Content uploaded by Jo-Ann Tsang
Author content
All content in this area was uploaded by Jo-Ann Tsang on Apr 30, 2014
Content may be subject to copyright.
This article was downloaded by: [Baylor University Libraries], [Ms Jo-Ann Tsang]
On: 29 April 2014, At: 10:56
Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK
The Journal of Positive Psychology: Dedicated to
furthering research and promoting good practice
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/rpos20
Tipping the scales: Conciliatory behavior and the
morality of self-forgiveness
Thomas P. Carpentera, Robert D. Carlislea & Jo-Ann Tsanga
a Department of Psychology and Neuroscience, Baylor University, Waco, TX 76798, USA
Published online: 22 Apr 2014.
To cite this article: Thomas P. Carpenter, Robert D. Carlisle & Jo-Ann Tsang (2014): Tipping the scales: Conciliatory behavior
and the morality of self-forgiveness, The Journal of Positive Psychology: Dedicated to furthering research and promoting
good practice, DOI: 10.1080/17439760.2014.910823
To link to this article: http://dx.doi.org/10.1080/17439760.2014.910823
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
Tipping the scales: Conciliatory behavior and the morality of self-forgiveness
Thomas P. Carpenter, Robert D. Carlisle and Jo-Ann Tsang*
Department of Psychology and Neuroscience, Baylor University, Waco, TX 76798, USA
(Received 30 May 2013; accepted 21 March 2014)
Two studies examined whether conciliatory behavior aids self-forgiveness and whether it does so in part by making it
seem more morally appropriate. Participants in Study 1 (n= 269) completed an offense-recall procedure; participants in
Study 2 (n= 208) imagined a social transgression under conciliatory behavior (yes, no) and receipt of forgiveness (no,
ambiguous, yes) conditions. Conciliatory behavior predicted (Study 1) and caused (Study 2) elevated self-forgiveness
and increased perceptions of the moral appropriateness of self-forgiveness. Perceived morality consistently mediated the
effect of conciliatory behavior on self-forgiveness. Received forgiveness and guilt were considered as additional mecha-
nisms, but received mixed support. Results suggest that conciliatory behavior may influence self-forgiveness in part by
satisfying moral prerequisites for self-forgiveness.
Keywords: self-forgiveness; apology; guilt; morality; amends; justice
Although research on interpersonal forgiveness has pro-
liferated (Fehr, Gelfand, & Nag, 2010), self-forgiveness
has received less attention. Forgiveness of self and
others are related processes, each involving a reduction
of negative motivations and emotions and an increase
in positive motivations and emotions. However, they
differ in one important respect: in self-forgiveness, the
offender is the primary granter of forgiveness. Conse-
quently, self-forgiveness may have different moral
dynamics than interpersonal forgiveness. For example,
offenders may be tempted to be lenient on themselves,
or they may feel obligated to refrain from forgiving
themselves so long as the victim suffers or until the
victim offers forgiveness. Not surprisingly, the moral
appropriateness of self-forgiveness has been repeatedly
questioned, especially in situations where remorse has
not been expressed, apologies have not been offered,
and amends have not been made (Lamb, 2002;
Murphy, 2002).
In the present research, we explore the importance
of offenders’perceptions of the appropriateness of for-
giving themselves and the role of these perceptions in
self-forgiveness. We test whether conciliatory behavior
(e.g. apology and making amends) makes self-forgive-
ness seem more morally appropriate and whether
this perceived morality may mediate any effect of
conciliatory behavior on self-forgiveness. We also
consider two other mechanisms (guilt reduction and
received forgiveness) and their interplay with perceived
morality.
Morality of self-forgiveness
Because of the moral ambiguity inherent in self-forgive-
ness, perceptions of its moral inappropriateness may be
important in the self-forgiveness process. By moral inap-
propriateness, we mean a moral sense that it would be
unjust to release oneself from resentment and make a
self-forgiving motivational transformation.
Theorists have long offered reservations about the
morality of self-forgiveness (Horsbrugh, 1974). For
example, Murphy (2002) argued that offenders morally
deserve to feel bad about their actions. However, they
have also argued that self-forgiveness becomes morally
appropriate following conciliatory behavior, such as resti-
tution (Holmgren, 2002; Horsbrugh, 1974) and repen-
tance (Murphy, 2002). Qualitative research suggests lay
individuals feel similarly, expressing reluctance to self-
forgive unless they have done ‘something to make it
right,’such as apologizing or making amends (Ingersoll-
Dayton & Krause, 2005, p. 279). No quantitative
research yet exists exploring whether conciliatory
behavior (or any factor) influences the perceived moral
appropriateness of self-forgiveness. To the degree that
self-forgiveness is a choice (Worthington, 2006), such
perceptions could be especially important.
Self-forgiveness and conciliatory behavior
We propose that the perceived moral appropriateness
of self-forgiveness may partly explain why concilia-
tory behavior influences self-forgiveness. Conciliatory
*Corresponding author. Email: Joann_Tsang@Baylor.edu
© 2014 Taylor & Francis
The Journal of Positive Psychology, 2014
http://dx.doi.org/10.1080/17439760.2014.910823
Downloaded by [Baylor University Libraries], [Ms Jo-Ann Tsang] at 10:56 29 April 2014
behavior is a class of reparative responses that includes
apologizing, making amends, restitution, and seeking for-
giveness (Fisher & Exline, 2010; Hall & Fincham,
2005). This overall conciliatory response serves the pur-
pose of expressing remorse and a desire to repair the
damage done by one’s wrongs, reversing the dynamic
caused by the offense (Exline et al., 2011). Conse-
quently, self-forgiveness theorists (e.g. Hall & Fincham,
2005, 2008) have generally treated conciliatory behavior
as an overall reparative response that is psychologically
important for self-forgiveness. According to Hall and
Fincham’s(2005) model, conciliatory behavior should
influence self-forgiveness both by reducing guilt and
leading to forgiveness by others.
Empirical support for an effect of conciliatory behavior
on self-forgiveness is mixed. Hall and Fincham (2008)
found conciliatory behavior predicted self-forgiveness for
a recalled transgression. Exline, Root, Yadavalli, Martin,
and Fisher (2011) found a self-forgiveness intervention
more effective for those who had previously made concil-
iatory gestures. They also found that post-intervention
conciliatory behavior predicted intervention efficacy, but
this did not replicate. Witvliet, Ludwig, and Bauer (2002)
found that imagining conciliatory behavior yielded more
self-forgiveness than did rumination; however, it is
unknown to what degree this reflects the benefits of concil-
iatory behavior or the detriments of rumination.
Other studies have offered less support. Examining
recalled transgressions, Cafaro and Exline (2003) found
no relationship between self-forgiveness and seeking for-
giveness or confession. Zechmeister and Romero (2002)
similarly found no association between self-forgiveness
and restitution and only a small association with apology.
Fisher and Exline (2006) found conciliatory behavior
predicted greater effort required to feel better about a
past offense. Finally, Rangganadhan and Todorov (2010)
found no relationship between trait self-forgiveness and
conciliatory behavior. The present research will further
test whether conciliatory behavior aids self-forgiveness
and whether this effect may in part depend on moral per-
ceptions of self-forgiveness.
Conciliatory behavior might affect the morality of
self-forgiveness by addressing justice. Transgressions
create unjust situations in which the individual who
experiences the most suffering (the victim) is not the one
who contributes the most harm (the offender). According
to equity theory (Adams, 1965), people should be moti-
vated to reduce this imbalance; offenders may respond
by foregoing self-forgiveness and allowing themselves to
suffer (Exline et al., 2011). Self-forgiveness without con-
ciliatory behavior may be seen as unjust, both because
the offender foregoes punishment and because the offen-
der benefits from self-forgiveness while the victim
remains hurt. Using Worthington’s(2006) terminology,
the ‘injustice gap’is widened as the situation becomes
more inequitable. By directly addressing injustice,
conciliatory behavior may alleviate these concerns.
