Full Terms & Conditions of access and use can be found at
Journal of Evidence-Informed Social Work
ISSN: 2376-1407 (Print) 2376-1415 (Online) Journal homepage: http://www.tandfonline.com/loi/webs21
Emotionally Focused Couples Therapy: A
Systematic Review of Its Effectiveness over the
past 19 Years
Candice C. Beasley & Richard Ager
To cite this article: Candice C. Beasley & Richard Ager (2019): Emotionally Focused Couples
Therapy: A Systematic Review of Its Effectiveness over the past 19 Years, Journal of Evidence-
Informed Social Work, DOI: 10.1080/23761407.2018.1563013
To link to this article: https://doi.org/10.1080/23761407.2018.1563013
Published online: 03 Jan 2019.
Submit your article to this journal
Article views: 18
View Crossmark data
Emotionally Focused Couples Therapy: A Systematic Review of
Its Eﬀectiveness over the past 19 Years
Candice C. Beasley and Richard Ager
Tulane University, School of Social Work, New Orleans, LA, USA
Purpose: A meta-analysis is the review of several qualifying studies
where the ﬁndings of each study is analyzed and is then pooled as to
determine if an intervention is eﬀective or not. The aim of this meta-
analysis was to evaluate if the intervention of Emotionally Focused
Couples Therapy (EFCT), also referred to as Emotionally Focused
Therapy (EFT), continues to be an eﬀective intervention, since its
last meta-analysis in 1999, and to determine whether the improve-
ments noted in EFCT continue to be eﬀective, over a period of time,
following the initial intervention.
Method: For the meta-analysis, nine studies which identify as rando-
mized control trials (RCTs), were extracted and utilized from the
original systematic search. These nine studies were used to evaluate
EFCT’s initial pre to post-treatment eﬀectiveness. The portion of the
meta-analysis, which evaluates whether EFCT sustained improvement
at follow-up, consisted of four studies that identify as RCTs.
Results: The results strongly suggest that the intervention of EFCT
not only improved marital satisfaction (Hedge’sgcoeﬃcient = 2.09)
but also, the improvement in marital satisfaction was sustained at
follow up. This sustained improvement was evident through the
results of both the Friedman’s repeated-measures and the post hoc
= 6.500, p= 0.039).
Conclusion: The ﬁndings provide preliminary support that, as it
relates to marital satisfaction, EFCT is an eﬀective treatment, both
in facilitating change during treatment, and in maintaining those
improvements following treatment.
therapy; couples therapy;
controlled trials; outcome
study; emotionally focused
Rationale for study
There is extensive evidence that couple discord not only causes pain for each partner, but
negatively impacts one’s psycho-social, familial, and health well-being resulting in depres-
sion (Denton, Wittenborn, & Golden, 2012), eating disorders (Maier, 2015; Wnuk,
Greenberg, & Dolhanty, 2015), and, of course, relationship dissolution. In a large sample
study, relationship discord was associated with psychological distress, suicidal ideation,
social impairment and employment dysfunction (Whisman & Uebelacker, 2006). It is
critical that eﬀective interventions are available to address the diverse needs of couples.
Based on results from a 2013 survey of expert therapists, couples therapy was predicted to
show more growth in the coming decade than any other approach, including individual,
group and family treatment (Norcross, Pfund, & Prochaska, 2013). Given Lebow’s(2014)
CONTACT Candice C. Beasley CBeasley@tulane.edu Tulane University at New Orleans School of Social Work
JOURNAL OF EVIDENCE-INFORMED SOCIAL WORK
© 2018 Taylor & Francis
ﬁnding that Emotionally Focused Therapy (EFT) represents one of just a few eﬀective
couples approaches, it is well positioned to play a prominent role in addressing relation-
Based on an evaluation of outcome studies, prior to 1999, Johnson, Hunsley,
Greenberg, and Schindler’s(1999)ﬁndings supported the eﬀectiveness of EFT with
couples. More speciﬁcally, a meta-analysis on four randomized control trials, using the
Dyadic Adjustment Scale (Spanier, 1976) as a common measure, yielded a large Weighted
Mean Eﬀect Size (d+) = 1.31, which was statistically signiﬁcant (Z = 6.42; p < .001). Three
of the four studies used in this analysis were conducted by at least one of the two founders
of EFT. However, there was no therapist overlap across these studies. The only informa-
tion Johnson and colleagues provided, about their search strategy, was that they restricted
their studies to randomized control trials (RCTs) on EFT with couples.
There have been two other meta-analyses involving EFT. Wood, Crane, Schaalje, and
Law (2005) conducted a meta-analysis on various behavioral couples interventions,
including EFT, focusing on the eﬀectiveness with diﬀerent severities of marital distress:
mild, moderate and severe. Their ﬁndings suggest that EFT demonstrated superior eﬀec-
tiveness compared to isolate Behavioral Therapy interventions in treating moderate
relationship distress. The other meta-analysis by Dunn and Schwebel (1995) compared
Behavioral Therapy, Cognitive Behavioral Therapy and a mix of what the authors referred
to as “Insight-oriented therapies”, of which EFT was included. Four of the Insight-
oriented Therapies (IOT) studies used EFT and two used what was referred to as “insight-
oriented marital psychotherapy”. The results suggested that IOT was more eﬀective than
both Behavioral Therapy and Cognitive Behavioral Therapy in improving relationships
functioning. However, Cognitive Behavioral Marital Therapy was the only approach
demonstrating improvement in post-treatment relationship-related cognitions. All of the
meta-analyses above-involved studies published in 1999 or earlier, and; therefore, do not
overlap with the current study.
Since 1999, EFT has expanded its application from couples to individuals (MacLeod &
Elliott, 2012), families (Stavrianopoulos, Faller, & Furrow, 2014), adult groups (Compare
& Tasca, 2016) and couples groups (Stavrianopoulos, 2015). The current study focused
exclusively on randomized control trials (RCT) with couples. However, Emotionally
Focused Couples Therapy (EFCT) RCTs have broadened their application to diverse
situations. Studies have continued to address couples with typical relationship diﬃculties
(Wiebe & Johnson, 2016). However, research-practitioners have used RCTs to test its
application with diﬀerent ethnic/cultural populations including Middle Eastern samples
(Najaﬁ, Soleimani, Ahmadi, Javidi, & Hoseini, 2015; Soltani, Shairi, Roshan, & Rahimi,
2014). EFCT RCTs have been employed with couples facing medical diﬃculties such as
infertility (Najaﬁet al., 2015), chronically ill children (Cloutier, Manion, Walker, &
Johnson, 2002), breast cancer (Naaman, 2009) and end-stage cancer (McLean, Walton,
Rodin, Esplen, & Jones, 2011). Other RCTs have focused on psychological challenges such
as wives who are either suﬀering from depression (Dessaulles, Johnson, & Denton, 2003)
or are survivors of childhood trauma (Dalton, Classen, Greenman, & Johnson, 2013).
Lebow (2014) recently reviewed couples therapy outcome studies to report on which
were found to be eﬀective. Two approaches stood out as having a notable collection of
studies showing signiﬁcant results –Cognitive-Behavioral/Behavioral Therapy and EFT.
He further reported that one of the primary challenges of couple treatment is not whether
2C. C. BEASLEY AND R. AGER
the improvement is notable just after treatment; but, whether it is maintained over time.
He cited evidence prior to Johnson et al.’s(1999) meta-analysis, suggesting that EFT
demonstrated maintenance of changeover short follow-ups. An intended contribution of
the current study is to systematically evaluate EFCT/EFT’seﬀectiveness over the past
19 years, to analyze its eﬀectiveness with a more diverse set of populations and problems,
and for the ﬁrst time to systematically evaluate through meta-analysis whether its eﬀec-
tiveness is maintained at follow-up.
Emotionally focused couples therapy: from theory to practice
Emotionally Focused Therapy (EFT) was created by Drs. Sue Johnson and Leslie Greenberg in
the 1980s (Greenberg & Johnson, 1988). It is increasing in popularity in the U.S. and
throughout the world. At its core, EFT oﬀers clinicians a vehicle to reduce conﬂict through
the creation of the safe emotional connection. Emotionally Focused Therapy has been used
with a diverse array of couples, individuals and families who face a variety of challenges. But,
its original focus was on couples, which is the focus of this analysis.
The theoretical foundation of EFT is derived from several sources. It has adopted
a Humanistic Experiential perspective, as described by Rogers (1951) and Perls (1973).
It draws from Systems Theory to understand how each partner’s response dictates the
reaction of the other, and the resulting interchange may evoke a characteristic problematic
‘dance”driven by unbridled reﬂexive feelings (Minuchin & Fishman, 1981). Emotionally
Focused Therapy embraces Gottman’s empirical work on healthy vs. unhealthy relation-
ships which highlights the destructive impact of cycles of interaction infused with criti-
cism, defensiveness, and complaining, among others. For example, Gottman reports that
men and women typically regulate emotions diﬀerently in interpersonal conﬂict. Women
tend to gravitate toward criticizing and complaining (a role which Johnson refers to as the
Pursuer), whereas men tend to pull away and stonewall (a role which Johnson refers to as
the Withdrawer) (Gottman, 1991; Gottman & Levenson, 1986).
Attachment Theory (Bowlby, 1988) lies at the heart of EFT where relationship distress
is best understood in terms of separation anxiety in reaction to an insecure bond. On the
ﬂip side, seeking and maintaining a connection with others is the central motivating factor
in our survival and growth. When the attachment is threatened, partners are compelled to
act out in predictable sequences of interaction beginning with protest and anger, followed
by seeking/clinging, and ﬁnally depression and despair. If the partner does not respond
and connect, detachment and separation results (Bowlby, 1969).
The treatment procedures for assisting couples are organized in several steps that fall within
three stages. Stage One focuses on de-escalating the couple’sconﬂict. Here, partners work to
identify and understand the nature of their negative cycle, which is repetitive and related to
underlying attachment issues. The therapist helps each partner to access underlying emo-
tions, which drives the intensity of his/her arguing. These emotions ultimately relate to
attachment issues, rather than merely the surface issues being debated. The EFT therapist
JOURNAL OF EVIDENCE-INFORMED SOCIAL WORK 3
helps each partner discover the underlying feelings that drive their arguments so both can
come to understand, acknowledge and accept their own as well as their partner’s underlying
feelings. At the end of this stage, couples understand the cycle, how they are unwittingly
involved and controlled by it, the role they and their partner play in the cycle, and the triggers
that set it oﬀ. Rather than blaming the partner for the cycle and demanding s/he change, the
cycle itself becomes the problem to overcome and ultimately change.
Stage Two focuses on restructuring the relationship bond so as to alter the negative
cycle. This begins with clients identifying their own attachment needs (e.g., need for
reassurance and comfort relating to fears of unworthiness). At the same time, partners
learn about and are helped to acknowledge, accept and develop empathy for their partner’s
vulnerabilities and needs. Finally, partners are helped to express their needs and wants. In
eﬀect, once partners can (a) identify the cycle as it emerges, (b) understand and express
their own emotions and needs that fuels their reactive involvement, and (c) empathize
with one’s partner’s needs, they can change the negative cycle over which they previously
had no or little control.
In Stage Three couples develop new healthier cycles of interacting and assume new
positions when dealing with their old problems. Their stories of conﬂict are now reﬂecting
less discourse and demonstrate enhanced capacity to repair. They further expand their
changes to establish new solutions to pragmatic issues. The therapist commends the
couple on their newfound capacity to more openly share their vulnerability, clearly express
their needs, and empathically respond to their partner. Their old cycle, which will begin to
emerge from time to time, is more readily interrupted and replaced with a diﬀerent dance,
one that is sensitive, supportive, and sustains stable attachment. The therapist directs the
couple to solve problems that have plagued them and stirred their unproductive negative
cycle in the past. They ultimately develop a newfound capacity to identify when the
negative cycle begins, understand their own and their partner’s underlying issues fueling
the cycle, and express their needs. This helps them to be less reactive and better able to
manage their negative aﬀect as well as the underlying fears and needs. Instead of expres-
sing negative often critical emotions, vulnerabilities can be shared, which make it easier
for each to listen, understand, and support each other. These new responses are con-
solidated in a “Resiliency Story”which represents the couple’s narrative of their new
bonding rituals (Johnson, 2008).
