Parents as Partners: A U.K. Trial of a U.S.
Couples-Based Parenting Intervention For At-Risk
PHILIP A. COWAN
CAROLYN P. COWAN
Despite the well-established links between couple relationship quality and healthy
family functioning, and burgeoning evidence from the international intervention field,
there is little or no evidence of the efficacy of couples-based interventions from the United
Kingdom (U.K.). This study explored whether the Parents as Partners (PasP) program, a
group-based intervention developed in the United States, brought about the same benefits
in the U.K. The evaluation is based on 97 couples with children from communities with
high levels of need, recruited to PasP because they are at high risk for parent and child psy-
chopathology. Both mothers and fathers completed self-report questionnaires assessing
parents’ psychological distress, parenting stress, couple relationship quality and conflict,
fathers’ involvement in child care and, importantly, children’s adjustment. Multilevel mod-
eling analysis comparing parents’ responses pre- and postintervention not only showed
substantial improvements for both parents on multiple measures of couple relationship
quality, but also improvements in parent and child psychopathology. Analyses also indi-
cated most substantial benefits for couples displaying poorest functioning at baseline. The
findings provide initial evidence for the successful implementation of PasP, an American-
origin program, in the U.K., and add support for the concept of the couple relationship as a
resource by which to strengthen families.
Keywords: Couple Relationship; Interparental Conflict; Parenting; Child Adjustment;
Fam Proc x:1–18, 2017
For decades, the British Government has invested in parenting programs as a way of
enriching the developmental environment for children in families not identified as in
distress, and as a way of providing remedial help for children in troubled families at high
*Tavistock Relationships, London, UK.
University of California, Berkeley, Berkeley, CA.
Correspondence concerning this article should be addressed to Polly Casey, Tavistock Relationships, 70
Warren Street, London W1T5PB, UK. E-mail: firstname.lastname@example.org
The authors thank Family Action staff for their commitment to the project, both as caseworkers and dili-
gent data collectors. Special thanks to Richard Burge for his careful combing of records to gather referral
information for this paper. Thanks also to TCCR groupworkers Anna Fitzgerald, James Ganpatsingh,
Krisztina Glausius, Hendrix Hammond, Anna Learmonth, and Liz Mawle, for their expertise and dedica-
tion. Finally, thanks to the families who gave their time and effort over their course of the project.
Family Process, Vol. x, No. x, 2017 ©2017 Family Process Institute
risk for maladaptive development (Department for Education and Skills, 2007). Although
the use of the descriptor “parenting” implies that these interventions are for both mothers
and fathers, in fact classes and therapy sessions are attended mostly by mothers (Panter-
Brick et al., 2014). The notion of the couple relationship as a resource to strengthen fami-
lies has been gaining traction in the United Kingdom (U.K.) and elsewhere, as reflected in
an announcement by previous Prime Minister David Cameron to double government fund-
ing for relationship support to £70 million over the remaining period of this Conservative
government (Prime Minister’s Office, 2016). The Prime Minister’s statement was followed
by a government-funded report issued by the Early Intervention Foundation (Harold,
Acquah, Sellers, & Chowdry, 2016), which reviewed conclusive evidence that unresolved
conflict between parents is associated with less effective parenting and more negative out-
comes for children and adolescents. Summarizing the existing findings, the report con-
cludes that there is some initial evidence from international studies that interventions for
parent couples in fact produce benefits for children, but no existing evidence-based pro-
grams have yet emerged in the U.K.
This paper describes an attempt to address that gap. Based on the results of successful
intervention trials in California, U.S. (Cowan & Cowan, 1992; Cowan, Cowan, Ablow,
Johnson, & Measelle, 2005; Cowan, Cowan, Pruett, Pruett, & Wong, 2009), the Depart-
ment for Education in the U.K. funded a feasibility study of the efficacy of the Parents as
Partners Program (PasP), comprising a 16-week couples curriculum modeled directly on
the latest version of the U.S. intervention, Supporting Father Involvement (SFI). This
report describes changes from pre- to postintervention for the first 97 couples to complete
the U.K. program. The first question addressed here is not whether participants fare bet-
ter than couples in a no-treatment control group, but rather whether the intervention pro-
duces the kinds of changes in U.K. participants that it did in the U.S. The second question
addressed here is whether the program produces differential effects for “high” and “low”
functioning couples, so categorized according to baseline scores on a number of indices of
COUPLE RELATIONSHIPS, PARENTING, AND CHILDREN’S DEVELOPMENT
Traditionally, the target of family intervention has been promoting positive parenting
practices and parent–child relationships. Harold and Leve (2012) argue that the most
“substantial dividends” can be gained by targeting investment at the level of the inter-
parental relationship because of the centrality of this relationship in the family system,
the high probability of spillover from the couple relationship into mother–child and
father–child relationships, and the potential for wide-reaching, short- and long-term, ben-
efits for families. These assertions are based on a large body of international research
amassed over several decades attesting to the robust association between the quality of
the couple relationship and other indices of family functioning (e.g., Cummings & Davies,
2002; Fincham & Beach, 2010).
Strong, supportive, and harmonious interparental relationships have consistently been
linked to children’s enhanced psychological wellbeing (Cowan & Cowan, 2002; Davies
et al., 2002; Grych, Harold, & Miles, 2003) and educational attainment (Harold, Aitken, &
Shelton, 2007). Mothers and fathers who are satisfied in their relationship with each other
tend to establish relationships with their children that include warmth, responsiveness,
limit-setting, and maturity demands (Cowan & Cowan, 1992; Fauchier & Margolin, 2004;
Kaczynski, Lindahl, Malik, & Laurenceau, 2006; Sturge-Apple, Davies, & Cummings,
2006), the hallmarks of authoritative parenting (Baumrind, 1971). Not only are harmo-
nious parents more positive in their separate engagements with their children, but they
are also more likely to establish a positive coparenting relationship (McHale &
Lindahl, 2011; Pruett & Pruett, 2009)—supporting rather than undermining each other’s
approach to the child. Coparenting is recognized as linked to, but distinct from, other
aspects of the couple relationship such as intimacy or couple-related conflict. Specifically,
the coparenting relationship (between parents in intact or separated relationships) refers
to “the way in which coparents work together in their role as parents” (Feinberg, 2003, p.
1499)—forming an alliance and providing support to one another in this respect, and
making shared decisions about child-rearing.
Furthermore, a strong couple relationship provides a buffer against negative influ-
ences from outside the nuclear family. For example, a positive relationship between the
partners disrupts the cycles of negative relationship behaviors that tend to be passed
from generation to generation without any intervention (Amato & Booth, 2001; Amato
& DeBoer, 2001). Similarly, strong couple relationships protect the family from the ten-
dency for economic hardship and job loss (Masarik et al., 2016) to result in increased
couple conflict, decreased parenting effectiveness, and problematic outcomes for chil-
dren. These correlational studies all suggest that interventions to help strengthen cou-
ple relationships and parents’ capacity to work together could produce a healthy return
on investment, due to the positioning of the interparental relationship at the fulcrum of
the family system.
A COUPLES INTERVENTION FOR LOW-INCOME FAMILIES WITH EVIDENCE OF
EFFECTS ON CHILDREN
Reviews of the international couples intervention literature make four important
points. First, in studies of family strengthening interventions, fathers are almost always
excluded, yet when fathers attend parenting interventions along with mothers, the results
are significantly stronger for parenting quality and children’s outcomes (Panter-Brick
et al., 2014). Second, several studies of couples group interventions show that interven-
tion-induced changes in couple relationship quality affect positive changes in parenting.
However, these studies did not provide data on whether the effects on parenting extended
to the children (Cowan & Cowan, 2014). Third, two recent comprehensive surveys of the
international literature searched for couples-based interventions that examine the effects
of parents’ participation on how their children fare. Both supported the hypothesis that
children benefit from their parents’ participation in a successful couples group interven-
tion. None of the intervention studies picked up by the reviews took place in the U.K.
(Cowan & Cowan, 2014; Harold et al., 2016).
A further point to note is that many of the these interventions are “universal” in their
offer, referring to the fact that recruitment into the program is not restricted to couples
with certain demographic characteristics or levels of psychological risk. These interven-
tions are typically evaluated across the entire sample of attendees, with differences
between average pre- and postintervention scores being interpreted as program effects for
all, regardless of baseline characteristics. However, authors from diverse disciplines have
recommended that the impact of intervention programs should be estimated according to
specific subgroups of participants, defined by pretreatment characteristics (e.g., Bloom &
Michalopoulos, 2013; Fournier et al., 2010; Quirk, Strokoff, Owen, France, & Bergen,
2014). An analysis of subgroups may help to clarify the intervention effects, and contribute
to the discussion of “what works best for whom?”, which would be relevant for those with
research, economic, and political interests. At present, very few couple intervention evalu-
ations have examined outcomes for specific subgroups; those that have tend to be from the
field of relationship education and they indicate stronger intervention effects for economi-
cally disadvantaged couples (Amato, 2014) or couples at high risk for relationship break-
down (e.g., Halford, Sanders, & Behrens, 2001).
Fam. Proc., Vol. x, xxxx, 2017
CASEY ET AL.
In their research program, the Cowans began with two relatively small studies
(N=100 families in each) of couples groups for primarily working-class and middle-class
mothers and fathers (i) making the transition to parenthood (Becoming a Family; Cowan
& Cowan, 1992) or (ii) having a first child entering the elementary school system
(Schoolchildren and their Families; Cowan et al., 2005; Cowan, Cowan, & Barry, 2011).
Next, along with Marsha Kline Pruett and Kyle Pruett, the Cowans recruited larger,
higher-risk samples of low-income Mexican American, African American, and European
American couples in five rural and urban California counties (SFI; Cowan et al., 2009;
Cowan, Cowan, Pruett, Pruett, & Gillette, 2014) that involved more than 800 families.
Taken together, the three U.S. studies included three RCTs and one pre–post replication,
and provide strong evidence that a couples’ group intervention could not only reduce dis-
tress between the parenting partners, but also improve parent–child relationships and
reduce the risks for children’s behavior problems. These reports have been followed by an
investigation by Epstein et al. (2015) into the differential impact of the SFI intervention
for couples initially reporting low, medium, and high levels of conflict, which showed the
greatest and most immediate intervention effects for couples with high levels of conflict at
THE PRESENT STUDY
This report describes a first exploration of the feasibility of adapting, implementing,
and replicating a U.S.–designed program in a U.K. context, which is offered to families
from communities with high levels of need. We report changes in couples participating in
the PasP intervention on a range of indicators of family functioning that include couple
relationship quality and conflict, father involvement, mothers’ and fathers’ psychological
wellbeing, and children’s emotional and behavioral difficulties. In the analysis of chil-
dren’s problem behaviors we set the analysis up in two ways: First, we examined changes
in children’s behaviors over the course of the program differentially according to mothers’
and fathers’ reports. In this analysis we controlled for child age and gender, because we
wanted to remove some of the variation in this relatively small sample while focusing on
reporter effects. Second, instead of removing variation due to child gender from the analy-
sis, we made this a focus of the analysis by examining changes over the course of the pro-
gram differentially according to child gender. This is because of the overwhelming
evidence that boys and girls show different levels of problem behaviors (Cummings,
Davies, & Campbell, 2000). In the second analysis we still controlled for the age of the
The current study adds significantly to the couples’ intervention literature by going
beyond an examination of average intervention effects across the sample to explore
patterns of change for different groups of parents distinguished by their levels of initial
The PasP program was located in Tavistock Relationships in central London, which is a
U.K. voluntary sector clinical training, practice, and research organization that special-
izes in delivering couple counseling and psychotherapy. Professors Philip and Carolyn
Cowan from the University of California, Berkeley, provided advice, training, support,
and ongoing feedback to the program in order to preserve fidelity to the intervention
model. Tavistock Relationships worked in close partnership with Family Action, a
voluntary organization with an established history of delivering family support to commu-
nities across the U.K. Couples were recruited from six London boroughs and, in the second
year, from Manchester as well. These were all areas with high levels of need; five of the
seven areas fall into the top 20 of 326 local authority districts in England with the highest
proportion of children from families in income deprivation (people who are out of work and
those who are in work but have low earnings; Department for Communities and Local
Government, 2015). In each of these areas, advertisements were published in the local
press and on social media, leaflets were sent through schools and community centers, and
professional networks were alerted. Because the program requires joint attendance by
mothers and fathers, particular attention was paid to the engagement of men, both by
direct publicity (e.g., in sports clubs) and by encouraging social workers and health profes-
sionals to focus on men.
As in the U.S. Supporting Father Involvement intervention, couples eligible for the pro-
gram were those with a child under 11, living together or separated. Exclusion criteria
were: couples or coparents currently involved in court proceedings; couples in current,
ongoing domestic violence; or couples with drug or alcohol addiction, or mental ill-health
problems that made participation in a group impossible. The adults also needed sufficient
facility in the English language to take part in a group conducted in English. A brief
screening interview with both parents was conducted by a Family Case Worker. Of 219
couples who expressed initial interest, 142 met the inclusion criteria (see Figure 1 Flow
After attending the Initial Screening Interview, each eligible couple was offered a
2-hour interview with the Group Leaders to introduce the program and assess risk. This
included an interview with the couple and individual, separate interviews with each par-
ent. During this process of assessment, Group Leaders seek to be satisfied that both mem-
bers of the couple can tolerate being in a group setting as a pair without this precipitating
extreme anxiety for them, and without distressing or disrupting the group too much (Stock
Whitaker, 2001). Additionally, Group Leaders may seek advice from referrers or mental
health professionals who have worked with the individual parent as to whether he or she
is suitable for the group. Finally, Group Leaders also take into consideration whether the
couple is able to contribute to the task of the group, and not distract from it (Foulkes,
1964). A high proportion of those interviewed agreed to participate in the 16-week pro-
gram and fill out the pretreatment questionnaires (73%) and almost all of those who com-
pleted the questionnaires went on to attend the group meetings. Retention was excellent.
Of those who started groups, 88% completed the program and 80% completed the postin-
tervention questionnaires. The median number of the 16 sessions attended was 14 (M 13.7
sessions, SD 2.14). Of the parents who dropped out before the end of the program, the
number of sessions attended ranged from 1–6, with an average of 2.9 sessions attended
(SD 1.61). There was no difference in the attendance of mothers and fathers (if parents
dropped out, this tended to be as a couple). Regardless of how many sessions they
attended, all participants were included in the analysis of pre- to postintervention changes
(i.e., akin to an intention-to-treat analysis).
The sample described in the present paper comprises 97 couples who attended the first
18 Parents as Partners groups offered. Sixteen groups took place across six London bor-
oughs, and two groups took place in Manchester. 89% of parents provided information
with regard to their ethnicity. Of these parents, the majority were White British/Irish
(43%), and a quarter (26%) described themselves as “White other.” 19% of parents
were Black (African or Caribbean), 6% were from a mixed ethnic background, 4% were
Fam. Proc., Vol. x, xxxx, 2017
CASEY ET AL.
Chinese, and the remaining 2% selected “Other.” 55.29% of parents were in an inter-ethnic
union with their partner, higher than the 9% average for England and Wales (Office for
National Statistics (ONS), 2014). Here we are using the ONS definition of an inter-ethnic
relationship, which refers to people in a couple relationship who each identify with an eth-
nic group different from the other partner. Therefore an inter-ethnic relationship could
describe a relationship in which one partner identifies as Black Caribbean and the other
identifies as White British, for example, as well as relationships between partners within
the same broad ethnic group categories such as White British and White Irish. Almost
98% of parents provided information with regard to their relationship status. The majority
of parents were married (45.4%) or cohabiting (31.4%). Others were in a relationship and
raising children together, but living apart (10.8%), and a small proportion (10.3%) had
ended their romantic relationship but were raising children together. 92.3% of parents
Initial screening interviews
Group Worker interviews
Completed pre-group measures
Completed post-group measures
FIGURE 1. Recruitment and Retention.
gave information about their employment status: Most were employed full-time (39.1%) or
part-time (22.9%); 15.6% were unemployed as they entered into the program; and a fur-
ther 17.9% were full-time homemakers/carers. Two parents (1.1%) were retired and 6
(3.4%) were full-time students. 7.9% (n=13) parents said that they were receiving bene-
fits, although 15.5% of parents did not provide this information so it is possible that this
number is higher. Similarly, 3.1% said that they were currently receiving statutory sick
pay (money paid to you by your employer if you are ill and unable to work). The partici-
pants were relatively highly educated. The majority had qualifications at least to GCSE
level (94.2%; equivalent to the U.S. High School Diploma); 26.9% were educated to under-
graduate level and a further 22.8% had postgraduate qualifications. The median age of
mothers was 39 years, and of fathers was 40 years. At baseline the number of children in
each household ranged from 1–5, with most parents having 1 (34.0%) or 2 children
(39.2%). The median age of the youngest child was 5.4 years, and ranged from 6 months
to 11 years.
The Parents as Partners Group Intervention
The PasP program underwent very few modifications in the process of implementing it
in the U.K. The model itself was unchanged; in content, structure, and delivery, the pro-
gram was identical to that of the original in the U.S. Groups of five or six couples met for
16 weekly 2-hour sessions. Cr
eches (child care) were provided for the children alongside
each couples group. Each group was led by a male–female pair, who were both qualified
social work or mental health professionals. Each couple also had an allocated Family Case
Worker, who, as well as administering the pre- and postgroup questionnaires, was avail-
able to troubleshoot any difficulties in attendance, and link families with additional sup-
port services where needed. All Group Leaders were trained in the PasP model of
intervention, and followed a manualized curriculum. The curriculum is organized around
five domains, based on potential risks and buffers for optimal family functioning, and each
session is focused on risks and protective factors in each of those domains: (1) individual
mental wellbeing; (2) couple relationship qualities; (3) the quality of relationship of each
parent to the child(ren); (4) the impact of intergenerational themes, the culture of each
parent’s family of origin, and relationships with their in-laws; and (5) supports and stres-
sors in the world outside the family. Of the 16 sessions, 3 were dedicated to individual
issues, 5 to the couple relationship, 4 to parenting, 2 to intergenerational issues, and 1 to
stresses and supports outside the family. The final session was used to tie all of the
The delivery of the model occupies a territory midway between a psychoeducational
parenting group and open–ended group psychotherapy. Here open–ended refers to the
first part of each of the 16 sessions, an open “check-in” for parents to bring their own ques-
tions and concerns that may have arisen since the previous session. Depending on the
urgency of the issues, this “check-in” may last from 20–50 minutes of the 2-hour session,
and the content is open-ended in the sense that parents can raise issues that are pertinent
to their particular situation (e.g., couple, individual, parent–child issues). The sessions
then go on to focus on one of the five domains using exercises, discussions, and activities
in pairs and small groups. Two of the sessions (session 5 and session 11, both covering the
parent–child relationship) are conducted separately for mothers and fathers; in the second
hour, fathers are joined by their youngest child. The reason for splitting coparents for
these sessions is to create the opportunity for smaller, more intimate groups, which can
allow for deeper or freer conversation. The groups were separated according to gender for
two reasons (a) so that it would facilitate the women’s staying engaged with each other
while the program supported men’s involvement, at the same time that the men played
Fam. Proc., Vol. x, xxxx, 2017
CASEY ET AL.
with and talked about being fathers to their children, and (b) to enable both partners to
raise issues that they were not willing or able to raise in the combined couples groups. The
only changes to the program in order to adapt for U.K. delivery were in the language used
in the manual and by the intervention team (cultural references to professional agencies
and systems) and to the name of the program. The decision to change the name of the pro-
gram from “Supporting Father Involvement” to “Parents as Partners” in part reflected the
locus of the U.K. program within a couple therapy organization (TR), for whom the pre-
dominant focus is strengthening the couple relationship. However, the name change was
also in recognition of the shift in prevailing patterns of gendered parental involvement in
families in the U.K. It was felt that there is more public awareness of the benefits of father
involvement in recent years (the U.S. program was instituted in 2002), that local authori-
ties had already done much to promote this, and thus, speaking to the “couple” as a
partner unit was more relevant.
Parents were asked to complete a booklet of questionnaires before starting group ses-
sions and again within a month of the final group session. Pregroup questionnaires were
administered by Caseworkers and Group Leaders at the initial interview stage (the major-
ity were administered by Caseworkers; only the Clinical Outcomes in Routine Evaluation
Outcome Measure [CORE-OM] and the Couple Communication questionnaire were
administered by Group Leaders because they were also used by them as screening aids),
and postgroup questionnaires were administered by Caseworkers at a visit with couples
after 16th session.
Parents’ psychological wellbeing
Clinical Outcomes in Routine Evaluation—Outcome Measure (CORE-OM; Evans et al.,
2002): The CORE-OM is a widely used 34-item self-report questionnaire designed to tap
into clients’ global psychological distress, including items that focus on risk to self and
others. Respondents answer questions about how they have been feeling over the last
week, using a 5-point scale ranging from “not at all” to “most or all of the time.” Total
scores range from 0 to 40, with higher scores representing greater distress. The cut-off
between clinical and nonclinical populations is a score of 10 on the CORE-OM (Connell
et al., 2007). Internal consistency for this sample was excellent for both mothers and
fathers (both with a Cronbach’s alpha of .93).
