Program

Sep 04, 2012

OVERVIEW

The Aban Aya Youth Project (AAYP), the name of which reflects the Ghanaian term for “protection and self-determination,” is a program designed to reduce rates of risky behavior among African-American children in 5th through 8th grades. In an evaluation conducted by AAYP and other researchers, 12 schools were randomly assigned to one of three conditions – a classroom curriculum, curriculum plus school and community-level interventions, or a control group that received a health-oriented intervention. The study found no impacts for girls. For boys, the school/community intervention resulted in significant impacts for all outcomes, including self-reported violence, provoking behavior, school delinquency, substance use, sexual activity, and condom use. Similar but statistically marginal impacts were found for boys in the schools that received only the classroom curriculum. Another analysis identified a significant reduction in the decline of empathy among students in the treatment groups, compared to the control group. Additional analyses examining subgroup differences found that, compared with the control group, boys in the combined school/community and classroom-curriculum-only groups reported less violent behavior, a reduction in the rate of growth of violent behavior, and a reduction in the rate of growth of substance use behavior. A different sub-group analysis showed that while the school/community intervention had a significant impact on the violent behavior of older children only.

DESCRIPTION OF PROGRAM

Target population: African-American students in metropolitan schools at high-risk for violence

The Aban Aya Youth Project (AAYP) was developed to compare three curriculum-based interventions designed to reduce the rate of growth of violence among African-American adolescents from grades 5 through 8. The interventions include principles that promote African American cultural values such as unity, self-determination, and responsibility, as well as culturally-based teaching methods, and inclusion of African American history and literature. All interventions are taught by trained university-based health educators.

The social development curriculum (SDC) focuses on reducing risky behaviors, such as violence, substance abuse, and unsafe sexual practices. The program teaches cognitive-behavioral skills, in order 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. The curriculum consists of 16 to 21 lessons per year from 5th through 8th grade.

The school/community intervention (SCI) includes the SDC plus parental support and school climate and community components. Each participating school forms a local task force, consisting of school representatives, students, parents, community members, and project staff, which proposes changes in school policy, develops school-community collaborations, and conducts program activities. Child-parent communication is promoted through the parental support element of the program, which provides training for parents. Linkages between parents, schools, and local businesses were formed.

The health enhancement curriculum (HEC) serves as the control condition and includes the same number of lessons as the SDC. The curriculum focuses on promoting healthy behaviors related to nutrition, physical activity, and health care. Like the SDC program, the HEC also attempted to foster students’ cultural pride and communalism.

The complete four-year package can be purchased for $820 as of 2012.

EVALUATION(S) OF PROGRAM

Study 1

Flay, B. R., Graumlich, S, Segawa, E., Burns, J. L., & Holliday, M. Y. (2004). Effects of 2 prevention programs on high-risk behaviors among African American Youth. Archives of Pediatrics and Adolescent Medicine, 158(4), 377-384.

Evaluated population: Twelve poor, African-American metropolitan Chicago schools with greater than 80 percent African-American and less than 10 percent Latino enrollment participated in the study. Other school criteria included housing grades kindergarten through eight, total enrollment greater than 500, and a student turnover rate of less than 50 percent per year. From the 12 schools, 1,153 fifth graders were recruited in 1994-1995 and 668 were followed until grade 8. The group was 49.5 percent male and had an average age of 10.8 years at the beginning of 5th grade.

Approach:Participating schools were stratified into four levels of risk, which was assessed using data on enrollment, truancy, mobility, family income, and achievement scores. Schools were then assigned using a randomized block design. Within schools, students were randomly assigned to two experimental groups (SDC and SCI) and one control group (HEC).

The curricula were taught by university-based health educators over the course of four years. In most cases, the health educators remained with the same school to prevent contamination across groups. Two training sessions were held for the educators before each lesson with senior staff providing feedback during training and observation during lessons. The classrooms’ regular teachers attended a four-hour workshop to review the curriculum and their support roles. Surveys administered by trained project staff were completed by the students at the end of each grade year.

Results: Compared to males across other conditions, males in SCI had lower rates of growth in risk behaviors. The SCI significantly reduced violence, provoking behavior, school delinquency, combined behavior, substance use, and sexual activity. The SDC had similar, but only marginally significant impacts, compared with the SCI and there was no impact on males’ condom use. The SCI had a larger impact in the total combined behaviors analysis than did the SDC. There were no significant program impacts for girls.

The HEC may have been more effective than a standard control condition because of the way in which the program integrated cultural pride into the curriculum. The effectiveness of the control group may explain the lack of impact among females. Also, the risk behaviors targeted by the program occurred at lower baseline levels for females, so it may have been more difficult for females to achieve reductions similar to males.

