Sexual Assault Prevention for Heavy Drinking College Men: Development and Feasibility of an Integrated Approach

Abstract

Despite the prevalence of sexual assault on college campuses, few interventions aimed at decreasing college men’s proclivity to perpetrate sexual aggression have been developed and tested. This article details the theoretical framework, content, and piloting of a sexual assault prevention program for college men who engage in heavy drinking, a high-risk group who may be particularly well positioned to intervene as proactive bystanders in drinking environments. In an open trial, male facilitators delivered the three-session Sexual Assault and Alcohol Feedback and Education (SAFE) program to 25 heavy drinking college men. Session 1 was a 90-min review of personalized normative feedback regarding alcohol use, sexual activity, alcohol-related sexual consequences, understanding of consent, and engagement in bystander intervention, delivered individually in a motivational interviewing style. Session 2 was a 2½-hr group-based sexual assault prevention workshop focusing on social norms, empathy, masculinity, consent, and bystander intervention. Session 3 was a 90-min booster group session that reviewed previous topics and included the active practice of bystander intervention skills. Analyses of postsession assessments of utility, therapeutic alliance, and satisfaction and examination of alcohol use and sexual assault–related outcomes from baseline to the 2-month assessment support the preliminary feasibility and acceptability of the SAFE program.

Rates of sexual violence are alarmingly high on college campuses (Krebs, Lindquist, Warner, Fisher, & Martin, 2007). Longitudinal studies indicate that between 30 and 35% of college men perpetrate sexual violence over the course of 4 years (White & Smith, 2004; Zinzow & Thompson, 2015), with many men perpetrating more than once (Gidycz, Warkentin, & Orchowski, 2007) without labeling their behavior as problematic (Edwards, Bradshaw, & Hinsz, 2014). Engaging men in sexual assault prevention, therefore, is vital to reducing rates of violence against women (Flood, 2011).

Public health approaches to preventing sexual violence focus on reducing perpetration of sexual aggression (McMahon, 2000), and include universal interventions as well as targeted interventions for individuals at high risk of perpetrating (Centers for Disease Control and Prevention [CDC], 2004). Over the past 30 years, numerous sexual assault prevention approaches have been developed and tested (see DeGue et al., 2014). Despite continued evolution in the field and the urgency to develop effective solutions, researchers and practitioners in the sexual violence field have been slow to produce efficacious programs (Tharp et al., 2011). The only sexual assault prevention approaches currently recognized as effective by the CDC (DeGue et al., 2014) include a 10-session curriculum for eighth- and ninth-grade youth (Safe Dates; Foshee et al., 2005), a six-session school-based intervention for sixth and seventh graders (Shifting Boundaries; Taylor, Stein, Mumford, & Woods, 2013), and improvements in state and local capacity for criminal justice enforcement and victim advocacy funded by the 1994 Violence Against Women Act (Boba & Lilley, 2009). There are currently no sexual assault prevention programs for college men that document long-term reductions in rates of sexual aggression in a rigorous program evaluation (DeGue et al., 2014). Although some programs for college men demonstrate short-term reductions in sexual aggression (Gidycz, Orchowski, & Berkowitz, 2011; Salazar, Vivolo-Kantor, Hardin, & Berkowitz, 2014), most sexual assault prevention programs for college men lack attention to basic principles of prevention and have been implemented without rigorous evaluation (Tharp et al., 2011). Furthermore, given that less than one third of existing sexual assault prevention approaches are geared toward male audiences (DeGue et al., 2014), research is needed to develop and rigorously evaluate sexual assault prevention programs for men that adhere to best practices in prevention (Nation et al., 2003).

There is also a need to more rigorously address the intersection between alcohol use and sexual violence in prevention programs. Alcohol use is considered to be an important component of sexual assault (Abbey, Wegner, Woerner, Pegram, & Pierce, 2014), with approximately half of all sexual assaults involving alcohol use by the victim and/or perpetrator (Abbey, Zawacki, Buck, Clinton, & McAuslan, 2004). Men’s alcohol use at the time of the assault is positively associated with the use of aggression (Parkhill, Abbey, & Jacques-Tuira, 2009) and with victim injury (Brecklin & Ullman, 2002). In addition, the most severe assaults often occur when perpetrators consume a moderate amount of alcohol, possibly because the impairments associated with high levels of alcohol use preclude completion of a rape, or because perpetrators limit their alcohol use to maintain control over the victim (Abbey, Clinton-Sherrod, McAuslan, Zawacki, & Buck, 2003). Despite these findings, sexual assault prevention approaches have yet to integrate evidence-based alcohol intervention, rigorously address alcohol use, or target heavy drinkers who may perpetrate or be in a position to intervene as proactive bystanders.

The complex interactions between alcohol use and sexual aggression suggest that a sexual assault prevention approach that also addresses alcohol use must be multifaceted in its approach. Men who engage in sexual aggression are a heterogeneous group who display various risk profiles (Tharp et al., 2013), utilize a range of perpetration tactics (Warkentin & Gidycz, 2007), and display varying patterns of perpetration over time (Abbey, Wegner, Pierce, & Jacques-Tiura, 2012; Swartout, Koss, White, & Thompson, 2015; Thompson, Swartout, & Koss, 2013). Sexually aggressive men also vary in their general use of alcohol (Abbey et al., 2014) and tendency to use alcohol at the time of an assault, with 48% perpetrating only when sober, 27% perpetrating only when intoxicated, and 25% perpetrating both when sober and when intoxicated (Parkhill & Abbey, 2008). These data highlight the importance of avoiding a “one size fits all” approach to addressing alcohol use within sexual assault prevention.

A sexual assault prevention program addressing alcohol as a risk factor should also consider the multiple pathways through which alcohol increases perpetration risk (Abbey, 2011). First, there are pharmacological explanations for why alcohol may increase risk. For example, the acute pharmacological effects of alcohol may contribute to sexual aggression by impairing judgment and decision making (Curtin & Fairchild, 2003), decreasing tension and anxiety (Greeley & Oei, 1999; Sayette, 1993), and hampering impulse control (Sher, Wood, Wood, & Raskin, 1996). Alcohol myopia theory suggests that alcohol use draws attention to the most salient cues in the environment, making it difficult to process complex cues (Steele & Josephs, 1990). As a result, when drinking, men who expect that an interaction with women will end in sexual activity may misinterpret women’s friendliness as a sign of sexual interest or disregard inhibitory cues indicating a partner’s disinterest (Farris, Treat, Viken, & McFall, 2008; Jacques-Tiura, Abbey, Parkhill, & Zawacki, 2007).

