• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Arch Pediatr Adolesc Med. Author manuscript; available in PMC May 15, 2008.
Published in final edited form as:
PMCID: PMC2386246
NIHMSID: NIHMS48877

Effects of 2 Prevention Programs on High-Risk Behaviors Among African American Youth

A Randomized Trial
Brian R. Flay, Dphil, Sally Graumlich, EdD, Eisuke Segawa, PhD, James L. Burns, MS, Michelle Y. Holliday, PhD, and for the Aban Aya Investigators

Abstract

Objective

To test the efficacy of 2 programs designed to reduce high-risk behaviors among inner-city African American youth.

Design

Cluster randomized trial.

Setting

Twelve metropolitan Chicago, Ill, schools and the communities they serve, 1994 through 1998.

Participants

Students in grades 5 through 8 and their parents and teachers.

Interventions

The social development curriculum (SDC) consisted of 16 to 21 lessons per year focusing on social competence skills necessary to manage situations in which high-risk behaviors occur. The school/community intervention (SCI) consisted of SDC and school-wide climate and parent and community components. The control group received an attention-placebo health enhancement curriculum (HEC) of equal intensity to the SDC focusing on nutrition, physical activity, and general health care.

Main Outcome Measures

Student self-reports of violence, provocative behavior, school delinquency, substance use, and sexual behaviors (intercourse and condom use).

Results

For boys, the SDC and SCI significantly reduced the rate of increase in violent behavior (by 35% and 47% compared with HEC, respectively), provoking behavior (41% and 59%), school delinquency (31% and 66%), drug use (32% and 34%), and recent sexual intercourse (44% and 65%), and improved the rate of increase in condom use (95% and 165%). The SCI was significantly more effective than the SDC for a combined behavioral measure (79% improvement vs 51%). There were no significant effects for girls.

Conclusions

Theoretically derived social-emotional programs that are culturally sensitive, developmentally appropriate, and offered in multiple grades can reduce multiple risk behaviors for inner-city African American boys in grades 5 through 8. The lack of effects for girls deserves further research.

Violence, substance use, and unsafe sexual practices are major public health problems challenging today’s urban African American youth.1,2 Urban African American youth are at high risk for violence owing to exposure to violence in their communities.36 They also experience more exposure, easy access, and daily pressure to use or traffic illicit drugs.79 Compared with white youth, African Americans are more likely to report earlier initiation of sex, higher lifetime rates of sexual intercourse, and more sexual partners in their lifetimes, with resulting high rates of pregnancy and human immunodeficiency virus infection.1012

Investigators have theorized that the seemingly separate behaviors of violence, substance use, delinquency, and risky sexual activity reflect an underlying “problem behavior” construct,1315 and empirical evidence increasingly supports this premise,16,17 regardless of ethnicity or race. Given the strong correlations among these behaviors and their predictors, prevention efforts may best be served by addressing multiple behaviors concurrently.13,18,19 Only a handful of interventions aimed at multiple behaviors have been tested,2023 and most have not used randomized designs. The current study was designed to overcome this methodological limitation and to meet recommendations for effective prevention programs.24

The Aban Aya Youth Project—which derives its name from 2 Ghanian symbols, aban, a fence signifying double (social) protection, and aya, an unfurling fern signifying self-determination—compared 3 interventions that were implemented in grades 5 through 8. Two experimental interventions (one a classroom-based curriculum and one that also included school and community-wide components) targeted the risk behaviors of violence, provoking behavior, substance use, school delinquency, and sexual practices (engaging in sexual intercourse and using condoms). The control program targeted health-enhancing behaviors (nutrition, exercise, and health care) and was of equal length and intensity. We hypothesized that both experimental conditions would result in reductions in the rate of increase of targeted behaviors compared with the control condition.

