A Popular Approach, but Do They Work? A Systematic Review of Social Marketing Campaigns to Prevent Sexual Violence on College Campuses

Abstract

College campuses continue to face high rates of sexual violence and social marketing campaigns have emerged as a common prevention strategy. However, there exists no summative research examining the effectiveness of this approach. A systematic search yielded 15 evaluations of eight unique prevention campaigns, which contributed to 38 individual outcome measures across four outcome categories (i.e., knowledge, attitudes, intentions/efficacy, and behavior). Summative results are mixed, but show promising campaign effects for increasing knowledge, modification of some attitudes toward sexual violence, intentions to participate, and actual participation in prevention activities. More evaluative research is needed for a comprehensive understanding of campaign effectiveness.

Introduction

Sexual violence on college campuses is a prevalent public health problem across the United States. Recent estimates suggest undergraduate students experience the highest rates of nonconsensual sexual contact, particularly among women (26%) and gender minority (transgender, nonbinary or genderqueer, gender questioning, or gender not listed) students (23%; Cantor et al., 2020). To reduce incidents of sexual violence, the 2013 Campus Sexual Violence Elimination (SaVE) Act (an amendment to the 1990 Clery Act) requires all Title IX schools to provide incoming students with sexual violence prevention programming (such as sexual assault and stalking; Griffin et al., 2017). The SaVE Act is intended to increase transparency in reporting requirements specific to sexual violence and harassment on college campuses, guarantee rights for survivors, establish disciplinary proceedings, and require the implementation of prevention education programs (RAINN, n.d.). Though compliance with this legislation has reportedly been inconsistent, there has been a sizeable increase in the number of sexual violence prevention initiatives being implemented across postsecondary institutions in the United States (Griffin et al., 2017).

While initiatives to prevent sexual violence on college campuses vary in format and approach, social marketing campaigns have emerged as a common option to raise awareness of campus sexual violence. This approach is appealing due to (1) the prevalence of social media usage among college students (Armstrong & Mahone, 2017), (2) the primarily passive and cost-effective method of delivery, and (3) it’s capacity to reach a wide audience (Bridges et al., 2015). Despite the popularity of this approach, no summative research has examined the effectiveness of social marketing campaigns as a preventative approach to campus sexual violence. Given the damaging outcomes that can result from experiencing sexual violence across a variety of psychological, social, and academic domains (e.g., suicidal thoughts, anxiety, depression [Carey et al., 2018; Dworkin et al., 2022], social withdrawal, emotional distancing [Georgia et al., 2018], decreased attendance, lowered GPA [Mengo & Black, 2015]), effective prevention initiatives are essential. While a sizeable literature has examined the effectiveness of various campus sexual violence prevention programs (e.g., Bringing in the Bystander; Banyard et al., 2009), research focused on social marketing campaigns is more limited. The current study presents the results of a systematic review of the existing literature on the effectiveness of college sexual violence prevention campaigns.

Strategies to Prevent Sexual Violence on College Campuses

Sexual violence prevention strategies aim to reduce or eliminate incidents of sexual violence within the campus community by addressing knowledge, attitudes, and behaviors related to sexual violence and misogynistic culture. Prevention strategies include, for example, awareness campaigns (to raise awareness of sexual violence as a problem; e.g., Red Flag Campaign; Carlyle et al., 2022), educational programs (to educate students about consent, and rape myths; e.g., Relationship Remix; Bonar, 2019), sexual assault risk reduction programs (i.e., teach students skills to reduce their risk of sexual assault victimization; e.g., EAAA Sexual Assault Resistance; Senn et al., 2015), or bystander education (i.e., teach skills to intervene when witnessing incidents of sexual violence; e.g., Bringing in the Bystander; Banyard et al., 2004). In addition, prevention strategies include environmental factors such as open-design campus spaces, increased lighting, increased security patrols at night, and “safe walk” programs (Griffin et al., 2017). Prevention initiatives also take several different forms, such as classroom-based education programs (single or multisession), large scale, assembly style presentations (e.g., during freshman orientation), or media/social marketing campaigns (Untied & Orchowski, 2013).

Social Marketing Campaigns for Sexual Violence Prevention

Social marketing campaigns have increased in prevalence in recent years due to their ease of implementation and dissemination (Wandersman & Florin, 2003). Compared to in-person programs, campaigns provide colleges with an efficient option for delivering sexual violence prevention content to the campus community. For example, educational programs such as multisession courses require a greater time commitment for participants and can range from 1 to 42 h of instructional time (DeGue et al., 2014). In-person instructional programs also require additional resources such as space and facilitators and may require increased financial support. Certain programs also have difficulty reaching a large proportion of students; while presentation style programs can be implemented with large groups (e.g., during freshman orientation), workshop-style programs typically require participants to register and are presented to much smaller groups (e.g., Amar et al., 2015; Banyard et al., 2007). Campaigns can address some of these gaps as they have the capacity to reach many members of the campus community due to their flexibility in dissemination; campaign materials can be displayed across campus in places where students congregate or in high-traffic, high-visibility areas (Bridges et al., 2015).

Campaign strategies have demonstrated effectiveness in conveying information and changing behaviors in several areas of public health education, such as high-risk drinking behaviors (Glider et al., 2001) and emergency contraception (Trussel et al., 2001). More recently, the use of campaigns has expanded into the area of sexual violence prevention (Stanley et al., 2017). Campaigns targeting sexual violence on college campuses are typically used to increase knowledge and address attitudes surrounding sexual violence among students (Gidycz et al., 2011; Untied & Orchowski, 2013). Some campaign initiatives also provide strategies for behavioral change (Randolph & Viswanath, 2004). To achieve this change, campaigns often target beliefs within the community by using a social norms approach. Rooted in the concept of conformity, this approach suggests that behaviors are influenced by our perception of others; we are more likely to engage in behavior that we think will be accepted by and match the behaviors of our peers and community (Burchell et al., 2013). As college students are often navigating increased independence and encountering new social experiences (Reingle et al., 2009), campaigns can prepare students for how to handle the challenges associated with new experiences such as those related to alcohol use or consent. Additionally, campaigns that advise students on how to respond to these situations can provide a framework for students to understand their social environment and what is considered to be acceptable behavior. By targeting attitudes and behaviors through the lens of social norms and providing convenient sources of information, campaigns have the potential to address the problem of campus sexual violence by enacting change from within the student culture.

Traditionally, campaigns have been implemented through print methods such as posters and flyers (Untied & Orchowski, 2013). More recently, dissemination approaches have been expanded to include online methods such as websites and social media (Armstrong & Mahone, 2017). Online methods of dissemination are a unique benefit of campaigns as students are exposed to campaign messages both on and off campus. In addition, presenting campaign materials via social media offers unique opportunities for students to interact with the information being disseminated (Armstrong & Mahone, 2017). For instance, students may have the opportunity to post their own experiences to the campaign sites, make comments, ask questions, or interact with other students who share similar experiences (DeMaria et al., 2018; e.g., see Define Your Line [Ortiz & Shafer, 2018]).

The Current Study

Substantial empirical research has focused on evaluating the effects of college sexual violence prevention initiatives, including many systematic reviews and meta-analyses that synthesize program impacts (e.g., Evans et al., 2019; Jouriles et al., 2018; Kettrey & Marx, 2020; Mujal et al., 2021). A notable gap in the summative review literature exists with respect to sexual violence prevention campaigns specifically (e.g., posters and social media). Given the prevalence of campaign methods across college campuses, understanding the effectiveness of such efforts is critical to our overall understanding of sexual violence prevention initiatives. The current study initially sought to conduct a systematic review and meta-analysis to investigate the impacts of existing sexual violence prevention campaigns. As will be described, the limited number of existing primary evaluations was not deemed sufficient for quantitative pooling. Instead, the current study presents a systematic review and description of the existing empirical evaluations of college sexual violence prevention campaigns.

Methods

A comprehensive search strategy was developed to identify all relevant evaluation studies of sexual violence prevention campaigns on college campuses. The search strategy (Table 1) included terms related to three constructs: (1) college, (2) sexual violence, and (3) campaign. The search was applied to 29 electronic databases (e.g., Social Sciences Full Text; see Supplemental Appendix 1) and key terms were searched in the Abstract field. To be included for selection, studies must have evaluated a sexual violence campaign that was implemented on a college campus. A campaign was defined as the passive dissemination of information intended for a wide audience and could include methods such as posters, banners, flyers, text messages, magazines, and social media. The campaign could also include additional components (e.g., a brief presentation), however, the campaign must have been the substantial focus of the overall intervention and the delivery must have primarily been passive in nature (i.e., minimal to no interaction with campaign leaders). If the campaign was presented in conjunction with a prevention program, the effects of the campaign component must have been assessed independently from the program (so as to not conflate the impacts of the campaign with those of the program).

Table 1.

Comprehensive Search Terms and Search Sources.

Construct	Key terms
University/college	“universit*” or “colleg*” or “higher education” or “campus*” or “undergrad*”
Sexual violence	“intimate partner violence” or “partner violence” or “IPV” or “partner abus*” or “dating violence” or “relationship violence” or “dating abuse” or “dating aggression” or “relationship aggression” or “sexual* violen*” or “sexual* aggressi*” or “sexual* abus*” or “sexual assault” or “rape” or “date rape” or “acquaintance rape” or “bystander*"
Campaign	“strateg*” or “interven*” or “prevent*” or “program*” or “campaign*” or “training or workshop” or “education” or “online” or “poster” or “social marketing” or “text message” or “mass communication” or “awareness or initiative*”
NOT	“online survey*” or “online questionnaire*” or “online study” or “collect* online” or “complet* online” or “recruit* online”

Articles must have been published in English or French between January 1, 2000, and February 10, 2021. Additional sources were searched to identify unpublished works such as technical reports, theses, and dissertations. These included Google and Google Scholar, 20 organization websites (e.g., National Sexual Violence Resource Center), 17 academic journals that often publish articles on sexual violence prevention (e.g., Journal of Interpersonal Violence and Journal of Community Psychology), and the curricula vitae of seven key authors in the field (e.g., V. Banyard, M. Moynihan, and S. Potter). See Supplemental Appendix 1 for details. As the search functions for the grey literature sources were more limited, shorter and modified sets of search terms were used and the first 100 hits for each search iteration were examined.

The systematic search was conducted by the three authors of the current study. Each reviewer independently searched a set of databases and/or grey literature sources and screened titles to identify relevant studies to be retrieved for further review; a merged list of potentially eligible studies was compiled from all reviewers following completion of the search. The reviewers then shared the task of applying the selection criteria to the study abstracts to determine which articles should be retrieved for further review; each study was screened by two reviewers. Last, two reviewers applied the selection criteria to the full-text articles to determine the final studies to be included for exhaustive coding.

Analytic Approach

Coding

Two reviewers extracted data for 58 variables from eligible studies, including general study information (e.g., program year and location), program characteristics (e.g., campaign elements and duration of campaign), study characteristics (e.g., research design), sample characteristics (e.g., demographics and sample size), outcome measures (e.g., tools/scales used to measure outcomes), and general study findings. Data were cross-validated by two reviewers, and disagreements were discussed until 100% agreement was reached on all variables.

Outcome Measures

To summarize the impacts of existing sexual violence prevention campaigns, the current study examines campaign impacts by separating outcome measures into four primary categories: (1) knowledge, (2) attitudes, (3) intentions/efficacy, and (4) behaviors. Knowledge outcomes refer to those that examine knowledge about sexual violence-related topics (e.g., types of violent behaviors, understanding of consent, and peer norms) and were split into subcategories of knowledge and perceived norms. Attitudinal outcomes include measures that assess participants’ beliefs, perceptions, or levels of acceptance of sexual violence-related norms or behaviors. Attitudinal outcome measures comprise four subcategories: rape myth acceptance, precontemplation, attitudes toward bystander behavior, and attitudes toward sexual communication. Measures were classified as intentions or efficacy if they addressed participants’ willingness or level of confidence to engage in a specific behavior; these include bystander intentions, intentions to obtain sexual consent, contemplation, and bystander efficacy. Last, behavioral outcomes include those that focus on changes in participants’ behaviors or active participation in certain activities; these include bystander behavior, action, and positive sexual behaviors. See Table 2 for details.

Table 2.

Outcome Measure Categories and Subcategories.

Overall category	Subcategory	Example measures
Knowledge	Knowledge	Whether participants gained and/or retained knowledge of information presented in the campaign.
	Perceived norms	Participant agreement on statements about their peers’ activities related to sexual assault prevention (e.g., “most students would support someone if they saw him/her trying to prevent sexual assault from happening”; Hust et al., 2017).
Attitudes	Rape myth acceptance	Participant level of endorsement of sexual violence myths and misconceptionsa (e.g., “If the rapist doesn't have a weapon, you really can't call it rape”; Payne et al., 1999).
	Precontemplationb	Participant awareness of the problem of sexual violence on campus (e.g., “I don't think there is much I can do about sexual assault on campus”; Potter et al., 2011).
	Attitudes toward bystander behavior	Participant ratings of how beneficial they believe certain behaviors are for preventing harmful events (Borsky et al., 2018) or their agreement with statements such as “I/most male students would admire someone who intervened to prevent abuse, sexual assault, or stalking” (Mennicke et al., 2015).
	Attitudes toward sexual communication	Participant attitudes about healthy sexual relationship behaviors such as “having a thoughtful conversation about sex with a prospective intimate partner” (Hovick & Silver, 2019) or establishing sexual consent before engaging in sexual activity (Ortiz & Shafer, 2018).
Intentions/efficacy	Bystander intentions	How likely participants are to intervene in a situation involving sexual violence (Borsky et al., 2018).
	Intentions to obtain sexual consent	Participant level of agreement with statements such as “I will stop if my partner asks me to stop during sexual activity, even if I was aroused.”
	Contemplationb	Participant willingness to get involved in reducing violence against women (e.g., “I am planning to learn more about the problem of sexual assault on campus”; Potter et al., 2011).
	Bystander efficacy	Participant ratings are. How certain they are that they could perform a series of behaviors such as, “Ask a friend if they need to be walked home from a party” (Cares et al., 2015).
Behavior	Bystander behavior	How often a participant intervened when witnessing an incident of sexual violence or related behaviors (e.g., removing a friend from an uncomfortable situation, speaking up against rape myths; Borsky et al., 2018).
	Actionb	Whether participants have taken action to address sexual violence (e.g., spoken up about relevant issues or in sexual violence situations; Potter et al., 2011).
	Positive sexual behavior	Specific behaviors relating to obtaining consent (e.g., “stopped the first time date said no”; Mennicke et al., 2015).

^a This measure was commonly assessed using the Illinois Rape Myth Acceptance Scale (IRMAS).

^b The Precontemplation, Contemplation, and Action subscales are from the Stages of Change Scale (Banyard et al.,2005).

Data Synthesis

Descriptions of the evaluated campaigns were developed using data on the intervention content, materials, dissemination methods, and locations. To assess campaign effectiveness, where possible, effect sizes were calculated as standardized mean differences using three types of reported outcome data: (1) pre test and post test means with standard deviations for a treatment and control group (n = 14); (2) pre test and post test means, standard deviations, and t-test for a single treatment group (n = 1); and (3) post test means with standard deviations for a treatment and comparison group (n = 8). Data required for standardized effect size calculation was only available for 23 of the 38 reported outcome measures; author-reported outcome data was used for the remaining 15 measures. Independent samples in each subcategory were required; if multiple studies were identified using the same sample, only one study was used for each measure. Results were classified into four primary outcome categories, each with specific subcategories as explained previously. Across each subcategory, results were summarized to estimate an overall treatment impact of the evaluated campaigns.

Results

Systematic Search Results

The systematic search of 29 databases resulted in a total of 12, 998 studies for review. Duplication was common across searches and the number of grey literature hits was not tracked. In total, 128 studies were identified for further review and retrieval. These studies were retrieved in full (nine were not retrievable) and assessed by three reviewers for inclusion. After the application of selection criteria, 28 studies were retained (e.g., 87 studies were excluded for not evaluating a social marketing campaign). During the data extraction and coding processes, an additional 13 studies were excluded, for a final sample of 15 studies and 38 individual outcomes. See Figure 1.

Figure 1.

PRISMA flow diagram.

^aThe report evaluated an intervention that did not include a social marketing campaign component (e.g., only evaluated an in-person workshop, interactive online program, theatre-based intervention, etc.).

Overview of Included Studies

Nearly all of the final sample of 15 studies were journal articles (93%). All were published between 2008 and 2020 and evaluated a campaign implemented at a post-secondary institution in the United States. The length of intervention exposure ranged from one week to 12 months. The methodological rigor of the research designs was assessed using the Maryland Scale of Scientific Methods (Sherman et al., 1998); scores ranged between 2 and 5 on the scale, with most studies rated at level 3 (40%, n = 6). The time between intervention and post test varied, with a majority of studies using an immediate post test (53%, n = 8). Sample sizes ranged from 95 to 1,650 participants (median: 284 students; mean = 406.3, SD = 404.7). Participants were predominantly White and female; 83% of studies (n = 10) used samples comprised of more than 70% White students and 87% of studies used samples comprised of mixed gender or majority female students. The 15 studies evaluated eight unique sexual violence prevention campaigns; each campaign is described narratively and study details are presented in Table 3.

Table 3.

Study Level Details.

