in some research i was doing for a class that i am in, i found a couple of articles that were really good in dealing with therapy for incest survivors, or those who've been sexually abused.

i know not everyone will like reading a research report, but some of you may, and you could find something in it to really encourage you. I am going to past it below, just in case you want to look at it.

Just to summarize:
To begin with they discuss what forgiveness is NOT...so don't immediately blow it off cuz you don't like what you think that term means. they have a really wonderful view of it (that it doesn't mean pretending nothing happened, letting the offender go without being held accountable, or reconciling/trusting the person who offended). Then it follows by telling you how they did their study, and what the therapy included.


Forgiveness as an Intervention Goal With Incest Survivors
By: Suzanne R. Freedman
Department of Educational Psychology, University of Northern Iowa
Robert D. Enright
Department of Educational Psychology, University of WisconsinMadison

Correspondence concerning this article should be addressed to: Suzanne R. Freedman, Department of Educational Psychology, University of Northern lowa, Cedar Falls, lowa 50614-0607

Interpersonal forgiveness is slowly emerging in psychology despite its long history of exclusive exploration in theology and philosophy (see, e.g., Nietzsche, 1886). Gartner (1988), recognizing the potential benefits of forgiving, challenged clinical psychology to build therapeutic models of forgiveness. The psychiatrist Fitzgibbons (1986) saw forgiveness as useful for anger reduction that enabled individuals to abandon guilt by expressing anger in healthy ways; Hope's (1987) ideas are similar. Many clinicians through case study reports see reductions in anger, depression, and anxiety for those who forgive their offenders (see also Cotroneo, 1982; Hunter, 1979; Kaufman, 1989).

Interpersonal forgiveness is defined as an unjustly hurt person's act of deliberately giving up resentment toward an offender while fostering the undeserved qualities of beneficence and compassion toward that offender (Enright and the Human Development Study Group, 1991; North, 1987). Forgiveness is in the context of deep injustice, in contrast to everyday annoyances. There is a decided paradoxical quality to forgiveness as the forgiver gives up the resentment, to which he or she has a right, and gives the gift of compassion, to which the offender has no right.

It is perhaps this paradoxical quality that led some to criticize forgiveness as morally unwarranted and psychologically unhealthy. For example, some who discuss forgiveness in the context of sexual abuse consider it unwise (Bass & Davis, 1988), potentially dangerous (Engel, 1989), and inappropriate (Forward, 1989). However, as Enright, Eastin, Golden, Sarinopoulos, and Freedman (1992) clarified, much of the criticism stems from a possible misunderstanding of forgiveness. For example, forgiveness is occasionally conflated with pardon. If we forgive, we let perpetrators free to cause more damage. When one forgives, one does not open a jail cell door but has an affective, cognitive, and possibly behavioral transformation toward the injurer; one can forgive and see justice realized. Forgiveness, condonation, and excusing are sometimes confused as well. In forgiving, the offended realizes that an offender has committed a serious wrong. The offer of lowered resentment and increased compassion are given nonetheless. In condoning, the offended comes to believe that there was no real injury in the first place. Forgiveness is at times equated with reconciliation, which is a misunderstanding of both constructs. In forgiving, the injured party may give up the qualities of resentment or even hatred but not necessarily enter into relation with an untrusted offender. In reconciliation, the offender realizes his or her wrong and takes steps to correct this behavior before the two enter once again into a relationship. It is the equating of forgiveness with reconciliation that seems to spark the majority of criticism against forgiveness as a moral and healthy response (Enright et al., 1992). The gist of such criticism is that the act of forgiving perpetuates abuse. When we realize that one may forgive without reconciling, this argument loses its weight. Given that the actual scientific study of forgiveness is rare and that forgiveness has its decidedly vocal critics in the area of sexual abuse, it is not surprising that no published study exists showing the psychological effects on incest survivors who forgive their perpetrators.