Conciliatory behavior may also increase the morality of
self-forgiveness by demonstrating that the offender has
taken appropriate responsibility. Vitz and Meade (2011)
note that individuals may need to prove their remorse to
themselves before ‘earning the right’(p. 253) to self-for-
give in order to demonstrate that they are not shirking
their obligations.
Received forgiveness and guilt reduction
Existing models emphasize guilt reduction and received
forgiveness as pathways by which conciliatory behavior
may aid self-forgiveness. We now consider how these
pathways may relate to the perceived morality of self-
forgiveness. Worthington (2006) argued that victims
influence offenders’self-forgiveness by reinforcing
voices of self-criticism or modeling forgiving responses.
Received forgiveness may reduce guilt by signaling that
relationship damage has been mitigated and providing
closure. Additionally, received forgiveness may increase
perceived morality of self-forgiveness. By embracing for-
giveness, the victim may communicate that forgiveness
is appropriate and that the offense and offender are wor-
thy of forgiveness. Evidence for the effect of received
forgiveness is limited. Some studies have reported asso-
ciations between received forgiveness and self-forgive-
ness (Cafaro & Exline, 2003; Hall & Fincham, 2008),
and others have reported null findings (Witvliet et al.,
2002; Zechmeister & Romero, 2002), signaling the need
for more research.
Conciliatory behavior may also impact self-forgive-
ness by reducing guilt (Hall & Fincham, 2005). Concilia-
tory behavior both follows (Cryder, Springer, &
Morewedge, 2012) and reduces guilt (Meek, Albright, &
McMinn, 1995). Whether or not this aids self-forgive-
ness is unclear, with some studies suggesting a unique
relationship between guilt and self-forgiveness (Hall &
Fincham, 2008), and others not (Exline et al., 2011;
Rangganadhan & Todorov, 2010). Because guilt provides
informational cues as to the moral valence of one’s
actions (Ketelaar & Au, 2003), guilt reduction might
make self-forgiveness seem more moral.
The present research
In the present research, we suggest that conciliatory
behavior may aid self-forgiveness in part by helping
individuals feel morally permitted to forgive themselves.
Specifically, we seek to test the hypothesis that morality
of self-forgiveness mediates the relationship between
conciliatory behavior and self-forgiveness directly as
well as potentially indirectly, following received forgive-
ness and guilt reduction (see Figure 1).
2T.P. Carpenter et al.
Downloaded by [Baylor University Libraries], [Ms Jo-Ann Tsang] at 10:56 29 April 2014
Study 1
Study 1 examined the impact of conciliatory behavior on
self-forgiveness for real-world offenses. Participants
recalled separate offenses for which they had and had
not apologized, allowing us to examine within-subjects
differences in self-forgiveness as a function of concilia-
tory behavior. Participants also recalled an offense for
which no apology criterion was specified. This allowed
us to conduct between-subjects mediation path analyses
without constraints on the range of conciliatory behavior
and to examine potential mediators.
Method
Participants
Participants were 269 undergraduate psychology students
(184 female, 80 male, and 5 unknown) at a mid-sized
private religious university.
1
Participants received course
credit for participation. Data collection was terminated at
the end of one semester.
Procedure
Participants completed the study online, which consisted
of three counterbalanced offense-recall prompts. For each
prompt, participants recalled a transgression in which
they were the perpetrator, followed by offense-specific
self-report measures. Participants completed all measures
for one offense before recalling subsequent offenses.
For the general-offense prompt, participants were
told:
Now we would like to ask you to remember a time in
which you seriously hurt or offended someone. Please
try to think of an especially important or memorable
experience rather than a trivial one. Take a few minutes
to remember the incident in as much detail as you can.
Participants were given a large text box in which to type
before completing the offense-specific measures. The
three prompts were identical, except for the first sen-
tence. For the apologized-offense prompt, participants
were told: ‘Now we would like to ask you to remember
a time in which you seriously hurt or offended someone,
for which you have apologized.’For the no-apology
prompt, this sentence ended ‘… for which you have not
apologized.’
Measures
Self-forgiveness
Participants’self-forgiveness for each offense was mea-
sured with the item, ‘To what extent have you forgiven
yourself for the offense?’(Hall & Fincham, 2008).
2
Par-
ticipants responded using a 7-point scale (1 = definitely
have NOT forgiven myself,7=have totally forgiven
myself).
Conciliatory behavior
For each offense, participants indicated on a 9-point
scale (1 = not very,9=very much) how much they had
engaged in five conciliatory behaviors: apology, confes-
sion, asking forgiveness, seeking forgiveness, and restitu-
tion. These items were culled from prior research (e.g.
Fisher & Exline, 2006). Prior to analysis, an exploratory
factor analysis using principal axis factoring was con-
ducted to verify that all items represented a general con-
ciliatory response. Consistent with predictions, one
‘conciliatory’factor was identified that explained 75.40%
of the variance, a strong one-factor solution. Cronbach’s
αs were excellent: 0.93 for the general offenses, 0.91 for
the apologized offenses, and 0.90 for the non-apologized
offenses.
Moral appropriateness of self-forgiveness
The degree to which participants saw self-forgiveness as
morally appropriate was assessed using the item, ‘To
what extent would it be morally right to forgive yourself
for the offense?’Participants responded using a 9-point
scale (1 = not very right,9=extremely right).
Received forgiveness
Received forgiveness was measured with the item, ‘To
what extent do you feel like the person/people you hurt
have forgiven you for the offense?’Participants
responded using a 9-point scale (1 = definitely has NOT
forgiven me,9=has totally forgiven me).
Guilt and shame
Offense-specific guilt and shame were assessed with the
remorse and self-condemnation scales (Exline et al.,
2011; Fisher & Exline, 2006). Participants read the
Conciliatory
Behavior
Received
Forgiveness Guilt Moral to
Self Forgive
Self
Forgiveness
Figure 1. Theoretical model depicting potential relationships
among primary study variables.
The Journal of Positive Psychology 3
Downloaded by [Baylor University Libraries], [Ms Jo-Ann Tsang] at 10:56 29 April 2014
prompt, ‘When you think about the offense you commit-
ted, to what extent do you feel …’ followed by 10 items,
each rated on an 11-point scale (1 = not at all,11=extre-
mely). The six remorse items measure offense-specific
guilt. Sample items included ‘guilty,’‘remorse,’and
‘regret about what you did.’Cronbach’sαs were excel-
lent: 0.93 for both the general and apologized offenses,
and 0.95 for the non-apologized offenses. The four self-
condemnation items measure offense-specific shame,
with items such as ‘like a bad person’and ‘hateful
toward yourself.’Reliabilities were good: 0.92 for gen-
eral offenses, 0.89 for the apologized offenses, and 0.91
for the non-apologized offenses.
Responsibility
Attributions of responsibility were assessed using items
developed by Fisher and Exline (2006). Participants
responded to five items on an 11-point scale (0 = com-
pletely disagree,10=completely agree) indicating the
degree to which they perceived themselves as responsi-
ble for each offense. Sample items include ‘I feel I was
responsible for what happened’and ‘this was clearly my
fault.’Cronbach’sαs were satisfactory: 0.79 for the gen-
eral offenses, 0.81 for the apologized offenses, and 0.82
for the non-apologized offenses.