METHODS

SCHOOL SELECTION AND RANDOMIZATION

The longitudinal trial of 3 interventions was conducted in a high-risk sample of 12 poor, African American metropolitan Chicago, Ill, schools (9 inner-city and 3 near-suburban) between 1994 and 1998. School inclusion criteria included enrollment of greater than 80% African American and less than 10% Latino or Hispanic students; grades kindergarten through 8 (or through 6 if students were tracked to 1 middle school); enrollment greater than 500; not on probation or slated for reorganization; not a special designated school (ie, magnet, academic center); and moderate mobility (<50% annual turnover, meaning approximately <25% transferred in and <25% transferred out). Eligible schools (n=141 inner-city and 14 near-suburban) were stratified into 4 quartiles of risk on the basis of a score that combined proxy risk variables using the procedures described by Graham et al.25 The proxies of risk came from school report card data (1991–1992) and included enrollment, attendance and truancy, mobility, family income, and achievement scores. Using a randomized block design, we assigned to each condition 2 inner-city schools from the middle of the highest risk quartile, 1 inner-city school from the middle of the second risk quartile, and 1 near-suburban school (also from the second quartile) per condition. One inner-city school refused to participate and was replaced with one from the same risk level. Schools signed an agreement to participate in the study for 4 years and agreed not to participate in another prevention initiative during that time. Study schools were 91% African American. Each school received the intervention free of charge (provided to all students in the appropriate grade levels) plus $250 for each participating classroom up to a maximum of $1000 each year of the study.

PARTICIPANTS

Participants were students in fifth-grade classes in the 12 schools during the 1994–1995 school year or who transferred in during the study; students who transferred out were not followed up, but their data from the times before they transferred out were included in the analysis sample. Students who transferred into study schools were similar to students who transferred out; both groups were more likely to engage in risky behavior than students who stayed in the same school for the duration of the project (significant only for violence and substance use). Parents or legal guardians were informed of the study and procedures and were provided with an opportunity to opt out in grades 5 through 7 and then again in grade 8. Less than 1% of parents denied consent during grades 5 through 7 and 1.7% did so at grade 8. The University of Illinois at Chicago Institutional Review Board approved the research protocol and informed consent procedures.

INTERVENTIONS

The conceptual framework of the experimental interventions was derived from established theories of behavior change14 to focus the interventions on risk and protective factors and skills related to the targeted behaviors. Activities and materials were either developed de novo or adapted from other theoretically derived prevention curricula (eg, New Haven Social Development Program,26 Youth AIDS Prevention Project,27 and Know Your Body28). New or adapted activities were piloted before being added to the curricula, and each grade-level curriculum was piloted the year before its use in the main study. As a result of piloting, minor changes were made to improve flow or language.

Studies suggest that programs for African American youth should incorporate components that (1) enhance growth of sense of self and cultural pride and (2) strengthen family and community ties.29,30 Hence, the interventions included the Nguzo Saba principles,31 which promote African American cultural values such as unity, self-determination, and responsibility; culturally based teaching methods32 (eg, storytelling and proverbs) and African and African American history and literature; and homework assignments that involved parents to encourage review and generalization of the information and skills and to expand the target of the intervention to parents.33

The 2 experimental conditions were the social development curriculum (SDC) and the school/community intervention (SCI) (Table 1). The SDC was classroom based, consisting of 16 to 21 lessons per year in grades 5 through 8. The SDC was designed to teach cognitive-behavioral skills to build self-esteem and empathy, manage stress and anxiety, develop interpersonal relationships, resist peer pressure, and develop decision-making, problem-solving, conflict-resolution, and goal-setting skills. It was structured to teach application of these skills to avoid violence, provocative behavior, school delinquency, drug use, and unsafe sexual behaviors.

Table 1
Intervention Content by Condition*

The SCI included the SDC plus parental support, school climate, and community components to impact all social domains of influence on children.34,35 The parent support program reinforced skills and promoted child-parent communication. The school staff and school-wide youth support programs integrated skills into the school environment. The community program forged linkages among parents, schools, and local businesses. Each SCI school formed a local school task force consisting of school personnel, students, parents, community advocates, and project staff to implement the program components,36 propose changes in school policy, develop other school-community liaisons supportive of school-based efforts, and solicit community organizations to conduct activities to support the SCI efforts. A goal of these linkages was to “rebuild the village” and create a “sense of ownership” by all stakeholders to promote sustainability of these efforts on completion of the project.37

The control condition was the health enhancement curriculum (HEC). It consisted of the same number of lessons as the SDC and taught some of the same skills (eg, decision making and problem solving), but with a focus on promoting healthy behaviors related to nutrition, physical activity, and general health care (see Table 1). It also integrated the importance of cultural pride and communalism.

STAFF TRAINING AND IMPLEMENTATION

University-based health educators delivered curricula in all 3 conditions, usually in social studies classes. Each health educator delivered 1 of the curricula to 1 or more schools and, in most cases, health educators stayed with the same school from year to year. This avoided contamination across conditions and developed continuity in relationships between educators and the schools and students. To ensure fidelity of implementation across health educators, experimental conditions, and times, 2 training sessions were held before each lesson. The health educators role-played each activity and senior staff provided feedback. Weekly debriefings were held to discuss issues that may have affected implementation. Senior staff also conducted observations to ensure fidelity and help target training needs. In addition, each year the regular classroom teachers received a 4-hour workshop to provide an overview of program philosophy, curriculum content, and clarification of their support roles.

