Is Skills Training Necessary for the Primary Prevention of Marital Distress and Dissolution? A 3-Year Experimental Study of Three Interventions

Abstract

Objective:
Evidence in support of skill-based programs for preventing marital discord and dissolution, while promising, comes mainly from studies using single treatment conditions, passive assessment-only control conditions, and short-term follow-up assessments of relationship outcomes. This study overcomes these limitations and further evaluates the efficacy of skill-based programs.

Method:
Engaged and newlywed couples (N = 174) were randomly assigned to a 4-session, 15-hr small-group intervention designed to teach them skills in managing conflict and problem resolution (PREP) or skills in acceptance, support, and empathy (CARE). These couples were compared to each other, to couples receiving a 1-session relationship awareness (RA) intervention with no skill training, and to couples receiving no treatment on 3-year rates of dissolution and 3-year trajectories of self-reported relationship functioning.

Results:
Couples in the no-treatment condition dissolved their relationships at a higher rate (24%) than couples completing PREP, CARE, and RA, who did not differ on rates of dissolution (11%). PREP and CARE yielded unintended effects on 3-year changes in reported relationship behaviors. For example, wives receiving PREP showed slower declines in hostile conflict than wives receiving CARE, and husbands and wives receiving CARE showed faster declines in positive behaviors than husbands and wives receiving PREP.

Conclusions:
These findings highlight the potential value of cost-effective interventions such as RA, cast doubt on the unique benefits of skill-based interventions for primary prevention of relationship dysfunction, and raise the possibility that skill-based interventions may inadvertently sensitize couples to skill deficits in their relationships.

Full text

Is Skills Training Necessary for the Primary Prevention of Marital Distress
and Dissolution? A 3-Year Experimental Study of Three Interventions
Ronald D. Rogge
University of Rochester
Rebecca J. Cobb
Simon Fraser University
Erika Lawrence
University of Iowa
Matthew D. Johnson
Binghamton University
Thomas N. Bradbury
University of California, Los Angeles
Objective: Evidence in support of skill-based programs for preventing marital discord and dissolution,
while promising, comes mainly from studies using single treatment conditions, passive assessment-only
control conditions, and short-term follow-up assessments of relationship outcomes. This study overcomes
these limitations and further evaluates the efficacy of skill-based programs. Method: Engaged and
newlywed couples (N⫽174) were randomly assigned to a 4-session, 15-hr small-group intervention
designed to teach them skills in managing conflict and problem resolution (PREP) or skills in acceptance,
support, and empathy (CARE). These couples were compared to each other, to couples receiving a
1-session relationship awareness (RA) intervention with no skill training, and to couples receiving no
treatment on 3-year rates of dissolution and 3-year trajectories of self-reported relationship functioning.
Results: Couples in the no-treatment condition dissolved their relationships at a higher rate (24%) than
couples completing PREP, CARE, and RA, who did not differ on rates of dissolution (11%). PREP and
CARE yielded unintended effects on 3-year changes in reported relationship behaviors. For example,
wives receiving PREP showed slower declines in hostile conflict than wives receiving CARE, and
husbands and wives receiving CARE showed faster declines in positive behaviors than husbands and
wives receiving PREP. Conclusions: These findings highlight the potential value of cost-effective
interventions such as RA, cast doubt on the unique benefits of skill-based interventions for primary
prevention of relationship dysfunction, and raise the possibility that skill-based interventions may
inadvertently sensitize couples to skill deficits in their relationships.
Keywords: couples, marriage, intervention, prevention
Efforts to prevent relationship distress assume that teaching
couples how to communicate effectively enables them to have
more fulfilling and durable unions. Controlled studies with 3- to
6-month follow-ups support these efforts (Hawkins, Blanchard,
Baldwin, & Fawcett, 2008), prompting scholars to recommend
new studies that follow couples for longer periods, that compare
interventions derived from competing theoretical perspectives, and
that replace passive assessment-only control groups with active
minimal interventions (e.g., see Halford, 2011). We address these
recommendations by comparing the 3-year relationship outcomes
of couples randomized to one of three active treatment conditions.
Review of Theory and Research
The most widely studied preventive interventions for couples
begin with the premise, adapted from social learning theory, that
judgments of relationship satisfaction are determined by the fre-
quency and patterning of the rewarding and punishing behaviors
that partners exchange (e.g., Jacobson & Margolin, 1979). Most
tests of this premise focus specifically on how partners manage
disagreements, and studies of newlywed couples support the con-
tention that displays of critical and hostile behaviors during
problem-solving conversations hasten the rate of relationship de-
terioration while constructive communication offsets these effects
(e.g., Johnson et al., 2005). The Prevention and Relationship
Enhancement Program (PREP; Markman, Stanley, & Blumberg,
1994) capitalizes on this model, providing couples with training in
communication and problem-solving skills, as well as relationship
expectations, friendship, and commitment. Studies of PREP with
This article was published Online First September 23, 2013.
Ronald D. Rogge, Department of Clinical and Social Sciences in Psy-
chology, University of Rochester; Rebecca J. Cobb, Department of Psy-
chology, Simon Fraser University, Burnaby, British Columbia, Canada;
Erika Lawrence, Department of Psychology, University of Iowa; Matthew
D. Johnson, Department of Psychology, Binghamton University; Thomas
N. Bradbury, Department of Psychology, University of California, Los
Angeles.
This research was supported by a grant from the John Templeton
Foundation awarded to Thomas N. Bradbury.
Correspondence concerning this article should be addressed to Thomas
N. Bradbury, Department of Psychology, University of California, Los
Angeles, CA 90095-1563. E-mail: bradbury@psych.ucla.edu
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Journal of Consulting and Clinical Psychology © 2013 American Psychological Association
2013, Vol. 81, No. 6, 949–961 0022-006X/13/$12.00 DOI: 10.1037/a0034209
949

3-year (Hahlweg, Markman, Thurmaier, Engl, & Eckert, 1998) and
5-year (Markman, Renick, Floyd, Stanley, & Clements, 1993)
follow-up assessments suggest that couples participating in this
program are more satisfied than couples who receive no interven-
tion, although couples’ self-selection into and out of the interven-
tion may exaggerate these differences (Halford, Markman, & Stan-
ley, 2008).
Though informative, comparisons between PREP and untreated
couples leave open the possibility that stronger effects can be
achieved with interventions derived from alternate theories. Re-
sponding to limitations of social learning models for treating
relationship distress, Jacobson and Christensen (1996) developed
an acceptance-oriented framework in which partners are encour-
aged to join around their differences and develop a shared but
detached perspective on the issues that divide them. Whereas
programs guided by social learning theory (e.g., PREP) address the
interpersonal skills couples need to negotiate competing demands
for change, acceptance-oriented programs target ways that partners
can empathize with one another’s perspectives. Other models that
prioritize prosocial exchanges to help couples maintain healthy
relationships have emerged in recent years, highlighting the im-
portance of interpersonal skills and capacities that extend beyond
conflict management (e.g., dyadic coping, Bodenmann, 2005; for-
giveness, McCullough, Worthington, & Rachal, 1997). The pres-
ent study compares PREP with a prosocial intervention—Compas-
sionate and Accepting Relationships Through Empathy, or CARE
(Rogge, Johnson, Lawrence, Cobb, & Bradbury, 2002)—in an
effort to clarify which theory-derived set of skills best promotes
satisfying relationships.
Like many preventive interventions, PREP and CARE assume
that improving skills is essential for healthy relationships. How-
ever, an alternative perspective holds that most couples already
possess the basic skills they need to keep their relationship healthy
and strong, requiring instead interventions that promote greater
relationship awareness (RA) and that prompt routine monitoring
of the relationship. Partners who are more aware of their relation-
ships and who identify strongly with the relationship are hypoth-
esized to make more efforts, and more effective efforts, at rela-
tionship maintenance (e.g., Acitelli, 2001). Greater relationship
identity and awareness reduce divorce rates over 5 years (Carrère,
Buehlman, Gottman, Coan, & Ruckstuhl, 2000), even after adjust-
ing for concurrent negativity. Interpersonal skills are not unim-
portant from the RA perspective but are seen as essentially intact
and secondary to partners’ motivations to identify with the rela-
tionship and to deploy the skills they already possess (see Snyder
& Schneider, 2002). Distinguishing between communication skills
and the motivation or desire to implement these skills suggests that
encouraging partners to attend to and maintain their relationship
may supplant their natural inclination to take the relationship for
granted as time passes.
Based on a study demonstrating that low-risk couples benefit as
much from discussing a book about relationships as from learning
specific communication skills (Halford, Sanders, & Behrens,
2001), we propose that partners’ motivation to maintain their
relationships can be enhanced and that relationship deterioration
can be prevented by heightening couples’ awareness of the impor-
tance of relationship maintenance. We propose further that, be-
cause no skill training is required, RA can be enhanced with
low-dose, low-cost interventions, particularly interventions that
couples can incorporate easily into their daily lives. In the present
study we implement an RA intervention that outlines this basic
principle in a single in-person session, followed by four self-
guided in-home practice sessions. Implementing the RA interven-
tion can clarify any incremental value of skill-based interventions
while addressing whether low-dose interventions can benefit cou-
ples who are unable or unwilling to participate in intensive skill-
based programs.
Study Aims and Hypotheses
The present study compares the 3-year outcomes of engaged or
newlywed couples randomly assigned to receive either the leading
prevention program derived from the social learning model
(PREP), a structurally similar program derived from an
acceptance-based model of relationships (CARE), or a brief, low-
cost intervention designed to increase RA. These three groups of
couples are compared to one another and to a group of comparable
couples receiving no intervention, to determine which of the three
active conditions, if any, prevent dissolution and slow rates of
decline in relationship satisfaction. Extending follow-up assess-
ments to 3 years allows us to sample dissolutions directly, thereby
expanding upon the more commonly studied changes in satisfac-
tion (see Hawkins et al., 2008).
Comparing three potentially active treatments allows us to ex-
amine another unanswered question in long-term intervention
studies with couples: Do interventions produce specific changes in
their main intervention targets, or do the effects expand beyond
these targets? By focusing primarily on behavioral outcomes that
are congruent with the intervention (e.g., PREP studies typically
focus on changes in problem-solving behaviors) and by relying on
treated-versus-untreated control comparisons, prior studies over-
looked the possibility that the interventions produce beneficial
effects in other interpersonal domains. An intervention that
strengthens problem-solving skills, for example, might yield ben-
efits in areas that it does not emphasize (e.g., improved support),
and conversely, an intervention designed to improve empathy and
acceptance might also improve conflict management. We explore
these possibilities by examining three active intervention condi-
tions and by assessing trajectories of self-reported 3-year change in
conflict, support, and affection.
We test three main hypotheses. First, we predicted that couples
receiving PREP and CARE would not differ in their outcomes but
would experience better relationship outcomes compared to cou-
ples receiving no treatment, where better relationship outcomes are
defined as lower divorce rates and more favorable 3-year relation-
ship satisfaction trajectories.
Second, in view of evidence that RA may promote relationship
functioning, we predicted that couples receiving the RA interven-
tion would experience better outcomes than couples receiving no
intervention. As RA is an untested approach, we made no predic-
tions comparing RA to PREP or CARE. On one hand, the focus on
skills training in PREP and CARE might yield stronger effects
than RA, while, on the other hand, the routine maintenance func-
tion of the RA intervention may yield comparable effects to PREP
and CARE.
Third, we predicted that PREP couples would experience
steeper declines in their reported use of negative behavior com-
pared to CARE couples and, conversely, that CARE couples
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
950 ROGGE, COBB, LAWRENCE, JOHNSON, AND BRADBURY

would experience smaller declines in their reported displays of
emotional support and affection than PREP couples would. We
expected the two skill-based interventions to outperform the RA
group in terms of negative behavior, emotional support, and af-
fection.
Method
Participants
Participants were 174 engaged or newlywed couples. Men av-
eraged 29.3 years of age (SD ⫽4.8) and 15.0 years of education
(SD ⫽2.5); 6% were unemployed. Women averaged 27.9 years of
age (SD ⫽4.9) and 15.0 years of education (SD ⫽2.3); 16% were
unemployed. Modal incomes for men and for women were be-
tween $30,000 and $50,000. Just over half of the participants were
Caucasian (55%), with 21% Latino, 11% Asian, 5% African
American, and 8% other. Most couples (80%) were engaged to be
married; they participated in the intervention for an average of 6.8
months (SD ⫽4.0) prior to their weddings. The remaining couples
were married at the screening and participated for an average of
3.2 months (SD ⫽2.7) after their weddings. Most (72%) couples
were cohabiting at the time of screening (M⫽2.6 years, SD ⫽
2.0). Forty-two couples (24%) had children, 18 of whom were
from a previous relationship.
Procedure
Recruitment and screening. Recruitment has been described
previously (Rogge et al., 2006) and is detailed in Figure 1. Fol-
lowing institutional review board–approved procedures, one
spouse from each couple was screened via a 20-min telephone
interview to obtain informed consent and to assess interest, eligi-
bility, demographics, and relationship satisfaction (using the Kan-
sas Marital Satisfaction Scale; Schumm, Nichols, Schectman, &
Grinsby, 1983). Eligibility requirements were (a) both partners
consented to participate, (b) partners were fluent in English, (c) the
couple was engaged to be married within the next year or married
fewer than 6 months, (d) the partners were starting first marriages,
and (e) the couple was not distressed (by the interviewee’s report).
Distressed couples received appropriate referrals. Eligible partners
received questionnaires (pretreatment, T0) with consent forms in
separate envelopes and written instructions not to discuss their
responses. Couples completing T0 (n⫽183) were no different
from couples who did not (n⫽155) on satisfaction, couples
counseling, presence of children at marriage, and proportion of
Asian spouses. There were fewer couples with an African Amer-
ican partner among couples who completed T0 (6%) compared to
those who did not complete T0 (15%), ␹
2
(1) ⫽7.6, p⬍.01.
After completing T0, couples were randomly assigned to
CARE, PREP, or the single-session RA condition and were sched-
uled by telephone for the workshops. Six couples assigned to
CARE or PREP with work or commute schedules that prohibited
attendance at the weeknight sessions accepted our invitation to
participate in the RA condition. Although 130 of 145 couples
(90%) attended at least the first treatment session, 15 couples could
never be successfully scheduled and were combined with the 29
couples who declined treatment to create the no-treatment (NoTx)
group. As Figure 1 shows, 52 couples received CARE, 45 couples
Phone Assessment
n = 831
T0 Packet
n = 338
Ineligible n = 493
351 – at least one spouse refused
152 – married longer than 6 months
139 – not first marriage
83 – wedding date more than 1 year away
59 – too severely distressed
7 – not fluent in English
Group Assignment
n = 183
155 failed to
complete packet
Not assigned n = 9
7 declined further participation
2 no longer met criteria
CARE
n = 52
PREP
n = 45
RA
n = 33
NoTx Group n = 44
29 declining active treatment
12 unable to schedule for RA
2 unable to schedule for CARE
1 unable to schedule for PREP
Figure 1. Flowchart of couple selection and participation. CARE ⫽Compassionate and Accepting Relation-
ships Through Empathy; NoTx ⫽no treatment; PREP ⫽Prevention and Relationship Enhancement Program;
RA ⫽relationship awareness; T0 ⫽pretreatment.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
951
PREVENTING MARITAL DISTRESS AND DISSOLUTION

received PREP, 33 couples received RA, and 44 couples were in
the NoTx group.
Treatment format. Groups of three to six couples completed
CARE and PREP workshops in an initial 6-hr weekend session
followed by three weekly 3-hr evening sessions (i.e., 15 hr over 1
month). Groups of 10 –15 couples completed RA workshops in a
single on-campus 4-hr session and four weekly home sessions. RA
participation was assessed by weekly telephone calls, and couples
provided brief notes on the movies and their semistructured dis-
cussions that were returned weekly to the project in self-addressed,
stamped envelopes. Doctoral students in clinical psychology with
at least 2 years of clinical training led the workshops. Advanced
undergraduate research assistants served as coaches for the CARE
and PREP exercises. Coach training and supervision are detailed
elsewhere (Rogge et al., 2002).
Treatment dropout. Of the 130 couples who participated in
active treatment conditions, 27 couples attended fewer than three
sessions, primarily because of time constraints and distance to
campus. Dropouts were evenly distributed across treatments; eight
couples withdrew from CARE, 10 from PREP, and nine from
RA, ␹
2
(2) ⫽1.8, ns. Repeated measures analyses of variance
(ANOVAs) with partners treated as a within-subject factor (and
chi-square analyses) indicated that withdrawing couples were very
similar to completers. However, withdrawers had slightly lower
levels of education (M⫽15.2 years, SD ⫽3.0) compared to
completers (M⫽16.4 years, SD ⫽3.1), F(1, 128) ⫽5.7, p⬍.02,
␩
2
⫽.043, and withdrawers had more children (M⫽2.6, SD ⫽
1.4, among couples with children) than completers (M⫽1.7,
SD ⫽0.6), F(1, 27) ⫽5.5, p⬍.03, ␩
2
⫽.170. Although it was
likely to underestimate treatment effects, we nevertheless retained
these couples in the outcome analyses.
Follow-up assessments. Couples in CARE, PREP, and RA
completed measures of relationship quality and relationship pro-
cesses a maximum of six times: at pretreatment (T0), at the start of
the workshop (T1, 1–2 months later), and at 6, 12, 24, and 36
months following the end of treatment (Times 2–5, respectively).
NoTx couples completed the T0 packet and provided Marital
Adjustment Test (MAT) and relationship stability data at T0, T3,
and T5. Couples received $25 for each assessment.
Attrition. Of the 174 couples, seven in CARE, eight in PREP,
and three in RA provided no follow-up data; they were evenly
distributed across conditions, ␹
2
(2) ⫽1.2, p⬍.55, ␸⫽.10.
Repeated measures ANOVAs with partners treated as a within-
subject factor, and chi-square analyses indicated that the couples
who failed to provide follow-up data were not significantly differ-
ent from couples who provided follow-up data on any measure.
Treatment Conditions
Table 1 summarizes the format and content of the three active
treatment conditions.
Prevention and Relationship Enhancement Program
(PREP). PREP is a psychoeducational program designed to
strengthen relationships by teaching couples communication skills.
PREP includes 14 lectures on a range of topics (e.g., problem
solving, time-outs, commitment, and sensuality); a forgiveness
Table 1
Description of Active Treatment Conditions
Program component CARE PREP RA
Program format
In-person sessions 5: 4 ⫹SAVE 5: 4 ⫹SAVE 1
Didactic lectures 16 (15–45 min) 16 (15–45 min) 1 (10 min)
Skill exercises 9 (20–45 min) 10 (15–50 min) 0
Homework 12 skill-based exercises 10 skill-based exercises 4 discussions
Contact time 15 hr ⫹3 hr for SAVE 15 hr ⫹3 hr for SAVE 4 hr
Conflict and problem-solving modules Hard vs. soft feelings and
acceptance; changing
perspectives and
problem reformulation;
detaching via empathic
joining; effective use
of time-outs;
preventing conflict
Structured communication and speaker–
listener technique; how filters distort
communication; altering negative
communication with XYZ statements;
danger signs, time-outs, and negative
conflict behaviors; effective
problem-solving; communicating
expectations; setting ground rules;
respecting core beliefs; issues vs. events
3 discussion prompts
about conflict; 2
discussion
prompts about
expectations
Support and positivity modules Listening as a friend and
empathic amplification;
managing moods;
committing random
acts of kindness and
affection
Maintaining positives via fun, friendship, and
sensuality; fostering commitment and a
long-term view
2 discussion prompts
about social
support
Forgiveness modules Discussing hurt feelings
with the language of
acceptance; effective
apologies; giving
partners the benefit of
the doubt
Discussing hurt feelings with the language of
acceptance; effective apologies; giving
partners the benefit of the doubt
2 discussion prompts
about forgiveness
Note. CARE ⫽Compassionate and Accepting Relationships Through Empathy; PREP ⫽Prevention and Relationship Enhancement Program; RA ⫽
relationship awareness; SAVE ⫽Stop Anger and Violence Escalation.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
952 ROGGE, COBB, LAWRENCE, JOHNSON, AND BRADBURY

module was excluded to minimize overlap with CARE. Couples
completed numerous exercises designed to practice the PREP
skills (see Table 1). Central to PREP is the speaker-listener tech-
nique, which slows the pace of communication by ensuring that
one spouse’s point of view is accurately reflected before moving
on to discuss the partner’s point of view; in many of the exercises,
couples discussed various topics while using the speaker-listener
technique. Three of the graduate students delivering PREP were
trained personally by the developers of the PREP program.
Compassionate and Accepting Relationships Through Em-
pathy (CARE). Designed for this study, CARE aims to
strengthen relationships by teaching couples supportive and em-
pathic skills. Based on integrative behavioral couple therapy
(IBCT; Jacobson & Christensen, 1996), CARE includes 16 lec-
tures covering a core set of acceptance-based skills. As with PREP,
the lectures were interspersed with exercises designed to help
couples practice new skills or to explore topics in their relationship
(see Table 1). CARE emphasizes skills designed to enhance em-
pathy, compassion, and acceptance. Building on strategies in
IBCT, couples were encouraged to use the language of acceptance
(e.g., focusing on understanding one’s partner, making soft rather
than hard disclosures) when discussing relationship problems,
individual problems, and relationship transgressions. Couples also
learned basic reframing skills and skills to help them empathically
join to tackle problematic patterns of dyadic behavior.
Relationship awareness (RA). Developed for this study, the
one-session RA condition was designed to heighten partners’
awareness of their relationship and the importance of regular
relationship maintenance. Rather than teach couples new skills,
RA drew partners’ attention to current behavior in their relation-
ship and encouraged them to decide for themselves if their behav-
ior was constructive or destructive. During an on-campus presen-
tation, couples were informed about the importance of RA and
maintenance, and they were introduced to the idea that regular
everyday events—particularly those captured in commercial
films— could be used as prompts to accomplish these goals. Cou-
ples then watched a movie, Two for the Road (Donen, 1967), in
which a couple revisits earlier scenes from their marriage and
recounts the joys and difficulties they experienced. In separate
rooms, couples then followed instructions for 50- to 60-min semi-
structured discussions in which they addressed the themes of this
film (including conflict, support, stress, and forgiveness) and how
they could reflect on these themes in their own relationship.
Coaches intervened minimally in the discussions and primarily
focused couples on the task, encouraged partners to engage the
questions thoughtfully, and answered any questions. After the
discussions, couples received a list of 47 movies with an inti-
mate relationship as a major plot focus with instructions to watch
one movie per week at home for the next month and to discuss the
same set of open-ended questions following each movie (see Table
1). Couples completed and returned a brief questionnaire for each
movie they watched; rental costs for all movies were reimbursed.
Additional elements in CARE and PREP. In our original
design, about half of the CARE and PREP couples were randomly
assigned to receive a specialized 3-hr forgiveness module. To keep
contact time similar for all CARE and PREP couples, remaining
couples received 3 extra hours of practice in the skills targeted by
their interventions. Couples who did and did not receive the
forgiveness module did not differ on any outcome measures, and
thus, all analyses focus on contrasts among CARE, PREP, RA, and
NoTx. CARE and PREP couples also attended a 3-hr session at the
end of treatment devoted to the Stop Anger and Violence Escala-
tion program (SAVE; Neidig, 1989). Couples in the RA condition
did not receive SAVE to avoid introducing any skill training into
that intervention.
Treatment adherence for CARE and PREP. Audiotapes of
16 CARE lectures and 15 PREP lectures, randomly sampled across
all workshop leaders, were coded by two independent judges for
the presence or absence of specific intervention elements (e.g.,
“Leader described three key ways expectations may cause prob-
lems”; range ⫽6 –16 elements per lecture, Mdn ⫽12 elements).
Interrater reliability was adequate (intraclass correlation coeffi-
cient ⫽.81), and computed means indicated that leaders adhered
to 93% of the PREP elements and 95% of the CARE elements.
Satisfaction with treatment. Satisfaction with the workshop,
assessed on a 7-point scale at the end of treatment, differed across
couples, as indicated by a significant repeated measures ANOVA
using partner as a within-subject factor, F(2, 79) ⫽4.1, p⬍.02,
partial ␩
2
⫽.094. Post hoc Tukey analyses showed that CARE
couples were more satisfied (M⫽5.6, SD ⫽0.8) than RA couples
(M⫽5.0, SD ⫽.8); PREP couples did not differ from the other
groups (M⫽5.4, SD ⫽1.0). CARE and PREP couples were
similar in their satisfaction with workshop leaders (M⫽6.5, SD ⫽
0.7, and M⫽6.4, SD ⫽0.8, respectively) and coaches (M⫽5.7,
SD ⫽0.5, and M⫽5.6, SD ⫽0.5, respectively; ns).
Measures
Table 2 provides psychometric data and descriptive statistics for
all key T0 measures.
Demographics. At T0, participants provided information
about their age, relationship, education, race, children, living ar-
rangements, and previous couples counseling.
Relationship satisfaction. As detailed above, participants
completed the Marital Adjustment Test (MAT; Locke & Wallace,
1959) and the Positive and Negative Quality in Marriage Scale
(PANQIMS; Fincham & Linfield, 1997), which asks respondents
to rate the positive aspects of their relationship while ignoring the
negative aspects and then rate the negative aspects of their rela-
tionship while ignoring the positives. Relevant items are averaged
to yield Positive and Negative subscale scores.
Hostile conflict behavior. The Conflict subscale of the Mar-
ital Coping Inventory (Bowman, 1990) assesses frequency of
antagonistic behaviors (e.g., “I yell or shout at my partner”).
Higher scores indicate higher levels of negative behavior.
Validation and affection. The Support in Intimate Relation-
ships Rating Scale (Dehle, Larsen, & Landers, 2001) assesses
frequency of social support provided by one’s partner in the prior
2 weeks. Based on a factor analysis (Barry, Bunde, Brock, &
Lawrence, 2009), we created subscales for emotional validation
(e.g., “Said he/she would feel the same way in my situation”) and
physical affection (“Hugged or cuddled with me”).
Forgiveness. The Wade Forgiveness Measure (McCullough
et al., 1997) assesses individuals’ tendencies to forgive their part-
ner after being unfairly hurt by them. Items focus on specific
behavioral reactions to being hurt (e.g., “I harbor a grudge,” “I
accept my partner”).
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
953
PREVENTING MARITAL DISTRESS AND DISSOLUTION

