Positive Youth Development in the United States: Research Findings on Evaluations of Positive Youth Development Programs. Effective Health Promotion Programs in Schools

11/13/1998

Know Your Body

Walter, Vaughan and Wynder (1989) evaluated the impact of Know Your Body, a school-based health promotion intervention that addressed eight positive youth development constructs including social, emotional, cognitive, and behavioral competencies, self-efficacy, recognition for positive behavior, positive identity, and prosocial norms.

Program Description

The program strategies encompassed a range of social, cognitive and behavioral skills training designed to promote children's competencies and self-efficacy on health and self-management issues.  These included educating youth on the connections between smoking-related decisions and self-image, values, anxiety, and stress (positive identity, prosocial norms); skills training in stress management, decision-making, communication and assertiveness (competence, self-efficacy, recognition, positive identity); and increased awareness of social influences in the initiation of smoking (prosocial norms).  The curriculum-based intervention was based on the principles of social learning theory and the health beliefs model.  The primary goal was to modify personal health behaviors related to the future development of cancer, including cigarette smoking, diet, and physical activity.

Research Design

The study used an experimental design, with eight schools (485 students) randomly assigned to the intervention group and seven schools (620 students) randomly assigned to the control group.  There were only two measurement points over six years, at baseline and at six-year posttest.  Both the attrition analysis and pretest measures of group equivalence showed that the 65% of original participants who remained at posttest measurement were not significantly different from those lost to the study on baseline levels of risk factors, knowledge, or behaviors.  The evaluation used schools as the unit of analysis and unit of assignment.  Implementation quality was assured, though not systematically measured, with teacher adherence to special teaching protocols monitored by research staff.  The study targeted a population of all fourth grade children (n=1105) in 15 New York city elementary schools serving Hispanic/Latino, Caucasian, African-American, and Asian ethnic groups, although the largest group represented was primarily middle class Caucasians.  Of the eligible subjects, 911 (82.4% overall, 92% of intervention school students and 75% of nonintervention school students) participated in the baseline measures.  By the end of the study, 593 (65%) of the sample had measurement data recorded at both baseline and at termination six years later.  In the baseline sample, the mean age was 8.9 years at baseline, and 79.3% were Caucasian, 13.8% were African American, 2.2% were Hispanic and 4.7% were another ethnic identity (mostly Asian).  The authors did not state the ethnographic composition of the final sample at posttest in this evaluation. The intervention provided two hours of curricula per week taught by the children's regular teachers throughout the school year over a six-year period.

Results of the Intervention

Behavioral outcomes were reported through a combination of physiological measures (e.g., serum tests to detect nicotine levels, examining and weighing school lunch portions), parent or caretaker reports, and 24-hour dietary recall interviews.  The results at six-year posttest showed that the program was effective in modifying two major risk factors associated with the development of cancer -- smoking and diet.  Ninth grade students (both females and males) in the intervention schools had 73.3% (p<.005) lower rates of smoking initiation than students in the nonintervention schools.  Also by the ninth grade, students (both females and males) in intervention schools had 19.4% (p<.05) lower levels of saturated fat, a 9.8% decrease in total fat, and a 9.5% (p<.05) net increase in consumption of carbohydrates.  Significant effects among sub-groups showed that males increased their consumption of total carbohydrates, while females showed net decreases in total and saturated fat plus increases in total carbohydrates and crude fiber.

Growing Healthy

The analysis of the Growing Healthy program (known originally as the School Health Curriculum Project, or SHCP) used information from several evaluations (Connell, Turner & Mason, 1985; Connell & Turner, 1985; Smith, Redican & Olson, 1992).  The study of main interest was an evaluation of the cumulative effects of the intervention.  This was a sub-study of the larger School Health Education Evaluation (SHEE), an extensive study of more than 30,000 children, grades 4 to 7, in 1071 classrooms in 20 states.  Growing Healthy addressed nine positive youth development constructs, including social, emotional, cognitive, and behavioral competencies, self-efficacy, opportunities for prosocial involvement, recognition for positive behavior, positive identity, and prosocial norms.

Program Description

The intervention was a comprehensive school health education curriculum that addressed emotional and social aspects of a child's growth and development.  The program strategies were implemented by regular classroom teachers, who were trained along with other educational staff in a team training format over three to five days.  Students were exposed to between 43 and 56 lessons over a one or two year period (depending on intervention condition).  Strategies targeted youth's skills and educational innovations in the classroom.  Skills training covered a range of social, cognitive, and behavioral instruction that promoted various positive youth development constructs (e.g., competence, self-efficacy) and enhanced self-image and school and home behavior (positive identity, competence, self-efficacy).  Educational innovations were introduced to positively change teacher practices (promoting recognition for positive behavior, opportunities for prosocial involvement, and prosocial norms).

