Personality and Individual Differences 38 (2005) 627–638 www.elsevier.com/locate/paid
Letting bygones be bygones: further evidence for the validity of the Tendency to Forgive scale Ryan P. Brown *, April Phillips Department of Psychology, University of Oklahoma, 455 W. Lindsey Street, DHT #705, Norman, OK 73019, USA Received 8 December 2003; received in revised form 5 April 2004; accepted 6 May 2004 Available online 28 July 2004
Abstract The present study examined the validity of three putative measures of dispositional forgiveness, as well as a measure of trait rumination, with respect to measures of mental health and forgiveness for a specific offense. Two hundred undergraduates completed the Tendency to Forgive scale, the Attitudes Toward Forgiveness scale, the Transgression Narrative Test of Forgivingness, and a multi-dimensional measure of trait rumination, along with measures of depression, life satisfaction, self-esteem, agreeableness, neuroticism, attachment, and aggression. Participants also recounted a specific offense they had experienced and reported the extent to which they had forgiven their offender. Results strongly supported the discriminant validity of the Tendency to Forgive scale with regard to both mental health (depression and life satisfaction) and state forgiveness. Ó 2004 Elsevier Ltd. All rights reserved. Keywords: Dispositional forgiveness; Mental health; State forgiveness; Rumination
1. Introduction As research on interpersonal forgiveness progresses, an important question is whether there are stable individual differences in the propensity to forgive others, and if so, what the antecedents and consequences of such a disposition might be. Early research on this question (e.g., Gorsuch & Hao, 1993; Mauger et al., 1992) was mired in psychometric and conceptual problems (see Brown,
*
Corresponding author. Tel.: +1-405-325-4526; fax: +1-405-325-4737. E-mail address:
[email protected] (R.P. Brown).
0191-8869/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.paid.2004.05.017
628
R.P. Brown, A. Phillips / Personality and Individual Differences 38 (2005) 627–638
2003, for a review). But more recently, researchers have begun to re-examine the issue of dispositional forgiveness with greater theoretical rigor and a corresponding level of empirical progress (e.g., Berry, Worthington, Parrott, OÕConnor, & Wade, 2001; Brown, 2003; DeShea, 2003; Tangney, Boone, Dearing, & Reinsmith, 2002). Perhaps the most common method of measuring dispositional forgiveness involves presenting respondents with offense vignettes and asking them to report how they believe they would respond to each hypothetical situation (Berry et al., 2001; DeShea, 2003; Tangney et al., 2002). This approach represents an improvement over previous attempts to measure trait forgiveness, if only by limiting the number of constructs involved. As Berry et al. (2001) noted, such measures can be designed to be relatively brief and theoretically agnostic with regard to precise definitions of forgiveness. The latter may be a benefit insofar as forgiveness researchers have yet to arrive at a consensus on what the necessary and sufficient elements of forgiveness are (Enright, Freedman, & Rique, 1998; Fincham, 2000; McCullough, Hoyt, & Rachal, 2000). Such scenario measures also allow for the easy rotation of respondentsÕ perspectives, from that of victim to that of offender. Such measures thus have much to offer. In contrast, Brown (2003) attempted to capture individual differences in dispositional forgiveness by asking respondents to report their typical responses to past offenses on the Tendency to Forgive scale (TTF), rather than projecting themselves into hypothetical scenarios. As with scenario-based measures, initial research using the TTF has been promising. As described by Brown (2003), self-reports on the TTF are corroborated by assessments made by knowledgeable others. In addition, scores on the TTF predict the ease with which people can recall past offense experiences under time constraints, suggesting that individuals with high scores on the TTF experience lower degrees of offense accessibility. Consistent with previous theorizing (McCullough, 2000) and experimental evidence (Freedman & Enright, 1996), high scores on the TTF were also related to lower levels of depression. Of note, this latter result remained even after controlling for pro-forgiveness attitudes, a construct that previous researchers (e.g., Mauger et al., 1992) have sometimes conflated with the tendency to forgive. Finally, Brown (2003) reported that scores on the TTF were positively, though modestly, related to scores on Berry et al.Õs (2001) scenario-based measure, although the latter was more highly correlated with pro-forgiveness attitudes than it was with the TTF. Given the fledgling state of research on trait forgiveness, it seems prudent to ask whether the various measures being developed to assess dispositional forgiveness all tap the same construct and demonstrate equivalent predictive power for theoretically related constructs, such as mental health and state forgiveness. The purpose of the present study was to address these questions by examining the validity of the TTF relative to related measures. The primary alternative measures against which the TTF was compared included Berry et al.Õs (2001) Transgression Narrative Test of Forgivingness (TNTF), BrownÕs (2003) measure of pro-forgiveness attitudes (the Attitudes Toward Forgiveness scale, or ATF), and a domain-general, multi-dimensional measure of trait rumination (or TR; Scott & McIntosh, 1999). The TNTF was chosen to represent the dominant, scenario-based approach to assessing dispositional forgiveness. This choice was made on the basis of the simplicity of the TNTF and its degree of validation evidence, relative to other scenario-based measures. The ATF was included to contrast the disposition to forgive with the related, but distinct, construct of pro-forgiveness attitudes, or the belief that forgiveness is a valuable, virtuous act. As Brown (2003) contended,
R.P. Brown, A. Phillips / Personality and Individual Differences 38 (2005) 627–638
629
pro-forgiveness attitudes and the tendency to forgive should be associated, but they are theoretically and empirically distinguishable constructs that should not be equated with one another. Finally, the trait rumination measure was included to address our concern that all any measure of trait forgiveness assesses is the tendency to ruminate, rather than a unique and important construct specific to interpersonal relations (McCullough, Bellah, Kilpatrick, & Johnson, 2001). These dispositional measures were used in the present study to predict respondentsÕ scores on two commonly used indicators of mental health: depression and life satisfaction. In addition, we compared the utility of these measures in relation to participantsÕ reports of the extent to which they had actually forgiven an offender for a real offense that they had experienced recently. The natural variability inherent in these real-life offenses also allowed us to examine whether individual differences in forgiveness interacted with certain offense-specific variables–– specifically, offense severity, the presence of an apology, and pre-offense intimacy with the offender––to predict state forgiveness (see McCullough & Hoyt, 2002, for indirect evidence for this possibility). To provide further evidence of discriminant validity, we also included a host of other measures that might conceivably relate to state forgiveness––specifically, measures of the big five personality dimensions of neuroticism and agreeableness, which McCullough et al. (2001) have suggested might be integrally related to individual differences in forgiveness; self-esteem, which relationship research has shown to predict responses to betrayal (e.g., Murray, Holmes, MacDonald, & Ellsworth, 1998); the relationship attachment dimensions of anxiety and avoidance, which are developmentally rooted in interpersonal security beliefs and predict anger reactions in adults (e.g., Mikulincer, 1998); and a multi-dimensional measure of aggressiveness, which Berry et al. (2001) found was related modestly to their measure of dispositional forgiveness. The addition of these variables allowed not only for more rigorous tests of the uniqueness of the dispositional forgiveness construct in predicting state forgiveness, but also for an examination of the relations between these other variables and the focal predictor variables themselves (the TTF, the TNTF, the ATF, and TR).
2. Method 2.1. Participants Two hundred undergraduates (73 males, 127 females) from The University of Oklahoma participated in this study in partial fulfillment of a course requirement. The majority of the sample was Caucasian (76.5%); 8.5% were Asian, 5.5% were African American, 3.5% were Hispanic, 3.5% were Native American, and 2.5% were of other ethnicities. Because of incomplete data on certain variables for several participants, degrees of freedom for some analyses reported here differ slightly. 2.2. Measures and procedure Participants first completed a packet of attitude and personality scales, as well as the mental health measures. Within this packet, dispositional forgiveness was measured with the TTF, the
630
R.P. Brown, A. Phillips / Personality and Individual Differences 38 (2005) 627–638
ATF and the TNTF (Berry et al., 2001; Brown, 2003). The TTF is a 4-item scale on which participants report how they usually respond when someone offends them (e.g., ‘‘I tend to get over it quickly when someone hurts my feelings’’). The ATF consists of six items measuring participantsÕ general attitudes about the merits of forgiveness (e.g., ‘‘It is admirable to be a forgiving person’’). Participants indicated their level of agreement with each statement on the TTF and the ATF on a scale ranging from 1 (strongly disagree) to 7 (strongly agree). The TNTF is a scenariobased measure of dispositional forgiveness developed by Berry et al. (2001). This scale presents participants with five hypothetical scenarios and instructs them to indicate the extent to which they think they would forgive the offender if the events in the scenario happened to them. These projected reactions are made on a scale ranging from 1 (definitely not forgive) to 5 (definitely forgive). The TTF, the ATF, and the TNTF have all demonstrated reasonable internal reliability in previous research (Berry et al., 2001; Brown, 2003), as well as in the present study (as ¼ 0.73, 0.69, and 0.75 for the TTF, AFT, and TNTF, respectively). Trait rumination (TR) was measured with the 9-item Scott-McIntosh rumination inventory, a domain-general measure that is divided into three subscales: motivation, distraction and emotionality (Scott & McIntosh, 1999). Participants indicate how well each item describes them on a scale ranging from 1 (does not describe me well) to 7 (does describe me well). The motivation subscale consists of 3 items assessing the degree to which participants are motivated to reduce ruminative thoughts. The distraction subscale assesses the degree to which participants are prone to experience ruminative distractions. Finally, the emotionality subscale consists of 3 items assessing participantsÕ emotional reactions to ruminative thoughts. Because this measure focuses mostly on the tendency to ruminate about failed attempts to accomplish goals, it provides a straightforward measure of dispositional rumination that does not measure simply negative affectivity, aggressiveness, or low self-esteem, unlike some other measures of rumination (e.g., the Dissipation–Rumination scale––Caprara, 1986). All three subscales demonstrated acceptable reliability (as > 0.68). Neuroticism and agreeableness were measured with the Big Five Inventory (BFI; John, Donahue, & Kentle, 1991). The BFI is a 44-item, multi-dimensional personality inventory measuring openness, extraversion, conscientiousness, agreeableness, and neuroticism. Only the scores for neuroticism and agreeableness were of interest in the present study, as these dimensions have been posited to underlie individual differences in forgiveness (e.g., McCullough et al., 2001). Participants responded to each item on this scale by indicating the extent to which they agreed or disagreed with each statement on a scale ranging from 1 (strongly disagree) to 5 (strongly agree). Both the agreeableness and the neuroticism scales demonstrated acceptable reliability (a ¼ 0:78 and 0.84, respectively). Participants also completed Brennan, Clark, and ShaverÕs (1998) 36-item Avoidance and Anxiety measures of relationship attachment. Each dimension is assessed with 18 items with response scales ranging from 1 (strongly disagree) to 7 (strongly agree). These subscales have demonstrated excellent internal consistency in previous studies, and were acceptably reliable in the present study (a ¼ 0:92 for Anxiety and 0.71 for Avoidance). As a measure of global self-esteem, participants completed a modified version of the SelfAttributes Questionnaire (SAQ) developed by Pelham and Swann (1989). Participants were asked to rate themselves relative to other students of their sex and age on 8 dimensions (intelligence/ academic ability, social skills/social competence, artistic ability, musical ability, athletic ability,
R.P. Brown, A. Phillips / Personality and Individual Differences 38 (2005) 627–638
631
physical attractiveness, leadership ability and common sense). These ratings were made on a 10point scale (1 ¼ ‘‘way below average’’ to 10 ¼ ‘‘way above average’’). This scale demonstrated acceptable reliability (a ¼ 0:71). Aggression was assessed with Buss and PerryÕs (1992) Aggression Questionnaire (AQ), a commonly used, multi-dimensional measure of aggressiveness. This scale consists of four components: physical aggression, verbal aggression, anger, and hostility. Participants indicated the extent to which they agreed or disagreed with each statement on the AQ on a scale ranging from 1 (strongly disagree) to 5 (strongly agree). All four subscales (physical aggression, verbal aggression, anger, and hostility) demonstrated acceptable reliability (a ¼ 0:83, 0.72, 0.81 and 0.79, respectively). To assess mental health, we included the 20-item Center for Epidemiological Studies Depression scale (or CESD; Radloff, 1977), on which participants reported the frequency with which they had experienced a variety of depressive symptoms during the previous week. In addition, we also included the 5-item Satisfaction with Life scale (or SWL; Diener, Emmons, Larsen, & Griffin, 1985). For this scale, participants report their overall satisfaction with life by indicating the extent to which they agree or disagree with 5 statements on a scale ranging from 1 (strongly disagree) to 7 (strongly agree). Both mental health measures, which have been well-validated in previous research, demonstrated acceptable reliability (as ¼ 0.89). Participants completed the above scales in the following standard order: CESD, AQ, TNTF, BFI, SWL, TTF, ATF, TR, Anxiety/Avoidance, and SAQ. We also included several other measures unrelated to the present study that we do not discuss here. After completing these questionnaires, participants were asked to recall a time within the last 3 years when someone wronged them, mistreated them, offended them or betrayed them. They were instructed to give a brief description of this event and then answer several questions about it and the person who offended them. They also reported offense-specific information such as their levels of pre- and post-offense intimacy with the offender (1 ¼ ‘‘not at all close’’ to 7 ¼ ‘‘extremely close’’), and whether or not an apology had been given. Offense severity was measured with two items in which participants reported how hurtful and serious the event was. This 2-item scale demonstrated acceptable reliability (a ¼ 0:75). Offense severity ratings were made on a scale ranging from 1 (not at all hurtful/serious) to 7 (extremely hurtful/ serious). State forgiveness for the offense was measured with a 7-item scale that we designed for this study. Although McCullough et al. (1998) have recently validated a measure of state forgiveness, this measure (the Transgression-Related Interpersonal Motivations Inventory, or TRIM) was designed for assessing forgiveness within close relationships. Hence, some of the items are less appropriate than others for offenses committed by relative strangers. Thus, we incorporated the concepts of avoidance and retribution desires captured by the TRIM with several items in our measure, in addition to items that measured more general feelings of hostility that respondents had toward their offenders. A complete list of these items is provided in the Appendix. Participants indicated their level of agreement or disagreement with each statement on a scale ranging from 1 (strongly disagree) to 7 (strongly agree). Negative items were reverse scored so that higher scores indicate higher levels of forgiveness. When combined, these 7 items resulted in a scale with excellent internal reliability (a ¼ 0:91).
632
R.P. Brown, A. Phillips / Personality and Individual Differences 38 (2005) 627–638
3. Results 3.1. Mental health As shown in Table 1, scores on all three forgiveness measures (the TTF, ATF, and TNTF) were significantly correlated with each other. Scores on the TTF were also negatively correlated with the distraction and emotionality components of trait rumination, whereas scores on the TNTF were negatively correlated with the distraction and motivation components. In contrast, the ATF was not significantly related to any of the dimensions of trait rumination, although it was substantially correlated with both the TNTF and the TTF. The first set of criterion variables that we examined were indices of mental health: depression and life satisfaction. As can be seen in Table 1, both mental health measures were significantly related to all of our predictor variables, as well as to each other. However, the most pertinent analyses for the purposes of establishing the discriminant validity of the predictor variables were simultaneous multiple regression analyses, which we describe next. In the first regression analysis, we regressed depression scores on the TTF, the ATF, the TNTF, and the three components of TR. This analysis revealed a significant negative association between depression and the TTF, b ¼ 0:19, p < 0:01. In addition, all three components of TR were significantly associated with depression: b ¼ 0:14 for motivation, b ¼ 0:14 for distraction, and b ¼ 0:27 for emotionality, all ps < 0.05. Neither the ATF (b ¼ 0:01) nor the TNTF (b ¼ 0:11) were significant predictors, ps > 0.13. In the second regression analysis, we regressed life satisfaction scores on the same dispositional measures, with similar results. The TTF was once again a significant predictor of satisfaction, b ¼ 0:28, p < 0:001, as were the motivation and emotionality components of TR, bs ¼ 0.38 and )0.20, ps < 0.01. The distraction component of TR was marginally significant, b ¼ 0:11, as was the ATF, b ¼ 0:11, ps < 0.10. However, scores on the TNTF were not significantly related to satisfaction, b ¼ 0:02, p > 0:80. These analyses thus reveal that scores on the TTF demonstrate consistent discriminant validity with regard to mental health, vis- a-vis several alternative measures of forgiveness and trait rumination. Table 1 Relations among dispositional forgiveness measures, trait rumination, and mental health indices ATF TTF ATF TNTF TRmot TRdis TRem CESD
0.33
TNTF
0.40 0.51
TRmot
TRdis
0.07 0.10 )0.18
)0.15 )0.03 )0.19 )0.23
TRem
)0.28 )0.06 )0.10 )0.07 0.19
CESD
SWL
)0.35 )0.15 )0.25 )0.21 0.25 0.37
0.41 0.25 0.21 0.45 )0.27 )0.32 )0.56
Note: TTF ¼ Tendency to Forgive; ATF ¼ Attitudes Toward Forgiveness; TNTF ¼ Transgression Narrative Test of Forgivingness; TRmot ¼ motivation component of Trait Rumination; TRdis ¼ distraction component of Trait Rumination; TRem ¼ emotionality component of Trait Rumination; CESD ¼ Depression; SWL ¼ Satisfaction with Life. p < 0:05, p < 0:01.
R.P. Brown, A. Phillips / Personality and Individual Differences 38 (2005) 627–638
633
3.2. State forgiveness 3.2.1. Offense-specific variables On average, participants reported offenses that occurred within relatively intimate relationships (M ¼ 5:65, SD ¼ 1.42), and that were relatively severe (M ¼ 5:88, SD ¼ 1.13). The average time since the offense occurred was just over 1 year (M ¼ 60:12 weeks, SD ¼ 71.10). Participants reported receiving an apology in 56% of the offenses. Offenses reported by females involved relationships with higher levels of pre-offense intimacy (M ¼ 5:88) than those reported by males (M ¼ 5:23), F ð1; 198Þ ¼ 10:06, p < 0:01, and were rated as being more severe (M ¼ 6:08) than those reported by males (M ¼ 5:56), F ð1; 198Þ ¼ 10:19, p < 0:01. Consistent with previous research (Finkel, Rusbult, Kumashiro, & Hannon, 2002; McCullough & Hoyt, 2002; McCullough et al., 1998), state forgiveness was associated with reports of preoffense intimacy (r ¼ 0:18, p < 0:01), the severity of the offense (r ¼ 0:19, p < 0:01), and whether the offender had apologized (r ¼ 0:37, p < 0:01). However, state forgiveness was not associated significantly with time since the offense (r ¼ 0:02, ns). 3.2.2. Dispositional variables The means, by gender, of each dispositional measure are shown in Table 2, along with their correlations with state forgiveness. State forgiveness was positively correlated with all three measures of dispositional forgiveness (TTF, ATF and TNTF), not surprisingly. State forgiveness was also significantly correlated with agreeableness, the motivation component of rumination, and the verbal and hostility components of aggression. A multiple regression analysis on state forgiveness that included only the TTF, the ATF, the TNTF, and the three dimensions of TR revealed that only the TTF (b ¼ 0:29, p < 0:001) and the Table 2 Zero-order relations between state forgiveness and dispositional variables Dispositional variables
r
MMales
MFemales
TTF ATF TNTF TRmot TRdis TRem Agreeableness Neuroticism Attachment–anxiety Attachment–avoidance Self-esteem (SAQ) Aggression–physical Aggression–verbal Aggression–anger Aggression–hostility
0.31 0.21 0.18 0.15 0.01 0.00 0.19 )0.12 )0.12 )0.10 0.00 )0.23 )0.12 )0.09 )0.24
3.59 4.34 2.76 4.63 4.23 3.02 3.69 2.71 2.76 3.23 6.23 2.57 2.69 2.26 2.43
3.38 4.66 2.97 4.86 4.23 3.17 3.98 2.95 2.58 2.96 5.99 1.73 2.37 2.07 2.20
(1.14) (0.82)a (0.74) (1.20) (1.25) (1.30) (0.49)a (0.83)a (1.14) (0.98) (1.25) (0.76)a (0.76)a (0.83) (0.79)a
(1.19) (0.79)b (0.77) (0.89) (1.16) (1.18) (0.55)b (0.70)b (1.14) (1.09) (1.14) (0.55)b (0.68)b (0.64) (0.71)b
Note: Significant gender differences ðp < 0:05Þ on predictor variables are denoted by different subscripts within a row. Standard deviations are reported in parentheses following each mean. p < 0:05, p < 0:01.