ASSESSMENT

Constructs were derived from the theory of triadic influence14 and program content, and included background covariates, process variables, mediating variables, and behaviors. Only student self-reports of behaviors (violence, provoking behaviors, school delinquency, substance use, and sexual behaviors) are reported in this article. Measures were based on instruments previously used with inner-city populations.20,27,3841 Survey questions were modified for grade 4 readability and cultural sensitivity by means of feedback from focus groups and piloting.

The items, response categories, scale score ranges, and reliability coefficients of each behavioral scale at each grade are available from one of us (B.R.F.). Violence, school delinquency, and substance use were measured from grade 5 onward; provoking behaviors, recent sexual intercourse, and condom use were added at grade 6. Each behavior was assessed with multiple items. For violence, provoking behavior, and substance use, scale scores were formed for each behavior by summing multiple items. For sexual behaviors (having sexual intercourse and use of condoms), single item scores were used. For school delinquency, a more complicated approach was necessary to produce a “scale score” because of “planned missingness.”42 To reduce respondent burden, starting in the spring of fifth grade, 3 versions of the survey, each containing the core and 2 of the modules, were randomly assigned to classrooms (evenly distributed across the 3 interventions) at each wave of data collection. The core unit, answered by every student, included items assessing demographics and all of the behavioral outcomes except school delinquency. Each of the 3 modules contained two thirds of the items from the measure of school delinquency. The scale and change scores were computed by fitting growth curves to each item simultaneously by means of mixed-effect models and summing them to form the intercept (baseline score) and growth (change) of delinquency. We created a combined behavior measure by adjusting the range of the variables to be the same (0–10) and reversing the direction of scoring for condom use.

DATA COLLECTION

Students completed surveys in classrooms at the beginning and end of grade 5 and at the end of each subsequent year. We took several precautions to ensure the validity of the data. To ensure even completion, staff read the survey aloud to students. To minimize underreporting of behaviors, trained project staff, not the teacher or health educator assigned to that classroom, administered the surveys. To emphasize the confidential nature of their answers, we assured students that results would not be shared with anyone and we used identification numbers rather than names to track students over time. Students without consent completed teacher-assigned tasks during survey administration.

ANALYTICAL METHODS

To estimate mean responses at baseline and in response to the program, we used hierarchical statistical models that accommodate nested observations (times within subjects, subjects within schools) and missing data.4345 For the major reported analyses, we included all students who provided one or more waves of data.

We used mixed models for continuous outcomes (violence, provoking behavior, school delinquency, and the combined behavior) and generalized estimating equations for ordinal outcomes (substance use, sexual activity, and condom use). We present 2-level models throughout, as school effects proved negligible in 3-level models for continuous outcomes (and the pattern of results were the same) and 3-level software for ordinal outcomes is not available. All models included terms for condition, sex, time (quadratic trends where necessary), and all interactions, except for condom use, which was estimated separately for boys (because of low rates of sexual intercourse for girls). Inference was based on tests of regression coefficients and contrasts among estimated means. Contrasts were used to test baseline differences between boys and girls and between conditions (HEC, SDC, and SCI), change from baseline to end point within condition, and differences between conditions in the amount of change, or program effects. All statistical tests are 2-tailed.

RESULTS

We first describe our sample, then baseline differences by sex, and finally program effects. Survey completion rates were 93.2% of students with consent at baseline, and between 89.5% and 92.7% at the other waves. Non-completions were due primarily to school absenteeism (4%–9%) or opting out. An average of 20% turnover occurred each year, resulting in an average sample of 644 students (range, 597–674) at each wave, with 339 (51%) of the 668 original grade 5 students still present at the end of grade 8 and a total analysis sample (students with one or more waves of data) of 1153. The final sample was 49.5% male, with an average age of 10.8 years (SD, 0.6 year) at the beginning of grade 5; approximately 77% received federally subsidized school lunches, and 47% lived in 2-parent households.

Table 2 shows baseline (grade 5 or grade 6), end point (grade 8), and change in scale scores or proportions engaging in behaviors, percentage relative reductions, significance levels, and effect sizes by condition for boys and girls. Boys engaged in higher levels of behaviors at baseline than girls for all behaviors (P<.001) except provoking (P=.17). The prevalence of all behaviors increased over time across sex and conditions. There was one significant baseline (grade 5) difference between conditions: boys receiving the SCI engaged in more violence than boys in the SDC (P=.02).