Parenting Stress Index (PSI; Abidin, 1997): The PSI is a standardized measure of stress
associated with parenting. Subscale scores are obtained for Parental Distress, Parent–
Child Dysfunctional Interaction, and Difficult Child, and combined to create a total score.
Total scores were used in the present analysis, and can range from 0–180, with higher
scores representing greater levels of stress. Scores greater than 90 indicate clinically sig-
nificant levels of stress. Internal consistency in the sample was excellent (Cronbach’s
alphas of .92 and .94 for mothers and fathers, respectively).
Couple relationship quality
Quality of Marriage Index (QMI; Norton, 1983): The QMI is a 6-item measure of each
partner’s satisfaction with the couple relationship. Maximum scores on the QMI can reach
45, with higher scores representing greater satisfaction. Cronbach’s alphas of .93 and .92
for mothers and fathers, respectively, indicated excellent internal consistency for this
measure in the current sample.
The Couple Communication Questionnaire (Cowan & Cowan, 1990a) is a 27-item mea-
sure of the amount of conflict between partners, specific areas of conflict, and the
strategies employed by the couple in dealing with conflict. We used three subscales (score
range in parentheses): (a) overall frequency of conflict (0–78), (b) conflict about the chil-
dren (0–12), and (c) violent problem solving (0–14), a 14-item scale measuring the com-
bined occurrence of physical and verbal violence between partners. Within the violent
problem scale are 7 behaviors referring to the respondent and the same behaviors
repeated but referring to the partner (e.g., when we attempt to solve a marital problem “I
push, grab or shove my partner” and “my partner pushes, grabs or shoves me”). In these
items, 1 of 7 is verbal (“I yell or insult my partner”), 2 are nonverbal, nonhitting (e.g.,
stomping out of the room, throwing things), and 4 involve physical violence against the
partner (e.g., hitting or slapping your partner). Internal consistency in this sample for the
three subscales was adequate, with alphas ranging from .76–.88 for mothers and .67–.76
Who Does What? (Cowan & Cowan, 1990b): Who Does What? is an 11-item question-
naire completed by both mothers and fathers to assess fathers’ relative involvement in the
care of the couple’s youngest child (e.g., feeding, laundry, responding to cries). Responses
to items are on a scale of 1 (she does it all) to 9 (he does it all), with midrange scores (5)
indicating that the couple shares child-rearing tasks equally. High scores represent
greater involvement by fathers in each task. Mothers’ and fathers’ responses indicated
good internal consistency for this measure in this sample (Cronbach’s alphas of .80 for
both mothers and fathers).
Children’s emotional and behavioural difficulties
Strengths and Difficulties Questionnaire (SDQ; Goodman, 1994, 1997): Mothers and
fathers completed the SDQ about their youngest child. The SDQ produces an overall
assessment of the child’s psychological state; total scores are used here, which can range
from 0–40, with high scores representing greater difficulties. Woerner, Becker, and
Rothenberger (2004) recommend that total SDQ scores of 16 and above are in the “abnor-
mal” range (approximately 10% of a community sample scores in the abnormal band);
scores of 13–15 are in the “borderline” range (8.4% of a community sample); and scores of
0–12 are in the “normal” range (81.6% of a community sample (Woerner et al., 2004).
Pre- and postgroup mean scores are presented in Table 1, model coefficients in Tables 2
Data were analyzed using Hierarchical Linear Modelling (HLM; (Raudenbush & Bryk,
2002), also called multilevel modeling (Snijders & Bosker, 1999). HLM allows researchers
to study the trajectory of individual change over time, and has several advantages over
more traditional statistical techniques. First, and perhaps most importantly, HLM is well
suited to the study of couples because it accounts for correlations in the data (between
repeated measures and between data from individuals within a couple). Second, HLM does
not require that all individuals provide data at every interval. Unlike more traditional sta-
tistical methods, HLM accounts for missing data provided that the data are “Missing at
Random.” This feature allowed us to include in the analysis parents who had started but
“dropped out” of the program (n=24; 1 of whom still completed postgroup measures), and
those who had completed the program but from whom we had not received postgroup data
Fam. Proc., Vol. x, xxxx, 2017
CASEY ET AL.
(14/194), or who had provided incomplete data (exact number varies from measure to mea-
sure). This makes the analysis far less susceptible to bias caused by only including data
from “completers,” and is therefore more stringent.
Pre- and Postgroup Means and Standard Deviations for all Measures, for Mothers and Fathers
Pregroup Postgroup Pregroup Postgroup
Mean SD Mean SD Mean SD Mean SD
CORE-OM 12.81 6.27 9.23 6.33 10.66 5.98 8.70 5.63
Parenting Stress Index 89.98 25.08 81.93 26.14 85.70 20.50 83.73 20.81
Quality of Marriage Index 24.03 7.40 27.30 8.90 26.09 7.4 28.70 8.84
Couple Communication Questionnaire
Couple conflict 33.64 13.61 24.07 14.69 32.90 12.21 29.55 15.68
Violent problem solving 3.37 2.28 1.53 1.49 3.13 2.35 1.33 1.65
Conflict about the kids 7.00 3.75 4.69 3.48 5.52 3.22 4.92 3.36
Who Does What?
Father involvement 36.92 13.45 38.41 11.45 44.49 13.91 43.91 13.53
Strengths and Difficulties Questionnaire
Total 14.69 7.10 13.10 7.33 12.72 5.58 11.66 6.36
Strengths and Difficulties Questionnaire
Total 14.79 6.96 11.27 6.16 13.09 6.08 13.08 7.24
Estimated Model Parameters for all Measures, Showing Effect of Parent Gender on the Intercept, Effect of
Time (slope), and the Effect of Parent Gender on Slope
B SEB z B SEB z B SEB z
CORE-OM 2.15 0.76 2.83** 2.37 0.49 4.82*** 1.42 0.86 0.10
Parenting Stress Index 4.29 2.74 1.56 4.43 1.61 2.75** 5.33 2.59 2.06*
Quality of Marriage Index 1.84 0.76 2.44*2.94 0.76 2.85*** 0.29 0.94 0.31
Couple Communication Questionnaire
Couple conflict 0.90 1.64 0.55 6.73 1.63 4.14*** 6.67 2.20 3.02**
Violent problem solving 0.24 0.24 1.01 1.82 0.22 8.36*** 0.04 0.27 0.87
Conflict about the kids 1.55 0.37 4.23*** 1.30 0.35 3.72*** 1.76 0.47 3.76***
Who Does What?
Father involvement 7.66 1.23 6.23*** 0.65 1.00 0.65 1.99 1.77 0.26
Strengths and Difficulties Questionnaire
Total 1.79 0.69 2.60** 1.36 0.61 2.23*0.51 0.92 0.56
Child gender 3Time
Strengths and Difficulties Questionnaire
Total 0.24 1.28 0.18 1.53 0.62 2.46*1.33 1.24 1.07
*p<.05, **p<.01, ***p<.001.
We used a three-level model, with repeated measures (pre–post), individuals, and the
couple unit as the three levels. Individuals’ data were fitted with an intercept, which indi-
cates their initial score on a given measure, and a slope, which indicates their rate of
change over time on that measure. Data were represented at the individual level by ran-
dom effects for the intercept, and by random effects for both the intercept and the slope at
the couple level as recommended by Atkins (2005). For each measure we examine parents’
scores at pre- and postintervention on each measure, and test for differences between the
amount of change reported by mothers and fathers by including an interaction term
between Gender and Time in the model. The statistics relevant to this interaction are
noted in Table 2, but we describe only statistically significant differences between males
and females in the text below. Following these tests, we then examine differences in the
amount of change reported by parents according to their baseline severity on each mea-
sure, comparing trajectories of parents with initially “high” and “low” distress. We chose
to examine the effect of gender and level of impairment separately so as to directly exam-
ine the influence of each independently of the other. Effect sizes are calculated by dividing
the coefficient for the main effect of Time by the square root of the combined between–couple,
between–individual, and residual variance (Taylor, 2014).
Analysis of Missing Data
Missing data analyses were carried out in order to test the assumption that data are
Missing at Random (MAR). Of the eight baseline questionnaires, logistic regression analy-
ses suggested that only CORE-OM scores predicted missingness, (B=0.08, Wald
(1) =2.88, p<.01) such that individuals with higher levels of psychological distress at
baseline were less likely to complete postgroup questionnaires. However, an odds ratio so
close to 1 (exp(B) =.92, 95% CI [0.87–0.97]) suggests that baseline CORE-OM scores may
have little substantive bearing on the odds of postgroup questionnaire completion (all
Estimated Model Parameters for all Measures, Showing Effect of Initial Functioning on the Intercept, Effect of
Time (slope), and the Effect of Initial Functioning on Slope
Initial distress 3Time
BSEB z BSEB z BSEB z
CORE-OM 9.60 0.61 15.62*** 2.12 0.49 4.34*** 5.36 0.82 6.53***
36.19 2.19 16.53*** 4.94 1.62 3.05** 10.85 2.80 3.88***
Quality of Marriage
11.20 0.74 15.18*** 3.05 0.78 3.93*** 6.07 1.20 5.07***
Couple Communication Questionnaire
Couple conflict 20.62 1.34 15.36*** 6.42 1.63 3.95*** 13.56 2.53 5.37***
3.48 0.21 16.64*** 2.05 0.18 11.61*** 2.49 0.30 8.24***
5.86 0.31 18.80*** 1.22 0.32 3.79*** 3.86 0.52 7.36***
Who Does What?
20.93 1.38 15.20*** 0.58 0.98 .55 10.97 1.83 6.00***
Strengths and Difficulties Questionnaire
Total 10.51 0.70 15.00*** 2.13 0.66 3.23*** 4.93 1.08 4.58***
Fam. Proc., Vol. x, xxxx, 2017
CASEY ET AL.
other factors being equal). Of the demographic variables, only relationship status pre-
dicted missingness (X
(3, N=190) =15.9, p<.01), with couples who were separated at
baseline being less likely to complete postgroup questionnaires. However, with only 10
separated couples in the sample, statistical power to detect differences was low. Analysis
with and without individuals with missing follow-up data revealed no differences in the
pattern of results, or in the estimates. Thus, we decided to proceed as if the data were
MAR, but the results reported below should be considered with this in mind.
Pre- to Postintervention Changes
We found significant pre–post improvement in 7 of the 8 measures (all but the Who
Does What?) and a consistent tendency for those at higher levels of stress or distress on
entering the study to show more positive changes than those with relatively little distress.
Parents’ psychological wellbeing
Following the end of group sessions, parents reported a significant decrease in general
psychological distress as measured by the CORE-OM, B=2.37, SE =0.49, z =4.82,
p<.001, 95% CI [3.34 to 1.41], d=.40. In order to determine whether parents in
higher or lower levels of distress improved most, we categorized participants using the rec-
ommended questionnaire cut-off of 10 (out of 40; Connell et al., 2007). Whereas parents in
the nonclinical range reported no change in their level of psychological distress, parents in
the clinical range reported a significant reduction after attending the group sessions
(B=4.80, SE =0.62, z =7.70, p<.001, 95% CI [6.03 to 3.58]). Of the parents
reporting clinical levels of distress at baseline, 50.6% were no longer within this range at
the end of the program.
Parents also reported a significant decrease in the amount of stress experienced associ-
ated with parenting (PSI), B=4.43, SE =1.61, z =2.75, p<.01, 95% CI [7.58 to
1.28], d=.20). Mothers reported a significantly larger reduction in parenting stress
than fathers, B=5.33, SE =2.59, z =2.06, p<.05, 95% CI [0.26–10.40]. Mothers and
fathers experiencing “clinical” (n=84) and “nonclinical” (n=104) levels of stress associ-
ated with parenting were distinguished using the recommended questionnaire cut-off of a
score 90 (Abidin, 1995). Parents’ initial level of parenting stress had a significant effect on
the slope, B=10.85, SE =1.10, z=87.79, p<.001, 95% CI [87.62–91.92]. While parents
in the nonclinical range maintained a stable (low) level of stress associated with parent-
ing, those in the clinical range reported a highly significant reduction in this regard,
B=10.36, SE =2.28, z =4.54, p<.001, 95% CI [14.84 to 5.89]. Of the parents
reporting clinical levels of parenting stress at baseline, 34.4% were no longer within this
range at the end of the program.
Parents’ relationship quality
Overall relationship satisfaction (QMI). HLM analyses indicated that both parents
reported significant improvement in the quality of their relationship with their coparent
after having attended the PasP program, B=2.94, SE =0.76, z=3.85, p<.001, 95% CI
[1.44–4.43], d=.41. Because mothers’ and fathers’ baseline QMI scores were significantly
different (B=1.84, SE =0.76, z =2.44, p<.05), cut-off scores were established sepa-
rately for mothers and fathers (medians of 23.0 and 27.0, respectively). Parents scoring
above this cut-off were categorized as being in a high quality relationship and those below
the cut-off in a “poor” quality couple relationship for the analysis. There was a significant
effect of initial relationship quality on the gradient of the slope (a significant interaction
term), B=6.07, SE =1.20, z=5.07, p<.001, 95% CI [3.72–8.42]. While parents who
began the program in high quality relationships reported no significant change in this
respect over the course of the program, parents in poor quality couple relationships
reported a significant improvement in the relationship quality, B=6.08, SE =1.01,
z=6.02, p<.001, 95% CI [4.10–8.06].
Parents reported a significant reduction in the amount of couple conflict, from before to
after having attended the program, B=6.73, SE =1.63, z =4.14, p<.001, 95% CI
[9.92 to 3.54], d=.52. Mothers and fathers changed differently over the course of the
program as indicated by the significant effect of parent gender on the slope (interaction
effect), B=6.67, SE =2.20, z=3.02, p<.001, 95% CI [2.34–10.99], with mothers
reporting a greater reduction in couple conflict than fathers. Baseline CCQ conflict scores
were statistically equivalent for mothers and fathers, and therefore the overall sample
median was used to establish the cut-off to distinguish individuals’ reporting high and low
couple conflict (a CCQ conflict score of 33.0). Whereas initially “low” conflict parents expe-
rienced a stable (low) level of conflict between the beginning and end of the program
(B=0.36, SE =2.12, z =.17, p=.87, 95% CI [3.79 to 4.50]), initially “high” conflict par-
ents reported a significant reduction in the level of couple conflict, B=13.20, SE =2.00,
z=6.60, p<.001, 95% CI [17.12 to 9.28].
Conflict about the children
Parents’ amount of conflict about matters to do with raising their children also declined
between pre- and postprogram measurements, B=1.30, SE =0.35, z =3.72, p<.001,
95% CI [1.98 to 0.62], d =.38. There was not only a significant effect of parent gender
on the gradient of the slope, B=1.76, SE =0.47, z=3.76, p<.001, 95% CI [0.84–2.68],
but also on the intercept, which indicated that mothers reported higher levels of disagree-
ment about child-rearing than fathers at baseline, B=1.55, SE =0.37, z =4.23,
p<.001, 95% CI [2.27 to 0.83]. The reduction in couple conflict about the children was
statistically significant only for mothers, B=2.18, SE =0.42, z =5.20, p<.001, 95%
CI [3.00 to 1.36]. The analysis also indicated a significantly different effect of the pro-
gram on initially “high” and “low” conflict parents, B=3.86, SE =0.52, z =7.36,
p<.001, 95% CI [4.89 to 2.83]. Whereas low conflict parents sustained this low level,
high conflict parents reported significant reductions in conflict about child-rearing
(B=3.16, SE =0.42, z =7.52, p<.001, 95% CI [3.98 to 2.33]).
Violent problem solving
After attending the program, both parents experienced similar, significant reductions
in “violent problem solving” (B=1.82, SE =0.23, z=8.36, p<.001, 95% CI [2.25 to
1.39], d=.93), also a subscale of the CCQ. There was a differential impact of the pro-
gram on parents’ use of violent problem solving for initially “high” and “low” violent prob-
lem solving parents, B=2.24, SE =0.31, z=7.26, p<.001, 95% CI [1.64–2.85]. Parents
with initially “high” and “low” levels of conflict about their children were distinguished
using median scores for mothers and fathers (8/12 and 6/12, respectively) on this subscale
of the CCQ. Although parents with both “high” and “low” scores on violent problem solving
reported a significant reduction in this respect, the reduction was greater for parents with
initially “high” violent problem solving scores.
On average, both parents reported little change with regard to father involvement after
having attended the 16 group sessions. However, there was a significant difference in the
responses to the program with respect to father involvement (FI) between parents who
described initially “high” and “low” levels of FI (B=10.97, SE =1.83, z=6.00,
Fam. Proc., Vol. x, xxxx, 2017
CASEY ET AL.
p<.001, 95% CI [14.56 to 7.39], with a significant increase in father involvement for
“low” FI parents only (B=6.22, SE =1.52, z=4.09, p<.001, 95% CI [3.24–9.20]).
Children’s psychological wellbeing
Children’s age (grand mean centered) was included as a control variable in the analysis
of SDQ data. Both mothers and fathers reported a reduction in their youngest child’s over-
all emotional and behavioral problems (total SDQ score), B=1.36, SE =0.61, z=2.23,
p<.05, 95% CI [2.54 to 0.17], d=.22. There was no significant effect of child gender
on either the intercept or the slope (B=1.33, SE =1.25, z=1.07, p=.29, 95% CI [1.11
to 3.77]), indicating that (i) there was little difference between the total SDQ scores of boys
and girls prior to their parents attending the program and (ii) both showed a similar
degree of improvement. Children experiencing emotional and behavioral difficulties
beyond that expected in the general population were distinguished using the cut-off values
recommended by Woerner et al. (2004). That is, children’s preintervention level of diffi-
culty was converted into a binary variable by combining children rated “normal” and “bor-
derline” into one group (n=98), and children in the “abnormal” range into another group
(n=45). There was a significant effect of children’s initial level of problems on the amount
of change reported by parents over the course of group sessions (B=4.93, SE =1.08,
z=4.58, p<.001, 95% CI [7.05 to 2.82]). Whereas the behavior of children with nor-
mal or borderline levels of emotional or behavioral difficulties remained stable according
to their parents, children with an initially “high” (abnormal) level of difficulty were
described as showing significant improvement, (B=4.93, SE =1.08, z=4.58,
p<.001, 95% CI [7.05 to 2.82]. Of the parents describing an “abnormal” level of diffi-
culty in their children at baseline, 47.4% now reported normal/borderline difficulties after
attending group sessions.
We examined the implementation of the U.S. Supporting Father Involvement (SFI)
intervention in the U.K., where it is known as Parents as Partners (PasP). This interven-
tion represents one of very few interventions, especially in the U.K., that addresses fam-
ily-wide issues by targeting the couple relationship, and is unique in its integration of
issues in the couple relationship, parenting, and the psychological wellbeing of parents
The results reported here indicated successful implementation of PasP in several
important ways. Both mothers and fathers were attracted to participate in the interven-
tion. They came to group meetings (most attended frequently) and participated in the fol-
low-up assessments. Their responses to the follow-up questionnaires administered about
one month after the groups ended showed positive changes in their psychological wellbe-
ing (global psychological distress and parenting stress), multiple measures of couple rela-
tionship quality (satisfaction, overall conflict, conflict about children, violent problem
solving), father involvement (for those initially less involved), and their children’s prob-
lematic behaviors. This pattern of results repeats those found in the randomized control
trial (RCT) assessing the impact of the Supporting Father Involvement (SFI) intervention
in the U.S., from which PasP has been adapted (Cowan et al., 2009). It was noteworthy
that participants in the U.K. groups showed a statistically significant increase in couple
relationship satisfaction, where previous U.S. trials showed no change in couple relation-
ship satisfaction, compared with control participants whose satisfaction declined over
time. Importantly, in the American study (Cowan et al., 2009), control group parents
reported little benefit on them as individuals, declining satisfaction as partners, and
increases in their children’s problem behaviors. In the absence of a control group in the
current evaluation, this consistency between the positive changes reported by parents
before and after attending PasP groups and those found in the RCT evaluation of the SFI
intervention is encouraging.
The analyses consistently indicated the greatest intervention effects for those in most
need on entering the program, that is, those who had described low levels of functioning
on each measure of individual and family functioning. Importantly, parents already
describing relatively high functioning in a given domain at baseline were able to sustain
that level and showed no signs of deterioration over the 16 weeks. For some, this begs the
question of inviting highly distressed couples after a screening process. Before introducing
such a screening criterion, it would be necessary to first compare the outcomes of a hetero-
geneous group to a group containing only highly distressed couples. Based on the experi-
ences of the group leaders of both the U.S. and U.K. groups, we would speculate that there
is great value in including couples at all points along a continuum of distress in the same
group and that this particular feature of the program is one of the reasons why the highly
distressed couples benefit from the group. While the program may serve to cement the
existing skills of high functioning couples (relatively speaking) and bolster protective fac-
tors in advance of future challenges, these couples may also act as effective models for
lower functioning couples. That said, even couples with initially low levels of violent prob-
lem solving strategies reported a significant reduction following the group sessions. This
is an important finding given the evidence showing that any degree of interparental vio-
lence in the family environment is particularly harmful for children (Wolfe, Crooks, Lee,
McIntyre-Smith, & Jaffe, 2003). With respect to fathers’ engagement in their children’s
care, whereas there was no aggregate level improvement, when analyzed according to
baseline levels of father involvement we found that parents reporting low levels of father
involvement before entering the group reported men’s greater involvement after the
There are a number of limitations inherent in this evaluation. First, it lacked a matched
control group who did not receive any intervention. The funding received for the PasP pro-
gram was earmarked for the successful adaptation, delivery, and implementation of the
U.S. intervention in the U.K., not for a full-scale RCT. Thus, we cannot attribute the posi-
tive pre–post group changes to parents’ program attendance with certainty. However, the
uniformity between results produced by the current evaluation and those produced by the
RCT in the U.S. allows a degree of confidence in the intervention impact. Second, self-
report questionnaires were used to measure each domain of family functioning in this
evaluation and this inevitably excludes the valuable objectivity of observational measures
or questionnaires completed by independent raters such as children’s teachers.
Third, the analysis examining the influence of baseline functioning on intervention
effects is problematic in some ways: (a) we took the decision to look into “high” and “low”
function on each measure individually, measure by measure. This may make clinical
interpretation of the results difficult since parents may be high functioning on one mea-
sure yet low functioning on another; (b) the impact of baseline impairment on program
outcomes is difficult to approximate for some measures because of the way in which “high”
and “low” functioning individuals were identified. For some measures, this threshold was
set according to the sample median scores, which is in essence a type of data manipula-
tion, and; (c) statistically speaking, this analysis is also problematic because splitting the
sample in this way creates extreme groups, and with that comes increased possibility of
regression towards the mean (heightened especially for those measures without a prede-
fined clinical cut-off score). That is, there is always the chance that the results of these
Fam. Proc., Vol. x, xxxx, 2017
CASEY ET AL.
analyses could be an artifact of the tendency of people who obtain extreme scores at either
end of a questionnaire scale to obtain scores that are closer to the average scores at subse-
Finally, these results can only speak to the immediate benefit of having attended the
PasP program because they are based on a comparison of data collected prior to the first
group session and follow-up data collected within one month of program completion.
Although it is not possible at this stage to speak to the program’s ability to bring about
sustained benefits for families, a second round of follow-up data collection is underway,
measuring the same aspects of family wellbeing six months after group completion. Also
underway is an exploration into the range of couples for whom the program can be benefi-
cial. We are conducting pilot trials with adoptive couples, separated couples, and same-sex
couples, but systematic data need to be gathered and analyzed before drawing conclusions
concerning the breadth of application of PasP.
The limitations that we have enumerated are all expectable in the early stages of explor-
ing the efficacy of an intervention developed in one country for the population of another.
What we have learned so far is that it has been possible to mount the Parents as Partner
program successfully in England, based on SFI developed in the U.S., with a great deal of
program fidelity, and with very little change in curriculum from its U.S. predecessor.
The results reported here are promising. After attending Parents as Partners groups
parents reported improvement on almost every index of family functioning assessed,
including parents’ and children’s psychological adjustment, results which are consistent
with the extensive testing of the same intervention model in three randomized control tri-
als in the U.S (Cowan & Cowan, 2000; Cowan et al., 2005, 2009, 2011, 2014; Schulz,
Cowan, & Cowan, 2006). The findings not only provide initial evidence for the successful
implementation of the PasP program in the U.K., but they also constitute support for
approaching the task of promoting parents’, children’s, and family wellbeing by strength-
ening the inter-parental couple relationship as well as the parent–child relationships. This
evaluation of the PasP program adds to mounting evidence for the innoculative effect of
strong couple relationships for families and children.
Abidin, R. R. (1995). Parenting stress index: Professional manual, 3rd ed. Lutz, FL: Psychological Assessment
Abidin, R. R. (1997). Parenting Stress Index: A measure of the parent-child system. In C. Zalaquett & R. Wood
(Eds.), Evaluating stress: A book of resources (pp. 277–291). Lanham, MD: Scarecrow Press.
Amato, P. R. (2014). Does social and economic disadvantage moderate the effects of relationship education on
unwed couples? An analysis of data from the 15-month Building Strong Families evaluation. Family Rela-
Amato, P. R., & Booth, A. (2001). The legacy of parents’ marital discord: Consequences for children’s marital qual-
ity. Journal of Personality and Social Psychology,81(4), 627–638.
Amato, P. R., & DeBoer, D. (2001). The transmission of divorce across generations: Relationship skills or commit-
ment to marriage? Journal of Marriage and the Family,63(4), 1038–1051.
Atkins, D. C. (2005). Using multilevel models to analyze couple and family treatment data: Basic and advanced
issues. Journal of Family Psychology,19(1), 98–110.
Baumrind, D. (1971). Current patterns of parental authority. Developmental Psychology Monographs,4(1, pt2),
Bloom, H. S., & Michalopoulos, C. (2013). When is the story in the subgroups? Prevention Science,14, 179–188.
Connell, J., Barkham, M., Stiles, W. B., Twigg, E., Singleton, N., Evans, O. et al. (2007). Distribution of CORE-
OM scores in a general population, clinical cut-off points and comparison with the CIS-R. The British Journal
of Psychiatry,190(1), 69–74.
Cowan, C. P., & Cowan, P. A. (1990a). Couple Communication Questionnaire. Berkeley, CA: Institute of Human
Development, University of California.
Cowan, C. P., & Cowan, P. A. (1992). When partners become parents: The big life change for couples. New York:
Cowan, C. P., & Cowan, P. A. (2000). When partners become parents: The big life change for couples. Mahwah,
Cowan, C. P., Cowan, P. A., & Barry, J. (2011). Couples’ groups for parents of preschoolers: Ten-year outcomes of
a randomized trial. Journal of Family Psychology,25(2), 240–250.
Cowan, P. A., & Cowan, C. P. (1990b). Who does what? In J. F. Touliatos, B. F. Perlmutter, & M. A. Straus (Eds.),
Handbook of family measurement techniques (pp. 447–448). Thousand Oaks, CA: Sage.
Cowan, P. A., & Cowan, C. P. (2002). Interventions as tests of family systems theories: Marital and family rela-
tionships in children’s development and psychopathology. Development and Psychopathology,14(4), 731–759.
Cowan, P. A., & Cowan, C. P. (2014). Controversies in couple relationship education (CRE): Overlooked evidence
and implications for research and policy. Psychology, Public Policy, and Law,20(4), 361–383.
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: Routledge.
Cowan, P. A., Cowan, C. P., Pruett, M. K., Pruett, K., & Gillette, P. (2014). Evaluating a couples group to enhance
father involvement in low-income families using a benchmark comparison. Family Relations,63(3), 356–370.
Cowan, P. A., Cowan, C. P., Pruett, M. K., Pruett, K., & Wong, J. (2009). Promoting fathers’ engagement with
children: Preventive interventions for low-income families. Journal of Marriage and Family,71(3), 663–679.
Cummings, E. M., Davies, P., & Campbell, S. B. (2000). Developmental psychopathology and family process:
Research, theory, and clinical implications. New York: Guilford Press.
Cummings, E. M., & Davies, P. T. (2002). Effects of marital conflict on children: Recent advances and emerging
themes in process-oriented research. Journal of Child Psychology and Psychiatry,43(1), 31–63.
Davies, P. T., Harold, G. T., Goeke-Morey, M. C., Cummings, E. M., Shelton, K., Rasi, J. A. et al. (2002). Child
emotional security and interparental conflict. Monographs of the Society for Research in Child Development,
Department for Communities and Local Government. (2015). The English Indices of Deprivation. https://www.gov.
Department for Education and Skills. (2007). Every Parent Matters. http://www.direct.gov.uk/everyparentmatters.
Epstein, K., Pruett, M. K., Cowan, P. A., Cowan, C. P., Pradhan, L., Mah, E. et al. (2015). More than one way to
get there: Pathways of change in coparenting conflict after a preventive intervention. Family Process,54(4),
Evans, C., Connell, J., Barkham, M., Margison, F., McGrath, G., Mellor-Clark, J. et al. (2002). Towards a stan-
dardised brief outcome measure: Psychometric properties and utility of the CORE-OM. The British Journal of
Fauchier, A., & Margolin, G. (2004). Affection and conflict in marital and parent-child relationships. Journal of
Marital and Family Therapy,30(2), 197–211.
Feinberg, M. E. (2003). The internal structure and ecological context of coparenting: A framework for research
and intervention. Parenting: Science and Practice 3(2), 95–131.
Fincham, F. D., & Beach, S. R. H. (2010). Marriage in the new millennium: A decade in review. Journal of Mar-
riage and Family,72(3), 630–649.
Foulkes, S. H. (1964). Therapeutic group analysis. London: George Allen & Unwin.
Fournier, J. C., DeRubeis, R. J., Hollon, S. D., Dimidjian, S., Amsterdam, J. D., Shelton, R. C. et al. (2010).
Antidepressant drug effects and depression severity: A patient-level meta-analysis. Journal of the American
Medical Association,303(1), 47–53.
Goodman, R. (1994). A modified version of the Rutter parent questionnaire including extra items on children’s
strengths: A research note. Journal of Child Psychology and Psychiatry,35(8), 1483–1494.
Goodman, R. (1997). The strengths and difficulties questionnaire: A research note. Journal of Child Psychology
and Psychiatry,38(5), 581–586.
Grych, J. H., Harold, G. T., & Miles, C. J. (2003). A prospective investigation of appraisals as mediators of the link
between interparental conflict and child adjustment. Child Development,74(4), 1176–1193.
Halford, W. K., Sanders, M. R., & Behrens, B. C. (2001). Can skills training prevent relationship problems in at-
risk couples? Four-years effects of a behavioural relationship education program. Journal of Family Psychol-
Harold, G. T., Acquah, D., Sellers, R., & Chowdry, H. (2016). What Works to Enhance Inter-Parental Relation-
ships and Improve Outcomes for Children. Early Intervention Foundation: http://www.eif.org.uk/publication/
Harold, G. T., Aitken, J. J., & Shelton, K. H. (2007). Inter-parental conflict and children’s academic attainment:
A longitudinal analysis. Journal of Child Psychology and Psychiatry,48(12), 1223–1232.
Fam. Proc., Vol. x, xxxx, 2017
CASEY ET AL.
Harold, G. T., & Leve, L. (2012). Parents and partners: How the parental relationship affects children’s psycho-
logical development. In A. Balfour, M. Morgan, & C. Vincent (Eds.), How couple relationships shape our world:
Clinical practice, research and policy perspectives. London: Karnac.
Kaczynski, K. J., Lindahl, K. M., Malik, N. M., & Laurenceau, J. P. (2006). Marital conflict, maternal and pater-
nal parenting, and child adjustment: A test of mediation and moderation. Journal of Family Psychology,20(2),
Masarik, A. S., Martin, M. J., Ferrer, E., Lorenz, F. O., Conger, K. J., & Conger, R. D. (2016). Couple resilience to
economic pressure over time and across generations. Journal of Marriage and Family,78(2), 326–345.
McHale, J. P., & Lindahl, K. M. (2011). Coparenting: A conceptual and clinical examination of family systems, 1st
ed. Washington, DC: American Psychological Association.
Norton, R. (1983). Measuring marital quality: A critical look at the dependent variable. Journal of Marriage and
the Family,45(1), 141–151.
Office for National Statistics. (2014). What does the 2011 census tell us about inter-ethnic relationships? http://
Panter-Brick, C., Burgess, A., Eggerman, M., McAllister, F., Pruett, K., & Leckman, J. F. (2014). Practitioner
review: Engaging fathers–recommendations for a game change in parenting interventions based on a system-
atic review of the global evidence. Journal of Child Psychology and Psychiatry,55(11), 1187–1212.
Pruett, K. D., & Pruett, M. K. (2009). Partnership parenting: How men and women parent differently –Why it
helps your kids and can strengthen your marriage. New York: Perseus.
Quirk, K., Strokoff, J., Owen, J. J., France, T., & Bergen, C. (2014). Relationship education in community set-
tings: Effectiveness with distressed and non-distressed low-income racial minority couples. Journal of Marital
and Family Therapy,40(4), 442–453.
Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models: Applications and data analysis methods.
Thousand Oaks, CA: Sage.
Schulz, M. S., Cowan, C. P., & Cowan, P. A. (2006). Promoting healthy beginnings: A randomised control trial of a
preventative intervention to preserve marital quality during the transition to parenthood. Journal of Consult-
ing and Clinical Psychology,74(1), 20–31.
Snijders, T. A. B., & Bosker, R. J. (1999). Multilevel analysis: An introduction to basic and advanced multilevel
modelling. London: Sage.
Stock Whitaker, D. (2001). Using groups to help people, 2nd ed. London: Brunner-Routledge.
Sturge-Apple, M. L., Davies, P. T., & Cummings, E. M. (2006). Impact of hostility and withdrawal in inter-
parental conflict on parental emotional unavailability and children’s adjustment difficulties. Child Develop-
Taylor, A. (2014). Standardised Effect Size in a Mixed/Multilevel Model. http://www.psy.mq.edu.au/psystat/docu
Woerner, W., Becker, A., & Rothenberger, A. (2004). Normative data and scale properties of the German parent
SDQ. European Journal of Child and Adolescent Psychiatry,13(2), 3–10.
Wolfe, D. A., Crooks, C. V., Lee, V., McIntyre-Smith, A., & Jaffe, P. G. (2003). The effects of children’s exposure to
domestic violence: A meta-analysis and critique. Clinical Child and Family Psychology Review,6(3), 171–187.