Gender Differences in Forgiveness and its Affective Correlates

Although women are believed to be more forgiving than men, the results of many studies comparing women with men vary. Moreover, little is known about unique correlates or differential patterns of experiencing forgiveness by gender. In the present study, we compared men and women in terms of their level of dispositional forgiveness and its emotional correlates, namely positive and negative affect, anxiety, and emotional control. The sample consisted of 625 individuals aged 19–69, of whom 478 (76.5%) were women and 147 (23.5%) were men. Polish versions of the Heartland Forgiveness Scale (HFS), the Positive and Negative Affect Schedule (PANAS), the Courtauld Emotional Control Scale (CECS), and the State-Trait Anxiety Inventory (STAI) were used. Men showed a higher level of general forgiveness and greater willingness to overcome unforgiveness than women, but there was no significant difference in positive facets of the disposition to forgive. In both genders negative affect, anxiety, and control of anger and of depression were negatively related to dimensions of dispositional forgiveness, and positive affect was positively associated with forgiveness. In females control of anxiety was negatively and in males it was positively related to facets of forgiveness. Gender moderated a number of links between affective traits and forgiveness of self and of situations beyond control, but not forgiveness of others.


Introduction
Women are generally believed to be more forgiving than men, due to their personality traits such as agreeableness and empathy, and their valuing relationships (Miller et al., 2008). In a meta-analytic review of 70 studies with 15,731 individuals, Miller 1 3 et al. (2008) found a gender difference in forgiveness with small to moderate effect size, showing greater willingness to forgive among female respondents. On the other hand, Fehr et al. (2010) who conducted meta-analyses across 53 studies with the total number of 8,366 participants, found no relationships between gender and forgiveness. There are many reasons for this discrepancy between findings reported by different researchers, including methodological approach, but the main one is the fact that majority of referenced studies were not aimed at comparing men and women in terms of multidimensional forgiveness and the way they experience this phenomenon. Also, very few attempts have been made to find unique correlates or differential patterns of experiencing forgiveness by genders (e.g., Toussaint & Webb, 2005). Thus, in the present study we directly compare males and females in terms of their level of dispositional forgiveness and its emotional correlates, namely positive and negative affect, anxiety, and emotional control, and we examine the moderating role of gender in the relationships between emotions and forgiveness.
However, the results of studies comparing dispositional forgiveness in men and women are ambiguous. For instance, Ka et al. (2006) found that women yielded dependent, connected, and less differentiated from others (Agerström et al., 2006;Cross & Madson, 1997). Such differences in defining oneself may have an impact on gender differences in responding to and dealing with transgressions. Women are likely to be more motivated to forgive because forgiveness helps them maintain a connection with others (Markus & Kitayama, 1991) by pushing people to reconciliation (Worthington et al., 2019). The link between interdependence and forgivingness was found to be notably stronger among women than men (Neto & Mullet, 2004). It might be particularly important when female stress and detachment following interpersonal transgressions are intensive and an individual needs to cope with them. Although there are different ways to reduce such distress (Wade & Worthington, 2003), only forgiveness results in pro-social acting and relating with others (Worthington & Wade, 1999). In confirmation, Neto and Mullet (2004) showed that loneliness was positively associated with propensity to forgive among women, whereas it was negatively correlated in men (Neto & Mullet, 2004).
Gender and different self-construals might not only affect the level of forgiveness, but they may also play a role in processing forgiveness, especially in the relationship between emotions and forgiveness. Different self-construals bring significant consequences for experiencing or/and suppressing emotions, both positive and negative (Markus & Kitayama, 1991). For instance, holding an interdependent view of the self inhibits overly expressed anger or frustration, which are emotional states closely related to forgiveness. Consequently, links between the tendency to experience and control different emotions and the ability to forgive might differ in men and women. Generally, negative affect and emotions such as anger, shame, and guilt may impede the process of forgiveness (Hall & Fincham, 2008;Macaskill, 2012;Mróz & Kaleta, 2017), and less forgiving individuals tend to experience higher levels of bitterness, hostility, fear, anxiety, or depressive affect (Barcaccia et al., 2020;Berry et al., 2005;Rye et al., 2001;Toussaint et al., 2008;Lawler-Row & Piferi, 2006;Burnette et al., 2009). In turn, positive emotions, like sympathy, compassion, or happiness, increase the likelihood of forgiving (Baker et al., 2017;Uysal & Satici, 2014). Still, there are gender differences in various affective traits. For instance, women often score higher in negative affect, especially for anxiety and depression (Fujita et al., 1991;Parker & Brotchie, 2010;Thomsen et al., 2005). In addition, emotion regulation which affects emotional experience and expression (Baker et al., 2017;Gross, 1998;Gross & John, 2003;Thomsen et al., 2005) and which is related to forgiveness (Baker et al., 2017;Hodgson & Wertheim, 2007;Mami et al., 2019;McCullough et al., 2001;Witvliet et al., 2010Witvliet et al., , 2011Witvliet et al., , 2015, might be gender-related (Gross & John, 2003;Koch et al., 2007;Matud, 2004;McRae et al., 2008;Thomsen et al., 2005). Summing up, our reasoning points to gender differentiation in forgivingness and its associations with emotional variables. Thus, the first goal of the present study was to examine gender differences in the levels of dispositional forgiveness. We put forward the hypothesis that women would score higher in multiple dimensions of forgivingness (H1). The second goal was to examine how emotional variables are associated with different aspects of dispositional forgiveness among women and men. We expected that negative affect, anxiety and control of anger, of depression and of anxiety, would be significantly and inversely correlated with forgivingness, whereas positive affect positively correlated with the ability to forgive in both genders. However, as shown, the link between emotionality and dispositional forgiveness may vary across genders in terms of direction or magnitude. Thus, we hypothesized the moderating role of gender in the analyzed relationships. From the self-construal approach positive emotions might motivate females to be more forgiving than men, whereas the experience of negative emotions may make women less reluctant than men to forgive in order to find peace after transgressions. Consequently, we hypothesized that the relationship between positive affect and dispositional forgiveness would be more positive for females (H2), while associations between negative affect and forgiveness, anxiety and forgiveness, and between suppression of emotions and forgiveness would be less negative for them (H3).

Forgivingness
Disposition to forgive was measured using the Polish version (Kaleta et al., 2016) of the Heartland Forgiveness Scale (Thompson & Snyder, 2003;Thompson et al., 2005). HFS is a multidimensional tool assessing dispositional forgiveness of self, others, and situations beyond anyone's control. Participants rate their responses to 18 items on a 7-point scale. The Polish version consists of two scales that allow to measure forgivingness in two separate domains-positive (P-scale; benevolent thoughts, feelings, and behaviors) and negative (N-scale; reduction of hostile thoughts, feelings, and behaviors), and six subscales with the distinction of forgiveness of self, others, and situations (P-self, P-others, P-situations, N-self, N-others, N-situations). Higher scores on each scale reflect a higher level of forgivingness in every domain. The total HFS score indicates how forgiving a person tends to be. Sample items include "Although I feel badly at first when I mess up, over time I can give myself some slack" and "If others mistreat me, I continue to think badly of them." Cronbach's alpha (internal consistency) values were found as follows: for overall HFS 0.76, for P-scale 0.70, and for N-scale 0.81, and from 0.50 to 0.81 for subscales (Kaleta et al., 2016).

Positive and Negative Affect
Positive-negative affect was measured using the Polish version (SUPIN C30; Brzozowski, 2010) of the Positive and Negative Affect Schedule (PANAS; Watson et al., 1988). The scale consists of 30 items, 15 items for positive affect (from 15 to 75 points) and 15 items for negative affect (from 15 to 75 points). We focused on measuring affect at a trait level. Using a 5-point scale, the participants are asked to indicate the degree to which they usually feel each emotion, e.g., ashamed, irritated, afraid, exited, proud, active. The higher the score, the higher the level of particular affectivity. Cronbach's alpha for PANAS ranged from 0.73 to 0.95 (Brzozowski, 2010).

Anxiety
Anxiety was measured using the State-Trait Anxiety Inventory (STAI, Spielberger et al., 1983), more specifically the Polish version of the tool (Wrześniewski et al., 2002). This is a widely used, valid instrument. In the present study, participants rated 20 trait anxiety items (STAI-X2) assessing how an individual generally feels, using a 4-point Likert scale. Sample items: "I worry too much over something that really doesn't matter," "I am a steady person." The score range is from 20 (low) to 80 (high anxiety). Cronbach's alpha measuring internal consistency of the Polish version ranged from 0.76 to 0.92 in the group of adult women and men.

