This study examined relations between aspects of family functioning and positive and
negative dimensions of forgiveness. Increased understanding of one’s partner and de-
creased anger about betrayal characterize positive forgiveness, whereas experiences such
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
as holding a grudge and desiring revenge indicate negative forgiveness. The sample
This document is copyrighted by the American Psychological Association or one of its allied publishers.
Conflict within relationships is inevitable. At one time or functioning, particularly in regards to psychological close-
another, partners might inadvertently make a hurtful com- ness, if partners are unable to forgive each other and effec-
ment, forget to pick up the dry-cleaning, or borrow their tively resolve their conflicts (Gordon & Baucom, 2003).
partner’s car and leave the gas tank empty. Most couples Although forgiveness is notoriously difficult to define,
resolve such conflicts on an ongoing basis, leaving little three elements of forgiveness appear most frequently in the
emotional residue to negatively impact their lives. However, psychological literature: (a) regaining a more balanced and
examples of more devastating relational conflicts include compassionate view of the offender and the event, (b)
infidelities, major lies, drastic unilateral financial decisions, decreasing negative affect towards and avoidance of the
and other similar humiliations and betrayals (e.g., Cano, offender, and (c) giving up the right to seek revenge or lash
Christian-Herman, O’Leary, & Avery-Leaf, 2002). These out toward the offender (e.g., Enright & the Human Devel-
conflicts frequently leave lasting emotional scars on marital opment Study Group, 1991; Gordon & Baucom, 1998;
McCullough, et al., 1998). Furthermore, recent views of
forgiveness suggest that forgiveness has at least two dimen-
sions (Fincham & Beach, 2002; Worthington, 2003): (a) a
Kristina Coop Gordon, Farrah M. Hughes, Nathan D. Tom- negative dimension, which involves the degree to which an
cik, Lee J. Dixon, and Samantha C. Litzinger, Department of
Psychology, University of Tennessee–Knoxville.
individual continues to hold grudges, withdraws from the
Farrah M. Hughes is now at Francis Marion University, Depart- relationship, and desires revenge or punishment against the
ment of Psychology. Nathan D. Tomcik is now at the Chalmers P. partner for a past betrayal (i.e., negative forgiveness), and
Wylie VA Outpatient Clinic, Columbus, OH. Samantha C. Litz- (b) a positive dimension, which involves the degree to
inger is now at Bellevue Hospital, New York, NY and at Monte- which an individual experiences a readiness to forgive, an
fiore Medical Center in Bronx, NY. increase in empathy, and a release from anger (i.e., positive
The authors wish to thank the numerous research assistants who forgiveness). We have used the terminology “negative for-
helped us collect this data and the families who generously shared giveness” and “positive forgiveness” to identify these di-
their time and experiences with us. mensions throughout this paper, as these terms are similar to
This study was funded in part by a Randy Gerson Memorial
the ones used in previous studies examining dimensions of
Grant from the American Psychological Foundation awarded to
the first author. forgiveness (e.g., Fincham & Beach, 2002; Fincham, Hall,
Correspondence concerning this article should be addressed to & Beach, 2006).
Kristina Coop Gordon, Department of Psychology, 310B Austin Many forgiveness researchers argue that that the “posi-
Peay Bldg., University of Tennessee, Knoxville, TN 37996. tive” and “negative” dimensions may play different roles in
E-mail: kgordon1@utk.edu different forgiveness situations and relational contexts (e.g.,
1
2 COOP GORDON, HUGHES, TOMCIK, DIXON, AND LITZINGER
Fincham & Beach, 2002; McCullough et al., 1998; Worth- struct or modify their former beliefs about their partner and
ington, 2005). For example, early forgiveness researchers the relationship, as well as their efforts to regain a sense of
primarily focused on the reductions in negativity, such as interpersonal safety in the relationship in order to effec-
vengefulness and avoidance, as primary indications of for- tively move beyond the event.
giveness; however, lay persons have focused more on pos- Furthermore, this model of forgiveness suggests that
itive or benevolent attitudes toward the offender as being early in the process, when people are likely to be most
most representative of forgiveness (e.g., Kearns & Fincham, unforgiving, individuals might experience emotional tur-
2004). Worthington (2005) suggests that in noncontinuing moil (e.g., anger, anxiety, depression) and cognitive confu-
relationships, the reduction in the negative dimension of sion about their partners, their relationships, and even them-
forgiveness might be most important in determining for- selves. They also are likely to seek revenge and withdraw
giveness, whereas Fincham et al. (2006) suggest that in from their partners. These experiences are consistent with
continuing relationships, forgiveness may require both a the negative forgiveness dimension described above. At the
reduction in the negative and an increase in the positive end of the forgiveness process, when people are most for-
dimensions. More specifically, Fincham and Beach (2002) giving, the injured partners are able to move away from
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
suggest that reductions in the negative dimension of for- blaming their partners, feel more at peace with their under-
giveness might lead to fewer avoidance behaviors, but this standing of the betrayal, and are able to move beyond the
change alone will not necessarily lead to an increase in betrayal; these experiences are consistent with the positive
approach behaviors that are necessary to resume closeness forgiveness dimension.