Table 2
Baseline, End Point, and Growth of Risk Behaviors With P Values From Growth Curve Analysis by Condition and Sex

There were no significant program effects for girls. Program effects for boys were significant for all 6 behaviors in the SCI and marginally so in the SDC (except for condom use); boys receiving the SDC and SCI increased these behaviors less (more for condom use) than boys in the HEC. The Figure shows the developmental pattern of behavior change and program effects for school delinquency. It exemplifies the nature of program effects for boys, occurring gradually between grades 6 and 7.

Figure
Changes in school delinquency for boys and girls by condition from the beginning of grade 5 to the end of grade 8. Baseline intercepts shown are the average of the 3 conditions for each sex. HEC indicates health enhancement curriculum; SDC, social development ...

Effect sizes for the comparison of SDC and SCI with HEC for boys ranged from 0.29 to 0.66, and relative improvements were 31% to 165%. For boys in the SDC and SCI, the increase in negative behaviors from fifth to eighth grade was less than in the HEC: violence by 35% and 47%, respectively; provoking behavior, 41% and 59%; school delinquency, 31% and 66%; drug use, 32% and 34%; and recent sexual intercourse, 44% and 65%. The relative improvement in the rate of condom use was 95% and 165%. The effect sizes for the combined behavior score were 0.52 for the SDC and 0.82 for the SCI, and the relative improvements were 51% and 79%.

In addition, all 6 behaviors increased less (more for condom use) for boys in the SCI than for boys in the SDC. This difference was significant for the combined behavior measure (mean difference in change, −3.35; effect size, 0.82 vs 0.52; or 79% vs 51% relative improvement) but only for one of the individual behaviors (school delinquency: effect size, 0.61 for the SCI and 0.29 for the SDC; relative reductions of 66% and 31%, respectively).

COMMENT

This randomized controlled study provides evidence that a prevention program that teaches skills and is theoretically derived, developmentally appropriate, and culturally sensitive can have concurrent effects on multiple risk behaviors for inner-city African American boys in grades 5 through 8. The effect sizes for violence (0.31 and 0.41) and substance use (0.42 and 0.45) are substantially better than those reported in meta-analyses for interactive school-based violence (0.16),23 drug (0.24),46 sex (0.05),47 and other problem behavior (0.16)48 prevention programs that address only 1 behavioral domain. Schools and communities should be encouraged to adopt programs that have effects on multiple outcomes. Public pressure on schools has resulted in school systems being mandated or expected to provide multiple prevention programs. Adoption of one effective multiple-behavior program would reduce the costs and burdens on school personnel. It may also lead to reduced school dropout rates and improved learning.13

Previous studies suggest that comprehensive programs that address multiple behaviors (like the SDC) and involve families and the community (like the SCI) are generally more effective than programs that address single behaviors or do not involve families or community.13,36,49,50 Both programs significantly reduced the rate of increase of multiple risk behaviors for boys. The significantly larger effect found for SCI in the combined behaviors analysis (and the generally larger effect sizes for SCI) suggest that the SCI may be even more effective than the SDC in reducing the targeted behaviors.

The effects of our programs may be underestimated because of the design of our control condition. We wanted to design a placebo-attention condition that would involve providing the same amount of attention to students as the SDC and be seen as equally interesting, engaging, and helpful by students. We probably erred on the side of making the HEC too similar to the SDC, in that the HEC included some of the same skills as the SDC, but with a focus on different behaviors. This means that the HEC might have been more effective than a standard placebo-attention condition or than “standard care” in most schools. If so, this would mean that our reported results underestimate the actual effectiveness of the SDC and SCI. It might also partially explain the lack of effects detected for girls.

Our program effects are of practical significance for public health and education. From a public health perspective, reducing these risk behaviors can decrease morbidity and mortality related to these behaviors. For example, a reduction in the use or carrying of weapons not only can prevent homicides, the leading cause of death for young African American males, but also can help decrease other crimes that impact African American communities. Reduced drug use and safer sexual practices can diminish the substantial morbidity and social problems associated with human immunodeficiency virus infection, unintended pregnancy, and sexually transmitted diseases.

