Emotionally Focused Couples Therapy:

Status and Challenges

Susan M. Johnson, University of Ottawa
John Hunsley, University of Ottawa
Leslie Greenberg, York University
Dwayne Schindler, University of Ottawa

This article presents the basis for, and the research ies (Alexander, Holtzworth-Munroe, & Jameson, 1994;
on, emotionally focused couples therapy (EFT), now Baucom, Shoham, Mueser, Daiuto, & Stickle, 1998):
recognized as one of the most researched and most behavioral marital therapy and emotionally focused cou-
effective approaches to changing distressed muit.1 re- ples therapy (EFT). Of these, EFT is the most recently
lationships. Drawing on attachment theoy and the re- formulated, being first described in the literature in 1985
search on interactional patterns in distressed relatlon- (Johnson & Greenberg, 1985a). This article summarizes
ships, we describe the theoretical context of EFT. We
the development of EFT over the last decade, in terms of
both outcome data and more clinical and theoretical
then outline the nature of the clinical intenentions
issues. We also consider the future challenges to EFT and
used in EFT and the steps hypothesized to be crucial
the field of couples therapy in general.
to couple change. The central role of accessing and
At the time EFT was formulated in the early 1980s,
working with emotional issues in the relationship con-
there were a number of particularly important questions
text is highllghted. Following this presentation, we re- facing the field of couples therapy. First, this modality had
view both the outcome and proc.sr rauNcch on EFT been almost exclusively practice driven. The essential ele-
and present mobanalytic data from randomized clini- ments of marital distress, and therefore the most appro-
cal trials to substantiate the clinical impact of EFT on priate targets for intervention, were s t i l l undelineated by
couple adjustment. Finally, the empirical and clinical empirical study. Second, there was a dearth of nonbehav-
challenges facing EFT are summarized. ioral, more dynamically oriented interventions that had
Key wods: emotionally focused couples therapy, been clearly described and tested. There was no clear
marital relationships. fain Prychd Sci h e 6~67-79, technology for relationship change outside the scope of
19993
the behavioral interventions (Gurman, 1978). There were
also concerns about the general efficacy of couples inter-
ventions and an acknowledged need to continue to
Although the failure to develop a satisfying intimate rela-
develop such interventions (Jacobson, 1978; Jacobson,
tionship with one’s partner is the single most frequently
Follette, & Elwood, 1984). Third, there was a lack of a
presented problem in therapy (Horowitz, 1979), couples
consistent, empirically supported theoretical perspective
therapy, the modality that most directly addresses this
on the nature of adult love and relationships that could be
problem, is a relatively young discipline. In this discipline,
used to clarifjr the goals and focus the process of therapy.
systematic approaches to changing distressed relationships
Fourth, couples interventions had focused on changing
are s t i l l being developed and evaluated. At present there
behavior and, in a limited way, on restructuring cogni-
are only two clearly delineated treatments for marital dis-
tions, such as the attributions partners make about each
tress that have been empirically tested in a number of stud-
other’s behavior. However, the role of affect had not been
Address correspondence to Susan M. Johnson, School of Psy- systematically addressed, although even behavioral ap-
chology, University of Ottawa, Ottawa, Ontario, Canada K1N proaches acknowledged that modifying affect was a neces-
6N5. sary part of treating distressed relationships (Jacobson &

0 1999 AMERICAN PSYCHOLOGICAL ASSOCIATION D l 2 67

 Margolin, 1979). EFT developed in response to these EFT has attempted to address the basic questions ofall psy-
issues and, as we describe below, reflects these key con- chotherapy research. The first question concerns efficacy;
cerns of the couples therapy field. that is, does it work? The second question concerns the
EFT is a brief systematic approach to modifying dis- process of change, or, how does it work? The third ques-
tressed couples’ constricted interaction patterns and emo- tion concerns the matching of client to treatment, or, for
tional responses and to fostering the development of a whom does it work? The research base supporting EFT is
secure emotional bond (Greenberg & Johnson, 1988; summarized in this article and a meta-analysis of the effects
Johnson, 1996). The specific targets of the EFT change ofEFT on marital functioning is presented.
process are the same variables identified in recent empiri-
cal research as the crucial elements in marital distress EFT: THE THEORETICAL M O D E L
(Gottman, 1979, 1993; Gottman, Coan, Carrere, & The Nature of Marital Distress
Swanson, 1998). Specifically, EFT targets absorbing states EFT assumes that the key factors in marital distress are the
of negative affect, that is, negative emotions such ongoing construction of absorbing states of distressed
as anger or fear that are difficult to quickly diminish. affect and the constrained, destructive interactional pat-
In the context of intimate relationships, interactional terns that arise from, reflect, and then in turn prime this
cues associated with these affective states tend to override affect. EFT combines an experiential, intrapsychic focus
other cues and become self-reinforcing. EFT also targets on inner experience, particularly affect, with a systemic
rigid self-reinforcing interaction patterns such as critical focus on cyclical, self-reinforcing interactional responses.
pursuit, followed by distance and defensiveness. EFT The focus on affect arises from the humanistic experiential
integrates the intrapsychic perspective afforded by perspective, as outlined by Rogers (1951) and Perls
psychodynamic approaches with an interpersonal sys- (1 973), and reflects the individual therapy training of the
temic perspective, and melds these perspectives into a originators of EFT (Greenberg &Johnson, 1988). The
technology for change that is formulated in a 9-step focus on how each partner’s responses constrain and dic-
change process. In this process, newly formulated emo- tate the other’s, and on interpersonal patterns, reflects the
tional responses are expressed in such a way as to create influence of systems theory, as exemplified by the work
specific shhs in interaction that prime bonding events. ofMinuchin (Minuchin & Fishman, 1981). The focus on
These events then create new constructive cycles of con- affect is supported by the empirical work of Gottman
tact and caring between partners. (1991), who emphasizes the power of negative affect, as
EFT views relationships from an attachment perspec- expressed in facial expression, to predict long-term stabil-
tive. This perspective has been recently identified as the ity and satisfaction in relationships and the destructive
most promising theory of adult love to date and already impact of repeated cycles of interaction, such as criticizing
has substantial empirical support (Bartholomew & Per- and defendmg oneself or complaining and stonewalling.
Iman, 1994; Collins & Read, 1990; Kobak & Hazan, The inability of distressed couples to sustain emotional
1991; Simpson, Rholes, & Nelligan, 1992), providing a engagement is also noted by Gottman (Gottman & Lev-
potential map of intimate relationships for couples thera- enson, 1986) and appears more central in maintaining dis-
pists. Attachment theory helps the therapist understand tress than disagreements per se or whether disagreements
partners’ needs and how particular responses to these can be resolved. Gottman (1991) notes that there appear
needs define close relationships. Lastly, in terms of the to be differences in affect regulation between men and
concerns identified above, EFT addresses the role ofaffect women. Women seem to be more able to regulate their
in close relationships and in changing those relationships, affect in interpersonal conflict and therefore more often
both on a theoretical level and on the level of clinical seem to take a critical and complaining position, whereas
intervention (Johnson & Greenberg, 1994). It is then part their male partners withdraw and stonewall to contain
of the recent zeitgeist that focuses on the facilitative role their affect. The cycle of critical complaint followed by
of emotion in human fbnctioning and therapeutic change. defense and distance is particularly destructive for couples
EFT also reflects the general context ofthe field ofpsy- relationships (Heavey, Christensen, & Malamuth, 1995).
chotherapy in the 1990sin that it is a brief, systematic inter- Gottman’s thorough and empirically based description of
vention that has been empirically validated. Research on marital distress and his model’s ability to predict marital