in a marriage. Similarly, their findings indicate that a reduc- The implications of this model for the present study are
tion in the negative dimension of forgiveness predicts re- that the failure to resolve the betrayal and adequately pro-
duction in psychological aggression, whereas only an in- cess disrupted relationship assumptions might lead spouses
crease in the positive dimension of forgiveness predicts an to have a lack of trust in their partners and experience
increase in constructive communication. Consequently, Fin- increased conflicts with their partners, which could then
cham et al. (2006) recommend that researchers interested in color subsequent dyadic interactions, even in areas of rela-
forgiveness in ongoing intimate relationships look at both tionship functioning that are unrelated to the betrayal (e.g.,
dimensions of forgiveness separately. parenting). In summary, it is expected that negative forgive-
Moreover, the current paper contends that it is possible ness, characterized by cognitive and emotional dysregula-
that the impact of partners’ inabilities to forgive each other tion, and positive forgiveness, characterized by a calmer and
does not just end with the couple; rather, the dimensions of more balanced relationship perspective, will both affect
forgiveness might exert influence on ever widening spheres aspects of dyadic and family functioning.
of family functioning. For example, negative forgiveness
between spouses might contribute to increases in marital Dyadic Correlates of Forgiveness
dissatisfaction that, in turn, spill over into the parenting
relationship and subsequently impact the children. Addi- Dyadic Functioning
tionally, it is possible that increases in the positive dimen-
sion may increase partners’ abilities to reconnect and trust Not surprisingly, the association between forgiveness and
each other again after a major betrayal. However, despite its relationship functioning has been well documented. Greater
potential for resonating throughout the family system, little general forgiveness is associated with higher levels of dy-
attention has been paid to the influence of the forgiveness adic adjustment (e.g., Gordon & Baucom, 2003; Paleari,
process in families beyond its impact on dyadic functioning; Regalia, & Fincham, 2005), greater empathy (McCullough
consequently, the goal of this study is to provide a more et al., 1998), and more constructive communication and
comprehensive look at the effects of major betrayals and lower levels of psychological aggression (Fincham &
their associations with other aspects of dyadic and family Beach, 2002). In addition, forgiveness has been related to
functioning. more positive cognitions regarding self and partner as well
as greater intimacy and closeness in the relationship (Finkel,
The Forgiveness Process Rusbult, Kumashiro, & Hannon, 2002; Gordon & Baucom,
2003). Similarly, lower levels of forgiveness are associated
Gordon and Baucom (1998; 2003) provided a process with poorer relational adjustment, intimacy, and closeness,
model of relational betrayal and forgiveness, outlining the and higher levels of psychological aggression (Fincham &
means by which betrayals might negatively impact dyadic Beach, 2002; Finkel et al., 2002; Gordon & Baucom, 2003).
relationships and, if unresolved, might have lasting effects. Despite these demonstrated associations, the mechanisms
Furthermore, they also described a forgiveness process that by which the more specific dimensions of positive and
may allow couples to resolve these issues in a healthy negative forgiveness affect marital adjustment require
manner. Briefly, Gordon and Baucom (1998) posited that a greater clarification. It is possible that negative forgiveness
major betrayal requiring forgiveness can be seen as an has its effects on marital adjustment primarily through its
interpersonal trauma that violates an individual’s assump- shared association with marital conflict (i.e., a spillover
tions about his or her partner and their relationship in effect); the negative affect and perceptions of the partner
general. Therefore, the need to engage in the forgiveness generated by a betrayal increase the likelihood of greater
process might result from individuals’ attempts to recon- conflict in the marriage, which in turn is associated with
FORGIVENESS AND FAMILY FUNCTIONING 3
decreased marital satisfaction. This possibility is supported Forgatch & DeGarmo, 1997; Krishnakumar & Buehler,
by findings indicating that less positive forgiveness is re- 2000). Similarly, we expect that if a parent has lingering
lated to higher levels of conflict in marriage (Fincham & anger about a betrayal and harbors a desire to punish his or
Beach, 2002). Similarly, it also is possible that more nega- her partner, then these attitudes are likely to weaken the
tive views of the partner may lead to decreased dyadic trust, parenting alliance. Thus, failure to forgive one’s partner can
which in turn leads to decreased dyadic satisfaction. More- contribute to poor parenting patterns and expose children to
over, an increase in positive forgiveness may influence parental conflict, which then has negative implications for
relationship functioning also by decreasing conflict levels child functioning (e.g., Gottman & Katz, 1989; Grych &
and increasing trust. A variety of theoretical and clinical Fincham, 1990).
publications on forgiveness suggest that forgiveness gener-
ally should be related to trust in relationships (e.g., Rusbult, Parental Functioning Hypotheses
Kumashiro, Finkel & Wildschut, 2002; Worthington, 2003).
Recent research suggests that these constructs are related Specifically, we expect higher positive forgiveness and
but distinct, as relational trust has been shown to be inde- lower negative forgiveness to predict a stronger parenting
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
couples both partners were betrayed, in 20.7% of the cou- moil and more desire for revenge and punishment, as well as
ples the wives only were betrayed, and in 5.4% of the more avoidance of the partner. Scores on the negative
couples husbands only were betrayed. The remaining indi- forgiveness subscale in the present sample ranged from 12
viduals either left the forgiveness measures blank or re- to 45 for wives and 12 to 42 for husbands. The positive
ported that no serious betrayal had occurred in their mar- forgiveness subscale measures the degree to which the
riages. Thus in our final sample, there were 87 wives who individual has put the betrayal behind him or her, feels
reported a betrayal and 74 husbands who reported a be- emotionally at peace, and is more understanding and less
trayal; these individuals, their partners, and their adolescent blaming of his or her partner. Higher scores on this subscale
children comprised the final sample used here (there were represent healthier emotional and interpersonal functioning.
