The Impact of Relationship Education on Adolescents of Diverse Backgrounds
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Family Relations, 56 (July 2007), 291–303. Blackwell Publishing.
Copyright 2007 by the National Council on Family Relations.
The Impact of Relationship Education on Adolescents
of Diverse Backgrounds
Francesca Adler-Baeder Jennifer L. Kerpelman David G. Schramm Brian Higginbotham Amber Paulk*
Abstract: Adolescent-focused marriage education is a relatively uncharted research area. Using a quasi-experimental
design, this study examined the effectiveness of an adapted version of the curriculum entitled, Love U2: Increasing
Your Relationship Smarts with an economically, geographically, and racially diverse sample of 340 high school stu-
dents. Findings suggest that participants showed increases in 5 dimensions of their relationship knowledge, includ-
ing their ability to identify unhealthy relationship patterns. Participants also had more realistic beliefs about
relationships/marriages and reported lower levels of verbal aggression use at postprogram compared to controls.
Moreover, these findings existed across race, household income, and family structure type, with all participating stu-
dents benefiting in similar ways. Implications for future programming and research are discussed.
Key Words: adolescent romantic relationships, dating violence, family life education and related areas, marriage,
program evaluation.
As part of the Federal Deficit Reduction Act of Yet, there is very little programmatic impact data
2005 (Senate Bill 1932), the U.S. government that provide support for these voluntary or man-
recently appropriated $500 million over the next dated program offerings.
5 years for marriage-strengthening activities. This We conducted a thorough search of several social
money will support new and existing educational science databases (e.g., PsychINFO, Social Science
programs designed to cultivate the skills necessary to Abstracts) and found only two published studies of
form and sustain healthy marriages. The legislation relationship or marriage education programs for
specifically authorizes relationship education in high youth (i.e., Gardner, 2001; Gardner, Giese, &
schools as one of the eight allowable activities. The Parrot, 2004). Because these programs contain
impact of youth-focused relationship and marriage information on basic relational skills for current dat-
education, however, is a relatively uncharted course ing relationships and for later adult relationships
of study (Pearson, 2000). It is commonly noted that and marriages, we searched for both the terms ‘‘rela-
relational attitudes and behaviors develop in adoles- tionship education’’ and ‘‘marriage education.’’ The
cence, that adolescents are interested in marriage terms are used interchangeably in this paper.
and in relationship/marriage education (Silliman & Participant characteristics of those studies limit
Schumm, 2004), and that youth should receive rela- the generalizability of the findings. Gardner et al.
tionship skills and ‘‘marriage readiness’’ courses (2004) evaluated a marriage education program,
while in school (e.g., Administration for Children Connections: Relationships and Marriage, with a pri-
and Families, 2005; Brotherson & Duncan, 2004; marily urban sample of high school adolescents,
Hawkins, Carroll, Doherty, & Willoughby, 2004). providing preliminary evidence that suggested
*Francesca Adler-Baeder is Associate Professor and Extension Specialist in the Department of Human Development and Family Studies at Auburn University, 203
Spidle Hall, Auburn, AL 36849 (fadlerbaeder@auburn.edu). Jennifer L. Kerpelman is Professor and Extension Specialist in the Department of Human Development
and Family Studies at Auburn University, 203 Spidle Hall, Auburn, AL 36849 (kerpejl@auburn.edu). David G. Schramm is a doctoral candidate in the Department of
Human Development and Family Studies at Auburn University, 203 Spidle Hall, Auburn, AL 36849 (schradg@auburn.edu). Brian Higginbotham is Assistant Profes-
sor and Extension Specialist in the Department of Family, Consumer, and Human Development at Utah State University, 2705 Old Main Hill, Logan, UT 84322-
2705 (brianh@ext.usu.edu). Amber Paulk is a doctoral student in the Department of Human Development and Family Studies at Auburn University, 203 Spidle Hall,
Auburn, AL 36849 (paulkal@auburn.edu).292 Family Relations Volume 56, Number 3 July 2007
a research-based relationship education curriculum of aggression, and research suggests that experienc-
can positively influence attitudes, knowledge, and ing dating violence in adolescence increases the like-
behaviors. Although the outcome of Gardner’s stud- lihood of experiencing future relationship violence
ies supported the value of relationship education, (Close, 2005; Wekerle & Wolfe, 1999; Wolfe,
results were based on a sample of White and His- 2006).
panic adolescents in an urban setting. The current Although there is risk of dating violence, it is
study adds to this limited literature base by more important to recognize that dating is a normal part
broadly considering the issue to include economic, of life, can be quite positive, and has developmental
geographical, and racial and ethnic diversity. We purposes (e.g., mate selection; Paul & White,
contend that regardless of race and ethnicity, income 1990). Often, adolescents do not realize that conflict
level, geographic location, and family structure, stu- in romantic relationships is inevitable and believe
dents will experience positive changes in relational conflict is negative because they use maladaptive
skills and thinking after participation in a relation- strategies to cope (Shulman, 2003). In fact, conflict
ship education program. Our study sample included and negotiation in adolescent romantic relationships
rural and urban (i.e., geographically diverse) African help maintain the important balance of emotional
American and White adolescents from diverse eco- closeness and individuality. When relationships are
nomic backgrounds and who lived in diverse family healthy, adolescent dating builds self-competence
structures. and self-worth, provides opportunities to practice
conflict management and negotiate trust, and allows
adolescents to learn lessons regarding how to form,
The Importance of Relationship Education for Youth
maintain, and end relationships (Collins, 2003).
It has been argued that an ideal time for marriage These skills and this knowledge have important
and relationship education is the high school years implications for later relational quality and stability.
