001 (.01) �0.38 (.47) 0.36** (.12)Spring Year 2 � PBI �0.01 (.03) 0.19* (.08) 0.01 (.10) 0.001 (.01) �1.1* (.50) �0.11 (.12)

Random effects varianceLevel 2, r0ij .021*** .108*** .158*** .007*** 36.0*** 1.58***Level 2, r2ij .012*** .120* NA .002*** 14.4*** 1.20***Level 2, r3ij .011*** .087** NA .005*** 24.6*** 1.72***Level 3, u00ij .000 .002* .002* .000*** .947*** .044***

Note. PBI � PeaceBuilders immediate-intervention schools; PBD � PeaceBuilders delayed-intervention schools; NA � not applicable.† p � .10. * p � .05. ** p � .01. *** p � .001.


gression, 31% and 44% of the students, respectively, reported notengaging in any aggressive behavior.

Baseline effects. With only one exception, students in PBI andPBD schools were not significantly different from each other atbaseline (see school effects for baseline in Tables 2 and 3).Teachers rated students in Grades K-2 in the PBI schools asslightly lower in aggressive behavior overall than students in thePBD schools (see Table 2).

As expected, there were several gender differences at baseline.Teachers rated boys as significantly lower in social competenceand higher in aggressive behavior than girls at baseline. This effectwas consistent for both students in Grades K–2 and students inGrades 3–5 (see Table 2). Child self-reports of gender differencesat baseline showed that among students in Grades K–2 and Grades3–5, boys rated themselves as significantly more aggressive, lessprosocial, and lower in peace-building behavior than did girls (seeTable 3).

Intervention effects. Given the unique design of our evalua-tion, in which the PBI schools received the intervention duringYear 1 and Year 2 but the comparison PBD schools received theintervention only during Year 2, our hierarchical linear modeloutlined above enabled us to examine the effects of the interven-tion at each data collection point (Time 2, Time 3, and Time 4)relative to baseline by examining the cross-level Semester �

School interaction effects. For example, a significant Time 2 �School interaction would indicate that, with baseline and genderdifferences controlled, the PBI schools were significantly differentfrom the PBD schools at Time 2. These effects are illustrated forsignificant outcomes in Figure 3 (teacher data) and Figure 4 (childself-report data).

Time 2. At Time 2 (spring of Year 1), compared with studentsin PBD schools, teachers rated students in PBI schools as signif-icantly higher in social competence (for Grades K–2; effects weremarginal for students in Grades 3–5, p � .10) and significantlylower in log aggression (for Grades 3–5; see Table 2 for coeffi-cients). Compared with students in PBD schools, students in PBIschools reported significantly greater peace-building behavior (forboth Grades K–2 and Grades 3–5; see Table 3 for coefficients) butself-reported less prosocial behavior (Grades 3–5).

Time 3. At Time 3 (fall of Year 2), compared with students inPBD schools, teachers rated students in Grades K–2 and Grades3–5 in PBI schools as significantly higher in social competenceand significantly lower on log aggression, although the effectswere stronger for students in Grades 3–5 ( p � .001) than for thosein Grades K–2 ( p � .10; see Table 2). Compared with students inPBD schools, students in PBI schools reported significantly greaterpeace-building behavior (Grades 3–5) and prosocial behavior(Grades K–2; see Table 3 for coefficients).

Figure 3. Means for teacher-reported social competence and aggression for PeaceBuilders immediate-intervention and PeaceBuilders delayed-intervention schools at baseline, Time 2, Time 3, and Time 4.

302 FLANNERY ET AL.

Time 4. At Time 4 (spring of Year 2), compared with studentsin PBD schools, teachers rated the students in PBI schools assignificantly higher in social competence (Grades K–2 and Grades3–5) and lower on log aggression (Grades 3–5 students only; seeTable 2 for coefficients). Compared with students in PBD schools,students in PBI schools reported significantly greater prosocialbehavior in Grades K–2 but lower prosocial behavior in Grades3–5 (see Table 3).

Differential effects. We conducted a series of linear regres-sions to examine the potential for differential effectiveness of theintervention—namely, that treatment effects would vary depend-ing on an individual student’s initial status on an outcome atbaseline. We conducted differential analyses only for Year 1 data(from baseline to Time 2) because this was the only year in whichwe had intervention (PBI) and nonintervention (PBD) comparisongroups. We conducted differential analyses on all main outcomesof interest.

Following the protocol developed by Stoolmiller et al. (2000),we conducted a linear regression of Time 2 (spring of Year 1)scores on baseline scores. We were interested in whether theregression slopes would significantly differ. If the slopes are notparallel, this is evidence that treatment effects vary according toinitial status (or that the treatment is differentially effective). For

teacher-rated dependent variables, we found significantly differentslopes for Grades 3–5 log aggression, t(1174) � 3.84, p � .001.For child self-reported dependent variables, we found significantlydifferent slopes for Grades K–2 peace-building behavior, t(649) ��2.46, p � .05; Grades K–2 prosocial behavior, t(649) � �2.48,p � .05; Grades 3–5 prosocial behavior, t(1494) � 1.97, p � .05;and Grades 3–5 aggression, t(1494) � 14.19, p � .001. There wereno differential effects for social competence, Grades K–2 teacher-reported or child self-reported aggression, or Grades 3–5 childself-reported peace-building behavior.

We can illustrate the difference on aggression by mapping theeffect sizes for aggression scores at four points: �1, 0 (mean), 1,and 2 SDs above the mean for the baseline sample. The meandifference is computed by plugging the preintervention score intothe fitted equation for both groups and then subtracting the pre-dicted intervention mean from the predicted control mean (Stool-miller et al., 2000). The obtained effect sizes for teacher-reportedGrades 3–5 log aggression were .00, .26, .52, and .78, respectively,and those for child self-reported Grades 3–5 log aggression were�.17, �.08, .02, and .12, respectively. Thus, the effect size (i.e.,treatment effect) was larger for students with higher aggressionscores at baseline.

Figure 4. Means for child self-reported peace-building and prosocial behavior for PeaceBuilders immediate-intervention and PeaceBuilders delayed-intervention schools at baseline, Time 2, Time 3, and Time 4.


For peace-building and prosocial behavior, we can illustrate theeffect sizes at four points: �2, �1, 0 (mean), and 1 SD above themean, because we would expect children lower at baseline toincrease their positive behavior after intervention. The obtainedeffect sizes were .48, .27, .07 and �.13, respectively, for GradesK–2 child self-reported peace-building behavior. Similarly, forGrades K–2 prosocial behavior, the obtained effect sizes were .57,.38, .18, and �.01, respectively. For Grades 3–5 child self-reportedprosocial behavior, the obtained effect sizes were .27, .16, .05, and�.06, respectively. The effect size was larger for students withlower baseline peace-building and prosocial behavior scores, sug-gesting a bigger treatment effect for increases in positive behaviorfor students who were lower at baseline.

Discussion

This study examined the initial behavior outcomes of Peace-Builders, a universal school-based preventive intervention pro-gram focused on reducing aggressive behavior and increasingsocial competence. We examined behavior change over 1 schoolyear in which half of our randomly assigned schools receivedimmediate intervention and half received no intervention. We alsoexamined change in Year 2, when the immediate-interventionschools continued treatment and the control schools received in-tervention for the first time. In general, we found consistentbehavior effects in Year 1, with students in Grades K–2 in theimmediate-intervention schools being rated significantly higher byteachers on social competence than control students (moderateeffects were obtained for students in Grades 3–5). Third- to fifth-grade students in the immediate-intervention schools were alsorated by teachers as significantly less aggressive than students innonintervention schools. As expected, students in the immediate-intervention condition also rated themselves higher on peace-building behaviors (Grades K–5) than control students. Thesebehavior changes occurred in intervention schools during Year 1,when no significant change in behavior was observed in noninter-vention schools.

Effects for increases in social competence and declines inteacher-reported aggressive behavior were maintained for all stu-dents in Grades K–5 in immediate-intervention schools in the fallof Year 2. Higher levels of peace building (Grades 3–5) andprosocial behavior (Grades K–2) were also maintained at Time 3and at Time 4 (Grades K–2 for prosocial). At Time 4 (spring ofYear 2), students from immediate-intervention schools were stillrated higher on social competence, higher on prosocial behavior(Grades K–2), and lower on aggression (Grades 3–5) relative tostudents in delayed-intervention schools. Our overall findings areconsistent with previous studies that have demonstrated the effi-cacy of elementary-school-based universal prevention programsfor increasing social competence and reducing aggressive behavior(CPPRG, 1999; Grossman et al., 1997; Kellam, Ling, et al., 1998;Reid et al., 1999). The one trend that ran counter to expectationswas for older students’ (Grades 3–5) prosocial behavior. At bothTime 2 and Time 4, students in the immediate-intervention schoolsrated themselves as less prosocial than students in the delayed-intervention schools. In general, however, we found consistentintervention effects for social competence and aggression. Theseeffects were realized using a conservative three-level hierarchical

linear model that accounted for school-level differences and vari-ability in individual student change over time.

From a policy perspective, it is important that early preventiveintervention focus on increasing positive skills and competenciesas well as reducing aggressive and other problem behaviors. Theseskills lay the groundwork for success in school, positive adult–child and peer relations, and long-term child adjustment and resil-iency. Interventions should contain strategies specifically designedto accomplish both of these behavioral goals in order to increasethe chances of sustained behavior change over time (Mayer, But-terworth, Nafpaktitis, & Sulzer-Azaroff, 1983; Tolan et al., 1995;Tremblay et al., 1995). Developmentally, children who displayaggressive and socially incompetent behavior at school are also athigh risk for rejection by their normative peer group. This in-creases their risk of associating with other deviant or rejectedpeers, which in turn increases their risk of subsequent delinquencyand other conduct problems (CPPRG, 1999; Reid et al., 1999).