Research Design

The Growing Healthy/SHCP study used a quasi-experimental design with comparison groups.  Four school districts (one mid-Eastern urban district, one mid-Southern suburban district, and two Western rural districts) were selected on the basis of meeting these conditions: all had a pool of children who had received one grade unit (43-56 sessions) of instruction in the preceding year; had a second pool of children who had received no instruction; had both pools take part in pre- and posttesting in the previous year; and had scheduled implementation for the next higher grade level after the first intervention year.  The study was designed to assess the cumulative effects of the intervention on 65 classrooms whose students fell into two groups, prior program exposure and no exposure.  The unit of analysis was partial classrooms defined by previous health instruction.  This resulted in eight groups of students (n=1397) in 130 partial classrooms with varying exposure levels: two groups receiving two units or doses of the curriculum, four groups receiving one unit, and two groups receiving no units.  In the 1982-83 school year, fourth and fifth grade students were enrolled in classrooms which either did or did not receive the program.  The performance of these groups at the beginning of the year was assessed on tests of knowledge, attitudes and practices by comparing 01 vs 03 and 05 vs 07.  Initial testing for effectiveness was done through comparing 02 vs 04 and 06 vs 08.  Youth from these classrooms were then promoted to the next higher grade: grade four SHCP and unexposed students were promoted to grade five and grade five SHCP and unexposed students were promoted to grade six.  Pretest measures of group equivalence were conducted and showed that groups were comparable at baseline measurement points.  The youth consisted of 47% who had been previously exposed to SHCP and 50.3% who were male.  Their ethnic identities were 40% African-American, 1% Hispanic, 57% Caucasian, and 1% from unspecified ethnic backgrounds.

Results of the Intervention

The results indicated significant effects for the intervention groups in both years compared to those with no exposure to the curriculum, shown by differences at posttest for knowledge, attitudes, practices, and smoking variables.  On all knowledge measures (growth and development, mental health, personal health, nutrition, family life, disease prevention, substance use/abuse, safety/first aid, consumer health, community health), attitude measures (accepting personal responsibility, maintaining a healthy body, safeguarding the environment, respecting the rights and roles of self and others), and measures of practices (decision-making skills, personal healthy practices, social adaptability), groups with two units of exposure performed better than those with one unit, while both groups scored higher than groups with no exposure to the curriculum.  At the 1982-83 posttest, intervention classrooms outperformed unexposed classrooms by 9% on the knowledge tests, and 5% on the attitude and practices scales (all findings significant at p<.01). The 1983-84 posttests found significant effects for knowledge by exposure and grade, and for attitudes by exposure group.  Significant effects for practices were found by exposure and grade, with the fifth grade having higher practice scores than the sixth.  Effects on smoking variables showed both exposure and grade differences.  Although the fifth grade children showed no significant differences for percent smoking, regardless of exposure (fifth grade students reported little smoking activity), students who had received two units of exposure indicated less intention to smoke in the future than unexposed groups. Among sixth grade groups, both exposure groups were lower for both current smoking (2 units = 2.8%; 1 unit = 3.7%, no exposure = 9.6%, p<.05) and for future intentions to smoke (2 units = 9.2%; 1 unit = 8.4%; no exposure = 17.1%, p<.05) than the unexposed group.

The Children of Divorce Intervention Program

Pedro-Carroll and Cowen (1985) evaluated the Children of Divorce Intervention Program, an intervention for elementary school age children that addressed 10 positive youth development constructs, including social, emotional, cognitive, and behavioral competencies, bonding, self-efficacy, resiliency, prosocial norms, opportunities for prosocial involvement, and recognition for positive behavior.

Program Description

The strategies in this study combined health and competence promotion techniques based upon a child development theory that hypothesized how children respond to parental separation.  The 10-session framework corresponded to an initial stage (two sessions) for building group cohesion and trust, (bonding, prosocial norms), an interim stage (four sessions) for structuring practice and rehearsal opportunities (competence, self-efficacy, opportunities, recognition) and a final stage (four sessions) for enhancing mastery and the generalization of skills to the youth's environment (resiliency, self-efficacy).  Specific strategies included skills for effective problem-solving, communications, decision-making, and anger and anxiety management.

Research Design

The intervention used an experimental design in which children were randomly assigned to an experimental or control condition within their own schools, and were matched by sex, grade, length of time since their parents' separation, and eight of 10 pre-adjustment measures.  There was no attrition in the study. A posttest was done for all measures two weeks after the intervention ended. No follow-up data were collected.  The study sample of third through sixth grade Caucasian students included 75 participants (42 boys and 33 girls) from four suburban schools who had been recruited from program descriptions sent to the parents.  Another program evaluation (Pedro-Carroll, Alpert-Gillis & Cowen, 1992) was reviewed, but the 1985 study is reported here because of its stronger research design.  We note that the later quasi-experimental study showed similar patterns of results.

Results of the Intervention

Results showed significant improvements at the two-week posttest for the experimental group compared with the control group on a number of behavioral outcomes.  Teacher ratings indicated that children in the experimental group had greater reductions in anxiety (31.5% vs 34.16%, p<.02) and learning problems (23.07% vs 27.75%, p<.05), as well as on an overall index of classroom adjustment problems (F(3,66) = 8.49, p<.001).  The children in the intervention group were rated higher by teachers on a total competence score (t = 6.50, p<.001), and on specific competencies including peer sociability (F(5,64) = 17.59, p<.001), frustration tolerance (F(5,64) = 6.04, p<.05), compliance with rules (F(5,64) = 7.72, p<.01), and adaptive assertiveness (F(5,64) = 5.10, p<.05).  Group leaders rated significant increases in skills for personal problem solving (t = 3.85, p<.001).