Because the incidence of all measured risk behaviors increased for girls, and no program effects were found, an obvious question is why. Others have also reported sex-specific results for these behaviors.5154 One possibility is that the targeted behaviors are more difficult to reduce among girls because they already occur at lower levels. The fact that the interventions reduced the frequency of the targeted behaviors for boys down to the levels for girls in some cases provides some support for this possibility. Nevertheless, the fact that the incidence of these behaviors increased for girls is still of concern.

In 2 previous studies, differential effects may have resulted from program implementers being male; they may have served as more effective role models for boys than girls.51,54 Findings from one review55 suggested that programs that provide positive female role models might improve intervention effects for girls. However, in our study, about equal numbers of classes were taught by men and women. Nevertheless, given the relative lack of male teachers in public schools, it is still possible that our male health educators contributed to the observed effects for boys.

Another possible explanation for the lack of female effects may be that, like at least one other intervention,51 the SDC did not address the types of aggressive behaviors used more by girls, ie, indirect aggressive behaviors, such as spreading rumors, and creating friendship alliances for the purpose of revenge.56 In addition, the SDC did not take into consideration the functions that violence may provide for girls in high-risk environments (ie, presenting a tough persona for protection).57

Our study supports evidence that the dominant prevention strategies may work better for boys than girls.58 Although research focusing on sex differences is sparse,58 current research is identifying factors that may enhance prevention strategies for young girls. This literature suggests that to be effective for girls, programs may need to focus more on internal manifestation of risks and on connectedness to school and family.54,59 In addition, further studies are warranted to help prevention researchers better understand when and how risk factors come into play at the various stages of female development so that programs can address these crucial variables.

A major strength of the SCI program was the strong partnership that was developed with community organizations, including a community-based mental health organization. All stakeholders, including academia, the schools, and their communities, had very different strengths and weaknesses that provided challenges as well as opportunities.59 The community mental health organization was instrumental in developing collaborative relationships and facilitating implementation of SCI components. However, it is not clear that this would be easily replicated in other communities because of the amount of coordination required, or worth the additional effort for what appears to be marginal improvement.

Some limitations of this study need to be noted. First, the number of schools was small. This leads to low statistical power to detect small differences, especially between the 2 intervention conditions. However, the significant effects have clear practical public health and educational relevance and application.

Second, as expected with a high-risk sample, student turnover in study schools was relatively high. This led us to adopt program and analytical strategies that included all students for whom we had 1 or more waves of data and who received at least some of the program, regardless of how much of the program they received. The idea was that the intervention would diffuse throughout the grade level and affect all students in that cohort. The curriculum was designed so that appropriate review and sequencing of content allowed new students to “catch up” reasonably well. Our results suggest that this approach was successful and needs to be adopted by future prevention studies in high-risk schools and communities.

The Aban Aya Investigators

Brian R. Flay, DPhil, principal investigator, Public Health and Psychology, University of Illinois at Chicago; Shaffdeen A. Amuwo, PhD, School of Public Health; Carl C. Bell, MD, Psychiatry and Public Health, and Community Mental Health Council; Michael L. Berbaum, PhD, Methodology Research Core, Health Research and Policy Centers; Richard T. Campbell, PhD, Sociology, and Methodology Research Core, Health Research and Policy Centers; Julia Cowell, RN, PhD, Nursing (now at Rush University, Chicago); Judith Cooksey, MD, Public Health (now at University of Maryland, College Park); Barbara L. Dancy, PhD, Nursing; Sally Graumlich, EdD, Health Research and Policy Centers; Donald Hedeker, PhD, Biostatistics, Public Health, and Methodology Research Core, Health Research and Policy Centers; Robert J. Jagers, PhD, African American Studies and Psychology (now at Morgan State University, Baltimore, Md); Susan R. Levy, PhD, Public Health; Roberta L. Paikoff, PhD, Psychiatry; Indru Punwani, DDS, Pediatric Dentistry; Roger P. Weissberg, PhD, Psychology.

What This Study Adds

Most school-based prevention programs are of short duration and address only one behavioral domain (eg, substance use) or one behavior (eg, smoking). High correlations among risky behaviors suggest the need for multibehavior programs, but few have been developed and even fewer have been tested in randomized trials. Risky behaviors are particularly problematic for African American youth; however, few school-based prevention interventions have been developed for them.

The Aban Aya Youth Project developed a culturally sensitive classroom curriculum and community program for inner-city African American students (grades 5–8) that targeted multiple risky behaviors (violence, provocative behavior, school delinquency, substance use, and sexual behavior). This study evaluated the curriculum and a combined curriculum plus community intervention in a school-based randomized trial. Results demonstrate that a single curriculum or intervention can have large effects on multiple behaviors, at least for boys, reducing their risky behavior to the levels observed in girls by the end of grade 8. A lack of effects for girls replicates other investigators’ findings, suggesting an area for new research.