Trait anger. The Multidimensional Anger Inventory (Siegel,
1986) assesses individuals’ cross-situational propensities to expe-
rience and react with anger (e.g., “I tend to get angry more
frequently than most people”).
Aggression. Respondents reported their own and their part-
ner’s aggression using the Conflict Tactics Scale–Revised (Straus,
Hamby, Boney-McCoy, & Sugarman, 1996) with four items at T0
and 16 items from T1 to T5. Aggression was treated as dichoto-
mous outcome, and an individual was considered aggressive if
either partner reported at least one aggressive act for the individ-
ual.
Results
As shown in Table 2, MAT and PANQIMS scores correlated in
expected directions, and higher levels of negative behavior cova-
ried strongly with lower levels of forgiveness and higher levels of
trait anger. Remaining correlations were moderate to weak in
magnitude and in the expected directions, suggesting low levels of
collinearity for the remaining analyses.
Equivalence of Treatment Conditions
As shown in Table 3, there were few differences between groups
at T0, with two exceptions: PREP participants were slightly older
than CARE and NoTx participants, and RA (26%) and NoTx
(31%) groups had higher proportions of Latino/a spouses than
CARE (15%) and PREP (13%) groups. Groups did not differ on 20
of 22 demographic and relationship dimensions examined (see
Table 3), indicating that they can be compared in a meaningful
way.
Do Dissolution Rates Vary by Treatment Group?
Of the 153 couples who provided follow-up data, 25 (16.3%)
ended their relationships (e.g., separation, divorce) by the 3-year
follow-up assessment: six CARE couples (13.3%), five PREP
couples (13.5%), four RA couples (13.3%), and 10 NoTx couples
(24.4%). A chi-square test showed that the CARE, PREP, and RA
couples had marginally lower rates of dissolution than NoTx
couples, ␹
2
(1) ⫽2.7, p⬍.10, ␸⫽⫺.132. This effect became
stronger when the analysis was restricted to the couples who
completed one of the three active treatments in comparison to the
NoTx couples—11% dissolution in treatment completers versus
24% in NoTx couples, ␹
2
(1) ⫽4.4, p⬍.03, ␸⫽⫺.175—where
completion was defined as participation in the first session as well
as two additional sessions (for PREP and CARE couples) or two
additional movies (for RA couples). Given the equivalence of
groups at T0 (see Table 3), these results suggest that all treatments
reduced risk of separation or divorce over the first 3 years of
marriage.
Do Changes in Relationship Outcomes Vary by
Treatment Group?
Using data summarized in Table 4, we examined whether tra-
jectories of relationship outcomes varied by treatment group using
multilevel modeling and the Hierarchical Linear Modeling Pro-
gram (HLM 6.0; Raudenbush, Bryk, & Congdon, 2004). We used
a three-level model in which repeated measurements over time
were modeled at Level 1 (using a slope-intercept format and
centering time at the start of treatment), individual partners were
modeled at Level 2, and dyads were modeled at Level 3. Trajec-
tories were allowed to vary across treatment conditions and gender
(using dichotomous variables identifying group membership), in-
tercepts were treated as random effects at Level 2, and slopes were
treated as random effects at Level 3.
Level 1 Relationship Quality ⫽␲
0
⫹␲
1
(time) ⫹E
Level 2 ␲
0
⫽␤
01
ⴱ
(MEN) ⫹␤
02
ⴱ
(WOMEN) ⫹r
0
␲
1
⫽␤
11
ⴱ
(MEN) ⫹␤
12
ⴱ
(WOMEN)
Level 3 ␤
01
⫽␥
010
⫹␥
011
ⴱ
(CARE) ⫹␥
012
ⴱ
(PREP) ⫹␥
013
ⴱ
(RA)
␤
02
⫽␥
020
⫹␥
021
ⴱ
(CARE) ⫹␥
022
ⴱ
(PREP) ⫹␥
023
ⴱ
(RA)
Table 2
Psychometrics and Correlations Among Variables at Pretreatment
Variable
No. of
items Range
Men Women Correlations between measures
MSD␣MSD␣123456789
1. Relationship satisfaction 15 2–158 119 21 .71 119 19 .70 .53 .60 ⫺.69 ⫺.36 .25 .23 .46 ⫺.22 ⫺.31
2. Positive relationship qualities 3 1–10 9.3 1.0 .82 9.5 0.7 .76 .35 .24 ⫺.46 ⫺.11 .18 .20 .30 ⫺.20 ⫺.28
3. Negative relationship qualities 3 1–10 3.7 2.0 .86 3.8 2.1 .87 ⫺.59 ⫺.26 .22 .46 ⫺.13 ⫺.11 ⫺.49 .27 .20
4. Hostile conflict behavior 15 0–60 14.0 8.8 .89 18.1 10.8 .91 ⫺.46 ⫺.22 .44 .41 ⫺.23 ⫺.06 ⫺.55 .51 .30
5. Emotional support 7 0–49 18.9 13.1 .94 19.1 13.5 .91 .12 .09 ⫺.10 ⫺.12 .20 .37 .29 ⫺.13 ⫺.03
6. Affectionate behavior 4 0–28 20.1 9.0 .94 20.2 8.7 .94 .19 .12 ⫺.03 ⫺.03 .40 .24 .13 ⫺.06 ⫺.05
7. Forgiveness toward partner 20 0–100 79.6 13.4 .87 73.7 16.8 .92 .41 .17 ⫺.35 ⫺.63 .10 .13 .19 ⫺.51 ⫺.22
8. Trait anger 13 0–52 16.8 8.9 .82 19.3 10.5 .78 ⫺.34 ⫺.12 .34 .47 ⫺.09 ⫺.02 ⫺.45 ⴚ.02 .26
9. Physical aggression 4 0–8 0.43 1.01 .74 0.74 1.43 .63 ⫺.33 ⫺.21 .28 .29 .01 .02 ⫺.22 .26 .44
Note. Correlations in men’s data are presented above the diagonal, correlations in women’s data are presented below the diagonal, and between-spouse
correlations are shown in bold text on the diagonal. All correlations above .14 or less than ⫺.14 are significant at the p⬍.05 level. Relationship satis-
faction ⫽self-reported Marital Adjustment Test scores; positive relationship qualities ⫽self-reported Positive and Negative Quality in Marriage Scale
(PANQIMS) scores; negative relationship qualities ⫽self-reported PANQIMS scores; hostile conflict behavior ⫽self-reported Marital Coping Inventory
scores; emotional support ⫽partner-reported Support in Intimate Relationships Rating Scale (SIRRS) scores; affectionate behavior ⫽partner-reported
SIRRS scores; forgiveness toward partner ⫽self-reported Wade Forgiveness Measure scores; trait anger ⫽self-reported Multidimensional Anger Inventory
scores; physical aggression ⫽partner-reported four-item scale scores from the Conflict Tactics Scales–Revised.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
954 ROGGE, COBB, LAWRENCE, JOHNSON, AND BRADBURY

␤
11
⫽␥
110
⫹␥
111
ⴱ
(CARE) ⫹␥
112
ⴱ
(PREP) ⫹␥
113
ⴱ
(RA) ⫹␮
11
␤
12
⫽␥
120
⫹␥
121
ⴱ
(CARE) ⫹␥
122
ⴱ
(PREP) ⫹␥
123
ⴱ
(RA) ⫹␮
12
Although the models allowed treatment groups to vary on the
outcome measures at the start of treatment (␲
0
, corresponding to
␥
010
through ␥
023
), these contrasts yielded trivial differences and
in the interest of space are not discussed further.
Marital satisfaction. Across all conditions, MAT scores
dropped an average of 3.6 points per year for men and 2.3 points
per year for women (ns) following the workshops. Coefficients
testing for differences from this general trend within the CARE,
PREP, and RA groups failed to identify any significant differences
in slopes of MAT-rated satisfaction for men or women (see Table
5and Figures 2A and 2B). Analyses modeling linear change in
perceived negative relationship qualities separately within the
CARE, PREP, and RA groups failed to yield any consistent
patterns of linear change, while analyses modeling linear change in
positive relationship qualities suggested that men and women in
the CARE condition and women in the PREP condition (margin-
ally) declined more rapidly on this dimension (see Table 5). These
results are surprising in light of the focus on acceptance, support,
and empathy in CARE intervention, yet, as we outline below, this
may be part of a broader pattern of results in which interventions
produce effects that are incongruent with their intended targets of
change.
Negative and positive behavior. We tested three-level mod-
els to examine linear change in self-reported hostile conflict be-
havior, emotional support, affection, and aggression, allowing
results to freely vary across the CARE, PREP, and RA groups. As
shown at the top of Table 6, women in the CARE and RA groups
declined in hostile conflict following the workshops. Contrary to
hypotheses, PREP women did not decline in hostile conflict de-
spite the central role that regulation of negative behavior plays in
this intervention (see Figure 2C). Hostile conflict slopes for men
were also relatively flat for men in the PREP group, but this did not
distinguish them reliably from men in the CARE and RA groups
(see Table 6).
Turning to positive behaviors, women receiving CARE ex-
perienced reliable linear declines in emotional support, while
women in the PREP and RA groups did not change reliably.
Again, this runs directly counter to the prediction that CARE
would stabilize or even improve participants’ perceptions of
support. Results were similar but weaker for men (see Table 6).
Results for women’s changes in emotional support are graphed
in Figure 2D.
Table 3
Equivalence of Groups at Pretreatment
Variable CARE PREP RA NoTx Treatment contrast Effect size
Demographics
Number of couples 52 45 33 44
Age 28 (4.5) 30 (5.0) 29 (4.8) 27 (5.0) F(3, 344) ⫽6.2
ⴱⴱⴱ
.051
Years of education 15.2 (2.5) 15.1 (2.2) 15.2 (2.3) 14.6 (2.6) F(3, 344) ⫽1.5 .013
Caucasian 63% 56% 47% 50% ␹
2
(3) ⫽5.0 .119
Latino/a 15% 13% 26% 31% ␹
2
(3) ⫽11.1
ⴱⴱⴱ
.179
Asian/Pacific Islander 11% 16% 12% 5% ␹
2
(3) ⫽5.9 .130
African American 3.8% 7.8% 6.1% 3.4% ␹
2
(3) ⫽2.3 .081
Cohabiting 75% 78% 66% 70% ␹
2
(3) ⫽1.7 .103
Years cohabiting 2.5 (1.8) 2.3 (1.3) 2.3 (1.6) 2.9 (2.9) F(3, 123) ⫽0.6 .014
Children at pretreatment 23% 24% 18% 30% ␹
2
(3) ⫽1.4 .089
Number of children 1.8 (0.6) 2.1 (1.4) 1.8 (0.4) 1.8 (1.0) F(3, 38) ⫽0.3 .022
Divorced parents 47% 39% 36% 42% ␹
2
(3) ⫽2.3 .082
Pretreatment relationship quality
Relationship satisfaction 121 (17) 116 (23) 123 (20) 116 (22) F(3, 344) ⫽2.3 .020
Negative relationship qualities 3.7 (2.0) 3.9 (2.2) 3.3 (2.0) 4.0 (2.1) F(3, 344) ⫽1.4 .012
Positive relationship qualities 9.6 (0.6) 9.3 (1.0) 9.5 (0.7) 9.4 (1.0) F(3, 344) ⫽1.7 .015
Previous couples counseling 5.8% 15.6% 9.1% 11.4% ␹
2
(3) ⫽2.6 .122
Hostile relationship conflict 17 (10) 15 (9) 15 (10) 17 (11) F(3, 310) ⫽1.3 .012
Emotional support 19 (13) 16 (12) 20 (15) 21 (13) F(3, 341) ⫽1.9 .017
Affectionate behavior 21 (8.4) 19 (9.6) 21 (9.2) 20 (8.8) F(3, 343) ⫽1.5 .013
Forgiveness 76 (17) 76 (15) 77 (15) 78 (15) F(3, 344) ⫽0.3 .002
Trait anger 19 (8.7) 18 (9.1) 17 (12.4) 18 (9.6) F(3, 344) ⫽0.4 .003
Physical aggression 25% 26% 20% 30% ␹
2
(3) ⫽1.9 .075
Amount of physical aggression 2.3 (1.4) 2.1 (1.4) 1.9 (1.4) 2.6 (1.7) F(3, 84) ⫽0.7 .024
Note. Values in parentheses are standard deviations. CARE ⫽Compassionate and Accepting Relationships Through Empathy; PREP ⫽Prevention and
Relationship Enhancement Program; RA ⫽relationship awareness; NoTx ⫽no treatment; effect size ⫽partial eta squared coefficients for all analyses of
variance presented and phi coefficients for the chi-square analyses presented; relationship satisfaction ⫽self-reported Marital Adjustment Test scores;
negative and positive relationship qualities ⫽self-reported Positive and Negative Quality in Marriage Scale scores; hostile relationship conflict ⫽
self-reported Marital Coping Inventory Conflict subscale scores; emotional support ⫽partner-reported Support in Intimate Relationships Rating Scale
(SIRRS) emotional support subscale scores; affectionate behavior ⫽partner-reported SIRRS affection subscale scores; trait anger ⫽self-reported
Multidimensional Anger Inventory scores; physical aggression ⫽partner-report on presence of any of the four screening items from the Conflict Tactics
Scales–Revised (CTS-R); amount of physical aggression ⫽number of aggressive acts reported on the four CTS-R screening items among respondents
reporting any aggression.
ⴱⴱⴱ
p⬍.01.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
955
PREVENTING MARITAL DISTRESS AND DISSOLUTION

Table 6 also shows results for changes in affection. Perceived
affection tended to decline among women in all three groups, to
comparable degrees. Men in the CARE group declined reliably,
while their counterparts in PREP did not change, again directly
contradicting our prediction that an intervention focusing on pos-
itive interaction would stabilize affectionate behaviors. Changes in
affection for men in the RA group were similar to those for men in
the CARE group. These results are graphed in Figures 2E (for
men) and 2F (for women).
Follow-up analyses directly comparing groups showed that
females participating in CARE had steeper declines in partner-
reported emotional support (␥
121
⫽⫺2.18, p⬍.015) than
females in PREP (␥
120
[intercept] ⫽⫺0.05, p⬍.945). Declines
in partner-reported affection were comparable among females
in CARE, PREP, and RA and among males in CARE and
RA.
Finally, groups were compared on self- and partner-reported
aggression. Across the 3-year follow-up, 54% of CARE men, 59%
of PREP men, and 63% of RA men were aggressive toward their
partners at least once; these groups did not differ, ␹
2
(2) ⫽0.55,
p⬍.76. However, RA women were more aggressive over the
follow-up (88%) than CARE (49%) and PREP women (61%),
Table 4
Means (and Standard Deviations in Parentheses) of Outcome Measures Across Waves of Assessment
Variable
CARE PREP RA NoTx
Males Females Males Females Males Females Males Females
Sample size
T0 52 52 45 45 33 33 44 44
T1 52 52 44 44 33 33 — —
T2 40 40 31 30 21 21 — —
T3 36 36 31 30 26 26 34 34
T4 43 43 33 33 26 26 — —
T5 45 45 37 37 30 30 41 41
Global relationship satisfaction
T0 121 (15) 121 (18) 113 (25) 119 (20) 126 (19) 120 (21) 116 (24) 116 (19)
T1 120 (15) 122 (18) 116 (23) 118 (22) 127 (19) 118 (26) — —
T2 121 (16) 120 (22) 111 (23) 117 (23) 122 (20) 126 (19) — —
T3 119 (14) 119 (21) 115 (22) 123 (15) 118 (22) 123 (21) 116 (24) 117 (19)
T4 103 (20) 107 (22) 104 (21) 111 (14) 112 (16) 111 (22) — —
T5 104 (18) 104 (22) 103 (22) 109 (21) 110 (17) 109 (23) 100 (18) 98 (20)
Negative relationship qualities
T0 3.6 (1.9) 3.7 (2.1) 3.9 (2.1) 3.8 (2.2) 3.0 (1.8) 3.6 (2.2) 3.9 (2.1) 4.0 (2.1)
T2 4.2 (2.4) 3.9 (2.4) 3.8 (1.9) 4.3 (2.5) 3.2 (2.2) 3.2 (2.2) — —
T3 4.3 (2.4) 4.1 (2.5) 3.5 (2.0) 3.5 (1.9) 5.8 (2.5) 4.1 (3.1) — —
T4 4.5 (2.5) 4.4 (2.6) 4.0 (2.3) 3.4 (1.8) 2.6 (1.3) 3.4 (1.8) — —
T5 3.8 (2.0) 3.9 (2.3) 4.0 (2.3) 3.8 (1.9) 4.0 (2.3) 4.0 (2.5) — —
Positive relationship qualities
T0 9.5 (0.7) 9.6 (0.6) 9.3 (1.1) 9.4 (0.9) 9.4 (0.8) 9.6 (0.6) 9.2 (1.2) 9.6 (0.6)
T2 9.5 (1.1) 9.6 (0.6) 9.1 (0.7) 9.2 (0.9) 9.0 (1.4) 9.2 (1.4) — —
T3 9.3 (1.0) 9.3 (1.0) 9.3 (0.8) 9.6 (0.6) 9.3 (0.7) 9.5 (0.7) — —
T4 9.0 (1.3) 9.1 (1.1) 9.0 (1.1) 9.6 (0.7) 9.2 (0.8) 9.3 (0.7) — —
T5 9.2 (0.9) 9.3 (0.8) 9.0 (1.5) 9.0 (1.3) 9.2 (0.9) 9.3 (0.8) — —
Hostile conflict behavior
T0 15.9 (9.1) 19.0 (11.0) 13.3 (8.6) 16.9 (9.1) 12.9 (8.7) 16.8 (11.3) 13.6 (8.6) 19.8 (11.8)
T2 11.0 (6.4) 16.6 (10.2) 12.6 (7.0) 15.4 (7.3) 12.1 (9.4) 13.9 (9.8) — —
T3 16.5 (12.7) 17.7 (11.7) 10.8 (6.1) 13.7 (7.9) 13.2 (8.8) 14.6 (11.9) — —
T4 12.7 (8.4) 15.2 (8.5) 11.7 (7.7) 14.7 (6.5) 10.4 (8.4) 13.5 (10.3) — —
T5 12.2 (9.5) 14.5 (8.3) 13.6 (7.9) 14.5 (7.9) 11.6 (8.5) 13.7 (9.9) — —
Emotional support
T0 17.9 (11.1) 19.8 (14.8) 17.9 (14.7) 15 (8.2) 19.3 (15.7) 21.6 (15.3) 21.6 (13) 20 (12.7)
T2 21.5 (14.9) 11.5 (10.2) 14.1 (11.4) 15.6 (12.8) 17.5 (14.3) 16.8 (13.9) — —
T3 14.6 (12.9) 13.1 (11.3) 14.2 (12.3) 19.5 (13.9) 17.5 (15.8) 19.4 (14.8) — —
T4 16.8 (12.4) 11.6 (10.8) 14.4 (10.1) 18.6 (13.5) 16 (16.8) 16.5 (14.3) — —
T5 17 (10.6) 11.2 (10.5) 12.4 (11.1) 14.8 (13.1) 16.7 (17.3) 19.3 (15.6) — —
Affection
T0 20.9 (8.6) 20.4 (8.3) 17.8 (10.0) 19.2 (9.2) 21.5 (8) 20 (10.4) 21 (7.5) 20.8 (8.7)
T2 20.7 (9.0) 17.5 (10.2) 19.5 (8.8) 19.7 (9.0) 20.5 (9.8) 19.3 (9.8) — —
T3 16.5 (10.6) 18.0 (10.1) 17.5 (8.0) 18.1 (10.0) 18.2 (11.1) 22.6 (7.6) — —
T4 15.8 (9.8) 16.7 (9.7) 18.3 (9.5) 19.9 (9.9) 14.5 (10.7) 16.6 (10.7) — —
T5 16.1 (10.2) 18.1 (9.9) 16.4 (10.0) 14.4 (10.7) 14.4 (11.0) 17.1 (9.6) — —
Note. Dashes indicate instances when data were not collected from participants in the NoTx group. CARE ⫽Compassionate and Accepting Relationships
Through Empathy; PREP ⫽Prevention and Relationship Enhancement Program; RA ⫽relationship awareness; NoTx ⫽no treatment; T0 ⫽completion
of initial packet; T1 ⫽first day of workshop; T2 ⫽6 months after workshop completion; T3 ⫽1 year after workshop; T4 ⫽2 years after workshop;
T5 ⫽3 years after workshop.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
956 ROGGE, COBB, LAWRENCE, JOHNSON, AND BRADBURY

␹
2
(2) ⫽9.819, p⬍.007. Aggression was much more likely to be
psychological than physical and tended to involve yelling at or
insulting the partner or stomping out of the room during an
argument. As only the CARE and PREP couples received the
SAVE intervention, this suggests that SAVE may have been suc-
cessful at reducing women’s relational aggression.
Discussion
Educational interventions for couples hold great potential, but
enthusiasm for this approach is lessened by the absence of sup-
portive evidence from experiments with longer term follow-up
assessments and theoretically informative treatment conditions.
This study aimed to clarify the effects of preventive interventions
on couples’ self-reported relationship processes and outcomes,
through three types of comparisons: between couples trained in
communication skills and conflict management (PREP) versus
skills in acceptance, support, and empathy (CARE); between these
two groups and couples instructed in the importance of RA and
maintenance but not in relationship skills; and between all three of
these groups and couples who received no treatment (NoTx). Key
strengths of the study include a relatively large sample of couples
(N⫽174) in which 45% of the participants were ethnic minorities,
a 3-year follow-up, multiwave assessments of satisfaction, and
direct assessment of relationship dissolution. Satisfaction declined
over 3 years, and 16.3% of relationships ended, a figure compa-
rable to the 2-year dissolution rate of 18% reported by Christensen,
Atkins, Yi, Baucom, and George (2006) in their comparison of
therapies for distressed couples.
Three main results were obtained. First, there were no differ-
ences among couples in the PREP, CARE, and RA groups on rates
of dissolution and on a global index of relationship satisfaction.
Couples trained to improve basic communication and conflict
management skills experienced 3-year outcomes similar to couples
trained in prosocial aspects of intimate communication. More
surprisingly, RA couples—who received information about attend-
ing to and maintaining their relationships but no direct skills
training— had similar outcomes to couples who received skills
training.
Second, couples in the three active interventions had lower rates
of dissolution than NoTx couples. (This dissolution effect was
significant for those who completed the interventions but only
marginally significant for those assigned to the active groups.)
Specifically, couples completing any of the three active interven-
tions dissolved their relationship at about half the rate of couples
who did not receive an intervention (11% vs. 24%). However, on
the MAT, couples in the three active interventions did not differ
from untreated couples (see Figures 2A and 2B). Interpretation of
these findings pivots on the comparability of the NoTx couples to
those in the other three groups, a point that we address later in our
limitations section.
Third, although the two skill-based interventions were expected
to produce effects congruent with their principal focus (i.e., PREP
would generate better outcomes than CARE on negative behavior,
CARE would generate better outcomes than PREP on emotional
support and affection), no such pattern emerged. CARE partici-
pants declined in their perceptions of globally positive features of
their relationship (see Table 5), and as shown in Figure 2, observed
changes in behavioral reports were incongruent with the goal of
the interventions. On reports of negative behavior, women receiv-
ing CARE performed better (i.e., they declined significantly in
negative behavior) than women receiving PREP. On reports of
emotional support, women in PREP performed better (i.e., they did
not decline in perceived emotional support) than their counterparts
in CARE. Results for affection followed this same counterintuitive
pattern for men—PREP men did not decline on affection, whereas
CARE men did— but this pattern did not hold for women: Wom-
en’s perceptions of affection tended to decline regardless of which
intervention they received. Thus, the skill-based interventions did
not produce the expected differences in reports of relationship
behaviors.
Several factors limit interpretation of these results. First, al-
though the sample was ethnically diverse, partners were in their
first marriage, and most reported good incomes. These factors
limit generalizability to other segments of the population, and they
are likely to lower overall risk for relationship distress (cf. John-
son, 2012). Second, all data were collected via self-report mea-
sures. This may underestimate the effects of the interventions on
the communication variables (e.g., Fawcett, Hawkins, Blanchard,
Table 5
Predicting Slopes of Relationship Quality Over 3 Years
Model BSEpSDunits
Predicting global relationship
satisfaction over time
In men
Intercept (NoTx group) ⫺3.59
ⴱⴱ
1.79 .046 .15
CARE participants ⫺2.85 2.53 .148 .11
PREP participants ⫺1.83 2.03 .386 .07
RA participants ⫺1.64 1.93 .395 .07
In women
Intercept (NoTx group) ⫺2.34 2.66 .382 .07
CARE participants ⫺3.86 2.82 .174 .10
PREP participants ⫺2.57 2.78 .357 .07
RA participants ⫺1.98 22.76 .475 .05
Predicting negative relationship
qualities over time
In men
CARE participants .06 .10 .536 .03
PREP participants .01 .12 .932 .01
RA participants .14 .13 .298 .06
In women
CARE participants .15 .10 .677 .06
PREP participants ⫺.05 .12 .540 ⫺.03
RA participants .08 .14 .536 .04
Predicting positive relationship
qualities over time
In men
CARE participants ⫺.14
ⴱⴱ
.05 .004 ⫺.14
PREP participants ⫺.03 .06 .632 ⫺.06
RA participants ⫺.05 .07 .385 ⫺.05
In women
CARE participants ⫺.14
ⴱⴱ
.05 .004 ⫺.13
PREP participants ⫺.10
ⴱ
.06 .088 ⫺.08
RA participants ⫺.09 .06 .136 ⫺.08
Note.SD units were calculated by standardizing the outcome measures
within the Level 1 file (across men and women and across all waves of
assessment) and then using those as outcome variables in secondary
hierarchical linear modeling analyses to put the effects presented into SD
units. CARE ⫽Compassionate and Accepting Relationships Through
Empathy; PREP ⫽Prevention and Relationship Enhancement Program;
RA ⫽relationship awareness; NoTx ⫽no treatment.
ⴱ
p⬍.10.
ⴱⴱ
p⬍.05.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
957
PREVENTING MARITAL DISTRESS AND DISSOLUTION

& Carroll, 2010). and with the exception of the dissolution data,
conclusions are limited to subjective perceptions of relationship
functioning.
Finally, comparisons between the three active interventions and
the NoTx control condition are limited because the NoTx group
consisted of 44 couples who either declined their assignment to an
active treatment or who could not be scheduled for an active
treatment. These couples may have possessed some risk factor that
led them to resist an intervention (e.g., difficulty communicating,
uncertainty about the relationship, low commitment) that, in turn,
brought about distress and dissolution. We therefore cannot rule
out the possibility that differences involving this group are arti-
facts. Three points, however, argue against this possibility. First,
like all couples, NoTx couples volunteered to participate in a study
of couple workshops, completed an extensive set of questionnaires
prior to group assignment, and completed the follow-up assess-
ments. Second, at T0, across 11 demographic dimensions and 11
aspects of relationship functioning, the NoTx couples were not
distinct from couples in the other groups (see Table 3). Third,
although the difference in dissolution between the NoTx couples
(24%) and the other three groups (11%) over 3 years is notewor-
thy, there were no differences in rates of change on the MAT. If
the NoTx couples were at elevated risk for adverse outcomes, it
seems likely that more and stronger differences in relationship
satisfaction would have emerged. In short, although counterargu-
ments make it less plausible, unmeasured factors may be generat-
Figure 2. Model-predicted trajectories of relationship quality and behaviors. CARE ⫽Compassionate and
Accepting Relationships Through Empathy; NoTx ⫽no treatment; PREP ⫽Prevention and Relationship
Enhancement Program; RA ⫽relationship awareness.
†
p⬍.10.
ⴱ
p⬍.05.
ⴱⴱ
p⬍.01.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
958 ROGGE, COBB, LAWRENCE, JOHNSON, AND BRADBURY

ing differences between the NoTx group and the remaining groups.
If true, this would not alter comparisons among the three active
treatments.
The results of this study lend qualified support to the view that
engaged and newlywed couples can experience lasting benefits
from educational interventions, yet they also pose significant chal-
lenges for the theoretical basis of these programs. Although PREP
and CARE arise from different conceptual perspectives, both begin
with the assumption that changing interpersonal processes is nec-
essary for preventing distress and dissolution. The fact that the less
intensive RA intervention, which merely instructs couples in the
importance of relationship maintenance, produced outcomes that
were similar to those obtained with the more intensive PREP and
CARE programs indicates that skill training may be unnecessary
for the prevention of adverse outcomes. This finding is consistent
with the distinction that Snyder and Schneider (2002) drew be-
tween the acquisition or possession of skills and partners’ moti-
vation to implement them, and it suggests that interventions could
be improved by (a) helping couples to appreciate the skills they
already have, (b) encouraging them to see the value in using these
skills routinely to keep their relationship strong, and (c) instructing
them to attend to the cues that signal the need to employ these
skills. If we are correct in speculating that couples benefit from
attending to everyday signals that encourage relationship mainte-
nance, then it may prove useful to distribute minimally invasive
interventions that support this goal over longer spans of time.
In considering this recommendation, however, it is important to
acknowledge that we have little direct evidence that the RA
intervention worked as intended. The possibility remains that the
RA couples performed as well as the PREP and CARE groups
simply because all three groups were exposed to professional
contact and a credible intervention and all devoted effort to rela-
tionship improvement. Regardless of whether the RA treatment is
an active versus placebo intervention, the theoretical contributions
of future outcome studies are likely to be enhanced if they include
some form of minimal intervention instead of or in addition to a
true no-treatment condition. Neglecting to include the RA group in
this study would have resulted in the erroneous conclusion that
skills training was uniquely responsible for the observed group
differences.
A second set of implications follows from results showing that
CARE, PREP, and RA couples were not reliably different from the
NoTx group on satisfaction (MAT) trajectories whereas they were
reliably different in relationship dissolution. Declines in satisfac-
tion are likely to foreshadow dissolution in newlyweds (Karney &
Bradbury, 1997), but these findings suggest that educational inter-
ventions can produce different kinds of effects on these two
outcomes. Even when satisfaction is relatively high and stable,
recently married couples may end a relationship in the face of
certain events (e.g., the first episode of physical aggression) or
disclosures (e.g., one partner revealing a prior debt) that lead one
or both partners to question the marriage. Couples participating in
an active intervention may be less likely to experience these events
or more capable of managing these disclosures, thereby providing
them an advantage in 3-year dissolution rates.
Although CARE, PREP, and RA couples diverged over time in
reports of hostile conflict, emotional support, and affection, the
reported behavioral changes were generally incongruent with the
intent of the interventions. This is troubling because it indicates,
for example, that women trained in communication and conflict
resolution skills (PREP) do not decline in negative behavior
(whereas women trained in CARE and RA do) and that women
trained in empathic and accepting communication (CARE) decline
in emotional support (whereas PREP and RA couples do not).
Disjunctures between the intent and impact of skills-based inter-
ventions are not new. Couples can experience better relationship
outcomes in the absence of any behavioral change (e.g., low-risk
couples in Halford et al., 2001), and changes in relationship
satisfaction can be inconsistently related to within-group changes
in problem-solving behavior following participation in PREP
(Stanley, Rhoades, Olmos-Gallo, & Markman, 2007). However,
by comparing two skill-based interventions and by assessing a
range of skills, we discovered that PREP and CARE can fail to
produce intended effects and can alter behaviors that are not the
primary targets of change.
These results suggest that presenting couples with intensive
training in skills believed to be necessary for successful relation-
ships may inadvertently sensitize them to specific inadequacies in
their relationship. Intensive interventions like CARE and PREP
not only aim to teach couples particular sets of skills believed to be
essential for healthy marriage but also emphasize the importance
and value of these skills as criteria for evaluations of relationship
health. Doing so may have unintended consequences. For example,
Table 6
Predicting Slopes of Self- and Partner-Reported Relationship
Behaviors Over 3 Years
Model BSEpSDunits
Predicting hostile conflict
over time
In men
CARE participants ⫺.60 .37 .111 .12
PREP participants ⫺.11 .44 .801 .02
RA participants ⫺.50 .49 .306 .10
In women
CARE participants ⫺1.02
ⴱⴱⴱ
.39 .009 .20
PREP participants ⫺.18 .43 .674 .05
RA participants ⫺1.04
ⴱⴱ
.50 .038 .12
Predicting emotional support
over time
In men
CARE participants ⫺1.02
ⴱ
.61 .098 .13
PREP participants ⫺.74 .72 .305 .09
RA participants ⫺.88 .82 .284 .07
In women
CARE participants ⫺2.17
ⴱⴱⴱ
.60 .001 .30
PREP participants .32 .70 .654 .04
RA participants ⫺.94 .80 .243 .09
Predicting affection over time
In men
CARE participants ⫺1.80
ⴱⴱⴱ
.49 .001 .32
PREP participants ⫺.47 .58 .422 .06
RA participants ⫺2.35
ⴱⴱⴱ
.66 .001 .22
In women
CARE participants ⫺.83
ⴱ
.48 .087 .14
PREP participants ⫺1.18
ⴱⴱ
.57 .039 .16
RA participants ⫺1.34
ⴱⴱ
.65 .041 .14
Note. CARE ⫽Compassionate and Accepting Relationships Through
Empathy; PREP ⫽Prevention and Relationship Enhancement Program;
RA ⫽relationship awareness.
ⴱ
p⬍.10.
ⴱⴱ
p⬍.05.
ⴱⴱⴱ
p⬍.01.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
959
PREVENTING MARITAL DISTRESS AND DISSOLUTION

skill-based interventions might make couples’ skill deficits more
salient to them and give them reason to doubt their ability to
maintain the relationship. Being informed by experts that conflict
resolution is fundamental to a good relationship, for example, may
generate concern for one or both partners if actual conflict reso-
lution fails to reach this high standard. These possibilities are
speculative, of course, but we believe that the incongruity between
the intent and impact of the skills-based interventions is notable
and may undermine any benefits that these interventions can
provide.
In conclusion, regardless of whether they were trained inten-
sively in conflict management skills (PREP) or in the provision of
empathy and support (CARE), the engaged and newlywed couples
in this 3-year study experienced lower dissolution rates when
compared to couples who received no intervention. Surprisingly,
these findings fail to support skill-based interventions for prevent-
ing adverse marital outcomes because (a) couples instructed in a
single session to simply attend to and maintain their relationship
achieved nearly identical relationship outcomes and (b) PREP and
CARE couples performed more poorly in their targeted commu-
nication domains than in other domains that the programs did not
target. Revisiting the conceptual rationale for skill-based interven-
tions may be timely, and learning more about low-dose, low-cost
interventions may open new avenues for enabling more couples to
have fulfilling and enduring relationships.
References
Acitelli, L. K. (2001). Maintaining and enhancing a relationship by attend-
ing to it. In J. Harvey & A. Wenzel (Eds.), Close romantic relationships:
Maintenance and enhancement (pp. 153–167). Mahwah, NJ: Erlbaum.
Barry, R. A., Bunde, M., Brock, R. L., & Lawrence, E. (2009). Validity and
utility of a multidimensional model of received support in intimate
relationships. Journal of Family Psychology, 23, 48 –57. doi:10.1037/
a0014174
Bodenmann, G. (2005). Dyadic coping and its significance for marital
functioning. In T. Revenson, K. Kayser, & G. Bodenmann (Eds.),
Couples coping with stress: Emerging perspectives on dyadic coping
(pp. 33– 49). Washington, DC: American Psychological Association.
doi:10.1037/11031-002
Bowman, M. L. (1990). Coping efforts and marital satisfaction: Measuring
marital coping and its correlates. Journal of Marriage and the Family,
52, 463– 474. doi:10.2307/353040
Carrère, S., Buehlman, K. T., Gottman, J. M., Coan, J. A., & Ruckstuhl, L.
(2000). Predicting marital stability and divorce in newlywed couples.
Journal of Family Psychology, 14, 42–58. doi:10.1037/0893-3200.14
.1.42
Christensen, A., Atkins, D. C., Yi, J., Baucom, D. H., & George, W. H.
(2006). Couple and individual adjustment for 2 years following a ran-
domized clinical trial comparing traditional versus integrative behavioral
couple therapy. Journal of Consulting and Clinical Psychology, 74,
1180 –1191. doi:10.1037/0022-006X.74.6.1180
Dehle, C., Larsen, D., & Landers, J. E. (2001). Social support in marriage.
American Journal of Family Therapy, 29, 307–324. doi:10.1080/
01926180152588725
Donen, S. (Producer & Director). (1967). Two for the road [Motion
picture]. United Kingdom: 20th Century Fox.
Fawcett, E. B., Hawkins, A. J., Blanchard, V. L., & Carroll, J. S. (2010).
Do premarital education programs really work? A meta-analytic study.
Family Relations, 59, 232–239. doi:10.1111/j.1741-3729.2010.00598.x
Fincham, F. D., & Linfield, K. J. (1997). A new look at marital quality:
Can spouses feel positive and negative about their marriage? Journal of
Family Psychology, 11, 489 –502. doi:10.1037/0893-3200.11.4.489-502
Hahlweg, K., Markman, H. J., Thurmaier, F., Engl, J., & Eckert, V. (1998).
Prevention of marital distress: Results of a German prospective longi-
tudinal study. Journal of Family Psychology, 12, 543–556. doi:10.1037/
0893-3200.12.4.543
Halford, W. K. (2011). Marriage and relationship education. New York,
NY: Guilford Press.
Halford, W. K., Markman, H. J., & Stanley, S. (2008). Strengthening
couples’ relationships with education: Social policy and public health
perspectives. Journal of Family Psychology, 22, 497–505. doi:10.1037/
a0012789
Halford, W. K., Sanders, M. R., & Behrens, B. C. (2001). Can skills
training prevent relationship problems in at-risk couples? Four-year
effects of a behavioral relationship education program. Journal of Fam-
ily Psychology, 15, 750 –768. doi:10.1037/0893-3200.15.4.750
Hawkins, A. J., Blanchard, V. L., Baldwin, S., & Fawcett, E. B. (2008).
Does marriage and relationship education work? A meta-analytic study.
Journal of Consulting and Clinical Psychology, 76, 723–734. doi:
10.1037/a0012584
Jacobson, N. S., & Christensen, A. (1996). Integrative couple therapy:
Promoting acceptance and change. New York, NY: Norton.
Jacobson, N. S., & Margolin, G. (1979). Marital therapy: Strategies based
on social learning and behavior exchange principles. New York, NY:
Brunner/Mazel.
Johnson, M. D. (2012). Healthy marriage initiatives: On the need for
empiricism in policy implementation. American Psychologist, 67, 296 –
308. doi:10.1037/a0027743
Johnson, M. D., Cohan, C. L., Davila, J., Lawrence, E. L., Rogge, R. D.,
Karney, B. R.,...Bradbury, T. N. (2005). Problem-solving skills and
affective expressions as predictors of change in marital satisfaction.
Journal of Consulting and Clinical Psychology, 73, 15–27. doi:10.1037/
0022-006X.73.1.15
Karney, B. R., & Bradbury, T. N. (1997). Neuroticism, marital interaction,
and the trajectory of marital satisfaction. Journal of Personality and
Social Psychology, 72, 1075–1092. doi:10.1037/0022-3514.72.5.1075
Locke, H. J., & Wallace, K. M. (1959). Short marital adjustment prediction
tests: Their reliability and validity. Marriage and Family Living, 21,
251–255. doi:10.2307/348022
Markman, H. J., Renick, M. J., Floyd, F. J., Stanley, S. M., & Clements, M.
(1993). Preventing marital distress through communication and conflict
management training: A 4- and 5-year follow-up. Journal of Consulting
and Clinical Psychology, 61, 70 –77. doi:10.1037/0022-006X.61.1.70
Markman, H. J., Stanley, S., & Blumberg, S. L. (1994). Fighting for your
marriage. San Francisco, CA: Jossey-Bass.
McCullough, M. E., Worthington, E. L., Jr., & Rachal, K. C. (1997).
Interpersonal forgiving in close relationships. Journal of Personality and
Social Psychology, 73, 321–336. doi:10.1037/0022-3514.73.2.321
Neidig, P. H. (1989). Stop Anger and Violence Escalation (SAVE) instruc-
tor’s guide. Beaufort, SC: Behavioral Science Associates.
Raudenbush, S. W., Bryk, A. S., & Congdon, R. (2004). HLM6 for
Windows [Computer software]. Lincolnwood, IL: SSI.
Rogge, R. D., Cobb, R. J., Story, L. B., Johnson, M. D., Lawrence, E.,
Rothman, A. D., & Bradbury, T. N. (2006). Recruitment and selection of
couples for intervention research: Achieving developmental homogene-
ity at the cost of demographic diversity. Journal of Consulting and
Clinical Psychology, 74, 777–784. doi:10.1037/0022-006X.74.4.777
Rogge, R. D., Johnson, M. D., Lawrence, E., Cobb, R., & Bradbury, T. N.
(2002). The CARE program: A preventive approach to marital interven-
tion. In A. S. Gurman & N. S. Jacobson (Eds.), Clinical handbook of
couple therapy (3rd ed., pp. 420 –435). New York, NY: Guilford Press.
Schumm, W. A., Nichols, C. W., Schectman, K. L., & Grinsby, C. C.
(1983). Characteristics of responses to the Kansas Marital Satisfaction
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
960 ROGGE, COBB, LAWRENCE, JOHNSON, AND BRADBURY

Scale by a sample of 84 married mothers. Psychological Reports, 53,
567–572. doi:10.2466/pr0.1983.53.2.567
Siegel, J. M. (1986). The Multidimensional Anger Inventory. Journal of
Personality and Social Psychology, 51, 191–200. doi:10.1037/0022-
3514.51.1.191
Snyder, D. K., & Schneider, W. J. (2002). Affective reconstruction: A
pluralistic, developmental approach. In A. S. Gurman & N. S. Jacobson
(Eds.), Clinical handbook of couple therapy (3rd ed., pp. 151–179). New
York, NY: Guilford Press.
Stanley, S. M., Rhoades, G. K., Olmos-Gallo, P. A., & Markman, H. J.
(2007). Mechanisms of change in a cognitive behavioral couples pre-
vention program: Does being naughty or nice matter? Prevention Sci-
ence, 8, 227–239. doi:10.1007/s11121-007-0071-8
Straus, M. A., Hamby, S. L., Boney-McCoy, S., & Sugarman, D. B. (1996).
The revised Conflict Tactics Scales (CTS2): Development and prelim-
inary psychometric data. Journal of Family Issues, 17, 283–316. doi:
10.1177/019251396017003001
Received August 13, 2011
Revision received May 14, 2013
Accepted July 22, 2013 䡲
Additional Journal Information
Copyright and Permission: Those who wish to reuse
APA-copyrighted material in a non-APA publication must
secure from APA written permission to reproduce a journal
article in full or journal text of more than 800 cumulative
words or more than 3 tables and/or figures. APA normally
grants permission contingent on permission of the author,
inclusion of the APA copyright notice on the first page of
reproduced material, and payment of a fee of $25 per page.
Libraries are permitted to photocopy beyond the limits of
U.S. copyright law: (1) post-1977 articles, provided the
per-copy fee in the code for this journal (0022-006X/13/
$12.00) is paid through the Copyright Clearance Center,
222 Rosewood Drive, Danvers, MA 01923; (2) pre-1978
articles, provided that the per-copy fee stated in the Pub-
lishers’ Fee List is paid through the CopyrightClearance
Center. Formore information along with a permission re-
quest form go to: www.apa.org/about/contact/copyright/
index.aspx
Electronic access: Individuals subscribers to this journal
have automatic access to all issues of the journal in the
PsycARTICLES® full-text database. See http://www.apa
.org/pubs/journals.ccp and click on View Table of Con-
tents.
Reprints: Authors may order reprints of their articles from
the printer when they receive proofs.
Single Issues, Back Issues, and Back Volumes: For in-
formation regarding single issues, back issues, or back
volumes, visit www.apa.org/pubs/journals/subscription-
s.aspx or write to Order Department, American Psycholog-
ical Association, 750 First Street, NE, Washington, DC
20002-4242; call 202-336-5600 or 800-374-2721.
Subscription Claims Information: A claim form to assist
members, institutions, and nonmember individuals who have a
problem with their subscription is available at http://forms.apa
.org/journals/subclaim/; or call 1-800-374-2721.
Microform Editions: For information regarding micro-
form editions, write to University Microfilms, Ann Arbor,
MI 48106.
Change of Address: Send change of address notice and a
recent mailing label to the attention of Subscriptions De-
partment, APA, 30 days prior to the actual change of
address. APA will not replace undelivered copies resulting
from address changes; journals will be forwarded only if
subscribers notify the local post office in writing that they
will guarantee periodicals forwarding postage.
APA Journal Staff: Susan J. A. Harris, Senior Director, Jour-
nals Program; John Breithaupt, Director, Journal Services;
Paige W. Jackson, Director, Editorial Services; Tyler Krupa,
Lead Editor; Richard Pepple, Account Manager; Jodi Ashcraft,
Director, Advertising Sales and Exhibits.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
961
PREVENTING MARITAL DISTRESS AND DISSOLUTION