Search strategy and inclusionary criteria
The authors completed the following steps as to gather qualifying studies for this meta-
analysis. In an attempt to locate the qualifying studies for this meta-analysis, a systematic
search was performed utilizing the following database: Academic Search Complete, CINAHL
Plus, E-Journals, ERIC, Family Studies Abstracts, Health and Psychosocial Instruments;
MEDLINE, PsycARTICLES, PsycINFO, Psychology and Behavioral Sciences Collection,
Social Sciences Full Text, Social Work Abstracts, and SocINDEX with Full Text were
searched from January 01, 1999 through December 31, 2017, screening for both “peer
reviewed”publications and those publications that were labeled as “dissertation”. The key-
words used were: “Emotion Focused Therapy”,“Emotionally Focused Therapy”,“Emotion
4C. C. BEASLEY AND R. AGER
Focused Couples Therapy”,“Emotionally Focused Couples Therapy”,“Emotion Focused
Therapy for Couples”,“Emotionally Focused Therapy for Couples”,“EFT”,“EFCT”,“EFT-
C”; along with “Randomized Control Trial”,“Outcome Study”and published in English.
Dissertations were included to address the common journal bias of only publishing studies
with successful outcomes (Campbell Collaboration, 2014). This was done by conducting two
separate searches, the ﬁrst including just peer-reviewed journal articles, the second including
just dissertations, and both using the same keywords listed above . The search also included
the International Centre for Excellence in Emotionally Focused Therapy site, which lists EFT
research on the following page: http://www.iceeft.com/images/PDFs/EFTResearch.pdf.The
authors did perform independent searches as to ensure that all potential studies were
captured. Finally, the authors examined the references of articles that met the inclusionary
criteria, mentioned below, as well as couples treatment outcome review articles.
The study only considered EFCT/EFT randomized control trials and outcome studies
with couples in which the subjects were administered the Dyadic Adjustment Scale (DAS)
(Spanier, 1976) or a similar relationship functioning instrument. The DAS was utilized as
inclusionary criteria because it is a self-reporting instrument, utilized by couples in
therapy, which measures levels of relationship improvement. Therefore, the data extracted
from the DAS, provides the base of analysis in determining the levels of the intervention’s
eﬀectiveness. Along with this, Randomized Control Trials, as well as Outcome Studies,
were utilized as inclusionary criteria because it is studies of this nature that will include
data which in turn will be used to perform the meta-analysis.
PRISMA Flow diagram
Through the use of the PRISMA Flow Diagram (Moher, Liberati, Tetzlaﬀ, & Altman, 2009),
records identiﬁed, through database searching, yielded 8160 possible articles with 86 articles
being listed as RCT and 816 articles being listed as outcome studies. Because part of the
inclusionary criteria is that a qualifying study (a) must be an outcome study or RCT, (b) must
utilize couples as the subjects of the intervention, and (c) the study used an instrument which
measured relationship functioning. This decreased the original 8160 articles to 902 articles.
Therefore, 902 full-text articles were assessed for eligibility. Of the 902 articles assessed for
eligibility, 894 full-text articles were excluded due to: (a) not meeting RCT criteria, (b) treating
entities other than couples (e.g., individuals, families, or couples groups), (c) study did not use
an instrument which measured relationship functioning, and (d) being identiﬁed as
a duplicate record. Therefore, there were a total of 9 studies, out of 902, included in the
ﬁnal meta-analysis synthesis (n = 9) as only 9 articles met all aspects of the inclusionary
criteria. Of the 9 included studies, only 4 included follow-up data which was utilized in the
intervention sustainment analysis.
Two meta-analyses were conducted. The ﬁrst of the meta-analyses consisted of a pre-post
analysis (i.e., an analysis that exams improvement from the onset of treatment until its
termination). This analysis was used to evaluate if the intervention allowed for the
improvement of marital satisfaction over the course of therapy. The second of the meta-
analyses consisted of a follow-up analysis. This analysis was used to evaluate whether
JOURNAL OF EVIDENCE-INFORMED SOCIAL WORK 5
improvements gained during treatment were maintained at follow-up. Hedge’sgwas
employed for the pre- post-treatment analysis because it utilizes pooled Standard
Deviation scores while correcting for population eﬀect size bias, especially in sample
sizes fewer than 20 (Hedges, 1981). The Hedges g is a measure of eﬀect size, which tells
you how much one group diﬀers from another. In this study, we are examining how much
the improvement of the group receiving EFCT/EFT diﬀered from the comparison group.
The Comprehensive Meta-analysis (CMA) statistical package was used to compute eﬀect
sizes and the overall Hedge’sgscore. The Naaman (2009) study failed to report the
standard deviations, which were required for computing Hedges’g. Lipsey and Wilson
(2001) indicate that a missing standard deviation can be substituted with a standard
deviation from a similar study which uses the same assessment instrument. Therefore,
scores from the Dalton et al.’s(2013) study were used.
The Friedman’s Test (Friedman, 1937) was employed for the follow-up meta-analysis
because it is a non-parametric statistical test used to detect diﬀerences in treatments
across multiple time points that does not require the dependent variable to follow
a normal distribution. As it relates to this study, “multiple attempts”includes Pre-EFT
treatment, Post-EFT treatment, and Follow-up. Friedman’s was calculated by utilizing
IBM SPSS Statistics (v.23) software. The Wilcox Test was included as a post hoc
analysis of the Friedman’s Test, to evaluate whether improvements gained during
treatment were sustained at follow-up. In other words, this test examines whether the
changes achieved at the end of treatment, are sustained for a certain period of time
In short, Hedges gis used to analyze if an intervention is eﬀective when there is a small
sample size (n = 9). Although Hedges g tells that the intervention caused an eﬀect, it does
not tell if the eﬀect was positive or negative. The Friedman’s test is used to analyze if the
intervention has a negative eﬀect (the treatment is ineﬀective) or a positive eﬀect. Finally,
the Wilcox test analyzes if the positive and/or negative eﬀect continues after the conclu-
sion of the intervention or does the intervention plateau at the conclusion of treatment.
Table 1 lists whether the study was included in the pre-post meta-analysis and/or the
Description of studies
Tables 1 and2list descriptive data for the 9 eligible publications included in the two analyses.
For the pre-post meta-analysis, sample sizes of studies were quite small with the mean of
approximately M=14 subjects in the experimental condition and approximately M=13
subjects in the control condition. Subject mean ages ranged across studies from approximately
33 to 56 years. The mean length of relationships ranged from 10 to 29 years; however, all but
one study’s participant relationships ranged from 10 to 14 years (see Tables 1 and 2).
For the follow-up meta-analysis, sample sizes of studies were quite small with the mean
of M=10 subjects in the experimental condition and M= 10 subjects in the control
condition. Subject mean ages ranged across studies from 33 to 37 years while the mean
length of relationships ranged from 11 to 14 years (see Tables 1 and 2).
6C. C. BEASLEY AND R. AGER
Relationship functioning instruments used in the study
One of the criteria, for inclusion, was that the study used an instrument which measured
relationship functioning. This inclusionary criteria was pertinent as to ensure that the
intervention was directed toward marital satisfaction. Almost all studies used the Dyadic
Adjustment Scale (DAS) (Spanier, 1976) or its abbreviated version, the Revised DAS. The
Dyadic Adjustment Scale (DAS) and Revised DAS have been used in research extensively
throughout the world and have been found to be reliable and valid measures (see Busby,
Crane, Larson, & Christensen, 1995; Montesino, Gómez, Femántiez, & Rodríguez, 2013)
in the testing of marital satisfaction. All but two of the studies, reported here, will use one
of these two instruments (see Tables 3 and 4).
Two studies used diﬀerent instruments: the Quality of Marriage Index and the Marital
Conﬂict Questionnaire. Like the DAS and revised DAS, the Quality of Marriage Index
measures marital discord (Norton, 1983). Psychometric analyses support its validity and
reliability (Johnson, White, Edwards, & Booth, 1986, Norton, 1983; Schumm et al., 1986).
Heyman, Sayers, and Bellack (1994) contend that it measures comparable constructs to the
Dyadic Adjustment Scale (DAS) and a score of 28 or less corresponds to a score of 97 or
less on the DAS.
Table 1. Sample sizes and meta-analysis assignment.
Included in Pre
Ahmadi etal., (2014) 15 15 N/A
Cloutier etal. (2002)
Dalton et al. (2013) 12 10 N/A
Denton et al. (2012) 12 12 4 7 Yes Yes
Dessaulles etal. (2003)
9 9 5 5 No Yes
McLean et al. (2011) 22 20 18 18 Yes Yes
Najaﬁet al. (2015) 15 15 N/A
Naaman (2009) 6 6 N/A
Walsh (2002) 15 10 N/A
Note: Meta-analytic results from the RCTs utilized in the Pre-Post analysis provided an overall eﬀect size of 2.09.
Data not reported and unable to be obtained.
Study was only included in the follow-up analysis as the pre-post analysis was already reported in the Johnson et al.
Study was not included in Pre –Post analysis; therefore, Hedges geﬀect size was not calculated.
Table 2. Sample and study characteristics.
Article Age (M)
Relationship Duration (M) Treatment Integrity Adequate
Ahmadi et al. (2014) N/A
Cloutier et al. (2002) 36.90 years 11.30 years Yes
Dalton et al. (2013) 43.00 years 14.00 years Yes
Denton et al. (2012) 32.90 years N/A Yes
Dessaulles et al. (2003) 37.00 years 10.85 years No
McLean et al. (2011) N/A
Najaﬁet al. (2015) 33.80 years 10.00 years No
Naaman (2009) 56.20 years 28.90 years Yes
Walsh (2002) 51.00 years N/A
Averaged Age of both partners together.
Data not reported and unable to be obtained.
JOURNAL OF EVIDENCE-INFORMED SOCIAL WORK 7
The Marital Conﬂict Questionnaire (MCQ) (Barati & Sanai, 1996)alsomeasures
marital functioning. It draws from similar constructs to western marital functioning
instruments. However, it is adapted to be sensitive to the norms of the Iranian
culture, particularly with regard to capturing conﬂict. It has been found to demon-
strate good internal reliability (Keikhayfarzaneh, Shahriari, Ghorbanshiroudi,
Sourizaei, & Keikhayfarzaneh, 2011). Unlike the other three instruments used in
the current study, higher scores on the MCQ indicate lower relationship functioning.
The studies showed a broad range of treatment integrity (TI). Treatment integrity refers to the
extent to which the treatment is accurately and consistently delivered. Beginning with Ahmadi,
Zarei & Fallahchai (2014), investigators described what appeared to be an inadequate level of
TI comprised of brief paragraphs summarizing what would be addressed in each of the nine 90
min treatment sessions. Cloutier et al. (2002) carried out a much more substantial
Table 3. Pre-post meta-analysis results.
Article (See superscript for marital
Ahmadi, Zarei, and Fallahchai (2014)
135.2 (4.86) 134.73 (3.35) 93.33 (8.93) 138.33 (4.29) 6.250
Cloutier et al. (2002)
99.15 (8.55) 101.10 (8.30) 108.38 (12.50) 99.10 (11.80) N/A
Dalton et al. (2013)
95.95 (13.29) 89.05 (16.82) 104.81 (15.15) 88.32 (25.54) 0.736
Denton et al. (2012)
15.90 (7.10) 20.40 (8.10) 36.00 (4.50) 26.20 (10.80) 1.144
Dessaulles et al. (2003)
87.0 (14.9) 81.20 (14.44) 99.9 (17.1) 115.81 (19.02) N/A
McLean et al. (2011)
44.91 (5.90) 43.58 (7.40) 55.29 (4.60) 42.91 (8.60) 1.763
Najaﬁet al. (2015)
21.27 (4.27) 22.17 (4.32) 41.03 (3.59) 22.57 (4.42) 4.461
21.08 (13.29) 4.05 (16.82) 5.87 (15.15) 3.66 (25.54) 0.105
97.40 (12.10) 93.30 (10.2) 92.90 (8.30) 94.10 (8.80) 0.137
Note: Meta-analysis results for RTCs: Hedges g= 2.09.
All scores are an average of both partners. Mean increases in all scores reﬂect relationship improvement except for
Ahmadi et al. (2014) and Naaman (2009).
Marital Conﬂict Questionnaire.
Dyadic Adjustment Scale.
Revised Dyadic Adjustment Scale.
Quality of Marriage Index.
Data not reported and unable to be obtained.
Table 4. Follow-up meta-analysis results.
Article (See superscript
for marital functioning
Cloutier et al. (2002)
99.15 (8.55) 101.10 (8.30) 108.38 (12.50) 99.10 (11.80) 108.31 (13.17) N/A
Denton et al. (2012)
15.90 (7.10) 20.40 (8.10) 36.00 (4.50) 26.20 (10.80) 27.0 (14.20) 23.6 (10.7)
Dessaulles et al. (2003)
87.0 (14.9) 81.20 (14.44) 99.9 (17.1) 115.81 (19.02) 100.70 (19.18) 90.20 (27.80)
McLean et al. (2011)
44.91 (5.90) 43.58 (7.40) 55.29 (4.60) 42.91 (8.60) 55.05 (6.05) 44.36 (10.25)
Note: Meta-analytic results from the RCTs utilized in the Follow-Up analysis provided a Friedman’s result of χ
(3) = 6.500,
p = 0.039; Wilcoxon signed-rank test: Z = −.730, p= .465.
All scores are an average of both partners. Mean increases in all scores reﬂect relationship improvement.
Dyadic Adjustment Scale.
Revised Dyadic Adjustment Scale used.
Quality of Marriage Index used.
Data not reported and unable to be obtained.
8C. C. BEASLEY AND R. AGER
standardization of treatment protocol beginning with a treatment manual. Seven senior-level
clinical psychology students, with a minimum of 1-year EFT, supervised training as well as
specialized training in couples therapy with chronically ill children (a focus of the study),
conducted the treatment. They received 3 h of supervision each week. Raters with similar
training to the therapists usedobjective criteria to assess EFT consistency using portions of the
audio-taped sessions. Only 3% of the session portions were found to include non-EFT therapist
activity and the inter-rater reliability mean Kappa coeﬃcient was .98 (Gordon-Walker,
Johlison, Manion, & Cloutier, 1996).
Dalton et al.’s(2013) study demonstrated a moderate level of TI, primarily because they
employed an EFT treatment manual written by Johnson (Johnson, 2004), one of the two
creators of this treatment approach. Therapists received 5 months of weekly EFT training
which included study of the treatment manual. Denton et al. (2012) similarly used
Johnson’s treatment manual which was the cornerstone of what appeared to be somewhat
better than moderate TI. Therapists received a weeklong externship, conducted by
Johnson, and they sustained an adequate score of at least 40 on the EFT-Therapist
Fidelity Scale, which is an instrument intended to measure TI. They received weekly
supervision by expert EFT supervisors during which video records of selected sessions
were reviewed to enhance integrity.
Based on the description provided in the Dessaulles et al. (2003) article, the TI was
inadequate. The authors only mentioned that therapy was conducted by six doctoral
clinical psychology interns with a minimum of one year of supervised EFT training
supplemented by 10 h of training in EFT with depressed populations (the targeted
subjects). Treatment Integrity described in the McLean et al. (2011) study was much
more substantial. They used an EFT manual adapted to address the issues faced by their
subjects where one partner had advanced metastatic cancer. A quarter of the sessions were
randomly selected for audio-tapings which were reviewed by Johnson to insure TI. All
treatment was delivered by the lead author, Linda McLean.
Based on Najaﬁet al.’s(2015) description of their study, Treatment Integrity was
inadequate. The researchers listed a table with a distilled version of the diﬀerent EFT
steps which therapists presumably followed. Naaman’s(2009) treatment integrity was
more substantial. Oversight of TI was provided by Johnson in consultation sessions.
The clinicians were master’s level psychologists with at least 7 years EFT experience,
and they used a 1996 manual created by Johnson. Walsh’s(2002) study also met adequate
TI standards. The EFT therapists included 3 masters level Marital and Family Therapy
year interns, a 3
year resident, and the author, who was getting a doctorate in
an MFT program. The therapists received 12 h of EFT training in a marital and family
therapy clinic where they studied an EFT text written by Johnson and used a treatment
manual (Denton, 2001). Therapists also received weekly or biweekly supervision from
Wayne Denton, an EFT researcher and director of a marital and family treatment
program. This supervision, which included monitoring tapes and reviewing progress
notes, was used to monitor adherence to the EFT model.
According to Cohen (1988), a Hedges gscore of .20–49, .50-.79, and .80 and greater is
interpreted as ‘small’,‘medium,’and ‘large’eﬀect sizes, respectively. As noted in Table 3,
seven of the nine studies listed provided suﬃcient data as to compute eﬀect sizes. Of those
JOURNAL OF EVIDENCE-INFORMED SOCIAL WORK 9
seven, ﬁve were far above the .20 minimum suggesting an adequate eﬀect size. The two that
were below the minimum were both dissertations that did not yield statistically signiﬁcant
ﬁndings on the marital functioning instrument, which was the Dyadic Adjustment Scale in
A random eﬀects model Hedge’sgcoeﬃcient was conducted to determine the eﬀect size of
the pre-post EFT treatment groups vs. control groups based upon a standardized mean
diﬀerence. As listed in Table 3, the results of the analysis supported the eﬀectiveness of
EFCT with the Hedge’sgcoeﬃcient of 2.09, 95% CI (0.04, 4.14).
For the follow-up analysis, the results of the non-parametric Friedman’s test of diﬀer-
ences among repeated measures suggested a statistically signiﬁcant improvement in
relationship adjustment: χ
(3) = 6.500, p = 0.039. Median marital satisfaction levels for
Pre-EFT, Post-EFT and EFT Follow-up were 65.9, 77.5, and 77.8, respectively. A post hoc
analysis, with Wilcoxon signed-rank tests, were also conducted. There were no signiﬁcant
diﬀerences between Post-EFT and EFT Follow-up trials (Z = −0.730, p = 0.465), suggest-
ing the improvements were maintained at follow-up (See Table 4).
The results of this analysis add to the growing support for the eﬀectiveness of EFT with
couples. Hedges gscores of .20–49, .50-.79, and .80 and greater are interpreted as ‘small’,
‘medium,’and ‘large’eﬀect sizes (Cohen, 1988). Given the score, in this meta-analysis, of
g= 2.09, the statistical support for the eﬀectiveness of EFT for couples appears strong. It
should be noted that this strong ﬁnding was generated despite the low Hedge g scores on
two of the seven studies. These two studies were dissertations which were included to
counter publication bias where only studies with positive ﬁndings tend to be accepted for
publication in juried journals. Not surprisingly, the two dissertation studies also failed to
achieve statistically signiﬁcant results for the marital functioning instrument.
There is tentative support for the eﬀectiveness of EFT to sustain change after treatment
(Z = −0.730, p = 0.465). Review of the mean scores of the four follow-up studies reveals
that all showed improvement over the course of therapy and, compared to pre-treatment,
all demonstrated improvement at follow-up. Three of the four studies showed that
improvements registered at post treatment were completely maintained at follow-up.
However, one study (Denton et al., 2012) showed a notable depression in score following
treatment. About 45% of the improvement gained during therapy was lost. This could be
related to the characteristics of the sample, which was comprised of couples in which the
wife suﬀered from the major depressive disorder. Severe depression may represent
a mediating factor in relationship improvement. Denton et al. (2012) also mentioned
that the relationships in this sample were very unstable with a quarter of them separating
during the study. In addition, they mentioned that there may have been a problem with
the therapist’s adherence to the EFT model. Although there was still marked improvement
in the mean scores, at follow-up compared to pre-treatment, further attention should be
given to the eﬀectiveness of EFT with this population and whether a supplemental
intervention might be considered.
10 C. C. BEASLEY AND R. AGER
Compared to Johnson et al.’s 1999 meta-analysis of EFT for couples, the studies
evaluated here expand the breadth of application. They include outcome evaluations of
Iranian samples (Najaﬁet al., 2015; Soltani et al., 2014), couples facing medical challenges
such as infertility (Najaﬁet al., 2015), chronically ill children (Cloutier et al., 2002), breast
cancer (Naaman, 2009) and end-stage cancer (McLean et al., 2011) and couples struggling
with psychological challenges such as depression (Dessaulles et al., 2003) and surviving
childhood trauma (Dalton et al., 2013). Additionally, this was the ﬁrst evaluation of RCT
studies with follow-ups and the results suggest that improvements are maintained follow-
ing treatment. In short, the ﬁndings add to the support for EFT as an eﬀective intervention
for couples with several diﬀerent characteristics and problems.
Limitations of study
Notable limitations of this evaluation are that (a) some studies failed to adhere to strict
treatment integrity standards, (b) some studies had small samples sizes, (c) there were
a limited number of studies, particularly for the follow-up analysis, (d) three studies were
not juried publications, and (e) the follow-ups varied in length. With regard to treatment
integrity (TI) as described in the Methodology, studies ranged from inadequate to very
good. From the descriptions provided, three studies appeared to have inadequate TI
(Ahmadi, Zarei, & Fallahchai, 2014; Dessaulles et al., 2003; Najaﬁet al., 2015). The
other six studies appeared to demonstrate TI that was at least acceptable. Clearly, the
present study was restricted by the limited number of RCTs available in the literature, and
therefore chose not to use TI as a criterion for selection. Still, problems in the TI of three
of the nine studies weakens the strength of the positive results.
Another limitation was the small sample sizes in the follow-up evaluation. Only four
studies were available and two of them had n’s in the experimental group of just 4
(Denton et al., 2012) and 5 (Dessaulles et al., 2003). This small sample size was addressed
statistically using the conservative nonparametric repeated measures Friedman’s Test
which still yielded a signiﬁcant ﬁnding that the intervention was successful. Also, the
ﬁnding of the post hoc Wilcox Test provided support for the maintenance of improve-
ment at follow-up. A related limitation was that there were a limited number of studies,
particularly for the follow-up analyses.
Another limitation was that two of the studies were not from juried journals. As
discussed earlier, Campbell Collaboration (2014) suggested that in order to address
journal bias in meta-analyses, where only studies with successful outcomes are published,
one should consider including studies from other sources, such as dissertations. Although
not juried, dissertation theses typically are subject to careful review by the committee to
insure an acceptable quality of research and scholarship. It should be noted that neither of
the two dissertations yielded statistically signiﬁcant eﬀect sizes (Naaman, 2009; Walsh,
2002). Despite their inclusion, the meta-analysis still demonstrated robust results.
Aﬁnal limitation is that the follow-ups varied in length. Cloutier et al.’s(2002) study
had the longest follow-up at 2 years, followed by Denton et al. (2012) and Dessaulles
et al.’s(2003) studies, which both had a 6-month follow-up, and the McLean et al.’s(2011)
follow-up, which was the shortest at 3 months. Similar to the limitations in availability of
RCT studies mentioned earlier, there were just 4 RTC studies available for this analysis. It
deserves mention that the maintenance of improvements is arguably the most challenging
JOURNAL OF EVIDENCE-INFORMED SOCIAL WORK 11
and possibly important result in treatment outcome research. This is addressed in the
following statement sometimes attributed to the Vaudevillian comedian, W. C. Fields, “It’s
easy to quit drinking. I know. I’ve done it a thousand times before”. Like drinking,
improving one’s intimate relationship has limited value if all gains are lost after a few
months. Although the number of studies is limited, all but one showed no change between
the post and follow-up scores in marital adjustment. The one that showed some digression
in score, still retained over half of its gains made during treatment. That study had the
smallest sample size of just 4 couples in the follow-up stage of the evaluation. One of the
larger follow-up studies, which had a sample size of 13, demonstrated the maintenance of
improvement at two years, a notable period. Adding these points to the earlier outcomes
reported, the results still suggest the maintenance of change, despite the limited number of
studies evaluated. Nevertheless, further studies with long follow-ups are still needed to
evaluate the extent to which improvements are maintained.
Clearly, EFT is becoming established as an evidence-based practice which needs to
become part of every couples therapist’s intervention repertoire. Social work education
programs, as well as programs from other treatment disciplines, are increasingly
interested in teaching evidence supported models, and including EFT would
strengthen those curricula. Although there has been progress in the breadth of
randomized control trials of EFT with couples, more expansion of the diversity of
target groups is necessary. For example, sorely missing are evaluation studies with
various cultures, ethnicities and racial groups such as African-Americans, about which
there is a paucity of outcome research (Harley & Stansbury, 2011).
Given EFT’s success in couples treatment, it would follow that EFT may have an
application with other systems. For example, preliminary ﬁndings suggest tentative
support for EFT eﬀectiveness in group therapy (Ancha, 2004;Compare&Tasca,2016),
family therapy (Robinson, Dolhanty,Stillar,Henderson,&Mayman,2016;
Stavrianopoulos et al., 2014) and individual therapy (MacLeod & Elliott, 2012).
However, more vigorous evaluation, such as randomized controlled trials with follow-
up evaluation, is necessary for more credible support.
Conﬂict of interest
Authors Candice C. Beasley and Richard Ager declare that there are no conﬂicts of interest as it
relates to this manuscript.
Compliance with ethical standards
This submission is a manuscript entitled: Emotionally Focused Couples Therapy: A Meta-Analysis
of Its Eﬀectiveness Over the Past 19 Years. There was no funding provided for neither the research
obtained in this manuscript, nor for the composition of this manuscript.
12 C. C. BEASLEY AND R. AGER
The authors conﬁrm that the data supporting the ﬁndings of this study are available within the
article and its supplementary materials.
Declaration of conﬂicting interests
The authors declare no potential conﬂicts of interest with respect to the research, authorship, and/or
publication of this article.
Because this study is categorized as a meta-analysis, the manuscript does not contain any studies
with human participants or animals. Therefore, the treatment of subjects being in accordance with
the ethical standards of the NASW and APA is not applicable.
The authors received no ﬁnancial support for the research, authorship, and/or publication of this
Authors Candice C. Beasley and Richard Ager declare that the manuscript entitled: Emotionally
Focused Couples Therapy: A Meta-Analysis of Its Eﬀectiveness Over the Past 19 Years, has not been
published elsewhere and has not been submitted simultaneously for publication elsewhere.
Ahmadi, F.S., Zarei, E. & Fallahchai, S.R. (2014). The eﬀectiveness of Emotionally-Focused Couple
Therapy in resolution of marital conﬂicts between the couples who visited the consultation centers.
Journal of Education and Management Studies,4,118–123.
Ancha, A. J. (2004). Program evaluation of a time-limited, abuse-focused treatment for child and
adolescent sexual abuse victims and their families. Dissertation Abstracts International,64, 5770.
Barati, T., & Sanai, B. (1996). Marital Conﬂict Questionnaire. In B. Sanai, S. Falahati, & A. Houman
(Eds.), Family and marriage scales (pp. 9). Tehran, Iran: Besat Publications.
Bowlby, J. (1969). Attachment and loss. New York, NY: Basic Books.
Bowlby, J. (1988). A secure base. New York, NY: Basic Books.
Busby, D. M., Crane, D. R., Larson, J. H., & Christensen, C. (1995). A revision of the Dyadic
Adjustment Scale for use with distressed and nondistressed couples: Construct hierarchy and
multidimensional scales. Journal of Marital and Family Therapy,21, 289–308. doi:10.1111/j.1752-
Campbell Collaboration. (2014). Campbell systematic reviews: Policies and guidelines. Campbell
Cloutier, P. F., Manion, I. G., Walker, J. G., & Johnson, S. M. (2002). Emotionally focused
interventions for couples with chronically ill children: A 2-year follow-up. Journal of Marital &
Family Therapy,28, 391–398. doi:10.1111/j.1752-0606.2002.tb00364.x
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed. ed.). New York, NY:
JOURNAL OF EVIDENCE-INFORMED SOCIAL WORK 13
Compare, A., & Tasca, G. A. (2016). The rate and shape of change in binge eating episodes and
weight: An eﬀectiveness trial of emotionally focused group therapy for binge-eating disorder.
Clinical Psychology & Psychotherapy,23,24–34. doi:10.1002/cpp.1932
Dalton, E. J., Classen, C. C., Greenman, P. S., & Johnson, S. M. (2013). Nurturing connections in the
aftermath of childhood trauma: A randomized controlled trial of emotionally focused couple
therapy for female survivors of childhood abuse. Couple and Family Psychology: Research and
Practice,2, 209–221. doi:10.1037/a0032772
Denton, W. H. (2001). SHARE study treatment manual. Unpublished manuscript, School of
Medicine, Wake Forest University, Winston-Salem, North Carolina.
Denton,W.H.,Wittenborn,A.K.,&Golden,R.N.(2012). Augmenting antidepressant
medication treatment of depressed women with emotionally focused therapy for couples:
Arandomizedpilotstudy.Journal of Marital and Family Therapy,38,23–38. doi:10.1111/
Dessaulles, A. (1992). The treatment of clinical depression in the context of marital distress.
Dissertation Abstracts International,53, 1605.
Dessaulles, A., Johnson, S. M., & Denton, W. H. (2003). Emotion-focused therapy for couples in the
treatment of depression: A pilot study. The American Journal of Family Therapy,31, 345–353.
Dunn, R. L., & Schwebel, A. I. (1995). Meta-Analytic review of marital therapy outcome research.
Journal of Family Psychology,9,58–68. doi:10.1037/0893-318.104.22.168
Friedman, M. (1937). The use of ranks to avoid the assumption of normality implicit in the analysis
of variance. Journal of the American Statistical Association,32, 675–701. doi:10.1080/
Gordon-Walker, J., Johlison, S., Manion, L., & Cloutier, P. (1996). Emotionally focused marital
interventions for couples with chronically ill children. Journal of Consulting and Clinical
Gottman, J. (1991). Predicting the longitudinal course of marriages. Journal of Marital and Family
Gottman, J., & Levenson, R. W. (1986). Assessing the role of emotion in marriage. Behavioral
Greenberg, L. S., & Johnson, S. M. (1988). Emotionally focused therapy for couples. New York, NY:
Harley,D.A.,&Stansbury,K.L.(2011). Diversity counseling with African Americans. In E. Mpofu (Ed.),
Counseling people of African ancestry (pp. 193–208). New York, NY: Cambridge University Press.
Hedges, L. V. (1981). Fitting categorical models to eﬀect sizes from a series of experiments. Journal
of Educational Statistics,7, 119–137. doi:10.3102/10769986007002119
Heyman, R. E., Sayers, S. L., & Bellack, A. S. (1994). Global marital satisfaction versus marital
adjustment: An empirical comparison of three measures. Journal of Family Psychology,8,
Johnson, D. R., White, L. K., Edwards, J. N., & Booth, A. (1986). Dimensions of marital quality:
Toward methodological and conceptual reﬁnement. Journal of Family Issues,7,31–49.
Johnson, S. M. (2004). The practice of emotionally focused marital therapy: Creating connection.
New York, NY: Brunner-Routledge.
Johnson, S. M. (2008). Hold me tight: Seven conversations for a lifetime of love. New York, NY: Little
Brown & Co.
Johnson, S. M., Hunsley, J., Greenberg, L., & Schindler, D. (1999). Emotionally focused couples
therapy: Status and challenges. Clinical Psychology: Science And Practice,6,67–79.
Keikhayfarzaneh, J. K., Shahriari, N., Ghorbanshiroudi, S., Sourizaei, M., & Keikhayfarzaneh, R.
(2011). The investigation of the eﬀectiveness of the solution-oriented counseling in reduction of
the marital conﬂicts of two-career couples. Journal of Basic Applied Science Research,(2011,
Lebow, J. L. (2014). Couple and Family Therapy: An integrative map of the territory. Washington,
14 C. C. BEASLEY AND R. AGER
Lipsey, M. W., & Wilson, D. B. (2001). Practical meta-analysis. Thousand Oaks, CA: Sage.
MacLeod, R., & Elliott, R. (2012). Emotion-focused therapy for social anxiety: A hermeneutic single
case eﬃcacy design study of a low-outcome case. Counselling Psychology Review,27,7–22.
Maier, C. A. (2015). Feminist-informed emotionally focused couples therapy as treatment for
eating disorders. American Journal of Family Therapy,43, 151–162. doi:10.1080/
McLean, L. M., Walton, T., Rodin, G., Esplen, M., & Jones, J. M. (2011). A couple based intervention for
patients and caregivers facing end-stage cancer: Outcomes of a randomized control trial. Psycho-
Minuchin, S., & Fishman, C. (1981). Family therapy techniques. Cambridge, MA: Harvard
Moher, D., Liberati, A., Tetzlaﬀ, J., & Altman, D. G. (2009). Preferred reporting items for systematic
reviews and meta-analyses: The PRISMA statement. PLoS Medicine,6(7), e1000097. doi:10.1371/
Montesino, M. C., Gómez, J. G., Femántiez, M. P., & Rodríguez, J. A. (2013). Propiedades
psicométricas de la Escala de Ajuste Diddico (DAS) en una muestra comunitaria de parejas.
Psicothema,25, 536–541. doi:10.7334/psicothema2013.85
Naaman,S.C.(2009). Evaluation of the clinical eﬃcacy of Emotionally Focused Therapy on
psychological adjustment and Natural Killer Cell Cytotoxicity in early breast cancer.
Dissertation Abstracts International,69, 5044.
Najaﬁ, M., Soleimani, A. A., Ahmadi, K. H., Javidi, N., & Hoseini, K. E. (2015). Theeﬀectiveness of
Emotionally Focused Therapy on enhancing marital adjustment and quality of life among infertile
couples with marital conﬂicts. International Journal of Fertility and Sterilization,9,238–246.
Norcross, J. C., Pfund, R. A., & Prochaska, J. O. (2013). Psychotherapy in 2022: A Delphi poll on its
future. Professional Psychology: Research and Practice,44, 363–370. doi:10.1037/a0034633
Norton, R. (1983). Measuring marital quality: A critical look at the dependent variable. Journal of
Marriage and the Family,45, 141–151. doi:10.2307/351302
Perls, F. (1973). The gestalt approach and eyewitness to therapy. San Francisco, CA: Science &
Robinson,A.L.,Dolhanty,J.,Stillar,A.,Henderson,K.,&Mayman,S.(2016). Emotion focused
family therapy for eating disorders across the lifespan: A pilot study of a 2 day transdiag-
nostic intervention for parents. Clinical Psychology & Psychotherapy,23,14–23. doi:10.1002/
Rogers, C. (1951). Client-centered therapy. Boston, MA: Houghton-Miﬄin.
Schumm, W. R., Paﬀbergen,L.A.,Hatch,R.C.,Obiorah,F.C.,Copeland,J.M.,&Meens,L.D.
(1986). Concurrent and discriminant validity oftheKansasMaritalSatisfactionScale.Journal
ofMarriageandtheFamily,48, 381–387. doi:10.2307/352405
Soltani, M., Shairi, M. R., Roshan, R., & Rahimi, C. (2014). The impact of emotionally focused
therapy on emotional distress in infertile couples. International Journal of Fertility and
Spanier, G. B. (1976). Measuring dyadic adjustment: New scales for assessing the quality of marriage
and similar dyads. Journal of Marriage & Family,38,15–28. doi:10.2307/350547
Stavrianopoulos, K. (2015). Enhancing relationship satisfaction among college student couples: An
emotionally focused therapy (EFT) approach. Journal of Couple & Relationship Therapy,14,1–16.
Stavrianopoulos, K., Faller, G., & Furrow, J. (2014). Emotionally focused family therapy: Facilitating
change within a family system. Journal of Couple & Relationship Therapy,13,2
Walsh, S. B. (2002, September). Emotion-focused couples therapy as a treatment of somatoform
disorders: An outcome study. Dissertation Abstracts International,63, 1579.
Whisman,M.A.,&Uebelacker,L.A.(2006). Impairment and distress associated with relationship discord
in a national sample of married or cohabiting adults. Journal of Family Psychology,20,369–377.
JOURNAL OF EVIDENCE-INFORMED SOCIAL WORK 15
Wiebe, S. A., & Johnson, S. M. (2016).AreviewoftheresearchinEmotionallyFocused
Therapy for couples. Family Process,55, 390–407. doi:10.1111/famp.12229
Wnuk, S. M., Greenberg, L., & Dolhanty, J. (2015). Emotion-focused group therapy for women with
symptoms of bulimia nervosa. Eating Disorders: the Journal of Treatment & Prevention,23,253–261.
Wood, N. D., Crane, D. R., Schaalje, G. B., & Law, D. D. (2005). What works for whom: A
meta-analytic review of marital and couples therapy in reference to marital distress. The
American Journal of Family Therapy,33, 273–287. doi:10.1080/01926180590962147
16 C. C. BEASLEY AND R. AGER