Development of a forgiveness scale, The

development of a forgiveness scale, The

Hargrave, Terry D

This paper reports on the development, validity, and reliability of a self-report instrument designed to assess a respondent’s perspective of pain resulting from relational violations and work toward relational forgiveness based on a framework proposed by Hargrave (1994a). Presented here is the,ive-stage procedure used in the development of the Interpersonal Relationship Resolution Scale. Construct validity and reliability were determined from an initial sample of 164 subjects. Concurrent validity of the scale was supported by another sample of 35 respondents who took the Interpersonal Relationship Resolution Scale, the Personal Authority in the Family System Questionnaire, the Relational Ethics Scale, the Fundamental Interpersonal Relations Orientation-Behavior scale, and the Burns Depression Checklist. Finally, a predictive validity study of the scale was performed with a clinical and nonclinical sample of 98 volunteers. Data are presented that support the validity and reliability of the instrument, as well as the final version of the scale.

The concept of forgiveness has been used since antiquity in the religious community as an essential factor in healing and restoring relationships between people (Hargrave,1994a). Recently, the psychotherapy literature has reflected a growing interest among clinicians in using forgiveness as an intervention to help families and individuals seek new beginnings in previously harmful relationships, resolve long-standing relational problems, and release anger and bitterness (DiBlasio & Proctor, 1993). Despite numerous informal or partial theoretical frameworks found in the family literature, DiBlasio and Proctor (1993) state that there is very little research on the phenomenon of forgiveness. One of the theoretical frameworks on forgiveness to emerge recently, reported by Hargrave (1994a,1994b), deals with the use of forgiveness in salvaging and restoring family relationships. Like most work in the area of forgiveness, however, there are no reports of empirical testing of the framework. The purpose of this study was to develop a reliable scale which would validate the constructs in Hargrave’s (1994a) theoretical framework and facilitate appropriate instrumentation to be used in further research on the concept of forgiveness.

General Information

Forgiveness usually refers to releasing resentment toward an offender (Hargrave,1994a), restoring relationships and healing inner emotional wounds (DiBlasio & Proctor,1993), or releasing the person who caused an injury from potential retaliation (Fitzgibbons, 1986; Smedes, 1984). Only a few conceptual articles have been published on the clin;ical use of forgiveness in journals that are not religiously oriented (e.g., Enright, Santos, & Al-Mabuk, 1989; Fisher, 1985; Fitzgibbons, 1986; Flanigan, 1987; Hope, 1987; Joy, 1985; Kaufman, 1984; Madanes, 1991; Schneider, 1989; Wolberg, 1973; Worthington & DiBlasio, 1990). Forgiveness is reported to be an effective intervention regarding problems stemming from anger and depression (Fitzgibbons,1986), family-of-origin issues (Hope,1987), sexual abuse and compulsions (Madanes, 1991; Schneider, 1989), personality disorders (Fisher, 1985; Wolberg, 1973), guilt (Joy, 1985), drug abuse (Flanigan, 1987), and broken marital relationships (Worthington & DiBlasio,1990). In addition, the ability of adolescents to forgive has been found to be associated with moral development (Enright et al., 1989).

Hargrave ‘s Forgiveness Framework

Hargrave (1994a, 1994b) conceptualizes the issues of relational pain from a contextual family therapy perspective (e.g., Boszormenyi-Nagy,1987; Boszormenyi-Nagy & Krasner, 1986; Boszormenyi-Nagy & Spark,1984). From the perspective of contextual family therapy, relationships exist in four dimensions: (a) facts, (b) individual psychology, (c) family or systemic transactions, and (d) relational ethics (Boszormenyi-Nagy & Krasner,1986). The fourth dimension, relational ethics, deals with the subjective balance of justice, trustworthiness, loyalty, merit, and entitlement between members of a relationship and is the most powerful and potentially the most therapeutic dimension of the family (Boszormenyi-Nagy & Krasner,1986). Relational ethics is rooted in the idea that people have an innate sense of justice that demands balance between what they are entitled to receive from a relationship and what they are obligated to give in order to maintain relational existence. When people engage in relationships that have a balance of give (obligations) and take (entitlements) over a period of time, the innate sense of justice is satisfied and trustworthiness is established in the relationship. However, when there is a consistent or severe imbalance between the relational give and take, the sense of justice is violated and individuals feel cheated or overbenefited by the relationship. The resulting lack of trust drives individuals to destructive entitlement (Boszormenyi-Nagy & Krasner, 1986), or the self-justifying effort to secure just entitlement. Destructive entitlement can manifest itself in many ways, including paranoid attitudes, hostility, rage, emotional cutoffs, and destructive harm to other individuals (Boszormenyi-Nagy & Krasner, 1986).

Hargrave (1994b) believes that it is this destructive entitlement that results in family pain and hurt. As Hargrave (1994a) states:

The justice and trust that is so necessary for a balanced relational ethic is violated, and family victims of the destructive behavior are left to draw painful conclusions:

The people on whom they are most dependent in the world cannot be trusted. (p. 341)

Using this theoretical construct, Hargrave (1994b) believes that individuals who are victims of family pain will transform the violations of love and trust into feelings concerning themselves and actions in future relationships. As Figure 1 illustrates, when individuals are violated, they are likely to feel (a) rage as they experience uncontrolled anger toward their victimizer, or (b) shame as they accuse themselves of being unlovable and not deserving of a trustworthy relationship. Similarly, violated individuals are likely to act in future relationships in ways that are (a) overcontrolling as they try to minimize their risk of hurt or (b) chaotic since they assume that little can be done to form trusting relationships and that they will eventually be hurt despite any effort. Some individuals who are victims of family pain experience a wide range of feelings and actions as they alternate in shame/rage, control/chaotic cycles (Hargrave, 1994b).

Contextual family therapy would maintain that victims of family injustice who are not compensated from the family of origin are likely to play out destructive entitlement in innocent relationships-specifically with spouses and children (Boszormenyi-Nagy & Krasner, 1986). In other words, victims of relational injustice are likely to become victimizers within other relationships. Hargrave (1994a) maintains that efforts aimed at forgiveness, such as the release of blame and reconciliation, are therapeutic opportunities to deal successfully with destructive entitlement and to heal family relationships.

The work of forgiveness, as outlined by Hargrave (1994b), is defined as effort in restoring love and trustworthiness to relationships so that victims and victimizers can put an end to destructive entitlement. The framework maintains that the work of forgiveness in families fits into two broad categories of exonerating and forgiving. Neither category is inherently better than the other, but the two divisions do make different demands on the relationship between the victim and victimizer. Hargrave (1994a) maintains that exonerating deals more with the internal processing of the victim and his or her ability to gain insight into how pain was perpetrated and to establish some form of identification with the victimizer. Forgiving, on the other hand, demands that the victim and victimizer involve themselves in interactions which work toward the establishment of love and trust in the relationship.

Exonerating is defined as the effort of a person who has experienced injustice or hurt to lift the load of culpability of the person who caused the hurt (Hargrave, 1994a) Hargrave (1994b) maintains that exoneration is achieved by insight and understanding. Insight refers to the ability of a person to objectify the mechanisms of family pain that have caused relational damage (Hargrave, 1994a). As the individual identifies these mechanisms and transactions which cause pain, he or she has an increased ability to block the transaction and stop relational damage from occurring in the future. Understanding involves identifying with the victimizer’s position, limitations, development, efforts, and intent. This understanding results in the victim acknowledging the fallibility of the victimizer, although it does not remove the victimizer’s responsibility for the destructive action. As understanding takes place, an individual feels a reduction in condemnation and blame toward his or her victimizer.

Hargrave (1994a) maintains that forgiving differs from exonerating in that forgiving requires some specific action regarding the responsibility for the injustice which caused the hurt. In a contextual framework, relational justice in a family demands that the person who is victimized and hurt is reasonable in holding the wrongdoer responsible for the hurt. Trust in the relationship has been damaged. In forgiving, the victimized person is given reason to believe that the wrongdoer accepts responsibility for the injustice he or she caused and promises to act trustworthy in the future. The relationship is re-established because trust has been restored. In Hargrave’s framework, forgiving is accomplished by giving the opportunity for compensation and through an overt act of forgiving. In giving the opportunity for compensation, the victim allows the victimizer to rebuild the status of trust in the relationship in a progressive manner by acting in ways that are trustworthy. In an overt act of forgiving, victim and victimizer discuss the relational violation openly and come to an agreement that they will seek a new trustworthy relationship in the future (Hargrave,1994a).

In sum, this forgiveness framework includes two broad divisions of exonerating and forgiving. Exonerating has two stations of insight and understanding; forgiving has two stations of giving the opportunity for compensation and the overt act of forgiving. This framework is illustrated in Figure 2. Hargrave (1994b) believes that both exonerating and forgiving are appropriate in different relationships at different times. Hargrave (1994b) further points out that the four stations should not be interpreted as stages. People oscillate between stations many times in the effort to forgive and re-establish relational trust (Hargrave, 1994a).

Rationale for Scale Development

In developing a forgiveness scale based on the framework outlined above, the authors wished to utilize a self-report instrument which would indicate not only individuals’ work in forgiveness as it applied to the four stations, but also the manifestations of individual pain as reflected in the characteristics of shame, rage, control, and chaos. Therefore, we developed a self-report instrument of two scales and eight subscales that measured individual work in forgiveness along the four identified constructs (insight, understanding, giving the opportunity for compensation, overt act of forgiving) and manifestations of pain (shame, rage, control, chaos). It should be pointed out here that much of the work of forgiveness as outlined by Hargrave in giving the opportunity for compensation and overt acts of forgiving are relational in nature. In developing a self-report instrument, therefore, only the victim’s perceptions of the interactions in the relationship dealing with these two constructs could be measured. The instrument was named the Interpersonal Relationship Resolution Scale (IRRS).

METHODS

A five-stage procedure was selected for developing the IRRS. In Stage 1, definitions were developed for the different constructs of forgiveness and manifestations of pain. In Stage 2, items were generated for use in the IRRS. A preliminary IRRS was tested in Stage 3 to determine reliability and construct validity. In Stage 4, concurrent validity of the revised IRRS was determined by correlating it with the Personal Authority in the Family System Questionnaire, the Relational Ethics Scale, the Fundamental Interpersonal Relations Orientation – Behavior Scale, and the Burns Depression Checklist. Finally, in Stage 5, a revised IRRS based on the preliminary results was tested with a clinical and nonclinical sample to determine predictive validity.

Stage 1: Definitions

The four forgiveness constructs and four manifestations of pain were identified by a review of the literature concerning Hargrave’s framework (Hargrave, 1994a, 1994b; Hargrave & Anderson, 1992). Definitions were formulated from composites of these writings and then reviewed by Hargrave, who was the final judge in determining face validity. The following definitions were adopted for this study.

Insight Insight is the ability to recognize transactional patterns and mechanisms by which relational damage was perpetrated and interrupt or intervene in the patterns or mechanisms to prevent relational damage in the future. Low involvement with this construct would indicate that a person would experience confusion concerning emotional pain and would be unable to make specific statements regarding painful interactions or how to avoid them. High involvement with this construct would indicate that a person experiences clarity in identifying pain and knows how to avoid painful interactions.

Understanding. Understanding is the ability to identify with the position, limitations, development, efforts, and intent of the person who caused relational damage. Low involvement with this construct would indicate that the person blames either him/herself or the perpetrator to an unreasonable degree without consideration of context or circumstances. High involvement with this construct would indicate that a person clearly understands the circumstances which must be considered in determining responsibility for relational deterioration.

Giving the opportunity for compensation. Giving the opportunity for compensation is the ability to engage in interactions and relationship with the former perpetrator in a way that is perceived by the victim as nonthreatening and builds emotional bonding. Low involvement in this construct would indicate that a person views a continued relationship with the perpetrator as nontrustworthy and that interactions are marked with communications that cause pain. High involvement in this construct would indicate that a person perceives that he or she is able to engage in a relationship with the perpetrator that promotes reasonable care and desires to continue the relationship in the future.

Overt act of forgiving. An overt act of forgiving is the perceived ability of a person to discuss past relational damage with the perpetrator and resolve issues of responsibility for specific violations to the point where the relationship can be secure and trustworthy. Low involvement with this construct would indicate that a person perceives him or herself as unsuccessful in discussing and resolving the relational damage. High involvement with this construct would indicate that a person perceives him or herself as being successful in overtly discussing the relational damage with the perpetrator and that a greater sense of trust resulted from the discussion.

Shame. Shame is the degree to which an individual internalizes painful or undesirable experiences. Shame is a global measure which assesses the overall manifestation of personal guilt. Low involvement with this construct would likely mean that the individual is comfortable with mild levels of confrontation and is secure with self. High involvement with this construct would indicate that the individual experiences excessive guilt and internalizes emotions which indicate self is unacceptable.

Rage. Rage is the degree to which an individual externalizes painful or undesirable experiences. Rage is a global measure which assesses internal feelings of anger and actions which are manifestations of anger. Low involvement with this construct would indicate that the individual does not express anger in an overt manner. High involvement with this construct would indicate that the individual expresses anger and resentment in external ways.

Control. Control is the degree to which an individual seeks to administer life to avoid or deal with situations. Control is a global measure which assesses overall effort in managing life. Low involvement with this construct reflects a relaxed style of conducting activities and relationships. High involvement with this construct reflects an authoritarian style of dealing with life goals or relationships.

Chaos. Chaos is the degree to which an individual seeks to avoid organization or responsibility in dealing with situations. Chaos is a global measure which assesses overall failure to manage life successfully. Low involvement with this construct would likely reflect a balanced effort in organizing life goals and being considered responsible. High involvement with this construct would likely reflect an inability to organize and manage life goals and relationships.

It is important to note here that the four manifestations of pain as defined by Hargrave (1994b) are not diametrically opposed. For instance, individuals experiencing low involvement with the construct of control do not necessarily experience high involvement with the construct of chaos. Likewise, it is possible for individuals who experience pain to display, at times, behavior in cycles of shame/rage and control/chaos (Hargrave, 1994b;. Therefore, these definitions reflect separate constructs from which subscales for the IRRS were developed.

Stage 2: Statement Formulation

Using the above definitions, the authors separately generated items to be used in the preliminary IRRS. We reviewed the items to ensure that there was adherence to the definitions. After discussion, we selected 162 items for inclusion in the preliminary IRRS. The preliminary IRRS consisted of a Forgiveness Scale which included the four subscales of Insight, Understanding, Giving the Opportunity for Compensation, and Overt Act of Forgiving, and a Pain Scale which included the four subscales of Shame, Rage, Control, and Chaos.

The items were written so that the IRRS would reflect clear extremes on each construct. Also, responses to the items were worded in such a way to produce a forced choice (yes or no) among respondents. In this way, the IRRS would reflect univariate linear scores on the eight subscales. The items that were used in the final version of the IRRS are contained in the Appendix.

Stage 3: Testing of the Preliminary IRRS

In order to determine construct validity and reliability of the preliminary IRRS, the scale was administered to a sample number of subjects. Construct validity, as used in this study, is the degree to which a test measures its intended hypothetical construct or nonobservable trait (Gay, 1981).

Subjects. A total of 164 participants completed the preliminary IRRS. The ages of the subjects ranged from 19 to 77 years (SD = 12.03, M = 31.92). The participants were recruited primarily from graduate and undergraduate psychology and counseling classes and various occupational sites, including a bank, a television station, and a construction site. A total of 118 females (71.06%) and 46 males (28.04%) participated in the initial testing. The ethnicity of the sample was 3.7% Black, 7.9% Hispanic, 6.7% Asian, and 81.7% White. The martial status of the sample was as follows: 42.8% married, 41.5% never married, 12.3% divorced, 3% re-married, and .4% widowed.

Factor analysis results. A factor analysis using a varimax rotation revealed that many items loaded on the eight constructs of the two scales. We then used the loading information to select items to be used in a final version of the IRRS. Items which loaded on only one construct at or above the .45 level were considered for possible use in the final version of the scale. Tables 1 and 2 show the loadings of the 44 items that were eventually used in the final version of the IRRS.

The items in the Forgiveness Scale accounted for 56.1% of the variance; the items in the Pain Scale of the IRRS accounted for 47.4% of the variance. The subscales of Giving the Opportunity for Compensation in the Forgiveness Scale and Rage in the Pain Scale accounted for the largest amount of variance. Eigenvalues and variance percentages are found in Table 3.

Reliability. A Cronbach’s alpha analysis was performed on the two scales and conesponding subscales of the IRRS using the items selected for the final version of the instrument. Reliability for the Forgiveness Scale and Pain Scale was computed at .92 and .95, respectively. Reliability for the Forgiveness Scale subscales was as follows: Insight, .85; Understanding, .78; Giving the Opportunity for Compensation, .86; Overt Act of Forgiving, .63. Reliability for the subscales of the Pain Scale was as follows: Shame, .74; Rage, .87; Control, .78; Chaos, .82.

A Pearson correlation coefficient analysis performed on the data from the initial testing of the instrument revealed that each scale of the IRRS had high, positive correlation with its corresponding subscales. There was, however, no significant correlation between the Forgiveness Scale and Pain Scale (r = -.15), indicating that the two scales measured different phenomena.

Stage 4: Concurrent Validity Test of the IRRS

One of the accepted methodologies of determining validity is to measure the degree to which the scores of the instrument are related to the scores of another instrument with established reliability and validity. This measure of validity is called concurrent validity (Gay, 1981). In order to determine concurrent validity of the IRRS, the revised instrument along with the Personal Authority in the Family System Questionnaire, the Relational Ethics Scale, the Fundamental Interpersonal Relations Orientation – Behavior Scale, and the Burns Depression Checklist were administered to sample of subjects.

Other instruments. The Personal Authority in the Family System Questionnaire (PAFSQ) is a 132-item instrument which measures eight nonoverlapping constructs of spousal intimacy, spousal fusion/individuation, intergenerational fusion/individuation, intergenerational intimacy, nuclear family triangulation, intergenerational triangulation, intergenerational intimidation, and personal authority (Bray, Williamson, & Malone, 1984a, 1984b). All items are rated on a 5-point Likert scale. Reliability for the PAFS-Q and subscales ranged from .82 to .95 with a mean of .90 (Bray et al., 1984a). Since the PAFSQ measures constructs dealing with issues of family interaction which are likely to result in either intimacy or pain, it was believed to be a sound concurrent validity instrument for the IRRS. Items on the PAFS-Q are scaled such that larger scores are reflective of the desirable characteristic or interaction. It was expected that some of the constructs of the Forgiveness Scale of the IRRS would have significant correlations with the PAFS-Q, reflecting that work and progress in the area of forgiveness are related to positive interactions and relationships with family. It was also expected that some of the constructs of the Pain Scale of the IRRS would have significant correlations with the PAFS-Q, reflecting that high manifestations of pain are related to negative interactions and relationships with family.

The Relational Ethics Scale (RES) is a 24-item instrument which measures the constructs of trust and justice, loyalty, and entitlement on two subscales relating to the family in which one was raised and one’s relation with a person of equal status (Hargrave & Bomba, 1993). All items are rated on a 5-point Likert scale. Reliability for the RES was established at .96; the two subscales had a reliability of .93 and .96 (Hargrave, Jennings, & Anderson, 1991). It was expected that the constructs of the Forgiveness Scale of the IRRS would have positive correlations with the Vertical Trust/Justice and Horizontal Trust/Justice subscales of the RES since these constructs involve family trust issues which are identified as essential to the work of forgiveness. Also, it was expected that the constructs of the Pain Scale of the IRRS would have negative correlations with the subscales of the RES since low scores on the RES would reflect some type of family violation.

The Fundamental Interpersonal Relations Orientation – Behavior Scale (FIRO-B) is a 54-item instrument which measures the constructs of expressed and wanted inclusion, control, and affection. The FIRO-B has two sets of response categories on a 6-point Likert scale. Although its validity as a measure has been increasingly questioned, the FIRO-B remains one of the more popular psychometric instruments (Fisher, Macrosson, & Walker, 1995). A variety of reliability data are available for the FIRO-B, but the author of the scale established the reliability of the expressed and wanted facets of inclusion, control, and affection in a range from .93 to .94 (Schutz, 1958). Of interest in the concurrent validity study were the control subscales of the FIRO-B. It was expected that the construct of control in the IRRS would have a positive correlation with the control subscales of the FIRO-B.

The Burns Depression Checklist (BDC) is a IS-item instrument which measures subject response to common symptoms of depression such as sadness, low self-esteem, and hopelessness (Burns, 1994). All items are rated on a 4-point Likert scale with higher scores reflecting more severe levels of depression. The author of the scale reports that the BDC was validated in an unpublished study correlating the instrument with the Beck Depression Inventory. Although the author reports that the correlation between the two measures was high, no validity or reliability information is available. We correlated the BDC and Beck Depression Inventory with a sample of 45 adults. Reliability of the BDC was .91, and the instrument correlated with the Beck Depression Inventory at the .85 level (p

Subjects. A total of 35 volunteer subjects participated in the concurrent validity study of the IRRS. The ages of the subjects ranged from 19 to 55 years (SD = 12.03, M=: 25). The subjects were recruited primarily from graduate and undergraduate psychology and counseling classes. A total of 23 females (65.8%) and 12 males (34.2%) participated in the study. The ethnicity of the sample was 17.15% Black, 20% Hispanic, 17.15% Asian, and 46.7% White. The martial status of the sample was as follows: 37.15% married, 34.29% never married, 11.43% divorced, 14.28% remarried, and 2.85% widowed.

Concurrent validity results. A Pearson correlation coefficient analysis was performed on all the instruments and applicable subscales. These correlations are reported in Table 4. Several of the IRRS Forgiveness Scale subscales had many expected correlations with the PAFS-Q. Insight was strongly correlated with Personal Authority, reflecting perhaps that individuals with high scores on the Insight subscale were comfortable setting boundaries to protect individual stances. Insight also had a strong correlation with Nuclear Family Triangulation, indicating that high insight reflected good nuclear family boundaries. The subscale of Understanding had a negative correlation with Spousal Fusion and Spousal Intimacy, suggesting that increased understanding of a violator causes more of a sense of individuation and less intimacy with a spouse. Understanding also correlated with Intergenerational Intimidation, indicating that as an individual experiences less intimidation and fear, understanding increases. Both Giving the Opportunity for Compensation and the Overt Act of Forgiving had strong correlations with Intergenerational Intimacy, indicating that as relational forgiveness progresses, intimacy is enhanced.

The Pain Scale had significant correlations to several of the subscales of the PAFS-Q. Both Shame and Rage had significant negative correlation with Intergenerational Triangulation, suggesting that individuals with higher levels of triangulation between parents also feel significant levels of these types of pain. Intergenerational Intimidation had negative correlations with the Pain Scale, perhaps indicating that when intimidation is present, individual pain is also evident. In addition, the Pain Scale correlated with Spousal Intimacy, indicating that individuals with these types of pain also experience higher levels of intimacy with a spouse.

Only two of the Forgiveness Scale subscales of the IRRS had significant correlations with the RES. Both Understanding and the Overt Act of Forgiving had a negative correlation with Vertical Entitlement, suggesting that higher scores on these Forgive]less Scale subscales indicate that an individual feels less violation from the family of origin. The Pain Scale of the IRRS had several subscales which correlated with those of the RES. Most notably, the subscale of Shame had negative correlations with all subscales of the RES except Vertical Loyalty, indicating that the absence of trust and justice in vertical and horizontal relationships corresponds to high levels of individual shame. Rage had negative correlations with Vertical Entitlement and Horizontal Trust/Justice, suggesting that individuals who have high scores on the Rage subscale also perceive violations from their families of origin and have low justice and trust in horizontal relationships.

The IRRS and the FIRO-B had several correlations between constructs and subscales. Most notable and most important for this validity study was the high correlation between the IRRS subscale of Control and the Expressed Control subscale of the FIRO-B, indicating that both reveal strong expressed control behavior. Giving the Opportunity for Compensation had negative correlations with both Wanted and Expressed Control subscales, suggesting that individuals who are engaged in relational reconciliation in this manner neither desire to be controlled nor want to control others. The negative correlation of Rage and Shame to the Wanted Affection subscale of the FIRO-B indicates that the presence of these types of pain is not accompanied by desired affection from others.

Three of the four constructs of the Pain Scale of the IRRS had significant correlations with the BDC. Shame, rage, and chaos all had high correlations with the BDC, indicating that these types of pain are also accompanied by depression. A negative correlation was found between Understanding and the BDC, suggesting that depression is not present when there are high Understanding scores.

Stage 5: Predictive Validity Test of the IRRS

Predictive validity is the degree to which a test or instrument can predict information concerning an individual (Gay, 1981). In order to test the predictive validity of the IRRS, the instrument was administered to a sample representing a clinical population which had experienced some type of family violation which caused pain and a nonclinical general population. There were two hypotheses in Stage 5. The first was that the clinical population would express significantly lower scores than the nonclinical group on all constructs of the Forgiveness Scale of the IRRS, indicating that the clinical group experienced less progress in the work of forgiveness. The second hypothesis was that the clinical group would record significantly higher scores than the nonclinical population on all constructs of the Pain Scale of the IRRS, indicating higher levels of pain.

Subjects. A total of 98 volunteer subjects participated in the concurrent validity study of the IRRS. Thirty-five of the subjects represented the clinical group and were recruited from therapy situations in which the presenting problem of the client was past sexual abuse, physical abuse, or family-of-origin issues. Sixty-three of the subjects represented the nonclinical group and were recruited from graduate and undergraduate psychology and counseling classes. The nonclinical group was recruited from a pool of subjects who had never participated in individual or family therapy. The ages of the subjects ranged from 19 to 61 years (SD = 9.6, M = 28). A total of 68 females (69.4%) and 30 males (30.6%) participated in the study. The ethnicity of the sample was 3.0% Black, 15.2% Hispanic, 13.1% Asian, and 68.7% White. The martial status of the sample was as follows: 46.4% married, 30.9% never married, 7.2% divorced, 14.4% re-married, and 1.1% widowed.

Predictive validity test results. A multiple analysis of variance (MANOVA) was used to determine whether there was significant difference between the scores of the clinical and nonclinical populations. The omnibus MANOVA produced significant Pillais at the .01 level, with an F(8, 89) = 13.16, p

DISCUSSION

This paper reports on the development of a scale to assess an individual’s perspective on (a) the work of forgiveness according to the constructs of insight, understanding, giving the opportunity for compensation, and the overt act of forgiving; and (b) the manifestations of pain according to the constructs of shame, rage, control, and chaos. Based on the fivestage procedure used in this study, the IRRS shows significant construct validity based on the factor analysis performed on the initial sample of 164 subjects. In addition, the IRRS has strong reliability, and the subscales have acceptable levels of reliability as assessed by Cronbach’s coefficient of internal consistency. Many of the IRRS constructs correlate in the expected direction with other family instruments (PAFS-Q and RES) as well as with an interpersonal instrument (FIRO-B) and a depression scale (BDC). In terms of predictive validity, the IRRS successfully discriminated between a clinical and nonclinical population. From this study, we can conclude that the IRRS is a reliable and valid instrument and an accurate measure of the forgiveness framework proposed by Hargrave (1994a).

One of the problems with creating an instrument that measures the work of forgiveness is that forgiveness is likely a process which involves a great deal of time and contains elements of both inter- and intrapersonal processing. It is conceivable, for example, that an individual could participate in overt forgiveness without ever achieving any level of understanding. Therefore, the subscale scores of Insight, Understanding, Giving the Opportunity for Compensation, and Overt Act of Forgiving may have more clinical use than the Forgiveness Scale total. In the same manner, Hargrave (1994b) points out that individuals experience a wide range of feelings and actions that may alternate between the extremes of rage and shame, control and chaos. Therefore, it is possible that an individual will score high in all four pain constructs.

The IRRS has impressive levels of reliability for both the Forgiveness Scale and the Pain Scale. Although the subscales of Overt Act of Forgiving and Shame have marginal levels of reliability, they are acceptable for new scale development (Gay, 1981). The subscale of Overt Act of Forgiving has a weaker level of reliability than the other subscales. Some of the items in this subscale focus on the victimizer apologizing for causing pain; other items focus on the victimizer accepting responsibility. It is possible that these two focuses represent different steps in the forgiving process and therefore are not necessarily representative of the same construct. Further study of this construct of forgiveness is warranted, and more items representative of this construct need to be generated to improve reliability.

Although the IRRS showed many of the expected correlations with the PAFS-Q, FIROB, and BDC, it did not have the expected correlations with the RES. The Vertical and Horizontal Trust/Justice subscales should measure the more essential elements and aspects of relational balance and trust. The constructs of the Forgiveness Scale of the IRRS are also directed at uncovering perceptions of relational trust, so significant correlations would have been expected. This fact warrants further study of the construct of trust from a contextual perspective. One of the limitations of this study was the small number of participants in the concurrent validity study. More participants in another concurrent validity study should yield additional possible explanations for correlations between the IRRS and other instruments and directions for the future development of the IRRS.

The power of the constructs to discriminate between clinical and nonclinical populations suggests that the IRRS is a good measure of Hargrave’s forgiveness framework. Individuals experiencing high scores reflective of pain from relational violations score lower in the work of forgiveness. Also, the low correlations between the Forgiveness Scale and Pain Scale of the IRRS clearly indicate that the subscales measure separate phenomena. In addition, the strong correlations of shame, rage, and chaos with the BDC may give clinicians insight into differentiating the distress of depression into specific pain. The IRRS, therefore, may be particularly useful to therapists working with individuals having difficulty with relational violations in assessing the potential for the work of forgiveness and the level of the client’s distress. The scoring technique used in this study, however, may limit clinical interpretations of scores since positive answers were given a value of one and negative responses were given a value of zero. Further consideration should be given in the scale development to creating a scoring technique that will yield clearly useful clinical information. Since the constructs of forgiveness according to Hargrave’s framework have been validated, the IRRS may also be useful as an empirical research tool in exploring the effectiveness of therapy aimed at forgiveness and individual pain resolution resulting from the work of forgiveness.

Forgiveness is a complex concept and resulting distress and pain from relational violation is a broad subject. The IRRS is a first attempt to identify these constructs in a brief and useful manner. However, it is important to note that this model of forgiveness is but one illustration of the process (Hargrave, 1994a) and that other substantial constructs dealing with forgiveness and pain are likely to exist. The IRRS, therefore, is neither comprehensive nor exhaustive as a clinical or research instrument.

Additional research is warranted to provide further validation for forgiveness constructs. Also, the effects of such variables as ethnicity, gender, marital status, family background, and severity of violating event on IRRS scores are yet to be determined. Larger respondent sets are needed to set clinical norms to provide specific meaning to scale scores. In spite of these limitations, however, the IRRS shows promise as an instrument which validates a forgiveness model.

REFERENCES

Boszormenyi-Nagy, I. (1987). Foundations of contextual therapy: Collected papers. New York: Brunner/Mazel.

Boszormenyi-Nagy, IL, & Krasner, B. (1986). Between give and take: A clinical guide to contextual therapy. New York: Brunner/Mazel.

Boszormenyi-Nagy, I., & Spark, G. (1984). Invisible loyalties. New York: Brunner/Mazel. Bray, J. H., Williamson, D. S., & Malone, P. E. (1984a). Personal authority in the family system: Development of a questionnaire to measure personal authority in intergenerational family processes. Journal of Marital and Family Therapy, 10, 167-178.

Bray, J. H., Williamson, D. S., 8 Malone, P. E. (1984b). Manual for the personal authority in the family system questionnaire. Houston, TX: Houston Family Institute. Burns, D. (1994). The therapist’s toolkit: Comprehensive treatment and assessment tools for mental

health professionals. (Available from David Burns, 11987 Morietta Ln., Los Altos Hills, CA 94022.)

DiBlasio, F. A., & Proctor, J. H. (1993). Therapists and the clinical use of forgiveness. The American

Journal of Family Therapy, 21, 175-184.

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

Fisher, S. F. (1985). Identity of two: The phenomenology of shame in borderline development and treatment. Psychotherapy, 22, 101-109.

Fisher, S. G., Macrosson, W. D. K., & Walker, C. A. (1995). FIRO-B: The power of love and the love of power. Psychological Reports, 76, 195-206.

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

Flanigan, B. J. (1987). Shame and forgiveness in alcoholism. Alcoholism Treatment Quarterly, 4, 181-195.

Gay, L. R. (1981). Educational research: Competencies for analysis and application (2nd ed.). Columbus, OH: Charles E. Merrill.

Hargrave, T. D. (1994a). Families and forgiveness: Healing intergenerational wounds. New York: Brunner/Mazel.

Hargrave, T D. (1994b). Families and forgiveness: A theoretical and therapeutic framework. The Family Journal, 2, 339-348.

Hargrave, T. D., & Anderson, W. T. (1992). Finishing well: Aging and reparation in the intergenerational family. New York: Brunner/Mazel.

Hargrave, T. D., & Bomba, A. K. (1993). Further validation of the relational ethics scale. Journal of Marital and Family Therapy, 19, 292-299.

Hargrave, T. D., Jennings, G., & Anderson, W. (1991). The development of a relational ethics scale. Journal of Marital and Family Therapy, 17, 145-158. Hope, D. (1987). The healing paradox of forgiveness. Psychotherapy, 24, 240-244. Joy, S. S. (1985). Abortion: An issue to grieve? Journal of Counseling and Development, 63, 375376.

Kaufman, M. E. (1984). The courage to forgive. Israel Journal of Psychiatry and Related Sciences, 21, 177-187.

Madanes, C. (1991). Sex, love and violence: Strategies for transformation. New York: W. W. Norton. Schutz, W. C. (1958). FIRO: A three-dimensional theory of interpersonal behavior. New York: Reinhart.

Schneider, J. P. (1989). Rebuilding the marriage during recovery from compulsive sexual behavior. Family Relations, 38, 288-294.

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

and Row.

Wolberg, A. R. (1973). The borderline patient. New York: Intercontinental Medical Book Company. Worthington, E. L., & DiBlasio, F. A. (1990). Promoting mutual forgiveness within the fractured relationship. Psychotherapy, 27, 219-223.

Terry D. Hargrave, PhD, is Associate Professor of Behavioral Studies, Amarillo College, PO Box 447, Amarillo, TX 79178.

James N. Sells, PhD, is Assistant Professor of Counseling, Northern Illinois University, DeKalb, IL 60115-2867.

Copyright American Association for Marriage and Family Therapy Jan 1997

Provided by ProQuest Information and Learning Company. All rights Reserved