The fact that we found effects for students both in Grades K–2and Grades 3–5 also underscores the importance of providingpreventive intervention services early in elementary school. Themajority of school-based violence prevention programs are inmiddle schools (e.g., Farrell & Meyer, 1997; Orpinas, Parcel,McAlister, & Frankowski, 1995) or high schools (Howard et al.,1999), but there is ample evidence that intervening earlier inelementary school can have greater effects on both educationaloutcomes and risk behaviors than can waiting to intervene later(CPPRG, 1999; Dolan et al., 1993; Kellam & Anthony, 1998;Tremblay et al., 1995) and that early and continued intervention inthe elementary grades can help put children on a positive devel-opmental course that is maintained through high school (Hawkinset al., 1999).

This study also reinforces the need to consider the potentialdifferential effects of preventive intervention trials (Stoolmiller etal., 2000). Although we found some significant effects for allchildren exposed to the intervention, which is important for uni-versal prevention efforts (compared with targeted interventionsthat focus on high-risk youth), we also found larger treatmenteffects for youth in Grades 3–5 who were higher on aggression atbaseline. We found these differential effects for both teacher-reported and child self-reported aggression. We did not, however,see differential effects for the aggressive behavior of youngerchildren (Grades K–2) even though teacher-rated aggression ofchildren in Grades K–2 was the only significant school-leveldifference at baseline. We did find differential effects for childself-reported prosocial behavior despite finding significant de-clines overall in prosocial behavior for children in Grades 3–5.Children who were the least prosocial at baseline improved themost after 1 year of intervention. In general, effect sizes were inthe moderate range (.27–.78) for the children at highest risk atbaseline, defined here as � 2 SD above the sample mean. Effectsfor children closer to the sample mean at baseline were not asdramatic. As Battistich and colleagues (Battistich, Schaps,Watson, Solomon, & Lewis, 2000) and others (Stoolmiller et al.,2000; Vazsonyi et al., 1999) have pointed out, however, fewchildren in early elementary school have begun to show seriousconduct problems. We adjusted for low base rates of aggression inour models and still found significant effects for aggressive be-havior. Even small early differences may lead to large preventiveeffects as children mature, a position that is consistent with models

304 FLANNERY ET AL.

of the developmental progression of conduct problems and violentbehavior (CPPRG, 1999; Reid et al., 1999; Tolan et al., 1995).

Although a specific focus on variations in the fidelity of treat-ment implementation is beyond the scope of this article, we pro-vide some evidence (a) that the program was implemented asdesigned and was provided by teachers with reasonable intensity,(b) that teachers were satisfied with their training and programmaterials, and (c) that students seemed to acquire the skills thatwere emphasized as part of the school-based program (e.g., peace-building behavior). First, over 90% of teachers who responded tosurveys indicated that the philosophy behind the program was easyto understand, and over 80% believed the ideas would be easy touse in the classroom. It is extremely important that teachers un-derstand the reason they are implementing a particular curriculumor activity and the intended impact of the intervention. If thematerials are difficult to implement, few teachers will take the timeor effort to adapt them. There exist too many demands on alreadybusy teachers for them to implement complicated programs thatthey do not understand or support. It is imperative that psycholo-gists continue to evaluate the implementation and effectiveness ofscientifically based, easy-to-implement, and cost-effective preven-tion programs. Few violence prevention programs systematicallyfocus on the importance of staff training or on assessing thefidelity of program implementation (Flannery & Seaman, 2001).

Another indicator of program implementation was the surveythat teachers completed at the end of the 2nd year of intervention,when all teachers surveyed had been trained and had been imple-menting the curriculum for at least 1 school year. Nearly allteachers in project schools completed the surveys, and 8 of 10reported that they used materials in their classrooms on a regularbasis. In some ways, observed changes in peace-building behavioralso acted as a validity check that the program was being imple-mented as intended. There was also strong consensus that theprogram had been integrated at the school level. In fact, by the endof the 2nd year, both participating school districts had formallyadopted the program as part of their regular curriculum.

For two of our outcomes of interest, peace-building behaviorand prosocial behavior, the data suggest a potential “summereffect” in that students in the delayed-intervention schools self-reported a significant increase in behavior at Time 3 (fall of Year2) relative to baseline. At least two factors may explain these“spikes,” which were not realized for aggression or social compe-tence. First, most teachers in the PBD schools received trainingimmediately at the beginning of the school year (given theirinterest in implementation). Teacher and student survey data weregathered about 1 month into the fall semester, so most students inthe PBD condition had at least some initial exposure to the inter-vention. Second, peace-building and prosocial behaviors are themost intervention-specific behaviors we assessed, so an increase inchild self-reports may reflect a response to initial interventionexposure.

Several important characteristics, particularly those related tothe likelihood of observing systematic behavior change, separatePeaceBuilders from other school-based violence prevention pro-grams. First, the focus of PeaceBuilders was to alter the entireschool climate, not just individual risk factors. Second, Peace-Builders was implemented in the immediate-intervention schoolsfor a longer period than is the case for most other time- orcurriculum-limited prevention programs, and once it was imple-

mented, it was maintained over time with no prespecified end pointto the intervention. Third, PeaceBuilders focused on universalprevention with children beginning in kindergarten. Persistentbehavioral change is more likely to occur when children areyounger, the behavior is more malleable, and the intervention ismaintained over time (Tolan et al., 1995).