Forgiveness therapy and education programs do exist in the context of older adults who have deep hurts (Hebl & Enright, 1993) and young people in late adolescence suffering from parental love deprivation (Al-Mabuk, Enright, & Cardis, in press). Those programs were based on the forgiveness model developed by Enright et al. (1991), which includes 17 units or processes. Enright et al. (1991) developed this specific model by synthesizing the processes involved in interpersonal forgiveness discussed in the forgiveness literature. The first seven units involve reexamining the injury and one's emotional, cognitive, and behavioral response to the injury. The last 10 units include one's decision and commitment to work on the forgiveness process and the active therapeutic regimen involved in forgiving someone. Details of this model will appear in the Method section.

It was in 1975 that the topic of incest became more frequently discussed in the literature. In the past 20 years, there has been a tremendous increase in reported cases, perhaps because of the greater number of adult survivors seeking treatment and society's gradual willingness to discuss the topic. Statistics suggest that over 20% of adult women are sexually abused before age 18 (Gold, 1986). Such numbers have challenged therapists to acquire knowledge about treatment goals and techniques for abuse victims. The majority of research has concentrated on identifying the negative psychological effects; only more recently have efforts turned to the issue of treatment.

Incest survivors are at a significantly greater risk than the general population for psychological health problems, such as depression, anxiety, low self-esteem, marital difficulties, suicidal ideation, self-blame and guilt, eating disorders, substance abuse, and conflictual interpersonal relations (Alexander, 1993; Mullen, 1993; Roth & Newman, 1993). Briere and Runtz (1993) refer to the effects of child sexual abuse as psychological toxicity. One explanation for the trauma is that a trusted person, who was supposed to love and protect the survivor, has instead manipulated her through lies and misrepresentations about moral standards (Finkelhor & Browne, 1985). This betrayal is often linked with interpersonal difficulty for the survivor, especially within family and intimate relationships. Poor social adjustment, feelings of isolation, and impaired judgement about the trustworthiness of others are frequent complaints of incest survivors (Daugherty, 1984). Anger is also a problem for many incest survivors who do not directly express it toward the abuser. Instead, anger manifests in the survivor's interactions with other individuals, damaging interpersonal relations.

Although psychotherapeutic treatments for sexual abuse survivors are discussed in the literature (Bagley & Ramsay, 1985; Ingram, 1985; Jehu, Gazan, & Klassen, 1985; Roth & Newman, 1993; Wyatt & Powell, 1985), none focuses on the survivor's unresolved feelings toward the abuser. That incest emerges out of a deep interpersonal injustice, the very issue that forgiveness tackles, suggests that forgiveness therapy may be beneficial for incest survivors. Many researchers call for more controlled treatment research with this population (Alexander, 1993; Beutler, 1993; Finkelhor, 1984; Follette, Alexander, & Follette, 1991; Mayer, 1983). Follette et al. (1991) noted the paucity of research on treating adult survivors of childhood incest. They identified only one controlled outcome study investigating the effectiveness of two different formats of group therapy for incest survivors (see Alexander, Neimeyer, Follette, Moore, & Harter, 1989).

The purpose of the present investigation is to assess the effectiveness of an intervention program for incest survivors using forgiveness as the goal. Because forgiveness intervention has proven to be effective in reducing depression and anxiety (Hebl & Enright, 1993) and raising self-esteem (Al-Mabuk et al., in press), the long-term effects that most survivors report, we thought it time to experiment with forgiveness for those individuals so deeply hurt. The sample was a group of adult women who were incestuously abused in childhood. We expected that the act of forgiving one's abuser would be associated with improved psychological health as illustrated in measures of self-esteem, depression, anxiety, and hopefulness toward the future.
Method
Participants

Participants were 12 women recruited from a Midwestern community to participate in an intervention for women sexually abused as a child by a male relative. Average age was 36 years (range, 2454 years), and average education was 15 years (range, 1219 years). All participants were Caucasian. Five participants were single, 3 were married, and 4 were divorced. For 50% (6) of the participants, their perpetrator was their natural father; for 8%, their stepfather; for 16%, their brother; for 12%, their grandfather; and for 8%, an uncle. Average age of onset of abuse was 6.3 years; average duration was a little less than 6 years (range, 6 months to 12 years). Forty-two percent of the women reported experiencing fondling only; 17%, fondling and oralgenital contact; and 42%, intercourse.
Instruments

Initial screenig

A screening interview and crisis symptom checklist were given to all respondents. Only those who met the criterion of having been sexually abused, involving contact, by a male relative, in which the abuse had not occurred within the past 2 years were included. In this way, the individual had time to normalize her reaction to the abuse. Participants also had to show evidence of experiencing psychological difficulty as illustrated on the screening interview, crisis symptom checklist, and pretest measures. We wanted a clinical sample that could benefit from the intervention. Exclusion criteria were severe substance abuse and severe psychopathology, such as psychosis. All participants were informed at the first interview of the possibility of being placed in a waiting-list control group and having to wait approximately 1 year before beginning the intervention. Participants also were told that we were interested in examining long-term effects of sexual abuse and applying the model we developed.

Psychological Profile of Forgiveness Scale

This is the same instrument used by Hebl and Enright (1993) and Al-Mabuk et al. (in press). This 30-item scale was used to measure the degree to which the person forgave the perpetrator. Each participant was asked to consider the one who offended her and rate that person on the basis of one's current emotions, cognitions, and behaviors. This scale is based on the theoretical premise that when one forgives there are six psychological responses: absence of negative emotions (I do not feel bitter toward the person), absence of negative judgments (I do not view the person as below me), absence of negative behaviors (I will not act negatively toward the person), presence of positive affects (I feel close to the person), presence of positive judgments (the person deserves fairness), and presence of positive behaviors (I'll show friendship). The word forgiveness is not used in the scale. Five items for each of the six subcategories were generated. Each item was placed into a 4-point Likert format ranging from strongly disagree (1) to strongly agree (4), generating 30 items with a range of scores from 30 (low forgiveness) to 120 (high forgiveness). Each of the six subscales has a range from 5 to 20. Validity to date shows significant correlations between those high on this forgiveness scale and high in self-esteem and low in psychological depression and anxiety (Al-Mabuk et al., in press). Validity also is evidenced by significantly increased scores following forgiveness interventions (Al-Mabuk et al., in press; Hebl & Enright, 1993). Internal consistency is usually in the .90.95 range (Al-Mabuk et al., in press; Subkoviak et al., in press).

Anxiety inventory

The StateTrait Anxiety Inventory (STAI) developed by Spielberger, Gorsuch, Lushene, Vagg, and Jacobs (1983) was used as a measure of anxiety. The STAI comprises separate self-report scales for measuring two distinct anxiety concepts: 20 state anxiety items that indicate how each feels at a particular moment in time and 20 trait anxiety items that indicate how a person generally feels. Each subscale is analyzed separately. Each item is given a 14 weight, yielding a total subscale range of 2080. Reliability and validity are adequate (Spielberger et al., 1983).

Depression scale

The Beck Depression Inventory, developed by Beck, Ward, Mendelson, Mock, and Erbaugh (1961) to measure the behavioral manifestations of depression, was used. The 21 items are scored on a scale ranging from 0 to 3, with a high score representing high depression. Beck et al. (1961) reported adequate psychometric properties.

Self-esteem inventory

The adult form of the Coopersmith Self-Esteem Inventory (CSEI) (Coopersmith, 1981), consists of 25 truefalse items adapted from the School Short form. It evaluates attitudes toward self in several respects: general self, social, self-peers, and homeparents. The raw score is multiplied by 4, generating scores ranging from 0 (low score) to 100 (high score). Coopersmith reported adequate validity and reliability for this scale.

Hope Scale

This 30-item scale reported by Al-Mabuk et al. (in press) assesses optimism toward the future in areas of parental relationships (there will be more trust in my relationship with my parents), friendship (I will have longer lasting friendships), achievement (I will be able to follow through on tasks), and affect (I will have more joy than I have now). The Likert format ranged from 1 (it probably won't happen) to 5 (it will happen to a greater extent than I now realize). Scores range from 30 (low hope) to 150 (high hope). Al-Mabuk et al. (in press), using an earlier version of the Hope Scale found positive correlations with this scale and forgiveness and also found a high internal consistency reliability of .93 of the Hope Scale.

Self-Report Forgiveness Measure

We developed the Self-Report Forgiveness Measure to assess whether the experimental participant had truly forgiven. Three definitions taken from Augsburger (1981), Cunningham (1985), Enright et al. (1991), North (1987), Richards (1988), and Smedes (1984) were given to the participants to read. Participants were then instructed to answer five questions that assessed their feelings toward the perpetrator in relation to the definitions of forgiveness. If the participant stated that she had forgiven and her rationale for doing so was valid, the forgiveness was believed to be genuine. Examples of two of the five questions are the following: 1. Do any of these statements describe your current attitudes or feelings toward your abuser? Which ones? 3. Do you believe that you have truly forgiven your abuser?
Design

We used a yoked, randomized experimental and control group design. Pairs of participants were matched as closely as possible on the following variables: nature of the abuse, abuser, current age of the survivor, education level, and socioeconomic status (SES) of the survivor. One participant of each pair was randomly selected to be in the experimental group (receiving the intervention first) and the other to a group receiving the intervention only after the matched experimental participant completed her program.
Procedure

Testing procedure

Initial assessment was the same for each individual responding to an advertisement, describing the research study as an intervention for women who had been sexually abused during childhood or adolescence by a male relative. Women still suffering from the abuse who were interested in individual sessions with a graduate student were asked to contact the interviewer. The ads appeared in local newspapers, agencies that work with incest survivors, doctor's offices, and so forth. A screening interview occurred first, in which the respondent was given the screening interviewresearch questionnaire and the Crisis Symptom Checklist. At this time, the study was described as an intervention designed to help incest survivors heal. Forgiveness was not mentioned, because we did not want to bias the respondent's attitude toward forgiveness on the forgiveness pretest measure. If the respondent met the criteria of having been sexually abused by a male relative, with no abuse occurring in the past 2 years and was experiencing negative psychological effects as illustrated on the Crisis Symptom Checklist, she was administered the five pretest measures. Before administration of the pretest measures, all participants signed an informed consent document. If the individual scored below the median on five of the six pretest measures, she was readministered the measures 2 weeks later and once again 2 weeks later. These reassessments were conducted with the intent of averaging a given participant's pretest scores for the purpose of reducing measurement error, which can be a problem with small samples. The measures were presented individually in randomized order to each participant for each testing period to prevent order effects.

The intervention began after the three pretesting periods for the experimental participants. Once the experimental respondent indicated that she had forgiven her perpetrator, as assessed by the Self-Report Forgiveness Measure, she was given the posttest measures. As with the three pretest sessions, there were three posttesting times with 2-week intervals. Posttesting was idiosyncratic for each participant as it depended on when the participant finished the intervention.

The same testing procedure was repeated when the control participants entered into the intervention. The three sets of measures taken by the control participants when their matched experimental participant reached criterion served as second pretest scores. The original experimental participants were given the posttests a second time as a follow-up when their matched control participant reached criterion. The only difference was that instead of taking the five measures three times, participants took them only once. Because the intervention had been over for some individuals for a full year, it was a difficult task to have them complete the measures three times.

Intervention procedure

After pretesting, the experimental participants engaged in the forgiveness intervention. The control group was a waiting-list control group in which the participants had contact once a month with the experimenter. Every third month, the contact was face to face, and the other months it was over the phone. The topic of forgiveness was never mentioned during these contacts. As these six participants had to wait on the average 14 months before beginning the intervention, it was possible that they may have started another type of therapy or treatment while waiting for the intervention to begin. This did not occur for any of the six women in the waiting-list control group.

The participants in the experimental group received 60-min weekly individual sessions. Each person participating in the intervention received a manual (available from Suzanne R. Freedman) that described the 17 units in the process model (Table 1) and offered examples applicable to incest survivors.

ccp-64-5-983-tbl1a.gif

Psychological Variables Engaged in a Process Intervention on Forgiveness

In Unit 1, the injured examines the psychological defense mechanisms she used to protect against pain. Although these defense mechanisms are often immediately adaptive, they prevent healing if held too long. The survivor's use of denial, displacement, and projection, for example, need to be recognized to confront true emotions. Unit 2 focuses on anger. Many individuals erroneously believe that getting angry is not part of the forgiveness process. In actuality it is one of the most important units. Before forgiving, one needs to express her justified anger at being personally, unfairly, and deeply injured. Unit 3 reflects the shame and guilt that one feels after being wronged. This is especially true in cases of sexual abuse. Units 4 and 5 focus on the injured's behavior of iattaching an excessive amount of emotional energy to the hurtful event (cathexis) and spending much time replaying the events of the injury in one's mind (cognitive rehearsal). Unit 6 addresses the injured party's behavior of comparing herself to the injurer. This can lead to the conclusion that the injurer is better off and, thus, life is unfair (Unit 7). Units 8 and 9 are turning points in the process. Unit 8 describes the change of heart that North (1987) recognizes as necessary to emotional healing. The commitment to forgive is made in Unit 9. It is a cognitive decision to work on the forgiveness process although one may not feel like forgiving at the moment. Units 10 through 16 make up the active therapeutic processes of forgiveness. Reframing, or rethinking who the injurer is, constitutes Unit 10. The injured views the offender in context and takes into consideration her developmental history, life experiences, and worth as a human being. One's feelings of empathy and compassion are analyzed in Units 11 and 12. Empathy is the affective counterpart to reframing and enables one to become more sensitive to the other person's internal frame of reference. This willingness to share the suffering enables the injured to accept the pain of the injury (Unit 13) rather than pass it on. One's realization that she has needed others' forgiveness in the past is the topic of Unit 14 and enables the injured party to see the imperfect nature of all human beings. In Unit 15, the injured realizes that she is not the same person as a result of being injured. These insights may lead to a decrease in negative feelings, thoughts, and behaviors, toward the offender (Unit 16) and the beginning of forgiveness. The feelings of release and improved psychological health the injured experiences (Unit 17) should result from the act of forgiving. As the journey of forgiveness is very personal, there will be marked individual differences in the way in which individuals move through the process. Not all people experience forgiveness in exactly the same way, and each person's process must be respected. There is no link in the published literature to date between forgiveness and incest, not only because empirical examination of forgiveness is rare but also because published literature on incest was sparse until recently.

All 17 processes were discussed during the intervention, and any given session focused on only one process. However, some individuals focused more on certain units than others depending on their personal issues. For example, one participant who felt particularly guilty and ashamed spent 4 weeks discussing Unit 3, Feelings of Guilt and Shame, whereas another participant only spent 1 week on the topic. The pace at which the individual passed through the processes determined how quickly she met criterion and finished the intervention. For this reason, individual sessions were held, as each individual was then able to work at her own pace without having to worry about the others' progress. We were also aware that severity of incest could play a part in intervention length; the greater the severity, the longer it may take to work through the processes. Although research illustrates that group therapy is a preferred treatment (Gold-Steinberg & Buttenheim, 1993), the advantages of individual sessions for this specific intervention outweighed the advantages of group therapy. Because each participant continued with the intervention to criterion, each participant had different amounts of time within the intervention. The average length of the intervention was 14.3 months, (range, 10 to 16 months).

Once an experimental participant indicated that she forgave her perpetrator, as illustrated on the Self-Report scale, the termination process began and both the experimental participant and matched control participant were given the dependent measures. The yoked control participant then began the intervention and followed the same procedures as the original experimental participant.

Sessions were audiotaped on a random basis, and tapes were randomly selected to assess the integrity of the treatment condition. The 17 psychological variables engaged in a process intervention on forgiveness (Table 1) was made into a checklist. Graduate students not involved in the intervention rated what specific unit was being discussed in the tapes they were examining. Agreement between the specific session and unit to be discussed was 88% over 30 tapes.

Qualifications of the interviewer

All participants were seen by the same female interviewer studying for a PhD in psychology and preparing for psychological licensure. She was supervised by 2 licensed psychologists. Before the research, the interviewer participated in 32 hr of training and received a certificate stating that she was qualified to work with individuals who were sexually abused. Before the intervention, she had served as a volunteer group facilitator for young girls who had been sexually abused by a family member.
Results

Because each participant was seen independent of all others, the individual rather than the group became the level of analysis and, thus, conventional parametric statistics were used when we compared the 6 experimental individuals with the 6 control participants. Cronbach's alpha of internal consistency was calculated for each scale. All showed high reliability with the exception of the Self-Esteem Scale, which had adequate reliability: Psychological Profile of Forgiveness Scale = .92 (Negative Affect subscale = . 66, Positive Affect subscale = .60, Negative Cognition subscale = .61, Positive Cognition subscale = .67, Negative Behavior subscale = .68, and Positive Behavior subscale = .81), Anxiety Scale = .90, STAI Trait subscale = .81, STAI State subscale = .89, Beck Depression Inventory = .84, Self-Esteem Scale = .69, and Hope Scale = .96. Means and standard deviations for all dependent measures at pretest, Posttest 1, and Posttest 2 are given in Table 2. As each scale was given three times for each testing period (pretest, Posttest 1, and Posttest 2), the score reported is the participant's average of the three test administrations for each period.

ccp-64-5-983-tbl2a.gif

Descriptive Statistics for the Dependent Variables

One-tailed matched pairs ttests were used to test the amount of change between experimental and control groups on the five dependent measures and their subscales. Despite Cronbach and Furby's (1970) early warning about change score unreliability, more recent works find change scores to be effective ways of measuring experimental effects (see Rogosa & Willett, 1983; Willett, 19881989). This is especially the case with reliably stable measures, which we have, particularly given the multiple-assessment procedures for each measure at each testing time. Each dependent measure was tested for four changescore comparisons of interest. Comparison 1 examined the change from pretest to posttest for the experimental group versus pretest to posttest for the control group. It was expected that there would be a significant difference between the groups, with the experimental group illustrating a greater positive change in psychological well-being for each measure. Comparison 2 examined the change within the control group from pretest to Posttest 1 versus Posttest 1 to Posttest 2. It was expected that there would be a significant difference, with more change in a positive direction occurring during the latter time for all dependent measures because the control group became an experimental group. Comparison 3 examined the change in the experimental group from pretest to post-test versus the change in the control-group-turned-experimental from Posttest 1 to Posttest 2. No significant differences were expected as the two interventions were the same for both groups. Comparison 4 examined the change from pretest to follow-up in the experimental group versus Posttest 1 to Posttest 2 in the control-group-turned-experimental group. It was hypothesized that there would not be significant differences between the groups on the five dependent measures because the experimental group should show stability in their scores rather than returning to baseline patterns. This would be evidence that there was maintenance of psychological well-being up to approximately 1 year after the intervention ended. The means and standard deviations for the change scores within each of the four comparisons are in Table 3 .

ccp-64-5-983-tbl3a.gif

Mean Change Scores (and Standard Deviations) Across the Four Statistical Comparisons

For Comparison 1, statistical significance was demonstrated on all the measures except for self-esteem. Thus, the experimental group showed a significantly greater reduction in anxiety, t ( 5) = 3.02, p < .05, state anxiety, t ( 5) = 2.79, p < .05, trait anxiety, t (5) = 2.74, p < .05, and depression, t ( 5) = 2.52, < .05, and greater increase in forgiving the perpetrator, t (5) = 2.41, p < .05, and in hope, t (5) = 4.32, p < .05. The unusually large standard deviation for the experimental group in the self-esteem score may be responsible in part for the lack of statistical significance. The following Forgiveness subscales were also statistically significant (p < .05) favoring the experimental group over the control group for the Comparison 1: positive affect, t (5) = 2.67; negative cognition, t (5) = 3.25; positive cognition, t (5) = 2.58, and negative behavior, t (5) = 2.94.

Comparison 2 showed significance for all five measures: Psychological Profile of Forgiveness, t (5) = 4.43, p < .05, anxiety, t (5) = 2.12, p < .05 [ as well as trait anxiety, t (5) = 2.36, p < .05 ], depression, t (5) = 2.76, p < .05, self-esteem, t (5) = 3.12, p < .05, and hope, t (5) = 7.64, p < .05. Thus, the original control group, once it became an experimental group, illustrated greater psychological health as evidenced on all the measures after participation in the forgiveness intervention compared to itself as control group. Similar to the aforementioned set of analyses, the following Forgiveness subscales were statistically significant (p < .05) favoring the participants in the control-group-turned-experimental over themselves as control group participants for Comparison 2: positive affect, t (5) = 4.10; negative cognition, t (5) = 5.11; positive cognition, t (5) = 2.10; and negative behavior, t (5) = 3.80.

Significance was not expected for the analyses of Comparison 3, which compared the change between the groups after participation in the intervention. Significance was not found for any imeasure. The analyses for Comparison 4 examined the change in the experimental group from pretest to follow-up in comparison with the change in the control-group-turned-experimental group from Posttest 1 to Posttest 2. Significance was not expected and was not found on any of the measures. After receiving the intervention, the control-group-turned-experimental group showed emotional well-being that did not differ significantly from the experimental group's emotional health 1-year postintervention. The experimental group participants maintained their change patterns on all scales, thus showing that there was no washout effect.

See the appendix for a case study showing some of the psychological processes of Table 1 involved in the forgiveness intervention (names have been changed to protect confidentiality).
Discussion

The present study demonstrates the effectiveness of a forgiveness intervention for female incest survivors. To our knowledge, this is the first study to empirically examine the role of forgiveness as a treatment for incest survivors and the first to empirically identify a relationship between forgiving and improved psychological health. Thus, this intervention was successful not only in its goal of promoting forgiveness in the participants but also in increasing psychological well-being as illustrated in the higher self-esteem and hope, and lowered depression and anxiety reported by the women following participation in the intervention. This relationship needs to be highlighted as forgiveness has been too quickly dismissed by critics in the past.

A potentially important finding concerns the pattern of change in forgiveness toward the abusers. Before the forgiveness intervention, the groups present a picture of unforgiveness. After participation in the intervention, both experimental groups showed stronger forgiveness, not only on the overall scale but also within the following subscales: Subtraction of Negative Behavior and Judgments and Addition of Positive Affect and Cognitions toward the injurer. The intervention, in other words, had an effect on all psychological systems of affect, cognition, and behavior toward the perpetrator.

The reduction in overall anxiety and in the Trait and State subscales after treatment is evidence of the important relationship between forgiving those who inflicted deep hurt and positive psychological health. Scores on the Beck Depression Inventory for both groups after intervention illustrate that the act of forgiveness is related to reductions in reported depression for incest survivors. Before the intervention, the participants reported moderate depression; after intervention, the average report indicated little or no depression. Because the experimental participants continued to illustrate decreased depression up to 1 year after the intervention had ended, forgiveness has shown itself to be psychologically beneficial both in the short term and in the long run.

The survivors' self-esteem was another key variable examined. Bagley and Ramsay (1985) found that women with very poor self-esteem were almost four times more likely to have been sexually abused. The mean self-esteem for this sample, preintervention, was lower than for the normed sample reported by Coopersmith (1981); 1 year after the intervention it was higher than average for the first experimental group and immediately after the intervention self-esteem was higher for the second experimental group. Verbal reports from the participants illustrate that the intervention was successful in positively changing their feelings about themselves. One woman who moved to another community after the intervention said that she felt much more confident about herself, which was a big reason she was able to set off on her own. Another woman who forgave her father felt confident enough to return to school, with plans to start her own business. One woman realized that she deserved more than she was getting from her live-in partner and ended the relationship, something she was not able to do earlier. As all 12 participants listed low self-esteem as an effect of the abuse, the relationship between forgiving and positive changes in the way one feels about oneself cannot be ignored. This is particularly true for incest survivors who suffer from extreme feelings of insecurity, self-doubt, and shame. We must realize, however, that such changes occurred in only one of the two experimental treatments (nonsignificant trends in the positive direction did occur in the other group).

The greater sense of hope that the participants experienced after the intervention is additional evidence that, when there is a change in forgiveness, there is improvement in psychological health. To date, optimism about the future has not been examined in the incest literature, but it may be a variable that has significant implications for determining how an individual's life view changes after going through the forgiveness process. Forgiveness shows individuals that there are alternatives to living one's life with anger, bitterness, and hatred. In certain cases, the hope for a restored relationship with the injurer exists, something that the survivor may have never conceived as possible. Because the choice to forgive is so significant for one's outlook on life, it would be interesting to administer the Hope Scale at various times during the intervention. This may help determine at what point a change in one's life view begins to appear.

The aforementioned results illustrate how forgiving may be psychologically beneficial to individuals who have suffered deep hurts. Not one participant in the study showed psychological deterioration or any negative effects as a result of forgiving. It must be mentioned that each participant experienced her own individual process of forgiveness. Some participants were able to feel love toward their abuser and enter into an improved relationship with him. Other participants felt relieved that forgiveness could just be a cessation of negative feelings, thoughts, and behaviors toward the abuser. Each individual's experience of forgiveness was unique to her specific situation although each had the same treatment model.

The criterion-referenced format used, in which every person has an idiosyncratic ending point, is an important part of this study. Learning toward mastery may be a key. The intervention continued for each participant until she stated that she had forgiven. This allowed each participant to progress at her own pace without being affected by another participant in the study. It is important to note that each case was relatively long term, over 1 year on average.

The benefits of using a manual on forgiveness should be noted as well. Having something concrete to take home and read was reported as helpful by the participants. They would frequently take notes in the manual and refer to them during the session. As Beutler (1993) pointed out, benefits of treatment manuals include allowing both the reliability and skillfullness of the therapy to be assessed, as well as assisting in the education of psychotherapists. Individuals who develop future interventions on forgiveness would be wise to develop a manual that fits the population.

A word on design is in order. A frequent criticism of intervention research is that replications are necessary before conclusions are reached. This is considered necessary because the vast majority of such interventions involve a therapist interacting with a group. Because each individual within the group may influence every other individual's responses on dependent measures, then one does not have independent observations of each participant. Because standard statistics assume independence, we would have violated statistical assumptions if we had used a group format. The present study circumvented this charge by intervening with one participant at a time, thus maintaining independence across the 12 participants.

Another issue frequently raised is Therapist Treatment interaction. The question posed goes something like this: Is the intervention the cause of the improvement, or is it simply the skills of this particular therapist? In our opinion, in the context of forgiveness therapy we cannot so clearly dichotomize an answer because the therapist must invest in the lives of the clients, not just implement a technique. The average time of intervention here was 14 months. Therapists who are convinced of the efficacy of forgiveness are more likely to make the time commitment necessary to see client improvement.

We, therefore, are claiming that the success of this program is based on the inextricable interaction between intervener and therapeutic content. The two should not be separated. Does this mean that only this intervener can be successful? We expect that only those with a deep knowledge of forgiveness and forgiveness therapy, as well as a deep commitment to incest survivors will be successful.

We already know, through an examination of the scientific record, that therapist skills alone are rarely responsible for incest survivors' improved psychological health. Many a skilled therapist has attempted myriad programs, devoid of forgiveness as a focus, to help this group. Most report results that range from disappointing to modest at best. If we could isolate skill alone, which we doubt is possible, then some of these past efforts should have demonstrated dramatic results that match the therapists' skills. Such results have never appeared in print.

Were we to devise a study to isolate forgiveness content alone, apart from therapist skills or commitment, we would be engaging in futility. We can only imagine the results if we placed a knowledgeable person who is skeptical about or disinterested in forgiveness into the therapy room armed with a forgiveness treatment manual.

The size of the sample also needs to be taken into consideration. Although results proved significant, a larger sample size would have provided a more solid assessment of the differences between groups. The lack of blind assessments as a result of the experimenter being the intervener is another limitation as is the possibility of intervener or therapist expectations. Future studies may want to use therapists who are not part of the research project. In this way, the possibility of experimenter influence as a cause of significant results is less likely and subtle therapist influence can be controlled.

Finally, we acknowledge the scientific weakness of not replicating the results with a different therapist. Replication will necessitate 3 years of work at the very least. In addition to the obvious time necessary to recruit clients and train therapists, the researcher must commit more than 1 full year to help the experimental group, another year to help the control group that begins the therapy, and perhaps many more months in data analysis. The time-consuming nature of this particular program and the consistently strong outcome effects are compelling reasons to report these initial results now rather than waiting years.

The results of this study are encouraging. The findings suggest that forgiveness can be psychologically beneficial for incest survivors, a group of individuals who have experienced one of the deepest hurts possible. Future studies may want to use a therapeutic intervention for the control group. This would be a strong test of the forgiveness model.