Greenberg 2008

Journal of Counseling Psychology Copyright 2008 by the American Psychological Association

2008, Vol. 55, No. 2, 185–196 0022-0167/08/$12.00 DOI: 10.1037/0022-0167.55.2.185

Differential Effects of Emotion-Focused Therapy and Psychoeducation in

Facilitating Forgiveness and Letting Go of Emotional Injuries
Leslie S. Greenberg, Serine H. Warwar, and Wanda M. Malcolm
York University
This study compared the effectiveness of emotion-focused therapy (EFT) involving gestalt empty-chair
dialogue in the treatment of individuals who were emotionally injured by a significant other with therapy
in a psychoeducation group designed to deal with these injuries. In addition, this study examined aspects
of the emotional process of forgiveness in resolving interpersonal injuries and investigated the relation-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
ship between letting go of distressing feelings and forgiveness. A total of 46 clients assessed as having
This document is copyrighted by the American Psychological Association or one of its allied publishers.
unresolved, interpersonal, emotional injuries were randomly assigned to an individual therapy treatment
of EFT or a psychoeducation group. Clients were assessed at pretreatment, posttreatment, and 3-month
follow-up on measures of forgiveness, letting go, depression, global symptoms, and key target com-
plaints. Results indicated that clients in EFT using empty-chair dialogue showed significantly more
improvement than the psychoeducation treatment on all measures of forgiveness and letting go, as well
as global symptoms and key target complaints.
Keywords: interpersonal injuries, forgiveness, letting go, emotion-focused therapy, psychoeducation
Forgiveness has recently been proposed as an important aspect Enright & Fitzgibbons, 2000; Fitzgibbons, 1986; Hope, 1987;
of emotional recovery following an interpersonal injury (Enright & Karen, 2001; Malcolm, Warwar, & Greenberg, 2005). One of the
Fitzgibbons, 2000; Worthington, 1998, 2001). Although the debate assumptions of EFT is that the blocking of primary biologically
continues as to whether forgiveness should be granted to injurers adaptive emotions subverts healthy boundary setting, self-
when they will not take responsibility for their actions or when respectful anger, and necessary grieving and that adequate pro-
they continue to perpetuate harmful acts (Worthington, 2005), cessing of unresolved emotion leads to its transformation (Green-
forgiveness has been shown to have a positive impact on physical, berg, 2002). Many clinicians (Akhtar, 2002; Baures, 1996; Enright
relational, mental, and spiritual health, whereas unforgiveness can & Eastin, 1992) have emphasized the value of facilitating in-
be distressing and may leave people ruminating about their injuries session expressions of adaptive anger at violation and suggested
and feeling hostile toward those who injured them (Witvliet, that facilitating forgiveness requires an acknowledgement of the
Ludwig, & Vander Laan, 2001). Most studies on the facilitation of legitimacy of emotions such as resentment and hatred toward the
forgiveness (Al-Mabuk, Enright, & Cardis, 1995; Hebl & Enright, offender. Baures (1996), Boss (1997), and Fincham (2000) con-
1993; McCullough & Worthington, 1995; Ripley & Worthington, sidered resentment and desires for revenge to be closely linked
2002; Wade, Worthington, & Meyer, 2005; Worthington & with self-respect, and Greenberg and Paivio (1997) suggested that
Drinkard, 2000) have involved psychoeducation group programs there may be times when it is therapeutic to encourage clients to
designed to promote the benefits of forgiveness to self and others,
talk about their revenge fantasies. From these perspectives, the
and these provide the knowledge and skills associated with a
desire to retaliate is normalized as a sign of how damaged the
particular model of forgiveness. To date, only two studies have
injured person feels. Encouraging such expressions in therapy is
been published on the effectiveness of individual therapy in facil-
not the same thing as promoting outer-directed blaming or hurling
itating unilateral forgiveness (Coyle & Enright, 1997; Freedman &
of insults. In encouraging clients to speak from their inner expe-
Enright, 1996). The present study examined the effectiveness of
riences of violation, the therapist is promoting ownership of the
emotion-focused therapy (EFT) compared with that of a psycho-
clients’ emotional experience and is empowering clients to appro-
education group (PG) in facilitating emotional resolution and
forgiveness. priately assign responsibility for harm done. Ownership of emotion
Numerous investigators have proposed that both emotion work also helps clients focus on their own needs and concerns rather
and empathy play important roles in forgiveness (Davenport, 1991; than getting stuck in blaming the other or feeling victimized. The
danger in short-circuiting expressions of anger as might occur in
some treatments is that a client may end up condoning or excusing
the injurer’s hurtful behavior or inappropriately take too much
Leslie S. Greenberg, Serine H. Warwar, and Wanda M. Malcolm, responsibility for the unfolding of events that surrounded the
Department of Psychology, York University, Toronto, Ontario, Canada.
injury.
This study was supported by Grant CRF 5202 from the Campaign for
Forgiveness Research awarded to Leslie S. Greenberg. Therapists also need to facilitate the process of grieving the loss
Correspondence should be sent to Leslie S. Greenberg, Department of of, or damage to, a significant relationship, as well as the shattering
Psychology, York University, 4700 Keele St., Toronto ON, M3J 1P3 of the client’s view of self and the world that may have been
Canada. E-mail: [email protected] caused by the injury. Akhtar (2002) addressed the relationship
185
186 GREENBERG, WARWAR, AND MALCOLM
between mourning and forgiveness, and Greenberg and Paivio could do under the same circumstances. In addition to assisting in
(1997) emphasized that work with betrayal and abandonment often the revision of how one sees the injurer, cognitive perspective
involves a process of facilitating normal grieving in which anger taking sometimes allows the injury itself to be recast within a
and sadness play central roles. Facilitating an imaginary dialogue broader understanding of the context of the unfolding of events.
with the injurer can help the client grieve and say goodbye to what However, cognitive perspective taking of this nature does not have
has been lost or irreparably damaged as a consequence of the to involve warm, benevolent feelings associated with forgiveness.
injury (Elliott, Watson, Goldman, & Greenberg, 2004; Greenberg, Something more is required for forgiveness, and this appears to be
Rice, & Elliott 1993). compassion for the injurer, or affective empathy.
Emotion-focused therapy (Greenberg, 2002), which uses empty- Affective empathy is best understood as a means of imaginative
chair dialogue as a major method for the resolution of unfinished entry into the world of the other, which generates a bodily felt
business, has been found to be an effective intervention in the sense of understanding what the other person may have been
treatment of depression, interpersonal problems. and trauma feeling, without actually sharing the same experience (Greenberg
(Greenberg & Watson, 1998, 2006; Paivio & Greenberg, 1995; & Rosenberg, 2002). Berecz (2001) has suggested that the task for
Paivio & Nieuwenhuis, 2001). Paivio and Greenberg’s (1995) the injured person is to imaginatively transpose himself or herself
comparative study of a PG intervention versus individual EFT into the other person’s place in an attempt to understand the
using gestalt empty-chair dialogue in the resolution of unfinished unfolding of events from the injurer’s perspective.
business supported the efficacy of the empty-chair intervention. Unforgiveness has been defined as the combination of a com-
The results showed a significant reduction in symptomatology, plex set of negative feelings toward an injurer, and it has been
target complaints, and interpersonal distress and more resolution of shown that people can decrease unforgiveness without increasing
unfinished business. The empty-chair method used in EFT is also forgiveness (Worthington, Sandage, & Berry, 2000; Worthington
a particularly effective tool in promoting empathy toward the & Wade, 1999). Unforgiveness is regarded as being stuck in
offender (Paivio & Greenberg, 1995; Paivio, Hall, Holowaty, negative emotions and experiencing a hyperaroused stress re-
Jellis, & Tran, 2001; Paivio & Nieuwenhuis, 2001). In imagina- sponse resulting from rumination (Harris & Thoresen, 2005). It is
tively bringing the injurer and injury alive, the client moves from noteworthy that reducing unforgiveness is not the same as promot-
a cognitive discussion with the therapist to an imaginal confron- ing forgiveness. Forgiveness seems to include the reduction of
tation and dialogue with the injurer. In so doing, the client is unforgiveness, or letting go, through decreasing negative feelings
helped to move reified inner representations of self and other and thoughts in relation to the injurer. In addition to, and in
(injurer) into a transitional space in conscious awareness where the contrast with, letting go or reducing unforgiveness, forgiveness is
representations can be re-examined, reworked, and resolved. also composed of the increase of positive emotions such as com-
Empirical evidence also is mounting in support of the impor- passion, empathy, or understanding felt toward the injurer.
tance of transforming emotions by changing one emotion with In our view, forgiveness thus appears to involve two important
another emotion (Fredrickson, 1998; Greenberg, 2002, 2004), and emotional processes: resolution of the hurt and anger involved in
this suggests that a maladaptive emotion state can be effectively the injury; and the possible generation of positive feelings of
transformed by undoing it with the presence of another, more compassion, loving, kindness, and empathic concern for the in-
adaptive emotion. More specifically, Fredrickson, Mancuso, jurer. People thus may be able to resolve emotional injuries by
Branigan, & Tugade (2000) have shown that positive emotion reducing or letting go of their bad feelings or by letting go of bad
undoes the cardiovascular aftereffects of negative emotion. Green- feelings and increasing positive feelings (i.e., by forgiving).
berg (2002) has suggested that the key to transforming maladap- The main purpose of the present study was to evaluate the
tive emotions is to access alternate, healthy, adaptive emotions to effectiveness of EFT (Greenberg et al., 1993) involving empty-
act as resources in the self. Thus, in an emotion-focused treatment, chair dialogue in the treatment of individuals who had been emo-
feelings related to unforgiveness such as anger, contempt, and pain tionally injured by a significant other. Empty-chair work was used
are eventually changed by accessing feelings of sadness, compas- to facilitate emotional transformation by asking clients to express
sion, empathy, and concern. and process their anger and sadness with the offender, thereby
McCullough and his colleagues have shown that empathy for promoting empathy, and by asking clients to play the role of the
the perpetrator mediates successful forgiveness (McCullough, injurer, thereby having them imagine what the injurer might feel if
Rachal, Sandage, & Worthington, 1997; McCullough, Worthing- he or she were capable of comprehending the consequences and
ton, & Rachal, 1997). This proposition is consistent with clinical impact of his or her actions on the client. The primary hypothesis
observation, theory, and empirical evidence concerning forgive- was that EFT using empty-chair dialogue to process unresolved
ness (Macaskill, Maltby, & Day, 2002; McCullough, Rachal, emotion would produce better outcomes than PG in the treatment
Sandage, & Worthington, 1997; McCullough, Worthington, & of interpersonal emotional injuries on measures of forgiveness and
Rachal, 1997; Worthington & Wade, 1999). When accessed, em- letting go and other indices of outcome. It was assumed that an
pathy involves understanding another’s feelings and is a complex experiential treatment such as EFT that works by evoking, pro-
cognitive/affective state that facilitates forgiveness of an interper- cessing, and transforming emotion would address the emotional
sonal injury. As Rowe et al. (1989) have pointed out, empathy causes of the injury more directly than a PG treatment that is less
toward the injurer involves being able to see the other person as emotionally activating.
acting in a quintessentially human manner, which flows out of the This study also examined the emotional process of forgiveness
context of his or her own self-focused needs and perceptions. This in resolving interpersonal injury and evaluated whether forgive-
includes (but does not require) the possibility of recognizing that ness was necessary to resolution of the injury. In the present study,
what the injurer did was similar to something one has done or recovery from an emotional injury therefore was conceptualized as
FACILITATION OF FORGIVENESS 187
occurring in one of the following two ways: (a) forgiving the nantly White, with 1 client of South Asian and 1 of East Asian
injurer, which was defined as letting go of unresolved bad feelings origins in each treatment group.
or reducing unforgiveness plus the development of empathy and In terms of pretreatment diagnosis on Diagnostic and Statistical
compassion for the other, or (b) letting go of the bad feelings, Manual of Mental Disorders (4th ed., text rev.; American Psychi-
which involves letting go of unmet needs and negative feelings in atric Association, 2000) Axis I and II, there were 9 clients (39%)
relation to the injury or injurer and changing negative perceptions with a least one Axis I diagnosis in the EFT treatment, and 7
of self in relation to the injury or injurer, without the development clients (30%) with at least one Axis I diagnosis in the PG treat-
of empathy and compassion for the other. The second hypothesis ment. In the EFT treatment, 6 clients (26%) had at least one Axis
was that all people who forgave would let go of bad feelings but II diagnosis, and in the PG treatment, 4 clients (17%) had an Axis
that not all people who let go would forgive. The third hypothesis II diagnosis. The mean score on the Global Assessment of Func-
was that reported emotional arousal would be higher in EFT than tioning was 77 for each treatment group, with the range being
in PG. 65–95 for the EFT treatment group and 60 –90 for the PG treat-
Treatment focused on facilitating the resolution of specific ment group. There were no statistically significant differences
unresolved interpersonal emotional injuries that had occurred at between treatment conditions on any of these variables.
least 2 years prior to the start of therapy and continued to be

distressing. The injuries that clients brought to treatment were both Emotional Injuries
emotional and interpersonal; they were emotional in that they
involved intense lingering unresolved feelings of hurt or anger and The types of emotional injuries clients presented in this study
betrayal, and they were interpersonal in that the injurer was a and the nature of the relationship with the significant other are
significant other in the injured person’s life. Injuries involved summarized in Table 1. Each participant targeted one injury.
abandonment, betrayals, or violations by significant others, such as Parents were the main perpetrators of the injuries. In the EFT
friends, bosses, family members, or intimate partners. treatment, 18 (78%) of the clients were dealing with an interper-
sonal injury in relation to at least one parent, and 2 (8%) with an
ex-partner, while in the PG treatment, 13 (57%) of the participants
Method were dealing with an injury regarding at least one parent and 5
(21%) with an ex-partner. The people in the “other” category
Participants included a boss, a neighbor, and people in nonparental positions of
The sample for the present study consisted of 46 clients who had power relative to children.
an unresolved interpersonal, emotional injury with a significant
other that had occurred at least 2 years prior to the client’s Therapists
commencing treatment. The requirement that the injury not be
There were eight therapists in the EFT condition, two of whom
more recent was to ensure that the natural process of recovering
were male and six female. Two of the therapists were registered
from hurts had been given time to work and that the injured person
psychologists, one had a doctorate, and five were advanced doc-
was not in the midst of coping with the immediate aftermath of the
toral students in clinical psychology. Before training for the treat-
injury. Participants were required to be 18 years of age or older.
ment study, all therapists were required to have had at least 1 year
Exclusion criteria for the study— based on the assumption a brief
of EFT training, including prior empathy training, and 1 year of
treatment program would be unsuitable for some people—were as
follows: victims of incest; individuals who had attempted suicide
or had lost a significant other in the past year; those currently in
physically violent relationships; individuals currently abusing Table 1
drugs or alcohol; and individuals diagnosed with antisocial, bor- Interpersonal Emotional Injuries Reported by Client Sample
derline, or narcissistic personality disorder, posttraumatic stress
disorder, or a psychotic disorder. Individuals who were already in Emotion-focused Psychoeducation
Variable therapy group
psychotherapy elsewhere were also excluded.
The average age for clients in the EFT treatment was 43 years. Issue
The average age for the PG treatment was 46 years. The overall Betrayal 5 8
population thus had a mean age of 44.5 years (SD ⫽ 8.3; range, Criticism 6 4
Neglect 4 3
22– 67 years). There were 7 men and 16 women in the EFT Abandonment 5 4
treatment and 13 men and 10 women in the PG treatment. In the Physical abuse 2 3
EFT treatment, 8 individuals had never been married, 13 were Sexual abuse 1 1
married, and 2 were separated or divorced. In the PG treatment, 5 Significant other
had never been married, 9 were married, and 9 were separated or Both parents 5 5
Mother 8 7
divorced. In terms of their level of education, in the EFT treatment, Father 5 4
1 individual had completed high school, 3 had some college or Ex-partner 2 2
university training, 12 had graduated from college or university, Sibling 2 2
and 7 had postgraduate experience. In the PG treatment, 5 indi- Child 0 1
Other 1 2
viduals had completed high school, 3 had some college or univer-
sity training, 6 had graduated from college or university, and 9 had Note. N ⫽ 23 in each treatment group. Each participant targeted one
postgraduate experience. Ethnicity in both groups was predomi- injury only.
experience as a therapist. Therapists in the EFT treatment received not to be done in every session but (if suitable) in at least half the
an additional 30 hr of specialized training that was based on a sessions in the evocation phase. The next two phases integrate
treatment manual for resolving emotional injuries developed for within and overlap with Phase 2.
this project (Greenberg, Malcolm & Warwar, 2002). The group in Phase 3: Self-interruptive work. The third phase of therapy
the PG condition had two leaders. One of the group leaders was a involves interventions facilitated by therapists at client markers of
registered psychologist who had devised the PG treatment and interruption such as emotional constriction, resignation, or hope-
conducted it on a number of prior occasions. The co-leader was a lessness. These interventions are aimed at turning the passive,
doctoral student and was trained by the first leader. The therapists automatic process of interruption into an active one. This phase
in both treatment conditions received weekly supervision through- aims to heighten clients’ awareness of how they interrupt them-
out the study to promote adherence to treatment manuals. selves and to promote change in these interruptive processes so
that emotions preventing resolution can be accessed and processed.
Treatments Phase 4: Empowerment and letting go or forgiving. This final
phase entails accessing previously unexpressed emotions and mo-
Both interventions involved 12 hr of treatment distributed over bilizing and promoting the entitlement of unmet needs. The ther-
approximately 12 weeks. apist promotes a change in the way the client views the injurer,
facilitated by emotional arousal and accessing of past unmet needs.
Emotion-Focused Therapy (EFT) This phase also involves helping clients grieve and let go of unmet
needs. Elaborating the worldview of the other aids in developing
The treatment manual for this study was based on the principles empathy toward the injurer, and the therapist helps the client
outlined for EFT (Greenberg et al., 1993; Greenberg & Rosenberg, understand or hold the other accountable.
2002), also known as process– experiential therapy. This therapy Homework. Clients were asked to complete homework
includes the implementation of the person-centered relational at- throughout the course of treatment. At the start, they were asked to
titudes of empathy, positive regard, and congruence, as well as keep a diary of their feelings and thoughts in relation to their
marker-guided, process-directive, experiential interventions. In injuries and to note how the therapy sessions played a role in their
EFT, the therapist uses the following interventions: gestalt two- change processes. At Session 6, clients were given a handout that
chair dialogues when clients present in-session self-evaluative instructed them to write an unmailed letter to the injurer, accusing
conflicts; client-centered systematic evocative unfolding for prob- the injurer of knowing the impact of his or her hurtful behavior.
lematic reactions over which clients are puzzling; and gestalt The second part of this homework exercise instructed clients to
empty-chair dialogue for resolving clients’ currently felt unfin- write a letter in response to themselves, denying the accusation
ished business with a significant other. Focusing (Gendlin, 1996) from the perspective of the person who injured them. This was
is also utilized in this approach to assist clients in attending to their done to highlight that change was to come from the client, not the
internal experience and to obtain a bodily felt sense of the issues perpetrator, since the perpetrator may never change. At Session 7,
they are exploring and struggling with (Gendlin, 1996; Greenberg clients were given some definitions of forgiveness and asked to
et al., 1993). The emphasis in EFT is on accessing primary think about their understanding of forgiveness and to consider
adaptive feelings and maladaptive emotion schemes in order to whether forgiveness was important to them personally with respect
make them amenable to change (Greenberg et al., 1993; Greenberg to their emotional injuries. At Session 8, clients were given a
& Paivio, 1997). handout that asked them to reflect on the bond that still held them
A specialized EFT treatment manual (Greenberg, Malcolm, & to the injurer and to write down their difficulties in letting go of the
Warwar, 2002) was developed for this project to focus on facili- emotional injury and what was sustaining the painful feelings.
tating the resolution of emotional interpersonal injuries. The treat- These were discussed in the session
ment protocol is summarized in the following four phases that
overlap rather than being purely sequentially. Psychoeducation Group (PG)
Phase 1: Creating an alliance. The first phase of treatment
involves creating a therapeutic alliance with the client by empathi- The PG manual was devised for this study and drew on various
cally responding to and validating the client’s pain and emotional sources for content (Bolger, 1999; Klassen, 2001; Paivio & Green-
experience of the interpersonal injury. This stage also entails berg, 1995; Smedes, 1984; Worthington and Drinkard, 2000). The
helping clients to identity the impact of the injury and articulate six workshops were facilitated by the two workshop leaders. The
and clarify the most problematic aspects of the injury for them. introductory session provided an overview of all the sessions,
Phase 2: Evocation and exploration. The second phase of along with a rationale for participating in the study and an expla-
treatment involves acknowledging, experiencing, and expressing nation of the differences between PG and group therapy. Sessions
the anger, sadness, pain, and other distressing feelings associated 2–5 included a discussion of the previous session’s homework, a
with the emotional injury. Empty-chair work is used to help clients topic presentation by the facilitators, a coffee break with a personal
process unresolved feelings toward the injurer. In our study, ther- reflection task, and then group discussion of the presentation and
apists were advised to begin work on evocation and exploration as personal reflection. Each session ended with the assignment of a
early as the second session if the injury was clear and if safety and homework task to be done between sessions and completion of
the bond seemed to be sufficiently strong and no later than the session measures. The content presentations covered the following
third session if the client appeared ready. In addition, therapists topics: the nature and structure of an emotional injury; understand-
were advised to continue evocation and exploration up until the ing unfinished business and how it disrupts adaptive functioning;
penultimate session if necessary. Empty-chair work, however, was aspects of forgiveness, including what it is and is not, and why one
would be motivated to forgive in the face of being hurt by another authors of the EFI did not use the term forgiveness in any other
person; the role of pain and other strong emotions in experiencing item of the EFI measure to avoid creating conceptual biases. For
and recovering from interpersonal emotional injuries; the process this reason, the EFI is referred to as the Attitude Scale during its
of reconciliation with the hurtful other and how it differs from administration. The Forgiveness Measure is thus used to directly
forgiveness; and finally how to resolve an injury. assess degree of forgiveness. In classifying clients’ degree of
The content presentation of the first session was short, and forgiveness, a score of 4 or above (i.e., meaning that they had
group discussion was intentionally structured to create safety and forgiven either a lot or completely) was used to indicate that
assure group members that their participation in the group discus- forgiveness had been attained.
sions was voluntary and at their discretion. In the final session, the Unfinished Business Empathy and Acceptance Scale (UFB EA;
material presented in the first five sessions was reviewed, and Singh, 1994). Singh developed the Unfinished Business Scale to
instead of receiving a homework assignment, clients interested in
measure resolution of unfinished business with a significant other.
further reading were given a bibliography on the topics covered in
For the present study, items were extracted from this scale and
the workshops.
adapted to create the Empathy and Acceptance Scale (UFB EA).
The group members in PG treatment received the same
The UFB EA comprises six items and measures the extent to

homework assignments as the participants in individual therapy

in the same order and at approximately the same time intervals. which clients feel acceptance and empathy toward the individual
They also received additional journaling and reflection exer- who injured them. Clients were asked to indicate their agreement
cises so that they would have homework assignments after each with the statements of empathy or acceptance on a 5-point Likert
of the first five workshops. It was assumed that any change scale (1 ⫽ not at all, 5 ⫽ very much). For example, empathy items
experienced in the PG members would be the result of discus- consist of statements expressing understanding of the injurer and
sion of and reflection on the information provided, which would the personal difficulties that he or she could be facing. Acceptance
produce change in attitude(s). It was further assumed that items include statements expressing receptiveness toward the other
attitude change would in turn change the feelings participants person and reverse-scored items expressing negative views of the
had toward the injurer and injury. other person. The items on the UFB EA have been found to
intercorrelate highly in a sample of clients in treatment for unfin-
Measures ished business (Paivio & Greenberg, 1995) and in a sample of
university students (Singh, 1994), and the overall UFN Scale has
A battery of self-report measures was administered before and been found to correlate with other outcome measures (Paivio &
after treatment to assess changes in specific domains. All clients Greenberg, 1995; Watson & Greenberg, 1996). Cronbach’s alpha
were assessed approximately 1 week prior to treatment and 1 week for the UFB EA in this sample was .87.
following treatment. Clients were also assessed at a 3-month
follow-up on measures of letting go, forgiveness, level of depres-
sion, global symptoms, and target complaints. Measures of Letting Go
Unfinished Business Feelings and Needs Scale (UFB FN).

Measures of Forgiveness
The Feelings and Needs Scale (UFB FN) was adapted from
The Enright Forgiveness Inventory (EFI; Enright, Rique, & Singh’s (1994) Unfinished Business Scale. It measures the client’s
Coyle, 2000). The EFI is a measure of the degree to which one resolution of feelings and needs as they relate to the injurer and
person forgives another who has hurt him or her deeply or unfairly. positive changes in the perception of self. This measure comprises
The first part of the inventory instructs individuals to visualize the eight items rated on a 5-point Likert scale (1 ⫽ not at all, 5 ⫽ very
emotional injury, focus on the offending person, and imagine what much). The UFB FN Scale contains three sets of items that refer to
happened. The participant is asked to write a description of the feelings, needs, and the self. The Feelings subscale contains items
injury in his or her own words. The second part of the inventory such as a statement describing the client’s unresolved feelings in
comprises 60 items from three 20-item subscales measuring affect, relation to the injurer. The Needs subscale includes items such as
behavior, and cognition in relation to forgiveness. Participants are a statement reflecting the client’s frustration about not having
asked to rate each item on a 6-point Likert scale (1 ⫽ strongly
needs met by the other person. Items pertaining to self include
disagree, 6 ⫽ strongly agree). The authors have reported a stabil-
statements such as one describing how negative treatment by the
ity coefficient in a community sample for total EFI scores of .86
injuring party has affected the client’s self-esteem. The UFB FN
during a 4-week test–retest reliability study. The subscale test–
has shown interitem reliability in a sample of clients in treatment
retest coefficients ranged from .67 to .91. With respect to concur-
rent validity, the EFI has been positively associated with other and a sample of university students and has been found to correlate
measures of forgiveness. There is also support for the EFI’s with other outcome measures in a sample of clients in treatment.
divergent validity (Enright et al., 2000). Cronbach’s alpha for the Cronbach’s alpha for the UFB FN in this sample was .79.
scale as a whole in this sample was .81. Letting Go Measure. This measure was constructed for this
Forgiveness Measure (Enright et al., 2000). This single-item study to parallel the single-item Forgiveness Measure. It is a
scale for assessing degree of forgiveness derives from the EFI and single-item self-report measure that assesses the extent to which
directly measures the extent to which clients have forgiven the clients have let go of their hurt and angry feelings toward the
person who injured them. Responses are indicated on a 5-point injurer. Responses are indicated on a 5-point Likert scale (1 ⫽ not
Likert scale (1 ⫽ not at all, 3 ⫽ in progress, 5 ⫽ completely). The at all, 3 ⫽ in progress, 5 ⫽ completely).
Measures of Other Outcomes community. The advertisements announced that a treatment study
was being conducted at the York University Psychotherapy Clinic
Target Complaints (TC) Discomfort and Change Scale (Battle et for individuals older than the age of 18 who had been emotionally
al., 1968). The TC Discomfort and Change Scale asks clients to injured by a significant other. The advertisements also indicated
specify three problems they would like to see change as a result of that the injury should have occurred at least 2 years prior to the
treatment. Clients were asked to rate each problem at three points time that the individuals responded to the advertisement and that
in time (pretreatment, posttreatment, and 3-month follow-up) in respondents should still be experiencing some lingering feelings of
terms of how distressed they were by the problem. In addition, at hurt or anger toward the other in regard to the injury.
two points in time (posttreatment and 3-month follow-up), clients Eighty-six participants who called in response to the advertise-
were asked to rate how much they felt the problem had changed ments were first briefly interviewed over the telephone to assess
since the beginning of treatment. Battle et al. (1968) reported high initial suitability regarding general inclusion and exclusion criteria
correlations with other outcome measures and test–retest reliability and to determine whether they were presenting with a specified
(r ⫽ .68) between pre- and posttreatment psychiatric interviews. unresolved target injury from at least 2 years earlier and were
Global Symptom Index (GSI) of the Symptom Checklist-90 — willing to participate in a videotaped research treatment. Fifty-nine
Revised (SCL-90 –R; Derogatis, 1983). The SCL-90 –R is a

suitable potential participants were invited to undergo a further

well-known instrument that measures general symptom distress, assessment process to ensure that the proposed treatment program
with high internal consistency (.77–.90) and test–retest reliability could meet their treatment needs. The initial 2-hr assessment
(.80 –.90) over a 1-week interval with people with a variety of interview was designed to obtain consent for assessment and
disorders (Derogatis, Rickels, & Roch, 1976). Calculations of treatment and to assess clients in terms of Axis I and Axis II
change on the GSI were used as an outcome measure at three disorders using the Semistructured Clinical Interview and Diagno-
points in time (pretreatment, posttreatment, and 3-month follow- sis protocol (SCID; Spitzer, Williams, Gibbon, & First, 1990).
up). Cronbach’s alpha for GSI in this sample was .82 Clients also completed the BDI and the SCL-90 –R at the first
Beck Depression Inventory (BDI; Beck, Ward, Mendelson, assessment appointment. If this initial assessment supported the
Mock, & Erbaugh, 1961). This 21-item inventory is widely used appropriateness of the proposed treatment for the client, he or she
to assess depression. It has high internal consistency and correlates was invited to participate in the treatment study and asked to come
highly with other self-report measures of depression and with in to complete the rest of the pretreatment measures. Forty-six
clinicians’ ratings of depression (r ⫽ .60 –.90; Beck, Steer, & clients were successfully assigned to treatment.
Garbin, 1988). Cronbach’s alpha for BDI in this sample was .87. Participants were randomly assigned to either the PG or EFT
treatment. Clients in the individual EFT treatment were seen for
Process Measures twelve 1-hr weekly individual therapy sessions. Clients in the PG
Working Alliance Inventory (WAI; Horvath & Greenberg, group were seen for six 2-hr sessions that were held biweekly for
1989). 12 weeks. Three PG groups (ns ⫽ 9, 8, and 6) were tested for a
The WAI is a 36-item scale rated on a 7-point Likert scale. The total of 23 participants. In the EFT treatment, one therapist saw 5
WAI is made up of three alliance subscales that assess the clients, one saw 4, two saw 3, and four saw 2 each. Sessions in the
therapist– client bond and agreement on therapy tasks and goals. PG group were audiotaped only, whereas sessions in the EFT
Internal consistency is high for the whole scale (.87–.93) as well as treatment were both audiotaped and videotaped. Clients in both
the subscales (.89 –.92) (Horvath & Greenberg, 1989). Cronbach’s groups completed questionnaires following each treatment session
alpha for WAI was .89. and were assigned homework at approximately the same time
Emotional Arousal Session Report Measure (Warwar & Green- intervals throughout the course of treatment. Clients completed
berg, 2002). The Emotional Arousal Session Report Measure outcome measures at pretreatment, at termination of treatment,
was developed to evaluate the intensity of emotional arousal during a posttherapy interview, and at a 3-month follow-up inter-
experienced during therapy. This measure consists of 18 emotion view. A short form of the WAI was given to the PG clients
items, which the client rates on a 7-point Likert scale (1 ⫽ not at following Session 1 and to the individual EFT clients after Session
all, 5 ⫽ moderately, 7 ⫽ very much). Clients are instructed to 3. Therapists completed a postsession questionnaire after each
indicate the degree to which they felt each of the 18 emotions session.
during their session. To account for the possibility that an emotion
that clients believed to be vital in describing their emotional Results
experience was absent from the 18 emotion categories, we added
an additional item (Question 19), which gives the client an oppor- Adherence to the therapy treatment manual in the EFT treatment
tunity to rate the intensity of any other emotion they may have felt. was monitored via therapist and supervisor reports. Adherence to
The Emotional Arousal Session Report Measure was completed by the PG group manual was monitored only via therapist reports.
clients after each session. This measure was constructed as a Therapists reported on a 5-point scale (ranging from not at all to
self-report form of the observer Emotional Arousal Scale (Warwar completely) the degree to which they judged themselves to have
& Greenberg, 1999), which has been shown to predict outcome. adhered to the treatment protocol, and in the individual therapy
condition, they reported whether they had used chair dialogues
during the session. The supervisor, using the same 5-point scale,
Procedure
reported an adherence judgment from viewing a videotape of at
Clients were recruited through advertisements in local commu- least four of the individual therapy sessions. Mean rating (with
nity newspapers and flyers distributed to the university and general standard deviation) of therapist reports for the individual treatment
over the sessions was 4.23 (SD ⫽ 1.11), and mean supervisor and d ⫽0.71 for PG. This shows that both treatments led practi-
rating was 4.14 (SD ⫽ 0.52), indicating good adherence. The cally to a large amount of change and that there was a meaningful
therapists reported a mean of 5.13 chair dialogues per treatment difference between groups on the major measure of forgiveness.
with a range of 4 –7 per client. The PG treatment leader reported Pre–posttreatment effects on the GSI were d ⫽ 0.62 for the EFT
successful implementation of the group manual for each session. treatment and negligible for the PG treatment. The between-
Given that clients were gathered in groups (of 9, 8, and 6) in the treatment effect size was d ⫽ 0.66, again showing large effects for
three PG groups and that each therapist saw a different number of EFT over PG. The pre–posttreatment effects on empathy and
clients, there was a nonindependence in the data that could have acceptance (UFB EA) were 1.73 and 0.74 for EFT and PG,
introduced a statistical bias in the analyses (Kenny, 1995). Thus, respectively, and the between-treatment effect size was 0.98. For
we investigated treatment outcome scores for bias by looking at feeling and needs (UFB FN), the pre–posttreatment effects were
intraclass correlations to measure homogeneity within groups in 3.22 and 1.62 for EFT and PG, respectively, with a between-
relation to variation between groups, and the F statistics used in the treatment effects size of 1.62. The pre–posttreatment effects on TC
analyses of variance were adjusted if the correlation was larger Discomfort were 3.84 and 2.27 for EFT and PG, respectively, and
than .10. For the analyses that did not compare treatments, effects the between-treatment effect was 1.56. All these effects are very
large. The between-treatment effects on BDI were very small, d ⫽

(the importance of ignoring statistical dependencies among obser-

vations) are likely to be relatively minor, and so in these cases 0.06, and the pre–posttreatment effects of 0.47 and 0.69 for EFT
dependence was ignored. and PG, respectively, although large, were much smaller than on
Correlations were conducted between all pretreatment vari- other measures.
ables.1 The pretreatment BDI did not correlate with any of the
other symptom measures at pretreatment. However, there were Forgiving and Letting Go
positive correlations between the EFI and the Forgiveness Measure
Cutoff scores were used on the relevant outcome measures to
(r ⫽ .35, p ⬍ .05) and between the EFI and the UFB EA (r ⫽ .56,
classify clients according to whether they forgave their injurers or
p ⬍ .001). There were no significant differences at the .05 level
let go of their negative feelings toward the injurer (Level 4 or
between treatments on any of the pretreatment, demographic, or
above on each measure). The classification of clients on this
other assessment variables examined using one-way analyses of
dichotomous classification is shown in Table 3. Dichotomy on
variance (ANOVAs). A t test that we conducted to evaluate
these single-item measures, however, is suggestive rather than
whether therapeutic alliances were different in the two treatments
definitive because the dividing lines between forgiving and not
showed that clients’ early working alliances were not significantly
forgiving and letting go and not letting go may not be that sharp.
different at the .05 level (EFT individual therapy, M ⫽ 5.89, SD ⫽
0.84; PG group, M ⫽ 5.75, SD ⫽ 0.78). 1
Readers can request a complete matrix of intercorrelations between all
variables by writing to Leslie S. Greenberg at [email protected].
Outcome 2
In order to test for possible dependence in the data, we computed
intraclass correlations (ICCs) on the output from two separate one-way
In order to test the hypothesis that EFT would be more effective
ANOVAs on the effect of the therapists in the EFT condition and of the
than PG, we performed repeated measures analyses of covariance. three groups in the PG condition. An ICC measures the relative homoge-
Pretest scores of the dependent measure were used as the covariate neity within groups in comparison to the between-group variation. ICC is
with posttreatment and follow-up scores as the repeated measures large and positive when there is no variation within the groups but group
and with the type of therapy received (either PG or individual means differ. It is at its largest negative value when group means are the
EFT) as the between-groups factor. TC Change, which does not same but there is great variation within groups. A negative ICC occurs
have a measure at pretreatment, was analyzed by a repeated when between-group variation is less than within-group variation. In this
measure ANOVA, with posttreatment and follow-up as the two situation, checking for independence, we would want the ICC to be close
occasions. The means for each of the measures for pretherapy, to 0 or negative, indicating that group mean differences are negligible
posttherapy, and follow-up are provided in Table 2. There were no relative to individual differences (i.e., the grouping does not matter). ICC
recently has been used in the context of hierarchical linear modeling
significant Group ⫻ Time interactions on any of the measures. The
(HLM) to measure the extent to which data clustering (i.e., nonindepen-
groups were significantly different on almost all measures such dence of observations) is present. If the ICC is near 0, then HLM gives very
that the EFT groups exhibited the highest levels of forgiveness- similar results to a traditional ANOVA or regression that assumes inde-
related gains (i.e. EFI, Forgiveness Measure, UFB EA, UFB FN, pendence (Raudenbush & Bryk, 2002; Shrout & Fleiss, 1979).
and Letting Go Measure) as well as the greatest levels of symptom ICCs at posttest were found to be low— below .08 or negative for both
reduction (i.e., TC Discomfort, TC Change, and GSI)2. However, treatment conditions for all of the dependent variables except for the
the differences for the BDI were not statistically significant. Unfinished Business Feelings and Needs Scale (UNB FN) in the PG group
(ICC ⫽.15). This provides evidence that the grouping of clients within
therapists and within groups was not a strong factor influencing the
Clinical Significance and Effect Size findings on the majority of measures in the study except possibly for the
UNB FN. Fs were adjusted according to the design effect, which is a
Cohen’s d, the standardized mean difference, which provides an
function of ICC and sample size (Skinner, Holt, & Smith, 1989) for posttest
index of the practical as well as the statistical significance of the scores on the UNB FN, and no significant difference between groups was
differences between treatments and between pre- and posttreat- found at posttest, F(1, 43) ⫽ 2.65, p ⫽ .112, thereby altering the conclu-
ment, was calculated. The effect size for differences between sion of difference on this measure. There was no difference in the between-
treatments on the EFI was moderate, d ⫽ 0.41; pre–posttreatment group comparisons on the other measures in which ICC had been lower
effect sizes were large for both treatments, with d ⫽ 1.0 for EFT than .08.
Table 2
Means and Standard Deviations of Outcome Measures at Pretreatment, Posttreatment, and Follow-Up
Prettreatment Posttreatment Follow-Up

Analysis of
EFT PG EFT PG EFT PG Covariance
Measure M SD M SD M SD M SD M SD M SD F(1, 43)
Enright Forgiveness
Inventory 199.22 60.58 197.09 58.14 261.13 47.24 237.30 51.28 277.52 51.41 243.43 55.01 4.98*
Forgiveness Measure 2.18 0.72 2.39 0.78 3.74 0.81 2.87 1.06 3.83 0.75 3.43 1.01 7.91**
Unfinished Business Scale
Empathy and Acceptance 14.04 4.43 14.26 4.97 22.17 5.42 17.78 5.48 22.13 4.38 18.87 4.78 9.28**
Feelings and Needsa 14.87 5.06 17.48 4.63 30.95 6.00 25.58 8.35 31.32 6.56 28.05 7.15 4.32*
Letting Go Measure 2.00 0.71 2.17 0.94 4.09** 0.79 2.17 0.98 4.06 0.71 3.42 0.76 437.69**
Target Complaints Scale

Discomfort 10.48 1.53 10.07 1.64 4.40 2.25 6.48 3.03 4.61 2.41 6.22 2.56 7.03*
Change — — — — 7.16 1.64 5.39 1.99 7.70 1.29 6.36 1.91 12.67**
Global Symptom Index .70 .43 .58 .36 .46* .47 .60 .45 .35 .31 .45 .32 6.73*
Beck Depression Inventory 12.18 10.32 10.79 6.34 7.26 7.59 6.39 5.88 6.79 7.13 5.47 5.43 1.1
Note. Pretreatment scores were used as covariates. EFT ⫽ emotion-focused therapy; PG ⫽ psychoeducation group.
a
See Footnote 2.
*
p ⬍ .05. ** p ⬍ .01.
All of the 9 clients in EFT who forgave the injurer also let go of Test of Emotional Arousal in Groups
their negative feelings, as did the 4 people in the PG treatment who
forgave the injurer. However, 5 people in the EFT treatment and 3 We compared clients in each of the treatment modalities on the
in the PG treatment let go but did not forgive. The result of a intensity of self-reported emotional arousal they experienced in the
chi-square analysis of the distribution comparing forgivers and session to test for the hypothesized difference between treatments
those who let go was significant, ␹2(l, N ⫽ 46) ⫽ 21.5, p ⬍ .001, in reported emotional arousal. The emotional arousal measure was
showing that 100% of those who forgave let go, while 38% of divided into positive and negative or unpleasant emotion clusters
those who let go did not forgive. This suggests that letting go may for the following periods: total duration of treatment and three
be a necessary requirement of forgiveness (i.e., everyone who phases of therapy (early, middle, and late). The positive emotion
forgave also let go), but it is not by itself sufficient for, nor is it the cluster included happy and content, while the negative cluster
same thing as, forgiveness, since clients who met the criteria for included sad, angry, afraid, and in pain.
letting go did not always consider themselves to have forgiven the Changes in reported intensity of negative/unpleasant and pleas-
injurer. It is important to note that in both groups, a large propor- ant in-session emotional arousal across three phases of each treat-
tion of those people who were classified as not forgiving or letting ment were examined. For treatment comparisons, the EFT sessions
go rated themselves as in progress (Level 3) on these tasks, so it were grouped to form early (three sessions), middle (five sessions),
was not the case that they did not benefit from treatment. They and late phases (last two sessions). This structure was thought to
simply had not yet reached as full a resolution, according to our best reflect the phase structure of the treatment. The six group
cutoff criteria, as those who indicated they had more fully forgiven sessions were broken into three sets of two sessions each. A 2 ⫻
or let go. 3 ⫻ 2 repeated measures ANOVA was used to compare reported
in-session intensity of positive and negative emotions in the two
treatment conditions overall and in each phase of therapy. Means
and standard deviations are shown in Table 4, and the graph over
Table 3 phases is given in Figure 1.
Frequency of Forgiveness and Letting Go of Negative Feelings There was a significant Emotion ⫻ Time interaction, F(2, 43)
⫽12.853, p ⬍ .001; a significant Emotion ⫻ Group interaction,
No. of clients who let go of negative feelings
F(1, 44) ⫽ 24.6, p ⬍ .001; and a significant three-way Emotion ⫻
Emotion- Psychoeducation Time ⫻ Group interaction, F(2, 43) ⫽ 3.29, p ⬍ .05. Post hoc
focused therapy group % of total analyses using a Bonferroni adjustment for multiple comparisons
Forgiveness found the groups differed significantly, p ⬍ .001, on the amount of
achieved Yes No Yes No Yes No
reported negative affect arousal overall in the treatment but that
Yes 9 0 4 0 28.26 0 there were no significant differences at the .05 level between the
No 5 9 3 16 17.39 54.34 treatment in self-reports of overall positive emotion experienced
in-session over all. This finding acted as a form of implementation
Total % 30.43 19.56 15.21 34.78 45.65 54.34
check showing that the individual EFT treatment was effective in
Note. A chi-square analysis of the distribution comparing forgivers and arousing more unpleasant emotion. The clients in the EFT treat-
those who let go was significant, ␹2(l, N ⫽ 46) ⫽ 21.5 p ⬍ .001. ment reported significantly higher levels of negative/unpleasant
Table 4 posttreatment changes as the standardized residuals by regressing

In-Session Emotion Clusters by Phase of Therapy for Clients initial scores onto final scores for all the symptom and outcome
Receiving Psychoeducation Group Treatment and Emotion- measures (Cronbach & Furby, 1970). As can be seen in the table,
Focused Therapy an increase in forgiveness, in empathy and acceptance, and in
feelings and needs all were related to a decrease in overall symp-
Phase of Therapy tom distress
Early Middle Late
Discussion
Emotion PG EFT PG EFT PG EFT
The results of this study show that clients in the individual EFT
Positive treatment experienced significantly more improvement than those
M 3.03 2.97 3.62** 2.73 4.17 3.82 in the PG treatment on all measures of forgiveness, on measures
SD 1.17 1.16 1.26 1.23 1.53 1.44
Negative assessing the degree to which clients had let go of distressing
M 2.82 3.58* 2.67 3.50** 2.11 2.60 feelings and unmet needs in relation to the injurer, and on target
SD 1.40 0.91 1.12 0.98 1.04 1.09 complaints at termination and follow-up. The two groups, how-
ever, were found to not differ significantly on the change in

Note. N ⫽ 23 for each treatment group. Asterisks indicate between group
feelings and needs at termination, when possible dependence in the
differences. PG ⫽ psychoeducation group; EFT ⫽ emotion-focused ther-
apy. PG group was taken into account. Greater improvement also was
*
p ⬍ .05. ** p ⬍ .01. reported in the EFT condition on the GSI of the SCL-90 –R.
However, there was no difference between groups on the BDI, but
given that the clients in this study were not depressed, this finding is
not surprising. The significant change in general psychological symp-
emotional arousal than clients in the PG treatment in the initial
phase of treatment, p ⬍ .01, and in the middle phase of therapy,
p ⬍ .01. The clients in the PG treatment reported significantly
higher levels of positive emotional arousal than those in the EFT 4.5
treatment in the middle phase of therapy, p ⬍ .01. Finally, during
the late phase of therapy, there was no significant difference at the
.05 level between the groups in reported intensity of negative/
unpleasant or positive emotional arousal. 4.0
The EFT clients’ reports of in-session negative/unpleasant emo-
tional arousal were found to be relatively stable over the first two
phases, showing no significant differences at the .05 level between
Means
early and middle phases. However, the EFT clients reported sig-
nificantly less in-session negative/unpleasant emotional arousal in 3.5
the final phase of therapy compared with that reported in the
middle phase, p ⬍ .05. Analysis of the PG clients’ reports followed
a similar pattern. Differences between early and middle phases of
treatment were not significant at the .05 level. However, a signif-
icant decline in reported negative emotional arousal from the 3.0
middle to the late phase of therapy was present, p ⬍ .05.
In an examination of the pattern of reported positive in-session
emotional arousal, EFT clients reported a significant increase in
the intensity of positive emotional arousal from the middle phase
2.5
to the final phase of treatment, p ⬍ .001. In the PG clients, a trend
of increasing positive emotional arousal from the early to late
phases of therapy was found. The PG clients’ ratings of positive
emotions increased significantly from the early to middle treat- 2.0
ment phase, p ⬍ .01, and from the middle to late phase, p ⬍ .05. Early Middle Late
Degree of reported in-session intensity of emotional arousal aver-
aged over the whole treatment, or averaged over any of the phases Phase of Therapy
of treatment, did not correlate significantly with change on any of
the outcome measures for the combined sample or for either PG
treatment. EFT
Negative Emotion
Correlational Analyses Positive Emotion
Correlations calculated between residual gains in the

forgiveness- and letting-go-related variables, and symptom vari- Figure 1. Emotional arousal across phases. PG ⫽ psychoeducation
ables are displayed in Table 5. We calculated the pre– group; EFT ⫽ emotion-focused therapy.
Table 5 initial assessment correlated significantly with symptom distress

Correlations Between Pre–Posttreatment Residual Gain Scores approximately 9 months later (after controlling for symptoms
on Forgiveness, Letting Go and Symptom Variables distress at initial assessment) and that the length of time since the
offence mediated the relationship between forgiveness and symp-
Posttreatment measure tom distress, suggesting a complex relationship between forgive-
Global Beck ness and symptoms. Data on our treatment-seeking population do
Symptom Depression not support the idea that forgiveness correlates with distress at
Pretreatment measure Index Index initial assessment but do show that an increase in forgiveness
certainly correlates with a reduction in distress.
Enright Forgiveness Inventory ⫺.41** ⫺.08
Forgiveness Measure ⫺.36* ⫺.22 In addition to clients showing better outcomes, the EFT treat-
Unfinished Business Scale ment was found to involve more client-reported emotional arousal,
Empathy and Acceptance ⫺.37* .05 especially of negative emotions in the mid phase of treatment,
Feelings and Needs ⫺.38** ⫺.29 confirming that the EFT treatment was more emotionally arousing
Letting Go Measure ⫺.23 ⫺.17
Target Complaints Scale—Discomfort .27 .14 than the PG treatment. The finding that there was more negative
emotional arousal in the early phase of therapy in the EFT treat-

Note. N ⫽ 46. ment than in the PG treatment was probably because EFT clients
*
p ⬍ .05. ** p ⬍ .01. were more likely to access negative emotions right from the start
as a function of therapy, rather than because they entered treatment
with more negative emotions than clients in the PG treatment. This
toms on the SCL-90 –R in the EFT clients, with no change in the PG finding also suggests that the emotional change processes in ther-
clients on symptoms, is notable, suggesting that an individual apy are not necessarily simply ones of replacing bad feelings with
emotion-focused treatment in addition to enhancing forgiveness and good feelings in a linear process but that at times feeling bad can
letting go of the emotional injury, has positive effects on a person’s in fact lead to feeling good. Thus, working through bad feelings by
general level of well-being that exceed that of a psychoeducation facing them—allowing and accepting them— can lead to change
group. Taken as a whole, this study provides support for the differ-
(Greenberg, 2002). It should be noted, however, that reported
ential effectiveness of an EFT approach for promoting forgiveness,
in-session emotional arousal did not relate to outcome in either
resolving emotional injuries, and reducing general symptoms over a
group. This is likely because not all arousal of emotion is the same.
PG treatment intervention of the same duration.
For example, some arousal may be a sign of distress rather than a
In addition, all those clients who reported that they forgave the
sign of working through distress (Greenberg & Watson, 2006).
injurer, regardless of group, also indicated that they had let go of
Recently, Greenberg, Auszra, and Herrmann (2007) showed that
the distressing feelings and unmet needs previously associated
arousal alone is not necessarily a measure of productive emotional
with the injury. In contrast, some individuals indicated that they
processing and that it is productive processing of aroused emotion
had let go of the distressing feelings and unmet needs associated
that best discriminates good from poor outcomes. In other studies,
with the injury but had not forgiven the injurer. This finding
the intensity of observed expressed emotional arousal, however,
suggests that letting go of persistent unresolved feelings such as
anger, grief, sadness, or hurt may be a necessary step in resolving has been shown to predict outcome in the working phase of the
past interpersonal hurtfulness but may not be sufficient for, nor EFT treatment of depression (Missirlian, Toukmanian, Warwar, &
equivalent to, forgiveness. Further research on this question is Greenberg, 2005; Warwar, 2003). The difference in the finding on
needed, but viewing the process of reducing unforgiveness or the relationship between emotional arousal and outcome in this
letting go of bad feelings as distinct from the process of generating study and the finding on the relationship between the two in the
positive feelings, such as loving kindness, may be useful when one studies on depression (if not attributable to differences between a
is examining the process of forgiveness and the possibility of the depressed and an emotional injury population) may be explained
resolution of an emotional injury in the absence of forgiveness. by the differences between the nature of the experience of emotion
Contrary to what has been suggested in the forgiveness literature as measured by postsession self-reports of emotional arousal ex-
(McCullough & Witvliet, 2001; Orcutt, 2006), forgiveness was not perienced in the session and the expression of emotions captured
initially correlated with the pretreatment variables measuring emo- by observational measures of arousal. In the present study, we
tional health such as the BDI or (particularly) the GSI of the observed that in the sessions in which some clients reported having
SCL-90 –R. Thus, clients who at the start of treatment were more felt high emotional intensity (experienced emotion), those clients
forgiving were not psychologically healthier. This finding, how- showed few visible signs of the reported arousal (expression), and
ever, may have been due to lack of range on the Forgiveness in fact often the clients were quite constricted in their expression
Measure as well as on the GSI and BDI and due to the fact that of emotion in the session. They appeared to have felt a great deal
participants were seeking help for presenting problems involving a of anger or sadness but did not necessarily express it. Therefore,
lack of letting go and forgiveness. In our study, pre–posttreatment although the self-reports were indicative of emotion experienced,
increases in forgiveness and letting go variables were related to a they were not indices of how much emotion was allowed and
decrease in GSI. Thus, it seems that increases during treatment in expressed in the session and made accessible for further process-
forgiveness and in empathy and acceptance toward the injurer and ing. As Greenberg et al. (2007) has suggested, emotion awareness
the ability to resolve feelings and needs were related to improved and expression are different emotion change processes, and the
health as measured by the GSI, suggesting that there are health latter— by revealing the self to the other, overcoming constriction,
benefits to resolving injuries both by forgiveness and by letting go. and altering physiology and neurochemistry—may be what is most
Recently Orcutt (2006) found that offence-specific forgiveness at therapeutic when one is dealing with unresolved painful emotions.
Our study had some limitations. The clients in this study were Berecz, J. M. (2001). All that glitters is not gold: Bad forgiveness in
volunteers who responded to advertisements and therefore may not counseling and preaching. Pastoral Psychology, 49, 253–275.
have been representative of the general population seeking help for Bolger, E. (1999). Grounded theory analysis of emotional pain. Psycho-
emotional injuries. Furthermore, equivalence of type and intensity therapy Research, 9, 342–362.
of injuries was not taken into consideration in assigning clients to Boss, J. (1997). Throwing pearls to the swine: Women, forgiveness, and
groups because we assumed that this would be handled by ran- the unrepentant abuser. In L. D. Kaplan & L. F. Bove (Eds.) Philosoph-
ical perspectives on power and domination: Theories and practices (pp.
domization. However, in small samples such as this, this may not
235–247). Atlanta, GA: Rodopi.
be the case. In addition, resolution or forgiveness was not assessed
Coyle, C. T., & Enright, R. D. (1997). Forgiveness intervention with
in relation to whether the injured person still had an ongoing postabortion men. Journal of Consulting and Clinical Psychology, 65,
relationship with the injurer (as in the case with a living parent vs. 1042–1046.
a deceased parent). It also would have been preferable to ensure a Cronbach, L. J., & Furby, L. (1970). How we should measure
balance of men and women in both groups. To the extent that there “change”— or should we? Psychological Bulletin, 74, 68 – 80.
were more women in EFT, this may have introduced a confound. Davenport, D. S. (1991). The functions of anger and forgiveness: Guide-
Taking all these factors into consideration will require further
lines for psychotherapy with victims. Psychotherapy, 28, 140 –144.

research with a larger sample of clients. Derogatis, L. R. (1983). SCL-90 –R administration, scoring, and proce-
Because therapist and researcher allegiance to treatment model dures manual for the revised version. Baltimore: Clinical Psychiatric
(and their possible ability to persuade clients of the relevance of Research.
the treatment process to their presenting problems) is a common Derogatis, L. R., Rickels, K., & Roch, A. F. (1976). The SCL–90 and the
factor known to improve outcomes regardless of specific treatment MMPI: A step in the validation of a new self-report scale. British
models used, one might attribute observed differences in this study Journal of Psychiatry, 128, 280 –289.
to therapist training/allegiance or to researcher allegiance, rather Elliott, R., Watson, J., Goldman, R., & Greenberg, L. (2004). Learning
emotion-focused therapy. Washington, DC: American Psychological
than to differences in specific treatment techniques. The EFT
Association.
therapists had at least 1 year of experience with EFT and had
Enright, R. D., & Eastin, D. L. (1992). Interpersonal forgiveness within the
received an additional 30 hr of training in this specific intervention helping professions: An attempt to resolve differences of opinion. Coun-
prior to participation in the study. Similar training did not occur for seling and Values, 36, 84 –103.
the PG treatment leaders although the one leader was the developer Enright, R. D., & Fitzgibbons, R. P. (2000). Helping clients forgive: An
of the psychoeducation group treatment. Investigators were pro- empirical guide for resolving anger and restoring hope. Washington,
ponents of EFT, and this too could have influenced results. DC: American Psychological Association.
In addition, differences between factors operating in group and Enright, R. D., Rique, J., & Coyle, C. T. (2000). The Enright Forgiveness
individual therapy modalities could account for the differences in Inventory (EFI) user’s manual. Madison, WI: International Forgiveness
effectiveness of treatment. Clients receiving individual therapy re- Institute.
ceive a treatment tailored to their needs, whereas in group treatment, Fincham, F. D. (2000). The kiss of the porcupines: From attributing
there is less flexibility and individualization. Spacing of sessions also responsibility to forgiving. Personal Relationships, 7, 1–23.
was different in the two treatments. Individual therapy sessions took Fitzgibbons, R. P. (1986). The cognitive and emotive uses of forgiveness
place weekly, whereas the group sessions were twice the length of in the treatment of anger. Psychotherapy, 23, 629 – 633.
individual sessions but met only once every 2 weeks. This could have Fredrickson, B. L. (1998). What good are positive emotions? Review of
General Psychology 2, 300 –319.
provided more continuity of treatment and support in the individual
Fredrickson, B., Mancuso, R., Branigan, C., & Tugade, M. (2000). The
therapy. On the other hand, group participants may benefit from other
undoing effect of positive emotions. Motivation and Emotion, 23, 237–258.
factors in groups that promote change that are not available to the Freedman, S. R., & Enright, R. D. (1996). Forgiveness as an intervention
individual participants, such as having two therapists, group support, goal with incest survivors. Journal of Consulting and Clinical Psychol-
and a sense of the universality of their problems. ogy, 64, 983–992.
Gendlin, E. T. (1996). Focusing-oriented psychotherapy: A manual of the
References experiential method. New York: Guilford Press.
Greenberg, L. S. (2002). Emotion-focused therapy: Coaching clients to
Akhtar, S. (2002). Forgiveness: Origins, dynamics, psychopathology, and
work through feelings. Washington, DC: American Psychological As-
technical relevance. Psychoanalytic Quarterly, 71, 175–212.
sociation.
Al-Mabuk, R. H., Enright, R. D., & Cardis, P. A. (1995). Forgiveness
Greenberg, L. S. (2004). Emotion-focused therapy. Clinical Psychology
education with parentally love– deprived late adolescents. Journal of
Moral Education, 24, 427– 444. and Psychotherapy, 11, 3–16.
American Psychiatric Association. (2000). Diagnostic and statistical man- Greenberg, L. S., Auszra, L., & Herrmann, I. R. (2007). The relationship
ual of mental disorders (4th ed., text rev.). Washington, DC: Author. among emotional productivity, emotional arousal, and outcome in ex-
Battle, C., Imber, S., Hoehn-Saric, R., Stone, A., Nash, C., & Frank, J. periential therapy of depression. Psychotherapy Research 17, 482– 493.
(1968). Target complaints as criteria of improvements. American Jour- Greenberg, L. S., Malcolm, W. M., & Warwar, S. H., (2002). Emotion-
nal of Psychotherapy, 20, 184 –192. focused therapy manual for resolving interpersonal emotional injuries.
Baures, M. M. (1996). Letting go of bitterness and hate. Journal of Unpublished manuscript, York University, Toronto, ON, Canada.
Humanistic Psychology, 36, 75–90. Greenberg, L. S., & Paivio, S. C. (1997). Working with emotions in
Beck, A. T., Steer, R. A., & Garbin, M. G. (1988). Psychometric properties psychotherapy. New York: Guilford Press.
of the Beck Depression Inventory: Twenty-five years of evaluation, Greenberg, L. S., Rice, L. N., & Elliot, R. (1993). Facilitating emotional
Clinical Psychology Review, 8, 77–100. change: The moment by moment process. New York: Guilford Press.
Beck, A. T., Ward, C. H., Mendelson, M., Mock, J., & Ebaugh, J. (1961). Greenberg, L. S., & Rosenberg, R. (2002). Therapist experience of empa-
An inventory for measuring depression. Archives of General Psychiatry, thy. In J. C. Watson, R. N. Goldman, & M. Warner (Eds.). Client-
4, 561–571. centred and experiential psychotherapy in the 21st century: Advances in
theory, research, and practice. Ross-on-Wye, Herefordshire, United focused group interventions to promote marital enrichment. Journal of
Kingdom: PCCS Books. Counseling and Development, 80, 452– 463.
Greenberg, L. S., & Watson, J. (1998). Experiential therapy of depression: Rowe, J. O., Halling, S., Davies, E., Leifer, M., Powers, D., & Van
Differential effects of client-centered relationship conditions and process Bronkhorst, J. (1989). The psychology of forgiving another: A dialogal
experiential interventions. Psychotherapy Research, 8, 210 –224. research approach. In R. S. Valle & S. Halling (Eds.), Existential–
Greenberg, L. S. & Watson, J. (2006). Emotion-focused therapy for de- phenomenological perspective in psychology: Exploring the breadth of
pression. Washington, DC: American Psychological Association. human experience. New York: Plenum Press.
Harris, A. H. S., & Thoresen, C. E. (2005). Forgiveness, unforgiveness, Shrout, P. E., & Fleiss, J. L. (1979). Intraclass correlations: Uses in
health, and disease. In E. L. Worthington, Jr. (Ed.), Handbook of assessing rater reliability. Psychological Bulletin, 86, 420 – 428.
forgiveness. (pp. 321–333). New York: Routledge. Singh, M. (1994). Validation of a measure of session outcome in the
Hebl, J. H., & Enright, R. D. (1993). Forgiveness as a psychotherapeutic resolution of unfinished business. Unpublished doctoral dissertation,
goal with elderly females. Psychotherapy, 30, 658 – 667. York University, Toronto, ON, Canada.
Hope, D. (1987). The healing paradox of forgiveness. Psychotherapy, 24, Skinner, C. J., Holt, D., & Smith, T. M. F. (1989). Analysis of complex
240 –244. surveys. West Sussex, England: Wiley.
Horvath, A., & Greenberg, L. S. (1989). Development and validation of the Smedes, L. B. (1984). Forgive and forget: Healing the hurts we don’t
Working Alliance Inventory, Journal of Counseling Psychology, 36, deserve. San Francisco: Harper San Francisco.
223–233. Spitzer, R. L., Williams, J. B., Gibbon, M., & First, M. B. (1990).
Karen, R. (2001). The forgiving self: The road from resentment to connec- Structured Clinical Interview for DSM–III–R Personality Disorders
tion. New York: Doubleday. (SCID–II). Washington, DC: American Psychiatric Press.
Kenny, D. (1995). The effect of nonindependence on significance testing in Wade, N. G., & Worthington, E. L., Jr. (2003). Overcoming interpersonal
dyadic research Personal Relationships, 2, 67–75. offenses: Is forgiveness the only way to deal with unforgiveness? Jour-
Klassen, D. (2001). The forgiveness workbook. Thunder Bay, ON, Canada: nal of Counseling & Development, 81, 343–353.
Listening Centre. Warwar, S. (2003). Relating emotional processes to outcome in experien-
Macaskill, A., Maltby, J., & Day, L. (2002). Forgiveness of self and others tial psychotherapy of depression. Unpublished doctoral dissertation,
and emotional empathy. Journal of Social Psychology, 142, 663– 665. York University, Toronto, ON, Canada.
Malcolm, W. M., Warwar, S. H., & Greenberg, L. S. (2005). Facilitating Warwar, S., & Greenberg, L. S. (1999). Client Emotional Arousal Scale–
forgiveness in individual therapy as an approach to resolving interper- III. Unpublished manuscript, York Psychotherapy Research Clinic, York
sonal injuries. In E. L. Worthington, Jr. (Ed.), Handbook of forgiveness. University, Toronto, ON, Canada.
(pp. 379 –391). New York: Routledge. Warwar, S., & Greenberg, L. (2002). Emotional Arousal Session Report
McCullough, M. E., Rachal, K. C., Sandage, S. J., & Worthington, E. L. Jr. Measure. Unpublished manuscript, York University, North York, ON,
(1997, August). A sustainable future for the psychology of forgiveness. Canada.
Paper presented at the meeting of the American Psychological Associ- Watson, J., & Greenberg, L. (1996). Pathways to change in the psycho-
ation, Chicago, IL. therapy of depression: Relating process to session change and outcome.
McCullough, M. E., & Witvliet, C. V. O. (2001). The psychology of Psychotherapy, 33, 262–274.
forgiveness. In C. R. Snyder and S. Lopez (Eds.), Handbook of positive Witvliet, C., Ludwig, T., & Vander Laan, K. (2001). Granting forgiveness
psychology (pp. 446 – 458). New York: Oxford. or harboring grudges: implications for emotion, physiology, and health.
McCullough, M. E., & Worthington, E. L., Jr. (1994). Encouraging clients Psychological Science, 12, 117–123.
to forgive people who have hurt them: Review, critique, and research Worthington, E. L., Jr. (1998). The pyramid model of forgiveness: Some
prospectus. Journal of Psychology and Theology, 22, 3–20. interdisciplinary speculations about unforgiveness and the promotion of
McCullough, M. E., Worthington, E. L., Jr., & Rachal, K. C. (1997). forgiveness. In E. L. Worthington, Jr. (Ed.), Dimensions of forgiveness:
Interpersonal forgiving in close relationships. Journal of Personality and Psychological research and theological perspectives (pp. 107–138).
Social Psychology, 73, 321–336. Philadelphia: Templeton Foundation Press.
Missirlian, T., Toukmanian, S., Warwar, S. &. Greenberg, L. (2005). Worthington, E. L., Jr. (2001). Five steps to forgiveness: The art and
Emotional arousal, client perceptual processing, and the Working Alli- science of forgiving. New York: Crown Publishing.
ance in Experiential Psychotherapy for Depression. Journal of Consult- Worthington, E. L., Jr. (2005). Initial questions about the art and science of
ing & Clinical Psychology. 73, 861– 871. forgiving. In E. L. Worthington, Jr. (Ed.), Handbook of forgiveness. (pp.
Orcutt, H. K. (2006). The prospective relationship of interpersonal forgive- 1–13). New York: Routledge.
ness and psychological distress symptoms among college women. Jour- Worthington, E. L., Jr., & Drinkard, D. T. (2000). Promoting reconciliation
nal of Counseling Psychology, 53, 350 –361. through psychoeducational and therapeutic interventions. Journal of
Paivio, S. C., & Greenberg, L. S. (1995). Resolving “unfinished business”: Marital and Family Therapy, 26, 93–101.
Efficacy of experiential therapy using empty-chair dialogue. Journal of Worthington, E. L., Jr., Sandage, S. J., & Berry, J. W. (2000). Group
Consulting and Clinical Psychology, 63, 419 – 425. interventions to promote forgiveness: What researchers and clinicians
Paivio, S. C., Hall, I. E., Holowaty, K. A. M., Jellis, J. B., & Tran, N. ought to know. In M. E. McCullough, K. I. Pargament, & C. E. Thoresen
(2001). Imaginal confrontation for resolving child abuse issues. Psycho- (Eds.), Forgiveness: Theory, research, and practice (pp. 228 –253). New
therapy Research, 11, 433– 453. York: Guilford Press.
Paivio, S. C., & Nieuwenhuis, J. A. (2001). Efficacy of emotionally Worthington, E. L., Jr., & Wade, N. G. (1999). The psychology of unfor-
focused therapy for adult survivors of child abuse: A preliminary study. giveness and forgiveness and implications for clinical practice. Journal
Journal of Traumatic Stress, 14, 115–134. of Social and Clinical Psychology, 18, 385– 418.
Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models:
Applications and data analysis methods (2nd ed.). Thousand Oaks, CA: Received January 16, 2007
Sage. Revision received November 29, 2007
Ripley, J., & Worthington, E. (2002). Hope-focused and forgiveness- Accepted November 29, 2007 䡲

Greenberg 2008

Uploaded by

Copyright:

Available Formats

Greenberg 2008

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Greenberg 2008

Uploaded by

Copyright:

Available Formats

Journal of Counseling Psychology Copyright 2008 by the American Psychological Association

2008, Vol. 55, No. 2, 185–196 0022-0167/08/$12.00 DOI: 10.1037/0022-0167.55.2.185

Differential Effects of Emotion-Focused Therapy and Psychoeducation in

Keywords: interpersonal injuries, forgiveness, letting go, emotion-focused therapy, psychoeducation

least 2 years prior to the start of therapy and continued to be

The UFB EA comprises six items and measures the extent to

homework assignments as the participants in individual therapy

Unfinished Business Feelings and Needs Scale (UFB FN).

Revised (SCL-90 –R; Derogatis, 1983). The SCL-90 –R is a

suitable potential participants were invited to undergo a further

large. The between-treatment effects on BDI were very small, d ⫽

(the importance of ignoring statistical dependencies among obser-

Prettreatment Posttreatment Follow-Up

Measure M SD M SD M SD M SD M SD M SD F(1, 43)

Target Complaints Scale

Table 4 posttreatment changes as the standardized residuals by regressing

ever, were found to not differ significantly on the change in

Correlations calculated between residual gains in the

Table 5 initial assessment correlated significantly with symptom distress

emotional arousal in the early phase of therapy in the EFT treat-

lines for psychotherapy with victims. Psychotherapy, 28, 140 –144.

You might also like