Author (date)	Campaign components and length of exposure	Campaign setting	Research design and Maryland Scale Rating	Sample gender and ethnic characteristics	Sample size and composition	Time of post test
Know Your Power (n = 7 studies) Posters modeling prosocial bystander behaviors in risky situations, designed to raise awareness about sexual/relationship violence and stalking
Cares et al. (2015)	Posters Accompanying presentation Promotional products Campus events Website 6 weeks of exposure	Residence halls, dining halls, student centers, libraries, bathroom stalls, athletic venues, academic buildings, local businesses, student computers, water bottles, keychains, buttons, bookmarks, and table tents	Single group, pre test–post test designa SMS = 2b	47.8% female; 85.2% White	188 students on 2 campuses	12 months following pre test
Moynihan et al. (2015)			Single group, pre test–post test designa SMS = 2	48% female; 85.2% White	482 students on 2 campuses	12 months following pre test
Katz et al. (2013)	Posters 6 weeks of exposure	Residence halls	Two-group, post test only design SMS = 5c	69% female; ethnicity not provided	95 first-year students living in residence (61 TX and 34 CTL)	Immediately after campaign
Potter et al. (2008)	Posters 2 weeks of exposure	Residence halls	Two-group, pre test–post test randomized control design SMS = 5	51% female; ethnicity not provided	145 first-year students living in residence (81 TX and 64 CTL)	Immediately after campaign
Potter et al. (2011)	Posters 4 weeks of exposure	Residence halls, Greek houses, recreation facilities, dining halls, student center, local businesses (e.g., bars, banks, and coffee shops)	Two-group, post test only design SMS = 3d	62% female; 87% White	372 students (291 TX and 81 CTL)	Immediately after campaign
Potter (2012)	Posters Promotional products Campus events Website 6 weeks of exposure	Academic and recreation buildings, residence hall lobbies, student apartments, bathroom stalls; dining halls, libraries, bookstores, local businesses, buses, water bottles, flashlights, table tents, bookmarks, and computer pop-up images	Two-group, pre test–post test design SMS = 3	71% female; 95% White	353 students (230 TX and 102 CTL)	Immediately after removal of campaign materials
Potter and Stapleton (2013)	Posters Promotional products Website 4 weeks of exposure	Academic, administration, recreation, & residence buildings, buses, dining areas, university computers, water bottles, highlighter pens, Frisbees, bookmarks, local and businesses	Two-group, pre test–post test design SMS = 3	78% female; 92% White	236 students (129 TX and 107 CTL)	Immediately after campaign
Red Flag Campaign (n = 2 studies) A bystander campaign to encourage students to take action when they see “red flags” of sexual violence, dating violence, and stalking
Red Flag Campaign (n = 2 studies) A bystander campaign to encourage students to take action when they see “red flags: of sexual violence, dating violence, and stalking
Borsky et al. (2018)	Posters 30-min presentation Promotional products 1 week of exposure (+ 1 day for presentation)	Campus wide	Two-group, pre test–post test design SMS = 3	80% female; 84.1% Whitee	288 students from 2 universities (207 TX and 81 CTL)	3 months following pre test
Carlyle et al. (2022)	Posters Social media presence 4 weeks of exposure	Bystander training, awareness events, movie night, scavenger hunt, lawn flags, posters, stall seat journals, social media, and emails from administration	Two-group, post test only design SMS = 3	74% female; 49% White	203 first-year students (167 TX and 36 CTL)	Approximately one month following pre test
Consent is Sexy campaign (n = 1) A poster campaign using sex-positive images and messages to define and promote consent both as mandatory and sexy among students
Hovick and Silver (2019)	Posters approx. 3–12 months of exposure (campaign was ongoing)	Campus buildings where undergraduates congregate, on digital screens, and in dormitories, fraternity and sorority houses, and buses	Two-group, post test only design SMS = 2	58.5% female; 80.9% White	284 students (181 TX and 103 CTL)	3–12 months after campaign start
Consent is Good, Joyous, & Sexy campaign (n = 1) Featuring provocative and brightly-colored banners created to increase awareness of consent and bystander behavior to prevent sexual violence
Thomas et al. (2016)	Posters/banners 1 week of exposure	Main pedestrian walkway on campus	Two-group, post test only design SMS = 3	78% female; 92% White	374 students (199 TX and 175 CTL)	Immediately after campaign implementation
Define Your Line A social marketing campaign focused on “unblurring the lines” of sexual consent through peer-to-peer and interactive messaging
Ortiz and Shafer (2018)	Posters Promotional products Campus events Website Social media 1-2 semesters of exposure	Campus wide, table tents, posters, bus signs, website, and social media	Single group, pre test–post test design 2	60% female; 63% White	992 students	7 months after campaign launch
Text messaging campaign A series of text messages containing facts, risk factors, and tips to prevent sexual assault
Chiriboga (2016)	Text message notifications 4 weeks of exposure (2 messages/week)	Phone	Randomized two-group, post test only design 5	67.9% female; 75.5% White	155 students (78 TX and 77 CTL)	Immediately after intervention
Mini-magazine campaign (n = 1) A series of 8-page mini-magazines featuring both general interest content and information related to the prevention of sexual violence
Hust et al. (2017)	Magazines 10 weeks of exposure (1 new magazine every 2 weeks)	Residence halls	Two-group, pre test–post test design 4f	66% female; 72% White	77 first-year students living in residence (40 TX and 37 CTL)g	Immediately after exposure to the last magazine
Social marketing campaign (n = 1) A poster campaign targeting men; promotes social norms and awareness regarding consent, bystander behavior, rape myths, and sexual activity
Mennicke et al. (2021)	Posters Campus events campaign ongoing; exposure unknown	Campus wide, posters, bus wraps, table tents, and billboards	Cohort pre test–post test design 2	100% male; 74% White	1650 students	4 years after baselineh

^a The original study employed a multigroup design; however, to isolate the effects of the campaign, we assessed the outcome as a single-group pre/post design.

^b Level 2. Temporal sequence between the program and the crime or risk outcome clearly observed, or the presence of a comparison group without demonstrated comparability to the treatment group (Sherman et al., 1998, p. 4).

^c Level 5. Random assignment and analysis of comparable units to program and comparison groups (Sherman et al., 1998, p. 5).

^d Level 3. A comparison between two or more comparable units of analysis, one with and one without the program (Sherman et al., 1998, p. 4).

^e Control group = 77.8% White.

^f Level 4. Comparison between multiple units with and without the program, controlling for other factors, or using comparison units that evidence only minor differences (Sherman et al., 1998, p. 5).

^g The study included four conditions total (three treatment and one control); to isolate the appropriate impacts we assessed the outcomes based on one treatment group and the control group, as such the recorded sample size does not reflect the study's total sample size.

^h Baseline occurred after the first year of campaign implementation.

Know Your Power

The Know Your Power bystander campaign was developed in 2004 by a multidisciplinary media campaign working group at a public university in the Northeast. The campaign includes four posters depicting “typical” college scenarios in which prosocial bystander behavior is modeled (e.g., a house party where two bystanders discuss how to prevent a man from following an intoxicated woman upstairs to a room); two scenarios related to sexual assault, one to sexual harassment, and one to intimate partner abuse. Each poster also included the tagline “Know your power. Step in, Speak up. You can make a difference,” as well as advice about steps to take in similar situations; posters were displayed in various locations across campus (e.g., academic buildings, residence halls, and dining halls). Other elements of the campaign include promotional items (e.g., water bottles and keychains), campus information tables, and a website.

The Red Flag Campaign

The Red Flag Campaign was developed by the Virginia Sexual and Domestic Violence Action Alliance in conjunction with various college faculty, staff, students, and community victim advocates across Virginia. The Red Flag Campaign features a series of eight double-sided posters aimed at encouraging members of the campus community to “say something” when they see “red flags” of sexual or dating violence. One side of each poster depicts a scenario illustrating various components of sexual and dating violence such as emotional abuse, coercion, excessive jealousy, isolation, sexual assault, victim-blaming, and stalking; the images include diverse groups of students such as those from ethnic backgrounds and heterosexual and same-sex couples (see Carlyle et al., 2022). The other side of each poster focuses on the key tenets of healthy relationships and ways bystanders can safely intervene to prevent violence and abuse. The posters were displayed in a variety of locations across campus and were at times supplemented with additional components such as interactive campus activities or presentations.

Consent is Sexy

The Consent is Sexy campaign features a series of four posters promoting consent as mandatory using sex-positive images. Each poster includes the message “Respect yourself. Respect your partner.” and depicts diverse groups of students (e.g., same-sex couples) in the scenarios. The intent of the campaign messages is to move beyond simply defining consent by promoting effective communication between sexual partners (e.g., how to gain consent, that consent can be “sexy”). Posters were distributed across campus, including common areas within campus buildings, dorm buildings, fraternity and sorority houses, buses, and digital screens.

Consent is Good, Joyous, & Sexy

The objective of the Consent is Good, Joyous, & Sexy campaign is to increase awareness of consent and bystander behavior among students; campaign messages were developed through a series of informal conversations with students on campus. The campaign includes flyers and a series of five bright, provocative banners displayed along the main walkway of the campus, featuring five messages: “Consent can be Revoked”; “Consent is Best When Consensual”; “Sex without Consent Isn't Sex”; “The Joy of Consent”; and “Consent is an Aphrodisiac.”

Define Your Line

The Define Your Line campaign was developed with input from undergraduate student groups (e.g., athletes, Greek members, and Gay Straight Alliance members) and faculty members of a university in the southwest. The campaign intended to “unblur” the lines of sexual consent by generating open conversation across campus regarding sexual violence and focused on peer-to-peer communication as the primary method of spreading awareness. The campaign includes a variety of components such as information booths and interactive question/messaging systems on the campaign website and social media pages. Promotional materials such as posters, table tents, and bus signs were displayed across campus and addressed various misconceptions regarding sexual consent.

Text Messaging Campaign (No Campaign Name Specified)

Chiriboga (2016) developed a text message-based awareness campaign to promote sexual assault education and prevention. The campaign consists of eight text messages delivered over a four-week period, with two messages per week. The messages provide general facts about sexual assault, risk factors, and ways to prevent sexual assault, and often included a weblink, for example, “Sex without consent isn't sex. It is rape. For more information visit: www.notalone.gov” or “Sexual assault perpetrators use drugs that have no odor, color or flavor. Know your limits, if u or a friend doesn't feel right after a drink get help. http://web.uri.edu/womenscenter/resources/.”

Mini-Magazine Campaign (No Campaign Name Specified)

Hust et al. (2017) collaborated with student interns and volunteers at a large Northwestern university to develop a series of five mini-magazine booklets; one magazine was distributed throughout residence halls every 2 weeks for a total of 10 weeks. The magazines were designed to encourage students to participate in activities related to sexual assault prevention on campus and to take action to prevent sexual assault. Each magazine included four pages of general interest content and four pages of sexual violence prevention content. The sexual violence prevention articles focused on topics such as sexual consent negotiation, bystander intervention, reporting of sexual assault, services after a sexual assault incident, support of sexual assault victims, and participation in sexual assault reduction activities.

Social Marketing Campaign (No Campaign Name Specified)

A social norms-focused marketing campaign was developed by Mennicke et al. (2021) with the intention of reducing sexual violence on a large campus in the Southeastern United States. The campaign was implemented yearly and targeted male students, featuring a series of two to three posters and other promotional materials (e.g., bus wraps, table tents, and billboards). Each year, the campaign messages generally adhered to four themes: consent, bystander behavior, rape myths, and sexual activity. Messages included those such as “Most men understand the importance of getting consent before sexual intimacy” or “Most men would intervene to prevent sexual harassment or sexual assault.”

Summary of Campaign Effectiveness by Outcome Measure

Next, we summarize the 38 outcomes presented in the sample of 15 studies. The 38 measures fall into the main categories of knowledge (n = 5), attitudes (n = 12), intentions/efficacy (n = 13), and behavior (n = 9). Due to extensive heterogeneity among outcome measures, a formal meta-analysis was not possible; within the four primary categories are 13 subcategories, each with only a handful of possible effect sizes as only 23 of the outcomes included data that allowed for effect size calculation (Table 4).

Table 4.

Findings by Outcome Type (N = 38).

Author (date)	Campaign name	Outcome measure	Findings	Standardized effect size and standard error	Summary conclusions
Knowledge (n = 4)
Cares et al. (2015) a	Know Your Power	Knowledge (lifetime sexual assault for women)	No significant difference between pre test and 12-month follow-up for the campaign-only group	ES calculation not possibleb	Overall positive and significant effect is evident for general knowledge gain
Potter et al. (2008)	Know Your Power	Knowledge of helpful bystander intervention skills	Significant increase from pre test to post test for the treatment group versus the control group	0.282 (0.246)
Ortiz and Shafer (2018)	Define Your Line	Knowledge of “grey” sexual assault situations	Significant increase from pre test to post test for the treatment group	ES calculation not possibleb
Hust et al. (2017)	Mini-magazine campaign	Perceived norms related to sexual assault prevention	Significant increase in knowledge from pre test to post test for the treatment group versus the control group	0.32 (0.286)	Significant and positive impact on perceived norms
Attitudes (n = 12)
Cares et al. (2015)	Know Your Power	Rape myth acceptance (rape myth 1)	Significant decrease between pre test and 12-month follow-up for the campaign-only group	ES calculation not possibleb	Slight positive impact on rape myth acceptance, with some decrease in acceptance noted
Katz et al. (2013)	Know Your Power	Rape myth acceptance	No significant difference between treatment and control groups at post test	ES calculation not possibleb
Potter and Stapleton (2013)	Know Your Power	Rape myth acceptance	No significant difference between those who reported seeing the campaign and those who did not	−0.048 (0.131)
Chiriboga (2016)	Text messaging campaign	College date rape attitudes	No significant difference between the treatment and comparison groups	−0.248 (0.232)
Mennicke et al. (2021)	Social marketing campaign	Rape myth acceptance	Year 5 reported significantly lower acceptance of rape myths than those in year 1	ES calculation not possibleb
Cares et al. (2015)	Know Your Power	Precontemplation	No significant difference between pre test and 12-month follow-up for the campaign-only group	ES calculation not possibleb	No impact on the precontemplation subscale
Potter et al. (2011)	Know Your Power	Precontemplation	No significant difference between those who did and did not report seeing the posters	−0.132 (0.126)
Potter (2012)	Know Your Power	Precontemplation	No significant differences between pre test and post test (at p< .05) between those who saw posters more than once a day versus less than once a day	−0.09 (0.119)
Borsky et al. (2018)	Red Flag Campaign	Attitudes toward bystander behavior	No significant difference from pre test to post test for the intervention group versus the comparison group	0.192 (0.199)	No impact on attitudes toward bystander behavior
Mennicke et al. (2021)	Social marketing campaign	Attitudes toward sexual violence and bystander behavior	No significant difference between year 1 and year 5	ES calculation not possibleb
Hovick and Silver (2019)	Consent is Sexy campaign	Attitudes toward sexual communication	Those who recalled campaign exposure reported significantly more positive attitudes than those who did not	0.256 (0.124)	Significant and positive impact on positive sexual consent attitudes
Ortiz and Shafer (2018)	Define Your Line	Positive sexual consent attitudes	Significant increase from pre test to follow-ups	ES calculation not possibleb
Intentions/efficacy (n = 13)
Cares et al. (2015)	Know Your Power	Intent to help strangers	No significant difference between pre test and 12-month follow-up for the campaign-only group	ES calculation not possibleb	Overall impact appears to be non significant
Borsky et al. (2018)	Red Flag Campaign	Bystander intentions	No significant difference from pre test to post test for the intervention group, but a significant decrease for the comparison group	0.598 (0.198)
Carlyle et al. (2022)	Red Flag Campaign	Bystander intentions	No significant effect between those exposed to the Red Flag Campaign and those who were not	0.308 (0.184)
Katz et al. (2013)	Know Your Power	Bystander intentions	Those exposed to the campaign reported significantly higher willingness to help than those not exposed	0.456 (0.276)
Ortiz and Shafer (2018)	Define Your Line	Intentions to obtain sexual consent	Significant increase for treatment group at post test	ES calculation not possibleb	Significant impact on intent to gain consent
Cares et al. (2015)	Know Your Power	Contemplation	No significant difference between pre test and 12-month follow-up for the campaign-only group	ES calculation not possibleb	Overall significant impact on contemplation subscale
Potter et al. (2011)	Know Your Power	Contemplation	Those who reported seeing the posters reported significantly higher scores than those who did not	0.413 (0.127)
Potter (2012)	Know Your Power	Contemplation	Significant increase at post test for those who reported seeing the campaign more than once a day versus those who saw less than once a day	0.132 (0.119)
Hust et al. (2017)	Mini-magazine campaign	Intentions to get involved in prevention efforts	Significant increase at post test for the treatment group versus the control group	1.018 (0.297)
Cares et al. (2015)	Know Your Power	Bystander efficacy	No significant difference between pre test and 12-month follow-up for the campaign only group	ES calculation not possibleb	A non significant effect was primarily found, though a slight positive effect on levels of efficacy exists
Borsky et al. (2018)	Red Flag Campaign	Bystander efficacy	No significant difference from pre test to post test for the campaign-only group and a significant decrease for the comparison group	0.434 (0.199)
Carlyle et al. (2022)	Red Flag Campaign	Bystander efficacy	Those exposed to the Red Flag Campaign reported significantly higher levels of efficacy than those who were not exposed	0.445 (0.185)
Hust et al. (2017)	Mini-magazine campaign	Bystander efficacy	Significant increase from pre test to post test in the treatment group versus the control group	0.441 (0.287)
Behavior (n = 9)
Moynihan et al. (2015)	Know Your Power	Bystander behavior toward strangers	Significant decrease between pre test and 12-month follow-up	−0.328 (0.088)	Results appear mixed, with no clear indication of overall effectiveness
Borsky et al. (2018)	Red Flag Campaign	Bystander behavior	Significant increase at post test for the treatment group and non significant decrease for control	0.276 (0.193)
Mennicke et al. (2021)	Social marketing campaign	“intervened–unwanted” hitting on	No significant difference between years 1 and 5	ES calculation not possibleb
Potter et al. (2011)	Know Your Power	Action	Significantly higher action scores for those who saw the campaign versus those who did not	0.323 (0.126)	Overall mixed effects, but a slight positive trend for the action subscale
Potter (2012)	Know Your Power	Action	Significant increase from pre test to post test for both groups; however, those who saw the campaign less than once a day reported higher scores at post test than those who saw it more than once a day.	−0.074 (0.119)
Potter and Stapleton (2013)	Know Your Power	Action	No significant difference between those who reported seeing the campaign and those who did not	−0.095 (0.131)
Hust et al. (2017)	Mini-magazine campaign	Action	Significant increase from pre test to post test for treatment group versus control group	0.771 (0.292)
Thomas et al. (2016)	Consent is Good, Joyous, Sexy	Proactive bystander activities	No significant differences at post test between those who saw the banners and those who did not	ES calculation not possibleb
Mennicke et al. (2021)	Social marketing campaign	“Got consent before sexual intimacy”	Significantly higher rates in year 5 compared to year 1	ES calculation not possibleb	Significant and positive effect on prosocial sexual behaviors

^a Cares et al. (2015) and Moynihan et al. (2015) present various results using data collected from the same sample and campaign implementation.

^b We were unable to calculate effect sizes based on the available information in the study.

Knowledge (n = 4)

Three outcome measures examined knowledge of campaign content related to sexual violence; results appear promising with two outcomes reporting significant and positive results (see Table 3). Perceived norms were examined by Hust et al. (2017); results indicate that participants exposed to the mini-magazines had significantly more accurate perceptions of peer norms at post test. Altogether, findings regarding campaign impact on measures of knowledge gain/retention suggest that campaigns positively impact participants’ knowledge and may be an effective tool for educating college students.

Attitudes (n = 12)

Five studies examining rape myth acceptance produced mixed results; three studies noted no significant differences as a result of program exposure, while two presented a significant decrease in rape myth acceptance at post test (Cares et al., 2015; Mennicke et al., 2021). No significant effects were found for precontemplation (n = 3) or for attitudes toward bystander behavior (n = 2). Measures examining campaign impact on attitudes toward sexual consent (n = 2) produced significantly positive results. When assessed as a whole, results suggest that sexual violence campaigns are less effective at changing attitudes than imparting knowledge, with some moderate success found for attitudes concerning rape myth acceptance and sexual consent.

Intentions/Efficacy (n = 13)

Four measures in our sample were related to bystander intentions to intervene; overall, intentions do not appear to be influenced by campaigns as only one study reported a significant increase in willingness to intervene (Katz et al., 2013). With respect to participants’ intentions to gain sexual consent, Ortiz and Shafer (2018) found a significant increase at post test. Results for the contemplation subscale (n = 4) were overall positive with all but one study reporting a significant increase in student intent to participate in prevention activities after campaign exposure. Four measures assessed participants’ feelings of efficacy or confidence to intervene as a bystander; two studies reported a non significant effect, while two (Carlyle et al., 2022; Hust et al., 2017) had significant, positive findings.

Behavior (n = 9)

Last, results across the three bystander behavior measures were mixed with one significant negative effect, one significant positive effect, and one non significant effect. Results were also mixed for the action subscale (n = 5); two studies showed significant positive findings while three studies did not present statistically significant differences. Mennicke et al. (2021) examined prosocial sexual behaviors (i.e., “got consent before sexual intimacy”) and found significantly higher frequencies by participants in year 5 compared to those in year 1. In sum, although the effects of the campaigns collectively trend in a positive direction, the non significant and negative findings suggest that campaign impact on behavioral change is limited.

Discussion

Our systematic review of the literature on college sexual violence prevention campaigns yielded 15 studies that met the criteria for inclusion. Although we initially sought to conduct a meta-analysis to synthesize campaign impacts, the finite number of empirical evaluations that were amenable to meta-analysis restricted our investigation to a narrative review of the existing evaluations. Nevertheless, the current study provides an important overview of the current state of the evaluation literature on college sexual violence prevention campaigns.

The systematic search identified evaluations of eight individual campaign initiatives targeting sexual violence prevention. Across the eight campaigns, there are a number of similarities in content presented, targeted behaviors, and dissemination methods. With regard to content, the campaigns commonly targeted issues surrounding consent and bystander behavior. Campaigns such as the Consent is Sexy, Consent is Good, Joyous, & Sexy, and Define Your Line each focused on promoting positive sexual behavior and awareness of consent, while the Know Your Power campaign and the Red Flag Campaign both target sexual violence prevention via the promotion of bystander behavior. Both topics are important to sexual violence prevention strategies as they promote prosocial attitudes and behavior related to sexual violence (Gidycz et al., 2011), and can help to create positive change in the campus culture. The campaigns overwhelmingly favored visual print methods with posters or banners being the most commonly used form of dissemination (six out of eight campaigns featured these materials). This is not unexpected considering that this approach has traditionally been used in other media marketing campaigns (Untied & Orchowski, 2013). Additionally, these materials can be disseminated widely and with ease (Bridges et al., 2015).

One component that appears underutilized (at least in the current set of evaluated campaigns) is online materials and methods of dissemination. Recent research suggests that the increase in social media and online usage offers unique opportunities for prevention campaigns (Armstrong & Mahone, 2017; DeMaria et al., 2018). Yet, only two of the included campaigns indicated having any social media presence (i.e., Define Your Line and the Red Flag Campaign). As the literature on social marketing campaigns suggests that students have a higher interest in receiving and interacting with information via online or digital methods, web or social media components can offer additional opportunities for student exposure and engagement with the campaign materials and content. By focusing primarily on physical print materials without any online connection, campaigns may be losing valuable exposure opportunities and potentially inhibiting the overall impact of the campaign. Printed materials can be used as an effective tool to connect students to available online components or resources by including a QR code, hashtag phrase, or weblink (DeMaria et al., 2018), thus bridging the gap created by print materials in a digital age. The use of these additional online components appears to be increasing across college campaigns (e.g., #HUStands; Howard University, n.d.; #ItsOnUs; It’s On Us, n.d.). The inclusion of more comprehensive dissemination methods should not only be considered for future campaigns, but campaigns that do include these methods require increased evaluation.

Our review also summarized the impacts of existing sexual violence prevention campaigns on four primary outcome measures (i.e., knowledge, attitudes, intentions/efficacy, and behaviors). Overall, the results presented by the studies in the sample appear promising. With respect to knowledge, the findings are positive, suggesting that campaigns effectively increase students’ knowledge of sexual violence information such as what constitutes sexual violence, or what prosocial bystander behavior entails. Although knowledge alone cannot create cultural change, increasing awareness of the problem of sexual violence and knowledge of ways to prevent such violence is an important and necessary step for impacting the campus environment (Banyard et al., 2004).

Evaluation findings pertaining to attitudinal measures appear mixed; while measures related to rape myth acceptance and attitudes toward sexual communication show a positive pattern of results (indicating that campaigns have the potential to positively impact how participants view these topics), participant attitudes toward bystander behavior and attitudes toward sexual violence as a problem (i.e., precontemplation) seem less positively impacted by prevention campaigns. As sexual violence has historically been viewed as an individual problem (as opposed to a social problem; Banyard et al., 2004), shifting attitudes toward sexual violence as a community-wide issue may require more active engagement in order for positive modification to occur. Overall, these findings suggest that simply reading short messages or viewing scenarios on a poster may not be sufficient and that the passive dissemination of information may not be the best approach for affecting attitudinal change. Some attitudes may be better addressed by more formal in-person education methods (e.g., Bringing in the Bystander; Banyard et al., 2004), with campaigns providing supplementary content.

With respect to intentions, results suggest that intentions to intervene are not significantly impacted by campaigns, but intentions to get involved with prevention activities (i.e., contemplation) are more strongly impacted. This finding implies that students may be willing to take steps to help prevent sexual violence, but perhaps not if it requires them to actively insert themselves into situations in which sexual violence is occurring. Relatedly, findings regarding bystander efficacy indicate a trend toward a non significant effect, suggesting that participants are perhaps less willing to intervene in sexual violence situations because they are not confident in their abilities to successfully do so. Similar to attitudes, confidence, and willingness to intervene may require a more assertive approach than campaigns can offer in order for change to occur. Methods that require participants to actively engage with the materials may be better able to help students understand the intervention process, thus increasing their confidence in their own abilities and likelihood of intervening.

Last, the impacts of sexual violence prevention campaigns on several behavioral measures were mixed, with no clear indication of the impacts the campaigns had on those exposed. The negative effects are concerning, with two measures indicating students engaged in fewer prosocial behaviors after exposure to the campaigns. However, the negative impacts were found using a 12-month follow-up period (Moynihan et al., 2015), suggesting that short-term behavioral change may require additional intervention to maintain the effects. There was a slightly more definitive trend with students’ engagement in prevention activities (i.e., action), suggesting campaigns may help encourage participants to actively take part in prevention efforts on campus. As with bystander intentions, it may be that students are more willing to engage in broader efforts to prevent sexual violence, but are less willing to become involved in actual bystander situations.

In summary, our findings indicate that campaigns may have more notable impacts on certain measures of sexual violence prevention than on others. Campaigns show promising effects for increasing knowledge, affecting change in certain attitudes, intentions to participate, and actual participation in prevention activities (contemplation and action). Other outcomes including attitudes toward bystander behaviors, beliefs in sexual violence as a problem (precontemplation), bystander intentions, and bystander efficacy need more research to determine whether campaign methods are sufficient or if alternative methods (such as in-person workshops) are better suited for effecting change.

Limitations and Directions for Future Research

Overall, our systematic review identified several limitations and areas that require further consideration. First, there is a considerable lack of evaluative research on sexual violence prevention campaigns. While a basic Google search points to an exceptionally large number of prevention and awareness campaigns implemented at various colleges within North America, our comprehensive review indicates that relatively few of these campaigns have undergone empirical evaluation. Additionally, of the 15 studies that met the criteria for inclusion in this review, only eight unique campaigns were identified. As such, the current sample represents a fraction of the college sexual violence campaigns that exist, and conclusions regarding effectiveness are naturally limited to those campaigns that have been evaluated. As well, because many studies evaluated a marketing campaign in conjunction with another program (which prevented us from isolating the effects of the specific campaign) several otherwise eligible campaign evaluations were excluded from our sample. Future research should focus on using methodologically rigorous designs to examine the effectiveness of existing sexual violence prevention campaigns, separate from any accompanying programs.

Existing research examining non campaign-based prevention programming has also found evidence of differential program impacts based on various individual characteristics (e.g., gender [Inman et al., 2018; Moynihan & Banyard, 2008]; year of study [Kettrey & Marx, 2019]). As these types of variables were examined in a limited capacity within the included sample of studies, future evaluative research of sexual violence prevention campaigns should consider examining individual characteristics more closely. Additionally, future research should also consider whether prevention campaigns are successfully reaching current perpetrators of sexual violence. Social marketing campaigns generally target the overall campus population, and as such the included messaging may or may not be appropriately targeting those more prone to committing sexually violent acts. Last, there was little consistency across studies with respect to outcome measures. While we synthesized the findings into four primary outcome categories, there remains considerable variability within each category of outcomes. The variability across dependent variables makes it challenging to interpret the impacts that campaigns are producing. Evaluators should consider using validated measurement tools when possible, in an effort to produce clear, commensurable data that would be more conducive to summative research.

Conclusion

Given the current lack of summative research on the topic, this study provides a valuable contribution to aid in understanding the impacts of sexual violence prevention and social marketing campaigns on college campuses. An increase in rigorous empirical research of prevention campaigns is needed to better understand the effectiveness of existing campaigns and allow for future structured summative research to guide evidence-based best practices.

Author Biographies

Chelsey Lee is a doctoral student in the School of Criminology at Simon Fraser University. Her recent work has focused on gender-based violence, including examinations of university sexual assault policy and prevention programs, media representations of domestic violence and homicide, meta-analyses of adolescent and college dating violence prevention programs. She is also the project coordinator of an evaluation of sexual violence prevention activities at Simon Fraser University. Additional work has included meta-analyses of interventions to reduce criminal recidivism, including halfway houses and day reporting centers, and an evaluation of the Metro Vancouver YWCA Youth Education Programs.

Jessica Bouchard is a PhD candidate in the School of Criminology at Simon Fraser University. Ms. Bouchard has served as a Project Coordinator on several community-based crime prevention evaluation projects, which include an antigang messaging campaign, the Metro Vancouver YWCA Youth Education Programs, and 12 intimate partner violence intervention programs across British Columbia. Jessica has also worked with Dr. Wong on systematically reviewing and meta-analyzing the effects of community-based interventions such as aftercare/reentry programs for juvenile offenders, home confinement programs, Day Reporting Centers, and Teen Court programs.

Jennifer S. Wong is an associate professor in the School of Criminology at Simon Fraser University. Her work employs applied, quantitative, and qualitative methods to study issues of delinquency/crime prevention and intervention, focusing primarily on program evaluation and meta-analysis in the areas of crime prevention and intervention. Recent work includes evaluations of intimate partner violence intervention programs, examining the impact of prior deportation of illegal immigrants on recidivism, and meta-analyses on gang prevention programs, restorative justice programs for at-risk youth, halfway houses, and intensive supervision programs.