Offense severity
Offense severity was measured by asking participants to
rate on a 9-point scale (1 = not at all,9=extremely) the
degree to which they considered their behavior ‘seri-
ous,’‘harmful,’and ‘immoral.’These items were drawn
from prior studies and loaded on a single factor (Exline
et al., 2011; Fisher & Exline, 2006). Cronbach’sαs
were good: 0.83 for the general offenses, 0.81 for the
apologized offenses, and 0.80 for the non-apologized
offenses.
Time since the offense
Participants indicated duration in time (in months) since
each offense. This was included as a control as time pre-
dicts self-forgiveness and conciliatory behavior (Hall &
Fincham, 2008).
Pre-offense closeness
Participants indicated how close they were to each
person they hurt prior to the offense on a 9-point scale
(1 = not very close, 9=extremely close). This was
included as a control because it covaries with guilt and
conciliatory behavior (Exline et al., 2011; Fisher &
Exline, 2006).
Results
Offenses and response rates
Of the 269 participants, 15 did not report a general
offense (6%), 9 did not report an apologized offense
(3%), and 31 did not report an offense for which they
had not apologized (11%). Response rates decreased for
successive prompts (97% for the first prompt, 95% for
the second prompt, and 91% for the third prompt). This
left 234 participants (155 female, 76 male, and 3
unknown) who completed both the apology and
no-apology prompts.
Offenses involved hurtful words (38%), inconsiderate
acts (23%), trust violations (12%), physical injury (7%),
social rejection (5%), break-ups (5%), romantic betrayal
(5%), and gossip (4%). Time since the offense and per-
ceived morality of self-forgiveness were square-root
transformed to improve normality prior to analysis. See
Tables 1and 2for descriptive statistics and correlations.
Rates of offense type did not differ by condition,
χ
2
(14) = 11.08, p= 0.68, Cramer’sV= 0.09, indicating
that heterogeneity of offense types did not differ system-
atically across prompts. The effects below were not
moderated by gender, thus results are reported for both
genders combined.
Comparing apology and no-apology prompts
We first compared the apology and no-apology prompts.
Manipulation check
We checked that the prompts elicited different levels of
conciliatory behavior. Apology was higher for the apol-
ogy prompt (M= 7.14, SD = 2.08) than for the no-apol-
ogy prompt (M= 3.47, SD = 2.47), t(233) = 16.96,
p< 0.001, d= 1.61. Also, overall conciliatory behavior
was higher for the apology prompt (M= 32.76,
SD = 9.19) than for the no-apology prompt (M= 17.83,
SD = 10.19), t(233) = 16.42, p< 0.001, d= 1.49.
Gender differences
We used mixed-model ANOVAs to test for gender
effects; gender did not interact with condition for any
variables (all interaction p’s = 0.09–0.99) and conse-
quently results were collapsed across genders. Further,
there was no main effect of gender on conciliatory
behavior, self-forgiveness, received forgiveness, shame,
guilt, responsibility, or severity, (Fs = 0.01–3.27,
ps = 0.07–0.93, ds = 0.01–0.25). Women did perceive
self-forgiveness as less morally permissible (M= 7.22,
SD = 1.74) than did men (M= 6.65, SD = 1.87),
F(1, 250) = 5.64, p= 0.02, d= 0.34. Women also reported
more recent offenses, (M= 27.96, SD = 24.53) than
4T.P. Carpenter et al.
Downloaded by [Baylor University Libraries], [Ms Jo-Ann Tsang] at 10:56 29 April 2014
Table 1. Descriptive statistics and correlations for responses to the apology and non-apology prompts in Study 1.
Apology No apology
Variable MSD MSD12345678910
1. SF 5.32 1.70 4.89 1.75 −0.05 0.52*** 0.16*−0.47*** −0.43*** −0.17** −0.29*** −0.08 −0.03
2. Concil 32.76 9.19 17.83 10.19 0.11
+
0.00 0.39*** 0.41*** 0.41*** 0.28*** 0.31*** 0.01 0.29***
3. Moral to SF 7.21 1.97 7.08 2.13 0.62*** 0.22*** 0.21** −0.16*−0.35*** −0.03 −0.11
+
0.02 0.06
4. Received F5.61 1.51 4.21 1.98 0.41*** 0.32*** 0.41*** 0.16*0.07 −0.01 0.19** 0.18** 0.19**
5. Remorse 34.42 16.48 28.75 17.61 −0.34*** 0.41*** −0.13*0.02 0.78*** 0.41*** 0.59*** 0.14*0.16*
6. Self-cond 15.35 10.83 12.87 10.74 −0.48*** 0.23*** −0.37*** −0.15*0.72*** 0.38*** 0.45*** 0.02 0.20**
7. Severity 15.96 5.74 15.00 5.84 −0.21*** 0.27*** −0.17** −0.04 0.49*** 0.42*** 0.30*** 0.10 0.08
8. Resp 35.71 10.84 31.82 12.80 −0.07 0.32*** 0.11
+
0.22*** 0.44*** 0.25*** 0.21*** 0.13*0.05
9. Time 30.88 37.14 34.59 37.62 −0.01 0.11
+
0.05 0.12
+
0.18** 0.10 0.15*0.20** −0.19**
10. Closeness 7.26 2.15 5.78 2.74 0.05 0.31*** 0.28*** 0.23*** 0.16*−0.03 0.18** 0.14*−0.09
Note: Apologized offenses are below the diagonal. SF = self-forgiveness; Concil = conciliatory behavior; Moral to SF = perceived morality of self-forgiveness. Received F= received forgiveness; Self-
cond = self-condemnation; Resp = responsibility; Closeness = pre-offense closeness.
***p< 0.001, **p< 0.01, *p< 0.05,
+
p< 0.10.
The Journal of Positive Psychology 5
Downloaded by [Baylor University Libraries], [Ms Jo-Ann Tsang] at 10:56 29 April 2014
did men (M= 42.08, SD = 37.27), F(1, 220) = 11.36,
p= 0.001, d= 0.48 and reported higher pre-offense close-
ness (M= 6.74, SD = 1.69) than did men, (M= 5.85,
SD = 1.97), F(1, 248) = 13.51, p< 0.001, d= 0.52. How-
ever, as noted above, gender did not moderate the effect
of condition on any of these variables. Similarly, gender
was not related to any variable in our mediation model,
p’s = 0.09–0.64, and the mediated effect did not vary
across genders. Consequently, all analyses reported
below were conducted with the sample as a whole, with-
out gender as a covariate.
Differences between apology and no-apology prompts
We first predicted that conciliatory behavior would
enhance self-forgiveness. As predicted, responses to the
apology prompt were significantly higher in self-forgive-
ness (M= 5.32, SD = 1.70) than to the no-apology
prompt (M= 4.89, SD = 1.75), t(233) = 3.17, p= 0.002,
d= 0.25. Because transgressions were not standardized
(i.e. participants recalled separate offenses in each
prompt), we next compared prompts to determine if
something other than conciliatory behavior might explain
differences in self-forgiveness.
Although we predicted that conciliatory behavior
would reduce guilt, participants reported significantly
more remorse (state guilt) when reflecting on responses
to the apology prompt (M= 34.42, SD = 16.48) than
to the no-apology prompt (M= 28.75, SD = 28.75),
t(233) = 4.18, p< 0.001, d= 0.31. Participants also felt
significantly more self-condemnation (state shame) when
responding to the apology prompt (M= 15.35,
SD = 10.83) than to the no-apology prompt (M= 12.87,
SD = 10.74), t(233) = 2.69, p= 0.04, d= 0.20. Although
more self-forgiven, offenses in the apologized-offense
prompt elicited more negative feelings.
One reason participants may have felt worse about
offenses for the apology prompt is that objectively worse
offenses were more likely to elicit apologies. Consistent
with this, participants perceived themselves as signifi-
cantly more responsible in the apology prompt condition
(M= 35.71, SD = 10.84) than the no-apology prompt con-
dition (M= 31.82, SD = 12.80), t(233) = 3.79, p< 0.001,
d= 0.32, and they perceived apologized offenses as
marginally more severe (M= 15.96, SD = 5.74) than non-
apologized offenses (M= 15.00, SD = 5.84), t(232) =
1.70, p= 0.09, d= 0.14. Further item-level analysis
revealed that participants perceived apologized offenses
as significantly more serious, t(232) = 2.15, p= 0.03,
d= 0.17, marginally more harmful, t(233) = 1.65,
p= 0.10, d= 0.13, but not more immoral, t(233) = 0.79,
p= 0.43, d= 0.06. Although responses to the apology
prompt received more self-forgiveness overall, partici-
pants perceived them as worse in other ways.
We also looked for differences in time since the
offense and pre-offense closeness. The two prompts did
not differ in the duration of time since the offense, t
(219) = 1.19, p= 0.234, d= 0.10. In contrast, offenses in
the apology prompt took place in the context of
significantly closer relationships (M= 7.26, SD = 5.78)
than in the no-apology prompt (M= 5.78, SD = 2.74),
t(232) = 6.59, p< 0.001, d= 0.59. However, pre-offense
closeness did not correlate with self-forgiveness in either
the apology or no-apology prompts, rs = 0.05 and −0.03,
ps > 0.40. Neither time since the offense nor pre-offense
closeness appeared to be a tenable explanation for differ-
ences in self-forgiveness.
We theorized that conciliatory behavior would make
self-forgiveness seem more moral. Contrary to hypothe-
ses, self-forgiveness was not seen as significantly more
moral in the apology prompt condition (M= 7.21,
SD = 1.97) than the no-apology prompt condition
(M= 7.08, SD = 2.13), t(233) = 0.93, p= 0.35, d= 0.06,
although the difference was in the predicted direction.
We suspected that a potential morality-boosting effect of
conciliatory behavior may have been suppressed by the
Table 2. Descriptive statistics and correlations for responses to the general-offense prompt in Study 1.
Variable MSD1 2 3 456789
1. SF 5.30 1.73
2. Concil 28.48 11.69 0.06
3. Moral to SF 7.33 1.99 0.57*** 0.15*
4. Received F 5.01 1.83 0.21*** 0.43*** 0.27***
5. Remorse 32.20 16.99 −0.42*** 0.37*** −0.24*** 0.19**
6. Self-cond 14.54 11.36 −0.52*** 0.25*** −0.41*** 0.06 0.76***
7. Severity 16.06 6.29 −0.24*** 0.33*** −0.12
+
0.08 0.45*** 0.41***
8. Resp 34.84 11.29 −0.26*** 0.31*** −0.11
+
0.15*0.57*** 0.40*** 0.30***
9. Time 33.15 38.23 −0.10 −0.04 0.02 0.15*0.17** 0.11
+
0.11
+
0.18**
10. Closeness 6.63 2.45 0.03 0.50*** 0.15*0.28*** 0.23*** 0.14*0.19** 0.09 −0.05
Note: SF = self-forgiveness; Concil = conciliatory behavior; Moral to SF = perceived morality of self-forgiveness; Received F= received forgiveness;
Self-cond = self-condemnation; Resp = responsibility; Closeness = pre-offense closeness.
***p< 0.001, **p< 0.01, *p< 0.05,
+
p< 0.10.
6T.P. Carpenter et al.
Downloaded by [Baylor University Libraries], [Ms Jo-Ann Tsang] at 10:56 29 April 2014
greater levels of negative affect among apologized
offenses. Consistent with this, the difference in perceived
morality between the apology and no-apology prompts
was re-analyzed with a repeated-measures ANCOVA
controlling for remorse and self-condemnation. The dif-
ference in perceived morality between the apology
prompt (adjusted M= 7.28) and the no-apology prompt
(adjusted M= 6.90) became significant, F(1, 252) =
10.20, p= 0.02, ω
2
= 0.02, after controlling for these
differences. Further consistent with this hypothesis,
conciliatory behavior and morality of self-forgiveness
correlated positively within both prompts after self-
condemnation and remorse were controlled, r= 0.29,
p< 0.001 for the apology prompt and r= 0.14, p= 0.03,
for the no-apology prompt. Conciliatory behavior thus
emerged as a predictor of morality of self-forgiveness
when confounding differences in emotions were con-
trolled.
Finally, we predicted that conciliatory behavior
would lead to more received forgiveness. Consistent with
this, participants reported receiving more forgiveness in
the apology prompt condition (M= 5.61, SD = 1.51) than
the no-apology prompt condition (M= 4.21, SD = 1.98), t
(233) = 3.86, p< 0.001, d= 0.82.
Mediational analysis
To test our primary hypothesis that morality of self-for-
giveness explains, in part, why conciliatory behavior is
related to self-forgiveness, we examined responses to
the general offense prompt using a multiple mediation
model using the PROCESS macro for SPSS (Hayes,
2013). This had the added benefit of allowing us to
better control for negative elements of the offenses (e.g.
emotions), which correlated positively with conciliatory
behavior. Time since the offense, pre-offense closeness,
shame, responsibility, and severity were controlled at all
steps. Pre-offense closeness was included because it
covaries with conciliatory behavior and guilt (Exline
et al., 2011).
Coefficients from the mediation analysis are given in
Figure 2. A summary of mediated effects is given in
Table 3. There was a significant total effect of concilia-
tory behavior on self-forgiveness, β= 0.23, p< 0.001.
The direct effect of conciliatory behavior was not signifi-
cant, β= 0.09, p= 0.16, but a significant portion of the
effect of conciliatory behavior was mediated, estimate =
0.14, 95% CI [0.056, 0.243]. As predicted, morality of
self-forgiveness played an important mediating role.
Morality of self-forgiveness uniquely mediated the rela-
tionship between conciliatory behavior and self-forgive-
ness, estimate = 0.07, 95% CI [0.014, 0.154]. In addition,
morality of self-forgiveness functioned as part of a three-
path meditational chain, with conciliatory behavior pre-
dicting received forgiveness, morality of self-forgiveness,
and self-forgiveness in sequence, estimate = 0.03, 95%
CI [0.011, 0.058]. Together, the mediated effects contain-
ing perceived morality accounted for 45% of the rela-
tionship between conciliatory behavior and self-
forgiveness. Morality of self-forgiveness appeared to
play a significant role in explaining why conciliatory
behavior predicted self-forgiveness.
Results for received forgiveness and conciliatory
behavior were mixed. Received forgiveness uniquely
mediated the relationship between conciliatory behavior
and self-forgiveness, estimate = 0.05, 95% CI [0.012,
0.104]. Contrary to predictions, no mediation effects with
guilt were significant.
Table 3. Standardized effects of conciliatory behavior on self-forgiveness in Study 1.
Mediated effect of conciliatory behavior Est. (95% CI)
Concil →Moral →SF 0.07*[0.014, 0.154]
Concil →RF →SF 0.05*[0.012, 0.104]
Concil →Guilt →SF −0.01 [−0.051, 0.003]
Concil →RF →Moral →SF 0.03*[0.011, 0.058]
Concil →RF →Guilt →SF −0.004 [−0.020, 0.001]
Concil →Guilt →Moral →SF 0.001 [−0.004, 0.011]
Concil →RF →Guilt →Moral →SF <0.001 [−0.001, 0.004]
Note: Concil = conciliatory behavior; RF = received forgiveness; Rem = remorse; Moral = perceived morality of self-forgiveness; SF = self-forgiveness.
*p< 0.05.
.05
Conciliatory
Behavior
Received
Forgiveness Guilt Moral to
Self Forgive
Self
Forgiveness
.42*** .08 .16*
.16**
.03
.09
.43***
-.18*
.12*
Figure 2. Standardized mediation path coefficients examining
the relationship between conciliatory behavior and self-forgive-
ness in the general-offense prompt in Study 1.
The Journal of Positive Psychology 7
Downloaded by [Baylor University Libraries], [Ms Jo-Ann Tsang] at 10:56 29 April 2014
Discussion
Study 1 offered support both for the relationship between
conciliatory behavior and self-forgiveness and for the
role of perceived moral appropriateness of self-forgive-
ness in mediating that relationship. Within-subjects anal-
yses revealed that self-forgiveness was higher for
apologized offenses, despite the fact that participants felt
worse about those offenses and saw them as more
severe. These differences suggest that the elevated levels
of self-forgiveness among apologized offenses came in
spite of, rather than because of, pre-existing differences
among offenses. Although in the predicted direction,
morality of self-forgiveness did not differ between the
apology and no-apology prompts, possibly because of
these pre-existing differences. Consistent with this,
reanalysis controlling for emotions revealed significantly
higher levels of perceived morality among apologized
offenses. In addition, conciliatory behavior correlated
positively with perceived morality in all prompts after
controlling for negative emotions.
Mediational analyses within the general-offense
prompt allowed us to further explore these relationships.
As predicted, morality of self-forgiveness directly medi-
ated the relationship between conciliatory behavior and
self-forgiveness, and the total mediated effect through
morality of self-forgiveness accounted for nearly half of
the total relationship between conciliatory behavior and
self-forgiveness and over half of the mediated effect.
Also as predicted, received forgiveness uniquely medi-
ated the relationship between conciliatory behavior and
self-forgiveness, with some of this effect going through
morality of self-forgiveness. Contrary to predictions, no
support was found for any guilt-reduction effects, and
guilt appeared to play no mediating role.
Although Study 1 provided good initial support for
our central hypotheses, one weakness was that offenses
were not standardized. Although we attempted to mea-
sure and control for potential confounds, it is possible
that some other difference may have explained these
effects. Pre-existing differences in the offenses may
have likewise weakened our ability to assess guilt-
reduction effects, evidenced by the positive overall
association between guilt and conciliatory behavior. To
address these limitations, in Study 2 we directly manip-
ulated conciliatory behavior in response to a standard-
ized transgression.
Another limitation was the use of a single-item self-
forgiveness measure. We chose this measure, studied in-
depth by Hall and Fincham (2008), because the existing
state self-forgiveness scale (Wohl et al., 2008) includes
content related to, but outside of, our definition of self-
forgiveness, such as shame and self-rejection (e.g. belief
that the self is bad). This may be advantageous in some
research contexts, but it makes it inappropriate for exam-
ining process variables related to self-forgiveness, such
as emotions. Single item measures are a common solu-
tion to this issue in the emotion literature (e.g. Ellsworth
& Tong, 2006), but they are not without issues (e.g.
questions of breath vs. narrowness and psychometric
concerns). Although the self-forgiveness item we used in
Study 1 is face valid and performed consistently with
theory in both Study 1 and prior research (e.g. Hall &
Fincham, 2008; Witvliet et al., 2002), for psychometric
reasons, we included a three-item assessment in Study 2.
Study 2
Study 2 tested the relationship between conciliatory
behavior, perceived morality of self-forgiveness, and
self-forgiveness, with greater experimental control. Par-
ticipants read a standardized offense vignette and were
randomly assigned to imagine either engaging or not
engaging in conciliatory behavior. The use of a standard-
ized experimental manipulation allowed us to better draw
causal conclusions about the effects of conciliatory
behavior. Given the importance of received forgiveness
in Study 1, we attempted to further investigate its role in
Study 2 by manipulating it alongside conciliatory behav-
ior.
In replication of Study 1, we expected both concilia-
tory behavior and received forgiveness to influence self-
forgiveness. Likewise, morality of self-forgiveness was
expected to mediate both relationships. Given that
offenses were now standardized, we sought to re-
examine whether reduced guilt would mediate either
effect or feed into morality of self-forgiveness, as
predicted in Study 1.
Method
Participants
Participants were 208 psychology undergraduates (143
female and 65 male) at a mid-sized private religious uni-
versity. Data collection was terminated at the end of one
semester. Eight participants took an extreme amount of
time on the experiment (z’s > 3.0) and were flagged as
outliers and excluded. This left a final sample of 200
(136 female and 64 male). Participants had a mean age
of 19.46 years (SD = 1.31) and were Caucasian (62%),
Hispanic/Latino (14%), African-American (12%), and
Asian (12%).
Procedure
Participants completed the study online and were pre-
sented with a single transgression vignette adapted from
Tangney, Boone, Fee, and Reinsmith (1999). The victim
was gender-matched to the participant to keep the
8T.P. Carpenter et al.
Downloaded by [Baylor University Libraries], [Ms Jo-Ann Tsang] at 10:56 29 April 2014
offender–victim dynamic consistent for men and women.
The vignette read:
Imagine that your friend was fired from [his/her] job for
something [he/she] did not do. Even though you actually
committed the act, you did not speak up to take the
blame. However, you both knew that you were responsi-
ble. [He/she] was very upset.
Participants in the conciliatory behavior condition were
then told, ‘Further imagine that you genuinely apolo-
gized, tried to fix things, and tried to make it up to [him/
her].’Participants in the non-conciliatory condition were
told, ‘Further imagine that you have not apologized or
tried to fix things or make it up to [him/her].’
In addition, participants were told one of three things
about the friend’s forgiveness. In the no-forgiveness con-
dition, participants were told, ‘[He/she] has not forgiven
you and appears to hold a grudge against you.’In the
no-information condition, participants were not told spe-
cifically about the friend’s forgiveness. In the received-
forgiveness condition, participants were told, ‘[He/she]
has completely forgiven you for what you’ve done and
no longer holds a grudge against you.’
Measures
Self-forgiveness
Participants answered three items assessing self-forgive-
ness, improving upon the single-item measure from
Study 1: ‘How likely would you be to forgive your-
self?’‘How forgiving do you think you would be
toward yourself?’and ‘How unforgiving do you think
you would be toward yourself?’Each item was
answered on a 7-point scale (1 = not at all,7=extre-
mely). Items loaded on a single factor, and reliability
was satisfactory, α= 0.76.
Morality of self-forgiveness
Participants completed six items assessing perceived
morality of self-forgiveness, improving on the single
item from Study 1. Participants were asked how ‘morally
right,’‘morally appropriate,’‘morally wrong,’‘morally
inappropriate,’‘unjust,’and ‘unfair’it would be to for-
give themselves. Each item was scored on a 7-point
scale (1 = not at all,7=extremely). Items loaded on a
single factor, and reliability was excellent, α= 0.91.
Guilt
Participant guilt was measured with the single item,
‘How guilty would you feel?’The item was scored on a
5-point scale (1 = not at all,5=extremely).
3
Received forgiveness
As a manipulation check, participants completed the
items: ‘To what extent do you think the person you hurt
in the story has forgiven you?’‘How much would you
feel forgiven by the person you hurt in the story?’and
‘How forgiving do you think the person you hurt in the
story would act toward you?’Each item was scored on a
7-point scale (1 = not at all,7=extremely). Reliability
was good, α= 0.89.
Results
The data were analyzed with a 2 (conciliatory behav-
ior) × 3 (received forgiveness) analysis of variance. Guilt
was square-root transformed to improve normality prior
to any analyses. Descriptive statistics and correlations
are given in Table 4. There were some gender
differences, with women reporting less self-forgiveness
(M= 8.99, SD = 3.75) than men (M= 10.20, SD = 3.64),
t(198) = 2.14, p= 0.03, d= 0.30 and expressing more
guilt (M= 4.39, SD = 0.97) than men (M= 4.02, SD =
1.12), t(198) = 2.43, p= 0.02, d= 0.62. However, women
did not report different perceptions of morality of self-
forgiveness (M= 20.51, SD = 8.77) than men (M= 21.70,
SD = 8.67), t(198) = 0.90, p= 0.37, d= 0.14. Importantly,
the effect of condition was not moderated by gender for
any variables (all interaction p’s > 0.36) and thus results
were collapsed across gender.
We first ensured that the received-forgiveness manip-
ulation influenced the degree to which participants
anticipated feeling forgiven. The effect was significant,
F(2, 194) = 26.20, p< 0.001, ω
2
= 0.197. Post-hoc
Table 4. Means and standard deviations, and correlations for Study 2.
Conciliatory Behavior
No Conciliatory
Behavior
Variable MSD MSD 1 2
1. SF 10.81 3.73 7.83 3.12
2. Moral to SF 23.10 8.37 18.50 8.53 0.45***
3. Guilt 3.94 1.15 4.35 0.92 −0.49*** −0.37***
Note: SF = self-forgiveness; Moral to SF = moral to self-forgiveness.
***p< 0.001.
The Journal of Positive Psychology 9
Downloaded by [Baylor University Libraries], [Ms Jo-Ann Tsang] at 10:56 29 April 2014
analysis using the Dunn-Šidák procedure (Kirk, 2013)
revealed that participants felt less forgiven in the
no-forgiveness condition (M= 7.63, SD = 4.29) than the
no-information condition (M= 9.21, SD = 3.99), t
DS
(194,
2) = −2.30, p= 0.04, ω
2
= 0.020, d= 0.38. Similarly, par-
ticipants felt more forgiven in the received-forgiveness
condition (M= 12.42, SD = 3.69) than the no-information
condition, t
DS
(194, 2) = 4.74, p< 0.001, ω
2
= 0.081,
d= 0.83. The trend among conditions was linear, F(1,
197) = 48.46, p< 0.001, ω
2
= 0.19, with no higher-order
components, F(1, 197) = 1.83, p= 0.18, ω
2
< 0.001,
indicating that the manipulation yielded three linearly
increasing levels of received forgiveness.
As predicted, there was a significant self-forgiveness
enhancing effect of conciliatory behavior. Participants
self-forgave more in the conciliatory condition
(M= 10.81, SD = 3.73) than the non-conciliatory condi-
tion (M= 7.83, SD = 3.12), F(1, 194) = 36.62, p< 0.001,
ω
2
= 0.15, d= 0.86. Participants also saw self-forgiveness
as more moral in the conciliatory condition (M= 23.10,
SD = 8.37) than the non-conciliatory condition
(M= 18.50, SD = 8.53), F(1, 194) = 14.75, p< 0.001,
ω
2
= 0.067, d= 0.27. Finally, participants felt less guilt in
the conciliatory condition (M= 4.05, SD = 1.12) than
the non-conciliatory condition (M= 4.50, SD = 0.87),
F(1, 194) = 11.07, p= 0.001, ω
2
= 0.05, d= 0.47. Concil-
iatory behavior thus had all the predicted effects.
Contrary to predictions, there were no main effects of
received forgiveness on self-forgiveness, F(2, 194) = 0.41,
p= 0.66, ω
2
< 0.001, perceived morality of self-forgive-
ness, F(2, 194) = 0.47, p= 0.63, ω
2
< 0.001, or guilt,
F(2, 194) = 0.19, p= 0.83, ω
2
< 0.001. There were no
interactions between treatments (all p’s > 0.15).
Mediational analysis
Mediation was examined using the PROCESS macro for
SPSS. As in Study 1, bias-corrected and accelerated
bootstrapped confidence intervals were estimated with
10,000 resamples. Conciliatory behavior was dummy
coded (−0.5 = no conciliatory behavior, 0.5 = conciliatory
behavior). Given the linear relationship among the
received-forgiveness conditions, received forgiveness
was dummy coded (−1 = no forgiveness, 0 = no specifica-
tion, 1 = forgiveness). All other variables were standard-
ized to facilitate interpretation. Thus, mediation estimates
are scaled analogously to Cohen’sd. Because gender did
not moderate any experimental effect (see above), media-
tion models were collapsed across gender.
A summary of mediated effects is given in Table 5.
Following Study 1, guilt and morality of self-forgiveness
(both uniquely and in-sequence) were examined as medi-
ators of the effect of conciliatory behavior on self-for-
giveness (see Figure 3). The total effect was significant,
β= 0.79, p< 0.001. As expected, morality of self-forgive-
ness uniquely mediated the effect of conciliatory behav-
ior on self-forgiveness, estimate = 0.13, 95% CI [0.045,
0.241]. As predicted, guilt reduction also mediated the
effect of conciliatory behavior on self-forgiveness, esti-
mate = 0.12, 95% CI [0.040, 0.245]. In addition, concil-
iatory behavior predicted guilt, which in turn predicted
morality of self-forgiveness and self-forgiveness in
sequence. This sequential mediation was significant, esti-
mate = 0.03, 95% CI [0.008, 0.071]. Together, effects
containing morality of self-forgiveness accounted for
55% of the mediated effect and appeared to explain, in
part, why conciliatory behavior led to self-forgiveness.
In addition, guilt reduction played an important
Table 5. Mediated effects on self-forgiveness in Study 2.
Mediated effect of conciliatory behavior Est. (95% CI)
Total mediated effect: 0.28*** [0.140, 0.445]
Concil →Moral →SF 0.13** [0.045, 0.241]
Concil →Guilt →SF 0.12** [0.040, 0.245]
Concil →Guilt →Moral →SF 0.03** [0.008, 0.071]
Mediated effect of received forgiveness
Total mediated effect: 0.02 [−0.073, 0.097]
RF →Moral →SF 0.03 [−0.018, 0.081]
RF →Guilt →SF −0.01 [−0.070, 0.049]
RF →Guilt →Moral →SF 0.03 [−0.020, 0.081]
Note: Because conditions were dummy coded and DVs and mediators were standardized, coefficients are in units of Cohen’sd. Concil = conciliatory
behavior; RF = received forgiveness; Moral = perceived morality of self-forgiveness; SF = self-forgiveness.
**p< 0.01, ***p< 0.001.
Conciliatory
Behavior
Guilt Moral to
Self Forgive
Self
Forgiveness
-.25***
.44***
-.39**
.52***
.29***
-.32***
Figure 3. Mediation paths coefficients depicting the effect of
conciliatory behavior in Study 2. Conciliatory behavior was
dummy coded; other variables were standardized.
10 T.P. Carpenter et al.
Downloaded by [Baylor University Libraries], [Ms Jo-Ann Tsang] at 10:56 29 April 2014
mediating role, both in conjunction with morality of self-
forgiveness and on its own. However, unlike Study 1, no
effects of conciliatory behavior was mediated.
Discussion
In replication of Study 1, conciliatory behavior caused
increases in both self-forgiveness and morality of self-
forgiveness. Also as in Study 1, morality of self-forgive-
ness mediated the effect of conciliatory behavior on self-
forgiveness.
Unlike Study 1, which used non-standardized
offenses, conciliatory behavior reduced guilty feelings.
This guilt-reduction effect is consistent with past
research (Meek et al., 1995) and explained some of the
effect of conciliatory behavior on self-forgiveness. As
predicted, this effect partly explained some of the effect
of conciliatory behavior on perceived morality, and in
turn, self-forgiveness.
Also in contrast to Study 1, received forgiveness did
not have any influence on self-forgiveness or perceived
morality of self-forgiveness. Although it is possible that
there is no causal connection between received forgive-
ness and self-forgiveness, it is also possible that the
manipulation failed because it merely informed partici-
pants they were forgiven. It did not model expressions of
compassion or relational repair, which may be important
for self-forgiveness (Worthington, 2006). Past research
has also found that merely informing people that their
transgressions are forgiven can increase perceived ineq-
uity (Kelln & Ellard, 1999). Such effects may have can-
celed out any forgiveness-enhancing effect. Consistent
with this, supplementary analyses using the received-for-
giveness manipulation check as a mediator revealed a
positive indirect effect of received forgiveness on self-for-
giveness, estimate = 0.28, 95% CI [0.188, 0.391] and a
negative direct effect, β=−0.23, p= 0.004, which can-
celed out to yield the null main effect. In other words, the
received-forgiveness manipulation appeared to boost self-
forgiveness by making participants feel more forgiven,
yet hinder self-forgiveness directly. Similar results were
observed with perceived morality and guilt as dependent
variables. Although unpredicted, these exploratory results
suggest a possible explanation for the null effect of the
manipulation and the discrepancy between findings in
Studies 1 and 2. Because received forgiveness was less
central to the present investigation, we considered these
findings interesting but did not explore them further.
General discussion
Across two studies, we found support for the role of per-
ceived morality of self-forgiveness in explaining why
conciliatory behavior may increase self-forgiveness. In
Study 1, when negative emotions were controlled, posi-
tive associations emerged in all Study 1 prompts
between conciliatory behavior and morality of self-
forgiveness, and differences emerged between the
apologized and non-apologized offense prompts. A stan-
dardized experimental manipulation of conciliatory
behavior in Study 2 further demonstrated the causal
effect of conciliatory behavior on morality of self-for-
giveness. As predicted, morality of self-forgiveness
served as a mediator in both studies, explaining some of
the relationship between conciliatory behavior and self-
forgiveness. In both studies, perceived morality uniquely
mediated the effect of conciliatory behavior on self-for-
giveness. In addition, morality functioned as part of
more complex indirect effects. In Study 1, conciliatory
behavior predicted received forgiveness, which, in turn,
predicted perceived morality and then self-forgiveness.
In Study 2, conciliatory behavior caused decreased guilt,
which predicted perceived morality and self-forgiveness
in sequence. This supports our central hypothesis that
perceived morality of self-forgiveness partially explains
why conciliatory behavior aids self-forgiveness.
The present research examined perceived morality of
self-forgiveness alongside two other proposed mecha-
nisms underlying conciliatory behavior: received forgive-
ness and guilt reduction. In support of Worthington
(2006) and Hall and Fincham (2005), received forgive-
ness played a mediating role on its own in Study 1 and,
as predicted, fed into morality of self-forgiveness. In
contrast, a manipulation of received forgiveness in Study
2 did not increase self-forgiveness. Exploratory analyses
suggested that the manipulation had both forgiveness-
promoting and forgiveness-inhibiting effects that can-
celed out. However, it is also possible that there is no
causal effect of received forgiveness on self-forgiveness
or that such effects are contingent upon how forgiveness
is expressed. Further research is needed to more defini-
tively specify when and how expressions of forgiveness
may enhance self-forgiveness.
We also found inconsistent evidence for Hall and
Fincham’s(2005) hypothesis that conciliatory behavior
aids self-forgiveness by reducing guilt, with effects
observed in Study 2 but not in Study 1. This is not sur-
prising, given that guilt correlated positively with concil-
iatory behavior in Study 1 and that conciliatory behavior
is a response to guilt (Cryder et al., 2012). Because the
guilt-reduction effects were observed with a standardized
offense in Study 2, we suspect that Study 1 may have
failed to detect underlying guilt-reduction effects.
This research benefited from a combination of
recalled-offense and vignette approaches. Participants in
Study 1 recalled more serious, real-world transgressions,
giving results external validity. However, the offenses in
Study 1 were not standardized, introducing the possibil-
ity that some unmeasured variables may have caused the
observed effects. This weakness was addressed in Study
The Journal of Positive Psychology 11
Downloaded by [Baylor University Libraries], [Ms Jo-Ann Tsang] at 10:56 29 April 2014
2, which used a controlled vignette approach. Although
the vignette had less psychological realism than an actual
offense, the replication of our primary predicted effects
lends further support for our conclusions.
One limitation of the present research is that both
studies were based on self-report measures. Future
research could attempt to extend this work by assessing
behavioral, physiological, and/or implicit indicators of
self-forgiveness for recalled or induced transgressions.
Another limitation is that both samples were drawn from
college populations, which may not be representative of
the population at large. A community sample, for exam-
ple, might have recalled more diverse transgressions than
were represented here, increasing generalizability. Simi-
larly, the present sample was drawn from a religiously
affiliated university. Although religion was not a focus
of the present investigation, it is possible that religious
beliefs could play a role in the moral appraisals of self-
forgiveness. Future research could deliberately explore
the role of religion in moral attitudes toward self-forgive-
ness or seek to compare and contrast moral attitudes
toward self-forgiveness among different populations.
The present research utilized a measured-mediator
cross-sectional approach. We found that imagining con-
ciliatory behavior increased both self-forgiveness and
perceived morality of self-forgiveness. However, it is
possible that self-forgiveness increased perceived moral-
ity (e.g. justification) rather than the other way around.
Similarly, although we found that guilt was associated
with decreased moral perceptions in Study 2, it is possi-
ble that the causal direction among mediators was
reversed (e.g. the sense that it would be morally wrong
to self-forgive could amplify moral emotions). Although
our interpretation is consistent with theory, future
research could attempt to explore such alternate possibili-
ties by directly manipulating the perceived morality of
self-forgiveness and/or moral emotions surrounding an
unforgiven offense. Finally, the present findings examine
state-level variables; future research could build on these
findings by examining whether individual differences in
moral attitudes or justice motives yield similar results.
One additional direction for research could be to
explore different kinds of conciliatory behavior. We
focused on overall conciliatory responses here, following
theoretical and empirical work by Exline et al. (2011)
that suggests that conciliatory behavior inverts the offen-
der’s relationship to the offense from one of transgressor
to one of repair. This was further justified by our analy-
sis of conciliatory behaviors in Study 1, which revealed
that conciliatory behaviors were highly homogenous and
tended to load on a single factor. However, it is also pos-
sible that different conciliatory behaviors function differ-
ently; future research could explore the different effects
of behaviors such as apology, restitution, and confession
on self-forgiveness.
The present research has implications for self-forgive-
ness theory. To date, there has been little discussion over
motivations behind self-unforgiveness and those factors
that prevent self-forgiveness (for an exception, see Fisher
& Exline, 2010). The present research builds on theoreti-
cal work by Exline et al. (2011) and qualitative work by
Ingersoll-Dayton and Krause (2005) and suggests that
one important factor in self-unforgiveness may be a moral
sense that it would be unjust to release oneself from un-
forgiveness. It may often be the case that individuals wish
to self-forgive yet find themselves resistant to it –at least
until they have first done something to make things right.
Such moral prerequisites for self-forgiveness may be
important aspects of the recovery process.
Self-forgiveness may reside within morally ambigu-
ous territory, and individuals may, at times, believe that
they deserve to continue to pay for their wrongs. How-
ever, by making amends individuals may find themselves
able to tip the scales of justice and give themselves
moral permission to self-forgive.
Funding
This research was supported in part by a generous grant from
the Fetzer Institute.
Notes
1. Due to an error, age and ethnicity data for these partici-
pants were not recorded.
2. Although a scale assessing self-forgiveness does exist
(Wohl, DeShea, & Wahkinney, 2008), it was not used here,
as items go beyond self-forgiveness as defined by Hall and
Fincham (2005) and include related antecedent constructs
(e.g. shame, self-rejection, and bad self-attributions). We
sought to separately measure these constructs and model
their relationships to self-forgiveness, and thus a narrower
measure of self-forgiveness was used.
3. We chose a single-item measure of guilt over a guilt scale
(e.g. as in Study 1) because guilt scales ask about the sub-
jective experience of emotion, which participants would
not experience during an imagination task.
References
Adams, J. S. (1965). Inequity in social exchange. In L. Berkowitz
(Ed.), Advances in experimental social psychology (Vol. 2,
pp. 267–299). New York, NY: Academic Press.
Cafaro, A. M., & Exline, J. J. (2003, March). Correlates of
self-forgiveness after offending God. Poster presented at the
first annual Mid–winter Research Conference on Religion
and Spirituality, Timonium, MD.
Cryder, C. E., Springer, S., & Morewedge, C. K. (2012). Guilty
feelings, targeted actions. Personality and Social Psychol-
ogy Bulletin, 38, 607–618. doi:10.1177/0146167211435796
Ellsworth, P. C., & Tong, E. M. W. (2006). What does it mean
to be angry at yourself? Categories, appraisals, and the
problem of language. Emotion, 6, 572–586. doi:10.1037/
1528-3542.6.4.572
12 T.P. Carpenter et al.
Downloaded by [Baylor University Libraries], [Ms Jo-Ann Tsang] at 10:56 29 April 2014
Exline, J. J., Root, B. L., Yadavalli, S., Martin, A. M., &
Fisher, M. L. (2011). Reparative behaviors and self-forgive-
ness: Effects of a laboratory-based exercise. Self and Iden-
tity, 10, 101–126. doi:10.1080/15298861003669565
Fehr, R., Gelfand, M. J., & Nag, M. (2010). The road to for-
giveness: A meta-analytic synthesis of its situational and
dispositional correlates. Psychological Bulletin, 136, 894–
914. doi:10.1037/a0019993
Fisher, M. L., & Exline, J. J. (2006). Self-forgiveness versus
excusing: The roles of remorse, effort, and acceptance of
responsibility. Self and Identity, 5, 127–146. doi:10.1080/
15298860600586123
Fisher, M. L., & Exline, J. J. (2010). Moving toward self-for-
giveness: Removing barriers related to shame, guilt, and
regret. Social and Personality Psychology Compass, 4,
548–558. doi:10.1111/j.1751-9004.2010.00276.x
Hall, J. H., & Fincham, F. D. (2005). Self-forgiveness: The
stepchild of forgiveness research. Journal of Social and
Clinical Psychology, 24, 621–637. doi:10.1521/jscp.2005.
24.5.621
Hall, J. H., & Fincham, F. D. (2008). The temporal course of
self-forgiveness. Journal of Social and Clinical Psychology,
27, 174–202. doi:10.1521/jscp.2008.27.2.174
Hayes, A. F. (2013). An introduction to mediation, moderation,
and conditional process analysis: A regression-based
approach. New York, NY: Guilford Press.
Holmgren, M. R. (2002). Forgiveness and self-forgiveness in
psychotherapy. In S. Lamb & J. G. Murphy (Eds.), Before
forgiving: Cautionary views of forgiveness in psychotherapy
(pp. 112–135). New York, NY: Oxford University Press.
Horsbrugh, H. J. N. (1974). Forgiveness. Canadian Journal of
Philosophy, 4, 269–282.
Ingersoll-Dayton, B., & Krause, N. (2005). Self-forgiveness: A
component of mental health in later life. Research on
Aging, 27, 267–289. doi:10.1177/0164027504274122
Kelln, B. R. C., & Ellard, J. H. (1999). An equity theory analy-
sis of the impact of forgiveness and retribution on trans-
gressor compliance. Personality and Social Psychology
Bulletin, 25, 864–872. doi:10.1177/0146167299025007008
Ketelaar, T., & Au, W. T. (2003). The effects of feelings of
guilt on the behaviour of uncooperative individuals in
repeated social bargaining games: An affect-as-information
interpretation of the role of emotion in social interaction.
Cognition and Emotion, 17, 429–453. doi:10.1080/
02699930143000662
Kirk, R. E. (2013). Experimental design: Procedures for the
behavioral sciences (4th ed.). Los Angeles, CA: Sage.
Lamb, S. (2002). Reasons to be cautious about the use of for-
giveness in psychotherapy. In S. Lamb & J. G. Murphy
(Eds.), Before forgiving: Cautionary views of forgiveness in
psychotherapy (pp. 3–14). Oxford: Oxford University
Press.
Meek, K. R., Albright, J. S., & McMinn, M. R. (1995). Reli-
gious orientation, guilt, confession, and forgiveness. Jour-
nal of Psychology and Theology, 23, 190–197.
Murphy, J. G. (2002). Forgiveness in counseling: A philosophi-
cal perspective. In S. Lamb & J. G. Murphy (Eds.), Before
forgiving: Cautionary views of forgiveness in psychother-
apy (pp. 41–53). Oxford: Oxford University Press.
Rangganadhan, A. R., & Todorov, N. (2010). Personality and
self-forgiveness: The roles of shame, guilt, empathy and
conciliatory behavior. Journal of Social and Clinical Psy-
chology, 29,1–22. doi:10.1521/jscp.2010.29.1.1
Tangney, J. P., Boone, A. L., Fee, R. L., & Reinsmith, C.
(1999). Multidimensional forgiveness scale. Fairfax, VA:
George Mason University.
Vitz, P. C., & Meade, J. M. (2011). Self-forgiveness in psychol-
ogy and psychotherapy: A critique. Journal of Religion
and Health, 50, 248–263. doi:10.1007/s10943-010-9343-x
Witvliet, C. V. O., Ludwig, T. E., & Bauer, D. J. (2002). Please
forgive me: Transgressors’emotions and physiology during
imagery of seeking forgiveness and victim responses. Jour-
nal of Psychology and Christianity, 21, 219–233.
Wohl, M. J. A., DeShea, L., & Wahkinney, R. L. (2008). Look-
ing within: Measuring state self-forgiveness and its rela-
tionship to psychological well-being. Canadian Journal of
Behavioural Science/Revue canadienne des sciences du
comportement, 40,1–10. doi:10.1037/0008-400x.40.1.1.1
Worthington, E. L., Jr. (2006). Forgiveness and reconciliation:
Theory and application. New York, NY: Brunner/Routl-
edge.
Zechmeister, J. S., & Romero, C. (2002). Victim and offender
accounts of interpersonal conflict: Autobiographical narra-
tives of forgiveness and unforgiveness. Journal of Person-
ality and Social Psychology, 82, 675–686. doi:10.1037/
0022-3514.82.4.675
The Journal of Positive Psychology 13
Downloaded by [Baylor University Libraries], [Ms Jo-Ann Tsang] at 10:56 29 April 2014