Further analyses are needed to determine whether the interventions enhanced student bonding with their parents, connection with their heritage, and attachment to their school and community. Analyses are also needed to explore the role of mediators (eg, intentions and attitudes) in reducing the growth of problem behaviors in African American boys. Finally, further research is needed on why programs like this are ineffective for girls.

Acknowledgments

We thank the principals, teachers, students, and parents of the Chicago-area schools that participated in the Aban Aya Youth Project. Without their willing participation, this project would not have been possible. We also thank the Community Mental Health Council/Foundation and the many health education, evaluation, data management (with special thanks to Ling Zou Gurnack, MSc), and other staff who worked on the project during the 10 years of its conception, implementation, and reporting.

This study was one of several funded in 1992 by the Office for Research on Minority Health to conduct research on the prevention of violence, unsafe sex, and drug use among minority students, administered by the National Institute for Child Health and Human Development, Bethesda, Md, grant U01HD30078 (1992–1997). Grade 8 data collection and statistical analyses were funded by grant R01DA11019 from the National Institute on Drug Abuse, Bethesda (1998–2003).

References

1. Herrenkohl T, Maguin E, Hill K, Hawkins JD, Abbott RD, Catalano RF. Developmental risk factors for youth violence. J Adolesc Health. 2000;26:176–186. [PubMed]
2. Dahlberg L. Youth violence in the United States: major trends, risk factors, and prevention approaches. Am J Prev Med. 1998;14:259–272. [PubMed]
3. Bell C, Jenkins E. Community violence and children on Chicago’s southside. Psychiatry. 1993;56:46–54. [PubMed]
4. Garbarino J, Kostelny K, Dubrow N. What children can tell us about living in danger. Am Psychol. 1991;46:376–383. [PubMed]
5. Shakoor BH, Chalmers D. Co-victimization of African-American children who witness violence: effects on cognitive, emotional and behavioral development. J Natl Med Assoc. 1991;83:233–238. [PMC free article] [PubMed]
6. Jenkins E, Bell C. Violence among inner city high school students and post-traumatic stress disorder. In: Friedman S, editor. Anxiety Disorders in African Americans. New York, NY: Springer Publishing Co; 1994. pp. 76–88.
7. Kann L, Kinchen SA, Williams BI, et al. Youth risk behavior surveillance—United States, 1999. MMWR CDC Surveill Summ. 2000;49(5):1–32. [PubMed]
8. Kosterman R, Hawkins DJ, Guo J, Catalano RF, Abbott RD. The dynamics of alcohol and marijuana initiation: patterns and predictors of first use in adolescence. Am J Public Health. 2000;90:360–366. [PMC free article] [PubMed]
9. Cherry V, Belgrave FZ, Jones W, et al. NTU: an Africentric approach to substance abuse prevention among African American youth. J Prim Prev. 1998;18:319–339.
10. Division of STD Prevention. Tracking the Hidden Epidemics: Trends in STDs in the United States. Atlanta, Ga: Centers for Disease Control and Prevention; 2000. Available at: http://www.cdc.gov/nchstp/dstd/Stats_Trends/Trends2000.pdf.
11. Centers for Disease Control and Prevention. Trends in the HIV & AIDS Epidemic. Atlanta, Ga: Centers for Disease Control and Prevention; 1998. Available at: http://www.cdc.gov/nchstp/od/Trends.htm.
12. Centers for Disease Control and Prevention. National and state-specific pregnancy rates among adolescents—United States, 1995–1997 [published correction appears in MMWR Morb Mortal Wkly Rep. 2000;49:672] MMWR Morb Mortal Wkly Rep. 2000;49:605–611. [PubMed]
13. Flay BR. Positive youth development requires comprehensive health promotion programs. Am J Health Behav. 2002;26:407–424. [PubMed]
14. Flay BR, Petraitis J. The theory of triadic influence: a new theory of health behavior with implications for preventive interventions. In: Albrecht GS, editor. Advances in Medical Sociology, Vol IV: A Reconsideration of Models of Health Behavior Change. Greenwich, Conn: JAI Press; 1994. pp. 19–44.
15. Jessor R, Jessor S. Problem Behavior and Psychosocial Development. New York, NY: Academic Press; 1977. p. 281.