Expectancy theory offers a second explanation for why alcohol increases perpetration risk. Specifically, men who believe that alcohol use will increase their level of aggression (Goldman, Darkes, & Del Boca, 1999) may also consume alcohol to justify sexually aggressive behavior (George, Stoner, Norris, Lopez, & Lehman, 2000). Although these findings would seem to suggest that reducing men’s alcohol use might lead to decreases in sexual aggression among some men, other research suggests that reducing men’s alcohol use—without addressing other risk factors for sexual aggression—would be insufficient to prevent sexual assault. Specifically, although some studies document greater alcohol consumption and alcohol-related problems among sexually aggressive men compared with their nonsexually aggressive peers (Abbey & McAuslan, 2004; Borowsky, Hogan, & Ireland, 1997; Zawacki, Abbey, Buck, McAuslan, & Clinton-Sherrod, 2003), other studies fail to support these findings (Loh, Gidycz, Lobo, & Luthra, 2005; Testa & Cleveland, 2017; Thompson, Kingree, Zinzow, & Swartout, 2015).

The aforementioned inconsistency in the person-level association between alcohol use and sexual aggression supports a third pathway linking alcohol use and sexual aggression. Specifically, there are numerous shared personality characteristics among heavy drinking and sexually aggressive men, including impulsivity, sensation seeking, and antisocial personality characteristics (Lansford, Rabiner, Miller-Johnson, Golonka, & Hendren, 2003; Testa et al., 2015; Thompson et al., 2015, 2013), suggesting that these men will perpetrate regardless of alcohol use. This third pathway is also highlighted by the shared associations between alcohol use and risk of sexual aggression, including rape myth acceptance, impersonal sexual activity, and dominance over women (Locke & Mahalik, 2005).

The co-occurrence of drinking and pursuit of sexual partners in party and bar environments also confers considerable risk of sexual violence (Testa & Cleveland, 2017). College students report attending parties and visiting bars to consume alcohol and seek sexual partners (Grazian, 2007; Lindgren, Pantalone, Lewis, & George, 2009), and sexually aggressive behavior is common in drinking environments (Graham et al., 2014, 2006). Feeling pressure to compete with other men for sexual partners (Graham, Wells, Bernards, & Dennison, 2010), sexually aggressive men may seek out drinking environments to target women who they perceive to be open or vulnerable to sexual advances (Armstrong, Hamilton, & Sweeney, 2006; Parks & Zetes-Zanatta, 1999). The commonality of coercive sexual advances in drinking environments may normalize men’s engagement in sexual aggression (Becker & Tinkler, 2015), thereby reducing the likelihood that individuals will act as proactive bystanders when they witness sexually aggressive behavior (Oesterle, Moreno, & Orchowski, 2017). Although relatively little is known about the association between alcohol use and bystander intervention, heavy drinking men who hold traditional beliefs about masculinity are unlikely to address the sexually aggressive behavior of other men (Orchowski, Berkowitz, Boggis, & Oesterle, 2016). Accordingly, it may be especially important to train heavy drinking men who are not inclined to be sexually aggressive—but who frequent drinking environments such as parties and bars—on how best to intervene as proactive bystanders against the coercive aggressive behavior among their peers.

Drawing from the aforementioned research, this study advances the science of sexual assault prevention by presenting the design and feasibility results from an open trial of a sexual assault prevention program for heavy drinking college men. The three-session Sexual Assault and Alcohol Feedback and Education (SAFE) program includes an individually administered personalized feedback report (PFR), group-based sexual assault prevention workshop, and booster session. This program was created by adapting a successful two-session sexual assault prevention program for college men (Gidycz et al., 2011), and integrating it with personalized normative feedback for alcohol use, sexual activity, consent, and bystander intervention adapted from the Brief Alcohol Screening and Intervention for College Students (BASICS; Dimeff, Baer, Kivlahan, & Marlatt, 1999) and motivational interviewing (MI; Miller & Rollnick, 2013). These approaches to alcohol intervention and sexual assault prevention share a nonjudgmental approach to challenging misperceived norms.

This open trial of the SAFE program represented Stage Ia of a larger treatment development study (see Onken, Blaine, & Battjes, 1997; Rounsaville, Carroll, & Onken, 2001). Following Leon, Davis, and Kraemer’s (2011) recommendations for the appropriate interpretation of small-scale nonrandomized pilot studies, the goal of this research was to examine the preliminary feasibility and acceptance of SAFE with a focus on markers of utility, fidelity, and preliminary program outcomes. We hypothesized that the SAFE program would be conducted with a high level of fidelity, and would show high retention rates and participant satisfaction. A series of exploratory analyses was conducted to examine domains of change that we thought would be most likely within SAFE. Specifically, we expected that participants would report postprogram increases in motivation and confidence to reduce their alcohol use and lower drinking intentions. At the 2-month assessment, we also examined the effects on (a) alcohol use and related consequences, and strategies to limit drinking; (b) attitudes commonly associated with sexual aggression (rape myth acceptance, hypergender ideology, labeling of consent); (c) perceptions of peer norms regarding alcohol use and sexually aggressive behavior; and (d) bystander intervention perceived norms, intentions, and confidence.

Method

Participants and Procedures

Participants included 25 men between the ages of 18 and 22 enrolled at a large north-eastern university. Men were included in the study if they exceeded the national recommended limits for daily alcohol use (five or more drinks for men) more than once in the past month and engaged in sexual activity (oral, vaginal, or anal sexual intercourse) with a female partner in the past 4 months. Men were excluded if they met criteria for antisocial personality disorder (ASPD), displayed symptoms consistent with current substance use withdrawal, or reported current suicidal or homicidal ideation.

All procedures were approved by the participating university and hospital institutional review boards, and a Certificate of Confidentiality was received from the National Institutes of Health. A multistep enrollment and screening process was utilized. Using a list of men between the ages of 18 and 22 from the university registrar, a random sample of 1,600 students was selected to receive an email invitation that contained a link to the participant in a confidential online screening for research addressing alcohol and dating experiences among college men. The online screening survey utilized a secure https connection with 128-bit encryption and a signed Secure Sockets Layer (SSL) certificate. Participants agreed to an electronic consent statement prior to enrolling in the screening, which consisted of the Graduated Frequency Measure (Hilton, 1989) assessing past-month drinking and a single item assessing the number of female sexual partners in the past 4 months. One in every 50 participants was randomly selected to receive a US$50 gift card.