Emotional control
Emotional control was measured using the Courtauld Emotional Control Scale (CECS) developed by Watson and Greer (1983) and adapted in Poland by Juczyński (2009). The questionnaire allows to evaluate the extent to which individuals report suppression as an emotion regulation strategy in difficult situations. The CECS consists of 21 items divided into three subscales, each of which contains seven statements concerning the way of suppressing or expressing anger, depression and anxiety (from 7 to 28 points in every subscale). Participants are asked to respond to the phrases "When I feel angry…", "When I feel anxious (worried)…", and "When I feel unhappy (miserable)…" with statements such as "I keep quiet," "I bottle it up," or "I tell others about it" rated on a 4-point scale. The total emotional control index is established by summing up the results of the three subscales. The higher the result, the more negative emotions are suppressed. Reliability of the Polish version (Cronbach's alpha) is: for the control of anger 0.80, depression 0.77, anxiety 0.78, and for the total emotional control 0.87 (Juczyński, 2009).

Statistical methods
First, we calculated descriptive statistics (means and standard deviations) for all the variables in women and men separately. Due to unequal sample sizes we used the Mann-Whitney U test to examine gender differences and the Glass'Δ estimate to evaluate the effect size. Second, correlational analyses were performed to explore relationships among emotional variables and multidimensional forgiveness in females and males. Finally, to test whether gender moderated the relationship between emotionality and disposition to forgive, we used a set of moderation analyses using the Process macro for SPSS (Model 1, version 3.5, Hayes, 2018). Moderation models have been proposed with positive-negative affect, anxiety, emotional control (overall and control of anger, of anxiety, and of depression) as predictors, dimensions of forgivingness as outcome variables, and gender as a moderator. In order to properly estimate interactions of emotional variables and gender on forgivingness, mean centering and standardizing were applied. Bias-corrected confidence intervals (CI 95%) and the bootstrapping procedure (samples = 5000) were used to calculate direct and indirect effects. Table 1 presents gender differences in dimensions of forgivingness and analyzed variables and provides descriptive statistics, statistical significance, and Glass' Δ effect size estimates.

Results
Men showed higher levels of general forgivingness and more willingness to overcome unforgiveness, especially toward oneself and situations beyond anyone's control. Gender did not differentiate the results in positive facets of the disposition to forgive. There were gender differences in affective traits. Women scored significantly higher in negative affect and anxiety. Men reported greater positive affect and emotional control, in particular the suppression of anxiety. It should be noted that according to the Glass' effect size estimate, gender differences were rather small to moderate in magnitude. The largest differences between women and men were reported for reduced unforgiveness of situations, negative affect, anxiety, and control of anxiety. Table 2 shows correlations (Pearson's r) between the variables in women and men. .11 .06 .08 Reduction   As shown in Table 2, women's negative affect and anxiety were negatively correlated with multiple dimensions of dispositional forgiveness, whereas positive affect was positively associated with all facets of forgivingness. Women's control of emotions, including control of anger and depression, was inversely related to total forgiveness and reduced unforgiveness, but not related to positive forgiveness. Control of anxiety was negatively associated with overcoming unforgiveness of situations beyond control.
In the case of male participants, negative affect and anxiety were inversely correlated with overall forgiveness and all facets of reducing unforgiveness. Positive affect was positively related to total and positive forgiveness (especially of situations and others). Control of anger was negatively correlated with overcoming unforgiveness of self; control of depression was negatively associated with general forgivingness and decreased unforgiveness, particularly of situations. Finally, control of anxiety was positively associated with total forgiveness and reduction of unforgiveness (general, as well as of oneself and situations).
Next, we examined the moderating role of gender in the relationship between affective traits and disposition to forgive. Sixty-three models adopting subsequent affective traits as predictors and facets of forgivingness as outcomes were tested, and eight of them confirmed the moderating role of respondents' gender. Table 3 presents interaction effects of the examined models. 1 The analysis revealed no significant interaction effect of positive affect and gender on any dimension of forgiveness, but confirmed a significant effect of negative affect and gender in relation to positive forgiveness of situations (Table 3). It showed that negative affect was negatively related to this type of forgiveness for women Thus, moderation analysis confirmed gender differences in processing forgiveness of self and of situations. Higher negative affect, anxiety, emotional control, including control of anger and of anxiety, made overcoming negative responses in situations beyond control more difficult for females than for males. In contrast, greater control of anxiety helped males to be more forgiving, especially toward themselves and situations. The moderating role of gender in relation to forgivingness of others did not occur.