91 couples total). The participating couples had an average Scores on the positive forgiveness subscale in the present
of 2.6 children and had been married 16 years, on average. sample ranged from 32 to 60 for wives and 18 to 60 for
Nine of the families were blended families (i.e., the parents husbands.
were a biological parent and stepparent). Husbands aver- Conflict Tactics Scale The Conflict Tactics Scale (CTS;
aged 43.5 years of age (SD ⫽ 6.0), and wives averaged 41.6 Straus, 1979), a self-report survey is an extensively used
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
sound parenting relationship with their child’s other parent. All procedures were approved by the relevant Institutional
For example, items assess spouses’ perceptions of their Review Board prior to beginning the study.
support for each other and their desire to communicate
about their child. Abidin and Konold (1999) reported ac- Results
ceptable reliability coefficients as well as satisfactory indi-
ces of validity. Alphas in this sample were .97 for both All analyses reported below were preceded by an analysis
husbands and wives. Higher scores represent a stronger of missing values. The assumption that data were missing
perceived parenting alliance. Scores in this sample ranged completely at random (MCAR) was evaluated with Little’s
from 20 to 100 for wives and 22 to 100 for husbands. MCAR Test (Little, 1988) in SPSS version 15.0, and was
Children’s Perceptions of Interparental Conflict Scale. found to be a reasonable assumption. Missing values were
The Children’s Perceptions of Interparental Conflict Scale imputed with each participant’s mean response to the other
(CPIC; Grych, Seid, & Fincham, 1992), a 49-item scale as- items of each respective scale, but only when the partici-
sesses children’s perspectives regarding their parents’ mar- pants had responded to at least 80% of that scale’s items.
This method of replacing missing values has been found to
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
ties, Threat, and Self-Blame (which was not used in this be statistically acceptable and appropriate (Schafer & Gra-
study). Higher scores on these subscales indicate more ham, 2002). In the few cases where missing values could
negative perceptions of interparental conflict (e.g., that con- not be replaced, listwise deletion was used. Means and
flict is frequent, often unresolved, and stable over time). The standard deviations for husbands, wives, and children are
CPIC has demonstrated good internal consistency and ex- presented in Table 1.
cellent concurrent and criterion validity (Grych et al., 1992). As also shown in Table 1, the correlations between hus-
The Conflict Properties and Threat subscales were used in bands’ and wives’ reports of positive forgiveness and neg-
the present study as measures of the children’s perceptions ative forgiveness, marital satisfaction, parenting alliance,
of marital conflict. The Conflict Properties subscale is cre- and trust, and their children’s perceptions of marital conflict
ated by summing the Intensity, Resolution, Frequency, and were largely associated in the expected directions (correla-
Stability dimensions; high scores indicate that children per- tions between the first two variables were only calculated
ceive conflict as more hostile, frequent, and poorly resolved. for couples in which both partners reported a betrayal). Thus
Scores on the Conflict Properties subscale ranged from 24 to we examined more complex associations among these vari-
69 in this sample. The Threat subscale consists of the ables in the analyses described below.
Threat, Coping Efficacy, and Triangulation dimensions;
high scores suggest that children feel more threatened and Dyadic Satisfaction
less able to cope with interparental conflict. Scores on the
Threat subscale ranged from 16 to 41. Coefficient alphas for As displayed in Table 2, the results of hierarchical mul-
this sample were .95 for the Conflict Properties subscale and tiple regression analyses supported the hypothesis that for-
.87 for the Threat subscale. giveness predicts marital satisfaction. The overall models
significantly predicted marital satisfaction in all models,
with a range of effect sizes (f2 ⫽ .18 – 2.13). When spouses’
Procedure own data were used to predict satisfaction (the top portion
of Table 2), a decrease in reports of negative forgiveness
Families were recruited randomly from a mailing list of uniquely predicted increased dyadic satisfaction for both
families in the area. The mailing list was purchased from a husbands and wives. When the cross-spousal data were used
direct mail company that compiles information from major to predict reports of satisfaction, both full models were
databases (e.g., auto registration, credit companies) to create significant: husbands’ forgiveness predicted wives satisfac-
mailing lists for commercial use. The mailing list provided tion (R2 ⫽ .31; F3,73 ⫽ 16.09; p ⬍ .01) and wives’ forgive-
the addresses of a large sample of families with children ness predicted husbands’ satisfaction (R2 ⫽ .48; F3,84 ⫽
between the ages of 11 and 16 who lived within designated 39.22; p ⬍ .01). However, reports of negative forgiveness
zip codes. Research assistants telephoned families to assess were the only unique predictors of satisfaction ( ⫽ ⫺.57,
their interest and eligibility for participating in the study. Of p ⬍ .01 and  ⫽ ⫺.68, p ⬍ .01, for husbands and wives
all eligible families, 18% participated, a response rate that is respectively).
