Preventing Sexual Violence Through a Comprehensive, Peer-led Initiative

Sexual violence (SV) impacts adolescents at rates higher than most other age demographics.
Although it is a critical public health issue for youth, there are few prevention initiatives
that lead to reductions in SV. There is increasing recognition that SV prevention efforts may
be most effective if youth are central to the development and implementation of such efforts.
Problematically, there are no rigorously evaluated youth-led SV prevention efforts, and
researchers also know little about the process by which youth create prevention initiatives.
Teen UP (TU) is a youth-led initiative in Rapid City, SD that focuses on addressing a variety
of public health issues, including awareness about SV. This "homegrown" initiative is well
integrated in the community and is the ideal platform on which additional multi-level,
evidence-based SV prevention strategies can be added. Researchers and SV prevention
specialists at the University of New Hampshire, in partnership with Rape Prevention
Education-funded agencies in South Dakota (SD) (i.e., Network Against Family Violence and
Sexual Assault [The Network] and Working Against Violence Inc. [WAVI]), will work together
along with a number of community stakeholders (e.g., educators, peers, elders) and expert
consultants to enhance TU to include additional SV prevention strategies, a project titled
Youth Voices in Prevention (Youth VIP). Although the plan for the programming enhancement
will be refined based on feedback from the Research and Programming Advisory Board (RPAB) and
data collected during the refinement and planning phase of the project, the current plan is
to convene a youth summit where high school youth will be trained as popular opinion leaders
to diffuse SV prevention messages in addition to youth joining a formal working group
specific to a SV prevention strategy that also aligns with youth's career aspirations. A
mixed method approach will be used to measure implementation processes, fidelity, and costs
over time, and to assess the impact of Youth VIP on rates of sexual violence among middle and
high school students and intermediate outcomes including positive bystander actions, and
indicators of positive youth development. Collaboration with the community of students,
parents, and professionals in Rapid City will be promoted by establishing a Research
Programming Advisory Board (RPAB) that will meet monthly during the project.

All students in grades 7 through 10 in the fall of 2017 will be invited to participate in two
baseline and four follow-up surveys that will be given in school. Surveys will take place in
fall and spring of each of three school years, and the cohort enrolled in fall of 2017 will
be followed over time. Recruitment of students and parents is being done in close
collaboration with the Rapid City Area School District. The investigators anticipate
enrolling 2,000 students.

A comparison community that is demographically similar to Rapid City but located in a
neighboring state will enroll a cohort of 7th through 10th graders during the 2017-18 school
year. These students will take three surveys, one each spring during the three years of the
study. This school district will not receive Youth VIP. The investigators anticipate
enrolling 1,000 students.

Qualitative data related to implementation process and fidelity will be examined using
content analysis. Continuous variables will be summarized with descriptive statistics (mean,
standard deviation); categorical variables will be described with counts and percentages.
Change scores and percent change scores will be calculated across several baseline time
points as well as across two years of program implementation. Multilevel modeling will be
used to examine changes among Rapid City students compared to students in the comparison
community. One data set will be change across time at the individual level and another will
be aggregated school-level data with the matched control school data merged in. Descriptive
statistics will be conducted to evaluate the distributions of variables, testing for out of
range values and outliers, and determining the need for transformation of non-normal
distributions. In addition, the investigators will evaluate the measures for their
psychometric properties for the current samples using exploratory and confirmatory factor
analysis or structural equation modeling. Evaluation will include inspection of factor
loadings, error variances, intercorrelations of factors, path coefficients and overall model
fit. Additionally, the investigators will test for treatment differences in the total sample
on key variables collected on each time point (e.g., gender, income, race/ethnicity, etc.).
The investigators will calculate Horvitz-Thompson weights or propensity scores to adjust for
bias in the models to compare treatment effects. For the primary outcome, SV rate of
victimization and perpetration, the investigators will use mixed effect logistic regression
models to estimate the difference of intervention and non-intervention groups
(pre-intervention vs. post intervention in single case, multiple baseline design and
treatment difference between intervention schools and matched non-intervention schools),
adjusting by other background variables including social, demographics, and community-level
characteristics. Secondary behavioral, dichotomous outcomes will be analyzed the same way SV
perpetration and victimization are analyzed. The model can also be slightly altered to
include time as a continuous variable to examine the trajectories of SV rate reduction as
well as changes over time in intermediary and secondary outcomes as a function of Youth VIP
exposure and engagement.

Inclusion Criteria:

- Must be able to complete online survey independently

Exclusion Criteria: