Bystander interventions, recognized as promising violence prevention strategies, are unique in their engagement of all community members to 1) recognize situations that may become violent and 2) learn to safely and effectively intervene to reduce violence risk. Based on their promise, the Campus Sexual Violence Elimination Act (SaVE) now requires all publicly-funded colleges to provide bystander intervention. With SaVE's policy intervention requiring bystander interventions, a "natural experiment" has arisen to determine the relative efficacy of students' bystander training across multiple colleges. Investigators propose a quasi-experimental design (using fractional factorials) to evaluate the relative efficacy of three bystander interventions to reduce violence in college communities. Green Dot will be one of three bystander interventions evaluated. A recent rigorous evaluation has found that Green Dot is associated with a 20-40% reduction in VAW in college and high school settings.

In aim 1, investigators will compare the relative efficacy of bystander interventions to a) increase bystander efficacy and behaviors, b) reduce violence acceptance, c) reduce interpersonal violence victimization and perpetration, and d) increase program cost effectiveness. The three main bystander groups compared will be: exclusively online training, Green Dot (speeches and intensive bystander training), and other skills-based bystander training. Program efficacy data will be obtained from student surveys, campus crime statistics, and surveys with college staff and administrators responsible for selecting and implementing bystander interventions.

In aim 2, investigators seek to grow communities of VAW prevention researchers. Researcher communities will form through researchers' engagement with college recruitment, survey design, data collection and analyses. Specifically investigators will determine the efficacy of this program to increase VAW prevention research productivity defined as a) increasing research skills and b) increasing research communications measured as manuscript submissions, presentations, and publications.

This natural experiment will generate new understanding into efficacy of how bystander programs work. This natural experiment will also provide the VAW research community an opportunity to increase our skill-sets and share our experiences with and help grow the next generation of VAW prevention researchers.

A. SPECIFIC AIMS Goal 1. Grow Colleges as Communities for Primary Prevention of Violence Against Women Research The opportunity: Because the 2013 Campus Sexual Violence Elimination Act1 (SaVE) now requires bystander training in all publicly funded colleges and universities (hereafter colleges), a "natural experiment" presents itself to determine what components of bystander interventions have greater efficacy. Inclusion of multiple college communities (n=24) provides sufficient statistical power for these comparisons. The timing of this U01 (RFA-CE-15-003) is strategically important for the work investigators propose: to evaluate the efficacy of bystander interventions components (BIC) to reduce violence against women (VAW: operationally defined as intimate partner violence (IPV) and sexual violence (SV)). Investigators propose a quasi-experimental design using fractional factorials. Data to measure program efficacy for bystander and violent behavior outcomes will be obtained from student surveys and campus crime data. Key informant interviews with college administrators responsible for selection and implementation of bystander programs will describe components and attributes of bystander programs important for colleges' adoption, implementation and retention of bystander programs over the 4 years.

The purpose of this quasi-experimental, multi-college study is to evaluate the relative efficacy of newly implemented bystander interventions over time. Because all colleges must adopt an intervention, comparisons will be made prospectively before and after implementation and across program interventions.

Aim 1. Compare the relative efficacy of bystander interventions to a) increase bystander efficacy and behaviors, b) reduce violence acceptance, c) reduce violence victimization and perpetration, and d) increase program cost effectiveness. Three BIC comparisons will be: exclusively online training, Green Dot, and other skills-based, interactive bystander programs. Whether training is mandatory will also be evaluated.

Hypothesis 1: Relative to colleges implementing exclusively online training, colleges implementing Green Dot (or other skills-based programs) will significantly a) increase bystander efficacy and behaviors, b) reduce violence acceptance, c) reduce violence victimization and perpetration, and d) demonstrate greater cost effectiveness to reduce VAW.

Goal 2. Grow Communities of Violence Against Women Prevention Researchers The opportunity: This "natural experiment" also offers a unique opportunity for more senior VAW prevention researchers to share their expertise with each other and with more junior researchers, i.e. grow a researcher community. Team science is increasingly required in violence prevention research yet funding mechanisms supporting cross-institutional and cross-disciplinary career development grants in violence prevention are limited.

The purpose: Investigators propose a researcher community collaborative to grow the next generation of VAW researchers. Through engaging multiple public universities as our research community partners (Goal 1), investigators have the added opportunity to partner as a researcher community to both a) conduct better research because a broader skill-set can be engaged and b) share our research skill-sets with more junior researchers.

Aim 2. Determine the efficacy of this collaborative to increase VAW prevention researchers' productivity defined as a) research skills and b) research communications measured as manuscript submissions, presentations, and publications. Both senior and junior researchers will be members of this collaborative and included in prospective qualitative and quantitative assessment.

Hypothesis 2: Participation within research communities will, over time, increase research productivity for both senior and more junior collaborative members relative to their productivity prior to study participation.

Expected outcomes:

(Aim 1) This study will identify components of bystander programs associated with greater violence prevention efficacy and greater cost-effectiveness. Currently colleges have limited data or experience to make informed decisions regarding selecting a bystander intervention, now required by SaVE. With the range of bystander program comparisons afforded by including up to 24 multiple college communities, investigators will have the ability to measure relative bystander program efficacy (which programs are associated with the greater reduction in violence), which programs are associated with lower program implementation costs, and the programs associated with the lower cost to violence reduction ratio.

(Aim 2) As both senior and junior researchers are engaged in college recruitment, mixed-method data collection, and data analyses, research skill-sets will be shared such that senior researchers learn newer research approaches and junior researchers learn skills in survey design and conduct, measurement and surveying on sensitive topics), data analyses, and research communication.

Benchmarks for Success will be measured: [Aim 1] as our ability to recruit 24 colleges, conduct key informant interviews and annual undergraduate surveys, and final analyses by Fall 2019; and [Aim 2] by our ability, as a collaborative, to recruit and retain VAW prevention researchers from identified colleges and provide mentoring and training needed to advance their research skills.

Inclusion Criteria at College level:

• Public universities with junior faculty to be mentored in violence prevention

Exclusion Criteria:

• Private universities, institutions with less than 10,000 undergraduate students

Inclusion Criteria at Student level:

• Undergraduate students at the public 4 year colleges or universities, ages 18-24

Exclusion Criteria:

• Graduate or professional students, non-degree seeking or not enrolled

Public universities with junior faculty to be mentored in violence prevention Note: sampling unit for primary analyses is the college: answers to questions below refer to students clustered at each institution