consistent with other studies using similar recruiting meth- In order to test whether reports of marital conflict and
ods (e.g., Davila, Karney, & Bradbury, 1999; Gordon & trust mediate the association between forgiveness and mar-
Baucom, 2003). The packet of measures was mailed to ital satisfaction, we followed procedures outlined by Baron
participants’ homes, with husbands’, wives’, and children’s and Kenny (1986), and we also calculated the Sobel statistic
surveys packaged in separate envelopes and with a cover (Sobel, 1982) as an additional test of the hypothesized
letter instructing participants to refrain from sharing their mediational models. We conducted multiple regression
answers with each other and to be open and honest in their analyses using reports of negative forgiveness and positive
responses. The completed surveys and signed consent forms forgiveness dimensions as the predictors, marital satisfac-
were then either mailed back to the research lab or were tion as the dependent variable, and conflict and trust as the
picked up by a research assistant at the family’s home. hypothesized mediators. Notably, in these analyses, we de-
Families were given a $20 gift certificate as compensation. viated slightly from the usual procedures for mediation in
6 COOP GORDON, HUGHES, TOMCIK, DIXON, AND LITZINGER
Physical Aggression subscales (i.e., sum of aggressive conflict tactics reported for self and partner). Conflict Properties and Threat refer to those respective subscales of the
Note. FI-Pos. ⫽ Positive forgiveness as measured by the Forgiveness Inventory; FI-Neg. ⫽ Negative forgiveness as measured by the Forgiveness Inventory; DSat ⫽ Dyadic
Satisfaction subscale of the Dyadic Adjustment Scale; PAM ⫽ Parenting Alliance Measure; DTS ⫽ Dyadic Trust Scale; CTS ⫽ Conflict Tactics Scale scores, sum of Verbal and
that we were testing mediation using a set of variables rather
Threat
24.65
6.12
than variables considered individually. Thus, because we
were interpreting at the block level when testing mediation,
we chose to provide the individual parameters for variables
properties
Conflict
.72ⴱⴱ
in a block that contributed to significant change in R2, even
34.80
10.19
—
—
—
—
—
—
—
—
—
—
—
—
though these variables are not individually significant. As a
final note, the effect sizes for the following regression
equations were all in the large range (f 2 ⫽ .75 – 2.70), with
Husband
.26ⴱⴱ
the exception of the models predicting husbands’ conflict
.26ⴱ
CTS
37.91
57.60
—
—
—
—
—
—
—
—
—
—
—
from husbands’ forgiveness (f 2 ⫽ .18).
As shown in Table 2, the negative forgiveness dimension
significantly and negatively predicted wives’ satisfaction
.45ⴱⴱ
.51ⴱⴱ
.30ⴱⴱ
Wife
CTS
flict and trust, see Models 2 and 3 for wives), whereas the
positive forgiveness dimension was not a significant predic-
Husband
⫺.34ⴱⴱ
⫺.45ⴱⴱ
⫺.39ⴱⴱ
⫺.24ⴱ
DTS
45.74
10.14
⫺.51ⴱⴱ
⫺.28ⴱ
Wife
DTS
45.86
9.86
—
—
—
—
—
—
—
—
.35ⴱⴱ
.52ⴱⴱ
⫺.43ⴱⴱ
⫺.23ⴱⴱ
⫺.18
84.17
12.84
—
—
—
—
—
—
—
⫺.56ⴱⴱ
⫺.37ⴱⴱ
⫺.27ⴱ
PAM
Wife
⫺.55ⴱⴱ
⫺.50ⴱⴱ
⫺.49ⴱⴱ
.38ⴱⴱ
.35ⴱⴱ
⫺.43ⴱ
⫺.24ⴱ
.33ⴱ
17.82
6.35
.64ⴱⴱ
⫺.66ⴱⴱ
⫺.44ⴱⴱ
⫺.71ⴱⴱ
⫺.37ⴱⴱ
.33ⴱⴱ
.37ⴱⴱ
.55ⴱⴱ
.33ⴱⴱ
19.03
8.17
—
—
—
—
⫺.31ⴱⴱ
.27ⴱⴱ
.39ⴱⴱ
.42ⴱⴱ
FI-Pos.
.07
⫺.11
⫺.03
⫺.23
⫺.19
46.91
8.59
—
—
—
⫺.47ⴱⴱ
⫺.31ⴱⴱ
FI-Pos.
⫺.25ⴱ
Wife
⫺.21
49.42
7.80
—
—
39.33
6.71
—
.60ⴱⴱ
⫺.83ⴱⴱ
⫺.55ⴱⴱ
.69ⴱⴱ
.48ⴱⴱ
.70ⴱⴱ
.47ⴱⴱ
⫺.69ⴱⴱ
⫺.56ⴱⴱ
⫺.42ⴱⴱ
.15
39.36
6.71
.78
Husband PAM
Husband DSat
Husband DTS
Husband CTS
Wife PAM
Wife CTS
Table 1
Table 2
Regression Analyses Testing Whether Conflict and Trust Mediate Associations Between Forgiveness and Dyadic
Satisfaction
Wives Husbands
B SE  t R2 B SE  t R2
Model 1: Predicting Model 1: Predicting
DSat (n ⫽ 87): .68ⴱⴱ DSat (n ⫽ 74): .53ⴱⴱ
FI-Pos. .01 .08 .01 .08 FI-Pos. .07 .07 .09 1.08
FI-Neg. ⫺.72 .08 ⫺.82 ⫺9.30ⴱⴱ FI-Neg. ⫺.74 .09 ⫺.69 ⫺8.08ⴱⴱ
Model 2: Predicting Model 2: Predicting
CTS (n ⫽ 79): .43ⴱⴱ CTS (n ⫽ 72): .18ⴱⴱ
FI-Pos. ⫺.54 .67 ⫺.10 ⫺.80 FI-Pos. ⫺.09 .65 ⫺.02 ⫺.14
FI-Neg. 2.92 .62 .58 4.67ⴱⴱ FI-Neg. 3.22 .87 .42 3.70ⴱⴱ
Model 3: Predicting Model 3: Predicting
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
FI-Pos. .12 .15 .09 .82 FI-Pos. .33 .12 .28 2.71ⴱⴱ
FI-Neg. ⫺.82 .14 ⫺.64 ⫺5.84ⴱⴱ FI-Neg. ⫺.67 .16 ⫺.42 ⫺4.09ⴱⴱ
Model 4: Testing Model 4: Testing
Mediation, Mediation,
Predicting DSat Predicting DSat
(n ⫽ 79): .73ⴱⴱ (n ⫽ 72): .62ⴱⴱ
FI-Pos. .01 .08 .01 .08 FI-Pos. .04 .07 .05 .66
FI-Neg. ⫺.51 .10 ⫺.57 ⫺5.24ⴱⴱ FI-Neg. ⫺.57 .10 ⫺.53 ⫺5.92ⴱⴱ
CTS ⫺.04 .01 ⫺.21 ⫺2.51ⴱⴱ CTS ⫺.02 .01 ⫺.17 ⫺2.01ⴱ
DTS .12 .06 .17 1.95ⴱ DTS .17 .06 .26 2.80ⴱⴱ
Note. Please see the note following Table 1 for an explanation of variable names.
ⴱ
p ⬍ .05. ⴱⴱ p ⬍ .01.
proximately 20% of the variance of positive forgiveness regressions were in the moderate to large range (f 2 ⫽ .19 ⫺
with dyadic satisfaction for husbands. 1.08).
When spouses’ own data were used to predict parenting
Parental Functioning alliance for both wives and husbands (the top portion of
Table 3 displays the results of the regression analyses Table 3), dyadic satisfaction was the strongest predictor of
testing whether spouses’ reports of positive forgiveness parenting alliance; however, the positive forgiveness di-
and negative forgiveness predicted spouses’ perceptions mension also uniquely predicted parenting alliance. Reports
of their parenting alliance. Dyadic satisfaction was in- of more positive forgiveness and more satisfaction predicted
cluded in these analyses to control for spouses’ feelings perceptions of a stronger parenting alliance. When the
regarding their marital relationship. Effect sizes for these cross-spousal data were used to predict parenting alliance,
Table 3
Regression Analyses Using Parents’ Reported Forgiveness to Predict Perceptions of Parenting Alliance, Controlling for
Dyadic Satisfaction
B SE  t R2 R2⌬
Wives’ PAM (n ⫽ 87):
Block 1: .46ⴱⴱ
Wives’ FI-Pos. .53 .24 .26 2.26ⴱ
Wives’ FI-Neg. ⫺.93 .23 ⫺.47 ⫺4.12ⴱⴱ
Block 2: .52ⴱⴱ .06ⴱⴱ
ⴱⴱ
Wives’ DSat .92 .30 .41 3.03
Wives’ FI-Pos. .53 .23 .26 2.33ⴱ
Wives’ FI-Neg. ⫺.27 .31 ⫺.14 ⫺.87
Husbands’ PAM (n ⫽ 74):
Block 1: .25ⴱⴱ
Husbands’ FI-Pos. .41 .17 .26 2.45ⴱ
Husbands’ FI-Neg. ⫺.74 .23 ⫺.35 ⫺3.28ⴱⴱ
Block 2: .31ⴱⴱ .05ⴱ
ⴱ
Husbands’ DSat .65 .28 .33 2.29
Husbands’ FI-Pos. .36 .17 .23 2.21ⴱ
Husbands’ FI-Neg. ⫺.26 .31 ⫺.12 ⫺.85
Note. Please see the note following Table 1 for an explanation of variable names.
ⴱ
p ⬍ .05. ⴱⴱ p ⬍ .01.
8 COOP GORDON, HUGHES, TOMCIK, DIXON, AND LITZINGER
both full models were significant (R2 ⫽ .31; F3,73 ⫽ 15.90; effect size for the regression model predicting children’s
p ⬍ .01 and R2 ⫽ .20; F3,84 ⫽ 10.47; p ⬍ .01, for husbands perception of parental conflict from wives’ forgiveness was
predicting wives’ parenting alliance and wives predicting in the moderate range (f2 ⫽ .47), and the effect size for the
husbands’ parenting alliance respectively). Husbands’ neg- model predicting children’s perceived threat to the marriage
ative forgiveness was the strongest predictor of wives’ was in the small range (f2 ⫽ .14). The bottom portion of
parenting alliance ( ⫽ ⫺.38, p ⬍ .01), with more negative Table 4 displays the results for the models predicting chil-
forgiveness predicting perceptions of a weaker parenting dren’s perceptions of the marital conflict (including inten-
alliance, and satisfaction was notably nonsignificant in this sity, resolution, frequency, and stability). Wives’ reports of
model. In the prediction of husband’s parenting alliance, negative forgiveness predicted children’s perceptions of
although overall regression model was significant, but no marital conflict in the expected direction. Also as shown,
individual predictor reached significance. Wives’ negative only husbands’ reports of negative forgiveness were signif-
forgiveness predicted husbands’ parenting alliance in the icantly predictive of children’s perceptions of marital con-
first block ( ⫽ ⫺.34, p ⬍ .05), but when wives’ satisfac- flict.
tion was added to the model, negative forgiveness became
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Table 5
Regression Analyses Testing Whether Conflict and Parenting Alliance Mediate Associations Between Forgiveness and
Child Perceptions
Wives’ forgiveness Husbands’ forgiveness
B SE  t R2 B SE  t R2
Model 1: Predicting Husbands’ Model 1: Predicting Wives’ CTS
CTS (n ⫽ 79): .25ⴱⴱ (n ⫽ 68): .24ⴱⴱ
Wives’ FI-Pos. .58 1.14 .07 .51 Husbands’ FI-Pos. ⫺.14 .51 ⫺.03 ⫺.27
Wives’ FI-Neg. 4.21 1.07 .55 3.92ⴱⴱ Husbands’ FI-Neg. 2.84 .66 .48 4.29ⴱⴱ
Model 2: Predicting Husbands’ Model 2: Predicting Wives’
PAM (n ⫽ 84): .20ⴱⴱ PAM (n ⫽ 74): .33ⴱⴱ
Wives’ FI-Pos. .24 .24 .14 .99 Husbands’ FI-Pos. .18 .20 .10 .95
Wives’ FI-Neg. ⫺.56 .23 ⫺.34 ⫺2.41ⴱ Husbands’ FI-Neg. ⫺1.39 .27 ⫺.54 ⫺5.23ⴱⴱ
Model 3: Testing Mediation, Model 3: Testing Mediation,
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
husbands, it is likely that facilitating an increase in under- a strong, cohesive parenting alliance (Abidin & Brunner,
standing and a release from anger and blame might help 1995) and a satisfying marital relationship. Therefore, it is
men develop more trust for their partners, and thus poten- not surprising that partners’ levels of negative forgiveness
tially regain greater satisfaction following a betrayal. How- have more impact on their spouses than do their levels of
ever, as these results are only cross-sectional, longitudinal positive forgiveness. Further, the presence of these kinds of
studies are needed to fully confirm mediation; it is possible vengeful cognitions and destructive behaviors are more
that trusting husbands are better able to forgive their wives. likely to have a more immediate and stronger corrosive
Furthermore, even when controlling for levels of marital effect on the partner’s sense of closeness and satisfaction,
satisfaction, positive forgiveness uniquely predicted parent- than would simply an absence of positive behaviors (Gott-
ing alliance, yet negative forgiveness did not. As negative man & Silver, 1999). An absence of positive forgiveness
forgiveness was uniquely predictive of marital satisfaction, may not bring couples closer together, but it may not be as
and was also correlated with parenting alliance, it is possible destructive as an increase in the negative forgiveness di-
that its association with the parenting alliance occurs solely mension.
through its relationship to marital satisfaction. However, Just as the effects of forgiveness resonate beyond marital
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
positive forgiveness, or the ability to put the event behind satisfaction into the parenting realm, they may similarly
oneself, move on, and focus on understanding one’s partner ripple out to children, although the impact appears to be-
rather than blaming, may allow one to be a good parenting come less direct and weaker. Both positive and negative
partner, regardless of the level of marital satisfaction. forgiveness are correlated with both children’s perceptions
Therefore, a wife may be dissatisfied with her marriage, but of threat from marital conflict and their assessment of mar-
to the extent that she is able to put a negative event behind ital conflict properties. In analyses in which both positive
her and work to come to peace with it, she may still be able and negative forgiveness are regressed on children’s per-
to develop a better parenting alliance with her partner than ceptions, only negative forgiveness is significantly predic-
would a wife who cannot put the event behind her. How- tive of children’s perceptions. As wives and husbands re-
ever, as these measures were all self-reported and because ported less negative affect and punitiveness regarding the
they were highly correlated, these results must be inter- betrayal, children reported perceiving their parents’ marital
preted with caution as they might reflect common method conflict as less aggressive, better resolved, less frequent, and
variance. At the same time, these findings and the interpre- less persistent. However, when forgiveness is considered
tation given here are consistent with Fincham and Beach’s along with levels of marital conflict and the parenting alli-
(2002) suggestion that the positive dimensions of forgive- ance, its associations with children’s perceptions of marital
ness are likely to facilitate an increase in other positive conflict are significantly decreased, suggesting that it has its
aspects of the relationship. effect primarily through disruptions in marital and parental
In contrast, when cross-spousal reports were evaluated functioning. Still, forgiveness, or more aptly lack thereof
(e.g., husbands’ forgiveness predicting wives’ parenting (i.e., more negative forgiveness and lowered positive for-
alliance and vice versa) the familiar pattern in which neg- giveness), is associated with decreases in marital satisfac-
ative forgiveness predicts outcome re-emerged. For exam- tion, a weaker parenting alliance, and children’s perceptions
ple, if a husband remains angry, punitive, and unforgiving, of poorer marital functioning. These findings are notable,
then the wife is likely to report a weak parenting alliance. particularly in light of past research suggesting that chil-
Notably, these findings hold true even when controlling for dren’s perceptions of their parents’ marital functioning have
feelings of marital satisfaction. Thus, in terms of the effects a variety of implications for their individual well-being
of one’s own forgiveness of one’s partner, it seems that the (Grych & Fincham, 1990; Grych, Seid, & Fincham, 1992).