(Gardner, 2001; Gardner et al., 2004; Silliman & If dating experience offers a primary way for ado-
Schumm, 2004). It is during this time period that lescents to learn how to become a socially competent
many adolescents begin dating and forming other dating partner, then providing relationship educa-
bonds and relationships with friends. In fact, tion during adolescence appears to be an optimal
research suggests that the quality of adolescent time period for instruction. Providing relationship
romantic relationships is one of the strongest predic- education is especially important given that adoles-
tors of adolescent well-being indicators, including cents tend to hold idealistic, rather than realistic,
self-esteem, depression, and suicide attempts and beliefs about romantic relationships (Montgomery,
completions (Brent et al., 1993; Joyner & Udry, 2005) and could benefit from the knowledge gained
2000). through accurate information provided in a relation-
Unfortunately, as adolescents enter the dating ships course. In sum, providing education that
years, many of them will experience relationship builds knowledge and skills among adolescents
problems, including relationship violence. Perpetra- regarding healthy relationships can help prevent
tion estimates of any type of adolescent physical dat- unhealthy dating relationships now and unhealthy
ing violence range from 11 to 41% with 4 – 14% and unstable relationships in the future.
of adolescents reportedly using forms of violence For adolescents from low-income families, edu-
that are likely to result in serious physical injury cation and training in healthy relationship behaviors
(Grunbaum et al., 2001). A recent study demon- and attitudes may be especially beneficial. Low-
strated that physical aggression in adolescent dating income and minority youth may be particularly
relationships was relatively stable over a 3-month disadvantaged as they are more likely to have experi-
period. This finding held across gender and was true enced family structures that are prone to unhealthy
when adolescents were reporting perpetration or vic- patterns of interaction and relationship instability
timization. It also was found that psychological (Ooms & Wilson, 2004). Similarly, where eco-
aggression (i.e., control and jealousy) was signifi- nomic conditions are poor, prospects for quality of
cantly associated with physical aggression both con- life, healthy human development, and family stabil-
currently and 3 months later (O’Leary & Slep, ity are poor as well (Ooms & Wilson). Among low-
2003). Clearly, many relationships in the adolescent income youth, African Americans are at the greatest
years are at risk for both verbal and physical forms risk for unhealthy and unstable future relationships,Relationship Education Adler-Baeder et al. 293
are the least likely to marry, and when they do, are the The RS adapted curriculum consists of 12,
most likely to divorce (e.g., Teachman, Tedrow, & 60 – 90 min lessons that encompass four units (see
Crowder, 2000). Research has yet to document if Table 1 for more details). Unit 1 (Lessons 1 – 4)
and how the suggested benefits of relationship edu- covers the concepts of maturity, values, infatuation,
cation are manifest for racially, economically, and and love; Unit 2 (Lessons 5 – 7) addresses dating
geographically diverse samples. The current study processes and strategies; Unit 3 (Lessons 8 – 9)
was the first, to our knowledge, that empirically focuses on relationship problems and identifying un-
assessed the efficacy of youth-focused relationship healthy relationships; and Unit 4 (Lessons 10 – 12)
education with a large proportion of rural and urban helps students learn and practice relationship/
African Americans in the sample. marriage skills. Included in all the lessons are spe-
cific activities aimed at getting the adolescents to
process how the information applies to their per-
Curriculum Overview
sonal relationships and current life experiences.
Our study focused on the evaluation of an adapted
version of the curriculum Love U2: Increasing Your Purpose and Hypotheses
Relationship Smarts (RS adapted). RS adapted covers
material that is consistent with a developmental per- The overarching program goals for participants in the
spective of romantic relationship formation during RS program focused on reducing the risk of maltreat-
adolescence (Furman & Shaffer, 2003) and is ment in dating relationships, increasing knowledge of
designed for schools, youth agencies, clubs, and
faith-based organizations that work with youth in Table 1. The RS Adapted Lessons
Grades 8 through 12. The curriculum is research
based, contains validated content, and incorporates Unit Lessons Content Description
materials and activities that are sensitive to diverse 1 1–4 Exploring the social, emotional,
backgrounds of youth. Specific studies are cited and mental dimensions of maturity,
throughout the curriculum to support program con- reflecting on values and determining
tent. The research-based information includes a com- the ones that are personally important,
bination of didactic material and experiential gaining knowledge about the nature
activities designed to enhance adolescents’ relation- of infatuation, and exploring the
ship knowledge and skills. dimensions of mature love.
The original Love U2: Increasing Your Relation- 2 5 – 7 Guidelines for ‘‘dating smart,’’
ship Smarts (Pearson, 2004) curriculum was chosen learning low-risk dating strategies,
because of features judged to be especially appropri- and gaining knowledge related
ate for lower resource, racially diverse youth. These to what healthy and unhealthy
include a limited amount of didactic material, com- relationships look like.
mon ‘‘teen language,’’ materials that show diversity, 3 8 – 9 Identification and discussion of
and language that assumes teens are living in diverse behaviors that demonstrate abuse,
family structures. Compared to the original version, recognizing the different types and
RS adapted includes increased activities and interac- warning signs of abuse, and learning
tion, more practice of specific relationship skills, when and how to end a dating
greater incorporation of video illustrations, specific relationship and move on.
discussions of future adult relationships and mar- 4 10 – 12 Understanding the practices associated
riage, and elimination of material designed for with healthy stable marriages,
a younger audience (e.g., experiencing a ‘‘crush’’). understanding the importance of
Thus, the participatory nature of RS adapted was commitment and positive
designed not only to be highly effective with adoles- communication skills, learning to
cents but also to be consistent with recommenda- manage conflict and understanding
tions for working with low-income individuals, such the role of forgiveness in relationships,
as actively engaging participants, addressing specific and determining values associated
challenges, and building on solid program content with financial management.
(Ooms & Wilson, 2004).294 Family Relations Volume 56, Number 3 July 2007
the characteristics of healthy relationships, including By agreeing to participate in the RS adapted pro-
modifying beliefs to align with research-based infor- gram, each teacher agreed to administer a pretest
mation, and the promotion of future healthy couple and posttest survey both to the students in the class
and marital relationships as they transition into adult- who received the RS adapted curriculum and to
hood and parenthood. This evaluation of RS adapted another class they taught who did not receive the
examined changes over time in select areas of stu- curriculum (control group). The FCS classes in
dents’ beliefs, knowledge, and behaviors. We tested which RS adapted was offered varied according to
several hypotheses that related directly to the goals of the kinds of courses being taught by the participat-
this specific curriculum. First, we hypothesized that ing teachers. Courses taught within the FCS Family
students participating in the classes receiving the area included Family Dynamics, Human Dynamics,
RS adapted curriculum would experience (a) increases Parent and Child Dynamics, Family Wellness, and
in knowledge about healthy and unhealthy relation- Life Connections. Teachers were free to determine
ships and (b) increases in their understanding of the which class would receive RS adapted and which
skills needed to facilitate healthy relationships. would serve as the control. At posttest (approxi-
Because part of the program goals involved the mately 2 months after the pretest), the students who
reduction of the risk of maltreatment in dating rela- participated in the RS adapted classes also were asked
tionships, we also hypothesized that compared to to provide subjective feedback by writing any com-
control students, the participants in the RS adapted ments they had regarding the classes (e.g., what they
group would report (c) greater increases in their use enjoyed about the class and what they would change
of reasoning strategies during interpersonal conflicts about the class). Prior to participation in the study,
and (d) greater decreases in their use of verbal and both participant and control group students and
physical aggression in their close relationships. parents were required to complete student assent
Finally, we hypothesized that (e) RS adapted partici- and parental consent forms, respectively. Nearly all
pants would show a positive trend toward healthier program participants completed surveys as class time
and more realistic beliefs about relationships com- was set aside for completion. The teachers collected
pared to control participants. Along with the all the forms and surveys and mailed them to the
hypotheses related to the curriculum, we also research team. Identification numbers were used to
hypothesized that the improvements from pre- to match student questionnaires across Time 1 and
postprogram assessment would hold across race, Time 2 data collections.
income level, and family structure, thus providing A total of 465 students completed and returned
evidence that the RS adapted program may prove questionnaires. Of those, 340 were retained for the
helpful for adolescent participants from varying final sample for analyses; 235 students participated
backgrounds. in the experimental (i.e., RS adapted) group and 105
students in the control group. Student surveys were
excluded (n ¼ 125) if they appeared to systemati-
Method cally mark large portions of the questionnaire (i.e.,
used visible patterns in response markings). This
was judged by independent raters; reliability was
Participants and Procedures
a ¼ .96. The data were not included if only a pretest
RS adapted was evaluated in nine public high schools or posttest (but not both) was completed or because
located throughout Alabama (Grades 9 – 12). the research team was unable to match up correct
Teachers were invited to participate after being pretests with posttests because of students failing to
nominated by local extension agents or responding properly complete the survey.
to listserv advertisements for the Relationship Of the sample of 340 students, 46% were African
Smarts facilitator training. The Alabama Child American, 50% were White, 1% was Hispanic/
Abuse and Neglect Board funded faculty from Latino, and the other 3% included Asian, Native
Auburn University to train Family and Consumer American, and those in the ‘‘Other’’ category. The
Science (FCS) teachers to deliver the curriculum. All participants were, on average, 16.1 years old (range
the nine participating teachers were female; seven 14 – 19 years, SD ¼ 1.17), 74% were female, and
indicated their race as White, one as African Ameri- 26% were male (typical of Alabama FCS classes).
can, and one as Hispanic. Forty percent of students resided in a nuclear family,Relationship Education Adler-Baeder et al. 295
whereas 30% were in single-parent households, 23% (1) retrospective pretest measure that assessed
were in stepfamilies, and the remaining 7% indi- changes in knowledge on specific curriculum learn-
cated ‘‘Other.’’ Half of the participants reported ing objectives outlined in each of the RS adapted les-
having experienced at least one parental divorce. sons. That is, each lesson had specific learning
Twenty percent of students reported household objectives that were translated into a question on the
incomes of less than $20,000 a year, another 25% student evaluation. These questions were designed
reported their parents earning $20,000 – $40,000, to tap students’ perceptions, understanding, and
with 12% between $40,000 and $60,000 a year, knowledge of the curriculum elements before and
and 42% reporting a household income of over after participating in the program. This self-reported
$60,000 per year. For the nearly one third of partici- measure of change avoids pretest sensitivity and
pants in single-parent households, approximately response shift bias that may result from pretest over-
65% reported a household income of less than estimation or underestimation (Pratt, McGuigan, &
$40,000. Thus, although approximately one fifth of Katzev, 2000). Pratt et al. (2000) demonstrated that
the participants could be considered living in pov- the post 1 retrospective pretest produced a more
erty, there are many more who might be considered valid assessment of their program outcomes than did
‘‘low resource.’’ Studies frequently use the 200% of the traditional pretest-posttest evaluation method.
poverty or below demarcation (approximately They argued that participation in the program
$40,000) when categorizing ‘‘low-income’’ families served to shift the program recipients’ frame of
(e.g., Karney, Garvan, & Thomas, 2003). However, knowledge about what they knew before receiving
caution should be taken when interpreting the the program that would not have been captured
income data as 124 participants did not provide any using the traditional evaluation method. Thus, tra-
information related to either parent’s income. These ditional methods alone may fail to capture the
students may not have known this information. change that has actually occurred as the result of an
Analyses of demographic variables indicated that intervention. Further, the post 1 retrospective pre-
the group participating in the RS adapted program test has been shown to be an effective measure of
and the control group did not differ on income, change that is less susceptible to social desirability
race, or family structure, but there were slight differ- than are other retrospective self-report methods.
ences in age and gender. Those who participated Lam and Bengo (2003) examined the effectiveness
in the RS adapted program were slightly younger of differing retrospective methods for assessing ele-
(M ¼ 16.06 years old, SD ¼ 1.15) than those not mentary teachers’ self-reported changes in instruc-
in the program (M ¼ 16.45 years old, SD ¼ 1.19), tional practices. They found that, although all the
t(326) ¼ 22.78, p , .01. Further, although the methods detected change, the post 1 retrospective
majority of the students were female, those who pretest method, compared to methods having partic-
participated in the RS adapted program were sig- ipants indicating postknowledge only or postknowl-
nificantly more likely to be female (M ¼ 1.79, edge plus estimates of the perceived amount of
SD ¼ .40) than those not in the program (M ¼ change, was the most conservative estimate of
1.60, SD ¼ .49), t(329) ¼ 3.77, p , .001. change. The authors recommended use of the post 1
retrospective pretest design because it is less sensitive
to socially desirable response bias than are the other
Measures retrospective methods.
For the current study, participants responded to
The survey assessed demographic variables, knowl- a total of 36 questions that were tied to specific
edge and awareness of key concepts related to learning objectives of the RS adapted course. These
healthy relationships, the frequency of behaviors questions, using the post 1 retrospective pretest
used during interpersonal conflicts, and beliefs asso- design, addressed what they understood now that
ciated with healthy relationships. they have been exposed to the content and, concur-
rently, what they understood with respect to the
same items and corresponding scales before they
Relationship Knowledge Scale
were exposed to RS adapted.