Conducting program evaluation on a large number of students inpredominantly urban, mobile school populations presents manyempirical and practical challenges not easily overcome. Attritioncan have an adverse impact on behavioral outcomes, especially iflongitudinal samples are not large enough to provide adequatepower to detect treatment effects over a long period of time.Attrition in our sample was not negligible, although our rates arecomparable to those of other studies conducted with higher risk,frequently mobile students and families (Hansen, Tobler, & Gra-ham, 1990).

Large-scale intervention studies also face attrition by teachers orattrition at the school level, with schools sometimes dropping outof a project. This may occur because of changes in administrators,changes in school district policy, reductions in resources, changingacademic demands (e.g., proficiency testing), or changes in teacherstaff to the point that there no longer exists a majority who arewilling to participate in training, to complete data collection in-struments, or to implement a program (e.g., CPPRG, 1999; Reid etal., 1999). Although we had no schools drop out of our study in thefirst 2 years, we did have one control school delay data collectionuntil the spring of the 1st year. There is a need to balance the gainsfrom doing large-scale preventive interventions with the limita-tions in research design and method that occur when attempting tobridge science and practice (Flannery & Huff, 1999).

Limits also exist on the extent to which one can control a child’sexposure to other school and community programs or events thatmay influence the outcome behavior being examined. We tookseveral steps to limit the cohort’s degree of exposure to theintervention. For example, PBD schools agreed not to implementthe PeaceBuilders program during the 1st year, and we removedfrom our sample children in the PBI condition in Year 2 who werenot also present in Year 1.

Methodologically, there exist significant challenges to doinglarge-scale preventive intervention work. Every school we ap-proached about participating in the project expressed a high needfor immediate intervention and was uneasy about the prospect ofeven a 1-year nonintervention period. Despite our strategy ofoffering monetary incentives to schools to remain in a 1-yearcontrol condition, it was difficult to withhold interventions fromschools that had a need for immediate help. Matters were furthercomplicated when we not only wanted to withhold intervention butalso requested detailed survey data from teachers and students.This influenced our design over the course of the multiyear lon-gitudinal study because of the absence of an ongoing noninterven-tion control or comparison group.

Another methodological challenge is presented by the increasedemphasis on school-level (i.e., climate or culture changing) pre-ventive interventions. Programs limited to a few schools maycompromise their chances of finding significant effects on behav-ior because of problems of limited sample size (e.g., the unit ofanalysis is the school rather than the individual child; see Battistichet al., 2000; Stoolmiller et al., 2000) and because individual


students in a school are not independent of each other with regardto the potential effects of a universal prevention program.

Despite the limitations inherent in applied evaluation research,this project also had many strengths. The sample was large andethnically diverse and included a significant number of Hispanicand Native American children, two groups rarely sampled inlongitudinal studies of violence prevention programs. The childrenwere younger and covered a broader age range than the childrenfound in many other previous longitudinal evaluations (e.g., Gross-man et al., 1997; but see CPPRG, 1999), and the schools werefrom both urban and nonurban districts. Although our focus herewas on the first 2 years of exposure to the intervention, we havecontinued to gather outcome data from the children as they maturethrough middle school (Grades 6–8), over a 5-year period. As-sessing outcomes such as aggression, delinquent and violent be-havior, and violence exposure/victimization as a function of yearsof exposure to intervention may yield more information about theeffects of age of first exposure, developmental trajectories forsubgroups of children (e.g., high-aggressive youth with low socialcompetence vs. high-aggressive youth with high social compe-tence), or differences in program effectiveness related to childgender or history of exposure to violence (Flannery, 2000).

In sum, this evaluation of a universal preventive interventionprogram for children in Grades K–5 showed significant improve-ments in child social competence and peace-building behavior, aswell as reductions in aggressive behavior, after 1 year of interven-tion relative to students in nonintervention schools. These effectswere largely maintained in a 2nd year of intervention. It is alsoimportant to examine the differential effects of treatment on ag-gression and prosocial behavior so that one does not falsely as-sume that a universal preventive intervention failed because itlacked the power to affect the population as a whole.

References

Achenbach, T. (1991). Manual for the Teacher Report Form and profile.Burlington: University of Vermont, Department of Psychiatry.

Battistich, V., Schaps, E., Watson, M., Solomon, D., & Lewis, C. (2000).Effects of the Child Development Project on students’ drug use andother problem behaviors. Journal of Primary Prevention, 21, 75–99.

Bryk, A. S., & Raudenbush, S. W. (1992). Hierarchical linear models.Newbury Park, CA: Sage.

Cohen, J., & Cohen, P. (1983). Applied multiple regression/correlationanalysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum.

Conduct Problems Prevention Research Group (1999). Initial impact of theFast Track prevention trial for conduct problems: I. The high-risksample. Journal of Consulting and Clinical Psychology, 67, 631–647.

Dahlberg, L. (1998). Youth violence in the United States: Major trends,risk factors and prevention approaches. American Journal of PreventiveMedicine, 14, 259–272.

Diggle, P. J., Liang, K., & Zeger, S. (1994). Analysis of longitudinal data.New York: Oxford University Press.

Dolan, L. J., Kellam, S. G., Brown, C. H., Werthamer-Larsson, L., Rebok,G. W., Mayer, L. S., et al. (1993). The short term impact of twoclassroom-based preventive interventions on aggressive and shy behav-iors and poor achievement. Journal of Applied Developmental Psychol-ogy, 14, 317–345.

Elliott, D., Hamburg, B., & Williams, K. (Eds.). (1998). Violence inAmerican schools: New perspectives and solutions. New York: Cam-bridge University Press.

Embry, D. D. (1980). Designing instructional materials for young children.New Directions for Exceptional Children, 3, 75–99.

Embry, D. D., & Flannery, D. J. (1999). Two sides of the coin: Multilevelprevention and intervention to reduce youth violent behavior. In D.Flannery & C. R. Huff (Eds.), Youth violence: Prevention, interventionand social policy (pp. 47–72). Washington, DC: American PsychiatricPress.

Embry, D. D., Flannery, D. J., Vazsonyi, A. T., Powell, K. E., & Atha, H.(1996). PeaceBuilders: A theoretically driven, school-based model forearly violence prevention. American Journal of Preventive Medicine,Supplement to 12(5), 91–100.

Englander-Golden, P., Jackson, J., Crane, K., Schwarzkopf, A., & Lyle, P.(1989). Communication skills and self-esteem in prevention of destruc-tive behavior. Adolescence, 14, 481–501.

Eron, L., & Huesmann, R. (1990). The stability of aggressive behavior—even unto the third generation. In M. Lewis & S. Miller (Eds.), Hand-book of developmental psychopathology (pp. 147–156). New York:Plenum Press.

Fagan, J., & Wilkinson, D. (1997). Firearms and youth violence. In D.Stoff, J. Breiling, & J. Maser (Eds.), Handbook of antisocial behavior(pp. 551–566). New York: Wiley.

Farrell, A., & Meyer, A. (1997). The effectiveness of a school-basedcurriculum for reducing violence among urban sixth-grade students.American Journal of Public Health, 87, 979–984.

Flannery, D. J. (1997). School violence: Risk, preventive interventions, andpolicy. New York: Columbia University and ERIC Clearinghouse onUrban Education.

Flannery, D. J. (2000, August). Longitudinal effectiveness of the Peace-Builders universal school-based violence prevention program. Paperpresented at the 108th Annual Convention of the American Psycholog-ical Association, Washington, DC.

Flannery, D., & Huff, C. R. (Eds.). (1999). Youth violence: Prevention,intervention and social policy. Washington, DC: American PsychiatricPress.

Flannery, D., & Seaman, M. (2001). Evaluating school and community-based violence prevention resource guides: What to look for, what to do.In E. Flaxman (Ed.), Evaluating school violence programs (UrbanDiversity Series No. 113, pp. 51–75). New York: Columbia University,Institute for Urban and Minority Education, and ERIC Clearinghouse onUrban Education.

Flannery, D., & Williams, L. (1999). Effective youth violence prevention.In T. Gullotta & S. McElhaney (Eds.), Violence in homes and commu-nities: Prevention, intervention and treatment (pp. 207–243). ThousandOaks, CA: Sage.

Gottfredson, D. (1997). School based crime prevention. In Preventingcrime: What works, what doesn’t, what’s promising (Report to the U.S.Congress prepared for the National Institute of Justice by L. Sherman, D.Gottfredson, D. MacKenzie, J. Eck, P. Reuter, & S. Bushway in collab-oration with members of the Graduate Program Department of Crimi-nology and Criminal Justice, University of Maryland, pp. 5-1–5-14).Washington, DC: U.S. Department of Justice, Office of Justice Pro-grams, National Institute of Justice.

Greenberg, M. T., Kusche, C. A., Cook, E. T., & Quamma, J. P. (1995).Promoting emotional competence in school-aged children: The effects ofthe PATHS curriculum. Development and Psychopathology, 7, 117–136.

Grossman, D., Neckerman, H., Koepsell, T., Liu, P. Y., Asher, K. N.,Beland, K., et al. (1997). Effectiveness of a violence prevention curric-ulum among children in elementary school: A randomized controlledtrial. Journal of the American Medical Association, 277, 1605–1611.

Guo, S., & Hussey, D. (1999). Analyzing longitudinal rating data: Athree-level hierarchical linear model. Social Work Research, 23, 258–269.

Hansen, W. B., Tobler, N. S., & Graham, J. W. (1990). Attrition insubstance abuse prevention research: A meta-analysis of 85longitudinally-followed cohorts. Evaluation Review, 14, 677–685.

306 FLANNERY ET AL.

Hawkins, J. D. (1995). Controlling crime before it happens: Risk-focusedprevention. National Institute of Justice Journal, 229, 10–18.

Hawkins, J. D., Catalano, R. F., Kosterman, R., Abbott, R., & Hill, K. G.(1999). Preventing adolescent health-risk behavior by strengtheningprotection during childhood. Archives of Pediatric Adolescent Medicine,153, 226–234.

Hawkins, J. D., Herrenkohl, T., Farrington, D., Brewer, D., Catalano, R.,Harachi, T., & Cothern, L. (2000, April). Predictors of youth violence(OJJDP Juvenile Justice Bulletin). Washington, DC: U.S. Department ofJustice, Office of Justice Programs, Office of Juvenile Justice andDelinquency Prevention.

Howard, K. A., Flora, J., & Griffin, M. (1999). Violence-preventionprograms in schools: State of the science and implications for futureresearch. Applied & Preventive Psychology, 8, 197–215.

Huesmann, R., Maxwell, C., Eron, L., Dahlberg, L., Guerra, N., Tolan, P.,et al. (1996). Evaluating a cognitive/ecological program for the preven-tion of aggression among urban children. American Journal of Preven-tive Medicine, Supplement to 12(5), 120–128.

Huesmann, R., & Moise, J. (1999). Stability and continuity of aggressionfrom early childhood to young adulthood. In D. Flannery & C. R. Huff(Eds.), Youth violence: Prevention, intervention and social policy (pp.73–98). Washington, DC: American Psychiatric Press.

Kachur, S. P., Stennies, G. M., Kenneth, E. P., Modzeleski, W., Stephen-son, R., Murphy, R., et al. (1996). School associated violent deaths in theUnited States, 1992 to 1994. Journal of the American Medical Associ-ation, 275, 1729–1733.

Kann, L., Kinchen, S. A., Williams, B. I., Ross, J. G., Lowry, R., Grun-baum, J. A., et al. (2000, June 9). Youth risk behavior surveillance—United States, 1999. Morbidity and Mortality Weekly Report Surveil-lance Summary (Centers for Disease Control), 49, 1–96.

Kaufman, P., Chen, X., Choy, S. P., Ruddy, S. A., Miller, A. K., Fleury,J. K., et al. (2000). Indicators of school crime and safety, 2000 (NCES2001–017/NCJ-184176). Washington, DC: U.S. Department of Educa-tion, National Center for Education Statistics and U.S. Department ofJustice, Bureau of Justice Statistics.

Kellam, S. G., & Anthony, J. C. (1998). Targeting early antecedents toprevent tobacco smoking: Findings from an epidemiologically basedrandomized field trial. American Journal of Public Health, 88, 1490–1495.

Kellam, S., Ling, X., Merisca, R., Hendricks, B. C., & Ialongo, N. (1998).The effects of the level of aggression in first grade classrooms on thecourse and malleability of aggressive behavior in middle school. Devel-opment and Psychopathology, 10, 165–185.

Kellam, S., Mayer, L., Rebok, G., & Hawkins, W. (1998). Effects ofimproving achievement on aggressive behavior and of improving ag-gressive behavior on achievement through two preventive interventions:An investigation of causal paths. In B. P. Dohrenwend (Ed.), Adversity,stress and psychopathology (pp. 486–505). New York: Oxford Univer-sity Press.

Koepke, M. D., & Flay, B. (1989). Levels of analysis. In M. T. Braverman(Ed.), New directions for program evaluation: Vol. 43. Evaluatinghealth promotion programs (pp. 75–87). San Francisco: Jossey-Bass.

Krug, E. G., Brener, N. D., Dahlberg, L. L., Ryan, G. W., & Powell, K. E.(1997). The impact of an elementary school-based violence preventionprogram on visits to the school nurse. American Journal of PreventiveMedicine, 13, 459–463.

Lindsey, J. K. (1993). Models for repeated measures. New York: OxfordUniversity Press.

Littell, R. C., Milliken, G. A., Stroup, W. W., & Wolfinger, R. D. (1996).SAS system for mixed models. Cary, NC: SAS Institute.

Lochman, J. E., & Dodge, K. A. (1994). Social-cognitive processes ofseverely violent, moderately aggressive, and nonaggressive boys. Jour-nal of Consulting and Clinical Psychology, 62, 366–374.

Loeber, R., & Farrington, D. P. (1998). Never too early, never too late:

Risk factors and successful interventions for serious and violent juvenileoffenders. Studies on Crime and Crime Prevention, 7, 7–30.

Mayer, G. R., Butterworth, T., Nafpaktitis, M., & Sulzer-Azaroff, B.(1983). Preventing school vandalism and improving discipline: A threeyear study. Applied Behavior Analysis, 16, 355–369.

Mayer, G. R., & Sulzer-Azaroff, B. (1990). Interventions for vandalism. InG. Stoner, M. R. Shinn, & H. M. Walker (Eds.), Interventions forachievement and behavior problems (pp. 559–580). Washington, DC:National Association of School Psychologists.

Mercy, J., & Potter, L. (1996). Combining analysis and action to solve theproblem of youth violence. American Journal of Preventive Medicine,Supplement to 12(5), 1–2.

Murray, D. M., & Wolfinger, R. D. (1994). Analysis issues in the evalu-ation of community trials: Progress toward solutions in SAS/STATMIXED. Journal of Community Psychology (Monograph Series, CSAPSpecial Issue), 140–155.

Neter, J., Wasserman, W., & Kutner, M. H. (1983). Applied linear regres-sion models. Homewood, IL: Richard Irwin.

O’Donnell, J., Hawkins, J. D., Catalano, R. F., Abbott, R. D., & Day, L.(1995). Preventing school failure, drug use, and delinquency amonglow-income children: Long-term intervention in elementary schools.American Journal of Orthopsychiatry, 65, 87–100.

Orpinas, P., Parcel, G. S., McAlister, A., & Frankowski, R. (1995).Violence prevention in middle schools: A pilot evaluation. Journal ofAdolescent Health, 17, 360–371.

Piper, D., Moberg, D. P., & King, M. (2000). The Healthy for Life Project:Behavioral outcomes. Journal of Primary Prevention, 21, 47–73.

Powell, K., & Hawkins, D. (Eds.). (1996). Youth violence prevention:Description and baseline data from 13 evaluation projects. AmericanJournal of Preventive Medicine, Supplement to 12(5), 1–134.

Reid, J., Eddy, J. M., Fetrow, R., & Stoolmiller, M. (1999). Description andimmediate impacts of a preventive intervention for conduct problems.American Journal of Community Psychology, 27, 483–517.

Rooney, B. L., & Murray, D. M. (1996). A meta-analysis of smokingprevention programs after adjustment for errors in the unit of analysis.Health Education Quarterly, 23, 48–64.

Rushforth, N. B., & Flannery, D. J. (1999). Role of firearms in youthviolence. In D. Flannery & C. R. Huff (Eds.), Youth violence: Preven-tion, intervention and social policy (pp. 201–230). Washington, DC:American Psychiatric Press.

Salend, S. J., Reynolds, C. J., & Coyle, E. M. (1989). Individualizing thegood behavior game across type and frequency of behavior with emo-tionally disturbed adolescents. Behavior Modification, 13, 108–126.

Satcher, D., Powell, K., Mercy, J., & Rosenberg, M. (1996). Violenceprevention is as American as apple pie. American Journal of PreventiveMedicine, Supplement to 12(5), v-vi.

Sherman, L. W., Gottfredson, D. C., MacKenzie, D. L., Eck, J., Reuter, P.,& Bushway, S. D. (1997). Preventing crime: What works, what doesn’t,what’s promising. A report to the United States Congress (NCJ 171676).Washington, DC: U. S. Department of Justice, Office of Justice Pro-grams, National Institute of Justice.

Sickmund, M., Snyder, H. N., & Poe-Yamagata, E. (1997). Juvenileoffenders and victims: 1997 update on violence. Washington, DC: U. S.Department of Justice, Office of Justice Programs, Office of JuvenileJustice and Delinquency Prevention.

Singer, M., Anglin, T. M., Song, L., & Lunghofer, L. (1995). Adolescents’exposure to violence and associated symptoms of psychological trauma.Journal of the American Medical Association, 273, 477–482.

Singer, M., & Flannery, D. J. (2000). The relationship between children’sthreats of violence and violent behaviors. Archives of Pediatrics andAdolescent Medicine, 154, 785–790.

Singer, M., Miller, D. B., Guo, S., Flannery, D. J., Frierson, T., & Slovak,K. (1999). Contributors to violent behavior among elementary andmiddle school children. Pediatrics, 104, 878–884.


Snijder, T., & Bosker, R. (1999). Multilevel analysis: An introduction tobasic and advanced multilevel modeling. Thousand Oaks, CA: Sage.

Snyder, H. N. (2000). Special analyses of FBI serious violent crimes data.Pittsburgh, PA: National Center for Juvenile Justice.

Snyder, H. N., & Sickmund, M. (1999). Juvenile offenders and victims:1999 national report. Washington, DC: U. S. Department of Justice,Office of Justice Programs, Office of Juvenile Justice and DelinquencyPrevention.

Stokes, T. F., & Baer, D. M. (1977). An implicit technology of general-ization. Applied Behavior Analysis, 12, 285–310.

Stoolmiller, M., Eddy, J. M., & Reid, J. (2000). Detecting and describingpreventive intervention effects in a universal school-based randomizedtrial targeting delinquent and violent behavior. Journal of Consultingand Clinical Psychology, 68, 296–306.

Thornton, T. N., Craft, C. A., Dahlberg, L. L., Lynch, B. S., & Baer, K.(2000). Best practices of youth violence prevention: A sourcebook forcommunity action. Atlanta, GA: Centers for Disease Control and Pre-vention, National Center for Injury Prevention and Control.

Tolan, P. H., & Gorman-Smith, D. (1998). Development of serious andviolent offending careers. In R. Loeber & D. Farrington (Eds.), Seriousand violent juvenile offenders: Risk factors and successful interventions(pp. 68–85). Thousand Oaks, CA: Sage.

Tolan, P. H., & Guerra, N. G. (1994). What works in reducing adolescentviolence: An empirical review of the field. Boulder, CO: Center for theStudy and Prevention of Violence.

Tolan, P. H., Guerra, N. G., & Kendall, P. C. (1995). A developmentalperspective on antisocial behavior in children and adolescents: Towarda unified risk and intervention framework. Journal of Consulting andClinical Psychology, 63, 579–584.

Tremblay, R., Masse, B., Perron, D., LeBlanc, M., Schwartzman, A., &Ledingham, J. (1992). Early disruptive behavior, poor school achieve-ment, delinquent behavior, and delinquent personality: Longitudinalanalyses. Journal of Consulting and Clinical Psychology, 60, 64–72.

Tremblay, R., McCord, J., Boileau, H., Charlebois, P., Gagnon, C., Le-blanc, M. & Larivee, S. (1991). Can disruptive boys be helped to become

competent? Psychiatry: Journal for the Study of Interpersonal Pro-cesses, 54, 148–161.

Tremblay, R. E., Pagani-Kurtz, L., Masse, L. C., Vitaro, F., & Pihl, R. O.(1995). A bimodal preventive intervention for disruptive kindergartenboys: Its impact through mid-adolescence. Journal of Consulting andClinical Psychology, 63, 560–568.

U. S. Department of Health and Human Services. (2001). Youth violence:A report of the Surgeon General (Report prepared by the U.S. Depart-ment of Health and Human Services for the Office of the SurgeonGeneral in partnership with the Centers for Disease Control, NationalCenter for Injury Prevention and Control; the National Institutes ofHealth, National Institute of Mental Health; and the Substance Abuseand Mental Health Services Administration, Center for Mental HealthServices). Rockville, MD: U. S. Public Health Service, Office of theSurgeon General.

Vazsonyi, A. T., Vesterdal, W. J., Flannery, D. J., & Belliston, L. (1999).The utility of child self reports and teacher ratings in classifying officialdelinquency status. Studies on Crime and Crime Prevention, 8(2), 225–244.

Walker, H. M., Colvin, G., & Ramsey, E. (1995). Anti-social behavior inschools: Strategies and best practices. Pacific Grove, CA: Brooks/Cole.

Walker, H. M., Irvin, L., Noell, J., & Singer, G. (1992). A construct scoreapproach to the assessment of social competence. Behavior Modifica-tion, 16, 448–474.

Walker, H. M., & McConnell, S. R. (1995). The Walker-McConnell Scaleof Social Competence and School Adjustment (SSCSA). Florence, Ken-tucky: Thomson Learning.

Webster, D. W. (1993). The unconvincing case for school-based conflictresolution programs for adolescents. Health Affairs, 12, 126–141.

Weissberg, R. P., & Bell, D. N. (1997). A meta-analytic review of primaryprevention programs for children and adolescents: Contributions andcaveats. American Journal of Community Psychology, 25, 207–214.

Received November 7, 2000Revision received April 22, 2002

Accepted May 3, 2002 �

308 FLANNERY ET AL.

http://www.paxis.org/campus/library

http://www.paxis.org/campus/library

Dr. Embry salutes these people and organizationsin Salinas who have made the success possible of his first generation work:

The children, teachers & families of SalinasAllan Styles, the former mayorAnna Cabrillio, current mayor The late Dr. Oscar Loya, superintendent of Alisal UnifiedSchool DistrictPatricia Skelton, Healthy Start Director, Salinas City UnifiedLupe Garcia, Partners for Peace City CoordinatorRev. Ken Feske, Partners for Peace, education coordinatorBob Rice, former general manager, KSBW-Channel 8Elizabeth Murdoch, former owner, KSBW-Channel 8Diana Jacobsen, Monterey Public Health DepartmentBill Deeb, former principal of Alisal Community SchoolJorge Rifa, Assistant City ManagerDan Nelson, Chief of Police"Sarge" Ernie Williams, Juvenile HallSalinas Valley Community HospitalNatividad County HospitalThe CalifornianLaurie Singer, mentor teacher Dena Ruiz Eastwood, former KSBW anchorAnd many others who have lived the peace builders pledge.

Dr. Embry honors these people and organizations in Tucson and nationally who have made the success possible of his first generation work:

The children, staff and families of Tucson Unified School DistrictsDr. Daniel Flannery, Kent State University, Kent, OhioDr. Alex Vazsonyi, Dept. of Psychology, Auburn UniversityDr. Ken Powell, formerly the science officer for the National Center for Injury Prevention, Centers for Disease ControlBarbara LaWall, The Pima County AttorneyThe Tucson Police Department & The Pima County SheriffDepartmentDr. Kris Boswoth, Smith Endowed Professor, College of Education, University of ArizonaDr. Linda Dahlberg, Jennifer Friday, Dr. Tom Simon, Jim Mercy,Dr. Mark Rosenberg, the National Center for Injury Prevention,Centers for Disease ControlJeff Sales & Mindy Blake, KOLD TV Channel 13, TucsonThe Arizona Daily Star and the Tucson CitizenCaptain Danny Sharp, Tucson Police DepartmentTom Rose, Hawaii, HawaiiDr. David Satcher, Surgeon General of the United StatesDr. George Comerci, past president of the American Academy of PediatricsThe scores of students and volunteers who collect data on one of the largest studies in the US on youth violenceInland Agency, Riverside, CA

peace builders research-combo

Documents

early violence prevention

schoolbased model

school nurse

use of peacebuilders

youth violence

consecutive school years

program articlesthe

major peacebuilders