angry and punitive behaviors characteristic of the negative We wish to emphasize one major caveat in interpreting
forgiveness dimension may have a unique impact on the findings of the present study and comparing them to
spouses’ feelings about being a parenting team. In sum, this other discussions of forgiveness in the literature: The con-
pattern of results suggests that grudge-holding and venge- ceptualization of positive forgiveness here differs from
fulness is likely to weaken the bond between parents in other conceptualizations of positive forgiveness in the lit-
many ways. It appears to be extremely difficult for spouses erature. Thus, its failure to predict dependent variables over
to work as a team if they are unable to put past betrayals and above negative forgiveness in some of the analyses may
behind them. be due to the way it was conceptualized and measured. The
These hypotheses also are consistent with Fincham et pattern of results may have been different if the measure of
al.’s (2006) suggestions that persons high on the negative positive forgiveness had included aspects of empathy,
forgiveness dimension are likely to be engaging in both warmth, compassion, and reconciliation. However, the ap-
more avoidance behaviors and increased psychological ag- propriateness of including these elements is debatable, par-
gression toward their partners. Thus, when individuals re- ticularly if one considers the well-established conceptual
main angry, grudge-holding, and vengeful about a betrayal, differences between forgiveness and reconciliation (e.g.,
they are more likely to withdraw and/or lash out at their Fincham et al., 2006). If forgiveness is considered as nec-
partners in destructive ways. These behaviors, which may essarily encompassing these qualities, particularly reconcil-
be very tangible and obvious to the partner, would be likely iation, then forgiveness begins to carry dangers with it, such
to alter the partner’s perceptions of trust and good will, and as increasing the likelihood that people remain in abusive
these positive feelings are considered necessary to establish situations (e.g., Gordon, Burton, & Porter, 2004). One way
FORGIVENESS AND FAMILY FUNCTIONING 11
to address this conceptual issue would be for the field to examine how generalizable these processes are across eth-
move to dimensional models of forgiveness, such as those nicity, culture, or socioeconomic status. Forgiveness in var-
suggested by Worthington (2005) or Fincham et al. (2006). ious populations sorely requires more study and the reader
A healthy form of forgiveness in continually abusive situ- should proceed with great caution when applying these
ations might mean low scores on the negative forgiveness findings to diverse groups. Furthermore, it is unclear how
dimension and low scores on the positive forgiveness di- individuals who declined to participate in the study might
mension (which encompasses warmer, more positive rela- have answered these questions, which introduces a self-
tional outcomes), thus putting individuals at a reduced risk selection bias. It is unclear at this time in what direction that
of staying in dangerous situations, but still allowing them bias might lean. Although the use of the database from the
release from negative emotional experiences. Alternatively, direct mail company provided a large list of potential fam-
healthy forgiveness in reconciling spouses might mean low ilies and thus avoided some of the selection bias that occurs
scores on the negative forgiveness dimension and high when recruiting solely from ads and fliers, it was likely not
scores on the positive forgiveness dimension, allowing for an exhaustive list and most probably was less representative
better marital satisfaction and parental functioning. of individuals in the lowest socioeconomic range. However,
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
There are several limitations to this study that bear con- despite the fact that the sample is not representative of the
sideration. First, the current study suffers from a lack of U.S. culture at large, it is representative of the location from
observational measures that might offer additional, behav- which it is drawn.
ioral measures of constructs studied here (e.g., conflict
behaviors). It would be interesting to assess these constructs Implications
via interactions between partners to observe how these more
abstract concepts, such as levels of positive and negative Forgiveness of specific, self-identified major marital be-
forgiveness, are manifested behaviorally in the ways in trayals appears to have a significant and unique association
which spouses talk with one another in both general conflict with both marital satisfaction and the parenting alliance, and
situations and in situations specific to the betrayal. One forgiveness also has an indirect association with child func-
would suspect that individuals who were truly forgiving tioning through its associations with marital conflict and the
should engage differently with their partners than would parenting alliance. As such, these findings underscore the
individuals who were not forgiving. However, as this study need for marital and family researchers to further examine
does have multiple reporters, and cross-spousal and child forgiveness. It is clear that forgiveness is likely to be an
reports are all consistent with one another, these results are influential part of family functioning and as such deserves
at least indicative that the self-report measures are capturing close attention, as there are many questions remaining to be
truly systemic constructs. answered.
Also, this study is cross-sectional, and as such is unable Forgiveness also deserves attention from couple thera-
to reveal which comes first, forgiveness or marital satisfac- pists. Much of the empirically validated couple therapies
tion, or forgiveness or parenting alliance. It is possible that available today (e.g., Gurman & Jacobson, 2002) place the
relationship satisfaction predicts forgiveness, and indeed majority of their emphasis on addressing current interac-
some researchers have argued this point (e.g., Fincham, tions in couples’ relationships; an exception is Emotions
Paleari, & Regalia, 2002; McCullough et al., 1998); how- Focused Couple Therapy (EFT), which addresses
ever, many agree that the association between these con- “attachment-related injuries” in order to help couples with
structs is likely to be bidirectional. It is just as theoretically their current problems (Makinen & Johnson, 2006). Most
satisfying to suggest that a major betrayal affects one’s level therapies do not spend a great deal of time addressing past
of forgiveness, which in turn affects how satisfied a person’s betrayals. These data suggest that the past may inform the
relationship is, and that these associations may then become present in the form of grudge-holding, resentment, avoid-
reciprocal. Notably, this view has been argued most recently ance, and lack of understanding and blame between the
in the literature (Paleari et al., 2005). Similarly, this study partners. It is unlikely that simply addressing current con-
examined the possibility that forgiveness is mediated by flicts will always resolve past betrayals or address the un-
conflict and trust and found some evidence of partial me- derlying disruptions in relational assumptions that result
diation by trust for the spouses; however, a better test of from betrayals (e.g., Gordon & Baucom, 1998).
