|
Correlates of Perceived Effectiveness of the Safe Schools and Healthy Students Initiative
|
| Presenter(s):
|
| Bruce Ellis, Battelle Memorial Institute, ellis@battelle.org
|
| Ping Yu, Battelle Memorial Institute, yup@battelle.org
|
| Aaron Alford, Battelle Memorial Institute, alforda@battelle.org
|
| Danyelle Mannix, United States Department of Health and Human Services, danyelle.mannix@samhsa.hhs.gov
|
| Sharon Xiong, Battelle Memorial Institute, xiongs@battelle.org
|
| Abstract:
A three-level growth curve model was applied to estimate school-perceived impact growth trajectories, using multi-year data on outcome and progrma implementation from project and school surveys collected from 40 grantees and 3,300 participating schools. Primary interest is to determine whether and how project-level correlates affect schools' perceptions of the initiatiative's effectiveness over time when effects of pre-grant community and otehr environmental conditions are considered. Comprehensive programs and activities were found to be a significant predictor of initial increases in perceived overall impact of the initiative, of safety and violence prevention activities, and of substance use prevention activities even when the effect of funding was considered. Coordination and service integration activities were significantly related to mean rate of growth for substance use prevention over a three-year period. The paper demonstrates that this type of longitudinal grwoth curve model is appropriate for evaluating large-scale prevention initiatives funded by the Federal Government.
|
|
County-level Versus State-level Results: Is It Worth the Effort?
|
| Presenter(s):
|
| Susan Saka, University of Hawaii, Manoa, ssaka@hawaii.edu
|
| Abstract:
Given limited resources, is it worthwhile to increase sample sizes to obtain county-level results instead of only state-level results? Reasons for doing so for the Hawaii Youth Risk Behavior Survey (HYRBS) include (a) schools are interested in their own students, (b) community groups and state agencies want relevant data for grant applications, (c) educators want to know if their interventions had any effect, and (d) people want to know if events such as several youth-aged suicides in a small community were reflected in survey results.
Using data from the 2003 HYRBS where 8,791 students from 43 schools statewide completed usable surveys, simple analysis of variance (ANOVA) was conducted, including the calculation of percent of variance accounted for, to compare the differences among four counties and the state on key indicators. Findings that helped a planning committee to determine whether or not to increase the sample size will be presented.
|
| |