Available via license: CC BY 4.0
Content may be subject to copyright.
Citation: Hecht, M.L.; Norris, A.E.;
Crowley, D.M.; Temple, J.R.; Choi, H.J.
Findings from the Mighty Girls
Efficacy Trial: Changes in Acceptance
of Dating Violence. Children 2024,11,
1331. https://doi.org/10.3390/
children11111331
Academic Editors: Matteo Fabris,
Heng Luo and Ikseon Choi
Received: 23 September 2024
Revised: 27 October 2024
Accepted: 28 October 2024
Published: 31 October 2024
Copyright: © 2024 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
children
Article
Findings from the Mighty Girls Efficacy Trial: Changes in
Acceptance of Dating Violence
Michael L. Hecht 1, * , Anne E. Norris 2, Daniel Max Crowley 3, Jeff R. Temple 4and Hye Jeong Choi 5
1REAL Prevention LLC, 130 Pearl Brook Dr, Clifton, NJ 07013, USA
2School of Nursing & Health Studies, University of Miami, P.O. Box 248153, Coral Gables, FL 33124, USA;
aen16@miami.edu
3College of Health and Human Development, Pennsylvania State University, 219 HHD Building,
University Park, PA 16802, USA; dmc397@psu.edu
4Center for Violence Prevention, Betty and Rose Pfefferbaum Chair in Child Mass Trauma and Resilience,
School of Behavioral Health Sciences, University of Texas Health Science Center, Houston, TX 77030, USA;
jeffrey.r.temple@uth.tmc.edu
5College of Health Sciences, University of Missouri, 318 Clark Hall, Columbia, MO 65211, USA
*Correspondence: hechtpsu@gmail.com; Tel.: +1-814-360-1893
Abstract: Background/Objectives: Test efficacy of the social emotional learning (SEL)-based Mighty
Girls program, a program culturally tailored for English-speaking Hispanic/Latino girls in seventh
grade comprised of classroom sessions and a virtual reality computer game. We hypothesized that
the curriculum would decrease risky sexual behaviors in a program that can be used as part of a
comprehensive sex education curriculum or as a stand-alone program. Methods: A randomized group
trial was conducted in which 22 low-income, predominately Hispanic schools within the Miami-
Dade County Public School System were randomly assigned to intervention (consented
n= 335)
and control (consented n= 217) conditions. All study activities occurred after school. Primary
outcome measures were resistance self-efficacy, acceptance of dating violence, sexual intentions,
and sexual behavior. Assessments occurred at baseline, immediately post-intervention, 3-, 12-, and
24-months post-intervention. Changes in outcomes from baseline to 24 months were modeled using
multi-level models to account for nesting of students within schools with full information maximum
likelihood to account for missing data and baseline school attendance and enrollment in free and
reduced lunch as covariates. Analyses are also controlled for multiple testing. Results: The program
had a significant effect on reducing acceptance of dating violence at 24 months post-intervention
(
estimate = −0.083
,p
≤
0.05), but no effect on resistance self-efficacy, sexual intentions, or sexual
behavior (p
≥
0.58). Conclusion: Study findings demonstrate that a social emotional learning (SEL)
curriculum can impact sexual behaviors such as susceptibility to dating violence. Low baseline levels
for sexual intentions and behaviors as well as a high baseline of efficacy may have impacted findings
for the other outcomes.
Keywords: social emotional learning; sexual health; dating violence; Hispanic adolescents; prevention
of risky sexual behaviors
1. Introduction
While the overall teen birth rate in the United States has declined recently, rates among
Hispanic adolescents remain high [
1
]. Indeed, national data indicate that proportionally
more Hispanic teens give birth compared to their same-aged, non-Hispanic peers [
1
]. To
effectively reduce teen pregnancies among this population, programs should be gender-
specific and culturally reflective of Hispanic/Latino cultural values [2,3].
Early adolescence is a critically important developmental period characterized by
rapid behavioral and biological growth, as well as the time when many health behav-
iors are acquired [
4
] and gender norms about behavior within romantic relationships are
Children 2024,11, 1331. https://doi.org/10.3390/children11111331 https://www.mdpi.com/journal/children
Children 2024,11, 1331 2 of 17
formed [
5
,
6
], Unsurprisingly, this is also the time when youth begin to experiment with
risky behaviors. Thus, interventions delivered during this timeframe have the potential to
make positive and enduring changes. Those that target multiple problematic behaviors—
for example, unsafe sexual behavior and dating violence—may be especially promising [
7
].
Because attitudes supporting violence have been linked to both being a victim and perpetra-
tor of dating violence [
8
,
9
], as well as risky sexual behavior [
9
,
10
] and teen pregnancy [
11
],
pregnancy prevention programs should address attitudes about dating violence.
There are racial/ethnic disparities in these risks, with Black and Latina women more
at risk for teen pregnancy [
12
] as well as for contracting HIV [
13
] and other sexually trans-
mitted infections [
14
], which can lead to infertility and even death [
12
]. These experiences
are then related to later problems and challenges, including failure to complete obtaining a
high school diploma, incarceration, and unemployment [
12
]. Clearly, Latinx adolescents
are at higher risk for negative sexual experiences. These risks appear heightened among
recent Latina immigrants to Miami, where the study was conducted [15].
Another major developmental concern for early adolescents, regardless of culture/ethnic
background, is peer acceptance [
16
], which makes tailoring interventions to address this
concern critical to their effectiveness [
17
]. Indeed, peer influence has been shown to
be a more powerful contributor to risky behavior than the buffering benefit of parental
monitoring [
18
]. Unsurprisingly, research has found a consistent link between peer pressure,
peer norms, and risky sexual behavior [19].
To guide intervention development, Mighty Girls is based on social emotional learning
theory/SEL [
20
–
23
]. SEL argues that a focus on basic competences is needed to promote
health development. The approach has been demonstrated to have wide-ranging effects
across several domains by teaching basic skills (i.e., self-awareness, self-management,
responsible decision making, relationship skills, and social awareness) across contexts (i.e.,
classrooms, schools, families and caregivers, and communities). The lessons teach the basic
skills and are taught in an after-school context to facilitate the transfer of learning from
schools to family and community settings. The implementation was highly interactive,
with discussions, role plays, and games, along with some didactic content.
While pregnancy prevention interventions developed for Hispanic/Latino early ado-
lescents typically include some of these components, they rarely address peer pressure
directly or attitudes about dating violence. In contrast, the Mighty Girls program aims to
build relationally competent communication skills for resisting peer pressure and media
literacy skills for challenging messages supporting power imbalance and violence in male–
female relationships [
23
]. The aim of this study is to develop an effective curriculum to
curtail the participation of adolescents in risky sexual behaviors. Our primary hypotheses
were that girls receiving Mighty Girls would report higher resistance self-efficacy, lower sex-
ual intentions, less sexual behavior, and less positive attitudes about relationship violence
at follow-up, relative to girls in the control condition. If successful, the project will provide
an evidence-based intervention targeting risky sexual behavior among Latinas that has
wide applicability due to its focus on sexual pressure resistance rather than more controver-
sial aspects of sex education. Moreover, it will demonstrate the utility of narrative-based
videogaming in health promotion and provide further support for the SEL approach.
2. Materials and Methods
2.1. Study Design
A group randomized trial design was implemented in 22 public schools within the
Miami-Dade County Public School system (MDCPS) over a 2-year period. Assessments
occurred at baseline, immediately after intervention in treatment schools, and at 3-, 12-, and
24-months post-intervention exposure. All study procedures were pre-approved by the Uni-
versity of Miami Institutional Review Board and the MDCPS Research
Review Committee.
Children 2024,11, 1331 3 of 17
2.2. Randomization Process
Schools were randomized within enrollment year (“cohort”) and geographic area
(upper, central, and lower Miami-Dade County). An odd number of schools within a cohort
resulted in random assignment of the “extra” school to the intervention condition. The
first year of enrollment (cohort 1) was comprised of five intervention and four control
schools. The second year of enrollment (cohort 2) was comprised of eight intervention
and five control schools. There were originally seven control schools, but two schools
withdrew after randomization (immediately prior to starting recruitment while blind to
study conditions). In total, the study consisted of 13 intervention and nine control schools.
2.2.1. Recruitment and Study Participants
Schools
Schools with Hispanic enrollment greater than or equal to 60% and reduced/free
lunch participation greater than or equal to 50% were eligible. Principals were free to agree
or disagree with their school’s involvement in the study. Participating schools (C = 22)
represent 58% of all MDCPS schools meeting inclusion criteria. PI, study staff, and schools
were blind to school assignment until after study staff obtained parental consent.
Students
Student recruitment occurred during in-school assemblies of girls meeting study
criteria, during which study staff distributed information packets containing parental
consent forms. Staff also made these packets available in the school office. After receiving
parental consent forms, study staff contacted parents of potential participants by phone
to verify and review elements of informed consent, address any questions, and schedule
participants. Finally, study staff obtained student informed assent.
The sample was comprised of 552 girls with consent and assent who also met partici-
pant inclusion criteria (at least one Hispanic or Brazilian parent or grandparent, enrolled
in Grade 7, and English speaking) and did not meet any exclusion criteria (English for
Speakers of Other Languages; developmental delay; severe hearing, vision, or speech
impairment). Exclusion criteria were evaluated as part of obtaining parental consent over
the phone using a checklist protocol.
Enrollment rates, defined as the percentage of eligible girls with parental consent who
agreed to participate in the study, were 23% and 29% for Cohorts 1 and 2, respectively. The
percentage of girls whose parents consented but refused to provide assent was 4% and did
not vary by study condition (p= 0.40). Figure 1illustrates the progression of participants
through study activities.
2.3. Data Collection
A survey containing study measures was completed electronically using a link pro-
vided by staff (all responses saved to a password-protected server using 128-bit SSL en-
cryption). Survey software (LimeSurvey version 2.05) tailored item wording according
to sexual orientation. Completion time varied (20–45 min) depending on reading ability
and additional follow-up questions if vaginal intercourse or substance use was reported.
Participants completed surveys after school, in a school classroom, using a Chromebook
provided by study staff. For times 1 to 4, participants were supervised by staff. At time
5, students completed the survey at home as they were no longer at study schools (i.e.,
transitioned to high school). Participants unable to attend any of the initial four assessments
were offered the opportunity to complete the survey at home. Staff assisting with at-home
survey completion followed a procedure in which they first established the participant
had privacy and sufficient time to complete the survey and provided a callback number
in case there were difficulties. Once established, participants were provided a web link
for the survey and a numerical code to access survey items (code used to link participant
data across time points). Whether at school or at home, each assessment began with staff
stressing the voluntary nature of participation. Participants received a $15 gift card and
Children 2024,11, 1331 4 of 17
a community service hour for each survey they completed. This amount was deemed
necessary to incentivize survey responses without being coercive. Dates of the study are
2015–2018. See Consort Diagram, Figure 1).
Children 2024, 11, x FOR PEER REVIEW 4 of 18
Figure 1. Consort diagram indicating flow of trial participants through study activities.
Note: Participants were enrolled in two cohorts. The first cohort participated in the study from Fall
2015 through Fall 2017. The second cohort participated in the study from Fall 2016 through Fall 2018.
Figure 1. Consort diagram indicating flow of trial participants through study activities.
Children 2024,11, 1331 5 of 17
2.4. Study Conditions
2.4.1. Intervention Condition
Mighty Girls is a program for 12–13-year-old Hispanic/Latino girls consisting of six
45-min classroom sessions over a 2-week period followed by playing a 15-min virtual reality
game (DRAMA-RAMA) twice during the two succeeding weeks. Sessions taught SEL
skills to groups of 16–25 by female facilitators and co-facilitators selected for (a) experience
teaching or working with groups of children in schools, summer camps, or afterschool
programs, and (b) responses to questions about strategies for handling challenging class-
room situations. All received an additional 28 h of training provided by the PI and project
manager. The sessions were highly interactive and focused on risks and choices as well as
communication and relationship skills. After completing the sessions, the girls participated
in DRAMA-RAMA, a live virtual reality simulation of early adolescent social encounters
involving peer pressure. It immerses the player in a situation derived from formative
interviews in which they make choices requiring them to use the SEL skills taught during
the classroom sessions. During the game, participants interacted with avatars “puppeted”
by “interactors (female actors trained in use of technology and interactive performance)
(see Figure 2). Game play occurred in small, private rooms, each equipped with a 42-inch
TV monitor. A small Web camera was clipped on top of each monitor (see Figure 3).
Games were played in English, one person at a time, for 15 min. DRAMA-RAMA used
an object-oriented graphics rendering engine game engine. Audio interaction and video
observation of the participant were delivered over the Internet via Skype. The Geppetto
interface allowed the game to be played by talking (no keyboard or similar input required).
The simulation was played twice as part of intervention, and twice 3 months later as
intervention booster. See Table 1for a more complete program description.
Children 2024, 11, x FOR PEER REVIEW 6 of 18
Figure 2. Interactor puppeting the avatars during Drama-Rama.
Figure 3. Girls playing Drama-Rama. Note: The young girl in Figure 3 is an actor, not a study par-
ticipant.
A total of 335 girls were enrolled in the intervention condition and 217 in the control
condition. However, 8 of those in the intervention condition did not complete a survey at
any study time point. All girls enrolled in the control condition completed at least one T1-
T5 survey.
Table 1. Mighty Girls program components.
Lesson Content
1. Choices and Results Personal goals, everyday choices, and consequences
related to goals
2. What’s Risky Risky behaviors, consequences of risky behaviors,
personal risk
Figure 2. Interactor puppeting the avatars during Drama-Rama.
Children 2024,11, 1331 6 of 17
Children 2024, 11, x FOR PEER REVIEW 6 of 18
Figure 2. Interactor puppeting the avatars during Drama-Rama.
