Practice Profile: Programs to Reduce Dating and Sexual Violence for Youth and Young Adults

De La Rue and colleagues (2014) conducted a meta-analysis to examine the effectiveness of intervention programs designed to reduce or prevent dating and sexual violence perpetration and victimization in intimate relationships among school-aged youth. Studies were included if they had a well-defined control group (randomized controlled trials, quasi-randomized controlled trials, and quasi-experimental designs); examined programs implemented in (public and private) middle and high schools (studies of programs that were not conducted in schools were excluded); had a primary goal of reducing or preventing teen dating violence or sexual violence; and measured the impact of the program on attitude change, frequency of intimate partner violence perpetration or victimization, teen dating violence knowledge, or the ability to recognize safe and unsafe behaviors in intimate partner disputes.

A comprehensive search of studies published between 1969 and 2013 was done to identify eligible studies. Electronic databases, government policy databanks, internet search engines, and grey literature databases were used. A total of 23 unique studies were found. Of these studies, 10 used random assignment, and 13 used non-random assignment. Participants in the studies were assigned at the individual, group, school, district, or state level. Except for one study, conducted in Canada, all the studies were conducted in the United States. There was a wide range of programs implemented. In most studies, teachers presented the curriculum to students (n = 15). Other studies used adults and professionals from the community. Regarding the age of study participants, 13 studies involved 9th–12th graders, and 10 studies involved 6th–8th graders.

There were a number of different outcomes reported in the identified studies, including dating violence knowledge, dating violence attitudes, dating violence perpetration, and dating violence victimization. The three-level random effects model was used to estimate the mean effect size for each outcome construct. Effect sizes were converted to the d-metric and Hedges’ g.

Fellmeth and colleagues (2013) conducted a meta-analysis to examine the effectiveness of educational or skills-based intervention programs designed to reduce or prevent dating violence perpetration and victimization in intimate relationships among youth and young adults. Studies were eligible if they used randomized, cluster-randomized, and quasi-randomized designs. Eligible participants included youth, ages 12 to 18, and young adults, ages 19 to 25. Programs could be universal or designed for high-risk groups and delivered in any setting and for any length of time. Programs were excluded if prevention of dating or relationship violence was not one of the stated aims or if the programs consisted of multiple programs in which it was not possible to isolate the relative effects of the violence prevention component. Programs were also excluded if participants were screened only for occurrence of dating or relationship violence and then referred to a support agency.

To identify potential studies, the authors searched online databases, hand-searched reference lists of key articles and two relevant journals and contacted researchers in the field. In total, 38 studies met the eligibility criteria. Of these, 5 studies were not included in the meta-analysis because they 1) provided only summary statistics (e.g., only F-tests, t-tests, p values); 2) used non-parametric tests; or 3) lacked sufficient information on the numbers of participants, means, and SDs in each group.

Of the 33 studies included in the meta-analysis, 16 were randomized controlled trials, 15 were cluster-randomized trials, and 2 were quasi-randomized trials. Except for one study conducted in the Republic of Korea, all included studies were conducted in the United States. Most studies were carried out in educational settings (19 in universities, 5 in high schools, and 3 in middle schools), three took place in community settings, one studied young pregnant women attending health clinics, one studied adjudicated youth in a prison and courtroom setting, and one studied teenagers with a history of maltreatment who had been recruited from community centers. Most programs were designed for general audiences; however, five were targeted at individuals at high risk of experiencing or committing relationship violence. Ten studies included only males, and five included only females. Programs were predominantly educational, although five included a self-defense component, and one provided a component on communication skills. The duration of the programs ranged from one 50-minute session to 18 sessions delivered over 4 months. Follow-up periods varied. Thirty studies assessed short-term outcomes (0 to 6 months following the program), four assessed medium-term outcomes (6 to 12 months following the program), and four assessed short-term outcomes (more than 12 months following the program). Participants ranged in age from 10 to 44 (Please note that studies were included if more than 80 percent of the participants included in the study were within the age range of 12 to 25 years resulting in a wide age range.). Information on the race/ethnicity of participants was not provided.

Outcomes included episodes of relationship and dating violence, behavior in relationship and dating violence, knowledge of relationship and dating violence, and attitudes toward relationship and dating violence. The researchers used random effects and fixed effects models to assess each outcome—except for the behavior outcome, which showed no heterogeneity and therefore was only assessed using a fixed-effect model.

Lee and Wong (2020) conducted a meta-analysis to examine the overall effectiveness of dating prevention programs on increasing knowledge about dating violence, changing attitudes toward dating violence behaviors, increasing bystander intentions and/or behaviors, and reducing incidents of dating violence perpetration and victimization. The search for studies was conducted in 23 electronic databases (e.g., Criminal Justice Abstracts, PsycINFO). Grey literature was searched for unpublished works such as theses and dissertations, conference presentations, technical reports, and independent projects. The reference lists of included studies were also hand-searched to identify additional studies, as were the reference lists of identified meta-analyses or literature reviews focused on dating violence prevention or relationship education programs. Studies were included in the analysis if they 1) evaluated a program focused on dating violence prevention or education for a youth population under the age of 18; 2) included at least one quantitative outcome measure relating to knowledge about dating violence, attitudes toward dating violence, dating violence victimization, perpetration, or bystander behaviors; 3) provided enough data and information to calculate an effect size; 4) used a control or comparison group and/or a pre-test–posttest design and scored a 2 or higher on the Maryland Scale of Scientific Methods (i.e., “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); 5) used a sample of at least 20 participants; 6) were published in English or French; and 7) were conducted in North America, Western Europe, Australia, or New Zealand. Because of the heterogeneity of participants’ ages, the authors did not consider pooling outcomes for programs that targeted middle/high school students with college students.

The systematic search produced 38 studies contributing to 73 independent effect sizes. Studies included in the meta-analysis were published between 1997 and 2019, and most were published in peer-reviewed journals (82 percent). More than half of the studies used a randomized controlled trial design (58 percent), 13 percent used a quasi-experimental design, and the remaining 29 percent of studies did not specify the type of study design used. The mean age of the participants was 14.5 years. Of the studies reporting the racial/ethnic composition of the sample, 55 percent used a sample that primarily included minority participants, and 29 percent used a white or mixed ethnicity sample. The programs were mostly delivered in a classroom (74 percent), small group (16 percent), or another setting (11 percent). Less than a third of the programs incorporated bystander training (29 percent). Most programs were delivered in 2 to 10 hours (60.5 percent).

The outcomes of interest for the meta-analysis were 1) knowledge of dating violence (e.g., recognition of dating violence behaviors); 2) attitudes toward dating violence (e.g., beliefs about if/when violence is appropriate, acceptance of rape myths); 3) perpetration of dating violence (e.g., hitting, slapping, name-calling); 4) experiences of dating violence victimization (e.g., being hit or verbally abused by a dating partner); and 5) bystander behaviors and intentions (e.g., “if your friend called their partner a rude name, how likely is it that you would say something?”). Individual study findings were pooled by outcome measure into a standardized format. A standardized mean difference was used as the primary effect size type. A random effects models was utilized.

Piolanti and Foran (2021) conducted a meta-analysis of randomized clinical trials to assess the efficacy of prevention programs for dating violence in adolescence. A systematic review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) reporting guideline. Several databases were searched through April 2021. Other search strategies included checking reference lists of earlier systematic reviews and meta-analyses, and independently searching for protocols in the ClinicalTrials.gov database. Included studies had a randomized design examining the efficacy of a program to reduce teen dating violence, compared with a control group. Additionally, at least one measure of sexual or physical dating violence perpetration or survivorship needed to be included in the study. Participants had to be 18 years or younger. Non-randomized comparative studies, observational studies, quasi-experimental studies, and studies published in languages other than English were excluded.

The systematic review resulted in 18 studies. One study was conducted in Europe, and the others were conducted in North America. A total of 22,781 youths were included in the sample. The mean age of participants ranged from 12.2 to 17.6 years. Thirteen programs were implemented in schools, and five were implemented in other settings (i.e., at home, a health department, a pediatric department, child protective services, or another community setting). Primary components of programs included group discussions or sessions, individual interviews, parent–child activities, classroom activities, and a combination of classroom–group activities and parent–child activities. The number of sessions ranged from 1 to 24, and the overall length of the program ranged from 1 day to 2 years.

Outcomes were pooled using a random effects design. Meta-analyses were performed on three different outcomes: 1) sexual dating violence, 2) physical dating violence, and 3) a composite measure of sexual and physical dating violence. Separate analyses were conducted on survivorship and perpetration for each outcome. The authors combined the measures of physical/sexual violence and perpetration/survivorship into a single composite overall outcome for all studies. The between-group effect size was computed as the difference between the treatment and control groups at posttest or follow-ups at 3 years or less, by calculating the odds ratios. Means and standard deviations were used to compute effect sizes for continuous data, which were subsequently converted to odds ratios.