16. Donovan JE. Problem-behavior theory and the explanation of adolescent marijuana use. J Drug Issues. 1996;2:379–402.
17. Resnick MD, Bearman P, Blum RW, et al. Protecting adolescents from harm: findings from the National Longitudinal Study on Adolescent Health. JAMA. 1997;278:823–832. [PubMed]
18. Allensworth D, Lawson E, Wyche J. Schools and Health: Our Nation’s Investment. Washington, DC: National Academy Press; 1997. p. 498.
19. Kolbe LJ. An essential strategy to improve health and education of Americans. Prev Med. 1993;22:544–560. [PubMed]
20. Hawkins JD, Catalano RF, Morrison DM, O’Donnell J, Abbott R, Day LE. The Seattle Social Development Project: effects of the first four years on protective factors and problem behaviors. In: McCord J, Tremblay R, editors. Preventing AntiSocial Behavior: Interventions From Birth Through Adolescence. New York, NY: Guilford Press; 1992. pp. 139–161.
21. Piper DL, Moberg D, King R. The Healthy for Life Project: behavioral outcomes. J Prim Prev. 2000;21:47–73.
22. Flay BR, Allred CG, Ordway N. Effects of the Positive Action program on achievement and discipline: two matched-control comparisons. Prev Sci. 2001;2:71–89. [PubMed]
23. Derzon J, Wilson SJ, Cunningham CA. The Effectiveness of School-Based Interventions for Preventing and Reducing Violence: Center for Evaluation Research Methodology. Nashville, Tenn: Center for Evaluation Research Methodology, Vanderbilt University; 1999.
24. Carnegie Council on Adolescent Development. Turning Points: Preparing American Youth for the 21st Century: Report of the Task Force on Education of Young Adolescents. Washington, DC: Carnegie Council on Adolescent Development; Jun, 1989.
25. Graham J, Flay B, Johnson CA, et al. Group comparability: a multi attribute utility measurement approach to the use of random assignment with small numbers of aggregated units. Eval Rev. 1984;8:247–260.
26. Weissberg RP, Barton HA, Shriver TP. The Social Competence Promotion program for young adolescents. In: Albee GW, Gullotta TP, editors. Primary Prevention Works. Thousand Oaks, Calif: Sage Publications; 1997. pp. 268–290.
27. Levy S, Perhats C, Weeks K, Handler AS, Zhu C, Flay BR. Impact of a school-based AIDS prevention program on risk and protective behavior for newly sexually active students. J Sch Health. 1995;65:145–151. [PubMed]
28. Williams C, Carter BJ, Eng A. The “Know Your Body” program: a developmental approach to health education and disease prevention. Prev Med. 1980;9:371–383. [PubMed]
29. Hilliard AG., III . Reintegration for education: black community involvement with black students in schools. In: Smith WD, Chunn EW, editors. Black Education: A Quest for Equity and Excellence. New Brunswick, NJ: Transaction Publishers; 1989. pp. 201–208.
30. Hammond WR, Yung B. Psychology’s role in the public health response to assaultive violence among young African-American men. Am Psychol. 1993;48:142–154. [PubMed]
31. Karenga M. The African American Holiday of Kwanzaa: A Celebration of Family, Community and Culture. Los Angeles, Calif: University of Sankore Press; 1988. p. 143.
32. Boykin AW, Ellison C. The multiple ecologies of black youth socialization: an Afro-graphic analysis. In: Taylor R, editor. African American Youth Socialization: Their Social and Economic Status in the United States. Westport, Conn: Praeger Publishing; 1995. pp. 83–128.
33. Sussman S. Curriculum development in school-based prevention research. Health Educ Res. 1991;6:339–351.
34. Friedman J. Empowerment: The Politics of Alternative Development. Cambridge, Mass: Blackwell Scientific Publications; 1992. p. 196.
35. Fawcett SB, Paine AL, Francisco VT, et al. Promoting health through community development. In: Glenwick DS, Jason LA, editors. Promoting Health and Mental Health in Children, Youth and Families. New York, NY: Springer Publishing Corp; 1993. pp. 233–255.
36. Comer J. Educating poor minority children. Sci Am. 1988;259:42–48.
37. Bell C, Fink P. Psychiatric Aspects of Violence: Understanding Causes and Issues in Prevention and Treatment. San Francisco, Calif: Jossey-Bass Publishers; 2000. pp. 37–47.