Figure 3. Girls playing Drama-Rama. Note: The young girl in Figure 3 is an actor, not a study par-
ticipant.
A total of 335 girls were enrolled in the intervention condition and 217 in the control
condition. However, 8 of those in the intervention condition did not complete a survey at
any study time point. All girls enrolled in the control condition completed at least one T1-
T5 survey.
Table 1. Mighty Girls program components.
Lesson Content
1. Choices and Results Personal goals, everyday choices, and consequences
related to goals
2. What’s Risky Risky behaviors, consequences of risky behaviors,
personal risk
Figure 3. Girls playing Drama-Rama. Note: The young girl in Figure 3is an actor, not a study
participant.
Table 1. Mighty Girls program components.
Lesson Content
1. Choices and Results
Personal goals, everyday choices, and consequences related to goals
2. What’s Risky Risky behaviors, consequences of risky behaviors, personal risk
3. Avoid Skill a
Define the avoid skill, 3 methods to avoid pressure, and the Mighty
Girl way of avoiding
4. Refuse Skill a
Aggressive, passive, and assertive communication, matching verbal
and nonverbal communication, and the Mighty Girl way of refusing
5. Media Influences
Define media, understanding media influence, positive and
negative media messages about girls, and the difference between
media teens and “real” teens.
6. Wrap-up and Review Results-based choices, practice Mighty Girl skills
DRAMA-RAMA b
Live videogame simulation of peer pressure with participants
deciding whether or not to engage in risky behaviors, practice of
Mighty Girls skills
a
Adapted from keepin’ it REAL [
23
];
b
DRAMA-RAMA was played twice as part of the intervention, and twice 3
months later as an intervention booster.
A total of 335 girls were enrolled in the intervention condition and 217 in the control
condition. However, 8 of those in the intervention condition did not complete a survey
at any study time point. All girls enrolled in the control condition completed at least one
T1–T5 survey.
Mighty Girls was co-created by two of the authors with 2 of 6 classrooms and assis-
tance, with sessions adapted from the evidence-based substance use prevention program,
keepin’ it REAL (kiR) [
23
], that is listed as an SEL-certified program [
24
]. The acronym
REAL stands for refuse, explain, avoid, leave, the names for four different resistance
communication strategies that were identified in formative research and taught in kiR.
Focused on building developing SEL skills, in particular self-awareness through goal set-
ting, self-management through planning, responsible decision making, social awareness of
Children 2024,11, 1331 7 of 17
peer pressure, and relationally competent peer resistance skills, kiR has proven effective,
particularly among Hispanic/Latinos, as noted in the 2016 Surgeon General’s report on
addiction [21].
While lesson content is based on SEL, the communication competence model (CCM) [
19
]
and social cognitive theory (SCT) [
25
] guided the curriculum design of kiR and Mighty
Girls. The CCM specifies how verbal and non-verbal communication skills can be used to
communicate messages that leave the message recipient feeling positive about the message
sender, even when the message is a refusal to go along with a request. CCM provides
insights into how the SEL skills should be taught. It stresses that communication must be
relationally competent—that is, effective for both parties. Teaching these communication
skills is also congruent with Hispanic/Latino cultural values such as Personalismo and
Simpatía, which emphasize protecting social harmony by having warm and conflict-free
interactions [
26
,
27
]. SCT stresses that teaching SEL skills should be performed through
models that increase a sense of efficacy. Combining the CCM and the SCT resulted in a
behavior change program that builds resistance self-efficacy through modeling of specific
communication skills.
Consequently, the program was also “adolescent-centered” because it addressed an
outcome valued in early adolescence [
28
]. Program development was also informed by
interactive focus groups [
29
] and consultation with various key informants and stakehold-
ers, including Hispanic/Latino parents, Hispanic/Latino female high school students, a
Planned Parenthood staff member, an Orlando public school’s public school administrator
responsible for HIV prevention education, an active member of the National Organiza-
tion of Women, and a Hispanic Seventh Day Adventist minister. Thus, the program was
not only “adolescent-centered,” but also informed by formative work and responsive to
stakeholder concerns.
2.4.2. Control Condition
Participants received a website address for downloading Science Valley, a virtual
reality game in which the goal is to solve “challenges” related to renewable energy,
agriculture/sustainable living, and robotics. There were no classroom sessions for this
condition, and game play occurred concurrently with intervention participants playing
DRAMA-RAMA.
2.5. Instrument
The study survey included items assessing demographic and cultural characteristics,
contamination, and potential outcomes (resistance self-efficacy, acceptance of dating vio-
lence, sexual intentions, sexual behaviors; Table 2). Resistance self-efficacy operationalizes
our application of efficacy from Social Cognitive Theory, with the relevant belief being that
they are capable of resisting sexual pressure. Acceptance of dating violence and sexual
behaviors are the most salient outcomes of a sexual pressure resistance intervention, opera-
tionalizing the key behaviors we anticipated would result from the intervention. Sexual
intentions were included as a predictor of these behaviors. All items were previously
used with this age group in feasibility testing [
19
], work conducted with substance use
prevention by study team members, or the CDC’s YRBS (Youth Risk Behavior Survey [
30
]).
Feasibility testing items were pre-tested to ensure comprehension. See Table 2for further
detail about the measures.
2.6. Statistical Analyses
Descriptive analyses were conducted to characterize the sample and assess demo-
graphic variables. Baseline analyses comparing the control and intervention conditions
were conducted. Multi-level outcome models evaluated the change in outcomes between
the control and intervention conditions from baseline to T5. Analyses accounted for stu-
dents nested within school and employed robust standard errors. Table 3presents the
baseline equivalences, and Table 4presents the means across waves.
Children 2024,11, 1331 8 of 17
Table 2. Outcome measures and respective Cronbach’s alpha statistic.
Construct Number
of Items Sample Item Response Options Source T1-T5 Cronbach’s
Alpha Values
Resistance
Self-Efficacy 16
Stop someone who is
pressuring you to do sexual
things without making
them angry.
5-point scale with labels for
mid- and endpoints only
(0 = Not at all sure I can do
this; 2 = Moderately; Sure, I
can do this; 4 = Completely;
Sure I can do this
Adapted
from DiIorio
et al. [19,31]
0.92–0.94
Sexual
Intentions 4
I would have sex now if a
girl aI cared about
pressured me to have sex.
4-point scale (0 = No,
definitely not; 3 = Yes,
definitely yes)
Kirby et al.
[32]0.87–0.92
Sexual
Behavior 11
Had a boy btouch you
below the waist, underneath
your clothing.
5-point scale (0 = Never; to
4 = 10 times or more) [33] 0.85–0.93
Acceptance of
Dating
Violence
6
A girl who makes her
boyfriend jealous on
purpose deserves to be hit c
4-point scale (1 = strongly
disagree; 4 = strongly agree)
Foshee et al.
(1998) [34]0.80–0.84
a
Same-sex orientation wording. Item tailored for opposite sex orientation using “boy.” Item tailored for bisexual
orientation, and unsure or refused to answer responses using “boy or girl.”.
b
Opposite sex orientation wording.
c
This item is not tailored for sexual orientation.
Table 3. Baseline demographics for intervention and control condition participants 1.
Variable Intervention
(n= 325)
Control
(n= 217)
Age in years
Mean (SD a)
Median (range)
12.31 yrs. (0.68)
12 (11–15)
12.36 yrs. (0.78)
12 (11–15)
Acculturation Score
Mean (SD)
Median (range)
3.73 (0.72)
3.75 (1.5–5.3)
3.73 (0.64)
3.75 (1.5–5)
Born in US 71% 70%
Generation b
Mean (SD)
Median (range)
2.03 (0.88)
2 (1–4)
1.9 (0.78)
2 (1–4)
Country of Origin c
Cuba 40% 43%
Mexico 6% <1%
Dominican Republic 3% 3%
Puerto Rico 2% <1%
Central America 12% 15%
South America 10% 9%
Multiple Countries 27% 29%
Reduced/Free Lunch 86% 82%
Started Menses 62% 58%
Note.
1
Group means compared using the Student’s t-test. Group proportions (i.e., percentages) were compared
using the chi-square test. No significant group differences were observed in any demographic characteristics
(
p≥0.10
).
a
SD = standard deviation.
b
Generation is defined as follows: 1 = Participant not born in the US (first
generation). 2 = Participant is first person in family born in the US (second generation). 3 = Participant and at least
one parent born in the US (third generation). 4 = Participant and at least one parent and at least one grandparent
born in the US (fourth generation).
c
Chi-square analysis compared the proportion of Cuban Americans in each
group against all other countries of origin combined.
Children 2024,11, 1331 9 of 17
Table 4. Means across waves.
Variable Wave
Control Intervention
Mean Std Dev Mean Std Dev
Acceptance of
Dating Violence 1 1.27 0.39 1.33 0.39
2 1.22 0.35 1.25 0.32
3 1.24 0.41 1.20 0.32
4 1.17 0.28 1.17 0.25
5 1.14 0.28 1.13 0.23
Resistance
Self-Efficacy 1 4.48 0.79 4.48 0.68
2 4.56 0.70 4.57 0.69
3 4.58 0.71 4.60 0.62
4 4.63 0.55 4.59 0.56
5 4.64 0.58 4.67 0.48
Sexual Intentions 1 0.10 0.34 0.09 0.30
2 0.09 0.26 0.13 0.44
3 0.10 0.33 0.10 0.34
4 0.08 0.28 0.14 0.37
5 0.15 0.39 0.17 0.37
Sexual Behavior 1 0.13 0.36 0.22 0.47
2 0.15 0.34 0.21 0.44
3 0.22 0.53 0.24 0.50
4 0.26 0.49 0.44 0.78
5 0.45 0.73 0.60 0.82
3. Results
3.1. Baseline Equivalence
No significant baseline differences were found between groups in study outcome
variables (Table 3). Only three participants reported having had sexual intercourse at
baseline, and all these participants were in the control condition. However, 101 intervention
participants (30.2%) were missing on this item at baseline as compared to 37 (17%) in the
control condition (p< 0.0001). See Table 3for demographics.
3.2. Contamination
Control participant responses to close-ended questions indicated fairly minimal con-
tamination occurring at T2 (n= 3), T4 (n= 5), and T5 (n= 9). Of the total number reporting
hearing about the intervention (n= 17), only nine reported hearing about relationally
competent resistance communication concepts, with three at T5 reporting hearing about
all five listed intervention concepts. Higher reporting at T5 is consistent with participants
from control and intervention schools mixing in high school. Including/excluding these
participants in tests of study hypotheses had no effect on results of analyses.
3.3. Implementation Quality
3.3.1. Dosage Received
Thirty-five percent (n= 116) of intervention participants attended all six classroom
sessions, with 75% (n= 252) attending at least three. A majority (79%; n= 263) played
DRAMA-RAMA twice, with half (51%; n= 171) receiving the booster (played four times).
Participation was impacted by competing activities (e.g., doctor’s appointments, music
lessons, school clubs or sports) and developmental concerns (e.g., friend(s) not wanting to
attend, wanting to watch the boys’ basketball practice).
3.3.2. Adherence
Facilitator classroom session adherence scores ranged from 3.6–5.7 (mean = 3.37,
SD = 0.47
) with a maximum of 6. Scores were comprised of points (0–2) for following
Children 2024,11, 1331 10 of 17
lesson content, incorporating review of previously learned skills or content into session,
and final session participant group performances and discussion in final session. Adding
content not part of the curriculum resulted in a deduction of 1–2 points. DRAMA-RAMA
adherence (percent of required game story plot elements and dialogue elements delivered)
ranged from 66–100%, with 81% of sessions achieving 90–100% adherence.
3.3.3. Quality of Program Delivery
Classroom session delivery quality scores ranged from 2 to 6 (mean = 5.60, SD = 0.34)
with a maximum of 6. Scores were comprised of points (0–1) assigned for (a) making the
session “Mighty Moment” (relationally competent communication kinesthetic learning
activity developed by the project manager), (b) classroom climate having a “call and
response” feel, (c) classroom management, (d) eagerness of class response, (e) friendly/fun
feel, and (f) facilitator’s modeling of Mighty (relationally competent) communication.
DRAMA-RAMA delivery quality was rated excellent on three different dimensions of
quality using a 3-point scale (poor, average, excellent) for a large majority of game play
sessions: (1) inter-actor performance energy (100%); (2) naturalness of conversation (98.5%);
and (3) distinct and consistent game character characterization (95.7%).
3.4. Attrition
Overall, 20% of enrolled participants (n= 111) dropped or were lost to follow-up
(i.e., phone disconnected, never returned phone calls, no longer enrolled at study school,
hospitalized, or in foster care). Intervention condition participants (22%; n= 73) were
not more likely to drop out of the study or be lost to follow-up than control condition
participants (18%, n= 38); p> 0.19), and attrition did not alter comparability of intervention
and control conditions at any time point (i.e., no differences with respect to receiving
a reduced/free lunch, acculturation, being born in the United States, being of Cuban
origin, generation, or puberty characteristics (p
≥
0.65)). Attrition was also not greater in
the intervention condition than in the control condition during the intervention period
(
p> 0.27
). Reasons for attrition, including lost follow-up, did not differ significantly by
study condition (p> 0.36) apart from providing no reason for dropping from the study
(intervention: 7%, n= 24; control: 2%, n= 4; Fisher’s exact, p< 0.005).
3.5. Missing Data
There was no association between missing an assessment and receiving a reduced/free
lunch, acculturation, being born in the United States, being of Cuban origin, immigrant
generation, or menses onset (p> 0.47). However, intervention condition participants
were more likely than those in the control condition to have missing data (p< 0.01). We
experienced a significant differential rate of missingness amongst participants, with 44.31%
of intervention participants compared to 30.88% of control participants missing data at the
end of the study (estimate = 0.28, p< 0.001).
3.6. Tests of Study Hypotheses
The effect of the intervention on change in outcomes from baseline to 24 months was
modeled using multi-level models to account for the nested structure of the data (students
within schools), and baseline school attendance and enrollment in free and reduced lunch
were included as baseline covariates. Proc Mixed in SAS 9.4 was employed to test models
and assess intervention effects [
35
]. Full information maximum likelihood procedures were
employed to account for missing data [36].
No significant baseline differences were found between groups on sexual intentions
(estimate = 0.041, p= 0.58), sexual behavior (estimate =
−
0.01, p= 0.99) or self-efficacy
(estimate = 0.014, p= 0.80). A significant effect was found for acceptance of dating violence,
with intervention condition participants less likely to support dating violence than those in
the control condition 24-months post-intervention (estimate =
−
0.083, p
≤
0.05). Figure 4
displays the amount of change in study outcome variables observed in each study condition
Children 2024,11, 1331 11 of 17
over a 24-month period. The Y-axis is the difference of means of the study variables between
time points. Additional statistical information is provided in Appendix A.
Children 2024, 11, x FOR PEER REVIEW 11 of 18
3.5. Missing Data
There was no association between missing an assessment and receiving a re-
duced/free lunch, acculturation, being born in the United States, being of Cuban origin,
immigrant generation, or menses onset (p > 0.47). However, intervention condition partic-
ipants were more likely than those in the control condition to have missing data (p < 0.01).
We experienced a significant differential rate of missingness amongst participants, with
44.31% of intervention participants compared to 30.88% of control participants missing
data at the end of the study (estimate = 0.28, p < 0.001).
3.6. Tests of Study Hypotheses
The effect of the intervention on change in outcomes from baseline to 24 months was
modeled using multi-level models to account for the nested structure of the data (students
within schools), and baseline school aendance and enrollment in free and reduced lunch
were included as baseline covariates. Proc Mixed in SAS 9.4 was employed to test models
and assess intervention effects [35]. Full information maximum likelihood procedures
were employed to account for missing data [36].
No significant baseline differences were found between groups on sexual intentions
(estimate = 0.041, p = 0.58), sexual behavior (estimate = −0.01, p = 0.99) or self-efficacy (esti-
mate = 0.014, p = 0.80). A significant effect was found for acceptance of dating violence,
with intervention condition participants less likely to support dating violence than those
in the control condition 24-months post-intervention (estimate = −0.083, p < 0.05). Figure 4
displays the amount of change in study outcome variables observed in each study condi-
tion over a 24-month period. The Y-axis is the difference of means of the study variables
between time points. Additional statistical information is provided in Appendix A.
Figure 4. Change scores (difference between baseline and T5) in outcomes.
4. Discussion
The aim of this study was to develop an effective program for reducing risky sexual
behaviors among adolescents. Contrary to expectations and results from a previous feasi-
bility trial [19], we were unable to detect program effects on self-efficacy and sexual be-
havior. Thus, the prediction about efficacy from social cognitive theory is not confirmed.
Figure 4. Change scores (difference between baseline and T5) in outcomes.
4. Discussion
The aim of this study was to develop an effective program for reducing risky sex-
ual behaviors among adolescents. Contrary to expectations and results from a previous
feasibility trial [
19
], we were unable to detect program effects on self-efficacy and sexual
behavior. Thus, the prediction about efficacy from social cognitive theory is not confirmed.
We suspect the analysis was underpowered due to ceiling effects in our self-efficacy data
resulting from the observed skewing of responses (i.e., the majority of responses were
strongly agree or agree) and low baseline rates of sexual intercourse, which may reflect
under-reporting of sexual behavior. The rates of sexual behavior we obtained were lower
than those reported by Hispanic female participants in Miami-Dade County YRBS middle
and high school surveys [
30
] (Centers for Disease Control and Prevention [CDC], 2020),
which may only partly be explained by when we collected data (fall semester for time
4 and time 5) vs. YRBS (spring semester). It is also possible that the voluntary nature
of the sample resulted in a study population that was less risky than the general school
population of Hispanic girls enrolled in Miami-Dade County public schools.
We did find evidence for a delayed effect of the Mighty Girls program on attitudes
about relationship violence, suggesting that the program may influence teen dating vi-
olence as participants grow into later adolescence and emerging adulthood. This effect
emerged at the final time point (fall semester of 9th grade), possibly reflecting the increased
dating pressure and expectations often witnessed with the transition to high school. Inter-
estingly, only one intervention session addressed violence in relationships, and this was
in the context of images in the media about male–female relationships. However, using
relationally competent communication skills to assert your opinion, refuse a request, and
resist peer pressure was stressed in every session. Thus, this finding argues for the potential
power of teaching relationally competent communication skills in combination with media
literacy skills regarding the portrayal of male–female relationships in advertising. It also
supports the application of the Communication Competence Model to the intervention
by demonstrating a protective effect through the communication and relationship skills
imparted during the lessons.
Children 2024,11, 1331 12 of 17
Our study findings generate a critical question for future research on the impact of SEL
interventions that promote relationally competent communication on conflict resolution,
involvement in violent relationships, and safer sex negotiation. Building personal (decision
making), communication (resistance strategies), and relationship (conflict reduction) skills
should foster healthy relationships by promoting knowledge and efficacy. Theoretically,
increasing these SEL competencies should manifest themselves in healthier lives. However,
interventions do not exist in isolation, and focusing on changing individual behavior is
limited by the outer layers of the social ecology. It is incredibly difficult to overcome the
unhealthy environments and models (i.e., social media content) that often surround our
youth (including current participants). Thus, we believe the next logical step is a larger
study that (1) allows subgroup analyses to determine if Mighty Girls and similar programs
work better for some groups than others and (2) considers additional components that
address multiple levels of social influence in the SEL model (e.g., relationship, family,
community). Further, effective interventions delivered during early adolescence may be
particularly beneficial to shape how people relate prior to the development of unhealthy
behavioral patterns. Such interventions can have a growing positive impact across the life
span because skills and ideas learned in early adolescence may become enduring patterns
that influence future behavior [
12
]. However, demonstrating such effects requires a longer
framework for longitudinal analyses that follow youth as they age into adulthood.
A second critical question is the viability of virtual reality games such as DRAMA-
RAMA. Not only is the technology expensive, costs and implementation complexity are
exacerbated by the “human element” in the form of interactants introduced to maximize
realism and interactivity. VR headsets are now available to replace the TV, but space is
still required for privacy, and the authors are not convinced that the affordances of AI
will adequately provide the nuance needed during game play that was gained from the
interactants puppeting the avatars. It is possible that a simpler technology may suffice.
For example, a narrative writing app would allow participants to construct stories about
pressure resistance and SEL skill enactment that not only reinforces skill development but
teaches narrative writing.
Limitations
There are several limitations of this study. First, all outcome data are self-reported,
which likely resulted in the underreporting of sexual behavior through issues like social
desirability and the capacity for self-perception. We attempted to mitigate this concern
by employing computer-delivered assessments, which have been shown to encourage
greater self-disclosure by adolescents relative to paper/pencil assessments [
37
]. In addition,
intervention activities did not describe, call out, or focus on specific sexual behaviors except
for “kissing a boy you did not know very well” in the first DRAMA-RAMA game. Finally,
staff were not involved with collecting survey data at any school in which they delivered
intervention sessions. Nevertheless, the observations of staff involved with assessments
indicated dislike or discomfort with the sexual behavior items (e.g., participants complained
they were inappropriate).
Second, missing data occurred, with significant differential missingness between study
conditions (i.e., in the intervention condition, approximately 25% of girls did not attend
half the sessions), which likely contributed to difficulty detecting intervention effects. It
is also possible that the various causes of missing data (e.g., technical issues, missing
assessments, missing/refusing to answer a survey item, attrition) may have unrecognized
effects on study outcomes. However, we employed recommended methods for handling
missing data [
38
], which enabled us to adhere to intention to treat (TIT) principles by
retaining all participants in the analysis within the groups to which they were originally
randomized, irrespective of their participation in intervention activities, completion of data
collection sessions, or attrition [
39
]. Adhering to ITT principles maintains comparability of
study conditions obtained through randomization and eliminates bias [
40
] and produces
conservative estimates of treatment effects [
41
]. This argues for the trustworthiness of study
Children 2024,11, 1331 13 of 17
results. It is, perhaps, endemic to designs implemented outside of school that youth data
will include high degrees of missingness. The “intent-to-treat” model, while standard, does
reduce power. Designs where data are collected under more controlled circumstances (e.g.,
during school) would decrease this limitation.
Related to missing data was the apparent hesitancy of many parents to provide signed
consent. Topics related to sex are controversial and have become increasingly politicized
in the United States and particularly in the state of Florida. It is likely the most religious
and/or conservative parents were overrepresented in the non-consented group, providing
a potential bias in the sample.
Third, the effect size for even the significant finding for attitudes about relationship
violence was small. That said, a finding for females this young puts them on a path to
healthier sexual relations, so even a small effect can have practical significance. The low
baseline levels of risk behaviors in this study limited effects and demonstrated the potential
for even greater impact on groups with higher levels of risk behaviors.
Fourth, the construct validity of the measures was not tested. Although previously
validated measures were used with low-income Hispanic girls in Florida and other parts
of the US, this may still be an issue regarding the accuracy of the measurement and
the findings.
Fifth, the goal was to assess the change between baseline and final measurement. As
a result, longitudinal analyses are not reported. Earlier posttests occurred during middle
school, a time of limited sexual exploration. Since long-term effects are the strongest
indicator of program outcomes, we chose to compare the T5 assessment, approximately
2 years after the intervention and during high school, to baseline.
5. Conclusions
This study addressed the question of how we can improve the sexual health of ado-
lescents through effective prevention interventions. The Mighty Teens program was de-
veloped based on social emotional learning theory and tested in a group randomized trial
to examine its efficacy. While all the expected results were not realized, the Mighty Girls
intervention demonstrated effects on sexual behavior in a previous randomized trial in
Orlando [
19
] and effects on attitudes about relational violence in the present study. Addi-
tionally, Mighty Girls has the advantage of being a sexual pressure resistance intervention
that focuses on SEL skill building rather than reproductive health content. This allows
the program to surmount the political and ideological issues surrounding sex education
that create barriers to program implementation [
42
]. While not meant as a comprehensive
sex education curriculum, study findings argue for Mighty Girls having demonstrated the
potential to impact issues of sexual health, including teen pregnancy and STIs, through its
effect on attitudes about teen dating violence.
Author Contributions: Conceptualization, M.L.H., A.E.N. and J.R.T.; formal analysis. D.M.C. and
H.J.C.; investigation, J.R.T.; methodology, M.L.H., A.E.N., D.M.C., J.R.T. and H.J.C.; project adminis-
tration, A.E.N.; supervision, A.E.N.; writing—original draft, M.L.H. and A.E.N.; writing—review and
editing, J.R.T. and H.J.C. All authors have read and agreed to the published version of
the manuscript.
Funding: Funding for this study was provided by the University of Miami School of Nursing
and Health Studies and National Institutes of Health (US)/National Institute of Nursing Research
(R01NR014851). Clinical Trial Registration at ClinicalTrials.gov (NCT02578147).
Institutional Review Board Statement: This study was approved by the University of Miami
Institutional Review Board. Approval code: 20140697; Approval date: 31 August 2015.
Informed Consent Statement: Active parental consent and youth assent were obtained prior to
participating in this research.
Children 2024,11, 1331 14 of 17
Data Availability Statement: The data presented in this study are available on request from the
second author. The data are not publicly available due to restrictions privacy.
Acknowledgments: The authors would like to thank the principals, school staff, and students who
participated in the study and the JUEGA! field research team whose efforts helped ensure project
success. They also wish to acknowledge Roxana Delcampo Thalasinos, as co-creator of the Mighty
Girls program.
Conflicts of Interest: Authors Michael L Hecht was employed by REAL Prevention LLC. The
remaining authors declare that the research was conducted in the absence of any commercial or
financial relationships that could be construed as a potential conflict of interest.
Appendix A
Table A1. Baseline and 24-month follow-up means for self-efficacy, sexual intentions, sexual behavior,
and acceptance of dating violence by study condition.
Variable Intervention Condition Control Condition
T1
Oct-Dec 2015 T5 T1 T5
Resistance Self-Efficacy
4.48
(0.68)
Range: 1–5
(n= 251)
4.47
(0.49)
Range: 2–5
(n= 261)
4.48
(0.73)
Range: 1–5
(n= 187)
4.64
(0.528)
Range: 1–5
(n= 177)
Sexual Intentions
0.09
(0.30)
Range: 0–3
(n= 244)
0.17
(0.37)
Range: 0–1.80
(n= 260)
0.10
(0.34)
Range: 0–3
(n= 185)
0.15
(0.39)
Range: 0–3
(n= 179)
Sexual Behavior
0.22
(0.47)
Range: 0–2.58
(n= 236)
0.60
(0.82)
Range: 0–4
(n= 259)
0.13
(0.36)
Range: 0–3.83
(n= 181)
0.45
(0.73)
Range: 0–4
(n= 177)
Acceptance of Dating
Violence
1.33
(0.39)
Range: 1–3
(n= 260)
1.13
(0.23)
Range: 1–2.56
(n= 261)
1.27
(0.39)
Range: 1–4
(n= 190)
1.14
(0.28)
Range: 1–2.72
(n= 179)
Table A2. Resistance self-efficacy model.
a. Solution for Fixed Effects:
Effect Estimate Standard
Error df t Value Pr > |t|
Intercept −0.7871 1.3310 19 −0.59 0.5612
Study Condition −0.00069 0.08768 314 −0.01 0.9937
Student Qualifies for Free/Reduced Price
Lunch 0.03232 0.06698 314 0.48 0.6298
School’s Average Daily Attendance 0.9443 1.3965 314 0.68 0.4994
b. Type 3 Tests of Fixed Effects
Effect df F Value Pr > F
Study Condition 1, 314 0.00 0.9937
Student Qualifies for Free/Reduced Price
Lunch 1, 314 0.23 0.6298
School’s Average Daily Attendance 1, 314 0.46 0.4994
Children 2024,11, 1331 15 of 17
Table A3. Sexual intentions model.
a. Solution for Fixed Effects
Effect Estimate Standard
Error df t Value Pr > |t|
Intercept −0.01458 0.7507 19 −0.02 0.9847
Study Condition 0.01232 0.05155 315 0.24 0.8113
Student Qualifies for Free/Reduced Price
Lunch −0.02878 0.03467 315 −0.83 0.4070
School’s Average Daily Attendance 0.1537 0.7897 315 0.19 0.8458
b. Type 3 Tests of Fixed Effects
Effect df F Value Pr > F
Study Condition 1, 315 0.06 0.8113
Student Qualifies for Free/Reduced Price
Lunch 1, 315 0.69 0.4070
School’s Average Daily Attendance 1, 315 0.04 0.8458
Table A4. Sexual behavior model.
a. Solution for Fixed Effects
Effect Estimate Standard
Error df t Value Pr > |t|
Intercept 0.02674 0.9952 19 0.03 0.9788
Study Condition 0.04129 0.06870 297 0.60 0.5483
Student Qualifies for Free/Reduced Price
Lunch 0.01131 0.04688 297 0.24 0.8096
School’s Average Daily Attendance 0.2036 1.0470 297 0.19 0.8459
b. Type 3 Tests of Fixed Effects
Effect df F Value Pr > F
Study Condition 1, 297 0.36 0.5483
Student Qualifies for Free/Reduced Price
Lunch 1, 297 0.06 0.8096
School’s Average Daily Attendance 1, 297 0.04 0.8459
Table A5. Acceptance of dating violence model.
a. Solution for Fixed Effects
Effect Estimate Standard
Error df t Value Pr > |t|
Intercept −0.5550 0.6356 19 −0.87 0.3934
Study Condition −0.08281 0.04160 324 −1.99 0.0473
Student Qualifies for Free/Reduced Price
Lunch 0.04165 0.03150 324 1.32 0.1871
School’s Average Daily Attendance 0.3125 0.6672 324 0.47 0.6399
b. Type 3 Tests of Fixed Effects
Effect df F Value Pr > F
Study Condition 1, 324 3.96 0.0473
Student Qualifies for Free/Reduced Price
Lunch 1, 324 1.75 0.1871
School’s Average Daily Attendance 1, 324 0.22 0.6399
References
1.
Martin, J.A.; Hamilton, B.E.; Osterman, M.J.K. Births in the United States, 2017. NCHS Data Brief. 2018, 1–8. Available online:
https://www.cdc.gov/nchs/data/databriefs/db318.pdf (accessed on 9 November 2020).
2.
Albarracin, D.; Gillette, J.C.; Earl, A.N.; Glasman, L.R.; Durantini, M.R.; Ho, M.-H. A Test of Major Assumptions about Behavior
Change: A Comprehensive Look at the Effects of Passive and Active HIV-Prevention Interventions Since the Beginning of the
Epidemic. Psychol. Bull. 2005,131, 856–897. [CrossRef] [PubMed]
Children 2024,11, 1331 16 of 17
3.
Albarracin, J.; Albarracin, D.; Durantini, M. Effects of HIV-prevention interventions for samples with higher and lower percents
of Latinos and Latin Americans: A meta-analysis of change in condom use and knowledge. AIDS Behav. 2008,12, 521–543.
[CrossRef] [PubMed]
4.
Blum, R.W.; Mmari, K.; Moreau, C. It Begins at 10: How Gender Expectations Shape Early Adolescence Around the World. J.
Adolesc. Health 2017,61 (Suppl. 4), S3–S4. [CrossRef]
5.
Lane, C.; Brundage, C.L.; Kreinin, T. Why We Must Invest in Early Adolescence: Early Intervention, Lasting Impact. J. Adolesc.
Health 2017,61 (Suppl. 4), S10–S11. [CrossRef]
6.
Sabol, B.M.; Kernsmith, P.D.; Hicks, M.R.; Smith-Darden, J.P. Attitudes about aggression and perpetration of Adolescent Dating
Aggression: A cross-lagged panel model. J. Adolesc. 2020,83, 100–111. [CrossRef]
7.
Shorey, R.C.; Fite, P.J.; Choi, H.; Cohen, J.R.; Stuart, G.L.; Temple, J.R. Dating violence and substance use as longitudinal predictors
of adolescents’ risky sexual behavior. Prev. Sci. 2015,16, 853–861. [CrossRef] [PubMed]
8.
Ybarra, M.L.; Langhinrichsen, R.J. Linkages between violence-associated attitudes and psychological, physical, and sexual dating
abuse perpetration and victimization among male and female adolescents. Aggress. Behav. 2019,45, 622–634. [CrossRef]
9.
Demissie, Z.; Clayton, H.B.; Vivolo-Kantor, A.M.; Estefan, L.F. Sexual teen dating violence victimization: Associations with sexual
risk behaviors among US high school students. Violence Vict. 2018,33, 964–980. [CrossRef] [PubMed]
10.
Fedina, L.; Howard, D.E.; Wang, M.Q.; Murray, K. Teen dating violence victimization, perpetration, and sexual health correlates
among urban, low-income, ethnic, and racial minority youth. Int. Q. Community Health Educ. 2016,37, 3–12. [CrossRef]
11.
Sebastian, C.; Viding, E.; Williams, K.D.; Blakemore, S.-J. Social brain development and the affective consequences of ostracism in
adolescence. Brain Cogn. 2010,72, 134–145. [CrossRef]
12.
Evans, R.; Widman, L.; Stokes, M.; Javidi, H.; Hope, E.; Brasileiro, J. Sexual Health Programs for Latinx Adolescents: A
Meta-analysis. Pediatrics 2020,146, e20193572. [CrossRef] [PubMed]
13.
Pahl, K.; Williams, S.Z.; Capasso, A.; Fuller Lewis, C.; Lekas, H.M. A longitudinal pathway from ethnic-racial discrimination to
sexual risk behaviors among Black women and Latinas: Ethnic-racial identity exploration as a protective factor. Soc. Sci. Med.
2023,316, 115061. [CrossRef] [PubMed]
14.
Sales, J.M.; Newton-Levinson, A.; Swartzendruber, A.L. Racial Disparities in STIs Among Adolescents in the USA. In Sexually
Transmitted Infections in Adolescence and Young Adulthood; Hussen, S., Ed.; Springer: Cham, Switzerland, 2020. [CrossRef]
15.
Ertl, M.M.; Dillon, F.R.; Cabrera Tineo, Y.A.; Verile, M.; Jurkowski, J.M.; De La Rosa, M. Sexual risk during initial months in US
among Latina young adults. AIDS Care 2017,30, 353–360. [CrossRef]
16.
Pedlow, C.T.; Carey, M.P. Developmentally Appropriate Sexual Risk Reduction Interventions for Adolescents: Rationale, Review
of Interventions, and Recommendations for Research and Practice. Ann. Behav. Med. 2004,27, 172–184. [CrossRef]
17.
Wang, B.; Deveaux, L.; Lunn, S.; Dinaj-Koci, V.; Li, X.; Stanton, B. The influence of sensation-seeking and parental and peer
influences in early adolescence on risk involvement through middle adolescence: A structural equation modeling analysis. Youth
Soc. 2016,48, 220–241. [CrossRef] [PubMed]
18.
Brechwald, W.A.; Prinstein, M.J. Beyond Homophily: A Decade of Advances in Understanding Peer Influence Processes. J. Res.
Adolesc. 2011,21, 166–179. [CrossRef]
19.
Norris, A.E.; Hughes, C.; Hecht, M.; Peragallo, N.; Nickerson, D. Randomized trial of a peer resistance skill-building game for
Hispanic early adolescent girls. Nurs. Res. 2013,62, 25–35. [CrossRef]
20.
Bogard, T.; Nenonene, R. Social Emotional Learning. In Trends and Issues in Instructional Design and Technology, 5th ed.; Reiser,
R.A., Carr-Chellman, A.A., Dempsey, J.V., Eds.; Routledge: London, UK, 2024. [CrossRef]
21.
Durlak, J.A.; Domitrovich, C.E.; Weissberg, R.P.; Gullotta, T.P. (Eds.) Handbook of Social and Emotional Learning; Guilford: New
York, NY, USA, 2015.
22.
Shi, J.; Cheung, A.C.K. Effective Components of Social Emotional Learning Programs: A Meta-analysis. J. Youth Adolesc. 2024,53,
755–771. [CrossRef]
23.
Hecht, M.L.; Shin, Y.; Pettigrew, J.; Miller-Day, M.; Krieger, J.L. Designed cultural adaptation and delivery quality in rural
substance use prevention: An effectiveness trial for the keepin’ it REAL curriculum. Prev. Sci. 2018,19, 1008–1018. [CrossRef]
24.
Substance Abuse and Mental Health Services Administration (US); Office of the Surgeon General (US). Facing Addiction in America:
The Surgeon General’s Report on Alcohol, Drugs, and Health; US Department of Health and Human Services: Washington, DC, USA,
2016. Available online: https://www.ncbi.nlm.nih.gov/books/NBK424857/pdf/Bookshelf_NBK424857.pdf (accessed on 26
April 2021). [PubMed]
25.
Bandura, A. Toward a psychology of human agency: Pathways and reflections. Perspect. Psychol. Sci. 2018,13, 130–136. [CrossRef]
26.
Galanti, G.A. The Hispanic family and male-female relationships: An overview. J. Transcult. Nurs. 2003,14, 180–185. [CrossRef]
[PubMed]
27.
Triandis, H.C.; Marin, G.; Lisansky, J.; Betancourt, H. Simpatía as a cultural script of Hispanics. J. Pers. Soc. Psychol. 1984,47, 1363.
[CrossRef]
28.
Krieger, J.L.; Coveleski, S.; Hecht, M.L.; Miller-Day, M.; Graham, J.W.; Pettigrew, J.; Kootsikas, A. From kids, through kids, to kids:
Examining the social influence strategies used by adolescents to promote prevention among peers. J. Health Commun. 2013,28,
683–695. [CrossRef]
29.
Norris, A.E.; Aroian, K.J.; Warren, S.; Wirth, J. Interactive performance and focus groups with adolescents: The power of play. Res.
Nurs. Health 2012,35, 671–679. [CrossRef] [PubMed]
Children 2024,11, 1331 17 of 17
30.
Centers for Disease Control and Prevention (CDC). 1995–2017 Middle School Youth Risk Behavior Survey Data. 2018. Available
online: http://nccd.cdc.gov/youthonline (accessed on 19 April 2020).
31.
DiIorio, C.; Resnicow, K.; Thomas, S.; Wang, D.T.; Dudley, W.N.; Van Marter, D.F.; Lipana, J. Keepin’ it REAL! Program description
and results of baseline assessment. Health Edu Behav. 2002,29, 104–123. [CrossRef]
32.
Kirby, D.; Korpi, M.; Barth, R.P.; Cagampang, H.H. The Impact of the Postponing Sexual Involvement Curriculum Among Youths
in California. Fam. Plan. Perspect. 1997,29, 100–108. [CrossRef]
33.
Barnett, J.E. Evaluating “Baby Think It Over” infant simulators: A comparison group study. Adolescence 2006,41, 103–110.
[PubMed]
34.
Foshee, V.A.; Bauman, K.E.; Arriaga, X.B.; Helms, R.W.; Koch, G.G.; Linder, G.F. An evaluation of Safe Dates, an adolescent dating
violence prevention program. Am. J. Public. Health 1998,88, 45–50. [CrossRef]
35.
Singer, J.D. Using SAS PROC MIXED to fit multilevel models, hierarchical models, and individual growth models. J. Educ. Behav.
Stat. 1998,23, 323–355. [CrossRef]
36.
Larsen, R. Missing Data Imputation versus Full Information Maximum Likelihood with Second-Level Dependencies. Struct. Equ.
Model. 2011,18, 649–662. [CrossRef]
37.
Turner, C.F.; Ku, L.; Rogers, S.M.; Lindberg, L.D.; Pleck, J.H.; Sonenstein, F.L. Adolescent sexual behavior, drug use, and violence:
Increased reporting with computer survey technology. Science 1998,280, 867–873. [CrossRef] [PubMed]
38.
Graham, J.W. Missing data analysis: Making it work in the real world. Annu. Rev. Psychol. 2009,60, 549–576. [CrossRef] [PubMed]
39.
Teesson, M.; Newton, N.C.; Slade, T.; Carragher, N.; Barrett, E.L.; Champion, K.E.; Kelly, E.V.; Nair, N.K.; Stapinski, L.A.; Conrod,
P.J. Combined universal and selective prevention for adolescent alcohol use: A cluster randomized controlled trial. Psychol. Med.
2017,47, 1761. [CrossRef] [PubMed]
40.
Ranganathan, P.; Pramesh, C.S.; Aggarwal, R. Common pitfalls in statistical analysis: Intention-to-treat versus per-protocol
analysis. Perspect. Clin. Res. 2016,7, 144–146. [CrossRef]
41. Gupta, S.K. Intention-to-treat concept: A review. Perspect. Clin. Res. 2011,2, 109–112. [CrossRef]
42.
Wiley, D.C.; Plesons, M.; Chandra-Mouli, V.; Ortega, M. Managing Sex Education Controversy Deep in the Heart of Texas: A Case
Study of the North East Independent School District (NEISD). Am. J. Sex. Educ. 2020,15, 53–81. [CrossRef]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.
