Sexual assault resistance education for university women: study protocol for a randomized controlled trial (SARE trial)

This study addresses the high incidence of sexual assault among university women, particularly first-year students, and evaluates whether a small-group, four-unit sexual assault resistance education program can reduce completed sexual assault compared to standard university practice (informational brochures). The research question is whether the intervention reduces one-year incidence of rape (completed sexual assault) and improves intermediate processes (risk perception, self-defense self-efficacy, knowledge of effective resistance, risk appraisal, and attitudes). The study is situated within evidence that young women (ages 14–24) are at heightened risk, campus assaults are common, and existing coeducational or informational approaches have limited impact on behavior or assault incidence. The purpose is to rigorously test a theoretically and empirically grounded program using an RCT across multiple universities, with outcomes measured up to 24 months. The importance lies in potential direct benefits to women’s physical and mental health and providing an effective, adoptable program for universities.

Prior campus rape prevention has largely been coeducational and informational, yielding at best temporary improvements in knowledge and attitudes, often without measuring assault incidence. Reviews suggest theoretically-based, single-sex programs show more promise, with newer bystander-focused approaches still commonly delivered in single-sex groups. Attempts to reduce male perpetration show limited success to date, making women-focused resistance education a necessary component. An ecological cognitive framework by Nurius and Norris integrates background, environmental, and intrapersonal variables and analyzes women’s cognitive processes in responding to sexual coercion. Rozee and Koss’s synthesis emphasized an AAA model (Assess, Acknowledge, Act), adding self-defense tactics supported by evidence as key elements. Previous RCTs and evaluations show mixed effects. Some programs briefly lowered completed rape rates for subsets of women, with effects that often did not persist beyond a few months, even with booster sessions. An added self-defense component produced short-term reductions that waned by four months. Bibliotherapy showed positive proximal outcomes but no reduction in assaults. Rigorous RCTs focusing on survivors exist but do not inform primary prevention among college women. Pilot work by the present team developed and tested an AAA-based program. Initial pilots showed short-term gains in risk perception, self-defense self-efficacy, and strategy generation, and lower 3-month coercion/assault in the program versus controls (21% vs 39%). A subsequent small RCT (N=214) comparing revised AAA, sexuality-enhanced AAA, and control found 40–50% lower completed sexual assault at 3 and 6 months for program versus control (3 mo: 12% control vs 6% program; 6 mo: 8% control vs 5% program), with similar effectiveness regardless of prior victimization history. Integrating positive sexuality content improved risk perception but initially increased attrition when presented first; reordering to place sexuality content last reduced attrition to levels comparable to other units.