Study 2

Segawa, E., Ngwe, J. E., Li, Y., Flay, B. R., & Aban Aya Coinvestigators. (2005). Evaluation of the effect of the Aban Aya Youth Project in reducing violence among African American adolescent males using latent class growth mixture modeling techniques. Evaluation Review, 19(2), 128-148

Evaluated population: A subset of 552 boys from the sample in Study 1 were evaluated. At baseline, the mean age of this sample was 10.9 years.

Approach:See Study 1 for randomization process.

Self-report pre-test data were collected from participating students at the beginning of 5th grade in the fall of 1994. Post-test self-report data were collected at the end of grades 5 through 8 in the spring of 1995, 1996, 1997, and 1998. Survey measures asked students about their violent behaviors over their lifetime and in the past three months (90 days). Together these measures comprised a violence scale consisting of seven items: 1) carrying a gun, 2) carrying a knife, 3) threatening to beat up siblings, 4) threatening to beat up someone else, 5) threatening to cut, stab, or shoot people, 6) cutting or stabbing someone, and 7) shooting someone. Each item was scored on a scale of 0 to 3 (0=never; 1=yes for lifetime, but not in the past 3 months; 2=once in the past 3 months; 3=more than once in the past 3 months). Item scores were added to produce violence scale scores ranging from 0-21.

The results for the SCI and SDC are combined into one treatment group because Study 1 did not find a statistically significant difference in the reductions of behaviors between the two treatment groups.

Results: The data for this evaluation were analyzed using a latent class growth modeling strategy to estimate growth or decline in outcomes over time. For boys, the treatment resulted in significant impacts for violence, provoking behavior, school delinquency, combined behavior, substance use, and sexual activity. At pre-test, the treatment group and the HEC group were not significantly different. At the last post-test, the HEC group had a significantly higher level of violence compared with the treatment group. In a subgroup analysis of low-, medium-, and high-risk for violence groups, program impacts were three times as large in the high-risk class as in the other two.

Study 3

Ngwe, J. E., Liu, L. C., Flay, B. R, Segawa, E., & Aban Aya Co-Investigators. (2004). Violence prevention among African American
adolescent males. American Journal of Health Behavior, 28,S24-S37.

Evaluated population: The 571 boys from the sample in Study 1 were evaluated.

Approach: See Study 1 for randomization and data collection process.

The study assessed violence behavior, behavioral intentions, attitudes toward violence, estimates of peers’ behaviors (for example, “How many students in your grade get into physical fights?”), estimates of best friend’s behaviors, and peer encouragement (for example, “Do your best friends want you to get into a physical fight?”).

As in Study 2, the SCI and SDC are analyzed as one treatment group.

Results: The violent behavior data were analyzed using a basic latent growth curve. Results revealed that the growth over time in violence among the intervention group was significantly lower than in the control group. When compared to the control group, the treatment group also had a lower rate of growth in behavioral intentions, attitudes toward violence, estimates of peers’ and best friends’ behaviors, and peer encouragement.

Study 4
Jagers, R. J., Morgan-Lopez, A. A., Howard, T-L., Browne, D. C., Flay, B. R. & Aban Aya Coinvestigators. (2007). Mediators of the Development and Prevention of Violent Behavior. Prevention Science 8, 171-179.

Evaluated population: A total of 668 students from the sample in Study 1 were evaluated. This excludes students who transferred into the school and joined the program after baseline data collection.

Approach: See Study 1 for randomization process and data collection procedures.

The survey questions analyzed in this evaluation intend to measure communal values, empathy, and violence avoidance efficacy beliefs. Communal values reflect students’ ratings of the importance of (1) cooperating with others, (2) thinking about family, (3) learning about African American culture and history, and (4) keeping the neighborhood clean, on a scale from one to five. Empathy reflects students’ ratings of their agreement with statements that reflect empathetic feelings such as “When I see someone getting used I feel badly.” Participants recorded their violence avoidance efficacy beliefs by indicating how sure they were they could (1) stay away from a fight, (2) seek help instead of fighting, and (3) keep from getting into a fight.

This study uses longitudinal growth modeling for cluster-correlated data in order to account for any correlation between students in the same school, without explicitly including variation between the schools.

Results: Students in both of the treatment groups experienced a slower decline in empathy than students in the control group. The decrease in empathy in the control group was not statistically significant.

The treatment and control groups’ changes in communal values and violence avoidance efficacy beliefs were not significantly different from each other.

Study 5
Jagers, R. J., Morgan-Lopez, A. A., Flay, B. R., and Aban Aya Investigators. (2009). The Impact of Age and Type of Intervention on Youth Violent Behaviors. Journal of Primary Prevention, 30, 642-658.

Evaluated population: A total of 1,044 students who provided at least one wave of violence data were evaluated, which includes youth who transferred into the school and joined the study after baseline data collection in 5th grade.

