ArticlePDF Available

Controversies in Couple Relationship Education (CRE): Overlooked Evidence and Implications for Research and Policy

Authors:

Abstract

Government interest in strengthening families in the United States and Great Britain has contributed to a rapid growth in couple relationship education (CRE) interventions, with a recent increase in programs for low-income families. We describe the policy contexts that initially led to increased support for CRE in both countries but now threaten its continuation. We summarize meta-analytic research and discussions of single studies by authors who draw opposing inferences about CRE effectiveness, often from the same studies. We discuss three sets of findings not featured in previous reviews, all of which focus on the potential benefits of CRE for the well-being of children. First, without intervention, average couple relationship satisfaction declines, with negative consequences for children. Second, including both parents in father involvement and parenting interventions results in value-added contributions to family functioning. Third, we describe 9 CRE intervention trials that include child outcomes, 8 of which support the assumption that CRE benefits children. These studies represent only a first step in determining what happens to children when their parents experience CRE. From both research and policy perspectives, there are too many positive findings to give credence to the claim that CRE programs should be discontinued and funding directed elsewhere. But there are too many negative findings to support the conclusion that CRE has been an unqualified success and that existing programs should be offered more widely. The negative findings and criticisms have much to teach us about potential modifications to CRE programs that will lead to more effective CRE approaches . Keywords: couple relationship education, impact on couples, fathers, and children, moderator variables in meta-analyses, policy implications
Controversies in Couple Relationship Education (CRE): Overlooked
Evidence and Implications for Research and Policy
Philip A. Cowan and Carolyn Pape Cowan
University of California, Berkeley
Government interest in strengthening families in the United States and Great Britain has contributed to
a rapid growth in couple relationship education (CRE) interventions, with a recent increase in programs
for low-income families. We describe the policy contexts that initially led to increased support for CRE
in both countries but now threaten its continuation. We summarize meta-analytic research and discus-
sions of single studies by authors who draw opposing inferences about CRE effectiveness, often from the
same studies. We discuss three sets of findings not featured in previous reviews, all of which focus on
the potential benefits of CRE for the well-being of children. First, without intervention, average couple
relationship satisfaction declines, with negative consequences for children. Second, including both
parents in father involvement and parenting interventions results in value-added contributions to family
functioning. Third, we describe 9 CRE intervention trials that include child outcomes, 8 of which support
the assumption that CRE benefits children. These studies represent only a first step in determining what
happens to children when their parents experience CRE. From both research and policy perspectives, there
are too many positive findings to give credence to the claim that CRE programs should be discontinued and
funding directed elsewhere. But there are too many negative findings to support the conclusion that CRE
has been an unqualified success and that existing programs should be offered more widely. The negative
findings and criticisms have much to teach us about potential modifications to CRE programs that will
lead to more effective CRE approaches.
Keywords: couple relationship education, impact on couples, fathers, and children, moderator variables
in meta-analyses, policy implications
Since the 1950s, and increasingly over the last two decades,
concerns about the state of family life in the United States and the
United Kingdom have led to the creation and proliferation of
programs designed to strengthen couple relationships and increase
fathers’ positive involvement with their children (P. A. Cowan,
Cowan, & Knox, 2010; Panter-Brick et al., in press). We begin
with an account of the U.S. and U.K. policy contexts that sup-
ported funding for intervention programs for couples, especially
for those who are parents. The focus of this article is not on
couples therapy but on the various prevention programs summa-
rized under the general heading of Couple Relationship Education
(CRE). Over the past decade, the question of whether CRE pro-
grams produce the desired effects has been the subject of many
conflicting research articles, media stories, and Internet blogs,
which make policy decisions challenging at best. A careful reading
of the arguments suggested to us that the field was suffering from
selective accounts of published data, often from the same studies,
cited to support opposing conclusions concerning the benefits of
CRE. Those who support CRE point to data showing small but
significant effects on participants, at least in the short run, whereas
those who criticize CRE focus on the small effects and the fact that
the benefits often disappear over time. Policymakers who are
willing to turn to systematic studies for evidence before making
decisions about funding family programs are understandably con-
fused. Public dialogue about the usefulness of CRE consists of a
muddle of statements and counterstatements. We believe that this
dialogue can be enhanced by paying more attention to the details—
research designs and measurement approaches, specific outcomes
in specific target populations, and the size of the intervention
effects associated with each.
Although assessments of CRE studies rely increasingly on the
results of meta-analyses, we suggest that combining a wide variety
of CRE studies to obtain a few estimates of effect size glosses over
or distorts critical information about the various factors that lead
some programs to be effective and others to show few benefits for
the participants. We conclude that we are not at the stage in which
definitive generalizations can be drawn about whether CRE pro-
grams “work” until we can answer specific questions about how
variations in central program features are related to outcomes for
mothers, fathers, and children. We make suggestions for new
studies that that will inform policy decisions about strengthening
relationships among family members in ways that foster parents’
and children’s development.
The Policy Context of CRE Programs
In the last half of the 20th century and continuing to the present
day, concerns have been raised about changes in the structure and
This article was published Online First September 8, 2014.
Philip A. Cowan and Carolyn Pape Cowan, Department of Psychology,
University of California, Berkeley.
Correspondence concerning this article should be addressed to Philip A.
Cowan, Professor of Psychology, Emeritus, University of California,
Berkeley, 140 Highland Boulevard, Kensington, CA 94708. E-mail:
pcowan@berkeley.edu
Psychology, Public Policy, and Law © 2014 American Psychological Association
2014, Vol. 20, No. 4, 361–383 1076-8971/14/$12.00 http://dx.doi.org/10.1037/law0000025
361
quality of family life by social analysts, social scientists, family
service providers, family clinicians, and politicians at all levels of
government. Although there is controversy about the interpretation
of social trends such as increasing divorce, single parenthood,
father absence, cohabitation, and declining rates of childbearing
(Coontz, 2005; Parke, 2013), the shift away from so-called tradi-
tional family structures has often been interpreted as evidence that
contemporary families are in a state of decline (Popenoe, 1993) in
ways that place children at increased risk for developmental prob-
lems. In 1996, when the U.S. Temporary Assistance for Needy
Families (TANF) program was established with passage of the
Personal Responsibility and Work Opportunity Reconciliation
Act, one of its four goals was “to encourage the formation and
maintenance of two-parent families” (p. 8).
Influenced by the high levels of concern about families in
general, and impressed by the U.S. Fragile Families study findings
from a 20-city study of low-income women giving birth (Harknett,
Hardman, Garfinkel & McLanahan, 2001), which revealed that a
majority of fathers were still in romantic relationships with these
women around the transition to parenthood, the first Bush Admin-
istration planned, and Congress authorized, awards of $150 million
per year for 5 years for healthy marriage and responsible father-
hood programs within the Deficit Reduction Act of 2005, a reau-
thorization of the TANF program. The Administration for Children
and Families of the U.S. Department of Health and Human Ser-
vices distributed Healthy Marriage and Responsible Fatherhood
Grants to state, local, and community-based service providers to
address both marital instability and father absence. The Obama
administration continued support for these activities with an addi-
tional round of grants distributed in 2010 and another round
planned in 2014 or 2015.
The demographic family changes documented in the United
States have also been occurring in families in the United Kingdom
over the same period of time, with an even greater movement away
from marriage, a large increase in the incidence of single women
becoming mothers, and increases in both cohabitation and divorce
(Statistics, 2007). At first, the policy response by the U.K. gov-
ernment was similar to that of the U.S. In 1996, government
revisions of family law mandated the giving of grants for marriage
support and the publication of the Marriage Support Services
Directory, but the trajectory leading toward the support of couple
relationship stability was not as linear as it has been in the United
States. A major policy shift instituted by the British Labour gov-
ernment resulted in decreased support for services for couples
(Boucher, 2008), and Every Parent Matters, a government policy
document issued by the British Department for Education and
Skills (2007), set out a new series of plans and programs for
families that focused more on parenting than on the relationship
between the parents, with the goal of increasing the involvement of
fathers in children’s lives, and providing parenting skills training
for parents. More recently, U.K. policy directions shifted once
again as the Conservative–Liberal coalition government returned
to a focus on promoting marriage, strengthening couple relation-
ships, and encouraging fathers’ family involvement.
Based on similar demographic changes in the structure of fam-
ilies, the United States and the United Kingdom have currently
arrived at similar policy decisions by somewhat different routes to
fund programs that will increase family relationship quality and
stability by strengthening couple relationships and increasing fa-
thers’ family involvement. The main hope in each country was to
reduce the incidence of single parenthood and relationship breakup
in order reduce poverty and create more supportive environments
for children.
We acknowledge at the outset that as creators or co-creators of
three programs that embody elements of both couple strengthening
and father-involvement interventions (C. P. Cowan & Cowan
(2000); P. A. Cowan, Cowan, Ablow, Johnson & Measelle (2005);
P. A. Cowan, Cowan, Pruett, Pruett & Wong (2009), we have
some stake in the conclusions reached about program outcomes.
1
In this review, we have made every effort to recognize conflicting
opinions and to present a differentiated picture of an emerging
field in which there is enormous heterogeneity of programs and
results that defy simple aggregation in meta-analyses. We find
evidence for both optimism and skepticism about the impact of
CRE programs, and ambiguities that lead to a host of as-yet
unresolved questions. We conclude that the evidence produced so
far, although not always demonstrating clear and long-lasting
intervention effects, reveals promising trends. Relatively new re-
search, not previously included in discussions of CRE effective-
ness, points to the value-added contributions to family life of
couple-focused interventions. Also left out of current CRE con-
troversies are studies showing that children benefit when their
parents participate in interventions to strengthen their relationships
as couples or coparents. This idea has been used as a justification
for providing these programs, but has rarely been tested in sys-
tematic research. We describe eight studies with nine intervention
trials in which all but one show positive effects on one or more
measures of the children’s adaptation. We argue that these findings
help to justify further efforts to design research that will provide a
more nuanced picture of what kinds of CRE interventions work for
which target populations.
The Current State of Controversies About CRE
Couple relationship strengthening programs emerged in the
1960s along with many other preventive interventions, stimulated
by a zeitgeist of political and social science optimism that early
interventions could prevent individuals and families from later
distress and dysfunction. Initially, most programs were targeted to
middle-class couples early in their marital careers and not already
in marital or relationship distress (e.g., engaged, recently married,
making the transition to parenthood). More recently, there has
been a concern with providing services for couples who are at risk
by virtue of low incomes but not otherwise identified as experi-
encing serious relationship difficulties.
CRE interventions typically involve group meetings, ranging
from a single informational presentation, to weekend workshops,
to ongoing weekly meetings spread over 3 or 4 months. Most CRE
programs have focused on improving couple communication, al-
though a very few also focus on parents’ individual well-being,
coping with stress, parenting, and/or intergenerational issues. With
the exception of a discussion of the results of adding fathers to
therapeutic interventions for mothers and children, we have not
1
Our stake is intellectual, not financial. We have not been involved in
any way in merchandising the interventions that we have created or
benefitting financially from the sales of material or curricula created by
other CRE programs.
362 COWAN AND COWAN
included studies of therapy for troubled couples or therapy with
parents of a child in treatment for diagnosed behavior problems
(e.g., attention-deficit hyperactivity disorder, aggression), because
the risk status of the participants, the intervention formats, and the
outcomes are not comparable with those of CRE, which focus
mainly on prevention and typically use a group rather than an
individual or couple-by-couple format. Nor did we include evalu-
ations of parenting classes, because they range widely in content,
length, and leaders’ qualifications; are attended primarily by moth-
ers; and rarely deal with couple relationship issues.
There have been thousands of couples groups conducted over
the last 50 years,
2
with the bulk of them springing up in the last
two decades. The vast majority produced no published reports of
program effects, either pro or con, including almost all of the
programs funded in 2005 and 2010 by the Administration for
Children and Families (for two important exceptions, see Hsueh et
al, 2012; Wood, McConnell, Quinn, Clarkwest, & Hsueh, 2010).
From meta-analytic studies, we estimate that there may be as many
as 150 systematic studies of CRE, with about one third to one half
being doctoral dissertations. The number of studies with random
control trial (RCT) designs is quite small, and most meta-analyses
include no more than 25 studies with follow-ups beyond an im-
mediate posttest, and fewer than 10 with follow-ups beyond 7
months postintervention.
Almost all studies cited in articles on CRE research were con-
ducted in the United States. There are isolated exceptions. CRE
programs based on the Prevention and Relationship Enhancement
Program (PREP) created and evaluated in the United States (Mark-
man, Stanley, & Blumberg, 2010), have also been evaluated in
Germany (Hahlweg & Richter, 2010), Denmark (Van Widenfelt,
Hosman, Schaap, & van der Staak, 1996), and Sweden (Eng-
sheden, Fabian, & Sarkadi, 2013). Programs with more emphasis
on emotional communication, conducted in France by Bodenmann
and colleagues (Bodenmann, Pihet, Cina, Widmer, & Shantinath,
2006) and in Australia by Halford and colleagues (Halford, Petch,
& Creedy, 2010), have also been described by their authors as
successful. The situation in the United Kingdom is puzzling. As a
result of the government’s recent allocation of £30 million
($50,400,000) for relationship support, many organizations with
long-standing interests in couples have expanded their programs
(e.g., Tavistock Centre for Couple Relationships, Relate, One Plus
One, Marriage Care), but until very recently, with the exception of
marriage preparation classes, most of their efforts have been de-
voted to couples counseling, therapy, or education on a couple-
by-couple basis for already troubled partners; all of these have
been reviewed as having quite successful outcomes (Spielhofer et
al., 2014). While the Spielhofer et al. (2014) and other reviews of
couple interventions in the United Kingdom (Chang & Barrett,
2008; Coleman & Glenn, 2009) often cite American CRE studies,
professionally led groups for couples with children have generally
not been offered in England until recently.
Reviews of CRE Programs
The most comprehensive qualitative summary of CRE programs
that we have seen (Markman & Rhoades, 2012) includes 31
studies, almost all of which are interventions with middle-class
families, and 21 of which were RCTs. Another review of 15
couples-focused programs (P. A. Cowan et al., 2010) included 13
RCTs, four with low-income participants. Halford and Bodenmann
(2013) reviewed 17 RCT studies with follow-ups of at least one
year. These and other literature reviews (e.g., Halford, 2011, who
includes a number of European intervention trials) concluded that
relationship education programs show small but statistically sig-
nificant effects on marital quality, measured mostly by partici-
pants’ self-reports, and in a few cases, by observer measures of
couple interaction quality. Halford and Bodenmann (2013) went
beyond the assertion that the preponderance of studies showed
significant intervention effects to assert that couples with elevated
modifiable risk factors benefited substantially, whereas benefits
for low-risk couples were not as clearly established.
Descriptive research reviews are valuable at the beginning of a
new field, or new directions in a field, to acquaint readers with
current research and intervention directions, and to provide a map
of what has been done and what still needs to be done. However,
descriptions of research studies contain a number of problems.
First, there is no way of knowing how selective the author has been
in reporting studies that support his or her point of view. Second,
there is the “file drawer problem,” in which studies that fail to find
statistical significance are less likely to be submitted and, if sub-
mitted, less likely to be published. Third, until very recently,
significance at the p.05 level has been accepted as the standard
for determining whether an observed effect is likely to be attrib-
uted to participation in the intervention or is simply a chance
finding, but until the increasing popularity of meta-analysis, there
was no way to provide an overall summary of how strong an effect
was attained by the average CRE intervention. Our review here
attempts to summarize what has been learned from seven meta-
analyses of couple relationship programs published during the past
11 years,
3
five by Alan Hawkins and his colleagues.
Meta-analyses begin with an exhaustive literature search of
adequately designed studies, including both published and unpub-
lished findings, followed by procedures that produce a statistic that
reflects the average effect size (typically expressed as Cohen’s d).
The inclusion of unpublished studies is intended to guard against
the “file drawer problem.” We have some concerns about how this
bias has been interpreted, which we discuss later in this article.
Articles that evaluate the merits of CRE usually refer to one or
two meta-analytic studies, either to assert support for the effec-
tiveness of couples-group interventions (e.g., Markman &
Rhoades, 2012), or to cast a skeptical eye on the endeavor by
arguing that the effect sizes are small and tend to disappear over
time (e.g., P. Cohen, 2014; Johnson, 2014). A more differentiated
discussion can be stimulated by examining this body of work in
detail and by considering the pattern of effect size results across
the studies, as we do in Table 1 and the discussion that follows. In
addition to presenting effect sizes for the impact of CRE interven-
tions on couple relationship satisfaction and communication, when
available, we include data concerning potential moderating vari-
ables that make a difference to the size of the obtained effect—
2
See http://www.smartmarriages.com/app/Directory.BrowsePrograms
(United States) and www.oneplusone.org.uk (United Kingdom).
3
There was one meta-analysis published before 2000 (Butler &
Wampler, 1999), but it contained 16 RCT studies of one intervention
approach—the Couple Communication Program. Only four of the studies
had been published. Effect sizes were given but not tested for statistical
significance.
363
CONTROVERSIES IN COUPLE RELATIONSHIP EDUCATION
Table 1
Overall and Moderator Effect Sizes From Seven Meta-Analyses of CRE
Target population
Immediate follow-up Posttest follow-up Source of data Dosage Publication
Low Moderate No Yes
Sat Com Sat Com Parent
report Obs. Sat Com Sat Com Sat Com Sat Com
Hawkins et al. (2008) Inclusive
Immediate OR
follow-up .36
ⴱⴱⴱ
(46) .44
ⴱⴱⴱ
(37) .31
(22) .45
ⴱⴱ
(18) .12 (8) .18 (6) .52
ⴱⴱ
(11) .70
ⴱⴱⴱ
(11) .03 (8) .70
ⴱⴱⴱ
Immediate AND
follow-up .24 (17) .54
ⴱⴱ
(13) .28 (17) .37 (13)
Blanchard et al. (2009) Inclusive 7 mos
.59
ⴱⴱ
(5) .14 (25) .83
ⴱⴱ
(19)
Hawkins et al. (2012) Inclusive .30
ⴱⴱⴱ
(69) .40
ⴱⴱⴱ
(75) .10 (35) .10 (21) .38
ⴱⴱⴱ
(64) .41
ⴱⴱⴱ
(70)
Carroll & Doherty
(2003) Premarital couples .99
ⴱⴱⴱ
(7) .64
ⴱⴱ
(7)
Fawcett et al. (2010) Premarital couples .22 (17) .45
ⴱⴱ
(13) .31
(10) 1.02
ⴱⴱ
(6) .02 (9) .05 (7) .58
(8) .99
ⴱⴱⴱ
(6)
Pinquart & Teubert
(2010) Transition to
parenthood .09
(23) .28
ⴱⴱ
(19) .12 (11) .29 (11) .04 (15) 1.0
ⴱⴱⴱ
(4) .01 (15) .27 (14) .28
(8) .32 (5) .12 (8) .08 (15) .04 (15) .36 (14)
Hawkins & Fackrell
(2010) Low-income
couples RCT
.25
(3) .25 (1) .33 (2)
Pre–post
.29
ⴱⴱ
(10) .41
ⴱⴱⴱ
(8) .26
ⴱⴱⴱ
(6) .34
ⴱⴱⴱ
(6)
Note. Sat Relationship satisfaction; Com Couple communication; Numbers in parens number of studies included in the meta-analysis.
p.05.
ⴱⴱ
p.01.
ⴱⴱⴱ
p.001.
364 COWAN AND COWAN
target population, timing of follow-up, source of data, dosage, and
whether the intervention results had been published at the time the
meta-analysis had been conducted. Although meta-analyses of
CRE programs usually include data from quasi-experimental and
single-sample pre–post studies, we include only the results of
studies with random assignment to experimental and control con-
ditions—with one exception, which is Hawkins and Fackrell
(2010). Of the 15 studies in Hawkins and Fackrell’s meta-analysis,
only three were RCTs, and so for this meta-analysis only, we
include the 12 pre–post design studies in our table.
Table 1 presents the effect size (Cohen’s d), significance level,
and numbers of studies included in the effect size calculation (k)of
each of seven meta-analyses. The top three meta-analyses (in
Table 1) aggregate a large number of studies of different partici-
pant populations. The next four include fewer studies and report
effect sizes of CRE interventions with targeted populations (pre-
marital couples, couples becoming parents, or low-income cou-
ples). Blank cells signify that there was no available information,
and merged cells represent cases in which couple satisfaction and
communication scores have been combined in the analysis. It is
conventional to regard a dof .2 as small, .5 as medium, and .8 as
large (J. Cohen, 1988).
Maintenance of intervention effects over time. The three
largest meta-analyses by Hawkins and colleagues (Blanchard,
Hawkins, Baldwin, & Fawcett, 2009; Hawkins, Blanchard, Bald-
win, & Fawcett, 2008; Hawkins, Stanley, Blanchard, & Albright,
2012) attempted to include a wide variety of CRE interventions,
but because of what was available in the literature at that time, the
participants were mostly young middle-class couples in the early
stages of forming their families. In the earliest meta-analysis,
Hawkins et al. (2008) showed that when studies were included that
had either an immediate or longer term follow-up, there were
statistically significant effects on immediate posttest relationship
satisfaction (d.31) and communication (d.45) in favor of the
intervention participants. A subset of these studies had both im-
mediate and follow-up posttests, that is, they reported results on
the same participants followed over time. In contrast with the
immediate posttests, the effect sizes of the longer term follow-up
assessments (satisfaction, d.28; communication, d.37) were
no longer statistically significant. Similarly, at the immediate
posttest, Pinquart and Teubert’s meta-analysis found small but
statistically significant effects of the intervention on couple rela-
tionship satisfaction (d.09) and communication (d.28), but
the longer term posttests failed to show statistically significant
results. These findings suggest that the early intervention effects
disappear over time.
The results of two meta-analyses that also contrasted immediate
and longer term posttests challenge the conclusion that CRE ef-
fects are not maintained over time. With a sample of premarital
couples enrolled in CRE, Carroll and Doherty (2003) found a large
effect of CRE on combined measures of satisfaction and commu-
nication immediately after the intervention (d.99), with a
smaller but still significant moderate-sized effect on follow-ups
later on (d.64). Blanchard et al. (2009) noted that the Hawkins
et al. (2008) meta-analysis had simply divided postintervention
assessments into immediate and follow-up, without specifying the
timing of the follow-ups, which was often very soon after the
immediate posttests. They identified five studies with follow-ups
more than 7 months after the completion of the intervention, and
found that the longer term effect size of .59 was statistically
significant and in the moderate range. These two meta-analyses,
then, could be cited in support of the conclusion that although
intervention effects may be smaller at longer term than at imme-
diate posttests, they are maintained for some not-well-specified
period of time.
The remaining three meta-analyses (Fawcett, Hawkins,
Blanchard, & Carroll, 2010; Hawkins & Fackrell, 2010; Hawkins
et al., 2012) did not compare immediate and longer term posttests,
but reported results only for posttests at some point beyond the end
of the interventions. All three found statistically significant effect
sizes for communication (ds.40, .45, and .41), and two also
found significant effects for relationship satisfaction (ds.30 and
.29). Going beyond measures of couple satisfaction and commu-
nication, Pinquart and Teubert (2010) reported large longer term
follow-up effects on relationship stability (d.81) and parental
mental health (d.81).
In sum, the data from seven meta-analyses of RCT studies of
couple relationship interventions suggest that immediate posttests
generally reveal small but statistically significant intervention ef-
fects on measures of couple satisfaction and communication, al-
though Carroll and Doherty’s (2003) immediate posttest effect is
large. In meta-analyses in which satisfaction and communication
effect sizes were reported separately, the intervention effects on
communication were larger than those for satisfaction, but none of
the investigators tested this difference for statistical significance
between the two outcomes. The question of whether CRE inter-
vention effects are maintained over time (i.e., whether the effect
size for longer follow-ups is statistically significant) has not been
resolved. Rather than searching for a “yes” or “no” answer, we
suggest that CRE researchers and program planners explore mod-
ifications that might prolong the immediate intervention impact
(e.g., adding a focus on coparenting, booster sessions, additional
interventions to reduce external stressors and increase social sup-
ports).
Moderator effects. The dstatistic represents an average of
effect sizes across studies. Rather than assuming that the variation
around the mean is attributable to “error,” all of the investigators
who report on CRE meta-analyses search for systematic factors
that might identify characteristics of interventions that work more
or less well (i.e., moderator effects). The results of moderator
analyses have not been given the attention they deserve. In addi-
tion to the timing of follow-ups, which we have discussed, poten-
tial moderators in the seven meta-analyses include the source of
data (self-report vs. observation), dosage (number of sessions or
hours of group meetings), qualifications or experience of the
interveners, and whether the study included in the analysis has
been published.
Of the three meta-analyses that examine differences in effect
size attributable to the source of data (Blanchard et al., 2009;
Fawcett et al., 2010; Pinquart & Teubert, 2010), the first two find
that parent reports show statistically insignificant effects, and the
third (Fawcett et al., 2010) shows a small, statistically significant
effect (d.31). By contrast, all three show large effect sizes on
observed communication skills (d.83, 1.02, and 1.02). In
addition, we have Pinquart and Teubert’s (2010) finding, cited
earlier, that there were large effects on variables not usually
measured in CRE studies (relationship stability and parental men-
tal health). Note that the results from these moderator analyses do
365
CONTROVERSIES IN COUPLE RELATIONSHIP EDUCATION
not distinguish between immediate and longer term follow-ups,
and so do not address the issue of maintenance of intervention
effects over time.
One might expect that more intense interventions (more hours of
group meetings) would produce stronger effects. Two of the meta-
analyses that examined this question in primarily middle-class
samples with heterogeneous target populations (Hawkins et al.,
2008, 2012) supported this expectation. Both reported nonsignif-
icant effects on couple satisfaction and communication at longer
term follow-ups for low-dosage interventions (up to 7 or 8 hr),
with significant moderate to high effects for satisfaction (d.52
and .38) and communication (d.70 and .41) in moderate-dosage
interventions (up to 20 hr). Pinquart and Teubert’s meta-analysis
of couples in transition to parenthood found no significant effect
sizes for low-dose interventions, and a significant effect only on
communication for high-dose interventions. Finally, the Hawkins
and Fackrell (2010) meta-analysis of CRE programs for low-
income families found nonsignificant effects when the sample of
three RCT studies was divided into one low-dose and two
moderate-dose interventions (too small a sample to draw conclu-
sions), and statistically significant effects for the six pre–post
design studies for both low- and high-dose interventions (d.26
and d.34). Here, as elsewhere, the findings are mixed. Across
meta-analyses, then, there is some indication that dosage may have
an effect on outcomes, but as we indicate in our suggestions for
future research, there are design issues in determining dosage
effects that have yet to be addressed.
Three meta-analyses compared effect sizes for published and
unpublished studies. Two (Fawcett et al., 2010; Hawkins et al.,
2008) found that the effect sizes for unpublished studies were
negligible (d.02 or .03), whereas the effect sizes for published
studies were moderate or large (d.70 and .99). Pinquart and
Teubert’s (2010) meta-analysis of interventions for couples in
transition to parenthood found no significant effects on measures
of couple satisfaction for either published or unpublished studies,
but did find that published studies reported a significant impact of
CRE on couple communication (d.36). These results suggest
that the overall estimates of CRE effect size that we have cited
would have been larger if only published studies had been included
in the meta-analyses.
More thought needs to be given to the file drawer problem
associated with the difference between published and unpublished
studies. The typical interpretation of a publication “bias” is that the
body of published studies presents an exaggerated picture of
intervention effectiveness because unpublished, unsuccessful in-
tervention studies stored in file drawers are hidden from public
view. Including unpublished studies in the meta-analyses is the
rationale for reducing the bias toward intervention effectiveness in
published studies. However, it is likely that there are also biases
associated with unpublished intervention studies not simply attrib-
utable to their lack of statistically significant findings. For exam-
ple, a high proportion of unpublished CRE studies are reported in
dissertations. Others may have been submitted and rejected. It is
possible that these studies are not administered as competently,
and do not use as sophisticated designs and measures as published
studies do. If this hypothesis is correct, the inclusion of unpub-
lished studies in meta-analyses would minimize estimates of the
effectiveness of adequately mounted CRE programs. One finding
from the Fawcett et al. (2010) meta-analysis supports this inter-
pretation: In a search for moderator effects, they found that al-
though both self-report and observational measures showed statis-
tically significant impacts, the effects on observational measures
were appreciably larger, and that the difference between published
and unpublished studies in their meta-analysis might be explained
by the fact that none of the unpublished studies (mostly disserta-
tions) used observational measures. It is possible, then, that many
of the studies residing in file drawers deserve to remain where they
are and not be used to dilute overall estimates of CRE effects. At
the least, all meta-analyses of CRE should report results separately
for published and unpublished studies, and report on indicators of
whether unpublished studies are equivalent in quality to those that
have been published.
One important moderator effect was tested only in a single
meta-analysis and was not listed in Table 1 for reasons of space.
Pinquart and Teubert’s (2010) meta-analysis of CRE interventions
for couples in transition to parenthood examined differences in
effect size for couples groups led by paraprofessionals and groups
led by leaders with professional clinical qualifications. Significant
effect sizes for both satisfaction (d.23, k11) and communi-
cation (d.56, k9) were found only when the leaders had
professional clinical training.
In sum, the pattern of moderator effect sizes for CRE interven-
tions is quite different from the overall findings of small effects
when the question is whether the interventions create measurable
differences between experimental and control groups. According
to the results summarized in Table 1, an array of effect sizes
ranging from low through medium to high is obtained with (a)
longer term follow-ups, (b) studies using observational measures,
(c) moderate rather than low-dose interventions, and (d) profes-
sionally trained group leaders. Several cautions come with this
general statement. First, the effect sizes for each moderator are
averages. Some of the interventions in the category with the more
positive characteristics do not produce desirable results. Second,
moderator analyses are performed one variable at a time. We need
studies that combine potential moderator variables to find out
whether there are additive or even synergistic effects of the mod-
erator effects we have described here and whether they combine
with other potential moderators that have yet to be examined
systematically.
Another possible caution in interpreting the fact that the same
moderator effects can be found in more than one meta-analysis is
that there is considerable overlap in the reference lists included in
the articles listed in Table 1. Table 2 describes the overlap in terms
of the proportion of references included in the later publications
(moving from left to right, or top to bottom) that had also been
cited in the earlier studies. The largest overlap (67%) is between
Hawkins et al. (2008) and Blanchard et al. (2009), essentially the
same research group publishing meta-analyses in adjacent years.
Although they included some new studies and dropped some from
the earlier article, the Blanchard et al. article did not simply repeat
the analysis of overall effect sizes for immediate and follow-up
posttests. Instead, they intended to make new points about (a) the
importance of how long the follow-ups were conducted after the
intervention was completed, and (b) the fact that observations of
the couple revealed greater intervention effect sizes than partner’s
self-reports about communication quality. The third meta-analysis
(Hawkins et al., 2012) did not provide a reference list of included
studies. The four remaining meta-analyses focused on specific
366 COWAN AND COWAN
target populations and ranged from moderate overlap to none at all.
While some of the replicated moderator results come from meta-
analysis publications with overlapping studies, others appear in
articles with different target populations and no studies in com-
mon. We conclude that there is no clear evidence that the pattern
of findings in Table 1 is primarily attributable to the overlaps
among studies included in the seven meta-analyses, but future
reviews and meta-analyses should pay attention to the proportion
of studies in their analyses that have already been included in
earlier reports.
Focus on Single Studies: Johnson’s Critique of CRE
Pointed questions about whether CRE programs are effective,
especially for low-income families, have been raised recently by
Johnson (2012, 2014), who asserted that the government should
not be funding marriage and relationship programs for low-income
families. His criticism, based primarily on his interpretation of
three large-scale, well-designed government-funded programs,
elicited equally critical responses in defense of CRE (Hawkins,
2014; Hawkins et al., 2013). In the exchanges between Johnson
and Hawkins, we can clearly see the polarization in the field.
Building Strong Families (BSF). The first large CRE pro-
gram, BSF (Wood, McConnell, Quinn, Clarkwest, & Hsueh,
2010), included three different couple relationship intervention
approaches conducted across eight U.S. sites, with 5,102 low-
income unmarried couples randomly assigned to intervention and
control conditions. Outside of their low-income status, couples
were not identified as having relationship difficulties. Note that
this intervention and the next one we describe, Supporting Healthy
Marriage, for low-income married couples, used support workers
in addition to the group experience to help partners with problems
associated with poverty. Johnson (2012, 2014) emphasized the fact
that BSF showed no statistically significant effects across the eight
sites. Hawkins and colleagues (2012) countered that Johnson failed
to acknowledge that the couples-group interventions did show
significant positive effects for African American participants
(more than half the sample) at the 15-month follow-up. Compared
with controls, African American couples increased their level of
support and affection, and their use of constructive and avoidance
of destructive conflict management techniques. At the 15-month
assessment, African American participants were also significantly
less likely to report infidelity and intimate partner violence (no
effect sizes given).
Hawkins and colleagues (2012) also pointed to the positive early
findings in Oklahoma, one of the eight BSF sites. In comparison
with couples in the control group, group participants in the Okla-
homa site reported significantly more satisfying couple relation-
ships (d.20), greater use of constructive conflict behavior (d
.17), less use of destructive conflict behavior (d.09), and higher
coparenting quality (d.12). Mothers reported fewer depressive
symptoms (d.22) and were more likely to report that fathers
provided substantial financial support (d.27). These details
explain why Hawkins and others were heartened by the initial
results, despite the lack of statistically significant overall effects at
the eight sites. Johnson (2014) responded that Hawkins et al.
omitted the fact that in another of the eight sites in BSF (Balti-
more), couples-group participants fared significantly worse than
the controls. He also noted that in a subsequent follow-up at 3
Table 2
Proportion of Overlap Between Adjacent Studies in the Table
kHawkins et al.
(2008) Blanchard et al.
(2009) Hawkins et al.
(2012) Carroll & Doherty
(2003) Fawcett et al.
(2010) Pinquart & Teubert
(2010) Hawkins et al.
(2010)
Hawkins et al. (2008) 117 67% 38% 54% 26% 0%
Blanchard et al. (2009) 97 38% 18% 13% 8%
Hawkins et al. (2012) —— —
Carroll & Doherty (2003) 13 38% 0% 0%
Fawcett et al. (2010) 28 0% 0%
Pinquart & Teubert (2010) 21 0%
Hawkins & Fackrell al. (2010) 12
367
CONTROVERSIES IN COUPLE RELATIONSHIP EDUCATION
years postintervention (Wood, Moore, Clarkwest, & Killewald,
2014), there were no longer significant effects on couple relation-
ship quality for African American participants or for couples at the
Oklahoma site.
Turning to another issue, Hawkins et al. (2012) noted that BSF
used an intent-to-treat analysis that included all couples assigned
to intervention and control groups, despite the fact that actual
attendance was very low, with only 55% of the couples attending
even one group meeting. Johnson (2012, 2014) countered that
intent-to-treat analyses are standard in the field and the only way
to avoid selection bias that would inflate intervention effects,
because couples who refuse to participate or who drop out early
remove potentially low-functioning participants from the interven-
tion condition. We agree with Johnson on merits of intent-to-treat
analyses, but if 45% of couples who come in for an initial inter-
view fail to attend the program offered at that site, there may be
implementation flaws that contributed to the fact that couples who
did attend received no measurable benefits.
Only two tiny rays of hope for CRE emerged at the final 3-year
follow-up of BSF families (Wood et al., 2014). Although the
impact on couple relationship satisfaction in the Oklahoma site did
not hold up over time, 3 years after the families entered the study,
the children of couples-group participants were significantly more
likely to have parents still together. Furthermore, as we show in the
section on child effects, across the eight sites, there was an overall
slight, but statistically significant, positive impact of CRE partic-
ipation on parents’ reports of children’s problem behavior.
There is no question that the BSF findings were a disappoint-
ment to the investigators and to those who advocate the wide-
spread dissemination of couple relationship interventions for low-
income families. Johnson’s (2012, 2014) view is that the fading of
even the few positive effects over time justifies throwing in the
towel on this kind of intervention, at least from the point of view
of large-scale government support. We believe that the failure of
this massive study to produce meaningful long-term change in the
participating couples is, understandably, responsible for much of
the current skepticism about government funding of CRE inter-
ventions.
Supporting Healthy Marriage (SHM). The other two exam-
ples cited by Johnson in his criticism of CRE (2014) are not as
on-point as his presentation implies. The SHM program (Hsueh et
al., 2012)
4
was a new RCT with 6,298 low-income married cou-
ples at eight sites assigned randomly to one of four couples-group
programs plus a family support worker, or a no-treatment control
condition. Johnson (2014) dismissed the results of this trial be-
cause it had only several “very small positive intervention effects”
(p. 302). Hawkins (2014) responded that although the intervention
effects were small, they were seen not only in self-reports at 12-
and 30-month follow-ups but also in observational data collected
at the first follow-up.
At the 12-month follow-up, compared with those in the control
group, program group members showed statistically significant
effects on higher levels of marital happiness (d.13), lower
levels of marital distress (no effect size given), greater warmth and
support, (d.09), more positive communication skills (d
men
.08; d
women
.11), and fewer negative behaviors and emotions in
their interactions with their spouses (d
men
.08; d
women
.12).
They also reported less psychological abuse (no differences in
physical abuse). The effect sizes for self-report measures at the
30-month follow-ups were equal to or slightly greater than those
obtained 18 months earlier.
Furthermore, observational measures of couple interaction
(done only at the 12-month follow-up) indicated that for program
couples, on average, both partners showed more positive commu-
nication skills (d
men
.10; d
women
.09), and wives showed
significantly less anger and hostility (d.10), than partners in the
control group. Taken together, the self-report and observational
measures of intervention effects suggest that the SHM program
changed not only the way participants viewed their marriages but
also the extent to which they were able to implement the skills
taught by the SHM curricula in ways that were observable by
others. Describing these effects as small does not entitle a critic to
dismiss them entirely.
Finally, in contrast with 55% of couples in the BSF study who
attended at least one meeting, 83% of the couples in the SHM
study attended at least one meeting, with an overall attendance rate
of just over 70%. This raises the question of whether the SHM
results were more positive than BSF, because the intervention was
addressed to married rather than unmarried couples, or whether the
differences resulted from lessons learned about implementation by
the later program designers from the earlier investigators. Wood et
al. (2014) speculate that the difference in populations could ex-
plain the results; cohabiting partners may be less committed to
each other, less trustful of each other as they enter the study, and
less willing and able to do the hard work of applying what is
learned in group meetings to their relationship. We lean toward an
explanation based on incremental progress in the field of CRE
implementation. Only future studies can enlighten us about this
issue.
Community Healthy Marriage Initiative. In our view, the
third program, the Community Healthy Marriage Initiative, a large
quasi-experimental program (Bir et al., 2012) cited by Johnson as
representing a failure of CRE, did not actually provide quantitative
assessments of CRE participation. Media messages and an array of
relationship skills classes and service referrals were delivered to
three targeted cities with large low-income populations, and results
were compared with those in three matched control cities (there-
fore, this was not an RCT). Surveys were administered to a
representative sample (more than 77,000 individuals) to assess
attitudes about marriage and family life, household and marital
status, quality of couple relationships, income and earnings, and
child well-being. There were no differences in the modest rates of
participation in relationship classes (6% to 8%) between targeted
and comparison cities, and no differences in any of the self-report
survey measures. Hawkins (2014) noted that the study had design
flaws and discussed it no further. We can add that the only direct
assessment of CRE participants in the Bir et al. study was con-
ducted in qualitative interviews with 750 individuals (not couples)
who attended a class, 80% of whom reported that the classes
improved their relationships with their partners, and 97% of whom
reported that they would recommend the classes to others. While
we do not believe that consumer satisfaction measures constitute
proof of CRE effectiveness, the limited data here are not consistent
with a “no effects on the participants” verdict. Our conclusion is
4
The authors were two of a large group of academic consultants on this
project.
368 COWAN AND COWAN
that this was a study of the impact on a community of a number of
programs designed to encourage healthy marriages, not an RCT to
evaluate the impact of CRE programs.
In terms of the controversy between Johnson and Hawkins, we
believe that Johnson was justified in raising the point about gov-
ernment spending in light of the failure of BSF to show overall
significant effects of CRE for unmarried couples. On the other
hand, we support Hawkins’s point that CRE should not be gener-
ally dismissed because of the failure of BSF with unmarried
couples, or based on the fact that the effects of SHM on married
couples were “small.” As we have noted, data from other single
studies and meta-analyses with moderators show effect sizes rang-
ing from small to medium, and occasionally large. The overall
pattern of findings indicates to us that it is time to stop arguing
about whether CRE programs are generally effective, and to find
out more about the characteristics of programs that have shown
positive effect on the couples who participate.
Critiques of CRE based on other small studies. The
Johnson–Hawkins et al. controversy is not the only one in which
questions have been raised about the efficacy of CRE. Findings
from three other intervention studies have been used to support
claims that CRE programs have no discernible effect (Doss,
Rhoades, Stanley, & Markman, 2009; Markman, Rhoades, Stan-
ley, Ragan, & Whitton, 2010; Rogge, Bradbury, Hahlweg, Engl, &
Thurmaier, 2006). The form of the three studies is very similar.
Doss et al. (2009) and Markman et al. (2010) followed up an
intervention study by Stanley et al. (2001), whereas the Rogge et
al. (2006) study is based on a German trial of PREP by Hahlweg
and colleagues (Hahlweg, Markman, Thurmaier, Engl, & Eckert,
1998). In the early follow-ups, both studies found statistically
significant intervention effects. In the later follow-up articles, the
authors predicted postintervention outcomes based on preinterven-
tion data, and included intervention participation as a control
variable. All three found that this statistical control did not alter the
strength of the predictions. That is, pretest data predicted posttest
data equally well for participants in the CRE and control groups.
From these findings, the conclusion has been drawn (e.g., Brad-
bury & Lavner, 2012) that the intervention has no effects, a
conclusion based on faulty logic. The finding of no intervention
effects on pre–post correlations is completely independent of
whether there are intervention effects on participants.
Four studies raise additional questions about whether couples
benefit from CRE. Laurenceau, Stanley, Olmos-Gallo, Baucom, &
Markman (2004) compared a university-based PREP program with
a religious organization PREP program and a no-treatment control,
and followed couples for 14 months after they entered the study.
Growth curve analyses revealed no effects of participation in any
of the programs on self-reported marital satisfaction; this has been
cited as a failure of CRE, but there were some effects of the
religious organization PREP compared with no-treatment controls
on wives’ observed communication (d.60) and positive behav-
ior (d.55).
Trillingsgaard, Baucom, Heyman, & Elklit (2012) offered Dan-
ish couples a PREP intervention for a total of 17.5 hr, an
information-based parent program (INFO) for a total of 11 hr, and
compared them 24 months later with couples receiving no inter-
vention. All three groups declined in marital satisfaction and there
were no differences in self-reported communication measures—a
blow to CRE advocates. This was not a random assignment study,
however, as the no-intervention controls were neither invited to
participate in this study nor informed about the possibility of
intervention.
Rogge, Cobb, Lawrence, Johnson, and Bradbury (2013) issued
a more serious challenge to CRE. Engaged and newlywed couples
(n174) were randomly assigned to a PREP intervention teach-
ing communication skills and conflict management, or a CARE
program teaching acceptance, support, and empathy—each con-
ducted in four meetings for a total of 15 hr. Both were compared
with couples receiving a one-session relationship awareness (RA)
intervention (instructions to watch and discuss four movies with
intimate relationship plots within the next month) and couples
receiving no intervention. Follow-ups conducted semiannually
over 3 years assessed relationship stability and satisfaction. There
were no differences among couples in the PREP, CARE, and RA
groups on rates of dissolution or on a global index of relationship
satisfaction. The surprising finding was that RA couples—who
received information about attending to and maintaining their
relationships but no direct skills training—had similar outcomes to
couples who received skills training. The fact that in the random-
ized part of the study (assignment to the three intervention condi-
tions) a 1-hr group meeting plus watching and discussing four
movies had the same impact on relationship satisfaction as a 15-hr
PREP or CARE intervention is disquieting for supporters of CRE.
Finkel and colleagues (Finkel, Slotter, Luchies, Walton, &
Gross, 2013) raise a similar issue about shorter interventions for
couples, this time without using a couples-group intervention. One
hundred twenty couples were enrolled in a study of marital quality
and marital conflict for 2 years, during which they were assessed
every 3 months, including a task to write about a recent argument
with their partner and to rate their level of anger. No interventions
were conducted during the first year. In the second year, half of the
couples were randomly assigned to a reappraisal condition, in
which they were asked at Months 12, 16, and 20 to write about a
recent conflict, describe it from the perspective of a third person,
and rate how angry they were. Growth curve analyses indicated
that all couples (on average) declined in marital quality during the
first year, but that the decline was halted for the intervention
participants during the second year. Both Rogge et al. (2013) and
Finkel et al. remind us that CRE is not the only way to prevent the
normative decline in relationship satisfaction. Further studies are
needed to study whether the different approaches to couples inter-
vention have different effects for different kinds of couples.
We have been discussing a set of descriptive reviews, meta-
analyses, and critiques of findings from single studies that have
been interpreted differently by CRE supporters and critics. We
believe that it is premature at this point to come to a definitive
conclusion about CRE’s general effectiveness for couples. Al-
though some studies do show a strong impact of couples-group
interventions on couple relationship quality, the overall average in
meta-analyses that include unpublished studies is consistently
small at immediate posttest and in question at longer term follow-
ups. However, it is also too soon to declare that CRE is moribund,
and that research and service funds should be allocated in other
directions. Enthusiasm for CRE has certainly been muted by the
failure of the large BSF intervention to produce measurable short-
term and long-term effects. The fact that some single studies do not
show significant effects, and a few others achieve positive out-
comes for couples with interventions much less complex than
369
CONTROVERSIES IN COUPLE RELATIONSHIP EDUCATION
CRE, does raise serious questions for advocates of couples-group
programs. In response, we have noted that some studies cited as
proof that CRE is not effective turn out to have flaws that reduce
the power of the criticism. Furthermore, tests of moderator effects
in meta-analytic studies indicate that some CRE programs do have
substantial effects. The main problem in evaluating the usefulness
of CRE is not that we lack research studies, but that for almost
every claim that CRE has positive or negative effects on couples,
there has been a counterclaim or a reason why the study or analysis
under discussion does not support the author’s position on CRE.
That is, the jury is out on the merits of CRE, in part because
supporters and critics are telling very different stories constructed
from the same events. To the extent that the differences arise from
ideological positions on the need for CRE programs, there may be
no way of coming to a mutually agreed-upon conclusion.
Studies Currently Missing From the CRE
Controversy: A Focus on Child Outcomes
Before concluding that the controversies concerning CRE are
unresolvable, we believe it is necessary to examine a number of
important studies that have not been considered in the debates
about CRE effectiveness. The disagreements about the effective-
ness of CRE we have described center on whether couples-group
participants fare better in relationship quality, and, occasionally, in
parenting skills, than nonparticipants. Missing from the discussion
so far is research concerning three important questions: What
happens to couples and their children without intervention? Does
a couples approach add value to interventions for mothers or
fathers? Do children benefit when couples, parents, or parenting
figures participate together in a CRE intervention?
What Happens to Couples Over Time
Without Intervention?
The implications of a well-accepted finding have often been
ignored in recent accounts of CRE interventions. In more than 50
studies in a number of industrialized countries, marital or couple
relationship satisfaction declines over time, even more quickly
after couples become parents (Twenge, Campbell, & Foster,
2003). This fact has been cited repeatedly as a justification for
creating and disseminating CRE programs. What has not been
brought into the discussion is that fact that correlational data
suggest that this decline increases risks for children’s cognitive,
social, and emotional development (e.g., Cummings & Davies,
2010), and therefore that interventions for couples could have a
salutary effect on their children. Of course, satisfaction does not
erode for all couples. Recent research suggests that couples ini-
tially highly satisfied decline least and those in distress decline
most (Lorber, Erlanger, Heyman & O’Leary, 2014). Because of
the effort and cost expended in the mounting of CRE programs, it
might be useful to think about identifying couples in distress and
providing interventions for those most in need.
Including Both Parents in Parenting Interventions
Adding fathers to interventions for mothers. The field of
CRE can learn from an emerging theme in studies of parenting
interventions, in which mothers are the usual “designated parent.”
Behavioral therapies for mothers of aggressive children have been
remarkably successful (Miller & Prinz, 1990), at least in the short
run. Three decades ago, confronted by substantial numbers of
families in which children reverted to baseline or did not change at
all, several parenting therapy programs in the United States and
Australia identified a potential explanation of the failures; in these
families, unresolved marital conflict was high. After recruiting
fathers and adding a new focus on coparenting and marital issues,
several studies found that a combined marital and parenting em-
phasis was more successful in reducing sons’ problem behaviors
than a parenting skills approach with mothers alone (Brody &
Forehand, 1985; Dadds, Sanders, Behrens, & James, 1987;
Webster-Stratton, 1985). Meta-analyses of several different types
of parent-training interventions by Lundahl and colleagues (Lun-
dahl, Tollefson, Risser, & Lovejoy, 2008) and Bakermans-
Kranenberg and colleagues (Bakermans-Kranenburg, van Ijzen-
doorn, & Juffer, 2003) came to the same conclusion. In future
studies, it would be useful to examine whether the increased
involvement of fathers with their children is responsible for the
positive intervention effects.
Two RCT studies illustrate the contribution made by fathers’
participation in parenting programs. A Canadian study (Besnard,
Capuano, Verlaan, Pulin, & Vitaro, 2009) that randomly assigned
participants to couples participation, solo mother participation, or
a control group found that the quality of mothers’ parenting
improved when both parents were involved. The second, in the
United States (Rienks, Wadsworth, Markman, Einhorn, & Etter,
2011), is the most relevant to our concerns with CRE because the
intervention involved group meetings with a CRE approach. The
Fatherhood Relationship and Marriage Education program, an
adaptation of PREP called “Within our Reach” and designed to
meet the needs of low-income couples (Stanley et al., 2006), also
included modules on parenting and coping with economic stress.
One hundred thirty-seven primarily low-income couples, diverse
in ethnic background, were recruited and randomly assigned to (a)
couples groups, (b) groups attended by only one of the parents, or
(c) a no-treatment control condition (average age: fathers 36,
mothers 31, focal child 9). Workshops were led by trained
leader pairs (usually male–female) selected for their professional-
ism and sensitivity regarding cultural and economic issues.
Of interest here is that father involvement remained stable in the
control group, increased in couples-group participants, and de-
clined when only mothers attended the groups and attempted to
convey the information to their partners at home. The effect size of
the father-involvement difference between couples and mothers
groups was d.34. The authors speculate that the mothers-only
condition may have placed the mothers in a gatekeeping role and
led to defensiveness in the fathers. The failure of the mothers-only
condition to promote father involvement is concerning, the authors
point out, because it is similar to most parenting interventions
attended only by mothers.
Adding mothers to interventions for fathers. Here we turn
the story around to examine what happens when mothers partici-
pate along with their partners in an intervention previously tar-
geted toward men. Interventions to enhance father involvement
began in the 1980s, but remarkably few received systematic eval-
uation. A diligent search by Avellar et al. (2011) found 150 studies
of “responsible fatherhood programs” since 1990—90 that in-
cluded low-income fathers, but only 15 that used well-accepted
370 COWAN AND COWAN
research design and measurement procedures. A meta-analysis
(Holmes, Galovan, Yoshida, & Hawkins, 2010) of 16 father-
involvement interventions for resident, low-income, married, or
cohabiting fathers showed an overall effect size of d.26,
suggesting that father-involvement interventions, like CRE inter-
ventions, produce small but statistically significant effects.
Regardless of whether relationship issues are introduced into
father-involvement programs, almost all that we are aware of work
with men, in group or individual sessions, with male group leaders
or individual counselors. The Supporting Father Involvement (SFI;
P. A. Cowan, Cowan, Pruett, Pruett, & Wong, 2009) program is
the only study in addition to Rienks et al. (2011) that used an RCT
research design to evaluate a couples approach to father involve-
ment, but SFI contrasted a couples group and a fathers-only group
with the same leaders and curricula. The SFI approach was influ-
enced by the consistent finding that across the economic spectrum,
the single best predictor of fathers’ family involvement is the
quality of the father’s relationship with the mother (Carlson,
Pilkauskas, McLanahan, & Brooks-Gunn, 2011), a finding that
holds for married, cohabiting, separated, and divorced coparents
(Pruett & Johnston, 2004).
During Phase I of the SFI study, parents who expressed interest
in the program were invited to take part in one of the following on
a randomly assigned basis, all with the same staff: a one-time
informational group meeting (3 hr; n98), a group for fathers that
met for 16 weeks (32 hr; n96), or a group for couples that met
for 16 weeks (32 hr; n95). All three variations of the SFI
program were delivered by clinically trained male–female pairs of
group leaders, and both fathers and mothers completed all assess-
ments. Every family was also offered the support of a case man-
ager or family worker to help with referrals to other services, as
needed, during their time in the project. The majority of the
families in Phase I of the SFI study were Mexican American, with
most of the remaining families White. Their youngest children
ranged in age from birth to 7 years, with a median age of 2.5 years.
Two thirds of the families had household incomes below twice the
federal poverty line.
Fathers and mothers who participated in the one-time meeting
showed no positive and some negative changes; for example, at the
18-month follow-up assessment, father involvement did not
change, and satisfaction as a couple declined significantly. Partic-
ipants in both the 16-week fathers and couples groups showed
significant increases in fathers’ involvement in the care of the
children. In addition, couples-group participants reported signifi-
cant reductions in parenting stress, and no decline in satisfaction as
a couple, unlike fathers-group and control-group participants. Ef-
fect sizes comparing the couples-group participants with those in
the single-meeting control group include increases in couple rela-
tionship satisfaction (d.36) and father involvement (d.34),
and reductions in parenting stress (d.31). A unique feature of
this study is that it also showed benefits for the children of
participants—a finding we discuss in the next section.
It seems clear that there is a value-added contribution of includ-
ing both parents when interventions hope to affect children. Com-
pared with parenting interventions for mothers, programs that
intervene with both parents show greater positive effects on family
relationships and children’s behavior. According to the only ex-
isting study to contrast work with couples and fathers alone (SFI),
couples groups have a more positive impact on the relationship
between the parents than fathers groups, although both enhance
men’s participation in children’s lives and appear to prevent a rise
in young children’s problem behaviors. These two findings speak
to the desirability of testing CRE interventions not only against
no-treatment controls but also against father-involvement interven-
tions and programs targeted to one parent at a time, the situation
for most parenting classes.
Do Children Benefit From Their Parents Participation
in CRE Interventions?
Almost all of the current dialogue about CRE effects has fo-
cused on the couples themselves, despite the fact that the primary
justification for funding these interventions rests on the argument
that benefits for parents would be also benefit their children. Based
on direct searches of the literature, research reviews, and meta-
analyses, we were surprised to find that of the more than 150
adequately designed CRE studies, we found only nine studies with
RCT designs and measures of how the participants’ children fared.
In our view, the question of whether children whose parents
participate in CRE interventions are better off than those who do
not is, or should be, central to policymakers and service providers
considering whether to allocate resources to couples-group inter-
ventions.
Our expectation that CRE interventions have positive effects on
children comes in part from family theory, and partly from em-
pirical studies that document correlations between couple func-
tioning or positive father involvement and children’s cognitive,
social, and emotional adaptation. Family systems models (Walsh,
2012) all suggest that positive changes in individual family mem-
bers, or in the relationships among them, set processes in motion
that result in improved adaptation for both adults and children. A
large number of studies document correlations between high levels
of unresolved conflict between the parents and their children’s
academic problems (see meta-analysis of multiple ages by Kitz-
mann, Gayford, Holt, & Kenny, 2003) and emotional distress or
behavior problems in infants and toddlers (McHale & Fivaz-
Depeursinge, 1999), preschoolers and school-age children (Cum-
mings & Davies, 2010; Grych & Fincham, 2001), and teenagers
(Emery, 1988). The parents’ conflict need not be loud and overt;
cold, distanced withdrawal in either or both parents also places
children at higher risk for problematic behavior. While much of
this research began in the United States, studies of U.K. families
provide similar findings (Grych, Harold, & Miles, 2003; Shelton &
Harold, 2008).
In the past decade, McHale and Lindahl (2011), among others
(Belsky & Jaffee, 2006), made a distinction between two aspects
of couple relationships: the interaction between partners around
their intimate relationship, and the ways partners work as a team,
or fail to do so, as they interact with the child. More collaborative,
less competitive coparenting is associated with positive outcomes
for children. Taken together, these findings would lead us to
expect that interventions to foster more effective couple and co-
parenting relationships would have important consequences for
their offspring, but before we accept the assumption of causal
linkage, this expectation must be tested empirically in the context
of RCT intervention studies to show that the impact of the inter-
vention on couple, coparenting, or parent–child relationships is
actually associated with beneficial outcomes for the child.
371
CONTROVERSIES IN COUPLE RELATIONSHIP EDUCATION
Of the nine CRE trials that included assessments of child out-
comes, three attempted to enlist partners making the transition to
parenthood for the first time, and the rest began with parents of
children in the preschool and early elementary school years. Meta-
analyses have been done with even fewer than nine studies, but we
have chosen not to combine this highly heterogeneous set of
studies into one or two overall summary statistics that describe the
average impact of parents’ participation on the child. It is true that
six of the nine studies contain measures of internalizing or exter-
nalizing behavior, but Table 3 shows variability across studies on
a number of different dimensions. Five of the intervention trials
included primarily middle- and high-income couples, whereas four
focused primarily on low-income couples. Intervention intensity
(number of hours in the different CRE interventions) varies from
3 to 48, and the age of the child at follow-up from 3 months to 15
years, with no clear indication that measures of these constructs at
such widely different ages have similar meaning. Analyses of
intervention effects used different methods (ANOVAs, regres-
sions), making it difficult to combine in a single effect size. Across
studies, data came from parent ratings, teacher ratings, or experi-
menter observations. Tests of moderator effects of income level,
intervention intensity, child age, or source of data with two or three
studies in each category would not make sense, especially because
some of the moderator variables are confounded. At this point, we
have chosen to describe the strengths and weaknesses of each
study, with the goal of stimulating discussion and further research
on whether and how parents’ CRE participation affects their chil-
dren.
Becoming a family (C. P. Cowan & Cowan, 2000). In this
study, with random assignment to no-treatment and intervention
conditions, clinically trained faculty and graduate students in clin-
ical psychology formed male–female coleader teams who met
weekly with small groups of working-class to middle-class couples
from midpregnancy to 3 months after the birth of their first child
(48 hr over a period of 6 months). Couples were assessed at
baseline, 9 months, 1.5 years, 3.5 years, and 5.5 years postpartum
once the children had made their transition to kindergarten. The
intervention with the parents had a long-term significant effect
over the course of the study on couples’ marital satisfaction
(Schulz, Cowan, & Cowan, 2006), preventing the decline over
time found in most nonintervention studies (Twenge et al., 2003).
Although both self-reports and observations of marital quality
were correlated with parent reports, laboratory observations, and
kindergarten teacher ratings of the children’s school adaptation,
there were no significant differences between intervention- and
control-group children 3 and 5 years after the intervention ended.
Bringing Baby Home (BBH; Shapiro & Gottman, 2005).
This program was originally presented as a weekend workshop for
expectant couples (16 hr), with psychoeducational classes led by
clinical psychologists that included lectures, demonstrations, vid-
eotapes, role plays, and communication exercises. Topics included
couple communication, promoting father involvement, sensitive
parenting, coparenting, and infant development. A subsequent
RCT (Shapiro, Nahm, Gottman, & Content, 2011) was conducted
with 142 expectant middle-income couples assigned to a no-
treatment control condition (n46), a weekend workshop (n
45), or a workshop plus a 12-week “support group” (n51) that
met 12 times over a 6-month period after the birth of a child (24
hr). The workshops and support groups were led by childbirth
preparation teachers, some of who had master’s or nursing de-
grees. The Shapiro et al. article reported on assessments occurring
at baseline and again when the infants were 3 months old, a time
when the support groups had just begun. Observations were made
of mother, father, and infant in the Lausanne Trilogue Play para-
digm (Fivaz-Depeursinge & Corboz-Warnery, 1999), in which
father and mother take turns playing with the infant while the other
parent remains passive, and then both participate together in
parent–infant play. The BBH program was successful in promoting
less competition between parents during family play. There were
no significant differences between control and intervention fami-
lies (workshop only and workshop plus support group combined)
on infant facial affect, vocalization, or gaze, but this result could be
attributed to the fact that the support group was in its early phases
when the babies’ outcomes were assessed. An in-progress report of
a follow-up when the children were 30 months (Gottman, Gott-
man, & Shapiro, 2010) asserted that there were
dramatically increased effects from the 12-session Cowan-type couples
support group...added to the workshop...[including] less negative
ratings of child behavior on the Child Adaptive Behavior Inventory [P. A.
Cowan, Cowan, & Heming, 1995] and better language development in
the toddlers [no effect sizes reported]. In the experimental groups, chil-
dren ask more questions, engage in more pretend play, and have better
and more complex spoken vocabulary. (p. 174)
We are not aware of published detailed results of this follow-up.
Family Foundations (Feinberg, Kan, & Goslin, 2009). This
program has a unique emphasis based on data from earlier corre-
lational studies showing that the coparenting relationship is not
identical to the couple relationship, and that it adds significant
predictive power to path models that chart the connections be-
tween couple relationships, parenting quality, and children’s out-
comes (Feinberg, 2003). One hundred sixty-nine middle-income
couples expecting a first child, most of whom were married and
White, were randomly assigned to intervention and no-treatment
conditions. The intervention couples attended a total of eight
classes (16 hr), pre- and postbirth, with a special emphasis on
managing disagreements about parenting through the development
of communication skills, problem-solving, and conflict manage-
ment techniques. The program also included modules on partners
sharing expectations for each other, and on parenting an infant.
Based on parent-report data at the 3.5-year postpartum follow-up
(Feinberg et al., 2009), parents of boys showed a positive effect of
group participation on relationship satisfaction (d.43),
5
but
there were no significant effects for parents of girls. Other results
were more consistently positive: The intervention participants de-
scribed significantly more positive coparenting (d.18) and more
effective parenting than the parents in the control group (less lax,
less likely to inflict physical punishment; dfrom .30 to .36).
According to the parents, children of CRE participants showed
fewer externalizing, internalizing, and attention problems, espe-
5
Calculated from the published data.
372 COWAN AND COWAN
Table 3
CRE Studies That Include Child Outcomes
Income NHrs. of
CRE Child age at
pre Child age at
follow-up Impact on relationship
quality Impact on parenting Child outcomes Effect sizes
Parent report Observer Teacher report
Becoming a family
(Cowan & Cowan,
2000)
Mid/hi 96 48 In utero 5 yrs. Sig. Not reported Ext. n.s.
Int.
Academic
Parent report & observer Parent report & observer Parent report & observer
Bringing baby home
(Gottman et al., 2010) Mid/hi 142 34 Newborn 3 mos. Sig. Sig. Infant facial affect,
vocalization, gaze n.s
Ext.
Int.
Lang devel.
2.5 yrs. Sig. Sig. Pretend play Sig.
Parent report Parent report Parent report
Family foundations
(Feinberg et al.,
2009)
Mid/hi 169 16 Pre and post
birth 3.5 yrs. d.43 d.30–.36 Ext. Mdn d .74
Int.
Attention
Observer Observer Observer
Emotional security
(Cummings et al.,
2008) Mid/hi 90 3 4–8 yrs. 6–10 yrs. Sig. Sig. Child adjustment Sig.
Observer Observer Teacher and test Path model
Schoolchildren and their
families (Cowan
et al., 2005)
Mid/hi 100 32 5 yrs. 7 yrs. Sig. Sig. Ext. 31%
Int. 12%
Achievement 12%
Child perc. of kg. adj. 21%
Ext. Sig.
15 yrs. Sig. Sig. Int. Sig.
Observer
Building strong families
(Wood et al., 2010) Lo 5,102 32 0–15 2–17 n.s. n.s. Socioemotional devel. d.08
Parent report observer Observe Self-reg. d.07
Supporting Healthy
Marriage (Hsueh
et al., 2012)
Lo 6,298 32 0–2 yrs. 2.5–4.5 yrs. Int. d.10
Sig. Ext. d.08
Cog./Acad. n.s.
0–15 yrs. 2–17 yrs. Sig. Social comp. d.07
Distress d.06
Dysreg. d.05
Involved in parents
fights d.06
Self-blame d.13
373
CONTROVERSIES IN COUPLE RELATIONSHIP EDUCATION
cially for boys, and higher levels of social competence for both
sexes (median of six measures, d.74).
Emotional Security Theory Program for Community Fam-
ilies (Cummings, Faircloth, Mitchell, Cummings, & Schermer-
horn, 2008; Cummings & Schatz, 2012). Based on emotional
security theory (Cummings & Davies, 2010) as a construct linking
marital conflict to family relationships and children’s outcomes,
Cummings and colleagues developed and tested a four-session (45
min per session for a total of 3 hr) group intervention for nonclini-
cal couples who were parents of 4- to 8-year-old children. Couples
were randomly assigned to one of three conditions—(a) a parent-
only group (n24), (b) a parent–child group (n33), and (c) a
self-study condition (n33)—with pre- and posttests and
6-month and 1-year follow-ups. Intervention staff members were
given a brief preparation but were not clinically trained. The four
sessions focused on improving parents’ ways of expressing dis-
agreement, with special attention to helping couples distinguish
between productive and nonproductive conflict, the effects of
marital conflict on children, and the importance of maintaining the
quality of emotional bonds among all family members during
stressful family situations. A 2-year follow-up (Faircloth,
Schermerhorn, Mitchell, Cummings, & Cummings, 2011) with
substantial attrition (now 43% of the original couples) found
positive, large effects of the intervention, compared with a ran-
domized control, on measures of knowledge about the conse-
quences of marital conflict for children (d1.35) and constructive
conflict behaviors (d1.06), but not on couple problem solving.
The authors report that there was no difference between couples
who participated in the follow-up and couples who did not on any
of the pretest measures.
The data for main effects of the intervention on child outcomes
were not reported. Multilevel growth curve models found that
increases in the constructiveness of marital conflict behavior dur-
ing the marital conflict resolution task were linked with greater
marital satisfaction, better parenting practices, and improved child
adjustment (Child Behavior Check List, Achenbach & Edelbrock,
1983).
Schoolchildren and Their Families Project (P. A. Cowan,
Cowan, Ablow, Johnson, & Measelle, 2005). One hundred
middle-income couples were recruited to participate in a random-
ized clinical trial of a program for couples with a first child about
to make the transition to elementary school. One third of the
parents were assigned to a low-dose control condition, in which
they had the option of consulting once each year with the clinically
trained male–female staff team (psychologists and social workers)
who interviewed them when they entered the study. The remaining
two thirds participated in couples groups of four of five couples
meeting for 16 weeks (32 hr) with the same staff teams. The group
meetings began with an open-ended check-in followed by an
agenda for each week that covered one of the topics of a five-
domain risk-protective model of family functioning (individual
well-being; couple communication, including coparenting; parent-
ing; what to carry over and what to avoid from parents’ families of
origin; and life stresses and social supports outside the nuclear
family). Couples were randomly assigned to one of two variations
of the couples-group condition, with the same pairs of group
leaders conducting each variation. In one set of groups, in the
open-ended part of each meeting, leaders helped participants focus
on parenting and their relationship with the child, and in the other
groups, on their relationship as coparents. Two years after the
couples groups ended when the children were in first grade,
observations of parent–child interaction revealed that participants
in the intervention that emphasized parenting were significantly
warmer and more structured with their child than parents in the
control condition; their children showed fewer internalizing be-
haviors in their first-grade classroom and reported a greater sense
of well-being in a puppet interview (C. P. Cowan, Cowan, &
Heming, 2005b). Parents in the groups that emphasized the couple
and coparenting relationship used significantly more effective par-
enting strategies than the control parents 2 years later, and in
contrast with controls, they also showed no increase in their level
of couple conflict. Reports from first-grade teachers revealed that
their children were less aggressive and showed significantly higher
levels of academic achievement than children of the control par-
ticipants when tested individually by a member of the research
team. Over and above the quality of marital and parenting rela-
tionships, interventions with the parents accounted for 12% of the
variance in children’s academic achievement scores in kindergar-
ten, 21% of the variance in children’s perceptions of their kinder-
garten adjustment, 31% of the variance in reductions in external-
izing behavior between kindergarten and first grade, and 12% of
the reductions in their internalizing behaviors over the same period
of time. In an exploration of mediating effects, the study found that
changes in marital and parent–child interaction in the year from
prekindergarten to kindergarten were linked with the child out-
comes assessed at the end of first grade. Finally, in an unusually
long-term follow-up assessment 10 years later, as the children
made the transition to ninth grade in high school, intervention
effects on mothers, fathers, and children were still apparent (P. A.
Cowan, Cowan, & Barry, 2011).
BSF. In their final report of the 3-year follow-up of BSF for
low-income unmarried couples at eight sites, Wood et al. (2014)
presented new data on the effect of the intervention on three
measures of child well-being. The first two, family stability and
economic well-being, are measures of potential stressors or pro-
tective factors that affect children. The third is socioemotional
development, for which they found a very small but statistically
significant impact of the intervention on parents’ reports of chil-
dren’s behavior problems (d.08). It is puzzling that this effect,
however tiny, occurred in the absence of effects on the parents.
One speculation provided by the authors of the BSF report is that
the behavior problem finding was significant only in the four sites
providing home visiting services that included instructions in
parenting skills.
SHM (Hsueh et al., 2012; Lundquist et al., 2014). We
described this RCT study of CRE at eight sites with more than
6,000 low-income married couples earlier, along with outcomes
for self-report and observational measures of couple relationship
quality. Here we focus on what happened to the children. All four
of the couples-group programs in SHM focused on couple com-
munication and also included modules concerning individual dis-
tress and coping with external stressors. Although only one dealt
directly with parenting issues, many of the couple communication
issues in all of the programs involved disagreements about copa-
renting. The groups were led by male–female teams with a mix of
backgrounds, including parent educators, doctoral graduate stu-
dents, nurses, social workers, and marriage and family therapists.
374 COWAN AND COWAN
Staff qualifications at some of the eight sites were not described in
detail.
The results for child outcomes were reported in two sections—
primary and secondary measures. Primary measures were four
composite measures (including observations and parent reports)
that the investigators predicted would show intervention effects
(i.e., children’s self-regulation, internalizing and externalizing be-
havior, cognitive competence). Analysis of the total sample at the
30-month follow-up, when children ranged from 4 to 17 years,
revealed two small effects on children’s behavior (self-regulation,
d.03, and internalizing, d.04), which were not statistically
significant when corrections for performing multiple tests were
applied. However, when the whole sample was divided by chil-
dren’s age categories, significant positive intervention effects were
found for the youngest sample (2- to 4-year-olds) on three of the
four measures (self-regulation, d.07; internalizing, d.10;
externalizing, d.08), even after correcting for multiple tests.
There were no intervention effects on these measures in the 5- to
8.5-year-olds or the 8.5- to 17-year-olds.
Secondary measures were conceptualized as more exploratory.
The investigators reported small but statistically significant effects
for 2- to 17-year-old children whose parents participated in the
couples groups: The children were more socially competent (d
.07); were less distressed (d.06) and dysregulated (d.05);
became less involved when their parents were fighting (d.06);
and were less likely to attribute blame to themselves when their
parents fought (d.13). Given that five of nine secondary
measures showed intervention effects, the results could not be
attributed to the number of statistical tests.
SFI. In addition to effects on couple relationship satisfaction
and father involvement, the SFI study of low-income couples
(P. A. Cowan et al., 2009) tested for intervention effects on
parent-reported child behavior on four factor-constructed dimen-
sions: aggression, hyperactivity, shy/withdrawn behavior, and anx-
ious/depressed behavior. In the single-session control condition,
parents reported a significant increase in problem behaviors on all
four child dimensions; there were no significant increases for
children of couples in which fathers participated in groups, or for
couples who participated in groups. Two of the effect sizes con-
trasting control-group and couples-group participants were statis-
tically significant: child hyperactivity (d.22) and child shy/
withdrawn, (d1.88).
A partial replication study of the SFI couples-group intervention
was conducted in Phase II (P. A. Cowan, Cowan, Pruett, Pruett, &
Gillette, 2014) with a new set of 239 low-income Mexican Amer-
ican, White, and African American couples. Because the low-dose
controls in Phase I showed negative changes over time and the
fathers groups were not as effective, Phase II analyses were con-
ducted as a single-sample pre–post assessment of participants in
couples groups. In the absence of a randomized control, the results
of Phase I were used as a benchmark comparison to evaluate the
Phase II findings. Statistically significant effect sizes in favor of
the intervention participants were found for father involvement
(d.38), couple relationship satisfaction (d.37), and child
aggression, (d.34).
Summary. In sum, eight of nine intervention trials that in-
cluded measures of child effects showed statistically significant
effects of CRE interventions on children of the participants on at
least some of the child outcome measures—only the BAF inter-
vention did not; lack of statistical power (n66) and parents’
fairly high level of functioning at entry to the study could account
for the failure to find child effects associated with parents’ partic-
ipation in the intervention 5 years earlier. In addition, there was
little in the BAF materials related to parenting a newborn that were
relevant to the parenting issues of toddlers and preschoolers. It is
also the case that, like every other set of intervention studies, CRE
programs had different effects on different child outcomes, even
when, as in the case of SFI, the two intervention trials were almost
identical.
The data we have reported in the text and listed in Table 3 do not
simply represent a box score. The results include an array of small
positive effects, and several medium or large effects, across many
measures from different sources in both higher and lower income
families. Although these nine CRE trials are not sufficient to draw
a general conclusion about the impact of couples-group participa-
tion on children, they provide some support for what had been an
untested assumption—that parents’ participation in these groups
can have a positive effect on their children. They also provide
suggestions for studies of program variables that might be associ-
ated with more robust child outcome effects.
Future Directions for CRE Research
Our review of research on CRE interventions reveals that eval-
uation studies of programs to strengthen couple relationships in
middle-income and low-income families provide evidence for both
supporters and critics. Almost every statement from either perspec-
tive is subject to some form of rebuttal, and sometimes a rebuttal to
the rebuttal. Descriptive reviews generally include studies favoring
CRE, but often do not do justice to negative findings. Overall effect
sizes from meta-analyses are small; some show that effects disap-
pear over time. Countering these results are moderator analyses
that show some medium and large effects of CRE. Like the large
BSF trial with unmarried couples that was disappointing to CRE
supporters, the large SHM trial with married couples has been
dismissed as having small effects by critics, but, in fact, the SHM
results show consistent 12- and 30-month follow-up findings of
effects on both parent reports and observer measures. Some single
studies have been cited as casting doubt on CRE but, like studies
supporting CRE, have flaws that reduce the force of their conclu-
sions.
Our review also included data from studies that have not been
considered much in discussions of CRE, especially to answer the
question of whether parents’ participation in CRE benefits their
children. Without intervention, average marital satisfaction de-
clines, and this decline affects children negatively. Adding fathers
to interventions designed for mothers, and adding mothers to
interventions designed for fathers (i.e., taking a couples approach
to intervention), increases positive effects on both couples and
children. Finally, we described nine RCT intervention trials of
CRE that examined child outcomes, with eight reporting statisti-
cally significant effects.
In our view, there are too many negative findings and cautions
associated with CRE findings to conclude the CRE is an unqual-
ified success and that existing programs should be offered more
widely to large sections of the population. There are also too many
positive findings and promising indications of factors associated
with stronger results to recommend that funding for CRE programs
375
CONTROVERSIES IN COUPLE RELATIONSHIP EDUCATION
be discontinued and that we should look to other ways of helping
families. The meta-analyses we have surveyed suggest that some
CRE programs are more effective than others—moderate rather
than low-dose interventions that use professionally trained group
leaders and include observational measures. We need to disentan-
gle the separate and combined contributions of these dimensions,
and other dimensions yet to be tested, in order to create a more
differentiated picture of what CRE interventions can realistically
be expected to accomplish for families, and what their limitations
may be.
Study Design
We focused on randomized assignment studies in the earlier part
of this article, but there are many more reports in the literature of
CRE programs that use single-sample pre–post designs. Although
any attempt to provide systematic evaluations of intervention
programs is welcome, and contributions to refining interventions
can be made without using RCTs, a critical problem occurs when
single-sample pre–post studies show no change over time. The
single-sample design fails to identify circumstances in which the
absence of change represents a positive finding, for example, when
marital satisfaction remains stable for intervention participants, but
would likely have shown a decline over time if a comparison or
control group had been included in the study.
We noted that CRE intervention studies have relied on relatively
short-term follow-ups. Especially because early preschool inter-
ventions found “sleeper effects” long after the interventions con-
cluded (Schweinhart et al., 2004), it is important for researchers to
return to participants at least 18 months to 2 years postintervention,
or later, to determine whether couples have integrated what has
been learned in these interventions into their daily lives.
Program Design and Measurement
Up to the present, there has been almost no effort to compare the
effectiveness of different kinds of programs. Only very occasion-
ally have investigators gone beyond intervention versus control
comparisons to evaluate differences between programs with dif-
ferent characteristics. Benefits to participants may vary with (a) the
structure of the program, (b) the content of the curriculum, and (c)
the process involved in the way the intervention is delivered. So
far, analyses of intervention dosage have been conducted by com-
paring studies with fewer and greater numbers of sessions, or by
correlating attendance rates and outcomes. Analyses of attendance
rates do not provide good measures of dosage because self-
selection is usually a factor in whether couples or fathers continue
or drop out, and so those who attend more sessions (receive higher
treatment dosage) may be those who are more motivated or find
the program useful. We need information about the effects of an
intervention when dosage is varied systematically, using random
assignment to shorter and longer versions of the program.
Programs with different types of curriculum content and differ-
ent approaches to teaching and learning are hidden within the
summary statistics yielded by meta-analyses. At present, the most
highly researched CRE intervention is the PREP program (Mark-
man et al., 2010). PREP focuses on teaching moderate to large
groups of couples effective communication skills. Other programs,
which we could call “communication plus,” include additional
aspects of family life including three-generational issues (Gordon,
DeMaria, Haggerty, & Hayes), parenting (C. P. Cowan & Cowan,
2000; Gottman et al., 2010), and outside-the-family life stress
(P. A. Cowan et al., 2009; Rienks et al., 2011).
Often confounded with curriculum content is the program’s
approach to intervention. PREP programs take a psychoeduca-
tional approach, with material presented by instructors and fol-
lowed by practice sessions for each couple with coaching from the
instructor or assistants. Other programs, such as SFI, have a
defined curriculum but place more emphasis on participants’ par-
ticular issues, and on group process and interaction among the
participants. As of now, we know that some programs with very
different approaches (e.g., PREP, Family Foundations, SFI) have
helped to improve or at least maintain couple relationship quality,
but we do not know how these programs compare with each other
when tested in the same participant populations. Only by adding
and subtracting curriculum modules and measuring outcomes is it
possible to determine whether specific aspects of the intervention
curriculum produce positive effects (e.g., couple communication
training, father-involvement discussions, parenting skills).
Differences in curriculum and intervention approach are often
confounded with two additional variables. First, programs that
train specific skills can accommodate a larger number of attendees,
whereas programs that emphasize personal issues and group dis-
cussion must be smaller (ideally about four to six couples). We are
aware of no CRE studies that attempt to link group size to out-
comes. Second, although there is one meta-analysis showing that
professionally trained group leaders are more effective than para-
professionals, much more investigation of this key topic is needed.
One important issue is whether the effectiveness of higher levels of
training varies with program structure. From observation, we be-
lieve that the more the program relies on teaching skills with
lessons derived from a highly specific manual and communication
problems chosen by the instructor(s), the less training is required
for those who deliver the program. The more the program relies on
group leaders who raise topics, present exercises, and foster dis-
cussion of spontaneous personal examples between partners and
among group members, the more training and clinical skills the
leader must have in order to deal with the complex personal issues
that couples bring and to contain the level of individual and couple
distress that often emerges. We are aware that hiring leaders with
less extensive training is less costly to those responsible for mount-
ing the program, which means that answers to questions about
training and staff experience have important practical and policy
implications.
Almost all evaluations of CRE interventions focus on couple
relationship satisfaction and communication skills. A few as-
sess other domains, such as parents’ adjustment, fathers’ in-
volvement, and parenting style. And, as we have described, a
very few examine effects on the children. Elsewhere (Cowan et
al., 2005), we have shown that couple relationship quality,
parents’ individual adjustment, and parent–child relationship
quality are three of five major dimensions of risks and buffers
affecting family adaptation. The other two are factors associ-
ated with relationship patterns in one’s family of origin, and the
balance between life stressors and social supports outside the
family. Very few studies target these factors in the CRE cur-
riculum or assess these dimensions as potential mediating or
moderating variables.
376 COWAN AND COWAN
The source of information about the family also varies from
study to study. A wide variety of self-report questionnaire mea-
sures are included in evaluation studies. Some evidence suggests
that observational measures might provide clearer evidence of
intervention effects (e.g., Fawcett et al., 2010). The question is not
only whether observational measures are better than self-reports in
looking for program effects but also whether observational mea-
sures show changes earlier than self-reports. It may be, for exam-
ple, that it is more difficult to change overall perceptions and
attitudes about a relationship than it is to make small changes in
behavior when the partners disagree. It would help immeasurably
if the field would commit to developing some widely accepted
self-report and observational measures to be used across studies so
that direct comparisons of other study variations can be made more
clearly.
Target Populations
Income, ethnicity/race, and culture. Initially, almost all
evaluations of CRE were conducted on middle-class, primarily
White samples; more recently, low-income participants have
been included from White, African American, and Latino sam-
ples, but until the BSF, SHM, and SFI evaluations, systematic
evaluations of couple-focused interventions on non-White par-
ticipants were in short supply. Even less evident in CRE re-
search is information about other ethnic groups such as various
subgroups of Asian, Middle Eastern, etc. Whether intervention
outcomes differ across income levels and ethnic groups has yet
to be determined.
So far, the bulk of the studies of CRE have been located in
the United States, although we have noted a few located in other
countries. We know that across countries and cultures, there are
wide differences in conceptions of marital roles and gendered
parenting roles, but we do not know whether some programs
may be more effective than others in reaching different cultural
groups, or anything about the extent to which CRE programs
must be tailored to fit with cultural expectations and norms for
couples.
Family life intervention points. Another important source
of variation across studies has been the point in the family life
cycle at which the intervention was conducted. For CRE stud-
ies, the greatest programmatic emphasis has been on premarital
preventive interventions, with a second emphasis on couples
having a baby, usually a first baby. Confounded with family life
stage is the age of the parents and the child. While very few
CRE programs have been offered to teens (see Florsheim, 2014,
for an important exception), the vast majority occur early in the
family cycle with young parents. We need to know much more
about whether there are critical periods or intervention points at
which these programs produce their strongest effects—for par-
ents and for their children.
The prevention focus of CRE has been shifting in recent efforts
to address populations at risk by virtue of their poverty status.
Other important extensions of CRE could profitably be explored,
including coparenting interventions for divorcing couples or for
families involved in the child welfare or juvenile justice systems,
and a few of these programs are emerging. We assume that as we
climb the ladder of risk, leaders of CRE programs would be more
highly trained, and programs would be conducted in combination
with other services addressed to meeting the needs of these mul-
tirisk families.
We should note that despite the intent of programs to target
homogeneous populations, the participants in any couples or
father-involvement program are likely to be heterogeneous on a
number of dimensions (age, race/ethnicity/culture, income, mental
health, quality of coparenting relationship, etc.). Future studies
could profitably address the question of whether there is an opti-
mal level of homogeneity or heterogeneity among CRE-group
participants to obtain positive effects.
Dynamic interactions among variables that influence
outcomes. Adding to the complexity of our recommendations
for future research is the reasonable assumption that some of these
variables may act in combination to facilitate or interfere with
intervention outcome. To choose one obvious possibility, longer
interventions, with longer follow-ups, more highly trained staff,
and increased opportunities for family support may be required to
provide effective CRE interventions for parents at higher risk by
virtue of poverty, clinically diagnosable mental health problems,
or fragile relationship status (e.g., never married, never lived
together). It is not possible to include all factors affecting CRE
outcomes in one study. Our hope is that this more inclusive map of
the variables most likely to affect the selected target population
will encourage investigators to select a combination of variables
that will provide more information than we have now about what
works for whom.
Testing Theories About How the Interventions Work
Only a few studies that we are aware of have gone beyond the
determination of whether intervention participants fare better than
participants in control groups to examine potential pathways of
influence. For example, using path analyses and regressions, C. P.
Cowan et al. (2005) showed that intervention effects on child
outcomes were mediated by intervention-induced changes in ob-
served couple communication and parenting effectiveness. Simi-
larly, reduction in marital conflict was associated with improved
parenting in the emotional security-based intervention described
by Cummings et al. (2008).
As in the rest of our review, the results do not always follow the
expected pattern. For example, in a German study, Rogge et al.
(2013) found that changes in variables targeted by the intervention
curriculum either did not occur or were unrelated to outcomes.
Specifically, wives in the German PREP intervention focused on
reducing couple conflict did not decline in conflict as much as
wives in an empathy-based program. Another of several examples
cited by Bradbury and Lavner (2012) is a French study by Boden-
mann and colleagues (Bodenmann, Bradbury, & Pihet, 2008) in
which increases in female negative communication were associ-
ated with positive relationship outcomes, counter to the assump-
tion in the PREP program that reducing negative communication is
one of the keys to successful CRE programs. Bradbury and Lavner
advance an interesting hypothesis: It may be that investigators are
not measuring some of the important variables that produce CRE
intervention effects. Rather than a focus on prescribing behaviors,
they argue, it may be necessary to identify principles that underlie
relationship-sustaining communication, and to examine whether
changes in how couples adopt these principles underlie positive
CRE intervention effects.
377
CONTROVERSIES IN COUPLE RELATIONSHIP EDUCATION
Intervention studies also have unique power to contribute to the
testing of hypotheses about causal relationships (P. A. Cowan &
Cowan, 2002). If we can show that CRE interventions affect
specific aspects of family relationships, and that intervention-
induced improvement is associated with positive changes in chil-
dren’s adaptation, we will have obtained strong evidence for our
theories of how family factors affect children’s development and
well-being.
Policy Implications
Cost-Benefit Analyses: Translating Findings Into
Outcomes of Interest to Policymakers
An unresolved issue at the interface of research and policy is the
question of potential costs and benefits of CRE interventions.
Policymakers want data showing that the benefits ascribed to a
proposed program will be greater than the costs, but so far, only
the SHM program provided an estimate of costs (between $9,000
and $11,000 per family); we are not aware of attempts to determine
the financial benefits of reducing couple conflict or decreasing
behavior problems in young children, although beginning attempts
are being made in the United Kingdom. The report by Spielhofer
et al. (2014) asserts that for every £1 spent for marriage prepara-
tion, the returns to society are £11.5 ($18.5 for every $1 spent), and
that the returns are similar to those for couples counseling. The
main problem in this endeavor for current CRE programs is that,
with the exception of facts like marital status, presence of father in
the home, or contribution to child support, the outcomes are
typically reported as continuous measures (couple relationship
satisfaction, positive parenting, child behavior problems) that have
yet to be assigned monetary values. What cost–benefit researchers
and politicians want to know is whether CRE interventions pro-
duce change in the percentage of individuals and families involved
in clinical or social categories: diagnosed depression, divorce/
separation, and children and adolescents needing therapy or in-
volved in child protective or juvenile justice systems. A further
difficulty in monetizing child outcomes is that policymakers want
to know about socially important outcomes when children are
older, whereas most of these intervention programs, especially
those with a preventive focus, are with parents of children from
newborn to the preschool and elementary school years.
Service providers with an eye on costs also need more informa-
tion about program parameters such as those we listed earlier
(number of meetings, level of staff training, curriculum content or
approach). The question of costs and benefits is especially relevant
to the discussion of CRE programs because, at best, the effect sizes
of many of these interventions are small. Until we know whether
the benefits to be gained by these programs result in substantial
savings (e.g., whether reductions in young children’s aggression
ultimately pay off in terms of less need for therapy or involvement
in the juvenile justice system), it will be difficult to evaluate the
meaning of the effect sizes that have been found.
One specific outcome of interest to policymakers has been a
reduction in the incidence of divorce. Given the political context of
“marriage promotion” and increases in federal funding for CRE in
the last decade, the prevention of divorce may be a primary goal of
politicians who vote to fund these programs. And yet neither the
BSF nor the SHM programs (except at one site), nor the SFI
project, found that CRE interventions keep low-income couples
from separating and divorcing. While there have been occasional
reports that the rate of separation and divorce was reduced after
participation in a CRE program (e.g., Stanley et al., in press), most
studies (a) have short-term follow-ups in which it would be un-
likely to find changes in family stability, (b) fail to find such
effects, or (c) do not report relationship stability data. The question
to consider is whether the findings represent a criticism of CRE
programs or a misunderstanding of program goals. The policy
intention of some supporters of CRE is to “promote marriage”
and/or “maintain two-parent families,” whereas the intervention
curricula are devoted to improving couple relationship quality—
for the benefit of the parents and the children. From the former
perspective, a divorce or separation of a participant couple would
count as a program failure. From the latter perspective, if the
couple were able to maintain a collaborative coparenting relation-
ship after making the decision to separate or divorce, the program
would be regarded as a success, and from all we know, both
parents and children should benefit. It is possible that parental
separation or divorce might be protective of the child’s well-being
if it lessened the conflict the child was exposed to, or that an
improved, collaborative coparenting relationship might benefit
children if it lessened the strain in either or both of the parent–
child relationships, regardless of whether their parents are married,
cohabiting, separated, or divorced (Pruett & Barker, 2009). None
of the studies that we are aware of go beyond the fact of parental
status to examine the quality of the coparenting relationship after
separation or divorce in experimental and control participants.
How Do CRE Interventions Apply
to Same-Sex Marriages?
In the funds made available for CRE programs, the moral
emphasis on marriage promotion in the United States at a time
when same-sex marriage was not legal led to the exclusion of
same-sex couples from federally funded CRE intervention pro-
grams. Especially in view of the changing legal climate in the
United States, but also as a personal value, we believe that this
exclusion is not warranted, and that same-sex couples should not
only be eligible but also sought out for inclusion in CRE programs.
Gay and lesbian couples struggle with challenging issues as part-
ners and as parents, just as male–female couples do, and by any
logic, helping to strengthen their relationships should foster more
attentive parenting and their own and their children’s well-being.
There is a question about whether current versions of CRE
would require alterations to fit specific needs of lesbian and gay
parents. Our impression from the literature on lesbian couples
(Blake, Casey, Jadva, & Golombok, 2012; Goldberg & Perry-
Jenkins, 2007; Gottman et al., 2003) is that role differences and
disagreements about issues lead to very similar communication
dynamics as those experienced by heterosexual couples. Regard-
less of sexual orientation, all couples face relationship issues
arising from attempts to deal with different family of origin pat-
terns, personal mental health challenges, differences in parenting
ideas, work–family balance, and other modern family stresses. The
speculations we have raised can only be addressed empirically,
with systematic evaluations of intervention trials with both male
and female same-sex couples.
378 COWAN AND COWAN
Marriage Promotion, CRE, and Alleviating Poverty
One objection to the funding of CRE programs is that the
original premise of “promoting marriage” in order to raise
families out of poverty is flawed. The fact that children of
single mothers are more likely to live in poverty does not mean
that marriage (any marriage, regardless of the relationship qual-
ity) will raise the economic level of the family. What is missing
from the correlation between marital status and poverty is an
answer to the question “What happens when single mothers
marry?” According to Graefe and Lichter (2007), who analyzed
data from 3,872 women who participated in the National Survey
of Family Growth, poor single mothers who marry tend to have
low marital satisfaction, divorce early, and be even worse off
economically than single mothers who do not marry. That is, in
this study, the transition from single motherhood to marriage in
low-income families did not result in the hoped-for economic
benefits.
In the controversies over government funding of CRE pro-
grams, a frequently raised issue is that there are other interven-
tions, especially for low-income families, that would be more
effective. The most frequent view (e.g., P. Cohen, 2014; Wil-
liams, 2014) is that poor families would be better served by
programs that elevate their economic circumstances than by
relationship-enhancement interventions. There is no question
that poverty affects couple relationships and family functioning
(Conger, Cui, & Lorenz, 2011). There is also no question that
the normative decline in marital satisfaction over time also
affects family relationships, with negative outcomes for chil-
dren. What is in question here is whether alleviating poverty
through income supplementation will increase the quality of
couple and family relationships. We have found older studies in
both the United States and the United Kingdom in which family
income supplement programs under various names have had
either a negative (Hannan, Tuma, & Groenfeld, 1977; Knox &
Redcross, 2000) or neutral (Cain, Wissoker, Hannan, & Tuma,
1990) effect on maintaining marital stability. Despite the cor-
relations between income level and marital satisfaction, we
could find no information about whether income supplements
improve the quality of relationship between the partners. It may
be that job training and other forms of interventions to reduce
poverty may have beneficial effects on the family, but there is
no evidence (yet) that they improve the quality of relationships
between partners, or between parents and children. It is curious
that the policy argument has been framed in terms of a choice
between relationship and economic interventions. Why not de-
velop a program that combines both, and tests them in additive
combinations: CRE alone, economic intervention alone, or re-
lationship and economic intervention combined? Only in this
way will policymakers be able to make empirically informed
decisions about the value added by either approach.
Current Policy Issues in the United States and the
United Kingdom
In our view, controversies in both the United States and the
United Kingdom about government funding of CRE programs
have been complicated by the confounding of four different
issues: (a) an argument about the wish to promote marriage and
reduce divorce, sometimes stated in moral-religious terms,
sometimes in economic terms; (b) an empirical argument about
whether increasing the stability of marriages will reduce family
poverty; (c) a social policy argument about the need to
strengthen couple relationships (largely noncontroversial); and
(d) a concern, relatively unexplored until recently, about
whether CRE interventions benefit couples and their children.
Some, but certainly not all, of the opposition to government
funding of programs to strengthen couple relationships have
come from critics who would support the funding of programs
to strengthen couple relationships, but oppose programs to
promote marriage as one solution to the poverty associated with
single-parent families.
Our recommendation to combine CRE interventions with
economic and other kinds of interventions to benefit families
runs into an immediate obstacle, because federal, state, and
county government departments serving families are organized
in independent silos, with support for mothers and children in
one department, father-involvement programs in another, and
jobs programs in yet another. The same is true of nongovern-
mental public and private agencies addressing needs of fami-
lies. The important of reducing or eliminating silos was brought
home to us in our summaries of CRE and father-involvement
programs. Our perception is that father-involvement interven-
tions became more successful when they included more focus
on family relationships. Reciprocally, CRE interventions, by
their very nature, include fathers and often produce enhanced
involvement of fathers in the rearing of their children. The data
from the SFI project (P. A. Cowan et al., 2009) indicate that a
couples approach to father involvement had a broader impact on
the family than group meetings that included only fathers, with
the same facilitators and curriculum. Given that digital media
now make it possible for individuals across the world to com-
municate with each other, might it be possible to establish
communication links across corridors and buildings, so that
multiple approaches to strengthening families can result in even
more widespread benefits for parents and their children?
We began this article by describing similar current U.S. and
U.K. government policy decisions about supporting preventive
interventions for couples. The similarity of goals—strengthen-
ing couple relationships to increase father involvement and
family stability in the service of providing more supportive
environments for children—has not led to the same decisions
about what kinds of programs to fund. In the United States,
given a historical absence of government and institutional sup-
port for couple relationships (e.g., most insurance companies
will not pay for couples therapy), beliefs about the need to
strengthen marriage led to the adoption of a currently popular
modality—groups for CRE. As we have seen, results from the
U.S. large-scale funding of the BSF and SHM programs have
recently been released, to mixed reviews. Reacting to the fact
that most of the smaller programs funded by the original $100
million in Healthy Marriage Funds and the more recent alloca-
tion of $75 million have produced no systematic evidence of
positive effects, the Administration for Children and Families
has hired Mathematica, a policy research evaluation company,
to lead Parents and Children Together, a systematic evaluation
of some of the newer CRE programs, with a specific focus not
only on potential benefits for couple relationship quality but
379
CONTROVERSIES IN COUPLE RELATIONSHIP EDUCATION
also for economic self-sufficiency and the well-being of chil-
dren.
6
A similar endeavor, the Fatherhood Research and Prac-
tice Network,
7
has been funded by the Administration for
Children and Families to offer small grants to encourage re-
search on father involvement, with one of three aims being to
(a) plan, initiate, and carry out a research agenda to build an
evidence base for effective responsible fatherhood interventions
and the development of appropriate measures; (b) develop
capacity within the evaluator and practitioner communities to
conduct and participate in high-quality evaluation studies; and
(c) disseminate findings and best practices. One of three work-
ing groups of this new organization focuses specifically on
defining new directions for the evaluation of coparenting inter-
ventions. It seems that although large-scale intervention pro-
grams are not going to receive government funds in the near
future, smaller studies aimed at providing a better evidence base
for couple relationship strengthening programs will continue to
receive federal government support.
The U.K. government, with a history of ongoing support for
professional couples counseling and therapy as part of the National
Health Service, recently chose to create additional funding for
services directed to one couple at a time, along with some support
for marriage preparation couples groups. This has not prevented
the government from looking to the couples-group interventions in
the United States and beginning to consider CRE as a possible
option, especially for low-income families. The Department for
Education has recently allocated £2.9 million ($4.84 million) to the
Tavistock Centre for Couple Relationships and to Family Action,
for a trial of the SFI intervention for low-income and vulnerable
families (P. A. Cowan et al., 2009), renamed “Parents as Partners.”
Because the emphasis of government funding has shifted from
couple support to parenting support, depending on whether the
Conservative Party or Labour Party is in power, future directions
in the United Kingdom will depend on the outcome of elections to
be held in May 2015.
It seems that despite the current controversies in the research
literature concerning the benefits of couple relationship strength-
ening interventions, support for trials of small, well-evaluated
CRE programs will continue. Our reading of the literature leads us
to the conclusion that these trials are warranted. A more extensive
research program is needed before final decisions are made about
whether a new generation of preventive interventions for copar-
enting partners can fulfill the original promise of CRE to
strengthen family relationships in ways that foster both parents’
and children’s healthy development.
6
See http://www.acf.hhs.gov/programs/opre/research/project/parents-
and-children-together-pact-evaluation
7
See http://www.acf.hhs.gov/programs/opre/research/project/the-father
hood-research-and-practice-network
References
Avellar, S., Dion, M. R., Clarkwest, A., Zaveri, H. H., Asheer, S., Borra-
daile, K., & Zukiewicz, M. (2011). Catalogue of research: Programs for
low-income fathers. OPRE Report 2011–20. Washington DC: Math-
ematic Policy Research.
Bakermans-Kranenburg, M. J., van Ijzendoorn, M. H., & Juffer, F. (2003).
Less is more: Meta-analyses of sensitivity and attachment interventions
in early childhood. Psychological Bulletin, 129, 195–215. doi:10.1037/
0033-2909.129.2.195
Belsky, J., & Jaffee, S. R. (2006). The multiple determinants of parenting.
In D. Cicchetti & D. J. Cohen (2006), Developmental psychopathology.
Vol 3:Risk, disorder, and adaptation (2nd ed., pp. 3885). Hoboken,
NJ: Wiley.
Besnard, T., Capuano, F., Verlaan, P., Pulin, F., & Vitaro, F. (2009). Effect
ajoute del la participation des pers a une intervention familiale [The
added effect of the fathers’ participation in a family intervention]. Revue
de Psychoéducation, 38, 45–71.
Bir, A., Corwin, E., MacIlvain, B., Beard, A., Richburg, K., Smith, K., &
Lerman, R. (2012). Impacts of a community healthy marriage initiative.
OPRE Report # 2012-34A. Washington, DC: Office of Planning, Re-
search and Evaluation, Administration for Children and Families, U.S.
Department of Health and Human Service.
Blake, L., Casey, P., Jadva, V., & Golombok, S. (2012). Marital stability
and quality in families created by assisted reproductive technologies: A
follow-up study. Reproductive BioMedicine Online, 25, 678683. doi:
10.1016/j.rbmo.2012.09.006
Blanchard, V. L., Hawkins, A. J., Baldwin, S. A., & Fawcett, E. B. (2009).
Investigating the effects of marriage and relationship education on
couples’ communication skills: A meta-analytic study. Journal of Fam-
ily Psychology, 23, 203–214. doi:10.1037/a0015211
Bodenmann, G., Bradbury, T. N., & Pihet, S. (2008). Relative contribu-
tions of treatment-related changes in communication skills and dyadic
coping skills to the longitudinal course of marriage in the framework of
marital distress prevention. Journal of Divorce & Remarriage, 50, 1–21.
doi:10.1080/10502550802365391
Bodenmann, G., Pihet, S., Cina, A., Widmer, K., & Shantinath, S. (2006).
Improving dyadic coping in couples with a stress-oriented approach: A
2-year longitudinal study. Behavior Modification, 30, 571–597. doi:
10.1177/014554455042699.02
Boucher, D. (2008). Marriage support services review 2008. London, UK:
CARE.
Bradbury, T. N., & Lavner, J. A. (2012). How can we improve preventive
and educational interventions for intimate relationships? Behavior Ther-
apy, 43, 113–122. doi:10.1016/j.beth.2011.02.008
Brody, G. H., & Forehand, R. (1985). The efficacy of parent training with
maritally distressed and nondistressed mothers: A multimethod assess-
ment. Behaviour Research and Therapy, 23, 291–296. doi:10.1016/
0005-7967(85)90007-5
Butler, M. H., & Wampler, K. S. (1999). A meta-analytic update of
research on the couple communication program. American Journal of
Family Therapy, 27, 223–237. doi:10.1080/019261899261943
Cain, G. G., Wissoker, D. A., Hannan, M. T., & Tuma, N. B. (1990). A
reanalysis of marital stability in the Seattle-Denver Income-Maintenance
Experiment. American Journal of Sociology, 95, 1235–1236. doi:
10.1086/229428
Carlson, M. J., Pilkauskas, N. V., McLanahan, S. S., & Brooks-Gunn, J.
(2011). Couples as partners and parents over children’s early years.
Journal of Marriage and Family, 73, 317–334. doi:10.1111/j.1741-3737
.2010.00809.x
Carroll, J. S., & Doherty, W. J. (2003). Evaluating the effectiveness of
premarital prevention programs: A meta-analytic review of outcome
research. Family Relations, 52, 105–118.
Chang, Y.-S., & Barrett, H. (2008). Couple relationships: A review of the
nature and effectiveness of couple support services. London, UK: The
Family and Parenting Institute.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences
(2nd ed.). Hillsdale, NJ: Erlbaum.
Cohen, P. (2014). This “Supporting healthy marriage,” I do not think it
means what you think it means. Retrieved from http://familyinequality
.wordpress.com/2014/04/07
380 COWAN AND COWAN
Coleman, L., & Glenn, F. (2009). When couples part: Understanding the
consequences for adults and children. London, UK: One Plus One.
Conger, R. D., Cui, M., & Lorenz, F. O. (2011). Intergenerational conti-
nuities in economic pressure and couple conflict in romantic relation-
ships. In F. D. Fincham & C. Ming (Eds.), Romantic relationships in
emerging adulthood (pp. 101–122). New York, NY: Cambridge Univer-
sity Press.
Coontz, S. (2005). Marriage, a history: From obedience to intimacy, or
how love conquered marriage. New York, NY: Viking Press.
Cowan, C. P., & Cowan, P. A. (2000). When partners become parents: The
big life change for couples. Mahwah, NJ: Erlbaum.
Cowan, C. P., Cowan, P. A., & Barry, J. (2011). Couples’ groups for
parents of preschoolers: Ten-year outcomes of a randomized trial. Jour-
nal of Family Psychology, 25, 240–250.
Cowan, C. P., Cowan, P. A., & Heming, G. (2005). Two variations of a
preventive intervention for couples: Effects on parents and children
during the transition to elementary school. In P. A. Cowan, C. P. Cowan,
J. Ablow, V. K. Johnson, & J. Measelle (Eds.), The family context of
parenting in children’s adaptation to elementary school (pp. 277–312).
Mahwah, NJ: Erlbaum.
Cowan, P. A., & Cowan, C. P. (2002). Interventions as tests of family
systems theories: Marital and family relationships in children’s devel-
opment and psychopathology. Development and Psychopathology, 14,
731–759. doi:10.1017/S0954579402004054
Cowan, P. A., Cowan, C. P., Ablow, J. C., Johnson, V. K., & Measelle,
J. R. (2005). The family context of parenting in children’s adaptation to
elementary school. Mahwah, NJ: Erlbaum.
Cowan, P. A., Cowan, C. P., & Heming, G. (1995). Manual for the Child
Adaptive Behavior Inventory (CABI) [Unpublished manual]. University
of California, Berkeley Institute of Human Development.
Cowan, P. A., Cowan, C. P., & Knox, V. (2010). Marriage and fatherhood
programs. The Future of Children, 20, 205–230. doi:10.1353/foc.2010
.0000
Cowan, P. A., Cowan, C. P., Pruett, M., Pruett, K., & Gillette, P. (2014).
Evaluating a couples group to enhance father involvement in low-
income families using a benchmark comparison. Family Relations, 63,
356–370. doi:10.1111/fare.12072
Cowan, P. A., Cowan, C. P., Pruett, M. K., Pruett, K., & Wong, J. J. (2009).
Promoting fathers’ engagement with children: Preventive interventions
for low-income families. Journal of Marriage and Family, 71, 663–679.
doi:10.1111/j.1741-3737.2009.00625.x
Cummings, E. M., & Davies, P. (2010). Marital conflict and children: An
emotional security perspective. New York, NY: Guilford Press.
Cummings, E. M., Faircloth, W. B., Mitchell, P. M., Cummings, J. S., &
Schermerhorn, A. C. (2008). Evaluating a brief prevention program for
improving marital conflict in community families. Journal of Family
Psychology, 22, 193–202. doi:10.1037/0893-3200.22.2.193
Cummings, E. M., & Schatz, J. N. (2012). Family conflict, emotional
security, and child development: Translating research findings into a
prevention program for Community Families. Clinical Child and Family
Psychology Review, 15, 14–27. doi:10.1007/s10567-012-0112-0
Dadds, M. R., Sanders, M. R., Behrens, B. C., & James, J. E. (1987).
Marital discord and child behavior problems: A description of family
interactions during treatment. Journal of Clinical Child Psychology, 16,
192–203. doi:10.1207/s15374424jccp1603_3
Deficit Reduction Act (2005). Public Law 171. 109th Congress. Congres-
sional record. Vol. 151.
Department for Education and Skills. (2007). Every parent matters. Not-
tingham, UK: DfES Publications.
Doss, B. D., Rhoades, G. K., Stanley, S. M., & Markman, H. J. (2009). The
effect of the transition to parenthood on relationship quality: An 8-year
prospective study. Journal of Personality and Social Psychology, 96,
601–619. doi:10.1037/a0013969
Emery, R. E. (1988). Marriage, divorce, and children’s adjustment. Thou-
sand Oaks, CA: Sage.
Engsheden, N., Fabian, H., & Sarkadi, A. (2013). Offering relationship
education (PREP) for couples during pregnancy: Self-selection patterns.
Family Relations: An Interdisciplinary Journal of Applied Family Stud-
ies, 62, 676685. doi:10.1111/fare.12032
Faircloth, W. B., Schermerhorn, A. C., Mitchell, P. M., Cummings, J. S.,
& Cummings, E. M. (2011). Testing the long-term efficacy of a preven-
tion program for improving marital conflict in community families.
Journal of Applied Developmental Psychology, 32, 189–197. doi:
10.1016/j.appdev.2011.05.004
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.
Feinberg, M. E. (2003). The internal structure and ecological context of
coparenting: A framework for research and intervention. Parenting,
Science and Practice, 3, 95–131. doi:10.1207/S15327922PAR0302_01
Feinberg, M. E., Kan, M. L., & Goslin, M. C. (2009). Enhancing copar-
enting, parenting, and child self-regulation: Effects of family founda-
tions 1 year after birth. Prevention Science, 10, 276–285. doi:10.1007/
s11121-009-0130-4
Finkel, E. J., Slotter, E. B., Luchies, L. B., Walton, G. M., & Gross, J. J.
(2013). A brief intervention to promote conflict reappraisal preserves
marital quality over time. Psychological Science, 24, 1595–1601. doi:
10.1177/0956797612474938
Fivaz-Depeursinge, E., & Corboz-Warnery, A. (1999). The primary trian-
gle: A developmental systems view of mothers, fathers, and infants. New
York, NY: Basic Books.
Florsheim, P. (2014). The Young Parenthood Program: A guide to Helping
young mothers and fathers become effective co-parents. New York:
Oxford University Press.
Goldberg, A. E., & Perry-Jenkins, M. (2007). The division of labor and
perceptions of parental roles: Lesbian couples across the transition to
parenthood. Journal of Social and Personal Relationships, 24, 297–318.
Gordon, L., DeMaria, R., Haggerty, V., & Hayes, E. (2007). PAIRS SHM
Facilitator’s guide and curriculum for managers, facilitators, and family
support staff. Weston, FL: PAIRS Foundation.
Gottman, J., Gottman, J., & Shapiro, A. (2010). A new couples approach
to interventions for the transition to parenthood. In M. Schulz, M. Pruett,
P. Kerig, & R. D. Parke (Eds.), Strengthening couple relationship for
optimal child development: Lessons from research and intervention (pp.
165–180). Washington, DC: American Psychological Association. doi:
10.1037/12058-011
Gottman, J. M., Levenson, R. W., Gross, J., Frederickson, B. L., McCoy,
K., Rosenthal, L.,...Yoshimoto, D. (2003). Correlates of gay and
lesbian couples’ relationship satisfaction and relationship dissolution.
Journal of Homosexuality, 45, 23–43.
Graefe, D. R., & Lichter, D. T. (2007). When unwed mothers marry: The
marital and cohabiting partners of midlife women. Journal of Family
Issues, 28, 595–622.
Grych, J. H., & Fincham, F. D. (Eds.). (2001). Interparental conflict and
child development: Theory, research, and applications. New York, NY:
Cambridge University Press. doi:10.1017/CBO9780511527838
Grych, J. H., Harold, G. T., & Miles, C. J. (2003). A prospective investi-
gation of appraisals as mediators of the link between interparental
conflict and child adjustment. Child Development, 74, 1176–1193. doi:
10.1111/1467-8624.00600
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
Hahlweg, K., & Richter, D. (2010). Prevention of marital instability and
distress. Results of an 11-year longitudinal follow-up study. Behaviour
Research and Therapy, 48, 377–383. doi:10.1016/j.brat.2009.12.010
381
CONTROVERSIES IN COUPLE RELATIONSHIP EDUCATION
Halford, W. K. (2011). Marriage and relationship education: What works
and how to provide it. New York, NY: Guilford Press.
Halford, W. K., & Bodenmann, G. (2013). Effects of relationship education
on maintenance of couple relationship satisfaction. Clinical Psychology
Review, 33, 512–525. doi:10.1016/j.cpr.2013.02.001
Halford, W. K., Petch, J., & Creedy, D. K. (2010). Promoting a positive
transition to parenthood: A randomized clinical trial of couple relation-
ship education. Prevention Science, 11, 89–100. doi:10.1007/s11121-
009-0152-y
Hannan, M. T., Tuma, N. B., & Groeneveld, L. P. (1977). Income and
marital events: Evidence from an income-maintenance experiment.
American Journal of Sociology, 82, 1186.
Harknett, K., Hardman, L., Garfinkel, I., & McLanahan, S. S. (2001). The
Fragile Families Study: Social policies and labor markets in seven cities.
Children & Youth Services Review, 23, 537–555.
Hawkins, A. J. (2014). Continuing the important debate on govern-
supported healthy marriages and relationships initiative: A brief re-
sponse to Johnson’s (2014) comment. Family Relations: An Interdisci-
plinary Journal of Applied Family Studies, 63, 305–308. doi:10.1111/
fare.12059
Hawkins, A. J., Blanchard, V. L., Baldwin, S. A., & 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
Hawkins, A. J., & Fackrell, T. (2010). Does relationship and marriage
education for lower-income couples work? A meta-analytic study of
emerging research. Journal of Couple and Relationship Therapy: Inno-
vations in Clinical and Educational Interventions, 9, 181–191. doi:
10.1080/15332691003694927
Hawkins, A. J., Stanley, S. M., Blanchard, V. L., & Albright, M. (2012).
Exploring programmatic moderators of the effectiveness of marriage and
relationship education programs: A meta-analytic study. Behavior Ther-
apy, 43, 77–87. doi:10.1016/j.beth.2010.12.006
Hawkins, A. J., Stanley, S. M., Cowan, P. A., Fincham, F. D., Beach,
S. R. H., Cowan, C. P., . . . Daire, A. P. (2013). A more optimistic
perspective on government-supported marriage and relationship educa-
tion programs for lower income couples. American Psychologist, 68,
110–111. doi:10.1037/a0031792
Holmes, E. K., Galovan, A. M., Yoshida, K., & Hawkins, A. J. (2010).
Meta-analysis of the effectiveness of resident fathering programs: Are
family life educators interested in fathers? Family Relations: An Inter-
disciplinary Journal of Applied Family Studies, 59, 240–252. doi:
10.1111/j.1741-3729.2010.00599.x
Hsueh, J., Alderson, D. P., Lundquist, E., Michalopoulos, C., Gubits, D.,
Fein, D., & Knox, V. (2012). The Supporting Healthy Marriage evalu-
ation: Early impacts on low-income families (OPRE Report 2012–11).
Washington, DC: Office of Planning, Research, and Evaluation, Admin-
istration for Children and Families, U.S. Department of Health and
Human Services.
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. (2014). Government supported healthy marriage initiatives
are not associated with changes in family demographics: A comment on
Hawkins, Amato, and Kinghorn (2013). Family Relations: An Interdis-
ciplinary Journal of Applied Family Studies, 63, 300–304. doi:
10.1111/fare.12060
Kitzmann, K. M., Gayford, N. K., Holt, A. R., & Kenny, E. D. (2003).
Child witnesses to domestic violence: A meta-analytic review. Journal
of Consulting and Clinical Psychology, 71, 339–352. doi:10.1037/0022-
006X.71.2.339
Knox, V., & Redcross, C. (2000). Parenting and providing: The impact of
parents’ fair share on paternal involvement. New York: Manpower
Research Development Corporation.
Laurenceau, J. P., Stanley, S. M., Olmos-Gallo, A., Baucom, B., & Mark-
man, H. J. (2004). Community-based prevention of marital dysfunction:
multilevel modeling of a randomized effectiveness study. Journal of
Consulting Clinical Psychology, 72, 933–943. doi:10.1037/0022-006X
.72.6.933
Lorber, M. F., Erlanger, A. C. E., Heyman, R. E., & O’Leary, K. D. (2014).
The honeymoon effect: Does it exist and can it be predicted? Prevention
Science. doi:10.1007/s11121-014-0480-4
Lundahl, B. W., Tollefson, D., Risser, H., & Lovejoy, M. C. (2008). A
meta-analysis of father involvement in parent training. Research on
Social Work Practice, 18, 97–106. doi:10.1177/1049731507309828
Lundquist, E., Hsueh, J., Lowenstein, A., Faucetta, K., Gubits, D., Mich-
alopoulos, C., & Knox, V. (2014). A family-strengthening program for
low-income families: Final impacts from the Supporting Healthy Mar-
riage evaluation. OPRE Report 2013-49A. Washington, DC: Office of
Planning, Research and Evaluation, Administration for Children and
Families, U. S. Department of Health and Human Services.
Markman, H. J., & Rhoades, G. K. (2012). Relationship education re-
search: Current status and future directions. Journal of Marital and
Family Therapy, 38, 169–200. doi:10.1111/j.1752-0606.2011.00247.x
Markman, H. J., Rhoades, G. K., Stanley, S. M., Ragan, E. P., & Whitton,
S. W. (2010). The premarital communication roots of marital distress
and divorce: The first five years of marriage. Journal of Family Psy-
chology, 24, 289–298. doi:10.1037/a0019481
Markman, H., Stanley, S., & Blumberg, S. (2010). Fighting for your
marriage. San Francisco, CA: Jossey-Bass.
McHale, J. P., & Fivaz-Depeursinge, E. (1999). Understanding triadic and
family group interactions during infancy and toddlerhood. Clinical Child
and Family Psychology Review, 2, 107–127. doi:10.1023/A:
1021847714749
McHale, J. P., & Lindahl, K. M. (2011). Coparenting: a conceptual and
clinical examination of family systems (1st ed.). Washington, DC: Amer-
ican Psychological Association.
Miller, G. E., & Prinz, R. J. (1990). Enhancement of social learning family
interventions for childhood conduct disorder. Psychological Bulletin,
108, 291–307. doi:10.1037/0033-2909.108.2.291
Panter-Brick, C., Burgess, A., Eggerman, M., McAllister, F., Pruett, K., &
Leckman, J. (in press). A game change in parenting interventions: A
systematic review of the global evidence base on father engagement –
and recommendations. Journal of Child Psychology and Psychiatry.
Parke, R. D. (2013). Future families: Diverse forms, rich possibilities.
Malden, MA: Wiley. doi:10.1002/9781118602386
Personal Responsibility and Work Opportunity Reconciliation Act of 1996,
Public Law 104-193. TANF Legislation, Part A, Section 401(a)(4)
(1996).
Pinquart, M., & Teubert, D. (2010). A meta-analytic study of couple
interventions during the transition to parenthood. Family Relations, 59,
221–231.
Popenoe, D. (1993). American family decline, 1960–1990. Journal of
Marriage and the Family, 55, 527–541. doi:10.2307/353333
Pruett, M. K., & Barker, R. (2009). Children of divorce: New trends and
ongoing dilemmas In J. H. Bray & M. Stanton, The Wiley-Blackwell
handbook of family psychology (pp. 463–474). New York, NY: Wiley-
Blackwell.
Pruett, M. K., & Johnston, J. R. (2004). Therapeutic mediation with
high-conflict parents: Effective models and strategies. In J. Folberg,
A. L. Milne, & P. Salem (Eds.), Divorce and family mediation: Models,
techniques, and applications (pp. 92–111). New York, NY: Guilford
Press.
Rienks, S. L., Wadsworth, M. E., Markman, H. J., Einhorn, L., & Etter,
E. M. (2011). Father involvement in urban low-income fathers: Baseline
associations and changes resulting from preventive intervention. Family
Relations: An Interdisciplinary Journal of Applied Family Studies, 60,
191–204. doi:10.1111/j.1741-3729.2010.00642.x
382 COWAN AND COWAN
Rogge, R. D., Bradbury, T. N., Hahlweg, K., Engl, J., & Thurmaier, F.
(2006). Predicting marital distress and dissolution: Refining the two-
factor hypothesis. Journal of Family Psychology, 20, 156–159. doi:
10.1037/0893-3200.20.1.156
Rogge, R. D., Cobb, R. J., Lawrence, E., Johnson, M. D., & Bradbury,
T. N. (2013). Is skills training necessary for the primary prevention of
marital distress and dissolution? A 3-year experimental study of three
interventions. Journal of Consulting and Clinical Psychology, 81, 949
961. doi:10.1037/a0034209
Schulz, M. S., Cowan, C. P., & Cowan, P. A. (2006). Promoting healthy
beginnings: A randomized controlled trial of a preventive intervention to
preserve marital quality during the transition to parenthood. Journal of
Consulting and Clinical Psychology, 74, 20–31. doi:10.1037/0022-
006X.74.1.20
Schweinhart, L. J., Montie, J., Xiang, Z., Barnett, W. S., Belfield, C. R., &
Nores, M. (2004). Lifetime effects: The HighScope Perry Preschool
Study through age 40. Ypsilanti, MI: HighScope Press.
Shapiro, A. F., & Gottman, J. (2005). Effects on marriage of a psycho-
communicative-educational intervention with couples undergoing the
transition to parenthood: Evaluation at 1-year post intervention. Journal
of Family Communication, 5, 1–24. doi:10.1207/s15327698jfc0501_1
Shapiro, A. F., Nahm, E. Y., Gottman, J. M., & Content, K. (2011).
Bringing baby home together: Examining the impact of a couple-focused
intervention on the dynamics within family play. American Journal of
Orthopsychiatry, 81, 337–350. doi:10.1111/j.1939-0025.2011.01102.x
Shelton, K. H., & Harold, G. T. (2008). Interparental conflict, negative
parenting, and children’s adjustment: Bridging links between parents’
depression and children’s psychological distress. Journal of Family
Psychology, 22, 712–724. doi:10.1037/a0013515
Spielhofer, T., Corlyon, J., Durbin, B., Smith, M., Stock, L., & Gieve, M.
(2014). Relationship support interventions evaluation. London, Eng-
land: Tavistock Institute of Human Relations, National Foundation for
Educational Research, Thomas Coram Research Unit, Department for
Education, UK Government.
Stanley, S. M., Markman, H. J., Jenkins, N. H., Rhoades, G. K., Noll, L.,
& Ramos, L. D. (2006). Within Our Reach—Leader manual. Denver,
CO: PREP Educational Products.
Stanley, S. M., Markman, H. J., Prado, L. M., Olmos-Gallo, P. A., Tonelli,
L., St. Peters, M.,...Whitton, S. W. (2001). Community-based
premarital prevention: Clergy and lay leaders on the front lines. Family
Relations: An Interdisciplinary Journal of Applied Family Studies, 50,
67–76. doi:10.1111/j.1741-3729.2001.00067.x
Stanley, S. M., Rhoades, G., Loew, B., Allen, E. S., Carter, S., Osborne, L.,
. . . Markman, H. J. (in press). A randomized controlled trial of rela-
tionship education in the U.S. Army: 2-year outcomes. Family Relations.
Statistics. (2007). General household survey. Newport, UK. http://www
.ons.gov.uk/ons/rel/ghs/general-household-survey/2007-report/index
.html
Trillingsgaard, T., Baucom, K. J. W., Heyman, R. E., & Elklit, A. (2012).
Relationship interventions during the transition to parenthood: Issues of
timing and efficacy. Family Relations: An Interdisciplinary Journal of
Applied Family Studies, 61, 770–783.
Twenge, J. M., Campbell, W. K., & Foster, C. A. (2003). Parenthood and
marital satisfaction: A meta-analytic review. Journal of Marriage and
Family, 65, 574–583. doi:10.1111/j.1741-3737.2003.00574.x
Van Widenfelt, B., Hosman, C., Schaap, C., & van der Staak, C. (1996).
The prevention of relationship distress for couples at risk: A controlled
evaluation with nine-month and two-year follow-ups. Family Relations:
An Interdisciplinary Journal of Applied Family Studies, 45, 156–165.
doi:10.2307/585286
Walsh, F. (2012). Normal family processes: Growing diversity and com-
plexity (4th ed.). New York, NY: Guilford Press.
Webster-Stratton, C. (1985). The effects of father involvement in parent
training for conduct problem children. Child Psychology & Psychiatry &
Allied Disciplines, 26, 801–810. doi:10.1111/j.1469-7610.1985
.tb00593.x
Williams, K. (2014). What can reduce unintended pregnancies and increase
stable relationships? Retrieved March 10, 2014, from https://
contemporaryfamilies.org/cohabitation-divorce-commentaries/
Wood, R. G., McConnell, S., Quinn, M., Clarkwest, A., & Hsueh, J.
(2010). Strengthening Unmarried parents’ relationships: The early im-
pacts of building strong families. Washington DC: Mathematic Policy
Research.
Wood, R. G., Moore, Q., Clarkwest, A., & Killewald, A. (2014). The
long-term effects of Building Strong Families: A program for unmarried
parents. Journal of Marriage and Family, 76, 446463. doi:10.1111/
jomf.12094
Received January 21, 2014
Revision received July 1, 2014
Accepted July 2, 2014
383
CONTROVERSIES IN COUPLE RELATIONSHIP EDUCATION
... An important differentiation should be made between 1. fathers' participation in family-based interventions in clinical practice (e.g., parenting trainings), where a large number of studies and reviews are available (Cowan & Cowan, 2014;Fagan & Pearson, 2021;Lechowicz et al. 2019;McKee et al. 2021;Panter-Brick et al., 2014), and 2. fathers' participation in empirical studies that do not focus specifically on fathers, such as studies evaluating prevention and treatment interventions or studies with developmental psychopathology research questions (Costigan & Cox, 2001;Edwards, 2012;Mathiesen et al., 2018). Fathers are underrepresented in both clinical practice and empirical studies. ...
Article
Full-text available
When compared with mothers, fathers continue to be considerably underrepresented in developmental research on child and adolescent psychopathology. The present study examined the factors contributing to fathers’ participation in such studies in which both mothers and fathers were asked to participate. The sample consisted of 477 families which were examined four times over the course of ten years (Pre to Follow-up after ten years [FU10]). The average age of the children at Pre was 4 years, and that of the adolescents ten years later was 14 years (FU10). On average, 70.4% of fathers participated in the assessments. The average transition probability for participation over four assessment points was 93.1%. Fathers from two-parent families participated significantly more often than fathers from single-parent families. Household income, a good father-adolescent relationship, and joint custody, if applicable, were related to paternal participation. For fathers from two-parent families, participation at previous assessment points was the decisive predictor. For fathers from single-parent families, the relationship between the adolescent and the father was the decisive predictor. Fathers who had participated in previous assessments remained involved in the study over the long run. Therefore, the primary goal should be to encourage as many fathers as possible to participate when starting clinical empirical studies. Monetary incentives could be used to recruit families with low household incomes.
... Department of Health & Human Services, 2020.). These grants supported educational groups, marriage checkups, and other relationship wellness programs with the aim of making them more widely available to a range of heterosexual, cisgender, married couples (Cowan & Cowan, 2014). Prevention of relationship distress and the promotion of family wellness was a noteworthy goal for these initiatives. ...
Article
Full-text available
Little research has been done on what motivates gender, romantic, and sexual minority (GRSM) couples to attend relationship wellness programs or what incentives they find appealing. For this study, 18 participants in dyadic partnerships responded to questions about their motivations to participate in a brief relationship wellness program as well as what financial incentives held the most appeal. All nine couples shared that they were motivated to do the wellness program to contribute to research for the GRSM community, suggesting that altruistic motivations are particularly relevant to GRSM couples. In regard to financial incentives, gifts that supported their relationship health, such as a gift card for a meal out together, held the strongest appeal. The results in this preliminary study may assist in refining incentives and other strategies to engage these often hard-to-reach couples.
... The interest and availability of marriage and relationship education programs have increased in the United States since the 1950s (Cowan & Cowan, 2014;Cowan, Cowan, & Knox, 2010), creating a movement in community education to strengthen marriage and decrease risk of divorce (Larson, 2004). Relationship and marriage education (RME) programs, sometimes called marriage enrichment programs, refer to preventative educational programs that work with couples before conflicts or problems become too serious, rather than remedial or therapeutic approaches that are aimed at relationship repair (Larsen, 2004). ...
Article
Relationship Marriage Education (RME) varies and studies show the effectiveness of single-session events. This program explores RME event logistics, and changes in participant perceived relationship knowledge following single-session RME events.
Article
We posit that evaluators of relationship education interventions can explore a policy relevant outcome by assessing relationship stability and conflict in a single index that is based on the literature on the effects of divorce and marital distress on children. We provide an empirical example from a randomized trial. The U.S. Administration for Children and Families funds community‐based projects using relationship education with a foundational goal of fostering stable and healthy relationships. Assessing this outcome requires an approach different from separately analyzing stability and relationship quality. We used data (N = 1,156 couples) from a randomized trial of Family Expectations to test the Stable Low‐Conflict Index, comparing couples assigned to the intervention to couples assigned to an untreated control group at a follow‐up 8 to 9 months postintervention. Intervention couples were more likely to be in a stable, low‐conflict relationship at the follow up than control couples (b = .36, SE = .15, odds ratio = 1.44, p = .014). An index based on empirical precedence showed evidence of an intervention impact in a community‐based program. Evaluators of family policy linked interventions may advance the field by exploring outcomes that encapsulate aspects of both relationship stability and quality.
Article
For decades, researchers, interventionists, and the lay public have subscribed to the notion that couples low in relationship satisfaction and/or experiencing psychological, physical, or sexual intimate partner violence (IPV) have communication skills deficits. In contrast, experimental studies of communication have concluded that differences were more likely due to partners' "ill will than poor skill." We revisited this debate by recruiting a fairly generalizable sample of couples (N = 291) via random-digit dialing and asking them to discuss two top conflict areas ("at your best" and "as you typically do"), thus measuring will-conscious inhibition of hostility and negative reciprocity and production of positivity (i.e., the "conflict triad"). The conflict triad was observed with the Rapid Marital Interaction Coding System, Second Generation. We found partial support for the hypotheses grounded in Finkel's I3 meta-model. Frequency of hostility was associated with a complicated satisfaction × IPV-extent × conversation type × gender interaction, indicating that couples' communication skills are multi-determined. Unhappier couples showed almost no change in positivity when at their best, whereas happier couples nearly doubled their positivity despite their considerably higher typical positivity mean. Negative reciprocity was associated with satisfaction and IPV-extent but not conversation type, implying that immediate instigation combined with risk factors overwhelms conscious inhibition. Intervention implications are discussed.
Article
The current study investigated changes in couple, parenting, and individual functioning following participation in Family Expectations, a relationship and parenting education program for new or expectant parents. The sample comprised 339 couples who participated in most sessions of the Family Expectations program and completed assessments at three different time points over a 12‐month period. Study analyses examined: (a) change shortly following completion of the program, (b) associations between short‐term change and subsequent change in outcomes at 12‐month follow‐up, and (c) differences in short‐term change between married and unmarried couples. Significant improvements were observed in all three domains at short‐term follow‐up. Short‐term changes, particularly for psychological distress, were predictive of long‐term change in multiple domains. Few moderation effects by marital status were evident; those that appeared suggested stronger effects for married participants compared to unmarried. Study findings inform ongoing discussions into the utility of federally‐funded relationship and parenting programming.
Article
The focus of this special section is impact studies that were conducted as part of the Administration for Children and Families' Healthy Marriage and Responsible Fatherhood initiative. This initiative has led to more than 2 million people receiving relationship education and fatherhood programs across the United States over the last two decades. Community organizations develop these programs, select their curricula and content, and run these programs in their local communities. Many programs funded in the 2015-2020 cohort of grantees included randomized controlled trials of program effectiveness; some of these studies are presented in this issue, with the goal of marrying the field of program evaluation and family science.
Article
The federal government, through the Administration for Children and Families (ACF), has funded community‐based relationship education programs for couples, individuals, and families, with a strong focus on serving economically disadvantaged and racially diverse families. This study evaluated the impact of a 36‐hour, workshop‐based couple relationship education program that was funded by ACF using a randomized controlled trial (RCT) design and intent‐to‐treat (ITT) analyses. Participants were 1320 couples who were either expecting a baby or had a baby within the past 3 months, at the time of enrollment. Follow‐up surveys were administered 12 months later. Analyses evaluated program impacts on relationship stability, constructive communication, and destructive conflict compared to a no‐treatment control group. Analyses showed a statistically significant impact of the program on destructive conflict (d = 0.10) but not on constructive communication (d = 0.06) or stability (dCox = 0.10). Based on findings from previous evaluations, we also examined whether participants’ levels of sociodemographic disadvantage moderated these effects. There was significant moderation by sociodemographic disadvantage on constructive communication and destructive conflict, but not on stability. Effects were observed for those at higher levels of sociodemographic disadvantage. 联邦政府通过儿童和家庭管理局(ACF)资助了以社区为基础的夫妻、个人和家庭的关系教育计划,重点是为经济困难和种族多样化的家庭服务。本研究采用随机对照试验(RCT)设计和意向治疗分析(ITT),评估了由ACF资助的36小时的讲习班式伴侣关系教育项目的影响。参与研究的1320对夫妇在注册时正在怀孕或在过去三个月内已经有了孩子。随访调查在12个月后进行。分析评估项目对关系稳定、建设性沟通和破坏性冲突的影响,与无治疗对照组进行比较。分析显示,该项目在破坏性冲突(d = 0.10)上有显著的统计影响,但在建设性沟通(d = 0.06)或稳定性(dCox = 0.10)上没有显著影响。基于先前评估的结果,我们还检查了参与者的社会经济背景不利水平是否会调节这些效应。社会经济背景方面的不利情况对建设性沟通和破坏性冲突有显著的调节作用,但对稳定性没有显著的调节作用。对那些处于较高社会经济背景不利水平的人,影响是明显的。
Article
The field of relationship science has called for more research on the impact of relationship education on child outcomes, yet studies in this area remain sparse, particularly regarding maternal and infant health at birth. Research on group prenatal care demonstrates that individual‐oriented group interventions have a positive impact on infant birth outcomes, suggesting the need to consider the impacts of other forms of group programming for women. The current study examined the impact of MotherWise, an individual‐oriented relationship education and brief case management/coaching program for minority and low‐income pregnant women, on birth outcomes. The study sample included 136 women who enrolled in a larger randomized controlled trial of MotherWise during early pregnancy. Although statistical power was limited due to the sample size and the effects were not outright significant at p < 0.05, results indicated that the effects of MotherWise on birth outcomes were small to moderate in size (0.23 for birthweight, 0.46 for preterm birth) and suggest important avenues for future tests of relationship education programs and their impacts on maternal and infant health. The current study suggests that relationship education during pregnancy could directly impact women's and infant's health. 关系科学领域呼吁有更多的研究来针对关系教育对儿童结果的影响。这一领域的研究仍然很少,特别是在产妇和婴儿出生时的健康方面。关于群体产前护理的研究表明,面向个人的群体干预措施对婴儿出生结果具有积极影响,这表明需要考虑其他形式的群体方案对妇女的影响。本研究调查了MotherWise对分娩结果的影响,这是一项针对少数族裔和低收入孕妇的以个人为导向的关系教育和简单案例管理/指导项目。该研究的样本包括了136名妇女,她们在早期怀孕期间参加了MotherWise的一项更大的随机对照试验。虽然由于样本量的原因,统计能力有限,且在p < .05时的影响并不完全显著。结果表明,MotherWise对出生结果的影响在样本量上是小到中等的(出生体重为23,早产为46),并为未来的关系教育项目测试及其对母婴健康的影响提出了重要的途径。本研究表明,怀孕期间的关系教育可以直接影响妇女和婴儿的健康。
Article
A long-standing university-community partnership used a longitudinal randomized control trial to implement and evaluate two couple relationship education (CRE) curricula, ELEVATE and Couples Connecting Mindfully (CCM), among an economically and racially diverse population of adult couples. Married and non-married couples (n = 929 couples) completed baseline surveys and were randomly assigned to either one of the two program groups or to the control group by implementation site. Follow-up surveys were collected at 2 months, 6 months, and 1 year after baseline. Using an intent-to-treat approach, growth curve modeling comparisons of trajectories indicated program effects at 1 year post-baseline in key outcome areas. Both the ELEVATE and the CCM group reported significant gains in couple relationship skills, couple quality, and family harmony over time compared to the control group that experienced either no change or declines. Further, the ELEVATE group also demonstrated positive program effects on measures of mental health and sleep quality. An assessment of the central premise of CRE indicated that the immediate post-program improvements in couple relationship skills predicted later couple quality for both program groups. This study indicates that both ELEVATE and CCM can be considered evidence-based CRE programs for use with a broad population of couples.
Chapter
This chapter draws on three important themes in contemporary research on families and close relationships to provide new insights regarding important precursors of the quality of romantic unions during emerging adulthood. As Arnett (2004) noted, emerging adulthood, a period from 18 to 25 years old or perhaps even older, is a time of exploring possibilities and opportunities before assuming adult roles. In particular, Arnett (2004) suggested that it is a time for young people to explore their options in romance with different partners to discover what kind of person they would like to marry and to gain relationship experience before choosing someone to marry. Therefore, romantic relationships during this period could be self-focused and unstable. However, as Collins and van Dulmen (2006) pointed out, there is substantial continuity in close relationships (e.g., parent–child relationships, friendships, romantic relationships), and development in one period of life is built on development during an earlier time. This view is consistent with the life course perspective (Elder, 1985) that people's life trajectories are determined by a series of linked stages in which transitions from one state to another are always embedded in and have an impact on those trajectories. Therefore, the establishment of stability, satisfaction, and closeness in romantic relationships is one of the major developmental tasks during emerging adulthood.
Article
The Family Context of Parenting in Children's Adaptation to Elementary School is a result of a longitudinal prevention study of 100 families begun the year before their first children entered kindergarten. Each family went through an assessment and then a subset was randomly chosen for group intervention. The children in both groups were then studied as they progressed through kindergarten and first grade to assess the quality of their adaptation to the school environment. The text focuses on how parent-child relationships are only one determinant of a child's academic competence, social competence, and behavior. Rather, these relationships must be understood in the context of the role they play within the family as a system. It also addresses the recent challenges to claims about the impact of parents on their children's development. The book sheds additional light on family influences within the larger social environment as a key determinant of the quality of children's adjustment to schooling. It appeals to scientists, professionals, and parents alike. © 2005 by Lawrence Erlbaum Associates, Inc. All rights reserved.
Article
The theme for this year's National Childminding Week is ‘supporting families’, which means childcarers can raise awareness of the valuable work they perform, not only with children but with parents also.
Article
The Supporting Healthy Marriage (SHM) evaluation was launched in 2003 to test the effectiveness of a skills-based relationship education program designed to help low-income married couples strengthen their relationships and, in turn, to support more stable and more nurturing home environments and more positive outcomes for parents and their children. The evaluation is led by MDRC, in collaboration with Abt Associates and other partners, and is sponsored by the Department of Health and Human Services. The SHM program is a voluntary, yearlong, relationship and marriage education program for lowincome, married couples who have children or are expecting a child. The program provides group workshops based on structured curricula; supplemental activities to build on workshop themes; and family support services to address participation barriers, connect families with other services, and reinforce curricular themes. The study’s rigorous random assignment design compares outcomes for families who are offered SHM’s services with outcomes for a similar group of families who are not offered SHM’s services but can access other services. This report presents estimated impacts on the program’s targeted outcomes about one year after couples entered the study.