38. Graham JW, Flay BR, Johnson CA, Hansen WB, Grossman L, Sobel JL. Reliability of self-report measures of drug use in prevention research: evaluation of the Project SMART questionnaire via the test-retest reliability matrix. J Drug Educ. 1984;14:175–193. [PubMed]
39. Hansen W, Johnson AC, Flay B, Graham JW, Sobel J. Affective and social influences approaches to the prevention of multiple substance abuse among seventh grade students: results from project SMART. Prev Med. 1988;17:135–154. [PubMed]
40. Pentz MA, Dwyer JH, MacKinnon DP, et al. A multicommunity trial for primary prevention of adolescent drug abuse: effects on drug use prevalence. JAMA. 1989;261:3259–3266. [PubMed]
41. Uehara E, Chalmers D, Jenkins EJ, Shakoor B. Youth encounters with violence: results from the Chicago Community Mental Health Council Violence Screening Project. J Black Stud. 1996;26:768–781.
42. Graham JW, Hofer SM, MacKinnon DP. Maximizing the usefulness of data obtained with planned missing value patterns: an application of maximum likelihood procedures. Multivariate Behav Res. 1996;31:197–218.
43. Bryk A, Raudenbush SW. Hierarchical Linear Models. Newbury Park, Calif: Sage Publications; 1992. p. 265.
44. Diggle PJ, Liang K, Zeger SL. Analysis of Longitudinal Data. New York, NY: Oxford University Press; 1994.
45. Schafer JL, Graham JW. Missing data: our view of the state of the art. Psychol Methods. 2002;7:147–177. [PubMed]
46. Tolber N, Stratton HH. Effectiveness of school-based drug prevention programs: a meta-analysis of the research. J Prim Prev. 1997;10:71–128.
47. Johnson BT, Carey MP, Marsh KL, Levin KD, Scott-Sheldon LA. Interventions to reduce sexual risk for the human immunodeficiency virus in adolescents, 1985–2000: a research synthesis. Arch Pediatr Adolesc Med. 2003;157:381–388. [PubMed]
48. Wilson DB, Gottfredson DC, Najaka SS. School-based prevention of problem behaviors: a meta-analysis. J Quant Criminol. 2001;17:247–272.
49. Elliot DS, Tolan P. Youth violence prevention, intervention and social policy: an overview. In: Flannery DJ, Huff CR, editors. Youth Violence, Intervention and Social Policy. Washington, DC: American Psychiatric Press; 1999. pp. 3–46.
50. Hawkins JD, Catalano RF, Kosterman R, Abbott R, Hill KG. Preventing adolescent health-risk behaviors by strengthening protection during childhood. Arch Pediatr Adolesc Med. 1999;153:226–234. [PubMed]
51. Farrell A, Meyer A. The effectiveness of a school-based curriculum for reducing violence among urban sixth-grade students. Am J Public Health. 1997;87:979–984. [PMC free article] [PubMed]
52. O’Donnell J, Hawkins DJ, Catalano RF, Abbott RD, Day LE. Preventing school failure, drug use, and delinquency among low-income children: long-term intervention in elementary schools. Am J Orthopsychiatry. 1995;65:87–100. [PubMed]
53. The National Site Evaluation of High-Risk Youth Programs: Making Prevention Effective for Adolescent Boys and Girls: Gender Differences in Substance Use and Prevention. Rockville, Md: Substance Abuse and Mental Health Services Administration; 2002. Monograph Series No. 4. DHHS publication SMAA-003375.
54. Perry CL, Komro KA, Veblen-Mortenson S, et al. A randomized controlled trial of the middle and junior high school D.A.R.E. and D.A.R.E. Plus programs. Arch Pediatr Adolesc Med. 2003;157:178–184. [PubMed]
55. Blake SM, Amaro H, Schwartz P, Flinchbaugh LJ. A review of substance abuse prevention interventions for young adolescent girls. J Early Adolesc. 2001;21:294–324.
56. Lagerspetz K, Bjorkqvist K. Indirect aggression in boys and girls. In: Huessman L, editor. Aggressive Behavior. New York, NY: Plenum Press; 1994. pp. 131–150.
57. Campbell A. Men, Women and Aggression. New York, NY: Basic Books; 1993.
58. National Center on Addiction and Substance Abuse. The Formative Years: Pathways to Substance Abuse Among Girls and Young Women Ages 8–22. New York, NY: Columbia University; 2003.
59. Tolan P, Keys C, Chertok F, Jason L. Researching Community Psychology: Issues of Theory and Methods. Washington, DC: American Psychological Association; 1990.
PubReader format: click here to try

Formats: