Evaluation of the Transfer of Training for a Sexual Assault Resistance Program Enhanced with Sexuality Education

The Enhanced Assess, Acknowledge, Act (EAAA) Sexual Assault (SA) Resistance Program is a theoretically sound, evidence-based program providing SA resistance education within a positive sexuality framework. It was shown to substantially reduce sexual assault victimization among university women who participate (Senn et al. in New England Journal of Medicine 372(24), 2326-2335, 2015). Staff training can either enhance or impede successful program scale-up and implementation. In this paper, we evaluate the transfer of training to implementation sites (i.e., postsecondary institutions) using a train-the-trainer model. Using pre- and post-training surveys and post-training interviews conducted from 2016 to 2020 with 33 implementation staff members from multiple sites, we answered the following research questions: 1. Did the training meet its overall goal of preparing implementation staff? 2. What training components were perceived to contribute to training effectiveness and implementation staff preparedness? Results suggested that our model of training was effective. Competence, confidence, and knowledge and ability increased significantly after training, and most staff perceived the training to be highly useful and effective (especially for preparing them to address EAAA participant issues). Practice and feedback from trainers through active learning techniques were especially important. Although implementation staff reported being well prepared to deliver the training or program, they reported being less prepared for handling other implementation-related activities and issues (that the training was not necessarily designed to address in-depth). Our findings suggest a need to enhance existing training on self-care and supporting program facilitators and for ongoing support and reminders from program purveyors to ensure that implementers are making use of existing resources. This study fills important gaps in the literature as few studies have examined the transfer of training for SA prevention programming.


Introduction
Sexual assault (SA) is a growing concern on postsecondary campuses, with as many as one in four Canadian university women being sexually assaulted in the previous year (similar to rates in the USA; Muehlenhard et al., 2017), many of whom experience a range of lasting negative physical, psychological, and academic impacts (Burczycka, 2020;Dworkin et al., 2017;Jeffrey et al., 2022). Given the pervasiveness and impacts of SA, there is urgent need for effective SA prevention programming.
Most SA-related programming currently offered on Canadian university campuses (according to our 2022 review of university websites) is not empirically supported and aims to increase awareness and knowledge rather than prevent. Moreover, most prevention programs that have been evaluated fail to effect lasting reductions in victimization or perpetration. Senn's Enhanced Assess, Acknowledge, Act (EAAA ) SA Resistance Program (also known as Flip the Script with EAAA ™) is one of the only SA prevention programs proven to effectively reduce victimization and has stronger evidence than more widely implemented bystander programs (Orchowski et al., 2020). The "enhancement" referenced in the title of the program refers to a sexuality unit added early in the program's development to enhance effectiveness (Senn et al., 2011). In a randomized controlled trial (Senn et al., 2015), we demonstrated the efficacy of EAAA with 893 first-year university women on three Canadian campuses. We found a significant 1-year reduction in rape (46%) and attempted rape (63%) and other forms of sexual assault victimization among women who participated in EAAA , with effects lasting for at least 2 years (Senn et al., 2015(Senn et al., , 2017. Additional positive outcomes included significant decreases in women's rape myth acceptance and woman-blaming and increases in risk detection and self-defense self-efficacy (Senn et al., 2017) which are responsible for (mediate) the reductions in victimization .
Closely controlled efficacy studies, even when conducted in the program's intended setting, do not reflect the realities of implementing in real-world conditions. Moreover, implementation success " [exists] quite independently of the quality of the program being implemented" (Fixsen et al., 2005, p. 12). Recently, our goal has been to develop a training and support system to foster self-sustaining implementation at postsecondary institutions across Canada (and other countries) and to establish the effectiveness of EAAA in these real-world conditions. Our implementation trial is one of the first of its kind in the SA prevention field (e.g., see also Davidov et al., 2020;Edwards et al., 2021;Noonan et al., 2009). In the current paper, we evaluate one component of our trial: the transfer of training from the nonprofit that disseminates EAAA to implementation staff at universities. Staff training is an important phase in the dissemination process that can either enhance or impede successful program implementation (Cahill et al., 2019;Carmody, 2014Carmody, , 2015Elliott & Mihalic, 2004), yet evaluation of such training in the SA prevention field is scant (Carmody, 2014(Carmody, , 2015.

Training for Program Scale-Up and Implementation
Effective implementation requires clearly defined program components (purpose, audience, activities, outcomes, etc.), carefully selected and qualified staff to manage and deliver the program, and organizational support for the program (Blase et al., 2012;Fixsen et al., 2005;Metz et al., 2013). Most relevant to the current paper, it also requires that "the people responsible for implementing the intervention have the competencies that the intervention requires," and that "they employ effective methods to carry out, improve, and sustain the intervention" (Easterling & Metz, 2016, p. 102). We employed a train-the-trainer (TTT) model to transfer these competencies and disseminate EAAA to implementing Canadian universities. The TTT model is a "pyramid training structure" that allows for the transfer of programs from lead trainers (who provide training across sites) to new site trainers (who learn program content and how to hire, train, and supervise others to deliver the program at their site) and from these new trainers to those who will deliver the program at their site (i.e., program facilitators; Cross et al., 2014;Nakamura et al., 2014, p. 11;Orfaly et al., 2005;Pearce et al., 2012). Research finds that this is generally an effective, relatively quick, and cost-effective way to disseminate knowledge and training (Pearce et al., 2012;Welber & Feinberg, 2002). The TTT model also allows for longterm sustainability because the program and its training are distributed across multiple people and organizations who no longer need to rely on the original organizing institution provided there is no staff turnover (though continued involvement of program developers or purveyors is often important for successful implementation; Fixsen et al., 2005;Orfaly et al., 2005).
Training staff to deliver programs on gender-based violence, SA, or sexuality may have unique demands related to the sensitive nature of these topics; the need to address personal attitudes and assumptions about gender, sexuality, relationships, and violence; and the need to provide the skills to work with audiences with diverse attitudes, assumptions, and experiences (Carmody, 2014(Carmody, , 2015Choi et al., 2020;Weingarten et al., 2018;Williams et al., 2014). For example, few authors have reported on the importance of teaching self-care, addressing trainees' woman-blaming attitudes, and teaching skills to address others' womanblaming attitudes when training staff to deliver SA prevention (Carmody, 2014(Carmody, , 2015Choi et al., 2020). SA prevention staff also need to be prepared to work with diverse audiences, some of whom will have victimization histories (Carmody, 2014). Addressing trainees' assumptions about sexuality is also important when content relates to explicit sexual topics (common in training staff for HIV prevention programs, for example; e.g., Williams et al., 2014).

Overview of the EAAA Program and Training
EAAA is an educational program designed to help young women (self-identified) of all sexual identities resist or reduce their risk of SA. It is trauma informed and offered to women with and without prior histories of SA. The four-unit program is delivered by pairs of peer facilitators to small groups of women (up to 20). The first three units (Assess, Acknowledge, Act) are based largely on recommendations from Rozee and Koss (2001) based on their review of theory and evidence. These units teach women how to identify risk for SA (particularly from male acquaintances in social situations), overcome emotional barriers to seeing danger in acquaintance situations and to resistance, and use effective verbal and physical self-defense techniques based on feminist Wen-Do Women's Self Defence (henceforth referred to as Wen-Do; wendo.ca). The fourth unit (Relationships and Sexuality) involves emancipatory sexuality education, which was shown to enhance women's ability to detect risk for SA and willingness to use forceful verbal resistance above and beyond the effects of the first three units (Senn et al., 2011). This unit helps women apply content from the first three units to their own lives by helping them to identify their sexual and relationship values, desires, and boundaries and practice negotiating them in consensual contexts (including one in which a partner wants to engage in desired sexual activity without a condom/dental dam). Relationship and sexuality education is an important feature of SA and intimate partner violence prevention as it can promote critical reflection on and changed beliefs about gender, relationship, and sexuality norms and development of sexual communication and assertiveness (Makleff et al., 2020;Radtke et al., 2020;Senn et al., 2011). As a whole, EAAA aims to make women aware of and resist not only SA but also the sociocultural context that enables it, including stereotypical assumptions about women's sexuality and strength (Radtke et al., 2020). Each unit includes mini-lectures, games, facilitated discussions, and application activities designed to foster engagement, knowledge, and skill development. Additional detail on the program content can be found in Senn et al. (2013Senn et al. ( , 2015Senn et al. ( , 2017. The SARE Centre (i.e., the program purveyor; Fixsen et al., 2005) is a nonprofit founded by the EAAA developer to train and support individuals and institutions (who sign a user licensing agreement) in their implementation of EAAA . Each implementing site has staff/students in two key roles: one or more staff member campus trainers (CTs) who oversee all implementation activities and peer (under age 30) program facilitators (PFs) who deliver the program and assist CTs with implementation activities such as participant recruitment. CTs are encouraged to follow the SARE Centre's protocol when hiring PFs to ensure commitment and good fit for the topics covered (e.g., comfort discussing sexual and sexual violence topics, openness to learn and receive feedback) and to hire diverse PFs (e.g., in body size, personality, race/ethnicity, sexual identity, academic discipline) who would present diverse models to EAAA participants. PFs are not required to have past experience with SA education or as group educators as the training is designed to address this knowledge and skill.
The training comprises three levels (see Fig. 1 for an overview). First, lead trainers are trained by the EAAA developer to provide training to others on behalf of the nonprofit (level 1). Lead trainers then train CTs across sites to hire, train, and supervise expert peer PFs at their sites (level 2). Finally, CTs train PFs to deliver the program at their sites (level 3). The current paper focuses on levels 2 and 3. The PF training overlaps substantially with the CT training; but with additional content and skill development related to group facilitation, handling difficult questions, and problem-solving; and without content related to management and supervision. Although not a primary aim, the CT training also provides some information to help prepare sites for other implementation and program sustainability procedures such as participant recruitment and program scheduling. The training continues to evolve based on feedback from implementers within and outside of the current trial. Some minor changes were made within the years of data collection (2016-2020; see Fig. 1). Additional information about the training is available upon request to the second author.
Senn designed the two main levels of training based on evidence-based practices (see Fisher et al., 2010;Fixsen et al., 2005;and Pearce et al., 2012 for reviews). They incorporate multimodal delivery methods and learning strategies including didactic instruction, active learning, independent readings, and accompanying learning materials (detailed CT and PF manuals). For example, at levels 2 and 3, trainers provide information about the background and philosophy of the program in lecture-and discussion-style sessions. Trainees then learn, practice, and receive feedback on needed skills and abilities through a series of behavioral rehearsals. CTs trained in level 2 learn how to teach effective facilitation, and PFs trained in level 3 learn to effectively deliver EAAA . Rehearsals are different from role plays in that they provide practice and preparation for scenarios that trainees will encounter in their future roles, rather than pretending to occupy someone else's role (Fixsen et al., 2005). Although most CTs do not ultimately deliver EAAA , they rehearse the program units and receive feedback during training as if they were PFs. The idea is to immerse them in the PF role so that they can then train and support actual future PFs. As noted, one of the four program units involves feminist self-defense instruction. Typically, such instruction is provided by certified instructors with years of training. However, this model is not feasible for widespread delivery of EAAA , and women do not need high-level training to increase self-efficacy or self-protective behaviors (e.g., Orchowski et al., 2008).
Thus, Senn partnered with Wen-Do to devise a model that includes 3 days of training by a certified instructor that would allow a well-supervised EAAA PF to teach select strategies with effectiveness and model women's strength and competence. Our recent research confirms the effectiveness Note. Current training model. Minor changes were made during and after data collection for this trial. For example, the original pilot training in 2016 did not include individualized Wen-Do training for CTs and instead included an overview of recruitment and implementation on Day 7 followed by 3 webinars. Recruitment information was moved to a webinar, and then a learning module in subsequent years (again based on CT feedback). In 2017, based on CT feedback, CTs only attended a specially created one-day Wen-Do course (later reverted to two days based on contrary CT feedback). Post-training learning module on scheduling and recruitment X (optional) Post-training learning module on planning for PF EAAA Act unit training X (optional) Post-training live webinar with SARE Centre staff on hiring, training, and supervision issues X P ost-training live webinar with SARE Centre staff on supervising PFs in delivery of the EAAA Act unit Fig. 1 Overview of the current EAAA program training process of PFs' instruction level; many women who participated in EAAA subsequently used strategies learned in the program to successfully protect themselves (Crann et al., 2021).