Study design: Multicentre, open-label randomized controlled trial at three Canadian universities (Windsor, Guelph, Calgary). Participants are randomized after baseline assessment via centralized, secure web-based allocation stratified by site. Due to the nature of the intervention, blinding is not feasible; prior pilots suggested knowledge of allocation does not bias responses. Participants: Inclusion criteria: female; 17–24 years; first-year university student (≤10 courses completed); able to attend one of four scheduled programs in the enrollment semester; informed consent provided. No exclusion criteria; all students have sufficient English proficiency per university requirements. Recruitment: Standardized recruitment across sites via registrar/residence emails, psychology pools, posters, washroom postings, research boards, flyers, classroom announcements, campus tables, and orientation events. Interested students are screened by phone or email and scheduled for in-person baseline assessment aligned with intervention timing. Randomization and procedures: After consent and baseline computerized surveys, participants are randomized, then immediately attend their first assigned session (control brochures or program unit). Research assistants remain unaware of allocation until after baseline completion; participants are directed to sessions without allocation disclosure during scheduling. Interventions: - Control (standard practice): Participants receive and are invited to read university-provided brochures on sexual assault, date-rape drugs, legal/medical advice, and services (site-specific). A research assistant facilitates limited Q&A (10–15 minutes), with group or individual questions; sessions are audio-recorded. - Intervention (Enhanced AAA program; four 3-hour units; delivery either weekly over 4 weeks or across one weekend): Methods include interactive games, mini-lectures, facilitated discussions, group activities, media clips, written scenarios, and role plays; sessions are audio-recorded. Resource kits provided at first session. Unit 1 – Assess: Improve recognition of situational and behavioral risk cues (e.g., alcohol, isolation), counter stereotypes, and develop non-restrictive strategies to reduce perpetrator advantage while maintaining social engagement; brief legal definitions. Unit 2 – Acknowledge: Accelerate recognition of danger in coercive situations, reinforce sexual rights, address competing social goals (relationship vs safety), debunk miscommunication myths, and practice responses to verbal coercion; reduce emotional barriers to resistance. Unit 3 – Act: Present effective resistance options matched to context; counter myths (e.g., resistance increases injury); emphasize forceful verbal and physical tactics supported by evidence; develop individualized strategy “toolboxes”; provide WenDo-based physical self-defense (stances, strikes, escapes) with attention to emotional barriers in acquaintance contexts. Unit 4 – Positive sexuality (OWL-adapted): Integrate prior units into longer-term romantic/sexual relationships; increase comfort discussing sexuality, clarify values, desires, boundaries, and safety needs; practice assertive communication to support healthy relationships and resistance to coercion. Assessments and follow-up: Computerized baseline surveys, immediate program initiation, then follow-ups at 1 week post-last session (knowledge acquisition; controls matched to timing), and at 6, 12, 18, and 24 months post-randomization. Baseline and 1-week assessments are in-person; subsequent follow-ups via secure web-based surveys (participant code-linked). Outcomes: - Primary: Completed sexual assault between assessments, measured by SES-SFV (nine items on oral/vaginal/anal intercourse without consent via force, threat, or incapacitation; month recorded for time-to-event). The SES-SFV has strong reliability and validity. - Secondary: (a) Perceived risk (“What are your chances of being raped by someone you know?” adapted from Gray et al.); (b) Self-defense self-efficacy (Marx et al.’s 7-item scale; plus two added items on defending against stranger and acquaintance attempts; high internal consistency); (c) Knowledge of effective strategies via intended behavior in scenarios (Testa et al. measure, focus on direct-resistance) and reported tactics during any assaults (coded per Ullman); (d) Risk appraisal via Testa’s risk perception questionnaire and Messman-Moore & Brown’s Risk Perception Survey (stranger and acquaintance vignettes; administered at spaced intervals due to familiarity effects); (e) Rape beliefs/attitudes via IRMA-SF and PCOR (focus on Female Precipitation subscale). - Tertiary: Incidence of forced sexual contact (non-penetrative) and sexual coercion (verbally coerced intercourse), both via SES-SFV; maintenance of improvements up to 24 months. Bias mitigation: - Selection bias: Anticipated attraction of more sexually experienced/victimized participants; assess by comparing sample with entering student demographics. - Outcome assessment bias: No blinding; use of the validated SES-SFV applied uniformly. - Differential dropout: Align follow-ups with academic calendar; multiple reminders; web surveys; minimal dropout in pilots. Provide incentives and offer full program to controls after trial to reduce differential attrition. - Loss to follow-up: Define as unknown 1-year sexual assault status; estimate 10% at 1 year, +15% by 2 years. Obtain multiple contacts and collateral contacts. - Compliance: Minimize no-shows by same-day baseline, randomization, and first session plus incentives (lotteries, passes, refreshments, certificates). Define non-response on surveys and employ tailored follow-up procedures. - Cross-contamination: Query control participants at 1 year about discussions with program participants; describe any contamination. Training and fidelity: Facilitators receive a detailed manual and centralized training (including WenDo basic training and specialized instruction). Research assistants trained on-site. Session fidelity monitored via facilitator protocol forms and random audio review; issues addressed in meetings. Sample size: Assuming 1-year completed sexual assault rates of 25% in control (brochures) vs 17.5% in intervention (absolute 7.5%, 30% relative reduction), 932 women (466/group) provide 80% power at two-sided alpha 0.05. Accounting for clustering (ICC 0.02, cluster size 20; design effect 1.38) and anticipated loss (10% at 1 year; additional 15% by 2 years), total N=1716 (932×1.38+0.75). Based on pilot RCT (3-month: 12% control vs 6% program; 6-month: 8% control vs 5% program). Statistical analysis: Intention-to-treat. Primary and tertiary outcomes compared via log-binomial regression (PROC GENMOD) with GEE to adjust for cluster correlation; covariates added if baseline imbalances; interaction test for weekend vs weekday delivery. Time-to-event via Cox regression with robust sandwich variance. Missingness assessed (MCAR vs MAR). Secondary outcomes analyzed with repeated-measures ANCOVA (PROC MIXED) adjusting for baseline; rank-based ANCOVA sensitivity analyses. Additional per-protocol and multiple imputation (five datasets; discriminant for binary, MCMC for continuous) assuming MAR. Explore effect modification by prior sexual assault history. Ethics and oversight: Ethics approvals obtained at all sites (May–Sept 2011). Anticipated risks minimal (possible distress from content; low physical risk with self-defense instruction). No adverse effects observed in pilots. Steering committee oversees trial; centralized project management; monthly teleconferences and annual meetings. Fidelity monitored as above. Trial status: First randomization September 28, 2011; 918 women enrolled at time of report.

As a study protocol, no trial efficacy results are reported. Key prior data and planned outcomes include: - Pilot RCT evidence: Program groups had 40–50% lower completed sexual assault than controls at 3 and 6 months (3 mo: 12% control vs 6% program; 6 mo: 8% control vs 5% program), with similar effects regardless of prior victimization history. Earlier pilot showed reduced coercion/assault at 3 months (21% program vs 39% control) and improvements in risk perception, self-defense self-efficacy, and strategy generation. - Primary planned endpoint: One-year incidence of completed sexual assault measured by SES-SFV; assessments at baseline, 1 week, 6, 12, 18, and 24 months. - Sample size and effect size: Target N=1716 to detect an absolute 7.5% reduction (25% to 17.5%; 30% relative reduction) with 80% power, accounting for clustering and attrition. - Trial progress: 918 participants enrolled at reporting time. - Secondary outcomes will assess changes in perceived risk, self-defense self-efficacy, knowledge of effective resistance, risk appraisal, and rape-related beliefs/attitudes; tertiary outcomes include forced sexual contact and sexual coercion, and durability of effects to 24 months.

The trial addresses a critical gap in campus sexual assault prevention by rigorously evaluating a theoretically grounded, skills-based, small-group program tailored to women. If effective, reductions in completed sexual assault are expected to positively affect women’s mental and physical health during a high-risk period. The study’s comprehensive outcome framework will clarify mechanisms of change (risk appraisal, self-efficacy, strategy knowledge, attitudes) and identify which components drive impact. Findings will inform optimization of content and delivery, determine duration of benefits, and guide timing for refreshers. Results can facilitate implementation at universities and potentially adaptation for high school and non-university populations of young women.

This protocol describes a multicentre RCT testing a four-unit sexual assault resistance education program for first-year university women against standard brochure-based practice. Building on strong theoretical and empirical foundations and promising pilot data, the trial will estimate the program’s effect on completed sexual assault and elucidate intermediary processes. Anticipated contributions include a maximally effective, adoptable program with training materials, evidence on durability of effects, and guidance on booster timing. Future research may extend to high schools and other young women’s populations, refine content based on mediator analyses, and evaluate integration with bystander and perpetration-focused initiatives.

- Open-label design precludes blinding; outcomes rely on self-report via the SES-SFV, though it is validated and uniformly applied. - Potential selection bias, as studies on sexual topics may attract more sexually experienced or previously victimized participants; sample characteristics will be compared with entering student demographics. - Differential attrition risk, particularly in the control arm; mitigated through incentives, personalized contact, web-based follow-ups, and offering the program to controls post-trial. - Possible cross-contamination through participant discussions; assessed at 1-year follow-up among controls. - Brochure content varies by site for the control condition (due to local services), though core content is similar. - Some secondary measures (scenario-based risk perception) cannot be repeatedly administered frequently due to familiarity effects, limiting measurement points. - Physical self-defense training carries minimal physical risk if instructions are not followed; addressed through training and supervision.