Study 1
Gilmore, Lewis, and George (2015) used a randomized controlled trial to assess the effectiveness of a web-based sexual assault risk reduction (SARR) intervention to reduce sexual assault and alcohol-related sexual assault outcomes in a sample of college women at the 3-month follow-up.
Participants were recruited from college-level introductory psychology courses and were eligible for the study if they were 1) female, 2) reported heavy episodic drinking (HED), or consumption of four drinks over a 2-hour period at least once in the past month, and 3) were from 18 to 20 years old. Participants completed the baseline screening survey online and were given course credit. At baseline, the full study sample were on average 18.8 years old. The majority of participants reported they were freshman (61.1 percent), were not members of a sorority (65.0 percent), lived on campus or in a sorority house (72.9 percent), and were not in a serious relationship (71.5 percent). Participants identified as 57.6 percent White, 20.5 percent Asian American/Pacific Islander, 14.1 percent multiracial, 9.5 percent Hispanic/Latina, 3.9 percent Black/African American, 2.9 percent other, and 1.0 percent Native American.
Individuals who began the follow-up survey (n = 207) were included in the analysis, whether or not they viewed their personalized feedback. Eligible participants were randomly assigned, stratified by sexual assault history, to either a minimal assessment (20 percent) or a full assessment (80 percent). Participants assigned to complete the full assessment were then randomly assigned, stratified by sexual assault history, to one of four conditions: 1) Alcohol Only Condition (n = 53 at randomization, n = 45 at the 3-month follow-up), 2) SARR Only Condition (n = 52 at randomization, n = 42 at the 3-month follow-up), 3) Combined Condition (n = 52 at randomization, n = 38 at follow-up), and 4) Full Assessment Only Control Condition (n = 54 at randomization, n = 40 at follow-up). There were no statistically significant differences between conditions on baseline characteristics. Those in the Alcohol Only, SARR Only, and Combined Condition received the web-based intervention immediately following the completion of the survey. All participants were contacted 3 months after completing the intervention to complete the follow-up survey. The CrimeSolutions review of this study focused on the SARR-only condition compared with the control condition (full assessment–only group).
Several measures of alcohol-related sexual assault, sexual assault, and alcohol use were employed to establish participant behavior and experiences, and for the condition randomization process. For alcohol-related sexual assault, items from the revised Sexual Experiences Survey (SES) assessed intoxicated attempted or completed rape at follow-up. Participants were asked to indicate how many times (0, 1, 2, or 3 or more) they had had attempted or completed penetrative sex by incapacitation in the past 3 months. Incapacitation refers to being unconscious, or passing out, because of alcohol; some men use this as a tactic for sexual assault (Gilmore et al. 2018). For risk perception of alcohol-involved rape, participants estimated (at baseline and at follow-up) the percentage likelihood that they would experience nonconsensual sex while being incapacitated by alcohol by a man that they knew while in college. Participants could enter numbers ranging from 0 to 100 to estimate percentage likelihood. Participants were also asked at baseline and follow-up how often they consumed alcohol before or during sexual activity, responding using a five-point scale (0 = never to 4 = always). This was an indicator of behavior in the past 3 months.
The SES was also used to assess participants’ experiences with sexual coercion at three time points: after their 14th birthday but before entering college (baseline), since entering college (baseline), and in the past 3 months (follow-up). Baseline experiences were combined. Participants indicated the number of times that a tactic or multiple tactics were used in sexual assault experiences including sexual contact, attempted penetration, and completed penetration (0 times, 1 time, 2 times, and 3 or more times). Sexual assault incidence and severity were scored using a 63-point scale (Davis et al. 2014) for each time point (baseline and follow-up), with high scores indicating more severe sexual assault experiences and zeros indicating no sexual assault experiences. Participants estimated the likelihood (in percentages) of experiencing verbally coerced nonconsensual sex by a man that they knew while in college at baseline and follow-up using an open-ended question.
Sexual assault protective behavioral strategies (PBS) were assessed using a revised version of the Dating Self-Protection Against Rape Scale. This consisted of 15 items for participants to indicate how often they engaged in behaviors (for example, “provide your own transportation” and “meet in a public place instead of a private place”) when they were with a date (and revised to include “or someone who is sexually interested in you”). Answer choices ranged on a five-point scale (1 = never to 5 = always). Scores were computed by creating an average of all items for baseline and follow-up.
Alcohol use was measured by drinks per week, HED frequency, drinking norms, and PBS. Participants indicated the number of drinks typically consumed each day of their average week in the past 30 days using the Daily Drinking Questionnaire at baseline. Average drinks per week were calculated by summing the drinks consumed each day. HED frequency was assessed by asking participants, “How often did you have four or more drinks containing any kind of alcohol within a 2-hour period” in the past month at baseline and in the past 3 months at follow-up. Answer choices ranged from zero times in the past month to every day. Participants were also asked 15 items from the Protective Behavioral Strategies measure, including while using alcohol or “partying” whether they engaged in certain behaviors (e.g. “determined not to exceed a set number of drinks,” “avoid mixing different types of alcohol,” and “know where your drink had been at all times”) with answer choices ranging on a five-point scale reverse scored (1 = never to 5 = always). Items were averaged for a total drinking PBS score at baseline and follow-up.
Hierarchical regressions were used to examine incapacitated attempted/completed rape frequency, alcohol-related sexual assault incidents and severity, and use of sexual assault protective behavioral strategies at the 3-month follow-up. No subgroup analysis was conducted.