Post plus retrospective approach. Adolescents who Factor analysis. It was expected that there would
participated in the program completed a post plus be some concept overlap across the lessons; therefore,296 Family Relations Volume 56, Number 3 July 2007
the 36 items were subjected to principal components or a best friend (female). The instrument consists of
analysis (PCA) with varimax rotation and the pres- three subscales derived from instrument develop-
ence of five components with eigenvalues exceeding ment studies (Straus et al., 1996). Nevertheless, the
1 (ranging from 1.21 to 13.61) was revealed, responses from our study’s participants to the
explaining a total of 50.4% of the variance, with 18 items were subjected to a PCA and the presence
Component 1 contributing 11.54%, Component 2 of four components was revealed with eigenvalues
contributing 11.35%, Component 3 contributing exceeding 1. However, the answer item ‘‘Cried’’
10.17%, Component 4 contributing 9.91%, and from the verbal aggression scale loaded quite high
Component 5 contributing 7.40% of the variance. (.90) on one component, whereas another item
Of the five factors that emerged from the explor- ‘‘Stomped out of the room or house or yard’’ cross-
atory factor analysis, three tapped knowledge of loaded with this component, in addition to the ver-
healthy/unhealthy relationships: attraction/mature bal aggression component. Therefore, these two
love (seven items; e.g., ‘‘my knowledge of social, items were removed. PCA with varimax rotation
emotional, and mental dimensions of maturity’’), was carried out, and the three-factor solution
expectations and behaviors (seven items; e.g., ‘‘my explained a total of 68.39% of the variance, with
understanding of how my expectations affect my Component 1 (physical aggression) contributing
behavior’’), and unhealthy relationships (five items; 35.64%, Component 2 (verbal aggression) contrib-
e.g., ‘‘my ability to recognize the signs of an uting 23.12%, and Component 3 (reasoning) con-
unhealthy relationship’’). The remaining two factors tributing 9.63% of the variance. The reasoning
tapped knowledge of relationship skills needed to subscale consisted of three items, with possible
facilitate healthy relationships: communication skills scores ranging from 0 to 18, with higher scores indi-
and (nine items; e.g., ‘‘my understanding of the cating using reasoning more frequently as a conflict
attack/defend method of communication’’) smart tactic. The final verbal aggression subscale had four
dating strategies (seven items; e.g., ‘‘my knowledge items, with possible scores ranging from 0 to 24,
of Ôlow-risk’ dating strategies’’). Cronbach’s coeffi- with higher scores indicating using verbal aggression
cient as ranged from .81 to .88 for the five factors, more frequently to settle differences. Lastly, the
indicating good subscale reliability. Factor scores physical aggression subscale consisted of six items,
were retained and used in subsequent analyses. For with possible scores ranging from 0 to 36, with
the ‘‘Before’’ questions, answers included four higher scores indicating the use of violent aggression
options formatted in a Likert scale that ranged from more frequently to resolve conflicts. Coefficient
Was Poor (1) to Was Excellent (4). ‘‘After’’ response alphas at posttest for this study were .54 for the rea-
options were written in a corresponding manner soning subscale, .85 for verbal aggression, and .94
(i.e., Is Poor [1]; Is Excellent [4]). for physical aggression. Although the reasoning scale
has a lower alpha than normally considered accept-
able, similar alpha levels have been reported by
Conflict Tactics Scales
Straus (1990) and Gardner et al. (2004). Therefore,
A revised form of the Conflict Tactics Scales (CTS2; we decided to retain this measure in analyses.
Straus, Hamby, Boney-McCoy, & Sugarman, 1996)
was used to assess both frequency of use and type of
Relationship Beliefs
behaviors used (i.e., reasoning, verbal, physical)
when dealing with conflict. Both the control and the This scale consisted of 17 items (Gardner, 2001)
participant groups of students completed this scale that were answered on a 4-point scale ranging from
both at Time 1 (pretest) and at Time 2 (posttest). YES! strongly agree (1) to NO! strongly disagree (4).
Students indicated how frequently they had used We chose this measure as the 17 items related well
each of the 18 tactics in settling differences within with the overall objectives of the curriculum. Both
the past 2 months, with items arranged on a 7-point the control and the participant groups of students
Likert scale ranging from never (0) to more than 20 completed this scale both at Time 1 (pretest) and at
times (6). Before students completed these questions, Time 2 (posttest). For purposes of this study, we
they were asked to indicate who they were thinking again subjected all the items to a PCA with varimax
about as they answered the questions. Their choices rotation and six components were revealed with
included a boyfriend, girlfriend, best friend (male), eigenvalues exceeding 1 (ranging from 2.42 to 1.10).Relationship Education Adler-Baeder et al. 297
However, an inspection of the screeplot revealed beginning to the end of the classes, participants
a clear break after the third component so it was completed a post 1 retrospective pretest measure
decided to retain three components for further anal- that assessed changes in knowledge on specific cur-
yses. These three components explained 15.15, riculum topics. Paired-samples t tests were con-
10.67, and 8.85% of the variance, respectively, with ducted on the five relationship knowledge subscales
the three-factor solution explaining a total of (i.e., attraction/mature love, expectations and behav-
34.67% of the variance. We labeled the three factors iors, unhealthy relationships, communication skills,
in the following manner: aggression beliefs (two smart dating strategies) to evaluate the impact of the
items; e.g., ‘‘in today’s society, slapping a spouse or curriculum on relationship knowledge from Time 1
dating partner is understandable under some cir- to Time 2. Results are presented in Table 2 for all
cumstances’’), faulty relationship beliefs (five items; relationship knowledge subscales. Overall, there was
e.g., ‘‘most long-term, happy marriages never have a statistically significant increase in perceived knowl-
conflict’’), and realistic relationship beliefs (four edge for all five relationship knowledge subscales
items; e.g., ‘‘your communication style is affected by scores from retrospective pretest to posttest scores.
your family members’ style of communication’’; in The magnitude of this change was quite substantial;
addition, several items were eliminated because of the mean difference exceeded the standard deviation
cross-loading). High scores on the aggression beliefs for each subscale.
subscale indicated higher levels of disagreement with
the aggression beliefs. Similarly, higher scores on the Pre- and Posttests
faulty relationship beliefs subscale indicated greater
disagreement with the faulty relationship belief Table 3 provides an overview of the means and stan-
items. Questions for the realistic relationship beliefs dard deviations for the three subscales from the
subscale were recoded so higher scores indicate Conflict Tactics Scales and the three relationship
higher levels of agreement with the healthy relation- beliefs subscales for both the control and the experi-
ship belief statement. mental groups. Overall, mean pretest scores for both
groups were very similar. The following sections
provide further findings pertaining to the specific
Results hypotheses of the study.
Conflict Behaviors
Post 1 Retrospective Pretest
One of the overall goals of the curriculum involved
Relationship knowledge. To test the first two reducing the risk of maltreatment in dating relation-
hypotheses, we focused only on those students who ships by encouraging students to utilize more rea-
received the RS adapted course. In order to deter- soning strategies in their interpersonal conflicts and
mine whether the adolescents’ perceived knowledge less verbal and physically aggressive strategies in
of healthy/unhealthy relationships and skills needed their close relationships. To address the third and
to facilitate healthy relationships changed from the fourth hypotheses, the individual Conflict Tactics
Table 2. Paired-Samples t Tests for Relationship Knowledge Subscales for Experimental Group
Pretest Posttest
Relationship Knowledge Mean
Subscale M SD M SD Difference SEM df t
Attraction/mature love 2.68 .54 3.36 .49 2.68 .03 220 218.96***
Expectations and behaviors 2.59 .56 3.38 .50 2.79 .04 216 220.69***
Communication skills 2.62 .60 3.36 .49 2.74 .04 220 218.28***
Smart dating strategies 2.62 .54 3.36 .46 2.74 .03 221 220.10***
Unhealthy relationships 2.71 .65 3.52 .44 2.80 .04 219 218.59***
***p , .001.298 Family Relations Volume 56, Number 3 July 2007
Table 3. Means and Standard Deviations of Relationship Beliefs and Conflict Tactics Scales’ Subscales
Experimental Group Control Group
Pretest Posttest Pretest Posttest
M SD M SD M SD M SD
Conflict tactics subscales
Reasoning 2.66 1.26 2.73 1.32 2.77 1.55 2.90 1.27
Verbal 1.91 1.45 1.75 1.42 2.05 1.52 2.29 1.61
Physical .76 1.21 .82 1.33 .99 1.42 1.17 1.67
Relationship belief subscales
Aggression beliefs 3.37 .83 3.37 .86 3.19 .99 3.26 .94
Faulty relationship beliefs 2.96 .53 3.03 .58 2.97 .50 3.02 .52
Realistic relationship beliefs 2.89 .47 3.01 .52 2.89 .50 2.85 .58
Subscales were used and comparisons were made frequently they used reasoning as a method for
between the experimental and the control groups of settling interpersonal differences, F(1, 292) ¼ .003,
students. Of the three subscales (reasoning, verbal p ¼ .96.
aggression, physical aggression), only the verbal Analyses also were conducted using race, income,
aggression subscale was found to have a statistically and family structure variables with the three Con-
significant Time Group interaction effect, flict Tactics Subscales. Repeated measures mixed
F(1, 297) ¼ 5.22, p ¼ .02. That is, while the two between-within subjects analyses of variance (RMA-
groups did not differ at Time 1 on use of verbal NOVAs) revealed no significant interaction effects
aggression, the RS adapted group demonstrated a sig- pertaining to race (1 ¼ African American and 2 ¼
nificantly lower level of use of verbally aggressive White), income (1 ¼ less than $40,000 and 2 ¼
tactics following the classes than the control group more than $40,000), or family structure (1 ¼ lives
(see Figure 1). For use of physically aggressive tactics with both biological parents, 2 ¼ lives with a biologi-
in interpersonal conflicts, results indicated no signif- cal parent and a stepparent, and 3 ¼ lives with a sin-
icant Time Group interaction effect between the gle parent), but significant differences were found in
groups, F(1, 274) ¼ .47, p ¼ .50. Similarly, the the between-subjects analysis. Specifically, regardless
results showed no significant Time Group inter- of being in the control or experimental group, Afri-
action effect between the groups in terms of how can American students reported a significantly
higher total score than White students on the physi-
cally aggressive subscale, F(1, 250) ¼ 12.32, p ,
4
Received curriculum
yes,
.001, partial g2 ¼ .12. This finding held true for
experimental both the pre- and the posttest scores. A similar trend
no, control
was observed for the verbal scale, with African
Estimated Marginal Means
3
American students reporting a significantly higher
frequency of using the verbally aggressive conflict
2
tactics in their interpersonal relationships than did
White students, F(1, 272) ¼ 21.54, p , .001, par-
tial g2 ¼ .09. No other interaction or between-sub-
1 jects effects were found for race, and there were no
significant differences according to income or family
structure.
0
1 2
Relationship Beliefs
Time
Another goal of this evaluation of RS adapted was to
Figure 1. Verbal Aggression. assess changes in relationship beliefs that may affectRelationship Education Adler-Baeder et al. 299
future behaviors and decisions in relationships. To hypothesized that all students who participated in
address the final hypothesis, we tested for changes the program, when compared to adolescents in the
across Time Group for each of the three relation- control group, would experience increases in rela-
ship belief subscales. Of the three subscales (aggres- tionship knowledge, decreases in destructive verbal
sion beliefs, faulty relationship beliefs, realistic and physical conflict strategies, increases in reason-
relationship beliefs), only the realistic relationship ing strategies, and positive changes in relationship
beliefs subscale was found to have a statistically signif- beliefs that support healthy relationships. The results
icant Time Group interaction effect, F(1, 317) ¼ of this study provide evidence that students who
4.71, p ¼ .03 (see Figure 2). That is, the RS adapted participated showed immediate gains in knowledge
group had more realistic relationship beliefs than the in several areas when compared to students who
control group at posttest; the groups did not differ at were in the control group.
pretest. As expected, the students who participated in
classes using the RS adapted curriculum experienced
significant improvements in scores from Time 1 to
Discussion Time 2 on all five of the relationship knowledge
subscales associated with the specific program con-
As relationship and marriage education programs tent. For each area, the student’s knowledge signifi-
cantly increased from pre- to posttest, providing
become more well known and are implemented with
a variety of audiences, it becomes imperative to mea- support for our hypothesis that students would gain
sure and document the impact on the participants. relationship knowledge as a result of taking the class.
Furthermore, analyses indicated that all the students
Although these programs are increasingly wide-
spread, there are surprisingly few published studies benefited in similar ways, regardless of race, income,
or family structure. This key finding implies that
of the impact of participation on adults (Carroll &
Doherty, 2003) and even fewer focused on youth. this curriculum has the potential to provide benefits
to a range of students with diverse ethnic and socio-
Gardner et al. (2001, 2004) documented positive
impacts of relationship education among White and economic backgrounds.
We also hypothesized that as a result of partici-
Hispanic high school students. The purpose of this
study was to add to the empirical basis for providing pating in the RS adapted classes, students would
these educational curricula, by examining the RS implement more reasoning strategies during con-
adapted program’s impact on African American and flict in their interpersonal relationships. Conversely,
White high school students from diverse socioeco- we expected that students would experience
decreases in their use of verbal and physical aggres-
nomic status and family structure backgrounds. We
sion strategies in their interpersonal conflicts, as
a result of participating in the classes. Although the
test and control groups were not different at Time
Received curriculum
4.0 yes, 1, the test group was significantly lower than the
experimental control group in verbal aggression at Time 2. This
no, control
3.5 finding was very encouraging given that research
Estimated Marginal Means
suggests adolescent aggression in dating relation-
3.0
ships is relatively stable and verbal aggression pre-
dicts physical aggression (O’Leary & Slep, 2003).
2.5
Introducing education that promotes learning to
2.0
resolve conflict and communicate without being
verbally condescending in early relationships may
1.5 provide a stronger foundation for healthier rela-
tionships in the future. Our finding is consistent
1.0 with Gardner et al. (2004) who found, using a dif-
1 2
ferent relationships education curriculum, that
Time adolescents who participated in the relationships
education course showed decreases in verbal (as well
Figure 2. Realistic Relationship Beliefs. as physical) aggression.300 Family Relations Volume 56, Number 3 July 2007
However, contrary to our hypothesis (and consis- strategies more frequently in their relationships than
tent with the findings of Gardner et al., 2004), the did White students, with moderate to large effect
use of reasoning strategies did not show significant sizes (.09 and .12, respectively; see Cohen, 1988).
improvement over time for the RS adapted group. Moreover, this tendency was true at both pre- and
This may be explained by the lower reliability of this posttest. It is unclear how to interpret this finding.
subscale (.54), which may indicate that this subscale There are some indications that African American
is not a satisfactory measure of reasoning among children, compared to White children, may be at
high school students. Alternately, the RS adapted slightly greater risk of exposure to aggression use in
course may need to be adjusted to more adequately family contexts. For example, research findings indi-
teach relationship reasoning skills. Because reason- cate that African American children experience more
ing involves regulation of emotions, it is also proba- frequent and severe physical punishment as children
ble that adolescents are building these skills and adolescents (Lansford, Deater-Deckard, Dodge,
developmentally and changes may occur over a lon- Bates, & Pettit, 2004). In contrast, a study using
ger period of time (Labouvie-Vief & Medler, 2002). data from Wave 1 and Wave 2 of the National Sur-
Recent brain development research supports this vey of Families and Households showed little evi-
notion, indicating that the frontal lobe, responsible dence of parenting practices differing between
for reasoning and judgment, continues to develop White and African American families (Amato &
during adolescence (National Institute of Mental Fowler, 2002). Relatedly, Sorenson, Upchurch, and
Health, 2001) and that substantial changes are Shen (1996), using a national data set, found that
occurring in the brain that impact perceptions of African Americans are more likely than Whites to
risk and reward and regulatory competence up report that marital arguments escalate to physical
through the late adolescent years (Steinberg, 2005). violence. However, there is some indication of dif-
In considering that a significant Time Group ferential effects of exposure to marital aggression. In
interaction effect was not found for the physical one recent study, marital conflict predicted children
aggression subscale, it is important to note the very with problem behavior in White families but not in
low mean scores at pretest, indicating very little vari- African American families (Nievar & Luster, 2006).
ability in response levels and making it difficult to Thus, we need more clarity on relative exposure to
detect a statistically significant change. Because the aggression use in family relationships on the basis of
desired direction for change is a decrease, we note ethnicity, and we need to understand more about
a ‘‘floor’’ effect in that detecting a significant the meaning, perceptions, and effects of verbal and
decrease from pre- to posttest when the pretest mean physical aggression within African American family
is very low is unlikely. Very few respondents and dating relationships.
reported high levels of physical aggression (the sub- Because we cannot derive specific meaning for
stantial variability in physical aggression is seen in this racial difference found in level of aggression use,
the larger standard deviations relative to the means). it would not be appropriate in practice to specifically
Although higher levels of physical aggression in rela- target African American teens with different infor-
tionships are expected among only a small propor- mation. Overall, with the high levels of dating vio-
tion of the adolescents (9%; Grunbaum et al., lence reported ranging from 11 to 41% (Centers for
2001), some variability in reported levels of the use Disease Control, 2001) and the findings here that
of physical aggression was expected, given recent a proportion of all teens sampled are using verbal
findings on the use of physical aggression in dating and physical aggression in dating relationships, it is
relationships (Ackard, Neumark-Sztainer, & suggested that any adolescent relationship education
Hannan, 2003). It is possible that students were program have specific modules that cover topics
reporting more socially desirable answers. such as conflict management strategies, emotion reg-
A noteworthy finding should be elaborated here. ulation, and communication skills. Information
When race was included in the analyses with the should also address dating aggression, violence, and
conflict subscales, significant between-subject differ- abuse. The information we provide here and the evi-
ences were evident. Namely, the African American dence from other research can serve to sensitize edu-
students, regardless of whether they participated in cators to the potential differences among diverse
the RS adapted classes, tended to use both verbally youth regarding the experiences, meaning, and per-
aggressive and physically aggressive conflict ceptions of aggression use in family relationships.Relationship Education Adler-Baeder et al. 301
Educators can facilitate discussions of family rela- backgrounds, in addition to those from a traditional
tional patterns and allow for student self-assessment two-parent household.
of the usefulness of learned patterns.