CLINICAL PSYCHOLOGY: SCIENCE AND PRACTICE V 6 N1, SPRING 1999 68

outcomes suggest that emotional responses and particular anger, but his flight in fact heightens her anxiety and
self-reinforcing interaction patterns are the most appro- primes her aggression toward him.
priate targets of intervention in marital therapy (Gott- Attachment theory provides a map for adult intimate
man, 1994). relationships. It outlines adaptive needs for contact, com-
fort, security, and closeness as the features that define this
The Nature of Adult Love landscape. This perspective focuses the couples therapist
To understand why and how emotional responses and the on attachment fears, longings, and needs, and stresses the
interactional patterns outlined above are so central to mar- significance of experiences ofloss of trust and connection.
ital distress, we need to place these empirical findings in It directs the process of therapy toward the creation of the
the context of a theory of relationships. Marital therapy accessibility and responsiveness that foster safe emotional
has, in general, lacked a clear theory of adult intimacy and engagement. In terms of the process of change, attach-
therefore a clear sense of the primary goals and targets for ment theory directs the therapist’s attention to the
the change process (Roberts, 1992; Segraves, 1990). Such accessing and reprocessing of attachment-related affect,
a theory would allow clinicians to define what specific the modifjnng of interactions that block contact, and the
changes are necessary to encourage recovery from distress creation of bonding interactions. This theory, like Gott-
and promote long-term health and resilience in relation- man’s research, stresses the importance of affect in the
ships. definition of close relationships. In attachment theory,
In recent years attachment theory has been applied emotion may be seen as alerting partners to the signifi-
more and more to adult relationships rather than to par- cance and nature of key relational experiences, evoking
ent-child bonds (Bowlby, 1988; Hazan & Shaver, 1987). working models in a state-dependent fashion (e.g., “when
From an attachment perspective, the description of mari- I’m anxious about my relationship, I experience all my
tal distress outlined above is best understood in terms of fears about myself’), and, most importantly for the cou-
separation distress and an insecure bond. A bond here ples therapist, priming attachment behaviors (Johnson,
refers to an emotional tie, a set of attachment behaviors to 1996).
create and manage proximity to the attachment figure and There are four key assumptions of EFT that arise out
a set of working models or what are usually termed sche- of these theoretical perspectives. First, emotional re-
mas. These schemas are concerned with the dependability sponses and interactional patterns are reciprocally deter-
of others and the worth or lovableness of self. mining and both must be addressed in therapy. Second,
Seeking and maintaining contact with others is viewed partners are stuck in negative patterns that preclude the
as the primary motivating principle in human beings and responsiveness necessary for secure bonding. They are not
as an innate survival mechanism shaped by the process of viewed as immature or unskilled but, rather, as needing
evolution. Secure attachment provides a safe haven and a support to formulate their attachment needs and fears in a
secure base in a potentially dangerous world (Bowlby, manner that promotes secure bonding. Third, emotion is
1988). When attachment security is threatened, compel- seen as a key element in the definition and the redefinition
ling affect organizes attachment responses into predictable of close relationships. New emotional experience and
sequences. Typically protest and anger will be the first new interactions are necessary for change to occur.
response to such a threat, followed by some form of cling- Fourth, adult intimacy is best viewed as an attachment
ing and seeking, which then gives way to depression and process. This process gives couples interventions a specific
despair. Finally, if an attachment figure does not respond, focus, target, and set of goals.
detachment and separation will occur (Bowlby, 1969).
The potential loss of an attachment figure, or an ongoing EFT Interventions
inability to define the relationship as generally secure, is The process of change in EFT has been delineated in nine
significant enough to prime automatic fight, flight, or treatment steps. The first four steps involve assessment and
fieeze responses that limit information processing and the de-escalation of problematic interactional cycles. The
constrict interactional responses (Johnson, 1996). So, for middle three steps emphasize the creation of specific
example, a husband evades and avoids his wife in an change events where interactional positions shift and new
attempt to calm down the interaction and reduce her bonding experiences occur. The last two steps of therapy

EFT STATUS JOHNSON ET AL. 69

address the consolidation of change and the integration of sense of helplessness by placing it within the context of
these changes into the everyday life of the couple. The the destructive cycle that has taken over the relationship.
therapist leads the couple through these steps in spiral The therapist will heighten this experience in the session
fashion, as one step incorporates and leads into another. In and then help his partner to hear and accept it, even
mildly distressed couples, partners generally work quickly though it is very different from the way she usually experi-
through the steps at a parallel rate. In more distressed, ences her spouse. Finally, the therapist moves to structur-
more insecure couples, the more passive or withdrawn ing an interaction around this helplessness, as in, “So can
partner is usually invited to go through the steps slightly you turn to her and can you tell her, ‘I feel so helplessness
ahead of the other. The increased emotional engagement and defeated. I just want to run away and hide.”’ This kind
of hs partner then helps the other more active, critical of statement, in and of itself, represents a move away fiom
partner shift to a more trusting stance. passive withdrawal and is the beginning of active emo-
The nine steps of EFT are as follows: tional engagement. The steps of EFT are described in
greater detail elsewhere, as are the specific interventions
Cycle De-escalation. associated with each step (Johnson, 1996). There are also
Step 1. Assessment-creating an alliance and explicat- a number of therapy transcripts in the literature (Johnson,
ing the core issues in the couple’s conflict using an attach- 1996, in press; Johnson & Greenberg, 1992, 1995) that
ment perspective. illustrate the stance of the therapist and the types of inter-
Step 2. Identifjmg the problem interactional cycle that ventions employed in EFT.
maintains attachment insecurity and relationship distress.
Step 3. Accessing the unacknowledged emotions ASSESSING THE CLINICAL EFFICACY O F EFT
underlying interactional positions. The central focus of the empirical work on EFT has been
Step 4. Refiaming the problem in terms of the cycle, to determine its efficacy as a treatment for marital distress.
the underlying emotions, and attachment needs. To date, seven studies have examined the impact of EFT
on distressed couples, as assessed by a wide range of mea-
Changing Interactional Positions. sures, including indices of psychological and dyadlc
Step 5. Promoting identification with dlsowned needs adjustment, intimacy, and target complaints about the
and aspects of self and integrating these into relationshp relationshp. The majority of these studies have been ran-
interactions. domized clinical trials (RCTs) in which EFT was com-
Step 6. Promoting acceptance of the partner’s new pared to pharmacological or psychological treatments, or
construction of experience in the relationship and new to waiting list controls (Dessaulles, 1991; Goldman &
responses. Greenberg, 1992; James, 1991; Johnson & Greenberg,
Step 7. Facilitating the expression of specific needs and 1985a; Walker, Johnson, Manion, & Cloutier, 1996); in
wants and creating emotional engagement. two studies, treated couples served as their own controls
(Johnson & Greenberg, 1985b; Johnson & Talitman,
Consolidation/Zntegration. 1997). Additionally, to examine the extent to which EFT
Step 8. Facilitating the emergence of new solutions to may affect relationship issues other than marital distress,
old problematic relationship issues. two RCTs have explored the ability of EFT to enhance
Step 9. Consolidating new positions and new cycles of intimacy in maritally adjusted couples (Dandeneau &
attachment behavior. Johnson, 1994) and to modify low sexual desire in female
In all of these steps the therapist moves between (a) partners (MacPhee, Johnson, & Van der Veer, 1995).
helping partners crystalhe their emotional experience in Table 1 presents summary information on the characteris-
the present, trachng, reflecting, and then expanding this tics of the couples who participated in these studies and
experience and (b)setting interactional tasks that add new the design characteristics of these studies. As evident from
elements to and reorganize the interactional cycle. The Table 1, all EFT trials have included treatment integrity
therapist might, then, first help a withdrawn, guarded checks that were performed on tapes of therapy sessions
spouse formulate his sense of paralyzed helplessness that and have had very low attrition rates, thus enhancing the
primes his withdrawal. The therapist will validate this internal vahdity of the studles. Across the nine studies of

CLINICAL PSYCHOLOGY: SCIENCE AND PRACTICE V6 N1, SPRING 1999 70

Table 1. Sample and study characteristics in EFT studies

Relationship n of Treatment
n of Age Duration Children nof n of Integrity
Study Couples (M years) OM years) (MI Therapists Sessions Check Attrition

Johnson & Creenberg (1985a) 45 34.0 8.6 1.8 12 8 Yes 0%

Johnson & Creenberg (1985b) 14 33.0 6.9 1.o 7 10 Yes 0%
James (1991) 28 37.0 9.6 1.6 14 12 Yes Unknown
Coldman & Creenberg (1992) 28 38.5 11.3 1.4 7 10 Yes 0%
Dandeneau &Johnson (1994) 36 40.9 15.7 1.6 10 9 Yes 3%"
MacPhee et al. (1995) 49 41.5 14.0 1.4 10 10 Yes 8%
Walker et al. (1996) 32 36.9 11.3 2.3 7 10 Yes 3%
Dessaulles (1991) 12 37.0 10.9 2.0 6 15 Yes 33 %b
Johnson & Talitman (1997) 34 42.0 13.0 1.4 13 12 Yes 5%

'One couple withdrew from the non-EFT treatment condition.

bTwocouples withdrew from the EFT condition and four couples withdrew from the pharmacologicalintervention condition.

Table 2. Effects of EFT on marital adjustment

DAS

Prethcrapy Posttherapy

study M SO M SD Recovered Improved Deterioration

Johnson & Creenberg (1985a) 92.8 8.8 112.7 10.8 46% 66% 0%
Johnson & Creenberg (1985b) 93.9 n/a 103.9 n/a nfa n/a n/a
James (1991) 87.6 10.2 103.3 14.6 79% 86% 14%
Coldman & Creenberg (1992) 86.3 8.3 100.1 13.1 67 % 71% 0%
Dandeneau &Johnson (1994)" 105.9 6.7 110.5 4.9
MacPhee et al. (1993). 98.6 nfa 105.1 n/a
Walker et al. (1996) 99.7 8.3 109.6 9.2 38% 69 % 0%
Dessaulles (1991)b 87.0 14.9 99.9 17.1
Johnson & Talitrnan (1997) 88.0 7.9 102.8 13.3 50% 79% 6%

Note: nfa means not available.

'The focus of these studies were, respectively, to enhance aspects of the relationshipsof maritally adjusted couples and to improve couple's sexual
functioning. Calculations of recovery, improvement, and deterioration based on DAS scores are therefore not relevant.
bAsonly six couples received EFT in this study, estimates of clinical improvement are not provided.

EFT outcome, therapy has been provided by both novice most commonly used measure of dyadic adjustment in the
and experienced therapists. As almost all of the studies literature. Table 2 presents information on couples' DAS
have been conducted by the two originators of EFT (S. scores prior .to and following treatment. In all studies in
Johnson and L. Greenberg), it is important to note that which the primary focus of treatment was marital distress
there has been only minimal therapist overlap across stud- (i,e., excluding the Dandeneau &Johnson and the Mac-
ies (three therapists were common to the Walker et d . and Phee et al. studies), EFT has been found to result in sig-
the Johnson and Talitman studies), thus enhancing the nificantly improved dyadic adjustment, compared both to
external validity of this program of research. waiting list controls and to couples' pretreatment DAS
As the main god of EFT is to alleviate couples' rela- scores. Using the criteria suggested by Jacobson and Truax
tionshp &stress, we concentrate most of our presentation (1991) for assessing clinically significant change, the over-
of the effects of EFT on the results of treatment on cou- whelming majority of EFT-treated couples reported clin-
ples' Dyadic Adjustment scale (DAS; Spanier, 1976), the ical improvement on the DAS, and in most studies over

EFT STATUS * JOHNSON ET AL. 71

 half of the EFT-treated couples met criteria for recovery
Table 3. Meta-analysisof EFT effects on marital adjustment
(i.e., no longer maritally distressed). Additionally, there
appear to be only infrequent instances in which EFT- Reported
Study Effect Size N (EFT) N (Control)
treated couples experienced deterioration in their rela-
tionship over the course of treatment. Overall, these Randomized clinical trials including a no-treatment condition
Johnson & Greenberg (1985a) 2.19 15 15
results generally meet or exceed the oft-reported finding James (1991) .70 14 14
that approximately half of couples seen in marital therapy Goldman & Greenberg (1992) 1.52 15 15
Walker et al. (1996) 1.27 16 16
outcome research are no longer maritally &stressed (Hah-
Other EFT studies with distressed couples
lweg & Markman, 1988;Jacobson & Addis, 1993). Johnson & Greenberg (1985b) .94 14
There have been a number of reviews of the couples Dessaulles (1991) 1.49 7
Johnson & Talitman (1997) 1.26 34
therapy outcome literature in recent years, including both
traditional literature reviews (Lebow & Gurman, 1995; Note: Meta-analytic results from RCTs were as follows: Weighted mean
effect size: (d+) = 1.31. Significance of combined result: 2 = 6.42, p <
Piercy & Sprenkle, 1990) and meta-analytic reviews ,001, Homogeneity of effect sizes: Q = 6.05, df = 3, p > .05. Fail-safe
(Dunn & Schwebel, 1995; Shadish et al., 1993). Because n = 49 studies.

of the recency of publication for some of the articles

describing the effects of EFT, these reviews do not pro- it provides limited information about the effect of EFT.
vide a comprehensive overview of the eficacy of EFT. In the couples therapy literature, different researchers,
Accordingly, we conducted a meta-analysis on the out- examining various forms of couples therapy, use differing
come measures (dyadic adjustment, intimacy, target com- batteries of measures. Meta-analytic estimates that com-
plaints) reported in the four RCTs of EFT in which bine these different measures may mistakenly give the
couples were seeking treatment for their relationship dis- impression that one can precisely compare studies or ther-
tress (i.e., Goldman & Greenberg, 1992; James, 1991; apeutic approaches. Instead, to make accurate cross-study
Johnson & Greenberg, 1985a; Walker et al., 1996). In comparisons, one must make comparisons based solely on
conducting this meta-analysis reported below, we in- measures that the studies have in common. In the couples
cluded only the EFT studies ofmaritally &stressed couples therapy literature, the only such measure is the DAS.
that involved RCTs, as there is evidence that effect Therefore, to provide more precise meta-analytic esti-
sizes derived &om other types of research designs may mates of EFTS impact for future comparisons with other
underestimate the true effects of treatment (Shadish & treatment approaches, we repeated the meta-analysis
Ragsdale, 1996). Additionally, given that the Dessaulles using only the DAS data from the four RCTs.
(1991) study had a very small sample and did not include Table 3 presents the effect sizes reported in these stud-
a waiting list control group, we excluded this study from ies for couples’ DAS scores following treatment. For com-
our analysis. Using Schwarzer’s (1990) meta-analysis pro- parison purposes, effect sizes from the three other studies
gram, we calculated several meta-analytic indices for the with maritally &stressed couples are included. As shown
four R C T studies of EFT. First, effect sizes were calcu- in Table 3, the weighted mean effect size ofthe four RCTs
lated for each dependent variable in each study. Average attained a statistically significant value of 1.31. Not sur-
effect sizes were then calculated for each study. Finally, the prisingly, given that all couples in these studies were mari-
mean effect size, weighted by sample size, was calculated tally distressed, the effect sizes across studies were
across studies. The resulting overall mean effect size was homogeneous. An effect size of 1.31 is very large for psy-
1.28. This is a large effect size for psychotherapy research chotherapy research, especially in light of Dunn and
and is statistically significant, Z = 6.32, p < .001; the Schwebel’s (1995) estimate of .90 for the average effect
effect sizes across studies were homogeneous, Q = 5.34, size of couples therapy on global inQces (including the
p > .05. For comparison purposes, we note that previous DAS) of marital quality.
meta-analytic estimates for the effect size of couples ther- Our results are based on the data derived from only
apy have ranged from .60 (ShaQsh et al., 1993) to .90 four studies. It is possible, therefore, that they may not
(Dunn & Schwebel, 1995). be stable, One method for examining this possibility is to
Although the effect size obtained from the preceding calculate Rosenthal’s (1984) hil-safe 11. This calculation
meta-analysis is useful in making general comparisons to yields an estimate of the number of studies reporting non-
previous meta-analyses of the couples therapy literature, significant findings that would be required to reduce the

CLINICAL PSYCWOLOGV: SCIENCE AND PRACTICE V6 N1, SPRING 1999 72

overall effect size estimate of the body of research to a rable to others reported by researchers addressing low sex-
nonsignificant level. The fail-safe n for the EFT findings ual desire (Schover & Lopiccolo, 1982) in that subjects
on the DAS is 49; that is, 49 studies reporting nonsignifi- did improve across time on some measures but significant
cant results would be required to yield an overall effect positive changes were less than optimal.
size that was statistically nonsignificant. Rosenthal (1984)
has suggested that a tolerance level be set to determine THE PROCESS O F CHANGE IN EFT
what would constitute an unlikely number of nonsigni- In general, there has been very little research addressing
ficant studies not included in a meta-analysis. Using his the process of change in couples therapy. To date, four
tolerance level equation (Rosenthal, 1984, p. 1 lo), this studies of the process of change in EFT have been
number is 30 studies. Therefore, it is extremely unlikely reported in the literature, with the goal of this line of
that the obtained results for EFT on the DAS are inflated research being a greater understandmg of the crucial
due to nonpublication of nonsignificant findings. Accord- ingredients of change, from the point of view of both cli-
ingly, based on our meta-analytic estimates for homoge- ent performance and therapist interventions. The first
neity of effect sizes and the fail-safe n, it appears that the study (Johnson & Greenberg, 1988) examined the process
obtained effect size of 1.31 is stable and consistent across of therapy in “best” sessions for three couples whose DAS
studies. scores increased by an average of 47 points (i.e., approxi-
Not only is EFT clearly effective in reducing marital mately 2.5 standard deviations) in the original EFT
distress, there also seems to be a tendency for couples to outcome study (Johnson & Greenberg, 1985a). T h s
continue to improve after the termination of treatment. successful process was compared with that of the three
For example, in the most recent study (Johnson & Talit- lowest change couples who did not show significant
man, 1997), 70% of couples were found to be recovered improvement on the DAS. Videotapes of best sessions
at 3-month follow-up, an improvement over the 50% (chosen by the couples) were independently rated for lev-
who were recovered at the end of therapy. The same kind els of experiencing (Klein, Mathieu, Gendlin, & Miesler,
of increase occurred in the first EFT study by Johnson and 1969) and for afiliative and autonomous responses in
Greenberg (1985a; 46% recovered at termination, 73% at interactions using the Structural Analysis of Social Behav-
follow-up) and in the Walker et al. (1996) study (38% ior (SASB; Benjamin, 1986). A particular change event, a
recovered at termination, 70% at follow-up). A 2-year “softening,” where a previously critical partner expresses
follow-up of the couples involved in the Walker et al. vulnerability and asks for comfort and connection from
(1996) study has also been completed, with very positive his or her partner, was also defined using these measures.
results (Walker & Manion, 1998), as all treatment effects The high-change couples showed significantlyhigher lev-
were maintained at the follow-up assessment. These els of experiencing in best sessions. A x2 analysis also
results are particularly encouraging given that these cou- found that blaming partners in the high-change couples
ples were the parents of chronically ill children and so were more likely to move to demonstrating a more affil-
coping with considerable ongoing stress that might be iative and less coercive position toward their spouse in the
expected to make treatment gains more difficult to session. O n average, five softening change events were
maintain. found in the sessions of the successful couples and none
Although EFT is primarily designed to alter marital were found in the sessions of the low-change couples.
hnctioning, there are initial data indicating that EFT These results confirmed the relevance of encouraging
interventions also reduce depressive symptoms (Dessaul- couples to explore their emotional responses and engage
les, 1991; MacPhee et al., 1995; Walker, 1994). However, in tasks in which they express their attachment needs to
10 sessions of EFT failed to significantly increase the sex- their partner in a manner that facilitates emotional
ual adjustment of couples in which the female partner was engagement.
experiencing low sexual desire (MacPhee et al., 1995). Three additional small studies ofEFT change processes
Clinically, we found that many of these low-desire part- (Greenberg, Ford, Alden, &Johnson, 1993) demonstrated
ners had been traumatized in past close relationships and the same kinds of effects, In one study EFT couples were
10 sessions was not an adequate number to create new found to demonstrate more shifts from hostility to affilia-
positive interactions, particularly in the area of intimate tion on the SASB than wait-list couples who were begin-
touch and sexuality. The results in this study were compa- ning treatment. In another study, peak or best sessions of

EFT STATUS JOHNSON ET AL. 73

EFT as identified by couples were characterized by more because initial &stress level is usually by far the best pre-
depth of experiencing and affiliative and autonomous dictor of long-term success in marital therapy, accounting
statements than were sessions identified by couples as for as much as 46% of the variance in marital satisfaction
poor. The final study demonstrated that intimate, emo- (Whisman &Jacobson, 1990). In addition, the task rele-
tionally laden self-disclosure (as coded on the SASB) was vance aspect of the alliance was more predictive of
more likely to lead to afiliative statements by the other improvement than was a positive bond or a sense of shared
partner than other randomly selected responses. Reveal- goals. EFT was more successful with couples who saw the
ing underlying experience in an intimate manner thus led relevance of formulating and expressing their attachment
to a change in interaction. In general, the results of these needs and fears and then addressing issues of connection
studies are consistent with the theory of EFT. They imply and trust.
that change in EFT is associated with the expression of Several other characteristics were associated with
underlying feelings and needs that leads to a positive shifi improvement in therapy. Older males (over 35 years of
in interaction patterns and fosters accessibility and respon- age) tended to report greater relationship adjustment at
siveness. follow-up and to make more gains in therapy, perhaps
finding issues of intimacy and attachment more relevant
P R E D I C T I N G SUCCESS IN EFT than did the younger male partners. This is an interesting
The program of research on EFT is now beginning to finding in that previous studies of couples therapy have
consider the question of who is best suited to this form found an inverse relationship between age and outcome,
of couples intervention. Clinically, EFT has never been leading some to suggest that treating older people appears
recommended for couples where abuse is an ongoing part to be more &&cult (Jacobson & Addis, 1993).
of the relationship. Abusive partners are referred to group A statistically and conceptually significant finding was
or individual therapy to help them deal with their abusive that a female partner’s initial level of faith that her partner
behavior. They are offered EFT only after this therapy is still cared for her predicted the couple’s adjustment and
completed and their partners no longer feel at risk. Addi- intimacy levels at follow-up. In a culture where women
tionally, for couples who are separating, EFT is used only have traditionally taken most of the responsibility for
in an abbreviated form to clarify the redefinition of the maintaining close bonds, this finding suggests that if the
relationship and support partners as they separate. female no longer has faith in her partner, the emotional
The results of a recent study on predictors of success investment necessary for change may be stifled. This is
in EFT (Johnson & Talitman, 1997) provide some initial consistent with accumulating evidence that emotional
evidence on who is likely to benefit most &omEFT. Hier- disengagement, rather than factors such as the inability to
archical multiple regression analyses were used in this resolve dsagreement, is predictive of marital unhappiness
study to assess the unique contribution of the predic- and instability (Gottman, 1994) and lack ofsuccess in cou-
tor variables to improvement or recovery fiom marital ples therapy (Jacobson & Addis, 1993). Low levels of this
&stress, beyond that due to initial satisfaction level. In element of trust may be a bad prognostic indicator for
addition to posttreatment assessment of participants, a 3- couples engaging in any form of couples therapy.
month follow-up was conducted. Several partner characteristics were found to be unre-
Based on regression analyses, the quality of the alliance lated to improvement in couple adjustment. Lack ofemo-
with the therapist was a strong predictor of success in EFT. tional expressivenessor awareness did not predict progress
This is to be expected, as it is a general finding in research in EFT; in fact, EFT seemed to be particularly helphl for
on all forms of psychotherapy that a positive alliance is males who were described by their partners as inexpres-
associated with therapeutic success. More important, sive. Traditionality in relationships (i.e., where the male
though, the quality of the d a n c e in EFT seemed to be a partner is oriented to independence and the female to
much more powerful and general predictor of treatment affiliation) also did not affect progress in EFT.
success than was initial distress level. In this study, initial Overall, then, the findings of this study suggest that a
distress level was not an important predictor of success at female’s level of faith in her partner’s caring and the cou-
the 3-month follow-up assessment (it accounted for only ple’s ability to engage in an alliance with the therapist and
4% of the variance in outcome). This is an unusual finding respond to the tasks of EFT are more important prognos-

CLINICAL PSYCHOLOGY: SCIENCE A N D PRACTICE V6 N I . SPRING 1999 74

tic indicators for the EFT therapist than are initial distress can also be delineated by conducting constructive studles,
level or factors such as emotional inexpressiveness. Future wherein particular interventions are added to the treat-
research is necessary to substantiate these important find- ment protocol and the effects on outcome are noted.
ings and to expand our knowledge on the range of client Only one study of EFT has used this constructive design
and couple characteristics that may influence the impact approach (James, 1991), finding that the addition of a
of EFT. communication skills component to the usual EFT in-
terventions did not increase the effectiveness of the
S U M M A R Y A N D FUTURE DIRECTIONS treatment.
From a theoretical perspective, there is more and more EFT change strategies have also been adapted and
evidence that the targets of EFT interventions, emotional applied to different populations, and some of these appli-
responses and patterned interactional cycles, are the most cations have been researched, such as the use of EFT with
significant features of marital distress. These variables reli- depressed spouses (Dessaulles, 1991; MacPhee et al.,
ably predict long-term relationship distress and dlsruption 1995; Walker, 1994). EFT is presently used in a wide vari-
(Gottman, 1994). Evidence is also accumulating that ety of settings, including private practices, university clin-
attachment theory can provide a theoretical basis for ics and counseling centers, and hospital outpatient clinics
understanding the nature of marital distress and adult love (Blanchard, 1994). EFT has also been taught to a large
in general. Attachment theory offers a theoretical basis for number of therapists, varied in age, sex, and experience,
interventions such as EFT. It would seem to be a consider- and has proved itself to be replicable in the outcome stud-
able advance for the field of couples therapy to begin to ies completed to date. In terms of outcome data, the effect
have systematic forms of intervention that are based on sizes for EFT are large and treatment outcomes generally
clear evidence of the nature of marital distress and the exceed the reported average 50% success rate for couples
nature of adult love relationships. Eficient, short-term therapy (Jacobson & Addis, 1993). The empirical evi-
change strategies require a clear set of targets and goals dence is consistent and clear: EFT is an effective treatment
that focus interventions on the variables most likely to for marital distress.
mediate recovery from marital distress. EFT offers a clear In terms of research design, the studles completed on
sequenced set of interventions and change processes, EFT have several strengths. They have, in general, used
based on the phenomenology ofmarital distress and a the- random assignment to group, valid control groups, treat-
ory of adult love that has empirical support. ment implementation checks with very acceptable
From a clinical perspective, perhaps the main general interrater reliability, reliable measures of process and out-
contribution the work on EFT has made to the field of come, follow-up analyses, and appropriate research meth-
couples therapy is to offer an orientation to and specific odology. The size of the groups in these studles was
ways ofworking with emotion. It also offers a way ofinte- relatively small, the largest being 45, but this is typical of
grating a focus on the individual and the relationship, on the field of couples therapy as a whole. Outcome mea-
both within- and between-person variables and processes. sures were mainly self-report; however, it has been argued
Clinical change processes in EFT seem, fkom the small that this is appropriate in this particular field (Jacobson,
number of studies described above, to be consistent with 1985a , 1985b). When EFT has been compared to other
the conceptualization of emotion in this model and how treatments (Dandeneau & Johnson, 1994; Johnson &
emotional experience and expression &e seen as hcilitat- Greenberg, 1985a), therapists have been nested within
ing change. The literature on psychotherapy process treatment, except for one, study (Goldman & Greenberg,
(Rice & Greenberg, 1984) emphasizes the need to specifj. 1992). This decision was based on the ecological validity
not just the variables associated with change but change of the treatment and the clinical perception that therapists
events where a number of variables occur in a specific do best when they are trained in and committed to the
context. This research can then be used to provide some interventions they use.
guidance for therapist interventions. This kind of research There are two important methodological limitations to
has begun with EFT in the description of a softening the current body of research on EFT. The first is that all
change event (Johnson & Greenberg, 1988)and will con- but one study involved one or both of the developers
tinue in the future. The active ingredients of a treatment of EFT as a research investigator. Many meta-analytic

EFT STATUS JOHNSON ET AL. 75

 reviews of psychotherapy outcome have found that EFT process. It is possible, for example, that EFT might
researcher allegiance to a treatment is positively associated be more effective for partners displaying an anxious
with the treatment’s effect size; moreover, this phenome- attachment style (i.e., those who are prone to hyperactiva-
non may be particularly important in the early stages of tion of attachment behaviors when distressed) than for
evaluating a treatment (Gaffin, Tsaousis, & Kemp- those displaying an avoidant attachment style (i.e., those
Wheeler, 1995). Accordingly, research is needed where who minimize attachment behaviors when &stressed).
EFT is evaluated by investigators less directly affiliated How EFT might be tailored to such individual differences
with its development. Such research is beginning to sur- is just beginning to be described in the literature (John-
face and report positive results (Denton, Burleson, Clarke, son & Whiffen, in press).
Rodriguez, and Hobbs, in press). The second limitation is Future research needs to address the limitations out-
one common to almost all current psychotherapies shown lined above concerning what therapy ingredients are nec-
to have significantclinical effects in research trials, namely, essary and sufficient for clinically significant change,
the generalizability of the research findings to the routine which couples are particularly suited to EFT, and how the
use of the treatment in clinical settings (see Clarke, 1995). process of treatment might be tailored to accommodate
An important task facing us is to begin examining whether individual differences. Additional questions remain as to
the strength of the reported EFT effects is maintained in how successful EFT can be in reducing individual symp-
clinical settings where therapists have received less training tomatology that accompanies relationship distress and the
and supervision in EFT than the therapists in the RCTs effectiveness of emotionally focused interventions for
and where clients may have greater marital and psycho- different populations. Some recent research has attempted
logical distress or concomitant problems such as eating to test the limits of EFT by exploring the type of thera-
disorders or anxiety disorders. peutic effects possible when couples present with other
The data on how change occurs in EFT are also lim- problems in addition to marital distress (e.g., depression,
ited. Dismantling studies, which identify and examine stress due to a child’s chronic illness). In our view, the test-
the effects of specific ingredients in a specific treatment ing of the limits of EFT’S clinical utility is a crucial ingre-
approach, have not been conducted. Although a link be- dient in further developing and refining an established
tween engagement in emotional experience, a specific treatment that has the promise to improve the ability of
kind of interactional shift, and treatment outcome was clinicians to alleviate marital distress.
found in one study (Johnson & Greenberg, 1988), it A pilot evaluation has been conducted on the use of
would enhance the model if the contribution of different emotionally focused interventions to create change in
interventions and processes to positive outcome was families of eating-disordered adolescents (Johnson, Mad-
delineated. In the process studies completed so far, the deaux, & Blouin, 1998) and a pilot is planned on the use
possibility exists that the process variables were the result ofEFT when one partner suffers from posttraumatic stress
of improvement rather than the mediator of improve- disorder. The work with traumatized couples is particu-
ment. Therefore, in a dismantling study, it would be useful larly interesting in that initial indications are that EFT
to determine the effectiveness of simply naming the nega- interventions appear to not only improve the couple’s
tive cycle and linking it to underlying attachment emo- relationship but to create a healing environment that
tions. These interventions constitute Steps 2 and 3 in EFT allows the trauma survivor to deal more constructively
and are designed to create a de-escalation of the negative with trauma symptoms such as flashbacks and emotional
cycle. The study might then examine how adding inter- numbing (Johnson & Williams-Keeler, 1998). However,
ventions designed to foster the reprocessing of emotional the change process with this population appears to be
responses (Step 5) and the shaping ofspecific new interac- longer and to contain specific pitfalls and difficulties, par-
tions (Step 7) might add to treatment effectiveness. It may ticularly regarding the creation of trust.
also be tiUitfLl to examine the process of change in spe- A second research direction involves the continuation
cific intermediate outcomes that partners face in therapy, of our efforts to understand the process of therapeutic
such as achieving closure on past betrayals and learning to change in EFT by examining the effect ofEFT on specific
depend on the other partner again. client cognitions, in this case schemas or working models
There has been little attention in the EFT research to concerning the acceptability and worthiness of self and the
the role of individual differences and how they affect the dependability of others. These types of cognitions are

CLINICAL PSYCHOLOGY: SCIENCE A N D PRACTICE V6 N1, SPRING 1999 76

cited in the attachment literature as crucial elements in and procedures. Journal qf Consulting and Clinical Psychology,
defining the security of an attachment bond and as being 63, 718-725.
significantly related to variables such as resilience in the Collins, N., & Read, S. (1990) Adult attachment, working mod-
face of stress and flexibility in information processing els and relationship quality in dating couples.Journal qfPerson-
ality arid Social Psychology, 58, 644-663.
(Mikulincer, 1997; Mikulincer, Florian, & Weller, 1993).
Dandeneau, M., &Johnson, S. (1994). Facilitating intimacy: A
One preliminary study suggests that a change in how the
comparative outcome study of emotionally focused and cog-
partner is perceived is associated with success in E F T
nitive interventions. Journal of Marital G. Family Therapy, 20,
(Greenberg et al., 1993). Other process-oriented research 17-33.
is also underway t o examine clients’ experience of E F T Denton, W., Burleson, B. R., Clarke, P. E., Rodriguez, C. l?, &
compared to more cognitively oriented marital interven- Hobbs, B. V., (in press). Outcomes of emotionally focused
tions. therapy for couples in a training clinic.Journal ofMarital and
One of the current criticisms of marital and family Family Therapy.
interventions is that the client’s perspective o n the change Ilessaulles, A. (1991). The treatment ofclinical depression in the con-
process has been largely ignored by researchers; this per- text ofmarital distress. Unpublished doctoral dissertation, Uni-
spective needs to be more hlly considered in subsequent versity of Ottawa.
studies of EFT. There are also plans to expand the work Dunn, R. L., & Schwebel, A. I. (1995). Meta-analytic review of
marital therapy outcome research. Journal of Family Psychol-
o n predictors of success in E F T in order t o facilitate the
Ogy, 9,58-68.
optimal matching of clients to treatments. Finally, we
Gaffin, E. A., Tsaousis, I., & Kemp-Wheeler, S. M. (1995).
believe that research examining the in-session effects of
Researcher allegiance and meta-analysis: The case of cogni-
specific therapist interventions, at specific points in ther- tive therapy for depression. Journal of Consulting and Clinical
apy where clients are involved in particular tasks, would y , 966-980.
l ’ s y ~ h ~ l ~ g63,
also be useful in this regard and would enhance the utility Goldman, A,, & Greenberg, L. (1992). Comparison of inte-
of E F T interventions to couple therapists. grated systemic and emotionally focused approaches to cou-
ples therapy.Journal of ConsuLing and Clinical Psychology, 60,
962-969.
REFERENCES Gottman, J. (1979). Marital interaction: Empirical investigations.
Alexander,J. F., Holtzworth-Munroe, A., & Jameson, P. (1994). New York: Academic Press.
The process & outcome of marital & family therapy: Gottman, J. (1991). Predicting the longitudinal course of mar-
Research review and evaluation. In A. Bergin & S. Garfield riages.Journal ofMarital and Family Therapy, 17, 3-7.
(Eds.), Handbook ofpsychotherapy and behavior change (pp. 565- Gottman, J. (1993). A theory ofmarital dissolution and stability.
607). New York Wiley. Journal ofFamily Psychology, 7, 57-75.
Bartholomew, K., & Perlman, D. (Eds.). (1994). Attachmentpro- Gottman, J. (1994). An agenda for marital therapy. In S. M.
cesses in adulthood: Advances in personal relationships (Vol. 5). Johnson & L. S. Greenberg (Eds.), The h e m qf the matter: Per-
London: Jessica Kingsley. spectives qfemotion in marital therapy (pp. 256-296). New York:
Baucom, D. H., Shoham, V., Mueser, K. T., Daiuto, A. D., & Brunner/Mazel.
Stickle, T. R.(1998).Empirically supported couple and fam- Gottman, J., & Levenson, R. W. (1986). Assessing the role of
ily interventions for marital distress and adult mental health emotion in marriage. Behavioral Assessment, 8, 31-48.
problems. Journal of Consulting and Clinical Psychology, 66, Gottman, J. M., Coan, J. Carrere, S., & Swanson, C. (1998).
53-88. Predicting marital happiness and stability from newlywed
Benjamin, L. (1986). Adding social and intra-psychic descriptors interactions.Journal of Mam’age and the Family, 60, 5-22.
to axis I of DSM-III. In T. Millon & G. Herman (Eds.), Con- Greenberg, L. S., Ford, C., Alden, L., & Johnson, S. M.
temporary directions in psychopathology (pp. 215-232). New (1993).Change processes in emotionally focused therapy.
York: Guilford Press. Journal of Consulting and Clinical Psychology, 61, 78-84.
Blanchard, J. J. (1994). Directory of research opportunitiesfor clinical Greenberg, L. S., &Johnson, S. M. (1988). Emotionallyfocused
psychology interns. Albany, N Y : Graywind. therapyfor couples. New York: Guilford Press.
Bowlby, J. (1969). Attachment G loss: Vol. I. Attachment. New Gurman, A. S. (1978). Contemporary marital therapies: A cri-
York: Basic Books. tique and comparative analysis of psychoanalytic behavioral
Bowlby,J. (1988). A secure base. New York: Basic Books. and systems theory approaches. In T. J. Poalino & B. S.
Clarke, G. N. (1995). Improving the transition from basic efE- McCrady (Eds.), Marriage and marital therapy (pp. 445-566).
cacy research to effectiveness studies: Methodological issues New York: Brunner/Mazel.

EFT STATUS JOHNSON € 1 AL. 77

 Hahlweg, K., & Marknian, H . J. (1988). Effectiveness ofbehav- ples therapy: An outcome study.Journal $Marital and Family
ioral marital therapy: Empirical status of behavioral tech- Therapy, 1 1 , 313-317.
niques in preventing and alleviating marital &stress. Journal qf Johnson, S . M., & Greenberg, L. S. (1988). Relating process to
Consulting and Clinical Psychology, 56, 440-447. outcome in marital therapy.Journal ofMarital and Family Ther-
Hazan, C., & Shaver, P. (1987). Conceptualizing romantic love apy, 14, 175-183.
as an attachment process. Journal ofPersonality and Social Psy- Johnson, S. M., & Greenberg, L. S . (1992). Emotionally focused
chology, 52, 5 11-524. therapy: Restructuring attachment. In S. Budman, M.
Heavey, C. L., Christensen, A., & Malaniuth, N. M. (1995). Hoyt, & S . Friedman (Eds.), First sessions in time-efectivepsy-
The longitudinal impact of demand and withdrawal during chotherapy: A casebook (pp. 204-224). New York: Guilford
marital conflict. Journal of Consulting and Clinical Psychology, Press.
6 3 , 797-801. Johnson, S. M., & Greenberg, L. S . (Eds.). (1994). The heart of
Horowitz, L. M. (1979). Cognitive structure of interpersonal the matter: Perspectives on emotion in marital therapy. New York:
problems treated in psychotherapy.Journal of Consulting and Brunner/Mazel.
Clinical Psychology, 4 7 , 5-15. Johnson, S. M., & Greenberg, L. S. (1995). The emotionally
Jacobson, N. S. (1978). A review of the research on the effec- focused approach to problems in adult attachment. In N. S.
tiveness ofmarital therapy. In T. J. Paolino & B. S . McCrady Jacobson & A. S. Gurman (Eds.), The clinical handbook ofmari-
(Eds.), Marriage and marital therapy (pp. 395-444). New York: tal therapy (2nd ed., pp. 121-141). New York: Guilford Press.
Brunner/Mazel. Johnson, S. M., Maddeaux, C., & Blouin, J. (1998). Emotionally
Jacobson, N. S. (1985a). The role of observational measures in focused family therapy for bulimia. Psychotherapy, 35, 238-
behavior therapy outcome research. Behavioral Assessment, 7, 247.
297-308. Johnson, S., & Talitman, E. (1997). Predictors of outcome in
Jacobson, N. S. (1985b). Uses versus abuses of observational emotionally focused marital therapy. Journal of Marital and
measures. Behavioral Assessment, 7, 323-330. Family Therapy, 2 3 , 135-152.
Jacobson, N. S., & Addis, M. E. (1993). Research on couples Johnson, S . M., & Whiffen, V. (in press). Made to measure:
and couple therapy: What do we know? Where are we Adapting emotionally focused couples therapy to partner’s
going?Journal $Consulting and Clinical Psychology, 61, 85-93. attachment styles. Clinical Psychology: Science and Practice.
Jacobson, N. S., Follette, W. C., &Elwood, R. W. (1984). Out- Johnson, S . M., & Williams-Keeler, L. (1998). Creating healing
come research in behavioral marital therapy: A methodologi- relationships for couples dealing with trauma: The use of
cal and conceptual reappraisal. In K. Hahlweg & N. S . emotionally focused marital therapy. Journal of Marital and
Jacobson (Eds.), Marital interaction: Analysis and modification Family Therapy, 2 4 , 25-40.
(pp. 113-129). New York Guilford Press. Klein, M. H., Mathieu, P. L., Gendlin, E. T., & Miesler, D. J.
Jacobson, N. S., & Margolin, G. (1979). Marital therapy: Strategies (1969). The experiencing scale: A research and training manual
based on social learning and behavior exchange principles. New (Vol. 1). Mahson: University of Wisconsin, Bureau of
York BrunnedMazel. Audiovisual Instruction.
Jacobson, N. S., & Truax, I? (1991). Clinical significance: A sta- Kobak, R., & Hazan, C. (1991). Attachment in marriage: Effects
tistical approach to defining meaningfkl change in psycho- of security and accuracy ofworking models.Journal ofperson-
therapy research. Journal of Consulting and Clinical Psychology, ality and Social Psychology, 60, 861-869.
59, 12-19. Lebow,J. L., & Gurman, A. S. (1995). Research assessing couple
James, P. (1991). Effects of a communication training compo- and family therapy. Annual Review of Psychology, 46, 27-57.
nent added to an emotionally focused couples therapy.Journa1 MacPhee, D.C., Johnson, S. M., & Van der Veer, M. C. (1995).
ofMarita1 and Family Therapy, 17, 263-276. Low sexual desire in women: The effects of marital therapy.
Johnson, S. M. (1996). The practice of emotionally focused marital Journal $Sex and Marital Therapy, 21, 159-182.
therapy: Creating connection. New York Brunner/Mazel. Mikulincer, M. (1997). Adult attachment style and information
Johnson, S. M. (in press). Emotionally focused couples therapy: processing: Individual differences in curiosity and cognitive
Straight to the heart. In J. Donovan (Ed.), Short term couples closure. Journal of Personality and Social Psychology, 7 2 , 1217-
therapy. New York Guilford Press. 1230.
Johnson, S., & Greenberg, L. (1985a). The mfferential effec- Mikulincer, M., Florian, V., & Weller, A. (1993). Attachment
tiveness of experiential and problem solving interventions in styles, coping strategiesand post-traumatic psychologicaldis-
resolving marital conflict. Journal $ Consulting and Clinical tress. Journal $Personality and Social Psychology, 6 4 , 817-826.
g ~ 175-184.
P ~ ~ c h o l o53, Minuchin, S., & Fishman, H. C. (1981). Family therapy tech-
Johnson, S., & Greenberg, L. (1985b). Emotionally focused cou- niques. Cambridge, MA: Harvard University Press.

CLINICAL PSYCHOLOGY: SCIENCE AND PRACTICE V6 N l , SPRING 1999 78

Perk, F. (1973). Thegestalt approach and eyewitness to therapy. San Shadish, W. R., & Ragsdale, K. (1996). Random versus nonran-
Francisco: Science & Behavior Books. dom assignment in controlled experiments: Do you get the
I’iercy, F. P., & Sprenkle, D. H. (1990). Marriage and family same answer?Journal of Consulting arid Clinical Psychology, 64,
therapy: A decade review. Journal ofMarriage and the Family, 1290-1305.
52, 1116-1 126. Simpson,J. A., Rholes, W. S., & Nelligan, J. S. (1992). Support
Rice, Z . N., & Greenberg, L. S . (1984). Patterns qfchange: Inten- seeking and support giving within couples in an anxiety pro-
sive analysis ofpsychotherapyproces~.New York: Guilford Press. voking situation: The role of attachment styles.Journal $Per-
Roberts, T. W. (1992). Sexual attraction and romantic love: For- sonality and Social Psychology, 62, 434-446.
gotten variables in marital therapy.Journal qfMarital and Fam- Spanier, G. B. (1976). Measuring dyadic adjustment: New scales
ily Therapy, 18, 357-364. for assessing the quality ofmarriage and similar dyads.Journa/
Rogers, C. (1951). Client-centered therapy. Boston: Houghton- qfMarriage and the Family, 38, 15-28.
Mifflin. Walker, J. (1994). A marital intervention program for couples with
Rosenthal, R. (1984). Meta-analytic proceduresfor social research. chronically ill children. Unpublished doctoral dissertation, Uni-
Beverly Hills, C A Sage. versity of Ottawa.
Schover, L. R., & Lopiccolo, J. (1982). Treatment effectiveness Walker, J., Johnson, S., Manion, I., & Cloutier, P. (1996). An
for dysfunctions of sexual desire. Journal of Sex and Marital emotionally focused marital intervention for couples with
Therapy, 8, 179-197. chronically ill children. Journal of Consulting and Clinical Pry-
Schwarzer, R. (1990). Meta-analysis Programs (Version 5.1) cholog~64, 1029-1036.
[Computer software]. Berlin: Author. Walker, J., & Manion, I. (1998). A two-yearfollow-up on an emo-
Segraves, R. J. (1990).Theoretical orientations in the treatment tionally focused interventionfor couples with chronically ill children.
ofmarital discord. In F. D. Fincham & T. N. Bradbury (Eds.), Manuscript in preparation.
The psychology of marriage (pp. 281-298). New York Guil- Whisman, M. A., &Jacobson, N. S. (1990). Power, maritalsatis-
ford Press. faction, and response to marital therapy.Journa1 ofFamily Psy-
Shamsh, W. R., Montgomery, L. M., Wilson, P, Wilson, chology, 4, 202-212.
M. R., Bright, I., & Okwumabua, T. (1993). Effects offam-
ily and marital psychotherapies: A meta-analysis. Journal of Received November 2, 1997; revised March 31, 1998; accepted
Consulting and Clinical Psychology, 61, 992-1002. March 31, 1998.

EFT STATUS - JOHNSON ET AL. 79