mediation would involve at least three waves of data that Thus, in addition to focusing on present conflicts, couple
could show that forgiveness at Time 1 predicts changes in and family therapists would likely be well-served in evalu-
trust and conflict at Time 2, which in turn predicts changes ating whether there are major unresolved betrayals in cou-
in marital satisfaction at Time 3. Currently, the correlations ples’ pasts, and then working with these couples to process
between these factors when measured concurrently do not these betrayals. Indeed, data from a large marital treatment-
allow for determining direction of causality. Unfortunately, outcome study suggests that unknown betrayals, such as
the field must await multi-wave, longitudinal studies of affairs, can have a very detrimental effect upon the outcome
marital functioning that include measures of forgiveness to of therapy (Atkins, Yi, Baucom, & Christensen, 2005). As
more fully tease these associations apart. the present study revealed, forgiveness has implications for
Finally, the sample of this study is limited primarily to aspects of family functioning that extend well beyond the
white, middle-class, educated individuals. Further, there couple’s marital functioning (i.e., parenting alliance, chil-
was a low response rate overall; thus it is difficult to dren’s perceptions). Thus, clinicians who remain unaware
12 COOP GORDON, HUGHES, TOMCIK, DIXON, AND LITZINGER
of lingering resentments or past betrayals might be failing to The moral development of forgiveness. In W. Kurtines & J.
address an important component of the family’s difficulties. Gewirtz (Eds.), Handbook of moral behavior and development
Furthermore, this study’s findings point to the need for (pp. 123–152). Hillsdale, NJ: Erlbaum.
therapists to directly assess for and address perceptions of Fincham, F. D., & Beach, S. R. H. (2002). Forgiveness in mar-
trust, and perhaps use strategies aimed at slowly and delib- riage: Implications for psychological aggression and constructive
communication. Personal Relationships, 9, 239 –251.
erately rebuilding trust within the couple after a betrayal. Fincham, F. D., Hall, J., & Beach, S. R. H. (2006). Forgiveness in
Once a betrayal has been uncovered, there are several ex- marriage: Current status and future directions. Family Relations,
isting treatments in the literature with some empirical evi- 55, 415– 427.
dence for therapists to use (e.g., Gordon et al., 2004; Maki- Fincham, F. D., Paleari, F. G., & Regalia, C. (2002). Forgiveness
nen & Johnson, 2006). in marriage: The role of relationship quality, attributions, and
Similarly, policy makers and therapists who are con- empathy. Personal Relationships, 9, 27–37.
cerned with conflict mediation, mediated divorces, premar- Finkel, E. J., Rusbult, C. E., Kumashiro, M., & Hannon, P. A.
ital interventions, or otherwise assisting couples either to (2002). Dealing with betrayal in close relationships: Does com-
prevent or ameliorate the effects of divorce should also mitment promote forgiveness? Journal of Personality and Social
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
forgiving in close relationships: II. Theoretical elaboration and Straus, M. A. (1979). Measuring intrafamily conflict and violence:
measurement. Journal of Personality and Social Psychology, 75, The conflict tactics scale. Journal of Marriage and the Family,
1586 –1603. 41, 75– 88.
Paleari, F. G., Regalia, C., & Fincham, F. D. (2005). Marital Thompson, L. Y., Snyder, C. R., Hoffman, L., Michael, S. T.,
quality, forgiveness, empathy, and rumination: A longitudinal Rasmussen, H. N., Billings, L. S., et al. (2005). Dispositional
analysis. Personality and Social Psychology Bulletin, 31, 368 – forgiveness of self, others, and situations. Journal of Personality,
378. 73, 313–359.
Preacher, K. J., & Leonardelli, G. J. (2001). Calculation for the U.S. Census Bureau (2000). Census 2000 survey profiles. Re-
Sobel test: An interactive calculation tool for mediation tests trieved June 14, 2002, from http://www.census.gov
[Computer software]. Available from http://www.psych.ku.edu/ Weiss, R. L. (1980). Strategic behavioral marital therapy: Toward
preacher/sobel/sobel.htm a model for assessment and intervention. In J. P. Vincent (Ed.),
Rusbult, C. E., Kumashiro, M., Finkel, E. J., & Wildschut, T. Advances in family intervention, assessment, and theory (Vol. 1,
(2002). The war of the roses: An interdependence analysis of pp. 229 –271). Greenwich, CT: JAI Press.
betrayal and forgiveness. In P. Noller (Ed.) & Feeney, J. A. (Ed.),
Worthington, E. L. (2003). Forgiving the devil: Coming to terms
Understanding marriage: Developments in the study of couple
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.