Approach: See Study 1 for randomization process and data collection procedures.

This evaluation analyzes only the responses to the self-reported violent behavior survey questions described in Study 4. A three-level hierarchical model was used to account for the similarity between students over time and students in the same school. Age of entry into the study was included in the analysis in order to identify whether the impact of the interventions were different for kids of different ages. This kind of analysis was made possible by including the students who joined the ongoing interventions after the initial data collection, some of whom were older when they transferred into the school.

Results: There was no significant change in violent behavior between the schools. Both intervention groups had significant changes in violent behaviors over time compared to the control group. Both the SCI and the SDC groups had lower rates of growth in violence than the HEC group.

The impact of the SDC intervention was not different for students of different ages. For the SCI group, the change in violence between the younger kids in the SCI treatment and control group was not significantly different. However, the older kids in the SCI treatment had lower growth in violence than the control group.

Study 6
Liu, L. C., Flay, B. R., & Aban Aya Investigators. (2009). Evaluating Mediation in Longitudinal Multivariate Data: Mediation Effects for the Aban Aya Youth Project Drug Prevention Program. Prevention Science, 10, 197-207.

Evaluated population: The 571 male students from Study 1 were evaluated.

Approach: See Study 1 for randomization process and data collection procedures.

This study evaluated the male participants’ responses to a survey questions related to substance use. A substance use behavior question asked about the number of times in the previous month a student (1) smoked cigarettes, (2) used alcohol, (3) got drunk or high, and (4) used marijuana. Other measures include intention to use substances, attitudes toward using substances, self-efficacy of ability to abstain from substances, estimate of grade-peers who use substances, friends’ substance use, and friends’ encouragement to use substances. Each of those measures was asked separately about alcohol, cigarette, and marijuana use.

Latent growth models were used to analyze the change over time in substance use behavior. A curve-of-factors model was used to evaluate the impact of the program on the other measures.

Results: The rate of growth in substance use was significantly smaller for the students in the SCI group compared with the HEC control group. The SDC also had a smaller growth in substance use than the control group, but the difference was only marginally significant. While students in two treatment groups did not have significantly different changes in their substance use behavior, there is some evidence that the rate of growth in substance use among the SDC group decreased over time.

Both treatment groups had significantly lower growth in their estimates of peers’ and friends’ substance use than the control group. The impact of the SCI intervention on students’ attitude toward substance use was marginally significant and the impact on reported friends’ encouragement of substance use was significant, compared with the control group. The treatment and control groups did not have significant differences in their intentions and self-efficacy of ability to abstain from substances.

SOURCES FOR MORE INFORMATION

References:

Segawa, E., Ngwe, J. E., Li, Y., Flay, B. R., & Aban Aya Coinvestigators. (2005). Evaluation of the effect of the Aban Aya Youth Project in reducing violence among African American adolescent males using latent class growth mixture modeling techniques. Evaluation Review, 19(2), 128-148.

Flay, B. R., Graumlich, S, Segawa, E., Burns, J. L., & Holliday, M. Y. (2004). Effects of 2 prevention programs on high-risk behaviors among African American Youth. Archives of Pediatrics and Adolescent Medicine, 158(4), 377-384.

Jagers, R. J., Morgan-Lopez, A. A., Howard, T-L., Browne, D. C., Flay, B. R. & Aban Aya Coinvestigators. (2007). Mediators of the Development and Prevention of Violent Behavior. Prevention Science 8, 171-179.

Jagers, R. J., Morgan-Lopez, A. A., Flay, B. R., and Aban Aya Investigators. (2009). The Impact of Age and Type of Intervention on Youth Violent Behaviors. Journal of Primary Prevention, 30, 642-658.

Liu, L. C., Flay, B. R., & Aban Aya Investigators. (2009). Evaluating Mediation in Longitudinal Multivariate Data: Mediation Effects for the Aban Aya Youth Project Drug Prevention Program. Prevention Science, 10, 197-207.

Ngwe, J. E., Liu, L. C., Flay, B. R, Segawa, E., & Aban Aya Co-Investigators. (2004). Violence prevention among African American adolescent males. American Journal of Health Behavior, 28,S24-S37.

Link to program curriculum: http://www.socio.com/passt24.php

KEYWORDS: Adolescents, Elementary, Middle School, Males and Females (co-ed), Male only, High-Risk, Black/African American, Urban, Community-Based, Manual is Available, Cost Information is Available, Skills Training, Aggression, Delinquency, Other Substance Use, Sexual Activity, Contraception and Condom Use, School-based, Children, Parent or Family Component, Tobacco Use, Marijuana/Illicit/Prescription Drugs, Alcohol Use, Other Behavioral Problems

Program information last updated 9/4/2012

Subscribe to Child Trends

Short weekly updates of recent research on children and youth.