Journal of Adolescence 92 (2021) 194–236

Contents lists available at ScienceDirect

Journal of Adolescence
journal homepage: www.elsevier.com/locate/adolescence

Review article

Intervention programs designed to promote healthy romantic

relationships in youth: A systematic review
Emily Hielscher a, b, c, *, Catherine Moores b, d, 1, Melanie Blenkin e, 1,
Amarzaya Jadambaa a, f, James G. Scott a, b, c, e, g
a
QIMR Berghofer Medical Research Institute, Herston, QLD, Australia
b
Centre for Clinical Research, Faculty of Medicine, The University of Queensland, Brisbane, QLD, Australia
c
School of Public Health, Faculty of Medicine, The University of Queensland, Brisbane, QLD, Australia
d
Department of Anesthesia, McMaster University, Hamilton, Canada
e
Metro North Mental Health, Royal Brisbane and Women’s Hospital, Herston, QLD, Australia
f
Australian Centre for Health Services Innovation and Centre for Healthcare Transformation, School of Public Health and Social Work, Queensland
University of Technology, Kelvin Grove, QLD, Australia
g
Queensland Centre for Mental Health Research (QCMHR), The Park Centre for Mental Health, Wacol, QLD, Australia

A R T I C L E I N F O A B S T R A C T

Keywords: Introduction and Objectives: To conduct a systematic literature review of intervention programs
Adolescent designed to promote healthy romantic relationships in youth (aged 12–25 years). The focus was
Young person on universal interventions that have the potential to be effective and widely implemented.
Dating
Methods: We systematically searched PubMed, PsycINFO, Social Science Database, and Embase.
Randomised control trial
Systematic review
Articles were included if they were a randomized controlled trial (RCT) or quasi-experimental
study and reported on a universal intervention aimed at promoting healthy romantic relation­
ship knowledge, attitudes, and behaviors among youth.
Results: The search strategy identified 27 studies (26,212 participants). Interventions were found
to be effective for improving healthy romantic relationship knowledge in the target population.
However, the findings were mixed for intervention effectiveness in changing relationship atti­
tudes/beliefs, and there was limited evidence to support change across behavioral outcomes.
Conclusions: This review highlights the need for future research, including high quality RCTs with
longer follow-up periods in a broad range of cultural and ethnic settings, to improve the gen­
eralisability of findings. Interventions for adolescents that improve knowledge and behavioral
change relating to healthy romantic relationships have the potential to reduce mental and
physical health problems during this phase of development.

1. Introduction

Youth, or young people under 25 years, commonly experience unhealthy romantic relationships with many reporting psycho­
logical, physical, sexual, and other forms of in-person and online abuse (Halpern et al., 2001; Stöckl et al., 2014). Youth relationship
abuse and violence is well researched, but what constitutes a healthy romantic relationship has received less attention. Just as it is

* Corresponding author. QIMR Berghofer Medical Research Institute, 300 Herston Rd, Brisbane, QLD, 4006, Australia.
E-mail address: [email protected] (E. Hielscher).
1
These authors contributed equally and have joint second authorship.

https://doi.org/10.1016/j.adolescence.2021.08.008
Received 29 November 2020; Received in revised form 27 June 2021; Accepted 10 August 2021
Available online 25 September 2021
0140-1971/© 2021 The Authors. Published by Elsevier Ltd on behalf of The Foundation for Professionals in Services for Adolescents. This is an
open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

useful to define wellness as more than the absence of disease, it is important to establish meaningful frameworks for understanding
what constitutes a healthy relationship. This informs public health initiatives that promote the knowledge and skills necessary for
experiencing healthy romantic relationships in adolescence and beyond.
Unhealthy relationship behaviors are highly prevalent in young people’s romantic relationships and are associated with a myriad of
poor health and psychosocial outcomes. Among youth aged 15–24 years, lifetime prevalence of any intimate partner violence in a
World Health Organisation (WHO) multi-country study ranged from 19% to 66% (Stöckl et al., 2014). This is comparable to rates
reported in a nationally representative sample of Australian youth, where almost 25% of respondents (12–20 years) reported expe­
riencing physical violence by a romantic partner (Halpern et al., 2001). Prevalence of abuse in young people’s romantic relationships
seems to differ by type of violence. Psychological abuse appears the most prevalent, where up to 96% of youth report being victims to
this (Finkel et al., 2009). This is followed by moderate physical abuse, severe physical abuse, and lastly sexual dating abuse (Foshee
et al., 2009; Halpern et al., 2001).
Youth exposed to unhealthy romantic relationships are at increased risk of experiencing a range of mental and physical health
problems, both in adolescence and later life e.g. depression, chronic pain, and gastrointestinal disorders (Noonan & Charles, 2009).
Other studies have found physical or sexual abuse in an adolescent relationship is associated with anxiety, suicidality, substance abuse,
eating disorders, sexually transmitted diseases, unwanted pregnancy, and reduced quality of life (Bonomi et al., 2013; Miller, 2017).
While there have been extensive investigations on the prevalence and impacts of youth relationship abuse, there are a paucity of
studies examining healthy romantic relationships. The few studies (Finkel et al., 2009; Noonan & Charles, 2009) that have focused on
healthy relationships, tend to frame them in terms of the absence of unhealthy dynamics. Beyond the academic literature, youth
organisations focused on the promotion of healthy romantic relationships (e.g. Love Is Respect, That’s Not Cool) define such re­
lationships in terms of common sense, with no or limited overarching framework or stated basis in theory.
In the absence of any studies that adequately define healthy romantic relationships, the authors of the current review propose a
definition based on existing romantic relationship theories, and wider literature on adolescents’ perceptions of a healthy romantic
relationship – to provide an overall framework for the review. Relatively contemporary theories such as the Behavioral Systems
Approach (BSA) Theory (Furman & Wehner, 1994), integrates Attachment Theory (Ainsworth, 1989; Bowlby, 1969) and Sullivan’s
Theory of Social Needs (Sullivan, 1953) to conceptualise the role that adolescent romantic relationships play in development, ac­
cording to four systems (i.e. attachment, affiliation, caregiving, and sexual/reproductive function). The Duluth Equality Wheel/Model
for Healthy Relationships (Pence & Paymar, 1993) also offers a contemporary and comprehensive framework, with six concrete
features (self-expression, respect, trust and support, honesty and accountability, shared responsibility, and negotiation/fairness) for
understanding healthy, non-violent relationships. Two elements of the BSA Theory (attachment and sexual/reproductive function) are
characterised by concepts also captured by the Duluth Model. Consequently, a comprehensive and theory-driven definition of a
‘healthy romantic relationship’ should integrate the remaining two BSA systems (affiliation, caregiving) with the six key features of the
Duluth Healthy Relationship Wheel. In summary, a healthy romantic relationship can be defined as one characterised by strong
communication and negotiation skills, caregiving behaviors, self-expression, respect, trust, honesty, and fairness. These characteristics
were considered necessary on top of the absence of an unhealthy romantic relationship, which was defined as a relationship char­
acterised by features referred to as Adolescent Relationship Abuse (ARA) (Dick et al., 2014; Miller, 2017), including verbal, emotional,
psychological, and sexual abuse.
Perhaps because a ‘healthy relationship’ has proven hard to define, much of the intervention work in this area has been over­
whelmingly focused on preventing and intervening youth relationship abuse. However, interventions designed to not only reduce or
prevent relationship abuse, but also actively promote healthy romantic relationship knowledge and skills have the potential to sub­
stantially improve youth health and wellbeing. In doing so, such intervention programs shift the focus towards a broader, public health
approach of healthy relationship promotion. Preliminary evidence suggests that interventions focused on fostering healthy romantic
relationships could be effective among young people (Bradford et al., 2016; Larson et al., 2007; Schramm & Gomez-Scott, 2012).
Moreover, interventions designed to promote healthy relationship skills and knowledge and/or reduce unhealthy relationship among
school students began to be rolled out and evaluated prior to the 2000s in the United States (Avery-Leaf et al., 1997; Foshee et al.,
1998) and have become increasingly common around the world (Carrascosa et al., 2019; Dos Santos et al., 2019; Munoz-Fernandez
et al., 2019).
To fully understand their efficacy and effectiveness, it is important to systematically review the findings across multiple studies;
with consideration for intervention-related change across a broad range of knowledge, attitudinal and behavioral outcomes. The
current review aimed to conduct a systematic literature search of intervention programs designed to foster healthy romantic re­
lationships among youth (aged 12–25 years) to summarise the present state of empirical research. This review focused on young people
aged 12–25 as during adolescence and young adulthood, fundamental cognitive, physical and emotional development processes take
place (Borzekowski, 2009) and health and relationship-related behaviors and skills develop; making it an ideal period to intervene and
support people in their emerging romantic lives. The focus was also on universal interventions that have the potential to be widely
effective and implemented.

2. Methods

This systematic review adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) (Moher
et al., 2009). It was registered with PROSPERO (registration number: CRD42017065191).

195
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

2.1. Definitions

2.1.1. Healthy romantic relationship

For this review, a healthy romantic relationship was defined as one characterised by: strong communication and negotiation skills,
caregiving behaviors, self-expression, respect, trust, honesty and fairness. These characteristics were considered necessary in addition
to the absence of relationship abuse.
There were a paucity of studies defining healthy romantic relationships in the literature (Finkel et al., 2009; Noonan & Charles,
2009). As mentioned previously, the authors of the current review proposed this definition based on existing romantic relationship
theories and wider literature on adolescents’ perceptions of a healthy romantic relationship (see Appendix A for more details on the
definition generation process).

2.1.2. Unhealthy romantic relationship

Because healthy relationship interventions generally measure their effectiveness by a reduction in negative or unhealthy rela­
tionship outcomes (e.g. decreased relationship abuse), an amplified search strategy was used (see Search Strategy/Eligibility Criteria
sections below). This was designed to initially capture studies that reported on either (1) an intervention aimed at promoting healthy
romantic relationships, or (2) an intervention aimed at reducing unhealthy relationship dynamics/adolescent relationship abuse
(ARA). Studies were carefully screened to determine if the intervention design included a healthy relationship component.
For this review, an unhealthy romantic relationship was defined as a relationship characterised by features referred to as ARA (Dick
et al., 2014; Miller, 2017), including: verbal, emotional, psychological, and sexual abuse. This review focused on more subtle forms of
abuse, considering physical violence is a well-recognised marker of unhealthy relationships and stands at the other end of the spectrum
(Miller, 2017) from a healthy romantic relationship. As discussed by Miller (2017), the term ‘ARA’ tends to be more accurate than
‘violence’ in capturing the varied unhealthy relationship dynamics experienced during adolescence. Considering this, physical forms of
intimate partner violence were not explicitly included in the search strategy.

2.2. Search strategy

The following electronic databases were searched from inception to February 2020: PubMed, PsycINFO, Social Science Database
and Embase (first round of searches were conducted from inception to September 2017 by the first reviewer (CM); a subsequent search
using the same strategy was conducted from September 2017 to February 2020 by the second reviewer (MB)). Additional unique
records were identified through citation databases, including Web of Science, Scopus, and Google Scholar. Reference lists of included
articles were hand-searched. Authors of included articles were contacted to obtain further details if required.
Search terms were sourced from previous systematic reviews on ARA (De Koker et al., 2014; Lundgren & Amin, 2015; Simpson
et al., 2017) and from terms identified in the definition of healthy/unhealthy romantic relationships. See Appendix B for the full
PubMed search strategy.

2.3. Eligibility criteria and study selection

Inclusion criteria were studies that (a) were peer-reviewed and published in the English language; (b) were randomized controlled
trials (RCTs) or quasi-experimental studies (i.e. studies with a non-randomly allocated comparison group); (c) reported on a universal
intervention aimed at fostering and promoting healthy romantic relationships, which could either be an intervention focused on (1) the
promotion of healthy relationship skills/knowledge, or (2) an intervention focused on reducing or preventing various “subtle” forms of
relationship abuse (i.e. verbal, emotional, psychological), but with a healthy relationship skills/knowledge component; (d) had a clear
pre/post evaluative component in order to be able to assess the effectiveness of an intervention; and (e) were conducted in a sample of
young people aged 12–25 years. If the mean age of the population sample fell within this range the study was included.
Exclusion criteria were studies limited to (a) people in relationships who were below the age of 12, or above the age of 25 years; (b)
platonic relationships or friendships; (c) interventions focused solely on physical relationship abuse/violence (e.g. hitting, choking,
burning); and (d) studies conducted in restricted samples i.e. in clinical, help-seeking, vulnerable or at-risk populations. In addition,
the review did not include sex education intervention studies, unless the intervention had a primary focus on the development of
broader romantic relationship knowledge, attitudes, and skills.

2.4. Data extraction and assessment of study quality

Eligible studies were downloaded into EndNote. Two reviewers extracted the data independently (CM, MB), which was reviewed by
a third reviewer (EH). Any discrepancies were resolved through discussion. Data on the study design, population, intervention and
control conditions, primary/secondary outcomes, key results, and completion rates were extracted and tabulated in excel by the first
and second reviewer.
The quality of RCTs was assessed using the Cochrane Collaboration Risk of Bias Tool (Higgins et al., 2019). Risk of bias judgements
(low, high, or unclear risk) were made by first and second reviewers, in discussion with a third reviewer (EH). If a study was a cluster
RCT, modifications were applied (Eldridge et al., 2016). For non-randomized studies, risk of bias was assessed using the ROBINS-I
(“Risk Of Bias In Non-randomized Studies - of Interventions”) tool (Jonathan et al., 2016). To maintain consistency with the
Cochrane risk of bias tool, the categories of ‘low’, ‘high’ and ‘unclear’ risk were assigned (instead of the original ROBINS-I judgement

196
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

categories of moderate, serious, and critical). Additionally, interventions of included studies were assessed against the proposed
healthy relationship definition (see ‘Definition’ section) by two reviewers (AJ and EH), independently, and disagreements were
resolved by consensus discussion. Each intervention was assessed to determine whether each of the key defintion characteristics (i.e.
‘strong communication and negotiation skills’, ‘caregiving behaviors’, ‘self-expression’, ‘respect’, ‘trust’, ‘honesty’, and ‘fairness’) was
covered or addressed as part of the (1) intervention design and (2) study outcome measure/s. Assesment categories included: “covered”
(indicated by green), “not covered” (indicated by red), and “not clear” (indicated by yellow) i.e. unclear as to whether intervention
targeted or promoted the healthy relationship characteristic/s.

2.5. Data synthesis

The overall literature was critically reviewed via a narrative synthesis. This was deemed more appropriate than a formal meta-
analysis because the results and outcomes of interest were too heterogeneous to be pooled. This synthesis focused on: target popu­
lation characteristics (including consideration for diversity of samples), intervention type and content, and intervention outcomes.
Intervention outcomes were broken down into (1) interventions with a primary focus on promotion of healthy relationship knowledge/
skills, and (2) intervention focused on reducing or preventing various “subtle” forms of relationship abuse (i.e. verbal, emotional,
psychological) but with a healthy relationship skills/knowledge component.

3. Results

3.1. Characteristics of included studies

This systematic search identified 27 individual studies (from 30 publications) reporting on 26,212 participants (M = 955; Range =
33–2605). Because the study authors presented different outcome variables of interest, and conducted different analyses with such
variables, the review authors present the results section in terms of number of publications, as opposed to individual studies/datasets.
This approach was needed as several publications reported on the same intervention (using the same study cohort) (e.g. Safe Dates
Curriculum) (Foshee et al., 1998, 2000, 2004, 2005).
Table 1 provides an overview of descriptive information for the 30 included publications. Most studies were RCTs (n = 21) and the
remainder were quasi-experimental or non-randomized control group, pre-test-post-test designs (n = 9) (see Fig. 1 and Table 1). The
majority of studies were conducted in the United States (77.8%), with others conducted in Spain, Brazil, Mexico, Canada, and the
Netherlands. Most were conducted in educational settings, including middle school (grades 6–8; n = 11), high school (n = 12), middle/
high school (n = 2), universities (n = 3), and other settings (n = 2). In terms of other settings, Bradford et al. (2016) recruited
adolescent participants from eight suburban communities through newspaper advertisements, and Foshee et al. (2012) recruited

Table 1
Descriptive information of 30 publications included in the systematic review.
Descriptive variable Number of publications

Study design

Randomized control trial 21

Quasi-experimental or non-randomized control group, pre-test-post-test designs 9

Study location

USA 23
Spain 3
Brazil 1
Mexico 1
Canada 1
The Netherlands 1

Study settings

High school (grades 9–12 in USA) 12

Middle school (grades 6–8 in USA) 11
College/University 3
Middle/High school 2
Othera 2

Intervention type

Healthy/unhealthy relationship interventionb 20

Healthy relationship interventionc 10

Note:
a
Participant over 18 years from urban/suburban communities or caregivers and their teenagers.
b
An intervention aimed at reducing unhealthy relationship dynamics/adolescent relationship abuse (but with a healthy relationship skills/
knowledge component).
c
An intervention aimed at promoting healthy romantic relationships.

197
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Fig. 1. PRISMA flow chart.

caregivers and their 13-15 year-old teenagers via a sampling frame of 3134 American households. Ten studies reported healthy
relationship interventions, while 20 studies focused on interventions targeting both healthy and unhealthy relationship components.
See Table 2 for a full description of each study’s design and sample.

3.2. Intervention design

There were a variety of universal interventions employed in the included studies (see Table 2). Nearly all interventions (83.3%)
were delivered in-person as part of the existing curriculum in school. However, two interventions were delivered via mail (Bradford
et al., 2016; Foshee et al., 2012), two were delivered online (Larson et al., 2007; Widman, Evans, et al., 2020), and another was
delivered via routine school health centre visits (Miller, Williams, et al., 2015). The dose of interventions ranged from 1 × 45-min
session up to 12 x 1-h sessions. Almost half of the studies reported a theoretical basis (n = 13; 43.3%), including: social learning and
cognitive theory, protection motivation theory, and theory of reasoned action.
Most studies (n = 20; 66.6%) employed an intervention focused on reducing or preventing various forms of relationship abuse (as a
primary aim), plus promoting healthy relationship skills/knowledge (secondary aim). Eight interventions focused solely on the pro­
motion of healthy relationship skills and knowledge (see Table 2). All of these were education curricula and included lessons which
covered topics such as: building healthy relationships, personal boundaries, signs of abuse, ending unhealthy relationships, and
relationship communication. In addition to such topics, the PARE intervention (Lederman et al., 2008) included the participants’

198
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 2
Study characteristics and intervention design (n = 30 publications).
Author (year) Country N Study type Study population Intervention (duration, components,
theoretical basis)

Healthy relationship interventions (RCT studies n = 6)

Coyle et al. (2019) Netherlands 911 Cluster RCT Nine participating urban middle A healthy relationship-based
schools (grades 6–8) in three urban program, You-Me-Us. It’s a theory-
school districts (mean age 12.4 based intervention with two
years). Both male and female components: (1) a 12-session
students. healthy relationships-based
curriculum in 7th grade science
classes taught by trained health
educators hired by the project, and
(2) a school-wide norms component
that featured a small group of
students (called a peer team) who
developed activities to reinforce key
program messages outside the
classroom. The curriculum,
grounded in Social Cognitive Theory
(Bandura, 1986) and Positive Youth
Development principles (Damon,
2004), included content related to
building healthy friendships and
relationships, communicating
effectively, influences on sexual
expectations in relationships,
personal boundaries, navigating
situations that could challenge
personal boundaries, ending
unhealthy relationships, sexually
transmitted infections (including
HIV), and the use of condoms and
other contraceptives.
Implementation was three to four
activities per year per school.
Kerpelman et al. USA 1824 RCT High school students in family and 6-week, 50-90-min classes of the
(2009) consumer science classes in Alabama Relationship Smarts Plus curriculum.
Topics include: (1) identity
formation and future planning; (2)
mature versus impulsive decision
making; (3) healthy and unhealthy
relationships; (4) dating aggression
and abuse; (5) communication
styles; (6) importance of adult
relationships for child well-being;
and (7) risk and protective factors
for healthy relationships in
adulthood.
Larson et al. (2007) USA 78 RCT College-aged students at a college in 1 session, 2 intervention groups: 1.
Western USA, who were in serious Self-Interpretation group: completed
relationships RELATE programming online and
interpreted their results without
therapist assistance, 2. Therapist-
Assisted group: completed RELATE
programming online and interpreted
their results with the assistance of a
therapist. Pre-marital assessment
questionnaire: designed for casually
dating to engaged. Assists couples in
clarifying perceptions, highlighting
areas of agreement and
disagreement, and inviting
discussion of individual and couple
strengths and challenges to later
marital satisfaction. Individuals rate
themselves and their partners of a
number of factors including
individual personality traits, couple
traits, current relationship, etc.
RELATE report generated organizes
(continued on next page)

199
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 2 (continued )
Author (year) Country N Study type Study population Intervention (duration, components,
theoretical basis)

and compiles responses to be

compared against one-another.
Premise that each relationship is
constructed of mutual and continual
interactions between partners in
various contexts.
Lederman (2008) USA 192 RCT Adolescents (11–15 years) and their 18 h over 2 years of
parents in middle schools in Parent–Adolescent Relationship
Southeast Texas Education (PARE), after-school
prevention education curriculum:
small-group sessions including
communication about sex, value of
others’ opinions, parental rules
about adolescent behavior, parental
monitoring, as well as knowledge,
self-efficacy, and behavioral options
for prevention of pregnancy and
STDs. Social learning and behavioral
cognitive theories
Scull et al. (2018) USA 1030 Cluster RCT One large school district (with nine Media Aware is a teacher-led
middle schools) located in the South- comprehensive sexual health and
eastern U.S. MLE (media literacy education)
program designed for middle school
with the aim of improving sexual
health outcomes by reaching
students prior to the onset of sexual
activity. Program development was
informed by the theoretical
frameworks of both the TRA (theory
of reasoned action)/TPB (theory of
planned behavior) and the MIP
(message interpretation processing
model) model; through changing the
way that media messages are
processed and by providing
medically-accurate health
information and skill development,
the program was designed to impact
key constructs outlined in the model
related to healthy sexual decision-
making. It includes ten lessons and
covers a wide range of sexual and
relationship health topics including
lesson 4 on healthy and unhealthy
relationships. The curriculum in the
intervention group was taught by
7th and 8th grade health teachers
who were trained to deliver the
media aware program prior to its
commencement.
Widman, Evans, et al. USA 226 RCT High school students from a rural Single 45-min interactive, skills-
(2020) high school in the South-eastern focused online sexual health
United States. Mean age = 16.3. 58% program for adolescents, called
girls, 39.8% boys. Health Education and Relationship
Training (HEART for teens online
program). The program was
originally developed and evaluated
among adolescent girls (HEART for
Girls); the current project describes
and evaluates a new version of the
program that was adapted for boys
and girls. The content is grounded in
psychological and health behavior
change theories, including the
Reasoned Action Model and Fuzzy
Trace Theory. They each include five
modules that target five areas of
sexual decision-making: 1) safer sex
(continued on next page)

200
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 2 (continued )
Author (year) Country N Study type Study population Intervention (duration, components,
theoretical basis)

motivation; 2) HIV/STD knowledge;

3) sexual norms/attitudes; 4) safer
sex self-efficacy; and 5) sexual
communication skills. Modules are
taught within a sexual health
paradigm emphasizing personal
values, positive aspects of sexuality,
and the importance of competent
interpersonal skills, as well as risk
reduction. Inside each module, users
receive age-appropriate audio/video
clips, tips from other teens,
interactive games/quizzes,
infographics, and skill-building
exercises with self-feedback given in
real-time.
Healthy relationship interventions (non-randomized studies n = 4)
Adler-Baeder et al. USA 340 Quasi- Grade 9 to 12 students at nine public 12 × 60 to 90-min lessons of an
(2007) experimental high schools in Alabama adapted version of Love U2:
Increasing Your Relationship Smarts
(RS adapted) Lessons cover: (1)
maturity; values, infatuation and
love; (2) dating processes and
strategies; (3) relationship problems
and identification of unhealthy
relationships; and (4) learning/
practicing relationship and marriage
skills. Theoretical basis:
developmental perspective of
romantic relationship formation
during adolescence.
Avery-Leaf et al. USA 193 Quasi- Grade 9 to 12 students in high school 5 sessions of dating violence
(1997) experimental health classes in a Long Island, New prevention curriculum. Topics
York include: (a) equity in dating
relationships; (b) attitudes towards
violence as a means of conflict
resolution; (c) identification of
constructive communication skills;
(d) identification of support
resources for victims of aggression;
and (e) other: help-seeking, and
alternatives to a violent dating
relationship.
Bradford et al. (2016) USA 1144 Quasi- Participants over 18 years from Premarital Interpersonal Choices and
experimental urban/suburban communities in a Knowledge (PICK) Program. Goals
Western state include: (1) Recognizing
characteristics of a potential partner
and (2) Pacing a relationship
appropriately. Multiple formats
offered, with delivery ranging from
over 1 day up to 6 weeks, but all
were 6 h total in length with
identical content.
Schramm and USA 623 Quasi- High school students in grades 8 to Relationship Smarts Plus curriculum
Gomez-Scott experimental 12 in a Midwestern state. + additional lesson on preventing
(2012) Intervention delivered during family child abuse and neglect. RS+ 13
and consumer science class block lesson curriculum. Teacher
discretion exercised regarding time
period for delivery of 14 modules: 14
weeks or condensed
Healthy/unhealthy relationship interventions (RCT studies n = 15)
Dos Santos et al. Brazil 33 Cluster RCT Participants were 16- or 17-year-old Peer- and bystander approach-based
(2019) public education enrolees and the intervention. Three weekly
majority possessed a partial high intervention sessions of 90 min each
school education. They were on the healthy versus violent
recruited from classes of the project romantic relationships, the quality
“First Step to Work Program,” of friendship in the peer network,
developed by the Socio-Professional and the role of the bystander.
(continued on next page)

201
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 2 (continued )
Author (year) Country N Study type Study population Intervention (duration, components,
theoretical basis)

Education Sector of one educational Session 1 was the two faces of dating
institution of the city of Brasília, and objectives were: discuss the
Brazil. This project promotes the relationship characteristics,
civic, personal, and profession differentiate healthy and unhealthy
education of socially vulnerable relationships; raise awareness about
adolescents. the nature, dynamics, prevalence,
causes, and consequences of dating
violence in health. Contents
included: Modalities of intimate
relationships between friends: one-
night stands, flings, long-term
relationships. Characteristics of
dating relationships: intrinsic
rewards (intimate self-revelation,
care noticed from the partner),
standards of influence and
interaction (time spent with the
partner, sexual intimacy,
perceptions of balance and power),
and problematic characteristics
(jealousy, betrayal, lack of support
to the partner, conflicts). Warning
signs for dating violence. Session 2
was on friendship network and
session 3, Bystander approach.
Foshee et al. (1998) USA 1886 at baseline RCT Grade 8 and 9 students at 14 public 4 months of the Safe Dates
(1700 at 1-month schools, rural eastern North Carolina curriculum. School activities include:
follow-up) (1) theatre production by peers; (2)
10 session curriculum; and (3)
poster contest. Community activities
include: services for adolescents in
abusive relationships (crisis line,
support groups, materials for
parents), and community service
provider training. Primary
prevention (school activities)
through changing norms
surrounding partner violence,
decreasing gender stereotyping, and
improving conflict management
skills. Secondary prevention (school
and community activities) through
changing these variables as well as
changing beliefs about need for help,
awareness of services for victims and
perpetrators, and help-seeking
behaviors.
Foshee et al. (2000) USA 1886 at baseline RCT As above As above
(1603 at 1-year
follow-up)
Foshee et al. (2004) USA 460 (random RCT Grade 8 students at 10 public schools 4-months of the Safe Dates
subsample of in rural North Carolina. Adolescents curriculum. 10 × 45-min health and
original treatment from Foshee et al.’s (1998) original physical education curriculum;
group in Foshee study were included if they poster contest based on curriculum
et al.’s [1998] completed wave 1 baseline and both content. Booster treatment to
study) waves 4 and 6 randomly allocated half of original
treatment group, three years after
originally implemented. 11-page
newsletter containing information
and worksheets based on content
from Safe Dates school curriculum
mailed to adolescents and a personal
telephone call from a health
educator four weeks after mailing.
Health educator answered
questions, provided additional
information if necessary, and
(continued on next page)

202
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 2 (continued )
Author (year) Country N Study type Study population Intervention (duration, components,
theoretical basis)

determined if worksheets completed

and information read.
Foshee et al. (2005) USA 1566 RCT As above; Analysis limited to See original Foshee et al. (1998)
adolescents who completed wave 1 study
baseline questionnaire and who were
in either the original control group,
or the treatment group that received
Safe Dates but not the booster. For
these analyses, the authors used data
from waves 1 (baseline) through to 5
(3-year follow-up).
Foshee et al. (2012) USA 324 RCT American families, including Families for Safe Dates. Family-based
caregiver and 13-15-year-old teen. program for preventing teen
problem behavior, including 6
booklets delivered to families by
mail: first for caregivers only; others
contain interactive activities for
teens and caregivers to complete
together. Topics include: (1)
introduction to program; (2) talking
about dating; (3) skills for handling
conflict; (4) recognizing dating
abuse, preventing dating sexual
abuse and rape; and (5) planning for
the future. Goal is to motivate and
facilitate caregiver engagement in
dating abuse prevention with teens,
moderating risk factors for teen
dating abuse. Theoretical basis:
based on social ecological approach
and uses Protection Motivation
Theory.
Jaycox et al. (2006) USA 2540 RCT Grade 9 health students in Los 3 session program of the Break the
Angeles, California Cycle’s Ending Violence curriculum.
Topics include: (1) Introduction to
domestic violence statistics, myths,
types, warning signs, and obstacles
to help-seeking; (2) Domestic
violence law; and (3) The legal
process, safety planning, and
healthy relationships. Theoretical
basis: Social Learning Theory.
Joppa et al. (2016) USA 225 RCT Grade 10 students with a recent 5 × 50 to 60-min sessions over 1
dating history in a large public week of the Katie Brown Educational
school in Massachusetts Program: a brief, manualized group-
based dating violence prevention
curriculum. Topics include: (1)
understanding violence; (2) wants
and needs in a relationship; (3)
expectations in dating relationships;
(4) communication skills; and (5)
cycles of violence and warning signs.
Activities involved: presentations,
games, discussion, role play.
Lazarevich et al. Mexico 232 RCT Students from the Nutrition Faculty An educational workshop, “Dating
(2017) of a public university in Mexico City. Relationships in College Students,”
Intervention group mean age was was developed to help students
23.2 (SD = 4.8) years; 28 (25.2%) recognize psychological abuse and
were males and 83 (70.3%) were to promote the construction of
females. Control group mean age healthier relationships. The same
was 23.3 (SD = 3.5) years; 2 (21.4%) tutor conducted this 4 h workshop
were males and 80 (78.6%) were with groups of students (14–16). The
females. content of the workshop was based
on the following components:
information on basic concepts,
prevalence of dating violence, its
main consequences, and common
erroneous stereotypes. Afterwards,
(continued on next page)

203
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 2 (continued )
Author (year) Country N Study type Study population Intervention (duration, components,
theoretical basis)

the results of the university survey

on the prevalence of dating violence
were presented. Finally, five case
studies were discussed. The last
exercise was carried out with
students working together in small
groups. The results of each group’s
discussion were shared and
presented in a general session,
conducted by the tutor. Students
played an active role in presenting
and discussing each case, so that the
tutor intervened as little as possible
to allow them to take an active role
in the general session. However, the
tutor was responsible for focusing
the discussion on the most relevant
elements of each case.
Levesque et al. (2016) USA 2605 Cluster RCT Grade 9 to 11 high school students in 3x, 25-30-min sessions of Teen
Rhode Island Choices Intervention delivered at
baseline, one, and two months.
Assessments and guidance
individualized to dating history,
experience of domestic violence, and
stage of readiness to use healthy
relationship skills. Based on
Transtheoretical Model of health
behavior change
Miller, Williams, et al. USA 1062 Cluster RCT Students aged 14–19 years from School Health Centre Healthy
(2015) eleven school health centers in Adolescent Relationships Program
Northern California (SHARP). Healthcare provider-
delivered intervention implemented
at routine school health centre visits.
Healthcare provider (1) conducts
adolescent relationship abuse (ARA)
assessment; (2) introduces brochure
discussing healthy relationships,
how to help a friend, and ARA
resources; and (3) makes a warm
referral to a victim service advocate,
if indicated.
Miller et al. (2020) USA 973 Cluster RCT Conducted with coaches and male Coaching Boys Into Men (CBIM), is a
middle school student athletes social norms theory-based program
recruited from 43 middle schools in developed by a national non-profit
Western Pennsylvania. Ages 11–14 violence prevention organization
years; grades 6–8. School districts (Futures Without Violence). It aims
included urban, suburban, and rural to prevent adolescent relationship
schools with a range of abuse and sexual violence by
socioeconomic status. Three schools training athletic coaches to talk to
were private (2 intervention and 1 male athletes about (1) disrespectful
control). and harmful vs respectful behaviors
among peers (including homophobic
teasing) and in relationships, (2)
myths glorifying male sexual
aggression and promoting more
gender-equitable attitudes and (3)
positive bystander intervention
when athletes witness peers’
aggressive male behaviors toward
girls. CBIM consists of weekly 15-
min coach-led discussions using
program training cards over 12
weeks.
Munoz-Fernandez Spain 1423 RCT Adolescents from state high schools Dat-e Adolescence is a school-based
et al. (2019) in the cities of Seville and Córdoba universal and multi-component
(Andalusia region). Ages ranged dating violence and bullying
from 11 to 19 years (M = 14.98; SD prevention program designed for
adolescents aged 12–19 years. It is
(continued on next page)

204
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 2 (continued )
Author (year) Country N Study type Study population Intervention (duration, components,
theoretical basis)

= 1.39). Around 95% of participants based on the Dynamic

were born in Spain. Developmental Systems Model. It
comprises seven 1-h long sessions
that can be implemented during
school hours. In general terms, the
program adopts a constructivist and
experiential approach that
encourages content learning through
different teaching and learning
experiences. It includes researcher-
and peer-led training. The first five
sessions are administered by
researchers during school hours. The
main aims of these lessons are to (a)
raise awareness of the concepts of
love, myths about romantic love,
and healthy behaviors in
relationships; (b) encourage greater
recognition, expression, and
emotional regulation; and (c)
promote enhanced self-esteem. In
terms of couple dynamics, the
program aims to (d) improve
communication skills; and (e)
promote the recognition of violent
behavior in both traditional and
online forms of violence. These
lessons comprise discussions, role-
playing, debates, and watching
videos to promote socio-emotional
skills and increase knowledge about
romantic relationships and abuse.
Sanchez-Jimenez Spain 1764 Cluster RCT Students from across seven state high Dat-e Adolescence program as
et al. (2018) schools (four from Seville and three described in Munoz-Fernandez et al.
from Cordoba) in Andalucía (2019).
(southern Spain) with medium
economic, social and cultural level.
52.3% were boys with ages ranging
from 11 to 19 years (average age =
14.73; SD = 1.34). Around 96% of
participants were born in Spain.
Wolfe et al. (2009) Canada 1722 Cluster RCT Grade 9 students in School Health 28 h of The Fourth R: Skills for Youth
and Physical Education classes in Relationships. 21 lesson curriculum
Southwestern Ontario delivered in sex-segregated classes
with slightly different activities
delivered to boys and girls.
Individual student-level
intervention comprises 21 × 75-min
lessons covering: (1) personal safety
and injury prevention; (2) healthy
growth and sexuality; and (3)
substance use and abuse. Gender-
strategic approach. School-level
intervention: additional teacher
training on dating violence and
healthy relationships, information
for parents, and student-led "safe
school committees".
Healthy/unhealthy relationship interventions (non-randomized studies n = 5)
Banyard et al. (2019) USA 340 Quasi- Middle school boys from four schools The Reducing Sexism and Violence
experimental in northern New England. The age of Program – Middle School Program
participants ranged from 10 to 15 (RSVP-MSP). Session topics include
years old, Mean = 12.5, SD = 1.0. empathy, healthy relationships, and
The participants were approximately information about gender-based
evenly distributed across 6th, 7th violence including bystander
and 8th grade. intervention training. Leaders
engage participants in active
learning through physical activity,
(continued on next page)

205
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 2 (continued )
Author (year) Country N Study type Study population Intervention (duration, components,
theoretical basis)

peer-to-peer dialogue, storytelling,

role play, multimedia, and group
discussions.
Braithwaite, Lambert, USA 380 Quasi- College students in committed 13 week, 1 × 50 min class/week of
Fincham, & experimental heterosexual romantic relationships Relationship U (RU). Topics include:
Pasley, (2010) (1) conscious relationship decisions;
(2) creating safe relationships; (3)
effective communication; (4)
negative consequences of
extradyadic involvement; and (5)
creating healthy boundaries to
prevent its occurrence.
Carrascosa et al. Spain 191 Quasi- 12–17 years (M = 14.13, SD = 1.05), The Developing Healthy and
(2019) experimental studying Compulsory Secondary Egalitarian Relationships in
Education at two schools (one public Adolescents program (DARSI) aims
and another semiprivate) in the to prevent peer and teen dating
Valencian region. Both secondary violence, by raising adolescents’
schools are located in the same city awareness of the consequences of
and are similar in sociodemographic violence, their critical thinking on
characteristics of the students. sexist attitudes and myths of
romantic love, and their personal
and social resources. This program is
implemented in school contexts and
consists of 12 1-h sessions.
Miller, Williams, et al. USA 1517 Quasi- Middle school (grade 7) students Start Strong: Building Healthy Teen
(2015) experimental Relationships using The Fourth R or
Safe Dates curriculum. Program
elements included: (1)
implementing school-based TDV
prevention curricula; (2) engaging
key influencers (parents/caregivers,
teachers, other mentors); (3) using
social media marketing strategies;
and (4) policy and environmental
change.
Savasuk-Luxton et al. USA 2167 Quasi- Sample of adolescents from an Relationship Smarts Plus (RS+;
(2018) experimental intervention project in the South- Pearson, 2007) curriculum - six
eastern United States focused on weekly core lessons lasting from 50
providing individuals with to 90 min at their respective high
relationship education in order to schools delivered by community
promote healthy relationships. 58% educators from local family resource
girls, 42% boys. The average age was centers. In the curriculum there is
15.66 years (SD = 1.13). one lesson specific to dating violence
and identifying warning signs and
abuse in romantic relationships (i.e.,
Lesson 7: “Dating Violence and
Breaking Up”). Other areas and
activities throughout the curriculum
also address unhealthy and abusive
behaviors (e.g., Lesson 4: “Principles
of Smart Relationships”). Although
the curriculum does not directly
cover gender role beliefs, common,
underlying themes of the lessons
emphasize mutual support and
respect and power sharing (e.g.,
Lesson 3: “Attractions and
Infatuation” and Lesson 4:
“Principles of Smart Relationships”)
in romantic relationships, which
may challenge more traditional
beliefs about power differentials in
romantic relationships.

parents in after-school group sessions; and covered topics including parental rules about adolescent behavior and guidance for
appropriate parental monitoring. The Media Aware (Scull et al., 2018) and HEART for teens program (Widman, Evans, et al., 2020)
both have a strong focus on healthy relationship communication and negotiation skills, as consistent with the core components of the
healthy relationship definition (see ‘Definitions’ section). In addition to these core components, both of these programs are embedded

206
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 3
Outcome measures and key results (n = 30 publications).
Author (year) Control condition Outcome Times of Key results Test statistic/Effect size
measures measurement

Healthy relationship interventions (RCT studies n = 6)

Coyle et al. (2019) Business as usual: 1) Delay sexual Pre-test at No statistically p > .05
instruction on initiation or baseline and two significant
sexually behavioral intent post-test surveys differences between
transmitted (outcome required at 6 and 18 intervention and
infections, by the funder, months control on intention
including HIV, National Institute following to have vaginal sex
often through of Nursing baseline. in the next 3 months
science classes; Research), levels or by the end of 8th
some schools also of sexual initiation grade, or on sexual
discussed (vaginal sex only) initiation or sexual
preventive and touching touching behavior.
measures such as behaviors. Sexual-
condom use. health related
knowledge,
attitudes, beliefs,
self-efficacy,
perceived norms,
parent–child
communication,
sexual limits, and
avoidance of risky
situations.
Of note, however, OR = .49, p = .09
the intervention
group reduced
vaginal sexual
initiation by about
half at 6-month
follow-up, which
approached
significance
Secondary Estimated effect size − 0.10 at 6-month follow up
psychosocial (p = .01). Estimated effect size − 0.03 at 18
outcomes - months (p = .51)
Intervention were
less likely to believe
it is ok for people
their age to have
vaginal sex without
using condoms if the
girl is on birth
control.
None of the p > .05
remaining
psychosocial
variables differed
significantly by
intervention
condition.
Kerpelman et al. Family and 1. Faulty Pre-test at Six of the seven Effect of “Love is enough” = -.370, p < .001;
(2009) Consumer Science relationship baseline models revealed a Effect of “One and only” = -.416, p < .001;
Curriculum as beliefs (one and Post-test at significant Effect of “Cohabitation” = -.509, p < .001;
treatment as only, love is completion of treatment effect in Effect of “Conflict management” = .114, p < .10;
usual enough, program the expected Effect of “Marriage perpetration” = .234, p <
cohabitation), 2. Follow-up 1 and direction suggesting .001;
Perceived ability 2 years later the treatment effects Effect of “Supportive partner” = .384, p < .001.
to manage conflict were consistent with
effectively in close curriculum goals.
relationships, 3. RMSEA <.05 for all
Openness to models, and CFI
future >.99.
relationship
education, 4.
Beliefs about
importance of a
supportive
(continued on next page)

207
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 3 (continued )
Author (year) Control condition Outcome Times of Key results Test statistic/Effect size
measures measurement

partner, 5.
Frequency of
verbal aggression
in relationships
Larson et al. Delayed-control: 1. Relationship T1: 14 days after Across all 5 The planned comparison involving coefficients
(2007) Completed satisfaction, 2. reviewing the relationship +1, +1, − 2 for the RELATE groups and the
RELATE Commitment, 3. RELATE report; measures (1–5 at control group, respectively, yielded 92% of the
programming Opinion about couples in all 3 left), both I groups interaction effect, F(1, 72) = 4.29, MSE = 39.87,
online after study marriage, 4. groups showed p < .05);
was completed Feelings about completed a improvement over
marriage, 5. relationship time while control
Readiness for survey T2: 60 group relatively
marriage, 6. days after unchanged. Therapy
Participant reviewing the assistance provided
satisfaction with RELATE report; a significant benefit
RELATE and two all 3 groups to users of the
interpretation completed RELATE
formats relationship questionnaire
survey compared to the
self-interpretation
and control couples.
The therapy-assisted
group (mean =
119.81) scored
consistently higher
across all five
dependent variables
than either the self-
interpretation
(mean = 113.44) or
control (mean =
114.42) groups.
Lederman (2008) Attention-control 1. Frequency of Student and 1. Frequency of β = -.24, p < .01
group: traditional communication parent communication
didactic teaching about sex and questionnaires about sex and drug
on similar content drug use with administered at: use with parents
but without social parents, 2. baseline, end of decreased in both I
learning exercises Comfort in intervention and C groups across
communication (last/4th the 2 year period
about sex and session), and at 2. Comfort in –
drug use with beginning and communication
parents, 3. end of each of about sex and drug
Parental rules, 4. the 3 booster use with parents: no
Parent sessions (at 4 to change
involvement in 6-month 3. Parental rules: β = .15, p < .05
youths’ activities, intervals) increased for
5. Valuing intervention group,
parents’ opinions, and decreased for
6. Communication control group
with peers about 4. Parent β = -.24, p < .001
sex and drug use, involvement in
7. valuing peer Youths’ activities:
opinions, 8. decreased in both
knowledge about groups
pregnancy 5. Valuing parents’ –
prevention, 8. opinions: no change
self-efficacy for 6. Communication β = .83, p < .001
pregnancy with peers about sex
prevention and and drug use
sex refusal, 9. increased in both
Behavioral groups.
options for 7. valuing peer
resisting pressure opinions: no change
to have sex 8. knowledge about β = -.13, p < .05;
pregnancy and STI β = .13, p < .01
prevention:
compared with
control, I group
(continued on next page)

208
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 3 (continued )
Author (year) Control condition Outcome Times of Key results Test statistic/Effect size
measures measurement

showed increases in
knowledge
Although
systematic, the
gains across time
were small because
the mean score at
pre-test on the
knowledge scale
indicated a 91%
correct response",
9. self-efficacy for –
pregnancy
prevention and sex
refusal: no change
across time for I or C
10. Behavioral β = -.50, p < .05
options for resisting
pressure to have sex:
C group showed
slight decrease over
time in diversity of
resistance responses
for the control
group; I group
remained
approximately the
same.
Scull et al. (2018) Health promotion 1) Primary sexual Pre-test and Significant p < .05
lessons (content health outcomes: post-test. differences in
determined by the intentions to change in favour of
school e.g., engage in the Media Aware were
bullying, following observed for
nutrition/BMI). behaviors: have improving primary
Were asked not to sex, use and secondary
teach sexual/ contraception/ sexual health
relationship protection, outcomes as well as
health or MLE. communicate media-related
with a partner outcomes compared
about sexual to the control group.
health, Sexual Health β(SE) = .19(.07), t = 2.51, Cohen’s d = .17, p =
communicate Primary Outcomes .012
with a parent or —Students in the
another trusted Media Aware group
about sexual had more intentions
health, and to use contraception
communicate and increased β(SE) = .30(.10), t = 3.00, Cohen’s d = .20, p =
with a doctor intentions to .002
about sexual communicate with a
health. Secondary doctor or other
sexual health medical
outcomes professional and
included: No changes were β(SE) = .23(.08), t = 2.76, Cohen’s d = .18, p =
attitudes, seen for intentions .005
normative beliefs, to have sex and
and self-efficacy intentions to
related to teen communicate with a
sexual activity and parents/another
sexual refusal, trusted adult.
contraception/ Sexual Health β(SE) = .19(.07), t = 2.82, Cohen’s d = .18, p =
protection use, Secondary .004
and sexual health Outcomes
communication —Attitudes that
with partners, were impacted by
parents or another the program
trusted adult, and included more
a doctor or positive attitudes
another medical about
(continued on next page)

