Science1 distinct publisher3 min readUpdated
A review of 40 studies since 2012 reports few if any programs changed victimization or perpetration. Thirty-seven of the 40 came from the US or Canada.
The Scientist · Science desk
Compiled by The ScientistSomething wrong?How this is made
A new review of sexual violence prevention programs at universities in high-income countries reports that the dominant model has not moved the outcome it is sold on: few, if any, of the programs studied produced significant changes in victimization or perpetration [3][5]. That finding lands in Australia at an awkward moment, because the National Higher Education Code to Prevent and Respond to Gender-based Violence, enacted earlier this year, now requires universities to implement evidence-based prevention programs [10].
The scale of the problem is not in dispute. The recent National Student Safety Survey found that one in six students at Australian universities had been sexually harassed and one in 20 had been sexually assaulted since starting university [1], which is roughly 17 percent and 5 percent respectively [17]. Students facing intersecting forms of inequality, including LGBTIQA+ students, international students and students of color, are more likely to experience it [2].
What the review challenges is the prevailing intervention design. Since 1985, most prevention work in any setting has aimed at changing attitudes, beliefs and knowledge, often through bystander approaches that train people to intervene safely when they witness harmful behavior [4]. Attitudes are measurable and cheap to measure. Behavior change has not followed [5].
The evidence base itself is narrow. The authors examined 40 studies of university prevention programs published since 2012 [6]. Only three came from outside the United States or Canada, leaving 37 from North America [7], or about 93 percent of the sample [8]. The studies make universal claims, but the authors say there is no guarantee they hold in different social and cultural contexts [11]. Participants also skewed toward majority-white, cisgender heterosexual women, with limited focus on LGBTIQA+ communities, international students or staff, and students of color [9]. So a mandate to use evidence-based programs points Australian universities at a literature built somewhere else, on a demographic slice that is not the one most exposed [9][10][11].
The authors benchmarked the studies against the whole-of-university framework from the primary prevention body Our Watch, which maps change across five domains (student life, teaching and learning, the workplace, research, and business and operations) and three levers (values and behaviors, structures and policies, and culture and norms) [12]. Measured that way, the programs clustered in teaching and learning while neglecting student life, business and operations, workplace culture and policies, and equity in research [13]. That is the shape of a compliance artifact: a module you can assign, track and report, sitting inside institutions whose hierarchies and power dynamics the review argues are themselves part of the problem [13][15].
Their prescription is structural rather than curricular. Prevention has to move past one-off interventions and piecemeal education to enable program and policy development, empower victim-survivors and bystanders, and break down power dynamics and gendered norms [16]. Pressure for that shift has been building since the 2015 US documentary "Hunting Ground" and a series of Australian reports that documented no improvement in rates [14].
Watch how regulators interpret "evidence-based" under the Code [10]. If the standard is a citation, the workshop survives. If the standard is demonstrated reduction in incidence, almost nothing in the current literature qualifies [5][7].
Follow any of these and your For You feed starts watching them — no settings page required.
Ranked by verification strength, evidence, and original report placement.
Behavior change has not followed the attitude-focused approach, and few, if any, of these programs have resulted in significant changes in victimization or perpetration.
A recent National Student Safety Survey found one in six students attending Australian universities had been sexually harassed and one in 20 had been sexually assaulted since starting university.
Students who face intersecting forms of inequality, including LGBTIQA+ students, international students and students of color, are more likely to experience sexual violence.
The authors' latest paper reviews research on sexual violence prevention programs at universities in high-income countries, assessing how effective they have been and what blueprint they set for future research-informed approaches.
Since 1985, most efforts to prevent sexual violence in any setting have focused on changing attitudes, beliefs and knowledge, often through bystander approaches that train people to intervene safely when they witness harmful, disrespectful or dangerous behavior.
The review examined 40 studies of sexual violence prevention programs in universities published since 2012.
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
Peer-reviewed scoping review, reported through one outlet
The underlying artefact is a named, citable scoping review (Ison et al., Trauma, Violence, & Abuse, 2026) with concrete counts — 40 studies since 2012, three from outside the US or Canada — and an explicit benchmark framework. That lifts it above commentary. But every fact reaches the cluster through a single author-written item; the central effectiveness finding is stated qualitatively as 'few, if any' with no effect sizes, program identities or study-level outcomes, and the prescriptive conclusions are argued rather than measured.
Programs widely run and now legally mandated, but narrowly scoped
Adoption of campus prevention programming is real and, in Australia, compelled: the new Code requires every university to implement evidence-based prevention, and 40 evaluated programs exist across high-income countries since 2012. What holds the score down is the shape of that adoption — concentrated in the teaching-and-learning domain and one behavioural lever, overwhelmingly evaluated in the US and Canada, and with no disclosed count of Australian institutions that have implemented anything under the Code.
Cluster framing generalises past what the review measured
The reporting is largely deflationary — it argues current programs do not work — so it is not promotional hype. The overstatement runs the other way: the review's finding that 'few, if any' evaluated programs produced significant change is presented as a settled conclusion that campus workshops do not cut incidence, when the corpus is 37 North American studies, no effect sizes are given, and absence of demonstrated effect is not the same as demonstrated absence of effect. The prescribed whole-of-university alternative is also advocated with confidence despite the article conceding primary prevention is relatively new and unproven.
Authors summarising their own paper and arguing for more prevention research
The item is written by the review's own authors and closes by calling for the development, implementation and evaluation of new prevention programs backed by funding and resources — an agenda their field directly benefits from. It also arrives at the moment a new legal mandate creates demand for exactly the 'right evidence' they say is missing. Nothing in the cluster suggests fabrication, and the underlying paper is peer reviewed, but the framing incentive is visible and undisclosed, and no adversarial voice is included.
Traceable artefact, single publisher, no corroboration
Confidence is moderate: the specific counts and the framework benchmark are precise and attributable to a citable peer-reviewed paper, which makes the factual spine reliable. But the cluster has one publisher and one item, the central effectiveness claim is unquantified, the interpretive claims about structural drivers and the whole-of-university remedy are unevidenced within the supplied material, and there is no independent confirmation of the Code's requirements or of any institutional response.
Distinct publishers with included, body-backed reporting in this cluster.
1 article · August 15, 2026