Discussion
Although neglected by most researchers, gender-related differences in forgivingness (and the associated variables) should be given more empirical attention. It is significant not only to gain more insight into the forgiveness phenomenon, but also to apply effective interventions and therapeutic techniques. Our study attempts to fill in this gap as it was aimed at direct comparison of men and women in their disposition to forgive and the way it is linked to affective traits.
The first hypothesis about a stronger tendency to forgive among women was not supported. Women revealed lower forgiveness in the total score and in its negative dimension. Few studies showed a similar trend, four of which included dispositional forgiveness (Brown, 2003;Charzyńska, 2015;Exline et al., 2004;Piątek, 2011).
Two other studies were focused on forgiveness of a specific offense in close relationships, showing male partners to be more forgiving toward female partners (Finkel et al., 2002;Miller & Worthington, 2010). However, they all measured forgiveness of others, while our results demonstrated that women are less able to overcome unforgiveness toward themselves and situations, but not toward other people. Also, there were no significant differences between women and men in positive facets of forgiveness. Multidimensional measures we used might have helped us investigate gender differences more accurately. To our knowledge, our study is the first to examine gender differences in terms of that many aspects of forgivingness.
Females' lower disposition to reduce unforgiveness, especially toward self and situations beyond control, might be interpreted as a function of different self-construals. An interdependent view of the self characterizes females as focused on maintaining connection to people and being less separated from the social context (Markus & Kitayama, 1991). Their own failures and wrongdoings might be seen as disturbing the harmony with relevant others and threatening their positive self-portrait (Fincham et al., 2004). Consequently, an interdependent self-construal makes females prone to feel shame (Markus & Kitayama, 1991) which is negatively related to forgiveness, especially of self (Hall & Fincham, 2005).
Consistently with previous studies (e.g., Rye et al., 2001;Thompson et al., 2005;Uysal & Satici, 2014), we found that negative affect, anxiety, control of anger, and of depression were negatively correlated with forgivingness, while positive affect was positively related to forgiveness, in both women and men. Gender significantly moderated a number of the relationships between affective traits and forgivingness, especially links involving emotional control. Few previous studies also demonstrated genderbased correlations between different variables and forgiveness. For instance, both Fincham et al. (2002) and Toussaint and Webb (2005) found that empathy is more strongly associated with forgiveness in men than in women. Fincham et al. (2002) also showed that the direct effect of attribution of responsibility on marital forgiveness proved to be stronger for wives than for husbands. Fincham et al. (2004) demonstrated that partner reports of inefficient conflict resolution were positively correlated with retaliation, a negative dimension of marital forgiveness, for husbands only.
We found-in contrast to our hypotheses-negative relationships between negative affect, anxiety, emotional control, including control of anger and of anxiety, and willingness to forgive of situations stronger for female than for male participants. For interdependent individuals, it is important to control intense experiences of negative emotions, especially anger, because it threatens their selves (Markus & Kitayama, 1991). However, in the light of our findings strong emotional control makes forgiving difficult for women. In turn, control of anxiety was positively related to forgiveness in men. For independent selves, it is important to express self and egofocused emotions, such as anger or frustration. Controlling anxiety in men might foster such expression and, consequently, favor forgiveness. Our results suggest that women and men need different interventions in facilitating forgiveness (Miller & Worthington, 2015). Females need more self-forgiveness interventions, releasing anxiety and promoting more open expression of negative emotions in the way they are able to accept, for instance in constructive communication. Males need work on emotion regulation, especially controlling of anxiety and of anger.

Limitations of the Study
Although the present study attempts to fill in the gap by addressing the scarcely examined question of gender differences in forgiveness, it is burdened with a number of limitations. First, there is a significant gender imbalance in the sample. It is a common problem that women tend to participate in forgiveness studies more often than men (Root & Exline, 2011;Worthington et al., 2000). We collected data only from interested individuals who were willing to participate in the study completely voluntarily (for no remuneration). The question is whether some bonuses would encourage more men to participate in the research and help minimize the imbalance. Secondly, the design of the research is cross-sectional and the statistics are merely correlational and cannot be used to draw conclusions about causality. Longitudinal studies would be more useful in exploring the causal relationships between analyzed variables more extensively. Thirdly, we did not control some important variables, such as social desirability bias which is related to forgiveness and which may moderate the link between affective traits and forgivingness (e.g., Rye et al., 2001). Finally, because the study uses measures of dispositional forgiveness, it does not in fact address actual forgiving. Consequently, the findings cannot be extended to forgiveness occurring in natural situations. Ability to forgive is just one of many more predictors of episodic forgiveness that is, in fact, more strongly related to contextual circumstances, such as severity of the wrongdoing, apology, relational closeness, and passage of time (Boon & Sulsky, 1997;Johnson et al., 2013;Koutsos et al., 2008). Thus, our findings still do not make it possible to predict whether women or men would be more willing to forgive a particular offense in a given context. Future research should apply more complex models in which both disposition to forgive and gender would mediate and moderate the relationships between different contextual factors, episodic forgiveness, and related variables. For instance, the relationship between apology and actual forgiveness could be mediated by forgivingness and moderated by gender.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.