209
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 3 (continued )
Author (year) Control condition Outcome Times of Key results Test statistic/Effect size
measures measurement

professional. In communicating
addition, the with a romantic
acceptance of partner, parents or
dating violence, another trusted
gender role norms adult
and sexual health Intervention β(SE) = .42(.09), t = 4.64, Cohen’s d = .31, p=
knowledge were students reported <.001
measured. 2) more self-efficacy β (SE) = .21(.07), t = 2.89, Cohen’s d = .19, p =
Media-related for contraception .003
outcomes: media use and
scepticism, communicating
enhanced media with a romantic
deconstructions partner (at post-test
skills, perceived compared with the
realism of sexually control group.
themed media Norms that were β(SE) = -.16(.06), t = -2.65, Cohen’s d = .17, p =
messages, and impacted by the .008
perceived program included a β(SE) = -.17(.07), t = -2.58, Cohen’s d = .17, p =
similarity to reduction in the .009
media messages. acceptance of dating
violence norms and
strict gender role
norms.
Finally, intervention β(SE) = .58(.14), t = 4.18, Cohen’s d = .27, p <
students had more .001
sexual health
knowledge at post-
test compared to
students in the
control group.
Media-Related β(SE) = 1.24(.20), t = 6.17, Cohen’s d = .40, p <
Outcomes: .0001
Participants’ media β(SE) = .17(.09), t = 2.03, Cohen’s d = .13, p =
deconstructions .042
skills and media
scepticism increased
as a result of the
program.
No changes were –
observed for
perceived realism of
sexually themed
media messages and
perceived similarity
to media messages.
Widman, Evans, Growing Minds - 1) Pre-test (at At post-test, effect size Cohen d range = 0.23 to 1.27
et al. (2020) an attention- Communication baseline) and students who
matched online Intentions, immediate post- completed HEART
control program Condom test. demonstrated
focused on Intentions, HIV/ improvements on
cultivating STD Knowledge, every outcome
academic and Condom examined:
social growth Attitudes, Sexual p < .001, ES 0.57
mindsets. Condom Norms, communication
Self-Efficacy, intentions
Sexual Condom use p = .020, ES 0.24
Assertiveness. intentions
HIV/STD p < .001, ES 1.27
knowledge
Condom attitudes p < .001, ES 0.55
Condom norms p = .001, ES 0.41
Self-efficacy to p = .031, ES 0.23
practice safer sex
Sexual assertiveness p = .023, ES 0.29
Girls who completed from pre-test (M = 3.28, SD = 1.39) to post-test
HEART showed (M = 3.88, SD = 1.08) than boys (pre-test
greater M = 3.40, SD = 1.07; post-test M = 3.49, SD =
0.93; b = − .65, SE = 0.28, p = .021).
(continued on next page)

210
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 3 (continued )
Author (year) Control condition Outcome Times of Key results Test statistic/Effect size
measures measurement

improvement in
condom norms
There were no (p range = .14-.75). p values ranged between
differences in sexual 0.14 and .75
communication or
condom use
intentions, HIV/STD
knowledge, sexual
attitudes, self-
efficacy, or sexual
assertiveness by
gender
There were no (p range = .27-.84). p values ranged between
statistically 0.27 and .84
significant
differences in any
program outcome
based on sexual
orientation
Healthy relationship interventions (non-randomized studies n = 4)
Adler-Baeder et al. Treatment as 1. Relationship Pre-test at Relationship F = 5.22, p = .02
(2007) usual: Family and knowledge (pre baseline knowledge:
Consumer Science and post-test only) Post-test 2 excluded as no
classes, including 2. Conflict months later control group
topics of family behaviors Conflict behaviors:
and human 3. Relationship 1 (verbal
dynamics, family beliefs aggression) of 3
wellness, and life subscales
connections (reasoning, verbal
aggression, physical
aggression)
demonstrated
statistically
significant
improvement.
Relationship beliefs: F=4.71, p = .03
1 (realistic
relationship beliefs)
of 3 (aggression
beliefs, faulty
relationship beliefs,
realistic relationship
beliefs) subscales
showed statistically
significant
improvement.
Avery-Leaf et al. Treatment as Dating violence Pre-test at Dating violence t (88) = 1.68, p < .09
(1997) usual: health class attitudes baseline attitudes:
Justification of Post-test 2 statistically
dating jealousy weeks later significant
and violence improvement in
attitudes justifying
male to female
aggression; no
group x time
interaction for
attitudes justifying
female to male
aggression
Justification of
dating jealousy and
violence: no
significant group x
time interaction
Bradford et al. Non-equivalent Perceived Pre-test at Perceived t = 18.88, p < .001
(2016) comparison group knowledge about: baseline knowledge about
(university 1. Relationship Post-test at relationship skills:
students) skills completion of the comparison
(continued on next page)

211
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 3 (continued )
Author (year) Control condition Outcome Times of Key results Test statistic/Effect size
measures measurement

2. Partner intervention (2 group x treatment x

selection weeks later for time variable
control group) differed
Perceived significantly,
importance of indicating
knowledge about: significant
3. Potential improvement for
partner’s treatment group
relationship versus comparison
patterns Perceived t = 17.31, p < .001
4. Potential knowledge about
partner’s partner selection:
relationship the comparison
behavior and group x treatment x
attitudes time variable
differed
significantly,
indicating
improvement in the
treatment group
versus comparison
Perceived t = 12.90, p < .001
importance of
knowledge about a
potential partner’s
relationship
patterns: the
comparison group x
treatment x time
variable differed
significantly,
indicating greater
improvement in
treatment group
over time compared
with comparison
group.
Perceived t = 7.31, p < .001
importance of
knowledge about a
potential partner’s
relationship
behavior and
attitudes: the
comparison group x
treatment x time
variable differed
significantly
suggesting
increased
improvement in
treatment group
versus comparison
Schramm and Treatment as 1. Relationship Pre-test at 1. Relationship ΔMean = -.77, − .95, − .82 t = -25.51, − 29.57,
Gomez-Scott usual: standard knowledge, 2. baseline knowledge: − 21.71 p < .001
(2012) family and Attitudes about Post-test at Statistically Cohen’s d = − 1.42, − 1.70, − 1.22
consumer science romance and mate conclusion of significant increase
curriculum selection, 3. intervention in perceived
Attitudes toward relationship
premarital and readiness, healthy
marital relationship
counseling, 4. knowledge, and
Marriage child abuse
attitudes, 5. knowledge
sexual attitudes, 2. Attitudes about F(1, 620) = 8.26, p = .004.
6. Resisting sexual romance and mate
pressure, 7. Verbal selection Significant
and physical time X group
(continued on next page)

212
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 3 (continued )
Author (year) Control condition Outcome Times of Key results Test statistic/Effect size
measures measurement

aggression during interaction for all 3

conflict, 8. Harsh subscales;
caregiving intervention
response scale, 9. students reported a
Spanking beliefs, significant decrease
10. Sudden infant in beliefs that there
death syndrome is only one person
SIDS knowledge, meant for them
(non-significant
change in control
group over time)
Similarly, the RS+ F(1, 620) = 4.12, p = .04.
students reported a
significant decrease
in their beliefs that
love is enough to
sustain a happy
marriage, whereas
the control group
did not experience a
significant change
over time
Finally, the RS+ F(1, 620) = 46.43, p < .001.
students
experienced a
significant decrease
in their belief that
cohabitation is a
wise way to test
whether a
relationship will
work, whereas the
control group of
students did not
experience a change
over time
3. Marriage F(1, 619) = .37, p = .54,
attitudes no F(1, 620) = 2.49, p = .11
significant
interactions for the
attitudes toward
counseling and
marital enrichment
scale or the
marriage attitudes
scale
4. Sexual attitudes –
and resisting sexual
pressure
Intervention group F(1, 619) = 11.73, p = .001
demonstrated
increased
understanding
about the risks of
sex and an increased
willingness to wait
until there is more
of an emotional
connection before
having sex, from
pre-test to post test,
compared with the
control group
The RS+ students F(1, 619) = 7.99, p = .005.
also reported a
significant increase
in their perceived
ability to resist
(continued on next page)

213
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 3 (continued )
Author (year) Control condition Outcome Times of Key results Test statistic/Effect size
measures measurement

sexual pressure,
whereas the control
group declined
somewhat
5. Aggression –
Students who were F(1, 407) = 6.97, p = .037.
currently involved
in a dating
relationship
reported a
significant decrease
in their use of verbal
aggression toward
their dating partner
across time,
whereas the control
group of students
did not experience a
significant decrease
in verbal aggression
toward their dating
partner
Healthy/unhealthy relationship interventions (RCT studies n = 15)
Foshee et al. Community-level 1. Psychological Pre-test at Outcome variables
(1998) intervention abuse baseline (full sample):
activities only (vs. victimization Post-test 1 Treatment group
school and 2. Non-sexual month after associated with
community physical violence completion of significant changes
activities for victimization intervention in:
intervention 3. Sexual violence Psychological abuse b = − .08, p = .001
group) victimization perpetration
4. Physical Sexual violence b = − .06 p = .009
violence in perpetration
current Violence b = − .06, p = .014
relationship: perpetrated in
victimization and current relationship
perpetration Mediating variables
Mediating (full sample):
variables: Treatment group –
1. Acceptance of was:
norms justifying Less supportive of
dating violence prescribed dating
2. Conflict norms
management skills More supportive of
3. Gender proscribed dating
stereotyping norms
4. Belief in need Perceived fewer
for help positive
consequences from
using dating
violence
Used more
constructive
communication
skills and responses
to anger. Were less
likely to engage in
gender stereotyping.
Were more aware of
victim perpetrator
services
Foshee et al. Community-level As above Pre-test at Treatment group:
(2000) intervention baseline Less accepting of p = .05
activities only (vs. Post-test 1 dating violence
school and month after Perceived more p = .05
community intervention end negative
activities for consequences of
(continued on next page)

214
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 3 (continued )
Author (year) Control condition Outcome Times of Key results Test statistic/Effect size
measures measurement

intervention engaging in dating

group) violence
Reported using less p = .08
destructive
responses to anger
Were more aware of p = .02
victim
Perpetrator services p = .02
Primary prevention –
subsample: number
of significant
differences were
found in any of the
mediating variables
between treatment
and control groups.
Victim subsample:
treatment group:
less accepting of p = .03
dating violence
Perceived more p = .02
negative
consequences from
engaging in dating
violence
Reported using less p = .08
destructive
responses to anger
Reported less p = .08
gender stereotyping
Were more aware of p = .05
victim services
Perpetrator
subsample:
treatment group:
Reported using less p = .02
destructive
responses to anger
More aware of p = .06
perpetrator services
Foshee et al. Community-level As above Baseline Oct Safe dates effects on β (SD) = − .23(.08), p = .01
(2004) intervention 1994 (n = 957, perpetration: (only
activities only (vs. wave 1), 1 safe dates) no
school and month (wave 2), statistically
community 1 year (wave 3), significant
activities for 2 years (n = 620, difference in
intervention wave 4) new psychological abuse
group) baseline, 3 years perpetration.
(wave 5), 4 years Safe dates effects on
(wave 6) victimization:
significant main p < .05
effect on sexual
victimization
Booster effects on
perpetration:
booster did not β(SD) = .05(.05), p = .26
improve the
effectiveness of Safe
Dates in preventing
sexual DV
perpetration
involvement in β(SD) = .34(.12), p = .003
psychological abuse
perpetration
moderated the effect
of the booster on
psychological abuse
perpetration
(continued on next page)

215
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 3 (continued )
Author (year) Control condition Outcome Times of Key results Test statistic/Effect size
measures measurement

adolescents high in p = .03

prior psychological
abuse perp who
were exposed to
booster reported
significantly more
psychological abuse
perp at follow-up
than those exposed
only to safe dates
Comparisons
between control and
booster group:
no significant β(SD) = .05(.05), p = .28
difference in sexual
DV perpetration (p
= .28),
no significant β(SD) = .70(.46), p = .38
difference in
psychological abuse
perpetration.
Booster effects on β(SD) = .21(.14), p = .16
victimization
(booster vs. safe
dates only):
no effects of booster β(SD) = .68(.91), p = .46
of psychological
abuse victimization
sexual victimization β(SD) = .50(0.11), p < .0001
moderated by prior
wave 4
victimization, such
that adolescents
reporting high prior
victimization and
exposed to booster
reported
significantly more
sexual victimization
at follow up than
adolescents
receiving only safe
dates.
Booster vs. control
(victimization):
no significant p = .70
differences between
booster and control
group in follow-up
psychological abuse
victimization
significantly less p = .03
sexual victimization
where there was no
prior sexual
victimization
Foshee et al. Community-level As above Baseline data Significant main b(95% CI) = -0.95(-1.48 to − .41), p = .0005; b
(2005) intervention (wave 1) were effects of treatment (95% CI) = -.36(-.66 – (− ).06); b(95% CI) = -.05
activities only (vs. collected in condition on (-.11-.00)p = .02, p = .04.
school and schools by psychological abuse
community trained research perpetration,
activities for staff in October moderate physical
intervention 1994 violence
group) Follow-up data perpetration, and
were collected sexual violence
from treatment perpetration.
and control That is, adolescents b(95% CI) = − .16(.30 to − .03), p = .01
adolescents 1 receiving Safe Dates
(continued on next page)

216
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 3 (continued )
Author (year) Control condition Outcome Times of Key results Test statistic/Effect size
measures measurement

month (wave 2), perpetrating less

1 year violence of all types
(wave 3), 2 years at all four follow-up
(wave 4), 3 years periods than
(wave 5), and 4 controls. 2.
years (wave 6) Significant main
after Safe Dates effects of treatment
was completed condition on
using the same moderate physical
procedures as violence
for baseline data victimization.
collection That is, adolescents b(95% CI) = -0.06(-.13 to − .00), p = .07;
who received Safe b(95% CI) = -.48(-1.16 to − .20, p = .17;
Dates reported less b(95% CI) = -.19(-.44–
victimization of .07), p = .14.
moderate physical
dating violence at
all four follow-up
waves than control.
3. Marginal program
effect on sexual
victimization (p =
.07) in expected
direction. 4.
Treatment condition
not associated with
psychological abuse
victimization (p =
.17), or severe
physical
victimization (p =
.14) at any of the
four follow ups.
Effects of mediators: b(95% CI) = -.71(-.99–(− ).42), p < .0001;
adolescents b(95% CI) = -.93(-1.29– (− ).56), p < .0001;
receiving Safe Dates b(95% CI) = -.93(-1.29– (− ).56), p < .0001;
reported less b(95% CI) = .13(.02–.24), p = .02;
acceptance of
prescribed dating
violence norms
(significant main
effect of treatment
condition, less
acceptance of
traditional gender-
role norms
(significant main
effect of treatment
condition, and
greater belief in
need for help
(significant main
effect of treatment
condition at all 4
follow-up periods
than controls.
Treatment and b(95% CI) = .07(.04-.09), p < .0001;
control differences b(95% CI) = .35(.05-.76), p = .09.
regarding
awareness of
community services
were greatest at
wave 2 and least at
wave 4 (significant
interaction between
treatment and time
squared. No effects
of treatment on
(continued on next page)

217
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 3 (continued )
Author (year) Control condition Outcome Times of Key results Test statistic/Effect size
measures measurement

conflict resolution
skills at any of the
follow-up waves (p
= .09)
Jaycox et al. Standard health All self-report Followed for 6 Intervention group
(2006) curriculum with measures months: pre-test vs control:
time-delayed 1. Knowledge on a Monday, Knew significantly Effect size (95% CI) = 0.73(.63–.82), F value =
control (received about dating intervention on more about laws 284.66, p < .001;
Ending Violence) violence (11 T/F Tuesday, Wed, related to dating Effect size (95% CI) = -.23(-.31-(− ).14), F value
when trial was statements) Thurs; Follow- violence, were less = 32.29, p < .001;
complete 2. Two scales used up surveys accepting of female- Effect size (95% CI) = .27(.19-.36), F value =
to capture help- administered in on-male violence, 45.78, p < .001;
seeking: class 6 months and showed higher
helpfulness, and later. likelihood of
likelihood of seeking help for
talking to each “if violence at post-test
you experienced relative to controls;
violence with a All students
date” reported a lower
3. Negative dating perception of
experiences in the helpfulness of
prior six months assistance sources
Revised Conflict on the post-test
Tactics Scale survey but the mean
decrease was
significantly smaller
for intervention
students than
control students.
Acceptance of male- Effect size(Victimization) (95% CI) = .10
on-female violence (-.10–.30), F value = 1.08;
in response to Effect size(Perpetration) (95% CI) = .06
provocation and (-.13–.25), F value = .49;
recent abusive/
fearful dating
experiences did not
differ significantly
across the
intervention and
control groups. 6-
month data on
dating violence
victimization and
perpetration
revealed no
significant
differences between
the two groups at
follow-up.
Of the help seeking –
subscales, 5 of 9
potential sources of
help were perceived
as more helpful
post-intervention in
the intervention
group, and 1 of 9
was reported as
more likely to be
consulted for help
post intervention in
the intervention
group.
Improvement in
knowledge persisted
at 6 months, as did
perceived
helpfulness of
(continued on next page)

218
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 3 (continued )
Author (year) Control condition Outcome Times of Key results Test statistic/Effect size
measures measurement

speaking with a
lawyer re: dating
violence.
Wolfe et al. (2009) Treatment as Primary: Pre-test at Main effects for 114 of 168 (67.9%) vs 65 of 111 (58.6%) (p <
usual: Grade 9 Perpetration of baseline secondary outcomes .01)
Health and physical dating Post-test 2.5 were not
Physical violence 2.5 years years later statistically
Education after baseline significant;
Ontario Secondary: 1. Peer however, sex by
violence, 2. group analyses
Substance use, 3. showed a significant
Unsafe sex difference in
(consistent use of condom use in
condoms during sexually active boys
sexual who received the
intercourse) intervention vs
controls.
Foshee et al. Randomly Primary (1): Pre-test at 1. Not relevant, 2. Coef (SE) = .84 (.27), Cohen’s d = .37, p < .01
(2012) allocated to either Factors related to baseline Targeted risk
a full or a partial motivating and Post-test at 3 factors: 3.
treatment group facilitating months Treatment condition –
and the other to a caregiver was significantly
control group. engagement in associated with teen
Those in the teen dating abuse acceptance of dating
control group prevention abuse in the
received no activities. expected direction.
program Targeted risk There were no Coef (SE) = − .50 (.33), p = .12;
materials. factors (2) (teen treatment effects on Coef (SE) = .47 (.34), p = .17;
acceptance of teens’ perceived
dating abuse, teen negative
perceived consequences of
negative dating abuse (p =
consequences of .12), teen conflict
dating abuse, teen resolution skills (p
skills in resolving = .17), or caregiver
conflict, caregiver date rule setting and
date rule setting & monitoring.
monitoring). Trends in treatment OR(95%) = .26 (.07–.94), p = .04
Secondary (3). effects on
Abuse psychological and
perpetration: physical dating
psychological and abuse victimization
physical, 4. Abuse and perpetration
victimization: were in the expected
psychological and directions, but only
physical one effect was
statistically
significant - that was
the effect of the
program on the
onset of physical
dating abuse
victimization (p =
.04).
Miller, Williams, Care as usual 1. recognition of Baseline and 3- Intervention versus (adjusted mean difference = 0.10 [0.01 to 0.18]).
et al. (2015) when attending abuse, 2. month post control adjusted
school health intentions to intervention mean differences
centre: standard intervene, 3. (95% confidence
social hx knowledge of interval) on changes
assessment resources in primary
without specific outcomes were not
prompts for ARA. statistically
significant.
Intervention
participants had
improved
recognition of
sexual coercion
(continued on next page)

219
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 3 (continued )
Author (year) Control condition Outcome Times of Key results Test statistic/Effect size
measures measurement

compared with
controls
Among participants (mean risk difference = 20.17 [20.21 to 20.12]).
reporting
relationship abuse
at baseline,
intervention
participants were
less likely to report
such abuse at
follow-up
Adolescents in (adjusted odds ratio = 2.77 [1.29 to 5.95]).
intervention clinics
who reported ever
being in an
unhealthy
relationship were
more likely to report
disclosing this
during the SHC visit
Joppa et al. (2016) Treatment as 1. Conflict in Baseline, end of Treatment vs
usual (health adolescent program (10 control at
class curriculum, relationships, 2. days later), 3 completion:
according to state Normative beliefs months post Significantly lower OR (95% CI) = -.20(-.33 –.07), X2 = 9.60, p < .01;
Comprehensive about aggression, intervention approval of
Health 3. Attitudes aggression,
Curriculum toward dating healthier dating
Framework violence, 4. DV attitudes, and more
knowledge and dating violence
healthy knowledge; effects
relationship sustained at 3-
attitudes (Katie month follow-up.
Brown specific Additionally, at 3- OR(95%CI)(Emotional/verbal DV perp) = -.27
questionnaire) month follow-up: (.12-.62), X2 = 9.68, p < .01;
treatment vs. OR(95% CI)(Total perp) = -.31(.13-.71), X2 =
control showed 7.66, p < .01;
significantly less OR(95% CI)(Emotional/verbal DV victim) = -.26
emotional/verbal (.1-.41), X2 = 32.40, p < .001;
DV and total DV OR(95% CI)(Total victim) = -.25(.15-.41), X2 =
perpetration and 30.13, p < .001;
victimization.
Levesque et al. Completion of Primary: Baseline, 6 Outcomes reported with unadjusted odd ratios ranging from .50 to
(2016) alternative online Emotional and months, 12 for the subsample of .72 at 6 months, and from .50 to .70 at 12 months
transtheoretical- physical dating months youth exposed to at
based violence least minimal risk
intervention, victimization and for dating violence:
Health in Motion, perpetration significantly
Secondary: reduced odds of all
Consistent use of four types of dating
healthy violence i.e.
relationship skills emotional
and rejection of victimization,
attitudes emotional
supporting dating perpetration,
violence physical
victimization, and
physical
perpetration
While rates of –
dating violence
increased in both
group from the 6–12
month follow up,
the odds ratios
remained stable,
suggesting stability
of the intervention
effect.
(continued on next page)

220
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 3 (continued )
Author (year) Control condition Outcome Times of Key results Test statistic/Effect size
measures measurement

The intervention OR = 1.68–2.60

was associated with
higher odds of
consistently using
healthy
relationships skills
at 6 and 12 months
intervention OR = 1.28; 1.02–1.60;
associated with OR = 1.13; 0.95–1.35.
significantly higher
odds of rejecting
attitudes supporting
dating violence at 6
months follow up
but not significant at
12 months follow-
up
Lazarevich et al. No workshop Perception and Pre-test at Intervention group
(2017) attitudes toward baseline and An improvement of p < .05. The mean score in pre-test was 13.27 (SD
dating violence. post-test the perception and = 2.27) whereas in post-test was 16.16 (SD =
attitudes related to 2.63)
dating violence was
observed. From 18
items of the
questionnaire, 13 of
them (72.2%) were
significantly
different between
the pre- and post-
evaluations;
Applying the p < .0001
matched pair test, a
significant
difference between
the two scores was
detected.
The most significant from 13.40 to 15.52 and from 12.86 to 15.10,
improvement was respectively (p = .045).
observed in
questions that
related violence
with educational
level and
employment and in
those that related
violence with
mental health
problems, such as
depression, low self-
esteem, and
negative personality
characteristics.
Considering the pre-
and post-violence
scores, women
showed a higher
improvement from
the first to the
second score than
men did
The intervention –
appeared to differ
by certain
sociodemographic
factors e.g. gender,
mother’s education
level; but not
(continued on next page)

221
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 3 (continued )
Author (year) Control condition Outcome Times of Key results Test statistic/Effect size
measures measurement

father’s educational
level nor age.
Control group
Only one item was p = .0482, The mean score in pre-test was 13.95
different between (SD = 1.26) whereas in post-test was 14.16 (SD =
the pre- and post- 0.94)
evaluations:
Violence is an
exclusively socio-
cultural
phenomenon. No
other differences
were observed.
Applying the p = .979
matched pair test,
no significant
difference between
the two scores was
detected
Sanchez-Jimenez No intervention. 1) Myths of Pre-test (at The program had a The effect sizes were large for better half myth
et al. (2018) School activities romantic love baseline) and significant impact (Cohen’s d = − .83), omnipotence (d = − .84),
as usual (jealousy, better post-test (2 on change across all passion, (d = − .94) and medium for jealousy (d
half, weeks after myths of romantic = − .56).
omnipotence, intervention and love, reporting a
passion); quality six months after reduction in the
of romantic baseline). acceptance of myths
relationship among those
(negative participants who
interaction scale, took the program
Intimacy f2f, compared with the
intimacy online) control group
and psychological participants.
variables (anger Program efficacy on
regulation and couple quality - a
self-esteem). 2) significant program
dating aggression effect on change was
and victimization not observed for
(psychological negative
aggression, interactions scales
psychological nor for both
victimization, intimacy measures.
physical Program efficacy on Cohen’s d = − .19.
aggression, anger regulation
physical and self-esteem -
victimization, those participants
cyber-aggression, who received the
cyber- intervention
victimization) increased their
confidence in the
self-regulation of
their own anger in
comparison to the
control group,
although the effect
size was small
In relation to self- Cohen’s d = − .15.
esteem, the efficacy
results showed that
the program had a
significant impact
on self-deprecation,
but not for self-
confidence. For self-
deprecation, the
experimental group
participants reduced
their scores
compared with the
(continued on next page)

222
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 3 (continued )
Author (year) Control condition Outcome Times of Key results Test statistic/Effect size
measures measurement

change observed in
the control group
participants, but the
effect size was small
Dos Santos et al. No intervention. 1) Intent to offer Pre-test (at There was not a p > .05
(2019) help at the level of baseline) and significant change
the individual in post-test (two in intention to help,
response to dating and half months bystander attitudes,
violence. 2) after the or empathy from
Empathy and intervention). pre-test to post-test
bystander between groups.
attitudes in A similar percentage
response to dating of both EG (86%)
violence. and CG (84%)
participants
reported an
intention to help in
dating violence
situations, a non-
significant
difference
Similarly, no p > .05
significant
differences were
observed between
intervention and
control groups for
empathy or
bystander attitudes
The hypothesis that –
the participants in
the intervention
group would intend
to help more
frequently than
participants in the
control group and
have more empathy
and more
sympathetic
bystander attitudes
in response to
dating violence at
two and half months
after the
intervention was
refuted.
Munoz-Fernandez No intervention - 1) Moderate Pre-test (at Correlations among
et al. (2019) the waiting list physical dating baseline) and outcome variables:
procedure was violence, severe post-test (six Dating violence and Cohen’s d = .05–.46
applied to the physical dating months after bullying were (.20;.18;.22;.16;.13;.05;.15;.10;.14;.05;.18;.46)
control schools violence and baseline) positively p < .001 and p < .05
that expressed sexual dating correlated. For each
interest in violence 2) form of violence,
receiving Bullying measures of
intervention in aggression and
future editions. victimization
showed a strong
correlation.
The effect size was Cohen’s d = .20; .22;.18; p < .001
large for the
relationship
between moderate
physical dating
aggression and
victimization, for
severe physical
dating aggression
(continued on next page)

223
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 3 (continued )
Author (year) Control condition Outcome Times of Key results Test statistic/Effect size
measures measurement

and victimization,
and for sexual
dating aggression
and victimization.
The effect size was Cohen’s d = .46; p < .001
medium for the
relationship
between bullying
aggression and
victimization.
Students Cohen’s d = .21 for severe physical dating
participating in the victimization; 0.25 for severe physical dating
Dat-e Adolescence aggression; 0.24 for sexual dating victimization;
program, exhibited 0.38 for sexual dating aggression; and 0.98 for
lower levels of bullying victimization.
severe physical
dating violence
(aggression and
victimization);
sexual dating
violence (aggression
and victimization);
and bullying
victimization six
months after the
intervention.
Miller et al. (2020) Standard 1) Positive Pre-test at Positive bystander relative risk, 1.51; 95% CI, 1.06–2.16 and 1.53;
coaching bystander baseline (time 1) behaviors increased 95% CI, 1.10–2.12, respectively
behaviors (i.e. and post-test at at end of sports
intervening in end of season season and at 1-year
peers’ (time 2) and 1 follow-up
disrespectful or year later (time At the end of the mean risk difference, 0.14; 95% CI, 0.01–0.27
harmful 3) sports season and at and 0.14; 95% CI, 0.00–0.28, respectively.
behaviors) 2) the 1-year follow-
Recognition of up, athletes in
what constitutes schools receiving
abusive behavior, CBIM reported
intentions to greater recognition
intervene, gender- of abusive behaviors
equitable than control
attitudes, and schools.
recent ARA/SV At 1-year follow-up, OR = 0.24 (95% CI, 0.09–0.65).
perpetration among those who
ever dated, athletes
on teams receiving
CBIM had lower
odds of reporting
recent ARA/SV
perpetration
Gender attitudes –
and intentions to
intervene did not
differ between study
arms.
In exploratory –
intensity-adjusted
and per protocol
analyses, athletes on
teams receiving
CBIM were more
likely to report
positive bystander
behaviors and to
endorse equitable
gender attitudes and
less likely to report
ARA and sexual
harassment
(continued on next page)

224
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 3 (continued )
Author (year) Control condition Outcome Times of Key results Test statistic/Effect size
measures measurement

perpetration 1 year
later.
Healthy/unhealthy relationship interventions (non-randomized studies n = 5)
Braithwaite Treatment as Sexual and Pre-test at 52% reduction in Hedge’s G = .06
Lambert, usual: university- romantic baseline extradyadic kissing
Fincham, & level course behaviors with Post-test at 6 (no change in
Pasley (2010) Introduction to someone other weeks and 12 control group).
Families Across the than an weeks 50% reduction in Hedge’s G = .04
Lifespan, which individual’s sexual intimacy
provides an committed without intercourse
overview of romantic partner (also 50% in control
research of in the previous 6 group).
marriage and weeks: 58% reduction in Hedge’s G = .05
families 1. Kissing extradyadic sexual
2. Sexual intimacy intercourse (33% in
without control group).
intercourse
3. Sexual
intercourse
Miller, Williams, Start Strong 1. TDV Fall and spring SHORT TERM:
et al. (2015) schools were perpetration and of grade 7; fall Parent-child interaction b = .10 [se = .04], p < .05.
compared to victimization and spring of communication
Comparison (including grade 8 decreased
schools. psychological and significantly in the
Comparison electronic) 2. comparison group,
schools were Attitudes towards whereas the Start
matched to gender Strong group
intervention stereotypes, 3. showed no such
schools Acceptance of DV, decline at wave 2
4. Perceived Start Strong group- by-time interaction b = .18 [SE = .09], p
negative students also < .05;
consequences of reported a and b = .19 [SE = .09], p < .05, respectively.
DV, 5. Reponses to statistically
anger, 6. Use of significant increase
positive in relationship
communication satisfaction and
skills, 7. Parent- support from waves
child 1 to 2, whereas
communication comparison
about students’ reports did
relationships, 8. not significantly
Boy/girlfriend change over time
relationship Effects for the –
quality dominance and
criticism scales were
not statistically
significant. Students
in the Start Strong
schools also
reported decreased
acceptance of TDV
from waves 1 to 2.
In contrast, change interaction b = − .15 [SE = .05], p < .01.
over time for
students in
comparison schools
was not statistically
significant
Similarly, negative b = − .08 [SE = .02], p < .01
gender stereotypes
decreased to a
significantly greater
extent among Start
Strong students than
comparison
students.
LONG TERM:
(continued on next page)

225
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 3 (continued )
Author (year) Control condition Outcome Times of Key results Test statistic/Effect size
measures measurement

Program effects b = -.08 [SE = .03], p < .05;

persisted for b = − .16 [SE = .03], p < .01
acceptance of TDV;
Although other
constructs showed
statistically
significant change,
their rates of change
did not differ
significantly
between
intervention and
comparison
conditions. Overall
levels of gender
stereotypes were
significantly lower
in the program
group. Acceptance
was also lower in
the intervention
group
Savasuk-Luxton No intervention 1) Gender role Pre-test and There was no Wilks’ Lambda = 1.00, F (1, 2048) = 0.69, p =
et al. (2018) beliefs and dating post-test statistically .41, ɳ2 = 0.000.
violence evaluations. significant
acceptance. (2.1) difference in change
Does the amount patterns in gender
of change in role beliefs between
traditional gender intervention and
role beliefs and in control adolescents
dating violence Additionally, there F (1, 2123) = 0.35, p = .55
acceptance were no significant F (1, 2086) = 0.10, p = .75
following RE differences between
participation groups in reported
differ by gender, gender role beliefs
race, and SES, at pre- or post-
independently? assessment
(2.2) Does the Although the mean Wilks’ Lambda = 0.994, F (1, 2063) = 13.37, p <
amount of change score for dating .001, ɳ2 = 0.006
in traditional violence acceptance
gender role beliefs was relatively low at
and in dating both time points and
violence there was limited
acceptance differ variability in
by the two-way responses, there was
interaction of a statistically
these identities (i. significant
e., gender x race, difference in change
gender x SES, and patterns in dating
race x SES) for violence acceptance
adolescents who between
receive RE? and intervention and
(2.3) Does the control adolescents
amount of change More specifically, t (1580) = 1.43, p = .15
in traditional intervention
gender role beliefs participants
and dating reported no
violence significant change
acceptance differ in dating violence
by the intersection acceptance.
of all three Whereas t (496) = − 2.86, p < .01
identities (i.e., adolescents in the
gender x race x control group
SES) for reported a
adolescents who significant increase
receive RE? in acceptance of
dating violence
(F (1, 2099) = 18.72, p < .001).
(continued on next page)

226
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 3 (continued )
Author (year) Control condition Outcome Times of Key results Test statistic/Effect size
measures measurement

significant
differences between
the two groups at
post-test such that
adolescents in the
intervention group
reported less dating
violence acceptance
(Mcomp = 1.81
(1.09) and Mpart =
1.59 (0.96)) than
adolescents in the
control group.
Banyard et al. Students at 1) Attitudes Pre-test at Boys who received ΔMean(SD) = − 0.08(0.41) p < .05
(2019) another middle reflecting male baseline and the RSVP-MSP
school received social dominance post-test. The demonstrated
no programming and support for time from significant
(they were an coercion in baseline to decreases over time
intent to treat relationships 2) follow-up varied in support for the
group). Emotional due to weather use of physical force
awareness, created school and violence in
attitudes in cancellations relationships.
support of gender and scheduling Contrary to ΔMean(SD) = 0.04(0.69)
equity, and issues with hypotheses, boys
intentions to take schools (from 56 showed increases in
action against days in the emotion
harassment. control, and 72, dysregulation, that
84, or 125 days is, feeling out of
in the treatment control about their
schools). negative feelings
(although only
marginally
significant at the
0.10 level).
Boys in both the ΔMean(SD) = − 0.08(0.39)
control and ΔMean(SD) = − 0.15(0.44)
treatment condition ΔMean(SD) = 0.02(0.41)
showed decreases in ΔMean(SD) = 0.20(0.58) p = .13
a measure of
support for male
power and increases
in support for
gender equity in
relationships.
There were no –
significant changes
for boys regardless
of condition on
social norms
supporting violence
prevention, intent to
take action, apathy,
or masculinity
stress.
Carrascosa et al. Two contrast Hostile sexism, Pre-test (at There was a F(2,174) = 16.19, p < .001, η2 = .157
(2019) groups were benevolent baseline) and significant
established: three sexism, beliefs in post-test (1 interaction effect in
other classrooms myths of romantic month after benevolent sexism,
(2nd, 3rd and 4th love, and peer intervention with a lower mean
of Secondary aggression (overt, completion). score after the
Education) in the relational, and intervention in this
same public online). variable in EG (M =
school (n = 62) 12.76), compared to
constituted CG1 (M = 26.53)
control group 1 and CG2 (M =
(CG1) and three 29.89).
classrooms (2nd, The interaction F(2,174) = 25.67, p < .001, η2 = .228
3rd and 4th of effect is also
(continued on next page)

227
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Table 3 (continued )
Author (year) Control condition Outcome Times of Key results Test statistic/Effect size
measures measurement

Secondary significant in hostile

Education) of the sexism, mean score
semi-private in post-test in EG (M
school (n = 72) = 13.03), compared
were assigned to to CG1 (M = 25.37)
control group 2 and CG2 (M =
(CG2). Although 27.20).
control group 1 In myths of F(2,174) = 11.136, p < .001,
(CG1) did not romantic love, there η2 = .113
implement the is a significant
program, it did interaction effect,
share some with a lower with a
classes and lower mean score in
facilities with the this variable in EG
experimental adolescents (M =
group, whereas 7.98) compared to
control group 2 the other two
(CG2) had no groups (CG1: M =
contact with the 12.24; CG2: M =
experimental 14.26).
group. In peer aggressive F(2,174) = 10.92, p < .001, η2 = .111
behaviors, F(2,174) = 9.96, p < .001, η2 = .103
significant F(2,174) = 8.874, p < .001, η2 = .093.
interaction effects
are observed in
overt aggression,
relational
aggression, and
cyber-aggression
In post-test (after –
program
implementation),
mean scores in
variables of overt
aggression,
relational
aggression, and
cyber-aggression of
EG participants are
significantly lower
than mean scores of
CG1 and CG2
adolescents.

NOTE: ARA - Adolescent Relationship Abuse; CI – Confidence Interval; CG – Control Group; CFI - Comparative Fit Index; CBIM - Coaching Boys Into
Men; DV – Domestic Violence; ES – Effect Size; EG – Experimental Group; HIV – Human Immunodeficiency Virus; OR – Odds Ratio; RS+ - Relationship
Smarts Plus; RCT- Randomized Control Trial; RELATE – Relationship Evaluation; RMSEA - Root Mean Square Error of Approximation; RE – Rela­
tionship Education; STD - Sexually Transmitted Disease; SE – Standard Error; SD – Standard Deviation; SES - Socio-Economic Status; SV - Sexual
Violence; TDV - Teen Dating Violence.

as part of a broader sexual health paradigm, including additional unique components such as HIV/STD knowledge, sexual norm­
s/attitudes, and safer sex self-efficacy (Scull et al., 2018).

3.3. Intervention effect

3.3.1. Healthy relationship interventions (n = 10 studies)

Table 3 reports on the eight healthy relationship interventions (reported in ten studies) and their key results. The studies had a
follow-up period ranging between two weeks to two years, and the majority (60.0%) had an RCT design. The key outcomes evaluated
among adolescent participants included: (1) relationship knowledge on self-reported questionnaires; (2) relationship attitudes and
beliefs (e.g. reduction of faulty relationship beliefs, gender role beliefs, and dating violence acceptance); (3) relationship satisfaction;
(4) perceived ability to manage conflict effectively in close relationships; (5) openness to future relationship education; (6) frequency
of verbal aggression and conflict behaviors in relationships; and (7) sexual health attitudes, knowledge, communication, and behavior.
The most consistent finding across studies was a significant improvement of perceived knowledge about romantic relationship skills
in the intervention group compared to controls (Bradford et al., 2016; Kerpelman et al., 2009; Larson et al., 2007; Schramm &
Gomez-Scott, 2012). Bradford et al. (2016) and Schramm and Gomez-Scott (2012) also found that perceived knowledge about partner

228
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

selection significantly improved in the treatment group compared to controls (t = 17.31, p < .001; F(1,620) = 8.26, p = .004). Sig­
nificant improvements in sexual health knowledge were observed amongst intervention participants in both the Media Aware (p <
.001) (Scull et al., 2018) and HEART (p < .001, d = 1.27) (Widman, Evans, et al., 2020) programs. These studies showed mixed results
on healthy relationship attitudes and beliefs (Scull et al., 2018; Widman, Evans, et al., 2020). However, there were a number of
significant findings which included intervention participants reporting a significant decrease in their belief that love is enough to
sustain a happy marriage compared to controls (F(1,620) = 4.12, p = .04) (Schramm & Gomez-Scott, 2012). Savasuk-Luxton et al.
(2018) found significant differences in dating violence acceptance at post-test (F (1,2099) = 18.72, p < .001), such that adolescents in
the intervention group reported less dating violence acceptance (Meancomparisongroup = 1.81 (SD = 1.09) and Meanintervention group = 1.59
(SD = 0.96)) than adolescents in the comparison group. Similarly, Media Aware (Scull et al., 2018) participants showed a reduction in
the acceptance of dating violence (p = .008), as well as strict gender role norms (p = .009). They also displayed more positive attitudes
about communicating with a romantic partner (p = .004) at post-test compared with the control group.
Adler-Baeder et al. (2007) found less consistent findings with respect to relationship beliefs, where out of three relationship belief
subscales (aggression, faulty relationship, and realistic relationship beliefs), only the realistic belief subscale showed statistically
significant improvement in the intervention group (F(1,317)=4.71, p = .03). Both Coyle et al. (2019) and Scull et al. (2018) studies
showed multiple non-significant results in terms of sexual-health and communication-related attitudes and beliefs. Savasuk-Luxton
et al. (2018) found there was no statistically significant difference in change patterns in gender role beliefs between intervention and
comparison adolescents (Wilks’ Lambda = 1.00, F(1,2048) = 0.69, p = .41, ɳ2 = 0.0).
Studies also found limited evidence in terms of behavioral outcomes. There was a reduction in verbal aggression, but not physical
aggression nor reasoning, at follow-up in the intervention group compared to controls (F(1,297) = 5.22, p = .02) (Adler-Baeder et al.,
2007). Kerpelman et al. (2009) found no change in verbal aggression between groups. In terms of sexual initiation and sexual touching
behaviors, Coyle et al. (2019) found no statistically significant differences between intervention and control groups.

3.3.2. Healthy/unhealthy relationship interventions (n = 20 studies)

There were 17 interventions (reported in 20 studies) which were primarily focused on reducing or preventing various forms of
relationship abuse (e.g. dating violence education curricula), with a component on the promotion of healthy relationship skills and
knowledge (see Table 3). The studies of interest had a follow-up period ranging between immediately post-intervention to two and half
years, and the majority (75.0%) were an RCT design. The key outcomes evaluated included: (1) dating violence or abuse knowledge
and attitudes; (2) abuse victimization and perpetration (psychological, physical, or sexual); (3) conflict management skills; (4) gender
norms and stereotyping; (5) healthy relationship attitudes; (6) use of healthy relationship skills; (7) anger management; (8) use of
positive communication skills; and (9) romantic relationship quality.
Few studies examined healthy or unhealthy relationship knowledge as an outcome. The exception was the ‘Coaching Boys into
Men’ program (Miller et al., 2020), which demonstrated an increased recognition of relationship abuse among male, middle-school
student athletes at post-test and one-year follow-up (mean risk difference, 0.14; 95% CI, 0.01–0.27 and 0.14; 95% CI, 0.00–0.28,
respectively).
In terms of healthy relationship attitudinal outcomes, the most consistent finding across studies was intervention groups were
significantly less acceptance of dating violence norms compared to controls (Avery-Leaf et al., 1997; Banyard et al., 2019; Foshee et al.,
1998, 2000, 2012; Jaycox et al., 2006; Joppa et al., 2016; Lazarevich et al., 2017; Levesque et al., 2016). Another consistent finding
was significantly less acceptance of prescribed dating norms (Foshee et al., 1998, 2005), less acceptance of traditional gender-role
norms (Banyard et al., 2019; Foshee et al., 1998, 2000, 2005; Miller, Williams, et al., 2015), and more constructive responses to
relationship-related anger reported in the intervention group compared to controls at follow-up (Foshee et al., 1998, 2000; Joppa et al.,
2016; Sanchez-Jimenez et al., 2018). Two studies assessed the ‘Date-e’ program (Munoz-Fernandez et al., 2019; Sanchez-Jimenez
et al., 2018). Sanchez-Jimenez et al. (2018) reported a significant impact on change across all myths of romantic love, reporting a
reduction in the acceptance of myths among intervention versus control group participants. The effect sizes were large for reductions in
the better half myth (i.e. the belief that we are incomplete without the other, and that the perfect person is out there for everyone; d =
− .83), omnipotence (d = − 0.84), and passion (d = − 0.94), with a medium effect size for reductions in jealousy (d = − 0.56) (San­
chez-Jimenez et al., 2018). A number of studies showed no significant difference between intervention and control groups across a
range of attitudinal measures including supporting violence prevention (Banyard et al., 2019) and gender attitudes (Miller et al.,
2020).
There were mixed findings in terms of behavioral outcomes. Some studies found that interventions were effective in reducing levels
of reported abuse victimization and perpetration (e.g. ARA at 1-year follow-up; OR = 0.24, 95% CI = 0.09–0.65 (Miller et al., 2020)
and small effect sizes for reduction in severe physical/sexual dating victimization and aggression were reported (Munoz-Fernandez
et al., 2019); across psychological, sexual, and physical measures (Carrascosa et al., 2019; Foshee et al., 1998; Joppa et al., 2016;
Levesque et al., 2016; Miller et al., 2015, 2020; Munoz-Fernandez et al., 2019). The Teen Choices Intervention was associated with
consistently higher odds of healthy relationship skill use (e.g. mentalisation, communication and conflict resolution) across the
12-month follow up (OR range:1.68–2.60) (Levesque et al., 2016). Other studies did not find a statistically significant difference
between intervention and control participants on these measures at follow-up, particularly with respect to psychological abuse
perpetration and victimization (Foshee et al., 2004, 2005, 2012; Jaycox et al., 2006; Munoz-Fernandez et al., 2019; Sanchez-Jimenez
et al., 2018). Improvements in conflict resolution skills were also less apparent in the included studies. Foshee et al. (2005) and Foshee
et al. (2012) found no significant improvements in conflict resolution across two studies investigating the Safe Dates Curriculum
intervention.
Table 3 presents outcome measures, key results, and effects sizes reported in the reviewed studies. There was considerable variation

229
 E. Hielscher et al.
230

Fig. 2. Risk assessment of 18 RCT studies (reported in 21 publications) using the Cochrane Risk of Bias Tool.

Journal of Adolescence 92 (2021) 194–236

 E. Hielscher et al.
231

Fig. 3. Risk assessment of 9 non-randomized studies using the ROBINS-I tool.

Journal of Adolescence 92 (2021) 194–236

E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

in measured outcomes and key results, and inconsistent reporting of test statistics/effect sizes across reviewed studies.

3.4. Risk of bias assessment

3.4.1. Risk of bias – RCT studies (n = 21)

3.4.1.1. Healthy relationship RCT interventions (n = 6). To assess the level of bias in RCT study designs (n = 6 studies), we employed
the Cochrane Risk of Bias Tool (see Fig. 2). The six healthy relationship interventions were compared to a control condition or
‘treatment as usual’ classroom condition. Three (50%) of the studies (Kerpelman et al., 2009; Larson et al., 2007; Lederman et al.,
2008) were rated as ‘high risk’ on at least three (of seven) risk of bias criteria including: (1) blinding of participants/researchers, and
(2) incomplete outcome data, where missing data was non-negligible or was not appropriately dealt with by the study authors. The
other three studies (Coyle et al., 2019; Scull et al., 2018; Widman, Evans, et al., 2020) were rated unclear or high risk for blinding of
participants/researchers and outcome assessment, but low risk for incomplete outcome data. Across all six intervention trials, it was
unclear if there was adequate concealment of participant allocation (to intervention or control group) prior to assignment.

3.4.1.2. Unhealthy/healthy relationship RCT interventions (n = 15). The 15 RCT interventions were largely compared to a ‘treatment as
usual’ classroom condition which covered similar content but lacked the applied, skill-based focus of the intervention condition. One
study (Wolfe et al., 2009) received a ‘low risk’ rating for all criteria, and four studies (Foshee et al., 2000; Miller, Williams, et al., 2015;
Munoz-Fernandez et al., 2019; Sanchez-Jimenez et al., 2018) received this rating for at least five (of seven) risk of bias criteria (see
Fig. 2). All other studies (66.6%) received mixed ratings, with incomplete outcome data and blinding of participants/researchers of
particular concern.

3.4.2. Risk of bias – non-randomized studies (n = 9)

3.4.2.1. Healthy relationship non-randomized interventions (n = 4). To assess the level of bias in non-randomized studies (n = 4), we
employed the ROBINS-I tool (see Fig. 3). Three (of the four) healthy intervention studies (Adler-Baeder et al., 2007; Avery-Leaf et al.,
1997; Bradford et al., 2016) were rated as ‘high risk’ on at least two (of eight) risk of bias criteria including: (1) bias due to confounding
(as without random assignment to groups it is plausible that other unmeasured factors may have contributed to the significant dif­
ferences found in these studies); and (2) bias in measurement of outcomes, where the measures incorporated had limited reliability and
validity to capture the constructs of interest, or lacked psychometric testing altogether. It was also unclear if there were deviations
from the intended intervention.

3.4.2.2. Unhealthy/healthy relationship non-randomized interventions (n = 5). The non-randomized studies (n = 5 studies) were
assessed using the ROBINS-I tool (see Fig. 3). Three of the five studies (Braithwaite et al., 2010; Miller, Williams, et al., 2015; Schramm
& Gomez-Scott, 2012) were rated as ‘high risk’ on at least three (of eight) risk of bias criteria. Similar to the healthy relationship
interventions, the main areas for concern were (1) bias due to confounding, and (2) bias in measurement of outcomes.

3.5. Key characteristics of proposed healthy relationship definition

3.5.1. Coverage in intervention designs

We assessed each intervention of included studies to determine if it was clear which specific feature/s of a healthy romantic
relationship they were intending to target/promote. The summary data extracted in relation to healthy relationship definition char­
acteristics covered or addressed by each intervention are shown in Fig. 4 (with reviewer ratings indicated by coloured squares in the
figure). We found ‘communication’ and ‘negotiation skills’ were (intended to be) targeted or promoted via all interventions reported in
reviewed studies, except for one study (Scull et al., 2018). Whereas ‘caregiving behavior”, self-expression’, ‘honesty’, and ‘fairness’
were not covered well by most interventions (see Fig. 4).

3.5.2. Coverage in outcome measures

We also assessed outcome measures of each included study in terms of whether change across each of the different characteristics/
elements of a healthy romantic relationship was examined (indicated by “+” sign in Fig. 4). We found that changes in ‘communication’
and ‘negotiation skills’ were measured as an outcome of many of the included interventions (n = 18, 60.0% of total interventions). For
example, adolescents in The Safe Dates Curriculum group, as compared with those in the control group, used more constructive
communication skills and responses to anger, were less likely to engage in gender stereotyping, and were more aware of victim and
perpetrator services (Foshee et al., 1998). The remaining definition characteristics, i.e. ‘caregiving behavior’, ‘self-expression’,
‘respect’, ‘trust’, ‘honesty’, and ‘fairness’, were inconsistently or sparsely covered. Most outcome measures in the included studies were
too broad to make a sound judgement as to whether they measured change across the specific characteristics of the healthy re­
lationships definition (i.e. strong communication and negotiation skills, caregiving behaviors, self-expression, respect, trust, honesty,
and fairness). See Fig. 4 for more details.

232
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Fig. 4. Coverage of key characteristics of proposed healthy relationship definition, in both the intervention design (coloured squares) and outcome
measure (+).

4. Discussion

This is one of the first systematic reviews to critically appraise interventions designed to foster and promote healthy romantic
relationships in youth (aged 12–25 years). This review found that, overall, healthy romantic relationship interventions were effective
at improving healthy relationship knowledge. Although there were numerous significant findings in the included studies, the gains in
knowledge across time (follow-up range two weeks to two years) tended to be small. Lederman et al. (2008) proposed that this was
likely the result of higher than expected relationship knowledge scores at baseline in both intervention and control groups. In terms of
healthy relationship attitudinal and belief outcomes, studies showed mixed results supporting intervention effectiveness. Despite
mixed results on these attitudinal measures, the range of notable and significant findings at post-test follow ups (Savasuk-Luxton et al.,
2018; Schramm & Gomez-Scott, 2012; Scull et al., 2018) were promising (e.g. improved communication with partner and reduced
dating violence acceptance). The most consistent finding across studies was intervention groups were significantly less accepting of
dating violence norms.
In general, included studies did not incorporate or measure behavioral outcomes. The few studies that did showed limited evidence
for interventions resulting in change across such measures (e.g. small to no reduction in verbal aggression) (Adler-Baeder et al., 2007;
Coyle et al., 2019). Whilst none of the included intervention studies focused solely on physical forms of relationship abuse, a number
reported on the impact or performance of these interventions (particularly for behavioral outcomes) using physical abuse-related
outcomes (e.g. physical dating abuse victimization, physical aggression). Overall, results were limited and mixed for behavioral
outcomes where some studies found interventions were effective at reducing abuse victimization and perpetration (Carrascosa et al.,
2019; Foshee et al., 1998; Joppa et al., 2016; Levesque et al., 2016; Miller et al., 2020; Munoz-Fernandez et al., 2019) and increasing
healthy relationship skill use (Levesque et al., 2016). Others did not find significant changes with respect to these outcomes (Foshee
et al., 2005, 2012; Jaycox et al., 2006; Munoz-Fernandez et al., 2019; Sanchez-Jimenez et al., 2018).
In summary, the studies reviewed show the strongest evidence for change in knowledge outcomes, mixed evidence for change in
attitudes/beliefs outcomes, and limited evidence for change in behavioral outcomes. One previous similar systematic review (among
those aged 13–29 years) (Simpson et al., 2018) found largely consistent results, where healthy relationship interventions showed
medium-sized impacts on romantic relationship knowledge and attitudes, at least in the short-term. This review differed from ours in
that it focused specifically on individually oriented relationship education programs (we did not make a distinction between individual
vs couple-focused interventions) and it was largely inclusive of pre-post, single group studies (76.5% of its studies) (Simpson et al.,
2018). Studies measuring the effectiveness of interventions for promoting healthy adolescent romantic relationships or preventing
adolescent relationship abuse require a comparison group. Adolescents’ knowledge and behaviors relating to relationships typically
develop with the rapid emotional and cognitive maturation that occurs during this developmental phase. Further, adolescents may be
exposed to new relationship experiences during follow up periods of an interventions. It is therefore essential for studies to have a
control/comparison group to measure intervention effectiveness. More broadly, previous studies and reviews have focused on in­
terventions aimed at primarily preventing adolescent relationship abuse. Limited or inconsistent evidence has been found to support

233
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

the effectiveness of these interventions across both knowledge, attitudes, and behavioral relationship outcomes (Fellmeth et al., 2013;
Lundgren & Amin, 2015; Rehana et al., 2016).
There was considerable variation in targeted or promoted healthy relationship components via interventions as well as measured
outcomes reported in reviewed studies. Only one of six characteristics of the proposed healthy relationship definition, that being
‘communication/negotiation skills’, was consistently targeted and measured in the intervention programs reported in the reviewed
studies; the remaining five definition characteristics were inconsistently or sparsely covered (Fig. 4). Furthermore, the results of these
studies demonstrated that ‘communication’ and ‘negotiation skill’ outcomes were improved via and sensitive to the intervention
(Foshee et al., 1998; Jaycox et al., 2006; Scull et al., 2018).

4.1. Strengths and weaknesses

Strengths of the studies reviewed included most incorporated an RCT design (70%) and were able to recruit relatively large sample
sizes. Furthermore, almost half the studies (43.3%) incorporated a theoretical basis into their design.
The quality of the studies was limited by bias due to confounding, bias in measurement of outcomes, and incomplete outcome data.
It was often unclear if there was adequate allocation concealment. It is worth noting that four of the studies administered the same
intervention and were evaluated by the same author group (Foshee et al., 1998, 2000, 2004, 2005). Moreover, healthy relationship
knowledge was not assessed as an outcome measure in these studies, making it difficult to draw conclusions about intervention
effectiveness on such measures. Also, many of the studies treated ‘healthy relationship skills/knowledge’ as a mediator between the
intervention and its effect on relationship abuse outcomes, and therefore, was not the primary outcome of interest in many of the
included studies. This may have implications for the fidelity of measures employed, as well as the rigorousness of statistical analyses.
Studies generally used self-report outcome measures which would be subject to recall bias.
Furthermore, the majority of studies were conducted in the United States (77.8%) making it difficult to draw definitive conclusions
about the interventions’ effectiveness in diverse groups of young people from other countries with varying religious, ethnic, socio-
economic, and cultural backgrounds. There was also limited consideration in existing interventions for adolescents with varying
sexual orientations and gender identifications. Moreover, most of the trials had short follow-up periods that were likely insufficient to
detect any meaningful change in attitudes or behaviors, and provided little evidence supporting intervention effectiveness in the
medium- and long-term. The short follow up time is particularly important because many of the adolescents to whom the intervention
is provided may not have experienced their first intimate relationship during the period of follow up. Finally, the current review only
reported interventions confined to either the middle school, high school, or university life stage. Future review studies should aim to
conduct age group analyses to find developmental and age-related differences.
In terms of the strengths and limitations of the review itself, a strength was the review’s comprehensive search strategy which
included interventions that also had an abuse focus. A meta-analysis was not performed because of the heterogeneity between study
interventions, including delivery settings, types of outcomes assessed, and duration of the intervention. The search was limited to the
English language and excluded restricted populations (i.e. clinical, help-seeking, vulnerable or at-risk populations). In terms of the
latter, it is possible that the interventions reviewed could have differing effectiveness in at-risk or lower socioeconomic populations, as
opposed to broader, universal settings (where adolescents may already have a considerable baseline understanding of broader social
and emotional wellbeing). This would require further inquiry.

4.2. Public health implications and future research

In terms of the effectiveness of programs, a key consideration was the presence versus absence of a therapist in interventions.
Larson et al. (2007) found therapy assistance provided a significant benefit to users of the RELATE questionnaire, compared to the
self-interpretation and control couples. This is an important consideration for future interventions, particularly with respect to
designing interventions intended to be delivered online which may not provide moderator support.
Encouragingly, the studies highlighted that embedding interventions into the existing education curricula appears to be a viable
mechanism to reach and implement change in ‘healthy romantic relationship knowledge’ among a broad range of young people.
However, further investigation is required regarding the acceptability (among teachers, parents, and students) and cost effectiveness of
these programs. Future researchers should also consider the use of feasibility and fidelity measures. Other considerations for future
research include the use of behavioral outcome measures and whether improvement in healthy relationship knowledge and attitudes
translates into actual behavioral change. Studies with longer follow up periods (3+ years) are also required to investigate whether
there are sustained effects as a result of these interventions. This would help to inform the need to potentially provide booster sessions
over several time points.

4.3. Conclusion

This review was comprehensive in identifying the relevant literature and supported intervention effectiveness in terms of
improving healthy romantic relationship knowledge in the target population. However, the findings were mixed for intervention
effectiveness in changing attitudes/beliefs, and there was limited evidence to support change across behavioral outcomes. This review
highlights the need for future research in this area, particularly the implementation of high quality RCTs in a broad range of cultural
and ethnic settings, to improve the generalisability of findings. Improving knowledge and changing adolescent behavior relating to
heathy romantic relationships has potential to reduce mental and physical health problems during this phase of development.

234
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Sources of funding and acknowledgements

C.M. was employed with funds from the generous gift of Ms Mia Pattison to The University of Queensland. JGS is supported by a
National Health and Medical Research Council Practitioner Fellowship Grant (grant number 1105807). EH is supported by the Dr F and
Mrs ME Zaccari Scholarship.

Declaration of competing interest

The authors declare no conflicts of interest.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.adolescence.2021.08.008.

References

Adler-Baeder, F., Kerpelman, J. L., Schramm, D. G., Higginbotham, B., & Paulk, A. (2007). The impact of relationship education on adolescents of diverse
backgrounds. Family Relations, 56(3), 291–303. https://doi.org/10.1111/j.1741-3729.2007.00460.x.
Ainsworth, M. S. (1989). Attachments beyond infancy. American Psychologist, 44(4), 709.
Avery-Leaf, S., Cascardi, M., O’Leary, K. D., & Cano, A. (1997, Jul). Efficacy of a dating violence prevention program on attitudes justifying aggression. Journal of
Adolescent Health, 21(1), 11–17. https://doi.org/10.1016/S1054-139X(96)00309-6.
Banyard, V. L., Edwards, K. M., Rizzo, A. J., Theodores, M., Tardiff, R., Lee, K., & Greenberg, P. (2019). Evaluating a gender transformative violence prevention
program for middle school boys: A pilot study. Children and Youth Services Review, 101, 165–173. https://doi.org/10.1016/j.childyouth.2019.03.052.
Bonomi, A. E., Anderson, M., Nemeth, J., Rivara, F. P., & Buettner, C. (2013). History of dating violence and the association with late adolescent health. BMC Public
Health, 13, 821. https://doi.org/10.1186/1471-2458-13-821.
Borzekowski, D. (2009). Considering children and health literacy: A theoretical approach. Pediatrics, 124, S282–S288. https://doi.org/10.1542/peds.2009-1162d.
Bowlby, J. (1969). Attachment and loss (Vol. 1). Basic Books.
Bradford, K., Stewart, J. W., Pfister, R., & Higginbotham, B. J. (2016). Avoid falling for a jerk(ette): Effectiveness of the premarital interpersonal Choices and
knowledge program among emerging adults. Journal of Marital and Family Therapy, 42(4), 630–644. https://doi.org/10.1111/jmft.12174.
Braithwaite, S. R., Lambert, N. M., Fincham, F. D., & Pasley, K. (2010, Dec). Does college-based relationship education decrease extradyadic involvement in
relationships? Journal of Family Psychology, 24(6), 740–745. https://doi.org/10.1037/a0021759.
Carrascosa, L., Cava, M. J., Buelga, S., & de Jesus, S. N. (2019, May). Reduction of sexist attitudes, romantic myths, and aggressive behaviors in adolescents: Efficacy of
the DARSI program. Psicothema, 31(2), 121–127. https://doi.org/10.7334/psicothema2018.245.
Coyle, K. K., Anderson, P., Franks, H. M., Walker, J. D., & Glassman, J. R. (2019, Dec). You-Me-Us: Results of a cluster randomized trial of a healthy relationships
approach to sexual risk reduction. Journal of Primary Prevention, 40(6), 607–629. https://doi.org/10.1007/s10935-019-00569-w.
De Koker, P., Mathews, C., Zuch, M., Bastien, S., & Mason-Jones, A. J. (2014). A systematic review of interventions for preventing adolescent intimate partner
violence. Journal of Adolescent Health, 54. https://doi.org/10.1016/j.jadohealth.2013.08.008.
Dick, R., McCauley, H., Jones, K. A., Tancredi, D. J., Goldstein, S., Blackburn, S., Monasterio, E., James, L., Silverman, J. G., & Miller, E. (2014). Cyber dating abuse
among teens using school-based health centers. Pediatrics, 136, e1560–1567. https://doi.org/10.1542/peds.2014-0537.
Dos Santos, K. B., Murta, S. G., do Amaral Vinha, L. G., & de Deus, J. S. (2019). Efficacy of a bystander intervention for preventing dating violence in Brazilian
adolescents: Short-term evaluation. Psicologia: Reflexão e Crítica, 32(1), 20. https://doi.org/10.1186/s41155-019-0133-4.
Eldridge, S., Campbell, M., Campbell, M., Drahota-Towns, A., Giraudeau, B., Higgins, B., Reeves, N., & Siegfried, N. (2016). Revised Cochrane risk of bias tool for
randomized trials (RoB 2.0): Additional considerations for cluster-randomized trials. https://sites.google.com/site/riskofbiastool/welcome/rob-2-0-tool.
Fellmeth, G., Heffernan, C., Nurse, J., Habibula, S., & Sethi, D. (2013). Educational and skills-based interventions for preventing relationship and dating violence in
adolescents and young adults. Cochrane Database of Systematic Reviews, (6)https://doi.org/10.4073/csr.2013.14. Article CD004534.
Finkel, E. J., DeWall, C. N., Slotter, E. B., Oaten, M., & Foshee, V. A. (2009). Self-regulatory failure and intimate partner violence perpetration. Journal of Personality
and Social Psychology, 97(3), 483–499. https://doi.org/10.1037/a0015433.
Foshee, V., Bauman, K. E., Arriaga, X. B., Helms, R. W., Koch, G. G., & Linder, G. F. (1998). An evaluation of safe dates, an adolescent dating violence prevention
program. American Journal of Public Health, 88(1), 45–50. https://doi.org/10.2105/AJPH.88.1.45.
Foshee, V. A., Bauman, K. E., Ennett, S. T., & Fletcher, L. G. (2004). Assessing the long-term effects of the safe dates program and a booster in preventing and reducing
adolescent dating violence victimization and perpetration. American Journal of Public Health, 94(4). https://doi.org/10.2105/AJPH.94.4.619.
Foshee, V. A., Bauman, K. E., Ennett, S. T., Suchindran, C., Benefield, T., & Linder, G. F. (2005). Assessing the effects of the dating violence prevention program "safe
dates" using random coefficient regression modeling. Prevention Science, 6(3), 245–258. https://doi.org/10.1007/s11121-005-0007-0.
Foshee, V. A., Bauman, K. E., Greene, W. F., Koch, G. G., Linder, G. F., & MacDougall, J. E. (2000). The safe dates program: 1-year follow-up results. American Journal
of Public Health, 90(10), 1619–1622. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1446374/pdf/11029999.pdf.
Foshee, V. A., Benefield, T., Suchindran, C., Ennett, S. T., Bauman, K. E., Karriker-Jaffe, K. J., Reyes, H. L., & Mathias, J. (2009). The development of four types of
adolescent dating abuse and selected demographic correlates. Journal of Research on Adolescence, 19(3), 380–400. https://doi.org/10.1111/j.1532-
7795.2009.00593.x.
Foshee, V. A., McNaughton Reyes, H. L., Ennett, S. T., Cance, J. D., Bauman, K. E., & Bowling, J. M. (2012). Assessing the effects of Families for Safe Dates, a family-
based teen dating abuse prevention program. Journal of Adolescent Health, 51(4), 349–356. https://doi.org/10.1016/j.jadohealth.2011.12.029.
Furman, W., & Wehner, E. A. (1994). Romantic views: Toward a theory of adolescent romantic relationships. In Personal relationships during adolescence (pp. 168–195).
Sage Publications, Inc.
Halpern, C. T., Oslak, S. G., Young, M. L., Martin, S. L., & Kupper, L. L. (2001). Partner violence among adolescents in opposite-sex romantic relationships: Findings
from the national longitudinal study of adolescent health. American Journal of Public Health, 91(10), 1679–1685. https://www.ncbi.nlm.nih.gov/pmc/articles/
PMC1446854/pdf/0911679.pdf.
Higgins, J. P. T., Thomas, J., Chandler, J., Cumpston, M., Li, T., Page, M. J., & Welch, V. A. (Eds.). (2019). Cochrane Handbook for Systematic Reviews of Interventions
(2nd Edition). Chichester (UK): John Wiley & Sons.
Jaycox, L. H., McCaffrey, D., Eiseman, B., Aronoff, J., Shelley, G. A., Collins, R. L., & Marshall, G. N. (2006, Nov). Impact of a school-based dating violence prevention
program among latino teens: Randomized controlled effectiveness trial. Journal of Adolescent Health, 39(5), 694–704. https://doi.org/10.1016/j.
jadohealth.2006.05.002.

235
E. Hielscher et al. Journal of Adolescence 92 (2021) 194–236

Jonathan, A. S., Miguel, A. H., Barnaby, C. R., Savović, J., Nancy, D. B., Viswanathan, M., Henry, D., Altman, D. G., Ansari, M. T., Boutron, I., Carpenter, J. R.,
Chan, A.-W., Churchill, R., Deeks, J. J., Hróbjartsson, A., Kirkham, J., Jüni, P., Loke, Y. K., Pigott, T. D., … Higgins, J. P. (2016). ROBINS-I: A tool for assessing risk
of bias in non-randomised studies of interventions. BMJ. https://doi.org/10.1136/bmj.i4919.
Joppa, M. C., Rizzo, C. J., Nieves, A. V., & Brown, L. K. (2016, Apr). Pilot investigation of the katie Brown educational program: A school-community partnership.
Journal of School Health, 86(4), 288–297. https://doi.org/10.1111/josh.12378.
Kerpelman, J. L., Pittman, J. F., Adler-Baeder, F., Eryigit, S., & Paulk, A. (2009). Evaluation of a statewide youth-focused relationships education curriculum. Journal
of Adolescence, 32(6), 1359. https://doi.org/10.1016/j.adolescence.2009.04.006.
Larson, J. H., Vatter, R. S., Galbraith, R. C., Holman, T. B., & Stahmann, R. F. (2007). The relationship evaluation (RELATE) with therapist-assisted interpretation:
Short-term effects in premarital relationships. Journal of Marital and Family Therapy, 33(3), 364–374. https://doi.org/10.1111/j.1752-0606.2007.00036.x.
Lazarevich, I., Irigoyen-Camacho, M. E., del Consuelo Velázquez-Alva, M., & Salinas-Avila, J. (2017). Dating violence in Mexican college students: Evaluation of an
educational workshop. Journal of Interpersonal Violence, 32(2), 183–204. https://doi.org/10.1177/0886260515585539.
Lederman, R. P., Chan, W., & Roberts-Gray, C. (2008). Parent-adolescent relationship education (PARE): Program delivery to reduce risks for adolescent pregnancy
and STDs. Behavoiral Medicine, 33(4), 137–143. https://doi.org/10.3200/bmed.33.4.137-144.
Levesque, D. A., Johnson, J. L., Welch, C. A., Prochaska, J. M., & Paiva, A. L. (2016). Teen dating violence prevention: Cluster-randomized trial of teen Choices, an
online, stage-based program for healthy, nonviolent relationships. Psychology of Violence, 6(3), 421–432. https://doi.org/10.1037/vio0000049.
Lundgren, R., & Amin, A. (2015). Addressing intimate partner violence and sexual violence among adolescents: Emerging evidence of effectiveness. Journal of
Adolescent Health, 56(S42eS50). https://doi.org/10.1016/j.jadohealth.2014.08.012.
Miller, E. (2017, Apr). Prevention of and interventions for dating and sexual violence in adolescence. Pediatric Clinics of North America, 64(2), 423–434. https://doi.
org/10.1016/j.pcl.2016.11.010.
Miller, E., Goldstein, S., McCauley, H. L., Jones, K. A., Dick, R. N., Jetton, J., Silverman, J. G., Blackburn, S., Monasterio, E., James, L., & Tancredi, D. J. (2015, Jan).
A school health center intervention for abusive adolescent relationships: A cluster RCT. Pediatrics, 135(1), 76–85. https://doi.org/10.1542/peds.2014-2471.
Miller, E., Jones, K. A., Ripper, L., Paglisotti, T., Mulbah, P., & Abebe, K. Z. (2020). An athletic coach-delivered middle school gender violence prevention program: A
cluster randomized clinical trial. JAMA Pediatrics. https://doi.org/10.1001/jamapediatrics.2019.5217.
Miller, S., Williams, J., Cutbush, S., Gibbs, D., Clinton-Sherrod, M., & Jones, S. (2015). Evaluation of the start strong initiative: Preventing teen dating violence and
promoting healthy relationships among middle school students. Journal of Adolescent Health, 56(2), S14–S19. https://doi.org/10.1016/j.jadohealth.2014.11.003.
Moher, D., Liberati, A., Tetzlaff, J., Altman, D. G., & The PRISMA Group. (2009). Preferred reporting Items for systematic reviews and meta-analyses: The PRISMA
statement. https://doi.org/10.1371/journal.pmed.1000097.
Munoz-Fernandez, N., Ortega-Rivera, J., Nocentini, A., Menesini, E., & Sanchez-Jimenez, V. (2019, Jan 31). The efficacy of the "Dat-e adolescence" prevention
program in the reduction of dating violence and bullying. International Journal of Environmental Research and Public Health, 16(3). https://doi.org/10.3390/
ijerph16030408.
Noonan, R. K., & Charles, D. (2009). Developing teen dating violence prevention strategies: Formative research with middle school youth. Violence Against Women, 15
(9), 1087. https://doi.org/10.1177/1077801209340761.
Pence, E., & Paymar, M. (1993). Power and control: Tactics of men who batter. Minnesota Program Development, Inc.
Rehana, A. S., Anadil, F., Nida, S., Zohra, S. L., Jai, K. D., Miriam, K., & Zulfiqar, A. B. (2016). Improving adolescent sexual and reproductive health: A systematic
review of potential interventions. Jounral of Adolescent Health, 59(S11-S28). https://doi.org/10.1016/j.jadohealth.2016.05.022.
Sanchez-Jimenez, V., Munoz-Fernandez, N., & Ortega-Rivera, J. (2018). Efficacy evaluation of "Dat-e adolescence": A dating violence prevention program in Spain.
PloS One, 13(10), Article e0205802. https://doi.org/10.1371/journal.pone.0205802.
Savasuk-Luxton, R., Adler-Baeder, F., & Haselschwerdt, M. L. (2018, Feb). Understanding change in violence-related attitudes for adolescents in relationship
education. Journal of Adolescence, 63, 153–164. https://doi.org/10.1016/j.adolescence.2017.12.012.
Schramm, D. G., & Gomez-Scott, J. (2012). Merging relationship education and child abuse prevention knowledge: An evaluation of effectiveness with adolescents.
Marriage & Family Review, 48(8), 792. https://doi.org/10.1080/01494929.2012.714722.
Scull, T. M., Kupersmidt, J. B., Malik, C. V., & Morgan-Lopez, A. A. (2018). Using Media literacy education for adolescent sexual health promotion in middle school:
Randomized control trial of Media aware. Journal of Health Communication, 23(12), 1051–1063. https://doi.org/10.1080/10810730.2018.1548669.
Simpson, D. M., Leonhardt, N. D., & Hawkins, A. J. (2018, Mar). Learning about love: A meta-analytic study of individually-oriented relationship education programs
for adolescents and emerging adults. Journal of Youth and Adolescence, 47(3), 477–489. https://doi.org/10.1007/s10964-017-0725-1.
Simpson, D. M., Leonhardt, N. D., & Hawkins, A. J. (2017). Learning about love: A meta-analytic study of individually- oriented relationship education programs for
adolescents and emerging adults. Journal of Youth and Adolescence. https://doi.org/10.1007/s10964-017-0725-1.
Stöckl, H., March, L., Pallitto, C., & Garcia-Moreno, C. (2014). Intimate partner violence among adolescents and young women: Prevalence and associated factors in
nine countries: A cross-sectional study. BMC Public Health, 14(75), 1–14.
Sullivan, H. S. (1953). The interpersonal theory of psychiatry. Norton.
Widman, L. K. K., Evans, R., Stewart, J. L., Choukas-Bradley, S., & Golin, C. E. (2020). Feasibility, acceptability, and preliminary efficacy of a brief online sexual health
program for adolescents. The Journal of Sex Research, 145–154. https://doi.org/10.1080/00224499.2019.1630800.
Widman, L., Kamke, K., Evans, R., Stewart, J. L., Choukas-Bradley, S., & Golin, C. E. (2020). Feasibility, acceptability, and preliminary efficacy of a brief online sexual
health program for adolescents. The Journal of Sex Research, 57(2), 145–154. https://doi.org/10.1080/00224499.2019.1630800.
Wolfe, D. A., Crooks, C., Jaffe, P., Chiodo, D., Hughes, R., Ellis, W., Stitt, L., & Donner, A. (2009, Aug). A school-based program to prevent adolescent dating violence:
A cluster randomized trial. Archives of Pediatrics and Adolescent Medicine, 163(8), 692–699. https://doi.org/10.1001/archpediatrics.2009.69.

236