Study 1
Coker and colleagues (2011) used a randomized experimental research design to evaluate the impact of the Green Dot program. The implementation of Green Dot was assessed to determine whether students who participated in the program had increased observed and/or active-bystander behaviors, compared with those who did not participate in the program. The program was implemented by the Violence Intervention and Prevention (VIP) Center at the University of Kentucky (UK).
A random sample of 2,000 UK undergraduate students from each class (freshmen, sophomores, juniors, and seniors) was selected from the university’s registrar list for spring 2010. Half of this sample was male and all students were between the ages of 18 to 26. The random sample was stratified using student classification (freshmen, sophomores, juniors, seniors) and gender (male and female). A total of 7,945 UK undergraduate students with UK email addresses were selected to complete an online survey.
Of the 7,945 invited to participate, 3,417 students completed the survey; however, only 2,504 students were included in the analysis because those with missing data were excluded. There were no differences between the demographic profile of UK students included in the random sample and the whole UK undergraduate population, except that 100 percent of students in the sample were permanent Kentucky residents and 80 percent of UK’s student body were permanent residents. When comparing those included in the invited sample with those who completed the online survey, those completing the survey were significantly more likely to be female, freshmen, in a fraternity or sorority, and in a dating relationship. However, there were no statistically significant differences between those who completed the online survey and those included in the analysis. The final analytic sample was 60.5 percent female and 85.7 percent white, and 29.0 percent were freshmen.
The impact of the Green Dot program was evaluated using three treatment groups: 1) the Students Educating and Empowering to Develop Safety (SEEDS) group, who listened to the Green Dot motivational speech and participated in the bystander training (n = 351); 2) the VIP engaged group, who were VIP Center clients/volunteers without bystander training, (n = 159); and 3) the Green Dot speech alone group, who did not receive training but only participated in the first component (n = 693). There were no significant differences among the three treatment groups. The control group (no intervention) consisted of 1,301 students who had not received bystander training, were not clients/volunteers of the VIP Center, and had not heard a Green Dot speech. The CrimeSolutions review looked at the difference between the SEEDS treatment group and the no-intervention control group.
Actual active-bystander behaviors, such as “talked to a friend who was raped or hit by a partner” or “made sure someone who had too much to drink got home safely,” were measured using the modified Bystander Behaviors Scale (Banyard, Plante, and Moynihan 2005). Observed bystander behaviors were measured using the same list of behaviors, with instructions to respond on the frequency with which the student had seen or heard someone else do the active bystander behavior. The same timeframe was used to ask about actual and observed active-bystander behaviors (Fall 2009 to Spring 2010). The data were analyzed using a multiple analysis of variance (MANOVA). The study authors did not conduct subgroup analyses.
Study 2
Coker and colleagues (2017) conducted a cluster randomized controlled trial to determine the impact of the Green Dot program on sexual violence and related forms of interpersonal violence. In 2010, 26 high schools were randomized to either the treatment condition, which implemented Green Dot programming (n = 13), or to the control condition, which received no intervention (n = 13). At the end of the study, schools in both conditions had the option to continue or adopt the program at no cost.
At baseline, 54.6 percent of the students in intervention schools were female, and 44.3 percent were currently receiving free or reduced meals. A majority (82.5 percent) were white, and approximately one quarter (24.7 percent) reported witnessing parental intimate partner violence (IPV) in their lifetime. Of the students in the control schools, 54.1 percent were female, and 45.5 percent were currently receiving free or reduced meals. Most students (87.2 percent) were white, and 25.7 percent reported witnessing parental IPV in their lifetime. There were no statistically significant differences in baseline characteristics between treatment and control schools.
Outcomes of interest included sexual violence perpetration and victimization, sexual harassment perpetration and victimization, psychological dating violence perpetration and victimization, and physical dating violence perpetration and victimization. All outcomes were measured using a 99-item, self-reported questionnaire on violence perpetration and victimization in the past 12 months. This survey was distributed at baseline (Spring 2010) and once each year during the implementation of the program, from 2011 to 2014. The CrimeSolutions review looked at the differences between the treatment and control groups from the year 4 survey.
From baseline to year 4, a total of 106,867 students were present on days when surveys were taken, and 83.9 percent completed the surveys. Students who did not provide demographic or violence information were excluded (n = 9,427), as were students determined to be potential “mischievous responders” such as those who self-reported that they had not been sexually active but were pregnant or had children, or that they had been in multiple relationships in the past 12 months but had no relationship within the same time frame for different measures (n = 6,485). The final analytic sample included 73,795 survey responses across 5 years from 8,099 students in the treatment group and 8,143 students in the control group. Linear mixed models were used to evaluate condition-time interaction and provide mean estimates by each year. Subgroup analyses were conducted to determine the potential effect of gender on program outcomes.
Study 3
Bush and colleagues (2019) followed up on the prior analyses from Study 2 (Coker et al. 2017) to examine how the Green Dot program worked to reduce violence perpetration. The study sample and baseline characteristics were the same as those described in Study 2.
Using primary data collected in Study 2, the study authors conducted a primary linear mixed model to determine the difference between treatment and control groups in acceptance of violence, in bystander actions through engaging peers, and in overall bystander actions. Outcomes were measured through a self-report survey. Acceptance of violence was measured through 7 items on the Illinois Rape Myth Acceptance Scale and 5 items on the Acceptance of Couple Violence Scale. Bystander actions through engaging peers was measured through 5 items on the survey, asking the students how often they communicated with their friends in certain situations. Overall bystander actions was measured through 7 items on the survey, asking the students how often they observed and engaged in positive bystander behaviors. The study authors did not conduct subgroup analyses.