Study Purpose
We examined if and how (i.e., through what means) the training provided to CTs and PFs during our implementation trial was successful. We answered the following research questions: 1. Did the training meet its overall goal of preparing CTs to train PFs and PFs to deliver EAAA ? 2. What training components were perceived to contribute to training effectiveness and CT and PF preparedness to deliver the training/program?
This study fills important gaps in the literature as few previous studies have examined the transfer of training for SA prevention programming (see also Choi et al., 2020;Noonan et al., 2009;Weingarten et al., 2018). Our evaluation was informed by the literature and included collection of quantitative and qualitative data from both CTs and PFs and at multiple stages of implementation assessing competence, confidence, and skill development, attitudes of relevance to EAAA , perceived preparedness resulting from the training, and perceived value, strengths, and limitations of the training (e.g., Connell et al., 2002;Fisher et al., 2010;Grenyer et al., 2004;Kirkpatrick & Kirkpatrick, 2006;O'Brien et al., 2015;Pearce et al., 2012;Pirrallo et al., 1995). Choi et al. (2020) also used similar methods in their SA program TTT evaluation.

Participants
Participants were 12 CTs and 21 PFs from seven sites who took part in the training and at least one of the following: a pre-training survey, a post-training survey, any of four posttraining interviews for CTs, or any of two post-training interviews for PFs. Participants represent 91.7% (n = 33/36) of the total number enrolled in the implementation trial. Four of the 33 participants occupied both CT and PF roles at their sites (something that was not anticipated), but, for the purposes of this paper, we consider and include data for only their primary role. Each analysis included a different number of participants because not all participated in all data collection methods (valid percentages that exclude those with missing responses are reported throughout). The interview data represent the five sites that implemented EAAA (two additional sites did not ultimately implement and were, therefore, withdrawn before the interviews).
CTs were aged 59 years and under (66.7% were under 40 and 25.0% were under 30; age ranges were used in the survey to help protect confidentiality). All CTs had college or university education (91.7% had university education). Most (91.7%) were employees of a university or college (e.g., administrators, educators, counsellors/advocates); one was an employee of a community organization working in partnership with a university. CTs had varied experiences relevant to prevention education. All had at least 1 year of experience working with students (1-26 years; M = 10.2; SD = 6.7). Most also had at least 1 year of experience with teaching (80.0%), group facilitation (80.0%), and working with SA survivors (80.0%); 63.6% had some training in self-defense. Fewer (30.0%) had 1 year or more of experience counselling (providing counselling is not part of the CT role).
PFs were aged 20-31 years (M = 24.2; SD = 3.6). Most PFs (83.3%) had at least one university degree or college diploma completed, and 73.7% were current undergraduate or graduate students. PFs had varied experiences relevant to prevention education. Most had experience working with postsecondary students (89.5%; 0.5-8 years; M = 3.0; SD = 1.8) and training related to violence against women or sexual assault (79.0%). Most also had at least 1 year of experience with teaching (84.2%) and group facilitation (79.0%). Roughly half had some training in self-defense (52.6%) and at least 1 year of experience working with SA survivors (47.4%) and/or counselling (57.9%; not part of the PF role).

Procedure
The procedures described here were part of the larger Canadian Institutes of Health Research funded implementation trial (2016-2021) designed to investigate the effectiveness of EAAA and key moderating factors on several Canadian campuses, the transfer of training (covered in this paper), and ways to maximize effectiveness. Data collection related to the transfer of training took place in the first 4 years of the trial and included the following: (a) pre-and post-training surveys, (b) post-training interviews, (c) reflection notes from PFs written after program deliveries, and (d) notes from meetings between the SARE Centre and CTs (e.g., email exchanges, phone and in-person meeting notes). We only report on the surveys and interviews because the other data sources contained little information relevant to the transfer of training.
Upon clearance from the institutional research ethics boards of each university enrolled in the implementation trial, we emailed all CTs and PFs inviting them to participate in the research. This email included a link to an online consent form detailing the procedures of the surveys and interviews. We also sent up to two reminder emails to CTs and up to three to PFs who had not yet responded to the consent form. All women who attend CT training are already asked to complete pre-and post-training online surveys by the SARE Centre administrative staff person as part of the organization's regular practice (independent from our research). Thus, our email invitation sought consent from CTs for the SARE Centre to share their survey responses with us for secondary data analysis and to be interviewed to provide additional training feedback. PFs who agreed to participate on the online consent form were directed to the pre-training online survey and then invited via email after training to complete the post-training survey. We emailed all CTs and PFs who agreed to participate to schedule each of the post-training interviews. We assigned all CTs and PFs a unique ID code so that we could match each person's surveys and interviews. We offered CTs up to $200 CAD and PFs up to $100 CAD (because there were fewer PF interviews) for completing/releasing both surveys and all interviews (prorated depending on tasks completed; additional incentives offered for meeting and reflection notes). Incentives were chosen to reflect the professional skill and time commitment required to participate in each stage without undue influence.

Pre-and Post-Training Surveys
CTs and PFs completed the pre-and post-training (typically within 1 to 2 weeks but up to 45 days post training) surveys online via Qualtrics. Although most completed the two surveys before and after training, respectively, nine PFs completed a single survey after their training that assessed both post-training capacities and pre-training capacities retrospectively (that is, asked them to think back to their capacities before training) since they had not agreed to participate in time for the true pre-training assessment (referred to as retrospective survey takers). The pre-training surveys for both CTs and PFs contained demographic and background questions. The pre-and post-training surveys contained parallel items from United Nations Population Fund (UNFPA) and Family Health International (FHI 360;, and Pirrallo et al. (1995) meant to assess changes in self-reported competence, confidence, and knowledge and ability and adapted to be relevant to EAAA . For CTs, competence items assessed competency to carry out training and implementation activities such as conducting PF training, facilitating rehearsals, and providing constructive feedback. Confidence items assessed confidence in their ability to conduct PF training and provide support to PFs for each of the EAAA units and problem-solve various issues. Knowledge and ability items assessed knowledge about SA, ability to identify and counteract rape myths and victim-blame, knowledge of and ability to describe why certain groups of women/girls are at higher risk of SA while avoiding victim-blame, benefits of EAAA , and so on. For PFs, competence items assessed competency to carry out various facilitation activities such as facilitating rehearsals and handling difficult questions. Confidence items assessed confidence in their ability to facilitate each EAAA unit and problem-solve various issues. Knowledge and ability items were the same as for CTs. See Table 1 for Likert scale ranges. All baseline competence, confidence, and knowledge and ability scales had high Cronbach's alphas (see Table 1). Pearson correlations between the three pre-training scale scores were all below 0.85, suggesting that discriminant validity likely exists between the three scales. Thus, we kept all three scales intact.
CT pre-and post-training surveys also included the perceived causes of rape scale (PCR; Cowan & Campbell, 1995;revised Cowan & Quinton, 1997), which asks the extent to which respondents agree with 32 statements (six subscales) about why rape occurs. We analyzed only the female precipitation subscale-six statements attributing rape to women's behavior-to assess post-training changes in CTs' womanblame as this is a critical attitude related to the ability to address woman-blame with EAAA participants. Finally, the post-training surveys for both CTs and PFs included six questions adapted from UNFPA and FHI 360 (2005) and Kirkpatrick and Kirkpatrick (2006) asking participants to rate the overall quality of the training received, the usefulness of the training for their work with EAAA , their trainer, the effectiveness of the activities at reinforcing the concepts discussed, the appropriateness of the training for their skill level, and the pace of the training (see Table 2).

Post-Training Interviews
We conducted up to four post-training interviews with each CT who agreed to participate: (a) one typically within 2 months of their own CT training, (b) one in the months following the training they provided to PFs at their site (almost a year later for one CT), (c) one in the early stages of implementation at their site, and (d) one in the later stages of implementation at their site. We conducted up to two post-training interviews with each PF who agreed to participate: (a) one typically within 2 (but up to 4) months of their own PF training and (b) one at the end of their first year of program delivery. All interviews were semi-structured, conducted by an experienced research assistant or the implementation trial manager, generally lasted 30-60 min, and took place via telephone or videoconference.
The immediate post-training interviews (first two CT interviews and first PF interview) were particularly relevant to the present assessment of the transfer of training; however, all interviews contained questions meant to assess the strengths and weaknesses of the training and CT and PF preparedness for delivering the training/program as implementation progressed. Particularly relevant questions included the following: (a) What aspect of the training is most memorable or had the most impact on you as a new CT/PF? (b) What did you like best (and least) about the training? (c) Many different (training) strategies were used -was there one or more of these activities that you think was/ will be particularly useful (and not useful) to you as a new CT/PF? (d) Now that you have conducted the full EAAA PF training, was your own training as a CT sufficient to help you effectively train facilitators? (e) At this point (early implementation), how well do you think your training as a CT prepared you for implementing EAAA on your campus?
All interviews were audio-recorded and summarized in detail by research assistants to protect participants' confidentiality (verbatim transcripts are more identifiable) and because we assumed participants might speak more honestly knowing their words were not being transcribed, and that the program/training developer would not be able to identify speakers. We do, however, include in our analysis some verbatim quotations later retrieved from the recordings.

Data Analysis
Inspired in part by O'Brien et al.'s (2015) evaluation, we integrated the quantitative and qualitative data to answer our two research questions: (1) Did the training meet its overall goal of preparing CTs to train PFs and PFs to deliver EAAA ? (2) What training components were perceived to contribute to training effectiveness and CT and PF preparedness to deliver the training/program?
We structure the Results around these two research questions; however, distinctions between whether and through what means CTs and PFs were prepared were sometimes artificial, and there is some overlap and leakage between the two sections. During our analysis process, it also became evident that while CTs and PFs generally reported being well prepared to deliver the training/program, they reported being less prepared for handling other implementation-related Table 1 Scale reliabilities and training impact on key competencies and perceived causes of rape A mean item score was calculated for each scale that a CT or PF completed, and their pre-and post-training mean item scores were compared using paired t-tests. The assumption of normality was violated for two out of seven of these tests, as assessed by Shapiro-Wilk's test, but the t-test is robust to deviations from normality. We retained all outliers (one in two of the tests) because results were not materially different when they were removed. Some previous research finds bias in retrospective pre-tests that can result in inflated intervention effects (e.g., Taylor et al., 2009). However, previous research supports the use of retrospective instruments in TTT evaluations, especially to maximize the response rate (Pirrallo et al., 1995). Moreover, PF pre-and post-training scores were not significantly different between retrospective and traditional pre-post survey takers in our study, and results of the paired t-tests comparing pre-and post-training scores were not materially different when retrospective survey takers were included. Thus, results presented here are for retrospective and traditional pre-post PFs combined ** Significant at the p < .001 level a 1, not competent to perform without extensive further training; 2, not competent to perform without moderate further training; 3, not competent to perform without brief further training; 4, minimally competent to perform without further training; 5, moderately competent to perform without further training; 6, fully competent to perform without further training b 1, not at all confident; 5, extremely confident c 1, lowest; 5, highest d 1, strongly disagree; 4, neither agree nor disagree; 7, strongly agree activities and issues (that the training was not necessarily designed to address in-depth). Moreover, factors beyond the training itself were also commonly reported as important to PF and CT preparedness. Although somewhat outside of our two research questions, we include these findings because they are integral to our overall goal of understanding and building CT and PF competency and successful implementation.
We performed all quantitative analyses in SPSS version 28.
We coded some open-ended survey data and all interview summaries using NVivo. The first author and principal qualitative analyst (who was not involved in data collection or implementation) listened to parts of the interview audio recordings only in cases where frequencies are provided for responses to specific interview questions, when clarification was needed, and to identify exemplary verbatim quotes. Our qualitative analysis combined elements of qualitative content analysis and thematic analysis. The first author first read the interview summaries and developed an initial set of codes. She then reread the summaries, coding relevant passages, developing new codes, and grouping existing codes under higher-order headings as needed. This initial process drew on some of the coding procedures of content analysis (Elo & Kyngäs, 2008;Schreier, 2014). Codes were mainly semantic (explicit, surface-level meaning) and organized around shared topics related to our research questions (whether and through what means the training was effective), for example, recruitment, responding to EAAA participant issues, and various training activities.
As she became familiar with the content within each code and the interviews overall, she drew on the principles and procedures of thematic analysis to examine shared meaning in the data (e.g., Braun et al., 2018) to help pull together the codes and construct an overarching story. She used the relevant codes and quantitative findings to write an initial analysis answering the two research questions and structured mainly around the overarching story. She included content from some open-ended survey questions to bolster some of the qualitative sections. At times, she also coded responses to single open-ended survey or interview questions (in which case she referred to the recordings to ensure accuracy in the summaries) to supplement existing interpretations with frequency information. Finally, she reread all summaries to identify additional exemplary responses and to ensure that the final written analysis still fit the data. We identify quotes by the speaker's role (CT or PF) and campus (identified as sites A through E).

Did the Training Meet Its Goal of Preparing CTs to Train PFs and PFs to Deliver EAAA?
Our quantitative and qualitative results suggested that our model of training was effective in preparing CTs to train PFs and PFs to deliver EAAA .

Training Impact on Key Competencies
Paired t-tests indicated that CT and PF competence, confidence, and knowledge and ability increased significantly after training, with large effect sizes for each pre-post comparison (see Table 1). CTs' own woman-blaming (i.e., perceptions of rape as caused by women's actions) was already low before training and did not change significantly after training, though their perceived ability to counteract other people's woman-blaming improved (see qualitative results).

Perceived Quality, Usefulness, and Appropriateness of the Training
Both CTs and PFs rated the quality, usefulness, and appropriateness of the training that they received highly (see Table 2). Mean ratings for overall quality of the training, overall usefulness of the training, their trainer, and effectiveness of the training activities at reinforcing the concepts discussed were all above 4 (5 being the highest or most effective). Mean scores for overall appropriateness of the training for their skill level and overall pace of the training were all close to the midpoint ("just right"). Over 85% of CTs and PFs were in the ideal range on all indices (see Table 2). Although only about 60% reported that the pace was "just right," relatively even numbers reported that some sections were covered too quickly and too slowly, and none reported entirely too quickly or slowly. These findings suggest that the pace of the training was appropriate for an audience with mixed backgrounds.

Perceived Preparedness for Delivering EAAA Training and Program
The interviews, in general, supported and bolstered the quantitative findings regarding the effectiveness of the transfer of training. Generally, most CTs and PFs reported that the training was "really good," and that they felt well prepared to "(train) the facilitators" (CT, site B) or to "teach most of the program" (PF, site B). For example, when asked after having conducted PF training whether their own training was sufficient to help them effectively train PFs, almost all CTs reported that it was. Many PFs also reported positive experiences with the training, and (although not directly asked) that the training was sufficient to help them effectively deliver EAAA . When asked after having facilitated EAAA what aspects of being a PF they thought they were best prepared for by their training, most described being quite well prepared to address or work through participant questions and issues. Most commonly, they described feeling well prepared to identify and address victim-or womanblaming comments and language. Indeed, almost all PFs who discussed addressing woman-blaming at some point in an interview reported that the training was sufficient, or that they generally felt prepared to deal with this issue: I could do that every day, all day because of the way that we were trained…I can identify victim-blaming. I can dismantle victim-blaming. That is something that just comes as second nature now…That was really drilled into us…don't let women leave with victimblaming attitudes. (PF, site D) Very few CTs reported gaps in their preparedness for training PFs after their own training (though this was not a primary focus of the interviews). Nevertheless, one (site B) described not feeling "polished" after her training, and another (site C) similarly explained that while the training was adequate, additional practice is needed given how much there is to remember and observe when conducting rehearsals with PFs. PFs also reported few gaps in their preparedness for delivering EAAA after their training (though more than CTs). When asked after having facilitated EAAA what aspects of being a PF they thought they were least prepared for by their training, the most commonly reported aspects of program delivery (we describe aspects beyond program delivery later) were teaching the physical self-defense (Act) unit and dealing with specific participant questions and issues such as victimization disclosures, participants being uncomfortable talking about sex, and questions surrounding drugs/alcohol and consent, and mental illness and SA perpetration (mental illness is not addressed in EAAA ). Compared to PF reports about addressing woman-blaming, there was more variability throughout the interviews in terms of whether participants reported sufficient or insufficient training surrounding these other participant questions and issues. Moreover, several of the PFs who described being least prepared to deal with specific participant issues noted that this did not necessarily reflect inadequate training or complete unpreparedness. For example, a few PFs described being unprepared for responding to disclosures because their campus had insufficient resources to offer or because the training did not prepare them for the emotional burden of hearing disclosures: You can't really prepare for that moment…There's no deficit within the training or anything that's lacking. We're provided with the tools, the resources, and the connections that we can make for the women. I think it's just maybe preparing for that is-maybe looking at what that can do to you emotionally or the burden of knowing this. (PF, site D) Others noted that their self-defense training was mostly adequate, but that they needed refreshers, additional practice, or more time in the individualized Wen-Do training: I didn't feel super unprepared to do anything, but… the part of the facilitation that was most challenging to me was definitely the Act piece [3-hour unit including physical self-defense]…The full basic Wen-Do course was really great, and I felt like I learned a lot. But I felt like the half day of facilitator specific training seemed a bit rushed…I think that [extending it] would help to…have been more confident in my actual ability to teach women to-it's one thing to be able to be comfortable with the techniques in my own body but then teaching it to somebody else I find is a bit more challenging. (PF, site B) This need for more time in the individualized training or for more time dedicated to training them how to teach (rather than just perform) the physical self-defense techniques was echoed by a few PFs in response to other interview questions as well: "There were moments when I was like, 'I wish we could focus on how to apply this more to EAAA.' So, a bit more, you know, personalized for the program training" (PF, site D). These PFs typically suggested extending the half-day individualized training to a full day (most sites opt for half-day when they have four or fewer PFs to save on cost; otherwise, a full day is recommended). One PF (site C) explained that she did not "feel qualified to teach basic self-defense moves just by doing a little bit of training" and was worried that participants could hurt themselves or use the moves ineffectively in real life. Given EAAA 's efficacy and safety record, these concerns point more to a need to boost PF confidence than any inadequacy in the self-defense training itself.

Perceived Preparedness for EAAA Implementation Beyond Training/Program Delivery
While CTs and PFs generally reported being well prepared to deliver the training or program, they reported being less prepared for handling other implementation-related activities and issues-activities and issues that the training was not necessarily designed to address: "…really well prepared with the material and understanding the philosophy and the evidence base and all of those pieces, but the actual 'making it happen' (laughs)…I was not prepared for that" (CT, site B). Most notably, CTs and PFs described ill-preparedness for handling participant recruitment, self-care, interpersonal and team dynamics, and the needs of diverse participants.

Participant Recruitment
Despite a lot of time and effort spent on recruitment, CTs and PFs at most sites reported challenges recruiting students to participate in EAAA and (less commonly) volunteers to participate in rehearsals. CTs and PFs often reported that the training did not adequately prepare them for recruitment and suggested a need for additional training and support from the SARE Centre and other implementing sites (e.g., hearing success stories and evidence-based best practices). Some PFs also reported that they did not receive sufficient recruitment support or resources from their CTs. Some recognized that the training likely could not have "changed the recruitment struggle" because "recruitment really wasn't what the [TTT] was about…it was really about…making sure the program was delivered properly" (CT, site B).

Self-Care
Many CTs and PFs reported mental and physical exhaustion resulting from receiving and providing the intensive EAAA training and facilitation. They reported that the training is lengthy and "emotionally demanding" (CT, site B) because of the upsetting content, the high level of engagement required, and the constant feedback received during rehearsals. PFs also described the emotional drain of receiving disclosures from participants. As a result, many noted the importance of and need for more training on selfcare to better support CTs as they receive and then provide training and PFs as they receive training and deliver EAAA . A few PFs reported that they were not prepared by the training for the emotional toll of facilitating. Specific suggestions from CTs and PFs included training CTs how to best support PFs and teaching techniques for individual and team self-care and for dealing with burnout and compassion fatigue. Some suggested ending training days with exercises such a fun group activity, walk, or debriefing conversation. Some noted the usefulness of the mandated team debriefing meetings after program deliveries and availability of a feminist counsellor, though a few PFs wanted more one-on-one support from their CTs or felt the counsellor was not available quickly enough.

Interpersonal and Team Dynamics
The importance of strong and positive interpersonal and team dynamics for successful implementation was a common thread in the interviews. CTs and PFs reported that cohesion, cooperation, trust, good communication, and clear roles and expectations within campus teams were integral to successful implementation. For instance, they reported that trust and good communication between co-PFs (part of the training) "are fundamental" (PF, site A) and allow for smooth facilitation, especially given the emotional weight of the work. A positive team relationship and being part of a like-minded community were also often reported as one of the most enjoyable aspects of implementation.
Participants generally reported strong and positive team relationships (e.g., established trust and cohesion between co-PFs), and most PFs reported feeling supported by their CTs. Likewise, when interpersonal challenges arose, many reported being able to work them out. In some cases, they described having built these relationships or dealt with interpersonal challenges through specific training activities (e.g., team building exercises, rehearsals, practice with their co-PF, group discussions about the readings, training lunch breaks together) and CTs' ongoing supervisory activities (regular formal and informal team meetings, team building exercises).
Nevertheless, some PFs reported less than ideal interpersonal and team dynamics such as not always feeling heard or supported by their CTs (especially when CTs were more laissez-faire or inaccessible) or experiencing challenges with their co-PFs. Although some of these concerns were related to inadequate CT supervision (e.g., not enough effort to foster team building through regular team meetings), some PFs reported being unprepared by the training for developing strong teams or dealing with interpersonal conflict. For example, one PF noted that the training needed to be "more explicit about creating time and space where co- [PFs] can come together and develop that type of trust and complicity" (site A). PFs and CTs provided specific suggestions for improving the training in terms of team building, including hosting social events; incorporating discussions and activities to help build trust, cohesion, and empathy (especially between PFs); and teaching how to give and receive feedback (e.g., how to frame feedback, how to not take feedback personally, how to approach a co-PF about facilitation issues). A few CTs also reported a need for training resources to teach them how to support and manage PFs.

Needs of Diverse Participants
The importance of inclusivity was consistently raised in the interviews, especially by PFs (including concerns about the program's ability to speak to the experiences of diverse students). Some reported that they were not well prepared by the training, and that they wanted more than what is currently covered in the troubleshooting section to help them better address the needs and questions of diverse participants or support students with disabilities, international students, indigenous and racialized students, students who are uncomfortable discussing sex (e.g., for religious reasons), and, particularly, LGBTQ+ students. It should also be noted, however, that hiring diverse PFs in terms of personality, race and ethnicity, sexual identity, body type, and academic discipline (as recommended) was identified as useful by CTs at some sites because it meant having PFs that diverse participants could relate to.

What Training Components Were Perceived to Contribute to Training Effectiveness and CT and PF Preparedness to Deliver the Training/Program?
The interviews pointed to several key components in and beyond the training that were particularly important for fostering preparedness. The themes described here were particularly salient. Results suggested that the training activities were generally highly effective in building competency and preparedness, but factors and processes beyond the training (e.g., related to CTs' ongoing coaching and supervision) were sometimes barriers.

Practice and Feedback
Practice and feedback were consistently reported as some of the most important, useful, or impactful aspects of the training. Four out of five CTs and 12 out of 15 PFs who responded to the open-ended survey question about which training activities were most useful described a component related to practice and feedback. When asked in an interview what aspect of the training they thought was most memorable or had the most impact on them as a new CT/PF, more than half of the interviewees described a practice-related component, and all reported on the importance or impact of practice and feedback at least once at some point in an interview.
Overwhelmingly, CTs and PFs described the worth of the behavioral rehearsal. Often, they reported being challenged by the rehearsals (especially given little pre-training exposure to EAAA ), and, for some, it was precisely this challenge that made the rehearsals effective: I was a little anxious about sort of like being thrown into the dry-run…but…it's a great learning opportu-nity…participating in that was…the most beneficial for me. (CT, site C) It was just like really interesting to start facilitating it without really knowing how the program's supposed to kind of unfold and how each session's really supposed to progress and you kind of figure it out and see how it works…There were people with prompts [audience planted difficult questions]…you were put on the spot…it was really good to be able to practice that. That definitely made us feel prepared to…facilitate. (PF, site C) CTs and, especially, PFs reported that the rehearsals made them "realize that [they] would be okay and comfortable facilitating" (PF, site C) and "[allowed them] to be effective when… actually facilitating" (PF, site D). Most commonly, they reported that the rehearsals were helpful because they provided an opportunity to practice and receive feedback from audience members and trainers. For example, for CTs, the rehearsals were an important opportunity to practice and see modeled how and what (i.e., common facilitation errors) to critique. For CTs and PFs alike, the rehearsals prepared them by providing an opportunity to practice dealing with logistical and participant issues, especially with the help of planted audience questions and issues: The people who were observing us would take on the role of like the loud participant or the participant who would woman-blame or the survivor. And so just like getting used to dealing with all the different types of participants…and what kind of situations and challenges that that might pose-I found that that was super, super helpful. (PF, site C) It made me feel more prepared because there were times when we were then facilitating that some of the situations did come up. And because we had already worked through them…I felt confident being able to deliver that message in a way that was not condescend-ing…but to kind of allow them to see a different point of view. (PF, site D) For some, the rehearsals helped them to better understand how the different units of the program fit together, what the activities are meant to look like, and the various practicalities involved in delivering the program. The rehearsals also helped to foster team building and address interpersonal challenges. Several reflected on how the rehearsals were effective because they provided an opportunity to see the program from the participant's perspective (for those who also acted as audience members) and a realistic, hands-on learning opportunity: Often training isn't real…I think the learning is different when it's more realistic…As opposed to talking about training, we actually did the thing. It wasn't just a breakdown of scenarios, it was the actual role-play piece…it was really helpful. (CT, site B) Although considerably less common than reports about the usefulness of the rehearsals, others reported on the general importance of facilitation practice for PFs or wishing they had more time for practice. Some PFs also noted that periodic refresher training and rehearsals (mandated each year after a summer pause) were or would be useful for increasing confidence and preparedness, especially as time passed since their initial training and especially for the Act unit (which most felt was the most challenging unit to facilitate). Similarly, some PFs explained that actually delivering the program was a key part of learning, practicing, and increasing their comfort and confidence because every facilitation is different and has different issues.
Finally, some CTs and PFs described the importance or usefulness of the Wen-Do training (especially the individualized training) in terms of providing practice and feedback for both performing and confidently teaching the physical self-defense techniques (though not all PFs felt fully or adequately prepared by this training). Likewise, some explained that the Wen-Do training provided a useful model; for example, PFs learned "through modelling" (PF, site D) how to respond to participant questions (e.g., firmly but empathetically), and CTs learned what the self-defense techniques are supposed to look like and how to offer feedback/corrections to PFs.
In contrast to the general consensus on the usefulness of the rehearsals and other practice-related training activities, most interviewed CTs reported that the optional webinars were not particularly helpful, usually because they were poorly attended and there was little engagement. Similarly, PFs (and CTs on behalf of PFs) sometimes described the pre-training readings as more useful when followed up by team discussions (which are recommended but were not always implemented). These results further speak to the importance of active learning and opportunities to practice skills and knowledge and receive feedback.

Ongoing Interpersonal and Informational Support
Earlier, we described positive interpersonal dynamics in terms of their importance to implementation, the extent to which CTs and PFs were adequately prepared by the training to build these dynamics, and some effective or suggested training strategies for fostering these dynamics. Here, we focus more specifically on interpersonal support and its importance as a component of effective training and implementation. Some CTs and PFs reported on the usefulness of ongoing support from the SARE Centre, other implementing sites, and their own campus team for successful implementation (e.g., building a sense of community, receiving clarification about the program background and support around PF dynamics, and learning how other sites have dealt with facilitation situations and successfully recruited participants).
Almost all CTs reported feeling well supported by the help and consultation of the program developer and SARE Centre, and some reported appreciating being able to reach out: The support of the SARE Centre is really good…everyone who I've worked with has said the SARE Centre's been lovely…That has been essential [emphasis original] to this program working, is the ongoing availability of the SARE Centre…I really think that it's a necessary resource…ongoing mentorship and connection is really needed. (CT, site D) Although it was mostly CTs who were in contact with the SARE Centre, a few PFs described having received helpful informational support, which also fostered a sense of community.
A few CTs and a few PFs also described the importance of support from other implementing campuses, some reporting positive experiences. For example, some CTs explained that the training provided a valuable opportunity for bonding (which enhanced learning), networking, and discussions among CTs about possible implementation issues. One CT said, likely referring to the SARE Centre's implementation community of practice Google Group: I appreciate that…centralized sort of repository of wisdom. I like that. I like that, you know, someone else… in [another location] had one of the issues that we hit, and we were able to get at the answer to what they did, and it was really helpful. (CT, site B) More commonly, participants reported wanting more support from and connection with other sites, possibly because they were not using the resources created by the SARE Centre to facilitate these connections, including the Google Group. For example, one CT explained that it was unfortunate that attendance and engagement were low during the webinars because they were useful forums for connecting and sharing best practices. A PF similarly suggested webinars (PFs do not currently attend the training webinars) as a potentially useful way to connect with the SARE Centre and other sites to help foster a sense of community and learn from other sites.
Finally, PFs and CTs often reported on the importance or impact of support from their own campus team. Although some PFs reported wanting more support from CTs or team building during training, most felt well supported. A few PFs reported that trust and good communication with their CTs and creation of a safe training space allowed them to feel safe to ask questions and learn from constructive feedback without feeling overwhelmed or harshly judged. Both were important for leaving the training feeling prepared. Some PFs also described the importance of having a CT who was welcoming and readily available to answer questions and provide emotional and practical support when needed. A few CTs noted that competitiveness or conflicting styles between PFs sometimes created a negative training atmosphere.

Background Knowledge, Information, and Skills
Background knowledge, information, and skills-both pre-existing the training and developing mostly from the pre-training readings-were also important for preparedness. For example, a few PFs and CTs reported that their existing knowledge and skills (e.g., violence against women background or facilitation or implementation experience) were useful and made the training process easier. Some CTs reported that existing background among PFs (especially facilitation experience) was not necessary but made training them easier (though one noted that it created challenges if PFs were not receptive to new learning). Some PFs reported that having a CT with prior experience in the area was beneficial in terms of the quality of training (e.g., feedback during rehearsals) and supervision they received. Although some PFs had concerns about the readings (e.g., saw them as outdated or as needing to cover more diverse populations to help them address participant issues), most PFs who discussed the readings reported that they were generally helpful, especially for those who were new to the field and, as noted earlier, especially when they also had team discussions or presentations about the readings as suggested. The readings were "helpful at providing background information as to why the program exists and proof of its effectiveness" (PF, site C). They also helped PFs build confidence, respond to difficult questions or scenarios during facilitations, and (when followed up by team discussions) build team rapport and trust. Some CTs noted that the readings and, especially, group discussions about them were helpful to their PFs and observed an improvement in their ability to effectively respond to questions during rehearsals: "I do feel like the quality of the facilitation is improved by them actually having a solid understanding of what undergirds the program" (CT, site D).

Discussion
Results suggest that our model of training was effective in achieving its primary aim of preparing CTs to train PFs and PFs to deliver EAAA . Both CTs and PFs rated the quality, usefulness, and appropriateness of the training highly and reported preparedness. Self-reported competence, confidence, and knowledge and ability also significantly increased after the training for both CTs and PFs. CTs' womanblaming (perceptions of rape as caused by women's behavior) was already about as low as possible before training, likely because many were already engaged with feminist and SA issues (in contrast to Choi et al. (2020) who found a decrease in rape myth acceptance post-training but whose sample may have had less prior engagement). Nevertheless, the qualitative findings suggest that the training did improve ability to identify and challenge other people's womanblaming-a key skill for successful EAAA facilitation.
Practice and feedback, especially through rehearsals, and ongoing support from campus teams, other implementing sites, and the SARE Centre were integral to preparedness for CTs and PFs in our trial. These findings are consistent with others' (Elliott & Mihalic, 2004;Noonan et al., 2009;Weingarten et al., 2018) findings that role plays or rehearsals were one of the most helpful parts of training for violence and SA prevention program delivery, as well as with the general implementation science literature advising that training should be skills based (rather than only providing information), and that practice and feedback from coaches are necessary for improving competence and confidence (Blase et al., 2012;Fixsen et al., 2005). Support from coaches, including emotional and personal support, is also important for successful implementation (Blase et al., 2012;Fixsen et al., 2005), maybe especially when programs are emotionally demanding. Although one of the benefits of the TTT model is the development of self-sustaining implementation sites, continued involvement of program purveyors is important for successful implementation (Fixsen et al., 2005), and CTs and PFs in our trial generally felt well supported by the program developer and SARE Centre. A community of practice among implementing sites who "share their collective wisdom and determine new courses of action that might benefit many of the members" is also useful (Fixsen et al., 2005, p. 77). However, our results suggest that program purveyors may need to more strongly encourage or mandate participation in such an initiative.
Some CTs and PFs reported gaps in their preparedness for handling other implementation-related activities and issues that the training was not necessarily designed to address in-depth, including participant recruitment, self-care, interpersonal and team dynamics, and the needs of diverse participants. Each of these topics was covered briefly in the training. For example, recruitment training was deliberately minimal (~ 20 min) and not prescriptive because context can deeply impact the success of different recruitment strategies, and this is not a normal part of TTT or facilitator training. Nevertheless, our findings highlight the difficulty of recruiting university students to participate in an otherwise wellreceived (i.e., high acceptability; Lewis et al., 2015) noncredit program, and that sites may not be able to adequately handle recruitment on their own. We have found in our other research that students are quite interested in attending EAAA after seeing advertisements, and that the time commitment is a primary barrier. The current results also suggest that EAAA PFs should be included in the recruitment learning module training (currently only offered to CTs) because they are often the ones assigned by their CTs to carry out recruitment. Their recruitment concerns in the interviews suggest that recruitment knowledge was not well passed down to PFs from CTs. We have also since conducted research on effective EAAA advertising (messaging, imaging, location) and scheduling and updated the recruitment training module accordingly.
The need for more instruction on self-care in training may be unique to SA prevention programs and those covering similarly sensitive topics. SA prevention staff trainees (also peer facilitators) in Choi et al. (2020) also reported on the importance of self-care during training. The EAAA training instructs CTs on the importance of ongoing debriefing with PFs and the need for PF access to a feminist therapist. Given the extent of PFs' self-care concerns and the length of the existing training, CTs may need to accentuate the availability of these supports and spend more time fostering self-and team care in debriefing meetings and throughout implementation. Ongoing check-ins and reminders from the SARE Centre about the availability of these and other resources (e.g., the community of practice) and enhanced CT training on supporting PFs might also help. The existing self-care sections of the training have also since been enhanced based on feedback to include discussion about specific self-care strategies with a handout provided. See also Choi et al. (2020) whose TTT model included a workshop on self-care (the authors noted that self-care was a new concept to trainees).
Issues related to interpersonal and team dynamics were complex and varied. Many reported strong and positive team relationships and the ability to manage these relationships, in many cases fostered by the training. Nevertheless, when asked how the training might be improved or what they did not feel prepared for, some still described interpersonal challenges or a need for team building activities and instruction on how to give and receive feedback. Although the rehearsals were clearly a key learning experience and model (successfully, according to many) how to provide feedback, it is possible that a more explicit discussion is needed, especially for those with limited facilitation experience.
Finally, some PFs reported being ill-prepared by the training to address the needs of diverse students, such as LGBTQ+ students and students who are uncomfortable discussing sex for religious reasons. Sexuality educators have similarly reported not feeling prepared by training to address a range of sociocultural norms and values (Eisenberg et al., 2010). To our knowledge, only Noonan et al. (2009) andCarmody (2015) have reported on SA prevention and/or sexuality education implementers' need for support addressing the needs of diverse groups. EAAA effectively reduces SA risk for women with diverse identities (e.g., Black, White, lesbian, bisexual, heterosexual; Senn et al., 2020), and the training incorporates troubleshooting (and practice with planted audience questions) on addressing a diversity of experiences. Nevertheless, our findings point to a need to strengthen the training of the CTs and/or PFs to increase PFs' confidence in their ability to support diverse students. The training has since been enhanced to include further discussion on intersectionality, who EAAA was designed for, and research evidence for its applicability and relevance to racialized women and women with other diverse identities (relevance to sexual minority women including those with asexual identities was always emphasized).

Limitations
Limitations of this study included the sole use of self-report indicators of training success, which may have been prone to social desirability bias. We did try to mitigate socially desirable responding in the interviews by ensuring participants that the program developer would not hear the recordings or be provided with identifiable feedback. However, it is still possible that participants inflated the success of the training. It is also possible that participants did not always know how to gauge their own learning or what training strategies supported their learning. Nevertheless, participants were still willing to report negative training experiences, and key findings related to effective training strategies generally matched the implementation science literature.

Implications and Conclusions
Our findings demonstrate that our TTT training model and the various training strategies used (e.g., active learning and ongoing support) were effective for promoting EAAA implementation on a national scale. They also highlight the challenges involved in preparing implementation staff on multiple levels, including program administration, program delivery, and personal well-being. Ongoing support from program purveyors, including ensuring that implementers are making use of existing resources, is clearly important. Our results also suggest that other SA and related programs using a TTT model would similarly benefit from using behavioral rehearsals, annual refresher training (especially critical for self-defense instruction), and ample practice time throughout implementation to ensure preparedness. In addition to relevant content and skills, training should include time spent on the importance of self-care and on each person finding strategies for self-care that work for them. CTs (and others in supervisory roles) should spend time fostering self-care throughout implementation (e.g., modelling their own practice, accentuating the availability of supports, and regular check-ins, reminders, and debriefing meetings with facilitators). Finally, implementers should hire facilitators who, in addition to the other position requirements, represent students on campus in terms of race/ethnicity, body size, sexual identity, and personality to present a range of empowered models to program participants. Training should include discussion on intersectionality or other topics related to participant diversity and aim to bolster facilitators' confidence in their ability to support diverse program participants through practice, feedback, and supervision.
Funding This study was funded by a Canadian Institutes of Health Research Project Grant (FRN: 148843) to the second author.

Declarations
Ethics Approval The University of Windsor's Research Ethics Board and the boards of all participating campuses (as needed) reviewed and cleared all procedures in accordance with the ethical guidelines of the Canadian Tri-Council Policy Statement (TCPS-2).

Consent to Participate
Informed consent was obtained from all individual participants included in the study.

Conflict of Interest
The SARE Centre is a nonprofit and uses a costrecovery model. The second and third authors hold pro bono positions with the SARE Centre, and the fourth author is a pro bono member of the board of directors. None receive financial gains from the SARE Centre or EAAA program. The other authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.