Of the 160 men who completed the online screening, 82 (51.3%) met the alcohol use and sexual activity inclusion criteria and were contacted via phone and invited to an in-person screening. Twenty-seven men (55.1%) attended the in-person screening and provided informed consent to participate. The interviewer administered the Alcohol Use Withdrawal Symptom Checklist (Pittman et al., 2007), a single item indicating suicide risk on the Beck Depression Inventory (Beck, Steer, & Brown, 1996), an item identifying homicidal ideation, and the ASPD module of the Structured Interview for Diagnostic and Statistical Manual of Mental Disorders (4th ed.; DSM-IV; American Psychiatric Association, 1994) Personality (Pfohl, Blum, & Zimmerman, 1997). Two men were excluded from the study: one as a result of reporting suicide risk and one as a result of reporting criteria consistent with ASPD. The 25 eligible participants were invited to participate in a program designed to promote healthy dating and sexual experiences. All 25 men provided informed consent and enrolled.

The mean age of the study sample was 19.44 years (SD = 1.33 years). Ninety-two percent self-identified as Caucasian (N = 23), 4% as Asian (N = 1), and 4% declined to answer (N = 1). No men reported their ethnicity as Hispanic or Latino. For relationship status, 52% were dating casually (N = 13), 36% were in a long-term dating relationship (N = 9), and 12% were not dating (N = 3). Thirty-six percent (N = 9) were a member of a social fraternity and 16% (N = 4) were a current or prior member of a college athletic team.

Theoretical Foundation and Evidence Base

Session 1 of SAFE was adapted from brief motivational interventions with personalized feedback to address alcohol use (Dimeff et al., 1999; Miller & Rollnick, 2013). This approach aims to correct participants’ overestimations of the extent to which people of their age and gender consume alcohol, an important predictor of personal drinking behavior (Korcuska & Thombs, 2003; Pederson & LaBrie, 2008). Feedback is delivered in a nonconfrontational manner aimed at revealing discrepancies between drinking behavior and personal values, exploring ambivalence toward change, examining readiness to change, and resolving any discrepancies between actual and ideal behavior (Prochaska, DiClemente, & Norcross, 1992). As the efficacy of brief motivational interventions (Barnett, Murphy, Colby, & Monti, 2007; Borsari et al., 2012) and personalized feedback (Larimer & Cronce, 2007; Walters & Neighbors, 2005) in reducing college student drinking is now well documented, this approach was ideal for integration with sexual assault prevention programming.

Session 2 and Session 3 of SAFE were adapted from the Men’s Workshop (Berkowitz, Lobo, Gidycz, Robison, & Zimak, 2006), a sexual assault prevention program for college men grounded in the Integrated Model of Sexual Assault (Berkowitz, 1994, 2003). This model suggests that a perpetrator’s attitudes, early experiences, and perception of peer norms interact to facilitate sexual aggression (Berkowitz, 2003). More specifically, misperceptions of social norms pressure men to be sexually active when they do not want to, engage in coercive behavior to garner sexual activity, and suppress their discomfort with other men’s inappropriate behavior (Berkowitz, 2003). Accordingly, the Men’s Workshop aims to debunk misperceptions of the extent to which other men engage in sexual activity (Lynch, Mowrey, Nesbitt, & O’Neil, 2004), support sexual aggression (Bohner, Siebler, & Schmelcher, 2006; Dardis, Murphy, Bill, & Gidycz, 2016), approve of impersonal sexual activity (Lambert, Kahn, & Apple, 2010), and disapprove of bystander intervention (Brown & Messman-Moore, 2010). When evaluated among 635 college men, participants in the Men’s Workshop reported less exposure to pornography, fewer associations with sexually aggressive peers, and lower rates of sexual aggression over 4 months in comparison with controls (Gidycz et al., 2011).

Overview of the SAFE Program

The development of SAFE followed a Stage Ia approach (Rounsaville et al., 2001). Modifications were made to adapt and integrate the existing intervention approaches based on reviews of the literature, the expertise of the investigator team, informant interviews (N = 12) and focus groups (N = 30) with college men, and interviews with campus administrators (N = 6; see Orchowski et al., 2014). The program was modified to ensure the developmental appropriateness of language, relevant handouts were designed, and new scenarios were developed. Normative data from a cross-sectional survey of men at the study site (N = 242) was also incorporated.

The resulting Sexuality and Alcohol Feedback and Education Program (SAFE) intervention is 5½ hr in length. The program is manualized and includes two core sessions and a booster session. Session 1 is a 90-min individual interview conducted with MI style that reviews personalized feedback on the intersection of alcohol use, sexual activity, alcohol-related risks/consequences, sexual consent, and bystander intervention. Session 2 is a 2½-hr group-based sexual assault prevention workshop targeting misperceived norms, masculinity, empathy for survivors, consent, and bystander intervention. Content addressing the intersection of alcohol use and sexual violence is integrated throughout the workshop. Session 3 is a 90-min booster session review of program material, which includes additional active practice of sexual communication and bystander intervention skills. Content addressing pathways through which alcohol use increases perpetration risk (e.g., alcohol myopia, misperception of sexual interest, alcohol expectancies, intoxication and capacity to consent, and the influence of alcohol and drinking environments on helping behavior) is integrated across the sessions (see Table 1 for intervention targets).

Table 1.

Intervention Targets Addressing Alcohol and Perpetration Risk.

Targeting the pharmacological effects

Enhance utilization of strategies to limit drinking.

Enhance protective behavioral strategies to reduce alcohol-related consequences.

Increase awareness of alcohol myopia as an influence on sexual behavior and bystander intervention.

Increase awareness of the likelihood to misperceive sexual intent, especially if a woman is drinking.

Targeting alcohol expectancies

Increase awareness of personal expectancies about alcohol. Specifically, do participants expect to feel more powerful, sexual, and aggressive after drinking?

Increase awareness of how expectancies influence sexual behavior and bystander intervention.

Debunk the use of alcohol as a justification for men’s sexual violence. Specifically, do participants drink purposely to experience the positive outcomes they expect from drinking—namely, a sense of disinhibition, increased sexual ease, and power? Do participants “write off,” discount, or brag about unacceptable acts they committed the night before by using alcohol as an excuse?

Targeting shared influences on alcohol use and aggression

Increase awareness of how alcohol use is associated with perceptions of traditional masculine norms.

Debunk misperceptions of the extent to which other men engage in sexual activity when intoxicated.

Reduce adherence to rape myths.

Increase empathy toward victims of sexual violence.

Session 1—Individually administered personalized feedback.

Session 1 is offered before the group-based sexual assault prevention workshop to promote men’s ability to personally relate to material in the prevention workshop. Participants are provided with a PFR on their alcohol use, sexual activity, alcohol-related risks and consequences, utilization of consent, and bystander intervention behavior. To maintain an explicit focus on the intersection between alcohol use and sexual activity, the PFR does not include some of the components commonly included in personalized feedback interventions (e.g., description of caloric intake, financial expenditures on alcohol, biphasic alcohol curve). The session is conducted by a single male interventionist. The interventionist first establishes rapport through empathic, concerned, nonauthoritarian, and nonjudgmental conversation regarding alcohol use, dating, and sexual activity. Participants next generate salient (positive and negative) aspects of alcohol use and its role in sexual activity and talk about the effects that matter most to them. The interventionist collaboratively reviews the norm-based and individualized feedback on alcohol use (e.g., quantity/frequency, blood alcohol level) and the intersection between alcohol use and sexual activity. Risk factors for alcohol use problems (e.g., risk of alcohol problems, tolerance, dependence, consequences) are presented. Feedback is provided regarding the participant’s current strategies for garnering sexual consent, confidence in garnering sexual consent when sober and when drinking, and engagement in proactive bystander intervention behavior when sober and when drinking. The session concludes with the establishment of a change plan and delineation of potential change strategies (see Table 2).

Table 2.

Session 1 Components and Aims.

Component	Aim
Rapport building	Establish rapport through empathic, concerned, nonauthoritarian, and nonjudgmental conversation.
Exploration of participant behaviors	Gather information about participant’s alcohol use and sexual behavior. Outline how alcohol use is involved in the participant’s recent sexual activity.
Identify pros/cons	Help the participant to identify both positive and negative aspects to alcohol use, sexual behaviors, and sexual activity involving alcohol. Highlight discrepancies between these behaviors and their goals and values.
Provide personalized feedback	Presented personalized information on
	Personal alcohol use in relation to their peers
	Co-occurrence of alcohol use and sexual activity
	Blood alcohol level (BAL; i.e., average, peak, average during sexual activity, peak during sexual activity)
	General consequences and risks of drinking
	Alcohol-related sexual consequences
	Sexual communication and consent
	While sober vs. while intoxicated
	Discuss potential effects of expectancies and alcohol myopia
	Utilization of bystander intervention strategies
	While sober vs. while intoxicated
	Discuss potential effects of expectancies and alcohol myopia
Discuss motivation to change	Attempt to elicit change talk for drinking and sexual behaviors and enhance participant commitment to change.
Discuss barriers to change	Increase self-efficacy for change by discussing potential barriers that might serve as roadblocks during the change process.
Discuss strategies for change	Provide guidance to participant in setting goals for reducing problems related to drinking, and its role in sexual activity, consent, and/or bystander intervention.

Session 2—Sexual assault prevention workshop.

Session 2 is delivered in a nonjudgmental manner by two male facilitators. The broad change strategies include (a) understanding the conditions of sexual consent, (b) increasing men’s empathy regarding the effects of sexual assault, (c) correcting misperceptions regarding masculinity and sexual behavior through normative feedback, and (d) increasing proactive bystander intervention (see Table 3).

Table 3.

Session 2 and Session 3 Components and Aims.

Component	Aims
Session 2
Definitions and facts	Increase men’s knowledge on the prevalence of sexual assault on college campuses.
Social norms	Correct men’s misperceptions regarding sexual behavior and alcohol use through the provision of social norms.
False accusations of sexual assault	Correct men’s misperceptions regarding the frequency of false accusations of sexual assault, and to increase men’s empathy regarding the effects of sexual assault.
Sexual communication and consent	Provide information and create discussion to help men in understanding the conditions of sexual consent.
Bystander intervention	Highlight ways men can increase their awareness of risk of sexually aggressive behavior among peers, discuss personal use of bystander intervention strategies, and practice strategies to intervene when witnessing dating and/or sexual violence.
Session 3
Utilization of program content	Discuss understanding of sexual assault on campus since participating in the program and create a discussion for how men have noticed applying program content in their lives since participating.
Sexual communication and consent	Create a discussion on men’s perception of the consent model and alcohol’s influence in sexual situations.
Correct misperceived social norms	Create discussion on norms supportive of sexual violence, including perceptions of traditionally masculine behaviors.
Bystander intervention	Discuss personal engagement in bystander intervention over the interim. Role-play a bystander intervention scenario and garner feedback on this approach from men.
Small group practice	Present scenarios to men and have them identify verbal and nonverbal strategies to intervene.

Session 3—Booster session.

Session 3 is a booster session review of program material delivered by two male facilitators. The session provides a review of the prevalence of sexual violence, the conditions for consent, and healthy norms regarding relationships and sexual activity. Participants are encouraged to share the ways in which they utilized the program material over the 2-month interim. The facilitators also engage participants in small group practice of bystander intervention strategies (see Table 3).

Training and Fidelity

One SAFE facilitator was a recent college graduate, and the second facilitator was a doctoral student in a clinical psychology training program. Training was standardized and administered by the first and third authors. Following training, each facilitator delivered three mock sessions, which were reviewed for adherence to the protocol and competency in facilitating using a nonjudgmental style. All sessions were digitally recorded and reviewed for supervision purposes. Session integrity was evaluated by an external rater using checklists of core content. Utilization of an MI-consistent style was assessed with the Motivational Interviewing Treatment Integrity 3.0 (MITI 3.0), a well-established method of assessing the integrity of brief motivational interventions (Moyers, Martin, Manuel, Miller, & Ernst, 2007). Ongoing supervision was provided by the first author, and MITI 3.0 ratings were provided by the fourth author.

Study Administration

Study sessions were conducted in private individual or group interview rooms in the psychology training clinic of the university. After completing the in-person screening and study consent, participants completed the 30-min baseline survey on a laptop computer; some measures were interviewer administered and some were self-administered. Participants were allowed a break after the completion of the baseline survey, during which the facilitator printed the PFR, which was programmed to generate automatically upon completion of the survey. After Session 1, participants completed pencil-and-paper questionnaires assessing program satisfaction, utility, and alliance. To minimize the influence of demand characteristics, participants were instructed to answer honestly and return the forms to the facilitator in a sealed envelope. Men were informed that the facilitator would not see their responses. Participants were provided with handouts, a list of resources, and a copy of the PFR. All procedures—including the in-person screening, informed consent process, baseline survey, Session 1, and postsession questionnaires—were completed in 3½ hr, and men were compensated US$40. Participants returned approximately 2 weeks later for Session 2, after which they again completed pencil-and-paper questionnaires. Study activities were completed in 3 hr, and men were compensated US$45. Men completed the 2-month assessment via an online survey prior to attending Session 3. After the session, participants completed the same set of postsession questionnaires. Study activities were completed in 2 hr, and men were compensated US$50. A 60-min exit interview was scheduled within 2 weeks of Session 3, and included a pencil-and-paper survey and semistructured interview facilitated by a male research assistant (RA). Participants received US$30 for the interview, and a bonus of US$30 if they completed all study components.

Measures

Participant characteristics.

Demographic characteristics were collected on a brief questionnaire at baseline.

Feasibility.

Program feasibility was evaluated with session attendance rates and program retention rates.

Satisfaction and acceptability.

At the completion of each session of SAFE, men completed the Client Satisfaction Questionnaire–8 (CSQ-8; Larsen, Attkisson, Hargreaves, & Nguyen, 1979, which includes eight items rated on a 4-point scale. Cronbach’s alpha was .82, .85, and .77 for Session 1, Session 2, and Session 3, respectively. At the exit interview, men completed a six-item Likert-type assessment of the SAFE program, which included satisfaction with the program, the number of sessions, and the facilitators; their likelihood to seek a program such as SAFE in the future; and the extent to which the program met their needs. Participants also completed assessments of the perceived usefulness of program components after each session using Likert-type scales adapted from Magill, Apodaca, Barnett, and Monti (2010). The survey included six items for Session 1 and two items for Session 2 and Session 3. Responses ranged from 0–3 (0 = topic not discussed, 1 = not useful, 2 = useful, 3= very useful).

Intervention fidelity.

Facilitator fidelity to the program protocol was assessed with adherence checklists, aligned with the content of each session of the SAFE program. Ratings were completed by an outside rater. Specifically, 29 items were rated for Session 1, 43 items were rated for Session 2, and 14 items were rated for Session 3. All sessions were recorded and rated. Adherence to ≥80% of the session content was deemed acceptable. Global ratings on the MITI 3.0 (Moyers et al., 2007) were utilized to establish competency in using MI style. The global rating consists of five behaviorally anchored ratings on 1–5 scale (1 = low, 5 = high). Domains assessed include the extent to which the facilitator prioritizes the participant’s reasons for change, collaborates with the participant, emphasizes autonomy, directs attention toward change, and displays evidence of understanding the participant’s point of view. The first two sessions for each facilitator were rated to establish competency. After each session, participants also indicated the extent to which the facilitator utilized a nonjudgmental style using a nine-item Likert-type scale (Magill et al., 2010). Item responses ranged from 1–4 (1 = strongly disagree, 4 = strongly agree). Cronbach’s alpha was .69, .79, and .83 for Session 1, Session 2, and Session 3, respectively.

Session 1 outcomes.

It was hypothesized that the personalized feedback would result in (a) increased motivation to change, (b) increased self-efficacy for reducing drinking, and (c) decreased personal drinking intentions. Assessments were competed at baseline and immediately following Session 1. The Contemplation Ladder (Biener & Abrams, 1991) is a single-item assessment of motivation to change drinking, with responses ranging from 0 (no thought of drinking less) to 10 (taking action to drink less). The Brief Situational Confidence Questionnaire (Breslin, Sobell, Sobell, & Agrawal, 2000) is a single-item assessment of an individual’s confidence in resisting drinking heavily in the future, with responses ranging from 0% (not at all confident) to 100% (completely confident). A weekly calendar was utilized to assess drinking intentions (LaBrie, Pederson, Earleywine, & Olsen, 2006). Participants were instructed to think about what their drinking pattern would be like over the next week, and then asked to enter the average number of drinks they planned to consume each day of the week. Responses were summed to represent participants’ estimation of the total number of drinks they intended to consume in the next week. Cronbach’s alpha was .73.

Alcohol use outcomes.

Change in the quantity and frequency of alcohol use, alcohol-related consequences, strategies to limit drinking, and perceived peer drinking norms were examined. All assessments were completed at baseline and 2 months. An interviewer administered the Time Line Follow Back (TLFB; Sobell & Sobell, 1992) measure, which assessed the number of standard drinks per day over the past 4 weeks, from which the average number of drinks per week and the number of heavy drinking days in the past month (five or more drinks for men) were derived. The Brief Young Adult Alcohol Consequences Questionnaire (Kahler, Strong, & Read, 2005) assessed experience of 24 possible alcohol-related consequences, to which participants respond “yes” or “no.” A summary score ranging from 0–24 was created to reflect the number of different consequences experienced over the past year (at baseline) and over the past 2 months (at the follow-up). Cronbach’s alpha was .86. Strategies to limit drinking were assessed with the Self-Control Questionnaire (Werch & Gorman, 1986). The extent of protective behavioral strategy use over the past 6 weeks was assessed at baseline, and use over the follow-up period was assessed at 2 months. Responses were provided along a scale of 1–5 (1 = never, 5 = always). Cronbach’s alpha was .91. The Drinking Norms Rating Form (Baer, Stacy, & Larimer, 1991) assessed perception of peer drinking norms. Participants estimate the alcohol consumption of typical same age and gender peers on each day of the week. Responses were summed to create an indicator of the perceived norm for other men’s weekly alcohol consumption. Cronbach’s alpha was .82.

Sexual aggression outcomes.

Engagement in sexual aggression was assessed at baseline (since the age of 14) and 2 months via the Sexual Experiences Survey–Short Form Perpetration (socioeconomic status [SES]-SFP; Koss et al., 2007). Participants indicate whether they used one of five tactics to bring about unwanted sex (i.e., verbal pressure, criticism, taking advantage of someone too drunk to stop it, threats of harm, force), or to attempt or complete seven different unwanted sexual acts, ranging from unwanted contact to penetration. Endorsement of rape myths was assessed with the short form of the Illinois Rape Myth Acceptance Scale (Payne, Lonsway, & Fitzgerald, 1999). Each of the 20 items are rated from 1–7 (1 = not at all agree, 7 = very much agree). Cronbach’s alpha was .81. Adherence to traditional beliefs about masculinity was assessed using the 19-item short form of the Hypergender Ideology Scale (Hamburger, Hogben, McGowan, & Dawson, 1996). Items are rated from 1–6 (1 = strongly disagree, 6 = strongly agree). Cronbach’s alpha was .88. Men’s labeling of consensual sexual activity was assessed with a scenario depicting the perpetration of sexual aggression (Pinzone-Glover, Gidycz, & Jacobs, 1998). Participants indicated the extent to which the scenario would be considered consensual sex (1 = consensual sex, 10 = rape). Perception of peers’ comfort with sexism and engagement in coercive sexual behavior was assessed with the two subscales of the Sexual Social Norms Inventory (Bruner, 2002). Items are rated from 1–5 (1 = strongly agree, 5 = strongly disagree), and the subscales were averaged and reverse coded so that higher responses reflected the belief that peers were more comfortable with sexism or engaging in more coercive sexual behavior. Cronbach’s alpha was .80 for both subscales.

Bystander intervention outcomes.

Perceptions of peer engagement in bystander intervention were assessed with a 20-item scale (Banyard, Moynihan, Cares, & Warner, 2014). Respondents indicate how likely their friends would be to engage in a range of bystander intervention behaviors, such as “ask a stranger if they need to be walked home from a party or get their friends to do so.” Items are rated from 1–5 (1 = not at all likely, 5 = very much likely). Cronbach’s alpha was .79. The 10-item Brief Intent to Help Friends and eight-item Intent to Help Strangers scales assessed confidence in engaging in bystander intervention (Banyard et al., 2014). Participants rate their confidence in performing each task ranging from 0–100 (0 = definitely cannot do, 100 = very certain can do). Cronbach’s alpha values were .81 and .92, respectively. The 51-item Bystander Attitudes Scale assessed likelihood to intervene in a risky situation (Banyard, Moynihan, & Plante, 2007). Items are rated from 1–5 (1 = not at all likely, 5 = extremely likely). Cronbach’s alpha was .93.

Results

Sample Characteristics

Participants reported 2.59 female sexual partners on average in the past 4 months (SD = 1.73), an average of 6.68 drinks (SD = 1.93 drinks) per drinking day, and 7.12 (SD = 3.42) heavy drinking days over the past month. Since the age of 14, 52% of the men reported engaging in some form of sexual aggression (N = 13), including unwanted sexual contact (4%, N = 1), sexual coercion (28%, N = 7), attempted rape (12%, N = 3), or rape (8%, N = 2). The majority of men who perpetrated did so more than once (69.2%, N = 9). When examining men’s most severe assault reported, 61.5% (N = 8) involved victim alcohol and/or drug use and 61.5% (N = 8) involved perpetrator alcohol use and/or drug use. Furthermore, 15.4% (N = 2) of men reported no relationship with the victim, 23.1% (N = 3) identified the victim as a friend/acquaintance, and 61.5% (N = 8) identified the victim as a steady date, girlfriend, or ex-girlfriend.

Acceptability and Utility

Of the 25 men who completed the baseline assessment and Session 1, 20 (80%) participated in the sexual assault prevention workshop (Session 2), 20 completed the booster session (Session 3), and 20 were retained at the 2-month assessment. Nineteen completed the exit interview (see Figure 1). Three workshops and three booster sessions were conducted with an average of six to seven men per group. Participants rated the six components of Session 1 as useful to very useful,” with mean ratings of session components ranging from 2.08 (SD = 0.76) to 2.84 (SD = 0.37). Ratings of the two main components of Session 2 and Session 3 were also high, with ratings ranging from 2.60 (SD = 0.75) to 2.75 (SD = 0.44; see Table 4).

Figure 1.

Participant recruitment and retention. Note. ASPD = antisocial personality disorder; SAFE = Sexual Assault and Alcohol Feedback and Education.

Table 4.

Utility of Program Components.

	Session 1 N = 25		Session 2 N = 20		Session 3 N = 20
	M	SD	M	SD	M	SD
Pros of drinking	2.36	0.7	—	—	—	—
Cons of drinking	2.64	0.49	—	—	—	—
Information of drinking norms	2.84	0.37	—	—	—	—
Information on consequences of drinking	2.52	0.77	—	—	—	—
Information on blood alcohol content	2.52	0.65	—	—	—	—
Information on personal risk factors	2.08	0.76	—	—	—	—
Ways to intervene in risky situations	—	—	2.60	0.75	2.60	0.75
Strategies for consent in sexual relationship	—	—	2.60	0.75	2.75	0.44

Note. Responses are shown from all men who completed the session. Responses range from 0–3 (0 = topic not discussed, 1 = not useful, 2 = useful, 3 = very useful).

Program Satisfaction

Among men completing Session 1 (N = 25), Session 2 (N = 20), and Session 3 (N = 20), satisfaction ratings on the CSQ-8 were 28.88 (SD = 2.83), 29.50 (SD = 2.59), and 30.20 (SD = 2.07), respectively, reflecting high scores on this 32-point scale. At the exit interview, 84.2% (N = 16), 73.7% (N = 14), and 94.7% (N = 18) indicated they were very satisfied with the program, the number of sessions, and the facilitators, respectively. Most participants (89.5% and 73.7%) indicated that they were likely to seek a program such as this for sexual relationships and alcohol use. All participants indicated that most (47.4%) or all (52.6%) of their needs had been met by the SAFE program (see Table 5).

Table 5.

SAFE Program Satisfaction at the Exit Interview (N = 19).

	N	%	N	%	N	%
Satisfaction ratings	Less than moderately satisfied		Moderately satisfied		Very satisfied
Overall, how satisfied were you with the program?	0	0	3	15.8	16	84.2
How satisfied are you with the number of sessions?	0	0	5	26.3	14	73.7
How satisfied are you with the facilitators?	0	0	1	5.3	18	94.7
Would you seek a program such as this in the future for	Probably no		Maybe		Yes
. . . your sexual relationships?	0	0	2	10.5	17	89.5
. . . your alcohol use?	2	10.5	3	15.8	14	73.7
To what extent did the program meet your needs?	<Most of my needs have been met		Most of my needs have been met		All my needs have been met
	0	0	9	47.4	10	52.6

Note. Overall program satisfaction was rated at the postprogram exit interview. Responses from men who completed Session 1, Session 2, and Session 3 are shown (N = 19). SAFE = Sexual Assault and Alcohol Feedback and Education.

Program Feasibility/Fidelity

Each administration of Session 1, Session 2, and Session 3 in the open trial was rated as adherent to protocol (≥80% of content included). On average, 93% of content within Session 1 was administered according to the protocol. In Session 2 and Session 3, review of recorded sessions indicated that 100% of content was administered according to protocol. Examination of MITI 3.0 global scores suggested that both facilitators demonstrated competency in the spirit of MI (average ratings on global measures ≥4). Participant mean ratings of alliance at Session 1, Session 2, and Session 3 were 33.68 (SD = 2.34), 33.10 (SD = 2.97), and 33.80 (SD = 2.59), respectively, suggesting that facilitators were successful in conveying the information in a nonjudgmental style (see Table 6).

Table 6.

Facilitator Alliance—Utilization of a Nonjudgmental Style.

	Session 1 N = 25		Session 2 N = 20		Session 3 N = 20
The facilitator(s)	M	SD	M	SD	M	SD
. . . was/were easy to talk to	3.92	0.28	3.90	0.31	3.95	0.22
. . . was/were concerned about me	3.16	0.90	3.20	0.83	3.25	0.71
. . . understood me	3.80	0.41	3.80	0.41	3.70	0.57
. . . asked my ideas before presenting his own	3.80	0.50	3.75	0.44	3.90	0.31
. . . helped me talk about my own reasons for change	3.44	0.65	3.25	0.91	3.55	0.51
. . . treated me like an equal	3.92	0.28	3.85	0.37	3.95	0.22
. . . respected my ideas about how change can occur	3.88	0.33	3.80	0.41	3.85	0.37
. . . did not push me into something I was not ready for	3.96	0.20	3.80	0.41	3.85	0.37
. . . accepted that I might choose not to change	3.80	0.41	3.75	0.44	3.80	0.41
Total score	33.68	2.34	33.10	2.97	33.80	2.59

Note. Item responses range from 1–4 (1 = strongly disagree, 4 = strongly agree), with total scores ranging from 9–36.

Session 1 Outcomes

A series of paired-sample t tests examined immediate postsession change in motivation to change drinking, confidence in resisting drinking heavily in the future, and weekly drinking intentions among the 25 men who completed Session 1. From baseline to posttest at Session 1, men intended to drink fewer drinks per week, t(21) = 4.09, p < .01, and reported increased motivation to change their alcohol use, t(24) = −3.50, p < .01 (see Table 7).

Table 7.

Session 1 Outcomes: Motivation to Change, Self-Efficacy, and Intentions to Drink (N = 25).

	Baseline		Session 1 posttest
	M	SD	M	SD	T	p
Outcome
Self-efficacy in changing drinking (N = 25)	70.40	31.22	74.84	29.01	−1.24	.22
Motivation to change drinking (N = 25)	2.48	2.50	4.12	3.30	−3.50	.002
Intended total drinks per week (N = 22)	21.64	10.55	16.68	9.18	4.09	.001

Note. Responses are shown from all men who completed Session 1 and provided complete data on the scales.

Alcohol Use and Sexual Aggression Outcomes at 2 Months

A series of paired-sample t tests examined program effects on alcohol and sexual assault outcomes among men who completed the baseline and 2-month survey (see Table 8). At 2 months, men reported increased use of strategies to limit drinking, t(19) = −2.12, p < .05, and fewer alcohol-related consequences, t(18) = 2.84, p < .05. Men reported less endorsement of rape myths at 2 months, t(19) = 2.10, p < .05, and lower perceptions of peer alcohol use, t(19) = 4.08, p < .01, and engagement in sexual coercion, t(18) = 2.78, p < .05. Men also reported greater likelihood of bystander intervention, t(19) = 2.31 p < .05, and greater confidence in intervening to help a stranger at risk of violence, t(19) = 2.11, p < .05.

Table 8.

Alcohol Use Outcomes and Sexual Aggression at Baseline and 2 Months (N = 20).

	Baseline		2 months
	M	SD	M	SD	T	p
Outcome
Strategies to limit drinking	101.60	17.80	109.00	20.63	−2.12	.04
Alcohol-related consequencesa	10.53	4.93	7.73	5.98	2.84	.01
Heavy drinking days in past montha	7.42	3.47	6.63	3.39	0.95	.35
Number of drinks per weeka	17.06	9.66	15.58	12.81	0.83	.42
Rape supportive attitudes and sexual consent
Rape myth acceptance	45.80	14.34	40.40	12.92	2.10	.04
Hypergender ideology	44.85	14.51	41.90	14.25	1.30	.21
Labeling of sexual consent	6.75	2.83	7.90	2.40	−2.33	.03
Perception of peer norms
Peer engagement in sexual coerciona	3.20	0.71	2.76	0.86	2.78	.01
Peer comfort with sexisma	3.47	0.60	3.23	0.54	1.81	.08
Peer norms: total drinks per week	29.16	10.81	23.59	8.00	4.08	.001
Bystander intervention
Perceptions of peer helping	71.20	7.83	74.15	10.09	−1.69	.10
Bystander intervention intentions	188.05	19.68	196.55	19.87	−2.31	.03
Confidence in helping a stranger	426,55	197.51	528.85	235.79	−2.11	.04
Confidence in helping a friend	868.44	64.88	871.22	86.56	−0.13	.90

^aN = 19.

In the 2 months between baseline and booster session, five of the 20 men (25%) perpetrated some form of sexual aggression. Four of these five men reported a history of sexual aggression at baseline. A series of exploratory between-groups t tests examined whether men who perpetrated after participating in the program varied on other 2-month outcome variables from those who did not perpetrate. No differences between groups were observed on alcohol-related outcomes. However, compared with men who did not perpetrate, men who engaged in sexual aggression were less likely to label a hypothetical scenario as rape, t(18) = 4.73, p < .001, indicated lower perceptions that their peers would engage in proactive bystander behavior, t(18) = 3.15, p < .01, and indicated greater endorsement of rape myths, t(18) = 2.60, p < .05.

Discussion

This open trial advances the science and practice of sexual assault prevention by documenting the preliminary feasibility, acceptability, and outcomes of a sexual assault prevention approach for heavy drinking college men that incorporates best practices in alcohol intervention and rigorously addresses the role of alcohol use as a risk factor for sexual aggression. To our knowledge, this is the first published study to formally integrate alcohol use intervention and sexual assault prevention for college men. Findings from this open pilot trial provide preliminary evidence of the feasibility and acceptability of this program for use among heavy drinking college men.

The baseline prevalence of sexual aggression (52%) and alcohol use (on average almost two heavy drinking days per week) confirmed that the inclusion/exclusion criteria and screening methods resulted in the recruitment of a high-risk sample. The sample recruited was also consistent with the demographics of the university, including representation of men in social fraternities (36%) and athletic teams (16%). Retention and program attendance rates were also excellent. Twenty of the 25 participants were retained at the 2-month assessment (80%), similar to the retention rate of the Men’s Workshop (Gidycz et al., 2011), which retained 83.4% and 77.8% of participants at 4 and 7 months, respectively.

The successful administration of a multisession intervention among a high-risk group of college men is noteworthy, given continued concern as to whether longer, multisession interventions can be implemented and sustained on college campuses. Identifying acceptable multisession sexual assault prevention programs is important, as single-session prevention programs are unlikely to be effective in producing behavior change in sexual aggression (DeGue et al., 2014; Tharp et al., 2011). With an eye toward dissemination, it would be useful to examine whether this intervention would be equally feasible and acceptable among college students most likely to violate institutional alcohol policies, such as those within fraternities, student–athletes, and students with prior alcohol violations. For example, brief alcohol interventions are routinely utilized among students who receive mandated referrals following alcohol violations (e.g., Borsari & Carey, 2005; Carey, Scott-Sheldon, Garey, Elliott, & Carey, 2016), and—if deemed efficacious in a larger trial—this integrated alcohol and sexual assault prevention program could offer a more comprehensive approach to addressing these synergistic health risks.

It is likely that several factors contributed to the successful retention and program attendance rates. The study team used rigorous follow-up procedures to maintain contact with participants, offering reminders via text, email, and phone. Sessions were also administered at the time participants deemed to be convenient. The compensation rate, which allotted US$135 for completing the 5½-hr program and US$30 for completing all study components, may also have facilitated retention. Continued work is needed to understand whether college students will volunteer to participate in effective interventions outside the context of paid research.

SAFE also appeared to be well accepted by heavy drinking men in this study. Participants indicated high satisfaction ratings, found the content to be useful, and reported that the program met most or all their needs. In addition, the sessions were delivered with high fidelity by male facilitators who were near in age to the study participants. Prior evaluations of the men’s program (Gidycz et al., 2011) document the successful training of undergraduate and graduate students to facilitate sexual assault prevention programming with college men. Participants reported that facilitators successfully conveyed a nonjudgmental style across all study sessions. Research is needed to examine how this communication style—which fosters recognition of group values through seeking consensus within a community, rather than enforcing the leaders’ values on the group—may contribute to the success of the program.

Study results were promising for immediate postsession effects, as well as several outcomes associated with alcohol use and risk factors for sexual assault. Specifically, within participants, positive effects were evidenced for motivation to change drinking, drinking intentions, peer drinking norms, and alcohol-related consequences. Positive effects were also evidenced for rape myth acceptance, labeling of sexual consent, perceptions of peer engagement in sexual coercion, bystander intervention intentions, and confidence to intervene to help strangers. Other outcomes shifted—though not significantly—in the direction that would be expected. Although these results are promising, they are subject to the limitations of a Stage Ia treatment development research methodology (Rounsaville et al., 2001), including small sample size, lack of a comparison group, short-term follow-up period, and limited generalizability. For these reasons, analyses were conducted within participants and did not generate effect sizes or document clinical significance. Findings, nonetheless, support further evaluation of this prevention approach within a subsequent pilot randomized trial.

One of the goals of a Stage Ia treatment development study is to utilize an open trial to refine an intervention (Rounsaville et al., 2001). As a result of this research, there are several potential changes to be made to the study procedures. For example, it is important to assess whether participants had the opportunity to engage in bystander intervention after participating in the program, and whether they actually engaged in proactive intervention when presented with the opportunity. Although speculative, it is possible that heavy drinking men have more opportunities to intervene than men who do not drink heavily, but are less likely to engage in prosocial intervention behavior when presented with a risky situation. It would also be useful to assess whether personality factors such as impulsivity, sensation seeking, and antisocial personality characteristics or general levels of aggression moderate program effects. There are different pathways through which alcohol use influences how and when individuals engage in sexual aggression (Abbey, 2011), and it is important to examine the possibility for varied responses to program participation. It would also be useful to better understand the profile of men who continue to perpetrate after program participation. In the present study, men who perpetrated after program participation, compared with men who did not perpetrate, reported higher levels of rape myths, had more difficulty labeling a nonconsensual sexual situation as rape, and perceived that other men were less likely to engage in proactive bystander intervention. Although preliminary, these findings provide some insight into how the program can be enhanced. Finally, given the heterogeneity of dating relationships reported by participants, with 36% in a long-term dating relationship and 61.5% of assaults perpetrated against an intimate partner, it is also important that sexual assault prevention programs use scenarios for bystander intervention that reflect the range of relationships reported.

The findings of the study should be considered in the context of some other limitations. This study did not employ a collateral report verification of self-reported drinking, alcohol-related problems, or sexual assault variables. That said, self-report is generally considered valid and reliable with little evidence of intentional bias (Borsari & Muellerleile, 2009). Second, the lack of racial and ethnic diversity in the sample limits generalizability of the current findings. Third, the study was limited to heavy drinking college men and applied a specific set of inclusion and exclusion criteria. One participant was excluded for displaying characteristics of ASPD, as we believed that ASPD-related attitudes and behaviors would be unlikely to change within the context of a multisession prevention program. It is possible that the SAFE program may demonstrate different effects among men who meet these criteria. Fourth, with the goal of reducing participant burden, the 2-month assessment was completed online prior to attending the booster session; the online methodology for assessments (as opposed to in-person) may have influenced study attrition and results. It is also necessary to include a follow-up assessment after the booster session to demonstrate whether this session serves to maintain program effects. Finally, the lack of a control condition precludes understanding whether the improvements were associated with participating in SAFE. It is possible that positive effects were due to the passage of time or demand characteristics. As noted by Tharp et al. (2011), the field of sexual assault prevention has been plagued by programs that lack a standard of evidence. Accordingly, this study should be considered a first step toward a more rigorous randomized trial.

In closing, this study provides preliminary evidence of the feasibility, acceptability, and promise of a sexual assault prevention program for heavy drinking college men with an integrated focus on the intersections between alcohol use and sexual violence. This open trial is particularly notable, as interventions that engage men as allies in violence prevention lag behind the development and testing of other intervention approaches (DeGue et al., 2014). As rates of campus sexual violence show little decline (Krebs et al., 2007), additional evaluation of the SAFE program is warranted, in conjunction with the continued development and testing of other innovative and rigorous intervention approaches.

Biographies

Lindsay M. Orchowski, PhD, is an associate professor (research) in the Department of Psychiatry and Human Behavior at the Alpert Medical School of Brown University and a staff psychologist in the adult outpatient division of the Department of Psychiatry at Rhode Island Hospital.

Nancy P. Barnett, PhD, is a professor (research) in the Department of Behavioral and Social Sciences in the School of Public Health of Brown University.

Alan Berkowitz, PhD, is an independent consultant specializing in sexual assault prevention, bystander intervention training, and social norms change.

Brian Borsari, PhD, is a clinical psychologist and health behavior coordinator at the San Francisco VA Medical Center. He is also professor-in-residence in the Department of Psychiatry at the University of California, San Francisco.

Dan Oesterle, BS, is a graduate of the University of Massachusetts Boston, and is a director of a clinical research laboratory within the Department of Psychiatry at Rhode Island Hospital.

Caron Zlotnick, PhD, is a professor in the Department of Psychiatry and Human Behavior, Department of Medicine, and Department of Obstetrics and Gynecology at Warren Alpert Medical School at Brown University.