634
R.P. Brown, A. Phillips / Personality and Individual Differences 38 (2005) 627–638
motivation component of trait rumination (b ¼ 0:15, p < 0:05) were significant predictors of state forgiveness. Neither the ATF nor the TNTF was even close to being statistically significant (ps > 0.35). When we added the other dispositional measures (self-esteem, attachment anxiety and avoidance, neuroticism, agreeableness, and aggressiveness), only the TTF was a significant predictor of state forgiveness, b ¼ 0:26, p < 0:01. Likewise, adding the offense-specific variables (preoffense intimacy, apology, time since offense, and severity of offense) as predictors had almost no impact on the regression weight for the TTF (b ¼ 0:25, p < 0:01), but did cause the motivation component of trait rumination and the physical aggression component of the AQ to be significant predictors (both bs ¼ 0.18, p < 0:05). All of the offense specific variables (except time) also accounted for unique variance beyond the dispositional variables in the model (all ps <0.05). In addition to the above regression analysis, three additional analyses were conducted to examine the possibility of interactions between dispositional forgiveness and the primary offensespecific variables (pre-offense intimacy, apology, and offense severity), as suggested by McCullough and Hoyt (2002). Prior to creating the interaction terms for these analyses, each continuous variable was first centered on its respective mean (Aiken & West, 1991). Next, the centered predictor variables were entered along with the interaction term. Results revealed a significant interaction between scores on the TTF and offense severity, b ¼ 0:17, p < 0:05. In this analysis, both the TTF (b ¼ 0:29, p < 0:001) and offense severity (b ¼ 0:16, p < 0:05) were significantly associated with state forgiveness as well. As shown in Fig. 1, individual differences in trait forgiveness were particularly predictive of state forgiveness when offenses were rated as more severe, b ¼ 0:44, p < 0:001. For less severe offenses, state forgiveness was generally high, so trait forgiveness (as measured by the TTF) was less important and not significant, b ¼ 0:14, p < 0:12. Controlling for pre-offense intimacy and apology did not change these results, nor did controlling for the ATF, the TNTF, and their interactions with severity (none of which was significant). Additional analyses revealed no significant interactions between scores on the TTF and
6.00
State Forgiveness Levels
5.50
5.00
4.50
4.00 Low Severity
3.50
High Severity
3.00 Low TTF
High TTF
Fig. 1. Predicted values of state forgiveness for a past offense as a function of offense severity and TTF scores. Low and high values on both the TTF and severity scales were calculated at ±1 SD of each variable.
R.P. Brown, A. Phillips / Personality and Individual Differences 38 (2005) 627–638
635
Table 3 Relations between dispositional forgiveness measures, trait rumination, and other dispositions Agreeableness Neuroticism Self-Esteem (SAQ) Attachment–anxiety Attachment–avoidance Aggression–physical Aggression–verbal Aggression–anger Aggression–hostility
TTF
ATF
TNTF
TRmot
TRdis
TRem
0.38 )0.50 0.14 )0.39 )0.25 )0.15 )0.15 )0.28 )0.46
0.27 )0.12 0.03 )0.20 )0.15 )0.40 )0.21 )0.22 )0.29
0.25 )0.19 0.00 )0.21 0.00 )0.25 )0.16 )0.18 )0.23
0.10 )0.12 0.32 )0.16 )0.21 )0.10 0.04 )0.13 )0.26
)0.22 0.25 )0.22 0.26 0.17 0.09 0.10 0.21 0.25
0.01 0.41 0.01 0.51 0.21 0.07 )0.06 0.15 0.45
Note: TTF ¼ Tendency to Forgive; ATF ¼ Attitudes Toward Forgiveness; TNTF ¼ Transgression Narrative Test of Forgivingness; TRmot ¼ motivation component of Trait Rumination; TRdis ¼ distraction component of Trait Rumination; TRem ¼ emotionality component of Trait Rumination. p < 0:05, p < 0:01.
pre-offense intimacy, b ¼ 0:04, p > 0:50; the presence of an apology, b ¼ 0:09, p > 0:50; or between the ATF, the TNTF, and any of the offense-specific variables. As a final examination of the correlates of the dispositional forgiveness measures and trait rumination, we examined the associations between these variables and the remaining dispositional variables included in this study. These correlations are displayed in Table 3. As can be seen in this table, the forgiveness measures were related similarly to many of the other dispositional measures. Although the TTF was correlated more strongly than the ATF and the TNTF with agreeableness, neuroticism, both attachment dimensions, and the anger and hostility components of aggressiveness, the ATF (and to a lesser degree the TNTF) were more strongly correlated than the TTF with the physical and verbal components of aggressiveness. Differences between all 3 forgiveness measures and the components of trait rumination also emerged across most of these other dispositions. With only a few exceptions (e.g., self-esteem), trait rumination was less strongly related to the other dispositions than were the forgiveness measures. In sum, the measures of dispositional forgiveness and trait rumination exhibited some degree of convergence, as well as substantial divergence, in their relations with a number of dispositions that have been conceptually and empirically linked with interpersonal interactions.
4. Discussion The present research demonstrates the relevance of dispositional variables to forgiveness. As a number of theorists have posited, individual differences in forgiveness do appear to exist and may enhance our ability to predict the aftermath of interpersonal offenses. Precisely how such individual differences should best be measured is a matter of some debate, and empirical support for different dispositional indices remains sparse. The present research is a step toward alleviating this lack of empirical guidance. The data we have reported here suggest that one measure, the TTF, may be particularly promising as a measure of dispositional forgiveness, as this measure predicted both mental health variables (depression and life satisfaction) and state forgiveness for an actual
636
R.P. Brown, A. Phillips / Personality and Individual Differences 38 (2005) 627–638
offense, independently of three other measures: pro-forgiveness attitudes, a scenario-based measure of dispositional forgiveness, and a multi-dimensional measure of trait rumination. In addition, a host of other dispositional variables that might also relate to state forgiveness did not substantially reduce the predictive validity of the TTF with regard to state forgiveness. These other dispositional variables also partially discriminated among the various forgiveness and rumination measures, further demonstrating the differences between them, though some consistencies were certainly present as well. Together, the findings presented here extend recent work by Brown (2003), who reported preliminary evidence for the construct validity of the TTF but did not present any discriminant validity evidence with respect to state forgiveness or trait rumination and only presented preliminary discriminant validity evidence with respect to mental health (contrasting the tendency to forgive with attitudes toward forgiveness). Although a theoretical distinction between the tendency to forgive and pro-forgiveness attitudes is easy to make, the distinction between retrospective measures like the TTF and scenariobased measures like the TNTF is less obvious. Why would reports of oneÕs tendency to forgive in the past not be nearly interchangeable with oneÕs predictions of forgiveness for hypothetical offenses? These measures certainly purport to assess the same construct. Nonetheless, across all of the criterion variables examined in this study, the TTF was a more effective predictor than the TNTF (as well as the ATF). In fact, neither the TNTF nor the ATF was independently significant as a predictor variable in any of the analyses that we conducted. Why might the hypothetical scenarios of the TNTF have been less effective than the TTF? We suspect that one important reason may be that when individuals project themselves into the scenarios of the TNTF, several variables contribute to the predictions they make about how forgiving they would be. One of those variables may be their actual proneness to forgive, but another is likely to be the value they place on forgiveness. Support for this contention was reported by Brown (2003), who showed that scores on the TTF and the ATF were uniquely related to respondentsÕ predictions on the TNTF, although the ATF was a somewhat stronger predictor than was the TTF. This finding, incidentally, was replicated in the present study (bs ¼ 0.26 and 0.42 for the TTF and ATF, respectively, ps < 0.001). This strong association between forgiveness attitudes and scores on the TNTF suggests that the latter might be influenced by the salience of the former. There are, of course, other variables that could contribute to peopleÕs forgiveness predictions, and we believe that exploring such variables remains an important issue, given the common use of hypothetical vignettes in psychological research, especially in studies of conflict and forgiveness (see also McCullough & Hoyt, 2002). One of the most glaring absences in the forgiveness literature to date is the experimental manipulation of offenses. Instead of creating offense experiences for participants and measuring their responses, researchers have relied primarily on self-reported past experiences or hypothetical scenarios. Thus, establishing that individual differences in forgiveness are truly meaningful and predictive would seem to demand that this gap in the literature be filled. However, the present results––derived from retrospective accounts of real offenses––sound a warning note for such an endeavor. The interaction between scores on the TTF and offense severity showed that dispositional forgiveness was only related to state forgiveness for relatively severe offenses. This interaction might pose a difficulty for experimental research on dispositional forgiveness, insofar as rather severe offense experiences might need to be created for dispositional variables to be predictive. The challenge that remains, then, is to demonstrate the predictive validity of dispositional
R.P. Brown, A. Phillips / Personality and Individual Differences 38 (2005) 627–638
637
forgiveness measures with experimentally manipulated offenses that are severe enough to matter, but not so severe as to violate the rights and welfare of participants. Achieving this goal promises to advance our understanding of state and trait dimensions of forgiveness and to promote the development of better models of the forgiveness process.
Appendix A. State forgiveness measure 1. 2. 3. 4. 5. 6. 7.
I have forgiven this person. I feel angry toward this person. (R) Even though his/her actions hurt me, I do not feel ill-will toward him/her. I dislike this person. (R) I feel warmly toward this person. I hope this person gets whatÕs coming to them for what they did to me. (R) If I saw this person again, I would try to avoid interacting with him/her. (R)
Note: Items marked with (R) were reverse scored, so that higher scores indicate more forgiveness. Items 6 and 7 were adapted from McCullough et al.Õs (1998) TRIM.
References Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Newbury Park, CA: Sage. Berry, J. W., Worthington, E. L., Jr., Parrott, L., OÕConnor, L. E., & Wade, N. G. (2001). Dispositional forgivingness: Development and construct validity of the transgression narrative test of forgivingness (TNTF). Personality and Social Psychology Bulletin, 27, 1277–1290. Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report measurement of adult attachment: An integrative overview. In J. A. Simpson & W. S. Rholes (Eds.), Attachment theory and close relationships (pp. 46–76). New York: Guilford Press. Brown, R. P. (2003). Measuring individual differences in the tendency to forgive: Construct validity and links with depression. Personality and Social Psychology Bulletin, 29, 759–771. Buss, A. H., & Perry, M. (1992). The aggression questionnaire. Journal of Personality and Social Psychology, 63, 452– 459. Caprara, G. V. (1986). Indicators of aggression: The dissipation–rumination scale. Personality and Individual Differences, 7, 763–769. DeShea, L. (2003). A scenario-based scale of willingness to forgive. Individual Differences Research, 1, 201–217. Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The satisfaction with life scale. Journal of Personality Assessment, 49, 71–75. Enright, R. D., Freedman, S., & Rique, J. (1998). The psychology of interpersonal forgiveness. In R. D. Enright & J. North (Eds.), Exploring forgiveness (pp. 46–62). Madison: The University of Wisconsin Press. Fincham, F. D. (2000). The kiss of the porcupines: From attributing responsibility to forgiving. Personal Relationships, 7, 1–23. Finkel, E. J., Rusbult, C. E., Kumashiro, M., & Hannon, P. A. (2002). Dealing with betrayal in close relationships: Does commitment promote forgiveness? Journal of Personality and Social Psychology, 82, 956–974. Freedman, S., & Enright, R. D. (1996). Forgiveness as an intervention goal with incest survivors. Journal of Consulting and Clinical Psychology, 64, 983–992. Gorsuch, R. L., & Hao, J. Y. (1993). Forgiveness: An exploratory factor analysis and its relationships to religious variables. Review of Religious Research, 34, 333–347.
638
R.P. Brown, A. Phillips / Personality and Individual Differences 38 (2005) 627–638
John, O. P., Donahue, E. M., & Kentle, R. (1991). The Big Five Inventory––Versions 4a and 54 (Tech. Report). Berkeley: University of California, Berkeley, Institute of Personality and Social Research. Mauger, P. A., Perry, J. E., Freeman, T., Grove, D. C., McBride, A. G., & McKinney, K. E. (1992). The measurement of forgiveness: Preliminary research. Journal of Psychology and Christianity, 11, 170–180. McCullough, M. E. (2000). Forgiveness as human strength: Theory, measurement, and links to well-being. Journal of Social and Clinical Psychology, 19, 43–55. McCullough, M. E., Bellah, C. G., Kilpatrick, S. D., & Johnson, J. L. (2001). Vengefulness: Relationships with forgiveness, rumination, well-being, and the big five. Personality and Social Psychology Bulletin, 27, 601–610. McCullough, M. E., & Hoyt, W. T. (2002). Transgression-related motivational dispositions: Personality substrates of forgiveness and their links to the big five. Personality and Social Psychology Bulletin, 28, 1556–1573. McCullough, M. E., Hoyt, W. T., & Rachal, K. C. (2000). What we know (and need to know) about assessing forgiveness constructs. In M. E. McCullough, K. I. Pargament, & C. E. Thoresen (Eds.), Forgiveness: Theory, research, and practice (pp. 65–88). New York: Guilford. McCullough, M. E., Rachal, K. C., Sandage, S. J., Worthington, E. L., Brown, S. W., & Hight, T. (1998). Interpersonal forgiving in close relationships: II. Theoretical elaboration and measurement. Journal of Personality and Social Psychology, 75, 1586–1603. Mikulincer, M. (1998). Adult attachment style and individual differences in functional versus dysfunctional experiences of anger. Journal of Personality and Social Psychology, 74, 513–524. Murray, S. L., Holmes, J. G., MacDonald, G., & Ellsworth, P. G. (1998). Through the looking glass darkly? When selfdoubts turn into relationship insecurities. Journal of Personality and Social Psychology, 75, 1459–1480. Pelham, B. W., & Swann, W. B., Jr. (1989). From self-conceptions to self-worth: On the sources and structure of global self-esteem. Journal of Personality and Social Psychology, 57, 672–680. Radloff, L. F. (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385–401. Scott, V. B., & McIntosh, W. D. (1999). The development of a trait measure of ruminative thought. Personality and Individual Differences, 26, 1045–1056. Tangney, J., Boone, A. L., Dearing, R., & Reinsmith, C. (2002). Individual differences in the propensity to forgive: Measurement and implications for psychological and social adjustment. Unpublished manuscript: George Mason University.