A final goal of this study was to assess relation-
ship beliefs that may affect future behaviors and
Limitations and Future Directions
decisions in relationships. Although only one of the
three relationship belief subscales (realistic relation-
ship beliefs) showed a statistically significant interac- Importantly, this is the first empirical evidence of
tion effect indicating differences between RS adapted short-term positive program impact of marriage
participants and controls, all the participants’ rela- education among a diverse sample of high school
tionship beliefs showed observed mean changes in adolescents that includes a significant number of
the expected direction from true pre- to posttest, African American students and a significant number
whereas those in the control group did not demon- of students from low-resource, geographically, and
strate such change. This suggests that those who par- structurally diverse families. Although positive
ticipated in the class may have gained insights that improvements have been detailed above, important
could lead to healthier relationships, knowledge that limitations should be noted. First, it is possible that
was not gained by those who did not take the class. the experimental and control groups differed from
Regarding the aggression beliefs and the faulty rela- the outset of the study. Without random assignment
tionship beliefs subscales, perhaps significant to groups, it is conceivable that unforeseen external
changes were not evident owing to the higher scores factors may have contributed to the significant dif-
at the onset for both groups. Likely, these adoles- ferences that were found, such as being more moti-
cents had an established grounded belief system in vated to learn the material. Students who
these areas that was consistent with desirable participated in the curriculum did self-select into an
responses at the onset. Thus, a ‘‘ceiling effect’’ may FCS class; however, controls also selected into an
be present for these subscales, as the mean scores for FCS class and therefore showed some interest in
both scales were fairly high to begin with. It is likely Human Science subjects as well. Another limitation
that program content affirmed these types of beliefs, is that the participants were slightly younger than
rather than changed them with these students. those in the control group and proportionately had
Because there are so few studies of this type, we are slightly more females. Future studies would do well
interested to see how the scale performs with other to include random assignment to groups to ensure
samples. Was this a group of students well informed greater confidence in interpreting the findings and
on these beliefs? If future studies replicate these find- increasing the reliability and generalizability of the
ings (i.e., students scoring high consistently at pre- results. Likewise, the teachers who implemented the
test), it might indicate that program content should curriculum were either self-selected or selected by
be adjusted. Rather than assuming these are beliefs a local extension agent, which may indicate they had
to be addressed/taught, it would be preferable to a greater motivation and perhaps had a more per-
affirm functional beliefs that teachers could assume sonal passion about the topics. This may have fur-
students already hold. ther affected a number of factors, including learning
Finally, for the majority of the findings, race, and teaching styles and positive impact on the par-
family structure, and income were not relevant to ticipants. This could be a limitation in generalizing
gains in relationship knowledge and beliefs. As these findings to classrooms in which teachers are
hypothesized, regardless of race, family structure, or directed to use the curriculum as a requirement
household income level, students taking the RS (e.g., as part of a mandated health curriculum). Less
adapted classes experienced gains in relationship motivated teachers could affect the curriculum’s
knowledge, including gaining a clearer understand- impact on students. This comparison remains an
ing of the linkages between relationship beliefs and empirical question.
knowledge and subsequent healthy dating behaviors Some potential next steps and future directions
and healthy romantic relationships. Additionally, for research include carrying out a more rigorous, ex-
less than half (40%) of the students came from perimental, longitudinal research design. Although
a two-parent nuclear family, which suggests that this we assessed students’ beliefs and knowledge with pre-
curriculum benefited those from a variety of family tests just prior to the first session, and posttests302 Family Relations Volume 56, Number 3 July 2007
shortly after the final session, a longer period of time enroll in FCS classes. It is suggested that the curricu-
is needed to determine the longevity of program lum be implemented in other classes in high schools,
impact on the adolescents’ future relationships, in- perhaps including it in mandatory core classes such
cluding potential marital relationships. Furthermore, as Health or Social Studies and in nonschool-based
long-term effects may be enhanced with booster ses- programs (e.g., 4-H, Boys and Girls Clubs, church
sions, so it would be helpful if future studies could youth groups, organized after-school programs).
assess the effects of the curriculum with and without In summary, educators have initial evidence that
such boosters. relationship and marriage education in high schools
Moreover, research will be necessary in order to is beneficial to White and Hispanic adolescent stu-
identify the components of the curriculum that are dents in an urban setting (Gardner et al., 2004).
most beneficial and whether this differs by gender, This finding is now expanded to include gains in
ethnicity, teacher, course type, or other variables. knowledge and positive changes in relationship
Given the ethnic differences found on use of verbal beliefs and behaviors for lower resource, geographi-
and physical aggression, it will be important to fur- cally diverse, and African American students.
ther explore the differential effects of specific areas Together, it can reasonably be said that relationship
of program content. It will also be important to programs show short-term benefit to adolescents
determine why some components of the program from different racial, family structural, and socioeco-
have a greater impact on students than do others. nomic backgrounds and from rural and urban set-
Determining which learning objectives are not being tings. Widespread research-based educational efforts
met and for whom and why this might be occurring aimed at teaching young people about healthy rela-
would implicate the areas of program content and tionships may prove valuable in future choices about
delivery that require adjustment. Expanding, alter- partnering and increase the chances for healthy rela-
ing, or tailoring the curriculum, or all, to the specific tionships and marriages, thus lowering rates of rela-
needs of the audience may facilitate expanded imple- tionship and marital instability and decreasing the
mentation to allow even greater effectiveness with risks for individual, family, and community dys-
a diverse group of students in different classroom function associated with family instability.
settings and delivered by teachers with varied
backgrounds.
References
Implications for Educators and Practitioners
In practice, it is recommended that educators con- Ackard, D. M., Neumark-Sztainer, D., & Hannan, P. (2003). Dating vio-
tinue to implement curricula that integrate an lence among a nationally representative sample of adolescent girls and
engaging and active learning process containing boys: Associations with behavioral and mental health. Journal of Gender
Specific Medicine, 6(3), 39–48.
experiential learning activities with more practical Administration for Children and Families. (2005). ACF healthy marriage
and experiential components. For example, teachers mission. U.S. Department of Health and Human Services. Retrieved
July 19, 2005, from http://www.acf.dhhs.gov/healthymarriage/about/
viewed the ‘‘bidding auction’’ for specific values that mission.html
adolescents could hold as much more engaging and Amato, P. R., & Fowler, F. (2002). Parenting practices, child adjustment,
and family diversity. Journal of Marriage and Family, 64, 703–716.
effective than simply discussing values with students. Brent, D. A., Perper, J. A., Moritz, G., Baugher, M., Roth, C., Balach, L.,
Similarly, adolescent relationship education pro- et al. (1993). Stressful life events, psychopathology, and adolescent suicide:
A case control study. Suicide and Life-Threatening Behavior, 23, 179–187.
grams should include extensive time devoted to Brotherson, S. E., & Duncan, W. C. (2004). Rebinding the ties that bind:
identifying abuse and positive healthy dating behav- Government efforts to preserve and promote marriage. Family Rela-
iors and options. Students often focused on this por- tions, 53, 459–468.
Carroll, J. S., & Doherty, W. J. (2003). Evaluating the effectiveness of pre-
tion of the curriculum in their written feedback, as marital prevention programs: A meta-analytic review of outcome
reflected in this comment: I enjoyed learning about research. Family Relations, 52, 105–118.
Close, S. M. (2005). Dating violence prevention in middle school and high
how to avoid being in an abusive relationship. school youth. Journal of Child and Adolescent Psychiatric Nursing, 18, 2–9.
There also lies a challenge in reaching a broader Cohen, J. (1988). Statistical power analysis for the behavioral sciences.
Hillsdale, NJ: Erlbaum.
audience with the curriculum. As with other pro- Collins, W. A. (2003). More than myth: The developmental significance
grams (Gardner et al., 2004; Nielsen, Pinsof, Ram- of romantic relationships during adolescence. Journal of Research on
Adolescence, 13, 1–24.
page, Solomon, & Goldstein, 2004), many who Furman, W., & Shaffer, L. (2003). The role of romantic relationships in
might benefit from relationships education do not adolescent development. In P. Florsheim (Ed.), Adolescent romanticRelationship Education Adler-Baeder et al. 303 relations and sexual behavior: Theory, research, and practical implications O’Leary, D. K., & Slep, A. M. S. (2003). A dyadic longitudinal model of (pp. 3–22). Mahwah, NJ: Erlbaum. adolescent dating aggression. Journal of Clinical Child and Adolescent Gardner, S. P. (2001). Evaluation of the ‘‘Connections: Relationships and Psychology, 32, 314–327. Marriage’’ curriculum. Journal of Family and Consumer Sciences Educa- Ooms, T., & Wilson, P. (2004). The challenges of offering relationship tion, 19 . Retrieved August 5, 2005, from http://www.natefacs.org/ and marriage education to low-income populations. Family Relations, JFCSE/vol19no1/v19no1Gardner.pdf 53, 440–447. Gardner, S. P., Giese, K., & Parrot, S. M. (2004). Evaluation of the con- Paul, E. L., & White, K. M. (1990). The development of intimate relation- nections: Relationships and marriage curriculum. Family Relations, 53, ships in late adolescence. Adolescence, 25, 375–399. 521–527. Pearson, M. (2000). Can kids get smart about marriage? A veteran teacher Grunbaum, J., Kann, L., Kinchen, S., Williams, B., Ross, J., Lowry, R., & reviews some leading marriage and relationship education programs . Kolbe, L. (2001). Youth Risk Behavior Surveillance—United States, Retrieved August 2, 2005, from http://marriage.rutgers.edu/publications/ 2001. MMWR 2002, 51(SS-4). Centers for Disease Control and Pre- pubcankids.htm vention. Retrieved August 5, 2005, from http://www.cdc.gov/mmwr/ Pearson, M. (2004). LoveU2: Getting smarter about relationships . Berkeley, preview/mmwrhtml/ss5104a1.htm CA: The Dibble Fund for Marriage Education. Hawkins, A. J., Carroll, J. S., Doherty, W. J., & Willoughby, B. (2004). Pratt, C. C., McGuigan, W. M., & Katzev, A. R. (2000). Measuring pro- A comprehensive framework for marriage education. Family Relations, gram outcomes: Using retrospective pretest methodology. American 53, 547–558. Journal of Evaluation, 21, 341–349. Joyner, K., & Udry, J. R. (2000). You don’t bring me anything but down: Shulman, S. (2003). Conflict and negotiation in adolescent romantic rela- Adolescent romance and depression. Journal of Health and Social Behav- tionships. In P. Florsheim (Ed.), Adolescent romantic relations and sexual ior, 41, 369–391. behavior: Theory, research, and practical implications (pp. 185–211). Karney, B., Garvan, C., & Thomas, M. (2003). Family formation in Florida: Mahwah, NJ: Erlbaum. 2003 baseline survey of attitudes, beliefs, and demographics relating to Silliman, B., & Schumm, W. R. (2004). Adolescents’ perceptions of marriage and family formation. Gainsville, FL: University of Florida. marriage and premarital couples education. Family Relations, 53, Labouvie-Vief, G., & Medler, M. (2002). Affect optimization and affect 513–520. complexity: Modes and styles of regulation in adulthood. Psychology & Sorenson, S. B., Upchurch, D. M., & Shen, H. (1996). Violence and injury Aging, 17, 571–589. in marital arguments: Risk patterns and gender differences. American Lam, T. C., & Bengo, P. (2003). A comparison of three retrospective self- Journal of Public Health, 86, 34–40. reporting methods of measuring change in instructional practice. Amer- Steinberg, L. (2005). Cognitive and affective development in adolescence. ican Journal of Evaluation, 24, 65–80. TRENDS in Cognitive Sciences, 9, 69–74. Lansford, J. E., Deater-Deckard, K., Dodge, K. A., Bates, J. E., & Pettit, Straus, M., Hamby, S., Boney-McCoy, S., & Sugarman, D. (1996). The G. S. (2004). Ethnic differences in the link between physical discipline revised conflict tactics scales (CTS2): Development and preliminary and later adolescent externalizing behaviors. Journal of Child Psychology psychometric data. Journal of Family Issues, 17, 283–316. and Psychiatry, 45, 801–812. Straus, M. A. (1990). The Conflict Tactics Scales and its critics: An evalua- Montgomery, M. J. (2005). Psychosocial intimacy and identity: From early tion and new data on validity and reliability. In M. A. Straus & R. J. adolescence to merging adulthood. Journal of Adolescent Research, 20, Gelles (Eds.), Physical violence in American families: Risk factors and 346–374. adaptations to violence in 8,145 families (pp. 49–73). New Brunswick, National Institute of Mental Health. (2001). Teenage brain: A work in NJ: Transaction. progress . Retrieved September 12, 2006, from http://www.nimh.nih. Teachman, J., Tedrow, L., Crowder, K. (2000). The changing demography of gov/Publicat/teenbrain.cfm America’s families. Journal of Marriage and the Family, 62, 1234–1246. Nielsen, A., Pinsof, W., Rampage, C., Solomon, A. H., & Goldstein, S. Wekerle, C., & Wolfe, D. A. (1999). Dating violence in mid-adolescence: (2004). Marriage 101: An integrated academic and experiential under- Theory, significance, and emerging prevention initiatives. Clinical Psy- graduate marriage education course. Family Relations, 53, 485–494. chology Review, 19, 435–456. Nievar, M. A., & Luster, T. (2006). Developmental processes in African Wolfe, D. A. (2006). Preventing violence in relationships: Psychological American families: An application of McLoyd’s theoretical model. science addressing complex social issues. Canadian Psychology, 47, Journal of Marriage and Family, 68, 320–331. 44–50.
You can also read
