IMPLICATIONS FOR PREVENTION OF RISK AND PROTECTIVE FACTORS AND ANTISOCIAL BEHAVIOR, BULLYING & AGGRESSIVE BEHAVIORS, PEER INFLUENCE AND AGGRESSIVE AND NON-AGGRESSIVE YOUTH
Chair: Tony Biglan
THE PREVENTION OF BULLYING AND OTHER AGGRESSIVE BEHAVIORS IN ELEMENTARY SCHOOL STUDENTS: INITIAL EFFECTS OF THE YOUTH MATTERS CURRICULUM. Jeffrey Jenson1, William Dieterich1, Jenifer Rinner1, Felicia Washington1, Kathleen Burgoyne2, 1University of Denver, Denver, CO United States; 2Comprehensive Health Education Foundation, Seattle, WA United States
Bullying and other forms of aggressive conduct in school settings exact significant individual, educational, and societal costs. Bullying behaviors pose risk to individual student victims and disrupt classroom learning processes. Students who bully their classmates consume a disproportionate share of school resources and have the capacity to adversely affect overall classroom and school climate. Bullying and other forms of early aggression are also associated with internalizing conditions such as depression and externalizing behavior problems such as delinquency and substance abuse. The relationship between early onset of aggression and subsequent involvement in antisocial behavior during adolescence reinforces the need for efficacious school-based prevention strategies targeting bullying and aggression during the elementary school years.
In this presentation, we report initial outcomes from a group-randomized trial of a prevention curriculum aimed at preventing and reducing bullying and other aggressive behaviors among elementary students in the Denver, Colorado public school system. Fourteen elementary schools in urban Denver were randomly assigned to receive selected modules of the Youth Matters curriculum that targeted bullying and aggression. Fourteen comparable schools were randomly assigned to a no-treatment control group. Intervention occurred in 39 experimental-group classrooms between 2003 and 2005. Based on the social development model articulated by Hawkins and Catalano, the Youth Matters curriculum incorporates knowledge of risk and protective factors associated with the onset of aggression and other antisocial behaviors in lesson design. The primary outcomes are measured with items from the Olweus Bullying Questionnaire. The efficacy of the Youth Matters curriculum was analyzed in mixed-model analysis of covariance and mixed-model logistic regression with adjustment for baseline values of the bullying measures. A significant interaction effect was found between gender and the Youth Matters intervention on a binary indicator of self-reported bullying behavior. Among boys, the odds of self-reported bullying behavior at follow-up were reduced by 50% in the Youth Matters condition, compared to the odds in the control condition, holding age at follow-up and the baseline value of bullying constant. Period-prevalence rates of bullying, time-3 results and implications of findings for the implementation of group-randomized trials in urban elementary school settings will be discussed.
WHAT RISK FACTORS PREDICT ANTISOCIAL BEHAVIOR? WHAT PREVENTION PROGRAMS CHANGE THOSE RISK FACTORS? USING META-ANALYSIS TO LINK RESEARCH EVIDENCE. Mark W. Lipsey1, Sandra Jo Wilson1, Kelly A. Noser1, 1Vanderbilt University, Nashville, TN United States
Risk-oriented prevention strategies for antisocial behavior aim to ameliorate risk and enhance promotive factors. Pursuing such strategies in programs for at-risk children requires, first, valid identification of significant and malleable risk factors and, second, effective interventions for reducing or buffering those risk factors. This presentation summarizes results to date from a project that applies meta-analysis techniques to (a) longitudinal research studies to identify the most predictive risk factors for antisocial behavior in adolescence and early adulthood, (b) intervention studies to determine which have the largest effects on different risk factors, and (c) the linkage between the outcome variables in intervention studies and the predictive variables in longitudinal studies. The objective of this project is to link and interpret these two large bodies of research in ways that reveal the most promising approaches to the prevention of antisocial behavior.
The meta-analysis of risk-outcome relationships examines the findings of longitudinal panel studies with samples of children age 4-18 on the first wave of measurement. Approximately 30,000 correlation coefficients representing both longitudinal and cross-sectional relationships between risk and/or antisocial outcome variables have been coded from 221 studies (each providing multiple study reports) and entered into a database for analysis. The meta-analysis techniques applied to integrate subsets of this data relating to different risk and outcome constructs at different developmental periods will be briefly described and findings about the most predictive risk factors will be reviewed.
The meta-analysis of research on the effects of intervention examines 425 studies with samples of children under age 15 and risk factors for antisocial behavior as outcome variables. The analysis focuses on identifying the interventions with the largest effects on the risk factors that are most strongly predictive of antisocial behavior. The challenges and procedures for linking outcome variables in one meta-analysis with predictor variables in another will be briefly described and the findings about what this research evidence shows about the most effective prevention programs will be summarized.
RELATIONSHIPS BETWEEN AGGRESSIVE AND NON-AGGRESSIVE YOUTH: WHAT DETERMINES THE DIRECTION AND DEGREE OF PEER INFLUENCE?. Kelly L. Rulison1, Scott Gest1, Janet Welsh1, 1Pennsylvania State University, University Park, PA United States
Relationships between aggressive and non-aggressive youth create a paradox in that they may be protective for aggressive youth but risky for non-aggressive youth. The goal of this poster is to identify conditions under which such relationships exacerbate or ameliorate behavior problems. We expect that the direction and degree of peer influence may depend on 3 factors: Children´s relative network centrality (more central youth being most influential), the overall proportion of peers who are aggressive or non-aggressive and the characteristics of a child´s closest peers.
To examine these issues, teacher-, peer- and self-reports were obtained in October and May from 382 students in 3rd, 4th and 5th grades (86% of enrolled students). Affiliation networks were defined from reciprocated friendship nominations and groups identified by Cairns´ Social Cognitive Map (SCM) method. Five teacher-rated items, standardized within class, were used to classify children´s aggression as High (z>1; N=56), Moderate (0< z <1; N=61) or Low (z<0; N=265). Network centrality, the total number of times each child was named to a group, was similarly standardized and categorized. Two peer nomination items of aggression were also obtained.
On average, children had 7 classmates who were either mutual friends or members of the same social group. Relationships between High and Low aggression youth were common: all High aggression children had at least one Low aggression peer and 70% of Low aggression youth had at least one High aggression peer. The usual practice of averaging across peers´ characteristics presumes all peers are equally influential, but this practice masks the diversity in peers´ network centrality, level of aggression and dyadic relationship strength. For example, 47% of Low centrality youth had at least one High centrality peer; the proportion of High aggression network members ranged from 0 to 100%; and on average only 2 of children´s 7 peers were especially strong (i.e., both mutual friends and members of the same group).
Preliminary analyses indicate that, on average, the presence or absence of aggressive peers was not associated with Fall-to-Spring changes in aggression. For example, after controlling for Fall aggression, there were no differences in Spring teacher- or peer-rated aggression between Low aggression children with no High aggression peers and those with at least one such peer. Final analyses will capitalize on the variability in network centrality, proportion of aggressive peers and dyadic relationship strength to further specify the conditions under which children with contrasting behavioral styles influence one another. Implications for interventions that include Low aggression youth in programs targeting High aggression youth will be discussed.
POSITIVE YOUTH DEVELOPMENT PROGRAMS: A CONCEPTUAL MODEL AND META-ANALYSIS OF PROGRAM OUTCOMES
Chair: Roger Weissberg
POSITIVE YOUTH DEVELOPMENT PROGRAMS: A CONCEPTUAL MODEL AND META-ANALYSIS OF PROGRAM OUTCOMES. Roger Weissberg1, Joseph Durlak1, Denise Gottfredson2, Mary Utne-O'Brien1, 1University of Illinois at Chicago, Chicago, IL United States; 2University of Maryland College Park, College Park, MD United States
The purpose of this symposium is to present the results of the first extensive and systematic meta-analysis of positive youth development (PYD) programs. Over 600 published and unpublished interventions are evaluated. Currently, there is much interest in PYD, which generally refers to helping youth to become contributing members of society by fostering their interpersonal and intrapersonal growth through various types of competency-enhancing interventions. Previous narrative reviews of this literature have concluded that PYD has merit (Catalano, Berglund, Ryan, Lonczak & Hawkins, 2002; National Research Council, 2003), but several questions remain. First, because several alternative models of PYD have been suggested, what is the best way to conceptualize and define the core features of PYD approaches? Second, to what extent have PYD programs been successful in improving young people´s adjustment on psychological, academic and social dimensions of functioning? Third, how do program effects vary as a function of participant characteristics and program components, and is it possible to identify some interventions as exemplars in terms of conceptualization, execution, and impact? Fourth, and finally, what is the relationship between enhancing competencies and preventing later problems?
The first presentation in the symposium presents a model that categorizes different types of PYD programs based on our extensive literature search and coding procedures. The second presentation summarizes our main post and follow-up findings for different types of programs, participants, and outcome domains. The third presentation focuses on dissemination of our results because one important phase of this funded project is to communicate our findings to different stakeholders in ways that can affect future research, practice and policy on PYD. A noted prevention researcher will serve as the Discussant, and substantial time will be available for audience reactions and discussion.
POSITIVE YOUTH DEVELOPMENT: HOW WELL DOES IT WORK?. Kei Kawashima1, Joseph Durlak1, Roger Weissberg1, 1Loyola University of Chicago, Chicago, IL United States
This presentation describes the main results of the meta-analysis at post and follow-up by describing the significance and magnitude of effects obtained for different types of programs, populations, and outcome measures. In addition to customary meta-analytic procedures, we present data for selected outcomes that we label as clinically or socially significant effects. That is, we identified salient outcomes that most audiences would perceive as clearly meaningful and practical benefits of intervention. These outcomes include academic achievement, school attendance, and changes in important indices of functioning such as levels of aggression, violence, or delinquency.
The types of programs able to achieve significant changes on these outcomes are described.
WHAT DOES POSITIVE YOUTH DEVELOPMENT ENCOMPASS?. Joseph Durlak1, Kei Kawashima1, Emily Preheim Dupre1, Molly Pachan1, 1Loyola University, Chicago, IL United States
This presentation describes our literature search and coding procedures and then presents a model by which different types of PYD programs were categorized and compared. For the past three years, multiple search procedures have been used to identify and obtain over 600 published and unpublished PYD outcome studies, each including some type of control or comparison group so that standardized mean effects can be computed to evaluate program impact on a variety of outcome domains. The challenges involved in addressing the diverse PYD literature and our methods of categorizing, coding and analyzing data from interventions are described.
KNOWLEDGE TO PRACTICE: AFTER THE COMPUTER ANALYSES AND THE JOURNAL ARTICLE, WHAT THEN?. Mary Utne-O'Brien1, Roger Weissberg1, 1University of Illinois at Chicago, Chicago, IL United States
A somewhat unusual feature of this meta-analysis project was that it was conceived with the ultimate plan of effectively disseminating its key findings in order to influence practice. That is, the funded project includes a major knowledge-to-practice phase, including funding for assessing impacts of dissemination efforts. Indications are that the project will yield information very much worth sharing, including a definitive up-to-date assessment of the current research evidence for positive youth development interventions, and information on variables that moderate or mediate positive program outcomes--critical for fashioning more efficient and effective interventions.
Our dissemination plan is based upon a four stage logic model of influence and change in youth policy and supports. The model describes our vision of the links between our analyses, our dissemination efforts, and our ultimate goals of influencing research, practice and policy.
This portion of the symposium will describe our stage model of dissemination and discuss some of the challenges that face the project team and our proposed solutions: How do we share the study results in a way that does justice to the complexity and richness of the findings, without being too complicated for different audiences? Is there a best order for dissemination targets—should we begin with the scientific community, the practitioner community, or both? What difficulties might be encountered as scientific team´s shifts its role from analysis and interpretation to advocacy? Is there a role for theater in the roll-out of information?
CONCURRENT 8, EMERGING OPPORTUNITIES FOR PREVENTION RESEARCH, Organized symposia
ADOLESCENTS AND OBESITY: INTERVENTIONS AND INFLUENCES
Chair: Matthew Farrelly
ADOLESCENTS AND OBESITY: INTERVENTIONS AND INFLUENCES. Matthew Farrelly1, Doug Evans2, James Hersey2, 1RTI, Research Triangle Park, NC United States; 2RTI International, Washington DC, United States
From 1980-2000 the percentage of adolescents aged 12–19 who are overweight has tripled from 5% to 15% (Ogden et al, 2002). This rise is troubling because it is associated with increases in Type 2 diabetes among adolescents (Rosenbloom et al, 1999) and overweight adolescents are more likely to become overweight or obese adults (Ferraro, Thorpe, and Wilkinson 2003). While increases in the prevalence of obesity are well-documented among adolescents, less is known about its social, environmental, policy, and attitudinal influences.
This symposium includes three papers that present data from a national survey of approximately 2000 adolescents ages 12-18 from the Healthy Youth Panel Survey (HYPS) and findings from a series of qualitative interviews and discussion groups with children, pre-adolescents and their parents to understand the role of television and videogame use in obesity. The HYPS questionnaire was designed using a social-ecological approach to capture a wide range of outcomes related to diet, physical activity, and weight and potential influences on these outcomes. These measures include individual attitudes, beliefs and intentions about physical activity, sedentary behaviors, healthy eating and weight and perceptions about body weight; parental encouragement of physical activity and healthy eating and limits on television and video game use; exposure to media messages promoting physical activity; and school-level factors such as the presence of vending machines in school, availability of healthy and unhealthy food choices in vending machines and school cafeterias, changes in food choice availability between the current and previous school year, physical activity options and changes in those options, and adolescents reactions to these changes.
The first paper sets the stage for the symposium by providing summary statistics on adolescents´ physical activity, weight and diet and related attitudes, beliefs and intentions. We then examine correlations between health behaviors and attitudes to illustrate potential salient messages for media campaigns and health education aimed at promoting physical activity and healthy diets.
The second paper explores the role of school interventions aimed at improving nutrition and physical activity options in schools and promoting healthy behavioral choices among students.
The third study uses HYPS data to explore factors associated with television and videogame use to complement in-depth qualitative interviews and discussion groups with children and their parents to understand families´ perceptions of the appeal and effects of television, videogames, and computers. Also discussed were rules about media use and potential strategies to reduce media use as a way to prevent obesity.
ADOLESCENT SELF-REPORTED EXPOSURE TO SCHOOL-BASED OBESITY PREVENTION INTERVENTIONS: RESULTS FROM A NATIONAL SURVEY. W. Douglas Evans1, Matthew Farrelly2, James Nonnemaker2, Kevin Davis2, James Hersey1, Jeanette Renaud1, 1RTI International, Washington, DC United States; 2RTI International, Research Triangle Park, NC United States
Recent studies have found that many school environments are not conducive to children maintaining healthy body weight. Low rates of physical education that involve vigorous physical activity, poor nutritional choices in school cafeterias, and availability of high calorie/low nutrient density (i.e., junk) and fast food in schools are potential risk factors for childhood obesity.
But many schools are implementing interventions to improve the nutrition and physical activity options in schools and promote healthy behavioral choices among students. These include increasing the proportion of fruits, vegetables, milk, water and other healthy choices in vending machines; reducing or eliminating junk and fast food availability in vending machines and cafeterias; and increasing PE requirements and intramural sports.
Recent research by the presenter (Evans et al. 2005) found strong support among U.S. adults for school-based obesity prevention interventions. This study found that support for the most restrictive interventions, such as removing all junk and fast food from school vending machines and cafeterias, was significantly higher among parents with school-age children. The current study asks to what extent youth are exposed to obesity interventions in schools, their attitudes and opinions about the interventions, and social norms about obesity prevention in schools.
We present data from the Healthy Youth Panel Survey (HYPS), a nationally representative, longitudinal survey of a cohort of approximately 2000 youth age 12-18. HYPS contains a series of questions about presence of vending machines in school, availability of healthy and unhealthy food choices in vending machines and school cafeterias, changes in food choice availability between the current and previous school year, physical activity options and changes in those options, and adolescent attitudes and opinions, and social norms.
We present descriptive statistics on these questions from the first HYPS longitudinal follow-up. We also regress intervention exposure, attitude, and social norms variables on socio-demographic variables such as income, race/ethnicity, and other obesity risk factors including TV and other media use. Results of these logistic regressions will inform future multivariate analyses of the effects of school interventions on obesity-related behaviors and behavioral determinants.
Learning Objectives: At the end of the presentation, the audience will: 1) understand and describe popular school-based obesity prevention interventions, and 2) know U.S. adolescent population estimates of exposure to such interventions, and 3) describe evidence for the effectiveness of these interventions in changing obesity prevention knowledge, attitudes, and behaviors.
INFLUENCING CHILDREN AND YOUTH TV AND VIDEOGAME USE TO REDUCE OBESITY: THE INFLUENCE OF THE FAMILY, PEER, AND MEDIA ENVIRONMENT. James Hersey1, Lucia Rojas Smith1, Christina An1, Jeanette Renaud1, Amy Jordan2, Jude McDivitt3, 1RTI International, Washington, DC United States; 2University of Pennsylvania, Philadelphia, PA United States; 3CDC, Atlanta, GA United States
Recent studies suggest that children´s heavy television viewing may be contributing to the rise in childhood overweight. However, we do not fully understand the factors that influence developmental changes in media use of children and adolescents, and we do not know what types of interventions to reduce media use will work best with different ages of children and adolescents and types of families.
To explore these issues, we conducted two interrelated studies. The first is a longitudinal survey of a national sample of approximately 2000 adolescents (ages 12 – 18) that explored factors that were associated with TV and videogame use. We use a social ecological model to inform multivariate analyses of the structural, social, and family, and attitudinal factors that influenced media use. We also modeled the changes in media use over time.
Then, to extend these survey findings, we conducted an in-depth qualitative study of factors in that influenced TV and media use among children and pre-adolescents. We conducted interviews or discussion groups with 180 with children aged 6-7; 9-10 and 12-13 and with 180 parents of these children. The sample was evenly divided between African Americans, Hispanics, and non-Hispanic white children. The sample in this in-depth study was similar to national samples in terms of the level of use of TV and videogames. These discussions explored families´ perceptions of the appeal and effects of TV, videogames, and computers. We also discussed rules about media use and potential strategies to reduce media use as a way to prevent obesity. .
The study revealed developmental changes in the social influences on media use. The viewing of younger children was largely determined by parental attitudes about content; media use of early adolescents was more influenced by peers, the appeal of TV and videogames; and the physical availability of media in the home. Parents´ rules about media use were motivated by education and behavioral concerns (rather than obesity or prevention). We investigated alternative strategies to reduce TV use. The analysis demonstrates the value of earlier interventions, social inoculation, and development of social competence in reducing media use and preventing obesity.