Psychology Journal Article Review: Foundational and Worldview Issues of Integration Assignment

JOURNAL OF COUNSELING & DEVELOPMENT • WINTER 2004 • VOLUME 82 79

A promising area of counseling research that emerged in the 1990s is the scientific investi- gation of forgiveness interventions. Although the notion of forgiving is ancient (Enright & the Human Development Study Group,

1991), it has not been systematically studied until relatively recently (Enright, Santos, & Al-Mabuk, 1989). Significant to counseling because of its interpersonal nature, forgive- ness issues are relevant to the contexts of marriage and dating relationships, parent–child relationships, friendships, professional relationships, and others. In addition, forgive- ness is integral to emotional constructs such as anger. As forgiveness therapies (Ferch, 1998; Fitzgibbons, 1986) and the empirical study of these therapies (Freedman & Enright, 1996) begin to unfold, it is important to ask if these inter- ventions can consistently demonstrate salient positive ef- fects on levels of forgiveness and on the mental health of targeted clients.

The purpose of this article is to analyze via meta-analysis the existing published interventions on forgiveness. Meta- analysis is a popular vehicle of synthesizing results across multiple studies. Recent successful uses of this method in- clude the study by McCullough (1999), who analyzed five studies that compared the efficacy for depression of stan- dard approaches with counseling with religion-accommo- dative approaches. Furthermore, in order to reach conclu- sions about the influence of hypnotherapy on treatment for clients with obesity, Allison and Faith (1996) used meta- analysis to examine six studies that compared the efficacy of using cognitive-behavioral therapy (CBT) alone with the use of CBT combined with hypnotherapy. Finally, Morris, Audet, Angelillo, Chalmers, and Mosteller (1992) used meta- analysis to combine the results of 10 studies with contra- dictory findings to show that the benefits of chlorinating drinking water far outweighed the risks. Although there may

be some concern that using forgiveness as an intervention in counseling is in too early a stage of development and that too few studies exist for a proper meta-analysis, the effec- tiveness of these recent meta-analyses supports this meta- analytic investigation. Certainly any findings must be tem- pered with due caution. However, this analysis may serve as important guidance for the structure and development of future counseling studies of forgiveness.

We first examine the early work in forgiveness interven- tions by examining the early case studies. From there, we define forgiveness, discuss the models of forgiveness in coun- seling and the empirically based interventions, and then turn to the meta-analysis.

EARLY CASE STUDIES

The early clinical case studies suggested that forgiveness might be helpful for people who have experienced deep emotional pain because of unjust treatment. For example, Hunter (1978) reported on Harriet, a 25-year-old woman with acute emo- tional distress. Harriet’s mother frequently condemned her daughter for the slightest deviation from her unreasonably high standards. Harriet’s anger toward her mother eventually led to symptoms of anxiety and depression. In addition, she started showing such externalizing symptoms as excessive anger and frustration directed at family members. With Hunter’s help, Harriet came to understand how she was reacting to her own victimization by victimizing others. In counseling, she was able to see her parents as capable of both good and bad behaviors. Forgiving her parents allowed her to take greater responsibil- ity for her own behavior; she did not have to belittle others. Forgiving her parents allowed Harriet to experience a greater self-acceptance and to establish meaningful friendships.

One of Kaufman’s (1986) early case studies involved Uri, an Israeli army officer in his 40s, who came to counseling

Thomas W. Baskin, Department of Counseling Psychology, University of Wisconsin–Madison; Robert D. Enright, Department of Educational Psychology, University of Wisconsin–Madison. The authors are grateful to Bruce E. Wampold for his encouragement and statistical insights. Correspondence concerning this article should be addressed to Thomas W. Baskin, Department of Counseling Psychology, 321 Education Building—1000 Bascom Mall, University of Wisconsin–Madison, Madison, WI 53706 (e-mail: [email protected]).

Intervention Studies on Forgiveness: A Meta-Analysis

Thomas W. Baskin and Robert D. Enright

In this meta-analysis, 9 published studies (N = 330) that investigated the efficacy of forgiveness interventions within counseling were examined. After a review of theories of forgiveness, it was discovered that the studies could logically be grouped into 3 categories: decision-based, process-based group, and process-based individual interventions. When compared with control groups, for measures of forgiveness and other emotional health measures, the decision-based interventions showed no effect, the process-based group interventions showed significant effects, and the process-based individual interventions showed large effects. Consequently, effectiveness has been shown for use of forgiveness in clinical and other settings.

© 2004 by the American Counseling Association. All rights reserved. pp. 79–90

JOURNAL OF COUNSELING & DEVELOPMENT • WINTER 2004 • VOLUME 8280

B a s k i n a n d E n r i g h t

because of an inability to establish positive relationships with women. Through forgiveness counseling, Uri realized how much unconscious and deep anger he had toward his father, who died when he was young, and his mother, whom he blamed for the family’s subsequent poverty. Uri realized that he had not yet matured, displacing his anger onto his relationships with women and showing regressive, rebellious behavior similar to behavior in his adolescence. Forgiving his father for dying allowed Uri to symbolically bury his father. Forgiving his mother for not providing a higher stan- dard of living allowed him to leave behind the debilitating anxiety that had plagued him since childhood. As a result, Kaufman observed Uri growing in courage and accepting adult responsibility. He married and was able to reestablish a loving relationship with his mother.

From his clinical practice, Fitzgibbons (1986) reported that forgiveness counseling seemed to reduce anger, anxiety, and psychological depression in his clients. He observed that as people learn to forgive, they also learn to express anger in more appropriate ways, similar to the observations by Hunter (1978) and Kaufman (1986). At the time of these clinical reports, however, the observations had not been tested.

FORGIVENESS DEFINED

If forgiveness was to become part of the scientific study of counseling, then an accurate, comprehensive definition had to be established. Forgiveness has been defined as the will- ful giving up of resentment in the face of another’s (or oth- ers’) considerable injustice and responding with beneficence to the offender even though that offender has no right to the forgiver’s moral goodness (see, for example, Enright & the Human Development Study Group, 1991). Forgiveness is an act freely chosen by the forgiver.

Forgiveness is distinguished from condoning and excusing, rec- onciling, and forgetting. When someone condones or excuses, he or she realizes that there was no unfairness. If, for example, Jack takes Mary’s car to drive an injured child to the hospital, Mary, on realizing what had happened, would not forgive Jack, but excuse him under the circumstances. Reconciliation involves two people coming together again in mutual trust, whereas forgiveness is one person’s choice to abandon resentment and offer beneficence in the face of unfairness. One can forgive with- out reconciling. When one forgives, he or she rarely forgets the event. People tend to recall traumatic events, but on forgiving, a person may remember in new ways—not continuing to har- bor the deeply held anger.

COUNSELING MODELS

Counseling models and measures of forgiveness emerged out of the aforementioned or similar definitions of forgiveness. Three basic intervention models have been developed.

Model One

The first model, by Enright and the Human Development Study Group (1991), encompasses 20 processes or units

within four phases: Uncovering, Decision, Work, and Deep- ening. Over the years, the model has been refined from 17, then to 18, and finally to 20 units as seen in Table 1. The different interventions via this model have used slight varia- tions of the model’s units over the years.

Denton and Martin (1998) asked more than 100 clinical social workers their opinion about the way forgiveness therapy usually proceeds. The findings closely approximated the model described as follows. Also, Osterndorf, Hepp-Dax, Miller, and Enright (1999) reported on a study in which people hurt unfairly by another ordered the variables in Table 1 according to their own experience of forgiving. The

TABLE 1

Processes of Forgiving Another

Phase

Kiel, 1986

Trainer, 1981

Patton, 1985 Droll, 1984

Droll, 1984

Kiel, 1986

Close, 1970

Flanigan, 1987

North, 1987

Enright, 2001 Neblett, 1974

Smith, 1981 Cunningham, 1985

Droll, 1984 Bergin, 1988

Frankl, 1959

Cunningham, 1985

Enright, 2001

Enright, 2001

Smedes, 1984

Note. Table is an extension of Enright and the Human Study Group (1991). The references listed for each item are prototypical examples or discussions of that item.

Reference

Uncovering phase 1. Examination of psychological defenses 2. Confrontation of anger; the point is to

release, not harbor, the anger 3. Admittance of shame, when it is

appropriate 4. Awareness of cathexis 5. Awareness of cognitive rehearsal of

the offence 6. Insight that the injured party may be

comparing self with the injurer 7. Realization that oneself may be

permanently and adversely changed by the injury

8. Insight into a possibly altered “just world” view

Decision phase 9. A change of heart, conversion, new

insights that old resolution strategies are not working

10. Willingness to consider forgiveness as an option

11. Commitment to forgive the offender Work phase

12. Reframing, through role taking, who the wrongdoer is by viewing him or her in context

13. Empathy toward the offender 14. Awareness of compassion, as it

emerges, toward the offender 15. Acceptance, absorption of the pain

Outcome phase 16. Finding meaning for self and others in

the suffering and in the forgiveness process

17. Realization that self has needed others’ forgiveness in the past

18. Insight that one is not alone (universality, support)

19. Realization that self may have a new purpose in life because of the injury

20. Awareness of decreased negative affect and, perhaps, increased positive affect, if this begins to emerge, toward the injurer; awareness of internal, emotional release

15566676, 2004, 1, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/j.1556-6678.2004.tb00288.x by L

iberty U niversity, W

iley O nline L

ibrary on [13/05/2024]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

JOURNAL OF COUNSELING & DEVELOPMENT • WINTER 2004 • VOLUME 82 81

F o r g i v e n e s s M e t a - A n a l y s i s

participants’ average correlation between their own rankings of the process (from first experience to last in the forgive- ness process) correlated r = .79 with the theoretically es- tablished order.

The eight units of the Uncovering Phase assist the partici- pant to explore the injustice he or she has experienced, as- sess the amount of anger, and understand the ways in which harboring that anger may be clinically compromising the person. For example, in the first unit, the person examines the various psychological defenses he or she may use to protect against emotional pain (Kiel, 1986). Although such defenses may be adaptive in the short run, they need to be recognized if the person’s true emotion about the unfairness is to be confronted and understood. Prior to forgiving, a person usually needs to express the anger over a genuinely hurtful offense (Unit 2). In Unit 3, the person acknowledges and assesses the amount of guilt and shame that he or she has over the incident (Patton, 1985). Incest survivors, for example, oftentimes experience guilt over pleasurable physi- cal sensations that occurred when they were being victim- ized; the person needs to realize that such feelings are normal and in no way implicates the person as cooperating with the offense. Units 4 and 5 focus on the person’s tendency to at- tach much emotional energy to the offense and to ruminate excessively on it in an attempt to find a solution (Droll, 1984). At times, the person begins to compare his or her less fortunate state with what is perceived to be the offender’s more fortunate state (Unit 6) and to realize that he or she has been permanently changed by the event (Unit 7). Both can deepen the person’s anger and distress. The point of Units 6 and 7 is to assess the extent to which these thought patterns are occurring. Finally, people oftentimes conclude that based on all of the emotional pain experienced, life is unfair (Unit 8). The insights from uncovering the pain lead to the Deci- sion Phase in which the person rethinks past attempts to regu- late emotions and solve the problem (Unit 9), explores the meaning of forgiveness and the option of forgiveness in deal- ing with the problem (Unit 10), and commits to forgiveness (Unit 11).

The Work Phase encompasses four units: a set of thinking exercises to see the offender in a new light, or reframing who he or she is (Unit 12), stepping inside the offender’s shoes to emotionally experience his or her confusion, vul- nerability, or stress (Unit 13), which can increase a sense of compassion for the offender (Unit 14) and lead to what Bergin (1988) and others called bearing the pain (Unit 15). Here the forgiver gives a moral gift to the offender by not seeking revenge and by showing respect for him or her, not because of what was done but despite what was done.

Finally, the Deepening Phase includes such units as finding meaning in what was suffered (Unit 16), realizing that he or she is imperfect and in need of others’ forgiveness from time to time, garnering support for forgiving, sometimes finding a new purpose in life (helping others in similar situations), and experiencing emotional relief (Units 17–20). All counseling programs done with this model have incorporated manuals as guides to the interventions (Al-Mabuk, Enright, & Cardis,

1995; Coyle & Enright, 1997; Freedman & Enright, 1996; Hebl & Enright, 1993).

Model Two

A second intervention model was described by McCullough, Worthington, and Rachal (1997). Presented as a way of fos- tering both cognitive and affective empathy, the model out- lined by the authors has nine different components. First, the participants built rapport with the intervener, and sec- ond, each participant explored what the hurtful event was and what his or her reaction was to it. The exercises here were similar, at least to a degree, to the first eight units of the aforementioned model, except that all the exercises were sum- marized in part of one session. The third step involved under- standing empathy through the use of vignettes and discussion. Fourth was a didactic unit in which the leader described the link between being empathic toward an offender and eventu- ally forgiving that offender. Fifth, through written and verbal exercises, the participants practiced cognitive reframing and focused on the offender’s psychological state and general situ- ation in life (similar to Unit 12 of Table 1). Next, the respon- dents considered times in which they, themselves, needed other people’s forgiveness (Unit 17, Table 1). The analysis of attribu- tion errors followed in which the participants were encour- aged to see the offender’s behavior in terms of its situational determinants (Unit 12 again). Next, came an emphasis on the offender’s needs (Unit 13, Table 1) and how forgiveness may enhance the offender’s well-being (Unit 14, Table 1). Finally, constructs such as repentance and reconciliation were distin- guished from forgiveness (Unit 10, Table 1), and strategies for generalizing the learning were discussed (Unit 19, Table 1).

Model Three

The third model, by McCullough and Worthington (1995), was designed to elicit forgiveness in a 1-hour session by focusing empathically on the offender and writing letters (which were not sent) in which feelings were expressed to the offender. Given the brief nature of the intervention, the model introduced people to the idea of forgiveness and served as a forum to consider a decision to forgive.

In all three models, the participants in the interventions are asked to think about one person who has hurt them unfairly and to do the work of forgiveness relative to that particular person.

PROCESS VERSUS DECISION

These models highlight an important underlying philosophical difference. The first two models are process based, whereas the third is decision based. The philosopher Neblett (1974) argued that the essence of forgiveness is in the decision to forgive, along with the proclamation “I forgive you.” As the person decides to forgive and so proclaims, several important things happen. First, the forgiver has crossed an important line, so to speak. He or she has moved from a position of resent- ment to one of not letting the resentment dominate the

15566676, 2004, 1, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/j.1556-6678.2004.tb00288.x by L

iberty U niversity, W

iley O nline L

ibrary on [13/05/2024]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

JOURNAL OF COUNSELING & DEVELOPMENT • WINTER 2004 • VOLUME 8282

B a s k i n a n d E n r i g h t

interaction. Although the one who forgives may still feel resentful, the person chooses not to let it be a controlling factor. Second, the decision and proclamation show that the forgiver is consciously aware of his or her new position. The forgiver, in other words, is not abandoning resentment be- cause of taking some memory-loss pill or simply letting time run its course. Instead, the decision is a defining mo- ment regarding who the forgiver is (“I am one who forgives”), who the forgiven is (“He/she is worthy of respect”), and what their relationship may be like as a result of this deci- sion. The emphasis on forgiveness as a decision, then, cen- ters the construct in the cognitive domain.

In contrast, the philosopher North (1987) argued that for- giveness is a process, with the defining-moment decision embedded within it. Forgiveness develops from resentment and anger, through the decision, to the struggles to love and feel compassion toward a person who is difficult to love. This process can take time and effort (see Smedes’s, 1984, early writing on this). From this perspective, a decision by itself leaves one with only half a story to tell and therefore cannot qualify as forgiveness per se, although it is a vital part of that story.

A similar difference emerged in the early psychological literature. Worthington and DiBlasio (1990) published an intriguing article in which they outlined a decision-based counseling plan for forgiveness. The essence of the interven- tion is to have one forgiveness session with two people in which each person takes turns offering and granting forgive- ness, along with a commitment to atone for wrongdoing and a genuine attempt to do better in the future. There is a sense of process here in that much preparation occurs between the therapist and the pair prior to the session. Yet, as with Neblett (1974), the session, the decision to offer and re- ceive forgiveness, is the defining moment of the counseling.

Another insightful contribution was made by Ferch (1998). He outlined a method of intentional forgiving such that there is first a psychoeducational stage, which prepares clients to forgive, and then there is a face-to-face processing of for- giveness when appropriate. The choice to forgive is described as both a decision, with immediate opportunity, and as the opening of doors to a journey that encompasses an entire forgiving process.

Perhaps another way to look at this difference of decision- based versus process-based models is the contrast between an exclusively cognitive approach and one that includes a more extended cognitive and affective/empathy approach. Fitzgibbons (1986) became aware early that clients tend to first approach forgiveness cognitively—saying they forgive— before they feel like forgiving and offering empathy.

EMPIRICAL STUDIES

This meta-analysis seems to be the first quantitative assess- ment of all existing published forgiveness counseling inter- ventions. A review of the literature indicates that nine em- pirical studies with a quantitative measure of forgiveness have been published thus far.

Hebl and Enright (1993)

Hebl and Enright (1993) implemented the first counseling forgiveness intervention following a treatment model based on Enright and the Human Development Study Group (1991). Participants were elderly women with a mean age of 74.5 years who qualified in all of four conditions for par- ticipation: (a) The participant had something to forgive, (b) the participant felt emotionally hurt by what happened, (c) there was a definite person in mind to forgive, and (d) the participant was not going through a grieving process. All 26 participants were randomly assigned to a control group in- tervention versus a forgiveness group intervention. The for- giveness group intervention consisted of eight 1-hour weekly sessions. A prepared manual was used, which was based on a process model of forgiveness, and clients went through 17 units related to forgiving another person. All 13 partici- pants assigned to forgiveness group intervention completed it. The control group intervention also consisted of eight 1- hour weekly sessions. Participants determined topics that they would like to discuss with each other during the first session, and successive sessions consisted of talking through these issues. Eleven of 13 participants assigned to this group completed it. Both groups completed the Psychological Profile of Forgiveness Scale to measure forgiveness. Furthermore, the mental health constructs of self-esteem, state anxiety, trait anxiety, and depression were measured.

McCullough and Worthington (1995)

McCullough and Worthington (1995) studied two different brief psychoeducational group interventions on participants’ forgiveness for an offender and compared them with a wait- ing list control group. Participants were recruited from an undergraduate psychology class, and they qualified by re- porting that they had not committed severe offenses, such as incest, sexual abuse, and family strife, at early ages and by arriving for the intervention at an appointed time. The 86 participants were grouped as follows: 30 participants in an interpersonal group intervention, 35 participants in a self- enhancement group intervention, and 21 participants into a waiting list control condition. In the interpersonal inter- vention, the participants were divided into two groups and given a 1-hour intervention to encourage them to decide to forgive. The rationale given was the restoration of partici- pants’ relationships with the offenders and significant oth- ers. The self-enhancement intervention was the same as that of the interpersonal intervention, except that participants were encouraged to forgive because forgiveness was seen as providing physical and emotional benefits for the forgiver. All three groupings were given Wade’s (1989) Forgiveness Scale to measure their level of forgiveness. No other mental health constructs were measured.

Al-Mabuk et al. (1995)

Al-Mabuk et al. (1995) examined two interventions with parental-love-deprived college students. In each case, effects

15566676, 2004, 1, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/j.1556-6678.2004.tb00288.x by L

iberty U niversity, W

iley O nline L

ibrary on [13/05/2024]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

JOURNAL OF COUNSELING & DEVELOPMENT • WINTER 2004 • VOLUME 82 83

F o r g i v e n e s s M e t a - A n a l y s i s

were compared with a control group intervention. In Study 1 and Study 2, participants were randomly selected from college students who scored 1 standard deviation above the mean on a parental love-deprivation screening questionnaire for at least one parent. In the first study, 48 college students were randomly placed in experimental intervention and control intervention groups. In the second study, the same was done with 48 different participants. In Study 1, the ex- perimental intervention group received four 1-hour group sessions designed to take participants through the decision to forgive (similar to 11 of 20 units of forgiveness in Table 1). It was reasoned that this would be effective based on Neblett’s (1974) argument that one’s commitment to for- give is the crux of forgiveness. The control intervention met for the same amount of time, yet the sessions included dif- ferent material. The control intervention consisted of a hu- man relations program that focused on leadership, commu- nication, self-discovery, and perception. Forgiveness and parent-relations were absent from the curriculum. All 24 participants completed each of these groups. In Study 2, the experimental intervention group received six 1-hour ses- sions designed to take participants through all forgiveness units (similar to the 20 units in Table 1). Here participants were exposed to the entire theorized process of forgiveness. In addition, while Study 1 had two sessions per week, Study 2 had one session per week. The control intervention met for the same amount of time as the experimental group and had the same topics as the control intervention of Study 1, with the added topics of avoiding vagueness in communica- tion and of personal affirmations in rewarding others. All 24 participants completed the experimental intervention, while 21 of 24 had complete data for the control intervention.

Freedman and Enright (1996)

Freedman and Enright (1996) compared a forgiveness inter- vention with female incest survivors to a waiting list con- trol. Participants were 12 women recruited from a midwestern community who were sexually abused, involving contact when they were children, by a male relative. Also, the abuse must not have occurred within the previous 2 years, and participants needed to show signs of experiencing psycho- logical difficulties. Average age was 36 years (range = 24–54 years). The intervention group was given weekly individual counseling sessions for an average of 14.3 months. Sessions followed a process model, giving a complete set of 17 units (similar to those in Table 1). During each session, no more than one unit was covered, and the intervention would re- main on one unit until the client felt ready to move on to the next. All 6 participants completed this intervention. The wait-listed control group waited an average of 14 months before receiving the intervention. These participants had a small amount of monthly contact with the experimenter to maintain the connection, although the topic of forgiveness was never mentioned. After a matched member of the inter- vention group finished her treatment, control participants were then given the identical full intervention. All 6 par-

ticipants assigned to this group completed the aforemen- tioned process. Both groups completed the Psychological Profile of Forgiveness Scale to measure forgiveness, as well as scales to measure the mental health constructs of hope, state anxiety, trait anxiety, self-esteem, and depression.

McCullough et al. (1997)

McCullough et al. (1997) conducted an empathy interven- tion group, a comparison intervention group, and a waiting list group with college students from an introductory psy- chology course. Participants wished to learn information and skills that might help them to forgive a specific person whom they wanted to forgive but had been previously unable to forgive. They were not taking psychotropic medications or receiving counseling; did not manifest substance abuse prob- lems, psychotic behavior, or personality disorders that might disrupt the groups; and agreed to being randomly assigned to either a seminar or a waiting list. Assignment to groups consisted of 13 participants to the empathy seminar, 17 to the comparison seminar, and 40 to a waiting list. The empa- thy intervention was a seminar that promoted forgiveness through encouraging a process of both cognitive and affec- tive empathy. The seminar consisted of eight 1-hour ses- sions conducted over one weekend. Each seminar consisted of between 5 and 8 participants. At follow-up, complete data were available for 12 of 13 participants assigned to the empathy intervention. The comparison intervention focused only on a cognitive understanding of the benefits of forgive- ness, the definitions of forgiveness, and hearing other people’s stories of how they forgave. The practices of reframing and empathy were omitted. The intent of these sessions was to commend forgiving as a health-promoting behavior with- out explicitly enhancing empathy for the offender. To this end, the cognitive decision to forgive was emphasized. The duration and size of the comparison seminars were the same as that of the empathy seminar. At follow-up, complete data were available for 15 of 17 participants assigned to the com- parison intervention. For the control group, 39 of 40 par- ticipants completed the assessments. A Forgiving Scale (FS) was given to measure forgiveness. The constructs of affec- tive empathy and cognitive empathy were also measured.

Coyle and Enright (1997)

Coyle and Enright (1997) implemented an intervention de- signed to foster forgiveness with “postabortion men.” Par- ticipants consisted of 10 men who self-identified as hurt by the abortion decision of a partner. They were randomly as- signed to either the treatment or the control (waiting list) condition, which received treatment after a 12-week wait- ing period. The treatment condition consisted of 12 weekly 90-minute individual sessions. Sessions were conducted by a graduate student in educational psychology under the su- pervision of a licensed psychologist. The intervention was based on psychological variables and units of a process model of forgiveness (similar to those in Table 1). The Enright Forgiveness Inventory (EFI) was used to measure forgive-

15566676, 2004, 1, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/j.1556-6678.2004.tb00288.x by L

iberty U niversity, W

iley O nline L

ibrary on [13/05/2024]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

JOURNAL OF COUNSELING & DEVELOPMENT • WINTER 2004 • VOLUME 8284

B a s k i n a n d E n r i g h t

ness. Other mental health variables measured included state anger, state anxiety, and grief.

CATEGORIZING THE STUDIES

Given the theoretical foundations of the interventions, we di- vided them into three groupings: (a) those studies that are pri- marily decision based; (b) those studies that are process based and had a group format; and (c) those studies that are process based and followed an individual format (see Table 2). The deci- sion subdivision contains single session interventions and par- tial interventions, which use a decision-based model. This in- cludes the first intervention in Al-Mabuk et al. (1995), both interventions in McCullough and Worthington (1995), and the second intervention in McCullough et al. (1997). The process- group subdivision includes group interventions of six to eight sessions that are process based. This includes Hebl and Enright (1993), the second intervention in Al-Mabuk et al. (1995), and the first intervention in McCullough et al. (1997). Last, the process-individual subdivision includes those interventions of 12 or more sessions of individual therapy, using a process- based model. Specifically, this is Freedman and Enright (1996) and Coyle and Enright (1997). All nine studies fit unambigu- ously into one of these three categories.

IMPORTANT QUESTIONS

Based on these forgiveness interventions and their differ- ences in theoretical foundations, this review has three im-

portant tasks that will be addressed through numerical analy- sis. First, is there evidence for grouping the studies into three categories versus considering them all as one category? Sec- ond, a philosophical difference between decision-based and process-based models has emerged: a onetime event or a series of procedures designed to elicit forgiveness. Looking across studies, can it be shown that interventions based on one theo- retical foundation give stronger results than those based on the other, or are they equivalent? Third, it is important to ascer- tain the nature of the effects of forgiveness therapy within counseling. Again, looking across studies, and according to their theoretical basis, are these counseling interventions effective in increasing forgiveness? Furthermore, do their benefits go beyond forgiveness to other well-established emotional health constructs, such as depression, anxiety, and others? These are the questions of our meta-analysis.

THE META-ANALYTIC PLAN

To answer our first question, we determine whether the stud- ies are more accurately viewed as one or three groups based on a test of homogeneity. Regarding the second question, we sum the effect size of the studies within each group for forgiveness and compare these among groups. To answer our third ques- tion, we sum the effect sizes of all nonforgiveness emotional health variables within a study, then sum these across studies within each group to assess a final level of effect for each group. We then compare these totals across groups.

TABLE 2

Overview of Studies in Meta-Analysis

Study and Author

1 Hebl & Enright, 1993

2 Al-Mabuk, Enright, & Cardis, 1995

3 Al-Mabuk et al., 1995

4 McCullough & Worthington, 1995

5 McCullough & Worthington, 1995

6 Freedman & Enright, 1996

7 McCullough, Worthington, & Rachal, 1997

8 McCullough et al., 1997

9 Coyle & Enright, 1997

Note. PPFS = Psychological Profile of Forgiveness Scale; PLD = parental love-deprived; 9 Wade subscales = Wade’s (1989) Forgiveness Scale subscales; FS = Forgiving Scale; EFI = Enright Forgiveness Inventory.

Group

Process–Group

Decision

Process–Group

Decision

Decision

Process–Individual

Process–Group

Decision

Process–Individual

Intervention

Elderly women, 8 group ses- sions, full intervention

PLD adolescents, 4 group sessions, 9/17 units

PLD adolescents, 6 group sessions, 17/17 units

Undergraduates (nonserious hurt), 1 group session, re- store relationship focus

Undergraduates (nonserious hurt), 1 group session, ben- efits for the forgiver focus

Incest survivors, 52+ individual sessions, full intervention

Undergraduates, 8 group ses- sions (one weekend), empathy- focused forgiveness

Undergraduates, 8 group ses- sions (one weekend), nonempathy forgiveness

“Postabortion men,” 12 individual sessions, full intervention

Forgiveness Measure

Emotional Health Dependent Variable

PPFS

PPFS

PPFS

9 Wade subscales

9 Wade subscales

PPFS

FS

FS

EFI

Self-esteem, state-anxiety, trait-anxiety, depression

Attitude toward father, attitude toward mother, hope, state-anxiety, trait-anxiety, self- esteem, depression

Attitude toward father, attitude toward mother, hope, state-anxiety, trait-anxiety, self- esteem, depression

None

None

Hope, state-anxiety, trait-anxiety, self- esteem, depression

Affective empathy, cognitive empathy

Affective empathy, cognitive empathy

Forgiveness, state anxiety, state anger, grief

15566676, 2004, 1, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/j.1556-6678.2004.tb00288.x by L

iberty U niversity, W

iley O nline L

ibrary on [13/05/2024]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

JOURNAL OF COUNSELING & DEVELOPMENT • WINTER 2004 • VOLUME 82 85

F o r g i v e n e s s M e t a - A n a l y s i s

METHOD

Establish Studies

To be included in this meta-analysis, a study had to have been empirical, with a quantitative measure of forgiveness, have had a control group, and had to have been published in a refereed journal. Furthermore, interventions had to have been based on some model of forgiveness. First, a search of electronic data- bases was conducted to find all relevant studies. Second, as studies were located, their references were used to check for further studies that may exist. Third, qualitative reviews of forgiveness literature were examined for references to empiri- cal studies. Nine empirical studies were found, all of which fit the outlined criteria, with a total N = 330. All studies were accomplished with well-trained therapists/leaders, and each fell within one of the forgiveness models described earlier.

Test for Homogeneity

To determine whether the studies could be more accurately viewed as one group or three, a test of homogeneity is needed. Because the variance of each statistic can be estimated, the modified medium chi-square test (Cramer, 1946) can be used to test for homogeneity (Hedges & Olkin, 1985). In this case,

Q = Σ (dSi – dS – agg) 2 / σ2(dSi) (1)

is distributed as a chi-square with k–1 degrees of freedom, where k is the number of studies yielding effect sizes, dSi is the obtained effect sizes for study i, and dSi – agg is the aggregated effect size across studies. Significantly large values of Q signal the rejection of the null hypothesis of homogeneity.

In investigating these three groups, we hoped to shed light on the effectiveness of decision-based versus process- based interventions. In addition, in our examination of the group and individual formats, we sought to deepen our un- derstanding of what is effective in forgiveness interven- tions that emphasize processes. Next we determined if, within each grouping, there is a preponderance of evidence that suggests that these interventions are effective in in- creasing forgiveness. Finally, we examined whether there is sufficient evidence to support the hypothesis that these benefits extend beyond forgiveness to other important mental health variables.

Calculation of Effect Size

The next task was to compute the effect sizes of the interven- tions for the dependent variables. As previously outlined, two groups of dependent variables were examined: (a) forgiveness and (b) all other emotional health dependent variables. These effect sizes were calculated using the methods outlined by Hedges and Olkin (1985).

Two studies presented a challenge about possible inclu- sion where no single measure of forgiveness was reported. McCullough and Worthington (1995), which included two studies, reported nine subscales of the Wade Forgiveness Scale but not the aggregate single measure. Statistically, simple

aggregation of the subscales would lead to an inaccurate result, because their correlation would not be accounted for. The subscales could, however, be aggregated using the same method we used to combine multiple measures within one intervention in this present analysis (see the following), as long as intercorrelations were known. Because McCullough and Worthington did report these intercorrelations for the subscales at preintervention, the aggregation was success- fully accomplished.

For each outcome, the effect size g was calculated by tak- ing the difference between the intervention mean and the control mean and dividing by the pooled standard deviation of the two, according to the following formula:

g = (MI – MC)/s, (2)

where MI and MC are the mean levels of measurements (with I denoting the intervention and C the control group, and with s the pooled standard deviation). The unbiased popula- tion effect size d for each result was calculated by correct- ing for the bias in g (Hedges & Olkin, 1985):

d = [1 – 3/(4N – 9)]g, (3)

where N = nI + nC´ the sum of the participants in the inter- vention and control groups. The variance of d was estimated by (Hedges & Olkin, 1985):

σ2(D) = [N/(nInC)] + [D2/2N]. (4)

Furthermore, Hedges and Olkin’s method of weighting the contribution of each study by the inverse of its variance was used. This corrects for random variation resulting from divergent sample sizes.

Dependent Measures and Correlation

In aggregating all nonforgiveness dependent variables, an addi- tional component must be accounted for: the correlation among measures. An estimate of the effect size for an entire study ds was derived from a vector of the effect size d for each dependent variable and the correlation between each of these variables, as described by Hedges and Olkin (1985). This is accomplished by taking di as the vector of effect sizes across all dependent variables in a study and R as the correlation matrix between the outcome measures. Because it is uncom- mon for correlations between outcome measures to be pub- lished with a study, an estimate of this relationship is needed. Taking the lead of Wampold et al. (1997), 0.5 was used as a standard correlation between all dependent variables. This is rooted in the knowledge that in any given study there are typically several measures of several constructs. Furthermore, depression and anxiety are very common constructs to measure. Shapiro and Shapiro (1982) found that 56% of outcome studies targeted depression, anxiety, or both. Furthermore, Tanaka- Matsumi and Kameoka (1986), in a comprehensive study of the validity of popular measures of depression and anxiety, found that the average correlation of the measures was slightly

15566676, 2004, 1, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/j.1556-6678.2004.tb00288.x by L

iberty U niversity, W

iley O nline L

ibrary on [13/05/2024]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

JOURNAL OF COUNSELING & DEVELOPMENT • WINTER 2004 • VOLUME 8286

B a s k i n a n d E n r i g h t

greater than 0.5. On this foundation, a correlation of 0.5 was chosen to aggregate the effect sizes of dependent vari- ables to properly take into account the relationship between constructs in each study.

Again, following Hedges and Olkin (1985), the integra- tion of a 0.5 correlation into the calculation is accomplished through the covariance matrix Σ = DiRDi, where Di is a diagonal matrix of the respective SDs of di. If e is a column vector of 1s and Λ is the inverse of Σ, then the aggregate estimate of the effect size for a comparison is given by

ds = [Λe/e´Λe]di (5)

with a corresponding estimated variance of

σ2(ds) = 1/e´Λe (6)

(Hedges & Olkin, 1985, pp. 212–213). These values, calculated with Equations 5 and 6, were used as the aggregate estimate of the effect size of nonforgiveness dependent measures within each study. Once the effect size of forgiveness for each study was calculated, the results were aggregated to determine the mean effect size for each of the three theory-based groupings.

Initial posttest measurements of intervention and control groups were used exclusively for data analysis. Regarding long- term posttests, three studies did not include follow-up mea- surements (Hebl & Enright, 1993, and the two studies in Al- Mabuk et al., 1995), four studies measured follow-up at 6 weeks (McCullough & Worthington, 1995; McCullough et al., 1997), and two studies measured follow-up at longer than 6 weeks (Coyle & Enright, 1997; Freedman & Enright, 1996). Consequently, follow-up measures were not included due to their variability of inclusion and duration within the studies.

RESULTS Test for Homogeneity

Forgiveness results were used to test for homogeneity. The results, taken as one group, did not pass the test for homoge-

neity. The aggregated Q value was 30.05, which, according to chi-squared values, should have been no more than 15.51. Consequently, the nine studies did indeed need to be di- vided into groups according to a logical method.

For the three groups into which the studies were divided using theoretical foundations, QBETWEEN GROUPS equaled 23.31, which was well above the chi-squared critical value of 5.99 (that Q would be expected to be below, if the groups were in fact homogeneous). In each of the three groups, an empirical test of homogeneity confirmed the value of this division. For decision-based interventions Q1 equaled 3.93, below the critical value of 7.82. For process-based group interventions Q2 equaled 2.05, below the critical value of 5.99, and for process-based individual interventions Q3 equaled 0.89, well below the critical value of 3.84.

Forgiveness as Dependent Variable

The mean effect size for levels of forgiveness in decision- based interventions, versus a control group (4 interventions with a total n = 188), was d1f = –0.04 (95% confidence interval [CI]: –0.24 to 0.16). Because this confidence interval encom- passed zero, the result could be considered to differ from zero. Therefore, these results suggested no significant difference in forgiveness between those receiving a decision-based intervention and those receiving no intervention. The mean effect size for levels of forgiveness in process-based group interventions (3 interventions, total n = 120) was d2f = 0.82 (95% CI: 0.43 to 1.21). This effect size can be consid- ered in terms of the average person in the intervention group doing as well as or better than 75% of the control group. The mean effect size for levels of forgiveness in pro- cess-based individual interventions (2 interventions, total n = 22) was d3f = 1.66 (95% CI: 0.68 to 2.64). This effect size can be considered in terms of the average person in the intervention group doing as well as or better than 95% of the control group. These results are shown in Table 3 and Figure 1.

TABLE 3

Quantitative Results of Studies in Meta-Analysis

Study and Author

2 Al-Mabuk, Enright, & Cardis, 1995 4 McCullough & Worthington, 1995 5 McCullough & Worthington, 1995 8 McCullough, Worthington, & Rachal, 1997

1 Hebl & Enright, 1993 3 Al-Mabuk et al., 1995 7 McCullough et al., 1997

6 Freedman & Enright, 1996 9 Coyle & Enright, 1997

aThe control group was the same for Studies 4 and 5. *p < .05.

Group

Decision

Total Process–Group

Total Process–Individual

Total

Treatment n Control n Forgiveness Effect Size

Emotional Health Effect Size

24 30 35 15

104 13 24 12 49

6 5

11

24 21 21 39 84a

11 21 39 71

6 5

11

–0.30 0.05 0.10

–0.46 –0.04

0.70 1.17 0.53 0.83* 2.16 1.21 1.66*

–0.14 — —

0.56 0.16 0.72 0.42 0.75 0.59* 1.44 1.40 1.42*

15566676, 2004, 1, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/j.1556-6678.2004.tb00288.x by L

iberty U niversity, W

iley O nline L

ibrary on [13/05/2024]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

JOURNAL OF COUNSELING & DEVELOPMENT • WINTER 2004 • VOLUME 82 87

F o r g i v e n e s s M e t a - A n a l y s i s

Emotional Health Dependent Variables

The mean effect size for all emotional health dependent variables in decision-based interventions, versus their con- trol groups (2 interventions with a total n = 102), was d1e = 0.16 (95% CI: –0.16 to 0.48). Because this CI encompasses zero, the result cannot be considered to differ from zero. However, for this subgroup, results were heterogeneous. The mean effect size for all nonforgiveness dependent variables in process-based group interventions (3 interventions, total n = 120) was d2e = 0.59 (95% CI: 0.28 to 0.90), with homo- geneous results. This effect size can be considered in terms of the average person in the intervention group doing as well as or better than 65% of the control group. In addition, the mean effect size for all nonforgiveness dependent variables in process-based individual interventions (2 interventions, to- tal n = 22) was d3e = 1.42 (95% CI: 0.66 to 2.18), with homogeneous results. This effect size can be considered in terms of the average person in the intervention group doing as well as or better than 92% of the control group. These results are shown in Table 3 and Figure 1.

DISCUSSION

The results of this study suggest a number of conclusions. First, the low scores of the studies in the decision grouping relative to those in the two process groupings (group and individual interventions) suggest support for the greater ef- fectiveness of the process models of forgiveness. Second, the significantly higher scores for the longer term individual counseling versus the medium-length group counseling sug- gest something about the time and energy required by cli- ents and counselors to fully and successfully forgive a per- son for a deep injustice. Third, the large effect size of the process-based individual counseling suggests the value of their continued use, especially with the specific groups of clients already assessed. Although caution must be exercised because of the numbers of studies, results include important evidence meriting a thoughtful examination at this time.

The empirical evidence currently does not endorse the predominantly cognitive decision-based interventions. As a group, the results were not shown to be significantly differ- ent from the control group. This was clearly true for for- giveness measurements. Emotional health measurements lacked homogeneity, leaving some level of ambiguity. How-

ever, taken as a whole, these interventions did not show a significant effect, either because the model is incomplete or because forgiveness is not likely to be affected by counsel- ing. The results are low, given that placebo psychological interventions are known to have an average 0.42 effect when compared with no treatment (Lambert & Bergin, 1994). The predominantly cognitive component may still be important but may properly be based in an expanded process model and not in isolation of that process.

The empirical evidence supporting process models of for- giveness is apparent from the second category of studies. Forgiveness can be affected by counseling. Again, the 0.82 effect size on forgiveness can be considered in terms of the average person in the intervention group doing as well as or better than 75% of the control group. The difference in find- ings between this grouping and the decision-based grouping supports the use of a process model of forgiveness in coun- seling. In addition, the results were extended from forgive- ness to emotional health constructs at an effect size of 0.59, again meaning that the average person in the forgiveness group did as well as or better than 65% of the control group. According to Lipsey (1990), empirical norms for describing the magnitude of effect sizes include less than 0.33 as small, between 0.33 and 0.55 as medium, and any value larger than 0.55 as large. Consequently, across mental health variables, the effect of this set of interventions can be considered large. This extends the findings to endorse not only the process model of forgiveness but also the effectiveness of forgiveness counseling as a treatment. This moves forgiveness beyond a study of improving people’s moral development, which was a key idea in the earliest study on forgiveness (Enright et al., 1989), to a factor in improving mental health.

The support of the process model of forgiveness counseling continues in the results from the individual interventions. With an effect size of 1.66 (average person in intervention group did as well as or better than 95% of control group) for forgiveness, process models appear to be working. Further- more, the emotional health effect size of 1.42 (average per- son in intervention group did as well as or better than 92% of control group) supports both the efficacy of process-based forgiveness counseling and the link between forgiveness and mental health. According to Lipsey’s (1990) assessment of effect sizes, the results are almost 3 times the minimum level for a large effect size. In a different assessment, Lambert and Bergin (1994) placed the standard effect size for effective psychotherapies across theoretical orientations at 0.82. The 1.42 effect size remains strong, being almost twice the stan- dard value for professional treatments. These results move forgiveness therapies into the realm of being important within the counseling community.

Perhaps a concern for endorsing forgiveness interventions, and indeed whether counselors should be giving this variable attention, is that forgiveness is not an established mental health variable. Anger resulting from an injustice or a lack of forgive- ness has yet to be included as a diagnosis in the Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV; American Psychiatric Association [APA], 1994),

FIGURE 1

Aggregate Effect Sizes

Mean Effect Size (With 95% CI)

Decision-Forgiveness Process (Group)-Forgiveness

Process (Individual)-Forgiveness

Decision-Emotional Health Process (Group)-Emotional Health

Process (Individual)-Emotional Health

15566676, 2004, 1, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/j.1556-6678.2004.tb00288.x by L

iberty U niversity, W

iley O nline L

ibrary on [13/05/2024]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

JOURNAL OF COUNSELING & DEVELOPMENT • WINTER 2004 • VOLUME 8288

B a s k i n a n d E n r i g h t

although a concomitance of anger with numerous psycho- logical disorders is now being recognized (Deffenbacher, Lynch, Oetting, & Kemper, 1996; Fauz, Rappe, West, & Herzog, 1995; Fauz & Rosenbaum, 1997). In addition, effec- tiveness of process-based forgiveness interventions may be distorted across the board both by self-selection for a for- giveness study and by such an emphasis on forgiveness that clients feel compelled (for the sake of the researchers) to improve in this area. Yet, these points become a strength of the forgiveness interventions when the scope of inquiry is expanded to include all measured emotional health con- structs. The large effect sizes in the nonforgiveness measures affirm the value of the interventions beyond a focus on for- giveness. Furthermore, if clients felt compelled to forgive at a posttest, then in all likelihood, they would not feel so compelled at a 12-week or 14-month follow-up. Positive results are maintained at such follow-ups (Coyle & Enright, 1997; Freedman & Enright, 1996).

It might be argued that the significant results are not sim- ply due to the effectiveness of the interventions but instead to the skill of the counselors. Counselor differences have been reported (Crits-Christoph et al., 1991; Garfield, 1997; Jennings & Skovholt, 1999; Luborsky, McLellan, Diguer, Woody, & Seligman, 1997; Orlinsky, 1999; Project MATCH Research GroupUSA, 1998). Although this is possible, it is the case that five different counselors were involved across studies in the process-based group and process-based indi- vidual forgiveness interventions. Forgiveness therapy does appear to be a valuable mental health option apart from the skills of a few counselors. Of course, all counseling is subject to some counselor variation; there is no reason to believe that forgiveness would be any different.

Another important consideration is the confound between the type of intervention and the duration of the intervention. The interventions with greater effects were consistently longer than those with lesser effects. Specifically, the decision-based interventions ranged from between 1 and 8 sessions, the pro- cess-based group interventions ranged from 6 to 8 sessions, and the process-based individual interventions ranged from 12 sessions to 60 sessions. It can be argued that the increased effects are merely a result of greater attention paid to cli- ents. This critique merits two important responses. First, the length of treatment is integral to the theoretical foun- dation of decision-based versus process-based interventions. By their very nature, decision-based interventions are shorter. According to this orientation, once the decision has been made, most of the work of forgiveness has been done. By contrast, process-based models have significant decision components subsumed within them, along with additional elements. Con- sequently, time factors accurately mirror theoretical founda- tions, and therefore correctly express their efficacy. Second, the conclusion of the confound—“more is better” regarding time spent with clients—is not problematic. Precisely the concern of this analysis is to determine if forgiveness counseling is efficacious. To determine that more counseling is more effec- tive contributes to the thesis that this counseling method is potent. The concern is not to prove that forgiveness measures

are superior across the board to what a counselor might oth- erwise be doing but that they are an equally effective ele- ment in the repertoire of a professional counselor.

Although forgiveness is not an intervention for every dis- order, its empirical showing in this meta-analysis is encour- aging. Empirical strength has been shown with traditionally challenging populations. For example, with incest survivors (Freedman & Enright, 1996), no consistently effective inter- ventions had yet been established, even after attempts by expert counselors. The gains with this group suggest the value of using this approach with certain select clients. For example, when problems such as sexual abuse, divorce, and family-of- origin concerns are considered, it is realized that a number of mental health issues are significantly related to anger. With benefits across a range of mental health constructs, counse- lors should be aware of the potential benefits that forgiveness can have with clients suffering significantly from issues that involve anger borne out of unfair treatment.

An important consideration is whether these results es- tablish forgiveness therapy as an empirically supported treat- ment. According to determined criteria (Chambless & Hollon, 1998; Chambless et al., 1996; Crits-Christoph, 1996; Garfield, 1996; Kendall, 1998; Shapiro, 1996), many of the studies in this analysis have significant aspects to be consid- ered “efficacious.” This includes the fact that some have been compared with a psychological placebo (Al-Mabuk et al., 1995; McCullough et al., 1997); others have been compared with established interventions, such as a support group (Hebl & Enright, 1993); many have been conducted with treat- ment manuals (Al-Mabuk et al., 1995; Coyle & Enright, 1997; Freedman & Enright, 1996; Hebl & Enright, 1993); and all of the studies clearly specify characteristics of their client sample. Over the corpus of studies, all criteria are met. However, because there are not two specific studies from two independent research settings demonstrating all of these criteria, we are not yet able to make a final assess- ment. Significantly, these studies do establish themselves among the older criteria of Probably Efficacious Treatments. A sufficient criterion for this designation is having two ex- periments that show the treatment to be more effective than a waiting list control group (Chambless et al., 1996). This is true for the process-based individual interventions and is independently true for the process-based group interven- tions. To be fully considered “efficacious” would require only a few modifications in research design. This is certainly the direction future research should take.

Given these encouraging results, more research is called for. From the foundation established by the current study, the research needs to progress in three ways. First, quality studies, building on the strengths of those mentioned here, should be conducted to establish process-based forgiveness counseling as an “efficacious” treatment. Second, given the success of the process-based model, more exploration is needed. Continued investigation, verification, and elabora- tion of the model itself would enrich our current understand- ing. Third, these encouraging results should be expanded to other populations. Among DSM-IV (APA, 1994) diagnoses

15566676, 2004, 1, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/j.1556-6678.2004.tb00288.x by L

iberty U niversity, W

iley O nline L

ibrary on [13/05/2024]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

JOURNAL OF COUNSELING & DEVELOPMENT • WINTER 2004 • VOLUME 82 89

F o r g i v e n e s s M e t a - A n a l y s i s

such as conduct disorder, oppositional defiant disorder, mood disorders, and anxiety disorders, a subset of clients merit investigation regarding whether etiology is rooted in anger issues and whether forgiveness therapy might offer relief.

In conclusion, forgiveness counseling is an addition to the repertoire of applications for the professional counseling com- munity. The large effect sizes establish forgiveness counsel- ing as a contribution to that community. Although it should not be seen as a cure for all psychological concerns, there are certain emotional health issues for which it is particularly well suited, such as incest survivors, adolescents hurt by emo- tionally distant parents, and men hurt by the abortion deci- sion of a partner. It is important within the counseling com- munity to have a diversity of options with a sound empirical base. In addition, forgiveness therapy reveals the strength of relationship-based versus psychopharmacology-based inter- ventions. For one incest survivor, emotional difficulty had remained for 50 years. Drug-based treatment may have pro- vided short-term alleviation of symptoms. However, few would desire to maintain a drug treatment over 50 years. Fourteen- month process-based individual forgiveness counseling brought about significant change that was maintained 14 months later. It is unclear whether a 14-month drug treatment would yield long-term gains that could be similarly maintained. The find- ings here suggest that the effects of forgiveness counseling on clients are worthy of further study.

REFERENCES References marked with an asterisk indicate studies included in the meta- analysis.

Allison, D. B., & Faith, M. S. (1996). Hypnosis as an adjunct to cognitive- behavioral psychotherapy for obesity: A meta-analytic reappraisal. Journal of Consulting and Clinical Psychology, 64, 513–516.

*Al-Mabuk, R. H., Enright, R. D., & Cardis, P. (1995). Forgiveness educa- tion with parentally love-deprived late adolescents. Journal of Moral Education, 24, 427–443.

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author.

Bergin, A. E. (1988). Three contributions of a spiritual perspective to counseling, psychotherapy, and behavioral change, Counseling and Values, 33, 21–31.

Chambless, D. L., & Hollon, S. D. (1998). Defining empirically supported therapies. Journal of Consulting and Clinical Psychology, 66, 7–18.

Chambless, D. L., Sanderson, W. C., Shoham, V., Bennett Johnson, S., Pope, K. S., Crits-Christoph, P., et al. (1996). An update on empirically validated therapies. The Clinical Psychologist, 49, 5–18.

Close, H. T. (1970). Forgiveness and responsibility: A case study. Pastoral Psychology, 21, 19–25.

*Coyle, C. T., & Enright, R. D. (1997). Forgiveness intervention with postabortion men. Journal of Consulting and Clinical Psychology, 65, 1042–1046.

Cramer, H. (1946). Mathematical methods of statistics. Princeton, NJ: Princeton University Press.

Crits-Christoph, P. (1996). The dissemination of efficacious psychologi- cal treatments. Clinical Psychology: Science & Practice, 3, 260–263.

Crits-Christoph, P., Baranackie, K., Kurcias, J. S., Beck, A. T., Carroll, K., Perry, K., et al. (1991). Meta-analysis of therapist effects in psycho- therapy outcome studies. Psychotherapy Research, 1, 81–91.

Cunningham, B. B. (1985). The will to forgive: A pastoral theological view of forgiving. Journal of Pastoral Care, 39, 141–149.

Deffenbacher, J. L., Lynch, R. S., Oetting, E. R., & Kemper, C. C. (1996). Anger reduction in early adolescence. Journal of Counseling Psychology, 43, 149–157.

Denton, R. T., & Martin, M. W. (1998). Defining forgiveness: An empiri- cal exploration of process and role. American Journal of Family Therapy, 26, 281–292.

Droll, D. M. (1984). Forgiveness: Theory and research (Doctoral disserta- tion, University of Nevada–Reno, 1984). Dissertation Abstracts Inter- national–B, 45(08), 2732.

Enright, R. D. (2001). Forgiveness is a choice. Washington, DC: APA LifeTools, American Psychological Association.

Enright, R. D., & the Human Development Study Group. (1991). The moral development of forgiveness. In W. Kurtines & J. Gewirtz (Eds.), Moral behavior and development (Vol. 1, pp. 123–152). Hillsdale, NJ: Erlbaum.

Enright, R. D., Santos, M., & Al-Mabuk, R. (1989). The adolescent as forgiver. Journal of Adolescence, 12, 95–110.

Fauz, M., Rappe, S. M., West, J., & Herzog, D. B. (1995). Anger attacks in eating disorders. Psychiatry Research, 56, 205–212.

Fauz, M., & Rosenbaum, J. F. (1997). Anger attacks in depression. Depres- sion and Anxiety, 6, 2–6.

Ferch, S. R. (1998). Intentional forgiving as a counseling intervention. Journal of Counseling & Development, 76, 261–270.

Fitzgibbons, R. P. (1986). The cognitive and emotive use of forgiveness in the treatment of anger. Psychotherapy, 23, 629–633.

Flanigan, B. (1987, September 25). Forgiving. Workshops at the Mendota Mental Health Institute, Madison, WI.

Frankl, V. (1959). The will to meaning: Foundations and applications of logotherapy. New York: World Publishing House.

*Freedman, S. R., & Enright, R. D. (1996). Forgiveness as an intervention goal with incest survivors. Journal of Consulting and Clinical Psychol- ogy, 64, 983–992.

Garfield, S. L. (1996). Some problems associated with “validated” forms of psychotherapy. Clinical Psychology Science and Practice, 3, 218–229.

Garfield, S. L. (1997). The therapist as a neglected variable in psycho- therapy research. Clinical Psychology Science and Practice, 4, 40–43.

*Hebl, J. H., & Enright, R. D. (1993). Forgiveness as a psychotherapeutic goal with elderly females. Psychotherapy, 30, 658–667.

Hedges, L. V., & Olkin, I. (1985). Statistical methods for meta-analysis. San Diego, CA: Academic Press.

Hunter, R. C. A. (1978). Forgiveness, retaliation, and paranoid reactions. Canadian Psychiatric Association Journal, 23, 167–173.

Jennings, L., & Skovholt, T. M. (1999). The cognitive, emotional, and relational characteristics of master therapists. Journal of Counseling Psychology, 46, 3–11.

Kaufman, M. E. (1986). The courage to forgive. Israeli Journal of Psychia- try and Related Sciences, 21, 177–187.

Kendall, P. C. (1998). Empirically supported psychological therapies. Journal of Consulting and Clinical Psychology, 66, 3–6.

Kiel, D. V. (1986, February). I’m learning how to forgive. Decisions, 12–13. Lambert, M. J., & Bergin, A. E. (1994). The effectiveness of psycho-

therapy. In A. E. Bergin & S. L. Garfield (Eds.), Handbook of psycho- therapy and behavior change (4th ed., pp. 143–189). New York: Wiley.

Lipsey, M. W. (1990). Design sensitivity: Statistical power for experimental research. Newbury Park, CA: Sage.

Luborsky, L., McLellan, A. T., Diguer, L., Woody, G., & Seligman, D. A. (1997). The psychotherapist matters: Comparison of outcomes across twenty-two therapists and seven patient samples. Clinical Psychology: Science & Practice, 4, 53–65.

McCullough, M. E. (1999). Research on religion-accommodative coun- seling: Review and meta-analysis. Journal of Counseling Psychology, 46, 92–98.

*McCullough, M. E., & Worthington, E. L., Jr. (1995). Promoting forgive- ness: The comparison of two brief psychoeducational interventions with a waiting-list control. Counseling and Values, 40, 55–68.

*McCullough, M. E., Worthington, E. L., Jr., & Rachal, K. C. (1997). Interpersonal forgiving in close relationships. Journal of Personality and Social Psychology, 73, 321–336.

Morris, R. D., Audet, A. M., Angelillo, I. F., Chalmers, T. C., & Mosteller, F. (1992). Chlorination, chlorination by-products, and cancer: A meta- analysis. American Journal of Public Health, 82, 955–963.

Neblett, W. R. (1974). Forgiveness and ideals. Mind, 83, 269–275.

15566676, 2004, 1, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/j.1556-6678.2004.tb00288.x by L

iberty U niversity, W

iley O nline L

ibrary on [13/05/2024]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

JOURNAL OF COUNSELING & DEVELOPMENT • WINTER 2004 • VOLUME 8290

B a s k i n a n d E n r i g h t

North, J. (1987). Wrongdoing and forgiveness. Philosophy, 62, 499–508.

Orlinsky, D. E. (1999). The master therapist: Ideal character or clinical fiction? Comments and questions on Jennings and Skovholt’s “The Cog- nitive, Emotional, and Relational Characteristics of Master Thera- pists.” Journal of Counseling Psychology, 46, 12–15.

Osterndorf, C., Hepp-Dax, S., Miller, M., & Enright, R. (1999, Novem- ber). The moral development of forgiveness: Research and education. Paper presented at the annual meeting of the Association for Moral Education, Minneapolis, MN.

Patton, J. (1985). Is human forgiveness possible? Nashville, TN: Abingdon. Project MATCH Research GroupUSA. (1998). Therapist effects in three

treatments for alcohol problems. Psychotherapy Research, 8, 455–474. Shapiro, D. A. (1996). “Validated” treatments and evidence-based psy-

chological services. Clinical Psychology: Science & Practice, 3, 256–259. Shapiro, D. A., & Shapiro, D. (1982). Meta-analysis of comparative therapy

outcome studies: A replication and refinement. Psychological Bulletin, 92, 581–604.

Smedes, L. B. (1984). Forgive and forget: Healing the hurts we don’t deserve. San Francisco: Harper & Row.

Smith, M. (1981). The psychology of forgiveness. The Month, 14, 301–307. Tanaka-Matsumi, J., & Kameoka, V. A. (1986). Reliabilities and concur-

rent validities of popular self-report measures of depression, anxiety, and social desirability. Journal of Consulting and Clinical Psychology, 54, 328–333.

Trainer, M. F. (1981). Forgiveness: Intrinsic, role-expected, expedient, in the context of divorce (Doctoral dissertation, Boston University, 1981). Dissertation Abstracts International–B, 45(04), 1325.

Wade, S. H. (1989). The development of a scale to measure forgiveness. Unpub- lished doctoral dissertation, Fuller Theological Seminary, Pasadena, CA.

Wampold, B. E., Mondin, G. W., Moody, M., Stich, F., Benson, K., & Ahn, H. (1997). A meta-analysis of outcome studies comparing bona fide psychotherapies: Empirically, “All Must Have Prizes.” Psychological Bulletin, 122, 1–13.

Worthington, E. L., & DiBlasio, F. (1990). Promoting mutual forgiveness within the fractured relationship. Psychotherapy, 27, 219–223.

15566676, 2004, 1, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/j.1556-6678.2004.tb00288.x by L

iberty U niversity, W

iley O nline L

ibrary on [13/05/2024]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense