Content uploaded by Gary W Lewandowski Jr.
Author content
All content in this area was uploaded by Gary W Lewandowski Jr.
Content may be subject to copyright.
The Journal of Positive Psychology, January 2007; 2(1): 40–54
Addition through subtraction: Growth following the dissolution of a
low quality relationship
GARY W. LEWANDOWSKI & NICOLE M. BIZZOCO
Monmouth University, NJ, USA
Abstract
Previous research has primarily focused on negative outcomes following relationship dissolution. The purpose of this study is
to add to the growing literature that shows the positive outcomes that are possible. It was hypothesized that growth and
positive emotions would be associated with self-related variables such as self-expansion and rediscovery of self, as well as
coping strategies. Participants were 155 undergraduates who had experienced the dissolution of a relationship in the past
6 months. Results confirmed the hypotheses that growth would be related to ending a relationship low in self-expansion and
that the relationship between these variables would be mediated by experiencing more rediscovery of the self, less loss of self,
and more positive emotions following dissolution.
Keywords: Growth, self, coping, relationship termination, relationship quality
Introduction
Relationship dissolution is a common occurrence that
is virtually unavoidable. Up to this point, research in
this area has focused primarily on experiences of
distress following dissolution (e.g., Frazier & Cook,
1993; Simpson, 1987; Sprecher, Felmlee, Metts,
Fehr, & Vanni, 1998). Unfortunately, this may
provide a narrow perspective of the post-dissolution
experience. In fact, negative outcomes may represent
only half of the picture, leaving positive outcomes
as an under-explored facet of post-dissolution
experiences. This is especially likely due to recent
research showing that positive emotional experiences
can be independent of negative emotional experiences
(see Feldman-Barrett & Russell, 1999, for a review).
The possibility that relationship dissolution can
produce positive outcomes has been addressed
empirically (e.g., Sprecher et al., 1998; Tashiro &
Frazier, 2003). The present research builds on
this work by examining how the experience of
post-dissolution growth relates to the quality of the
pre-dissolution relationship, potential mediators
of this experience, and on enhancing the prediction
of post-dissolution growth by accounting for several
potential predictors.
Negative outcomes of dissolution
The majority of research on outcomes following
dissolution has focused on negative aspects of the
experience. Typically this has been in the context of
determining relationship characteristics that result in
the greatest post-dissolution distress (e.g., Choo,
Levine, & Hatfield, 1996; Fine & Sacher, 1997;
Frazier & Cook, 1993; Simpson, 1987; Sprecher
et al., 1998), while others have focused on the
negative impact dissolution has on attachment
(Ruvolo, Fabin, & Ruvolo, 2001). Perhaps because
two of the most stressful events a person can
experience (death of a spouse and marital separation)
involve relationship dissolution (Holmes & Rahe,
1967), few studies have examined the possibility that
relationship dissolution may also produce positive
outcomes.
Growth following dissolution
Positive outcomes following dissolution seem likely
in light of research dealing with outcomes following
traumatic events. Indeed, crisis theory suggests
that negative events have the potential to promote
personal growth (Caplan, 1964). Several studies have
Correspondence: Gary W. Lewandowski, Jr. Department of Psychology, Monmouth University, West Long Branch,
NJ 07764, USA. Fax: 732 263 5159. E-mail: glewando@monmouth.edu
ISSN 1743-9760 print/ISSN 1743-9779 online/07/010040–15 ß2007 Taylor & Francis
DOI: 10.1080/17439760601069234
discovered an association between traumatic events
and positive outcomes in general (e.g., Calhoun,
Cann, Tedeschi, & McMillan, 2000; McMillen
& Fisher, 1998), as well as in the context of
bereavement (e.g., Frazier, Steward, & Mortensen,
2004; Park & Cohen, 1993).
Growth is defined as a cognitive process in which
a person actively strives to improve the self through
the discovery of new knowledge and perspectives
(Tedeschi, Park, & Calhoun, 1998). Other com-
monly reported examples of posttraumatic growth
include enhanced social resources, enhanced
personal resources, and improved coping skills
(Schaefer & Moos, 1992). In a set of three studies
examining growth following the most stressful event
they had encountered in the past year, the dissolution
of a romantic relationship was the most common
traumatic event cited by the sample (Park, Cohen,
& Murch, 1996). Results show that growth can
follow negative events, and that growth tends to
manifest itself in the form of life lessons learned
from the negative experience. Findings from
the third study indicate that growth correlates
highly with utilizing coping strategies (COPE;
Carver, Scheier, & Weintraub, 1989) that involve
positive reinterpretation and acceptance (Park et al.,
1996). This suggests that the coping strategies
implemented after a negative life event can influence
the experience of positive outcomes. Additionally,
findings across each of these studies indicate that
growth was positively associated with positive
emotions.
More recent research has furthered the notion
that relationship dissolution can produce positive
outcomes (Tashiro & Frazier, 2003). This study
consisted of 92 college students who completed
a questionnaire containing the Posttraumatic Growth
Inventory (PTGI; Tedeschi & Calhoun, 1996),
an open-ended growth question, as well as potential
correlates of growth such as the Big Five personality
characteristics and attributions. In response to the
open-ended growth item, participants reported an
average of five positive changes post-dissolution,
indicating that growth may be a more common
experience than originally thought (Tashiro &
Frazier, 2003). Positive changes were placed into
several categories, with positive changes in the self
being reported most frequently. This implies that
post-dissolution growth involves changes to one’s
characteristics, traits, and perspectives. Taken
together, these findings, along with Park et al.
(1996), indicate that relationship dissolution can
yield positive outcomes and suggest some potential
correlates of the experience.
In these studies, growth following relationship
dissolution is consistent with a crisis-growth
pathway (Tashiro, Frazier, & Berman, 2006).
The crisis-growth model suggests that dissolution is
an upsetting and potentially devastating experience
that can have a substantial impact on a person.
However, dissolution also provides the opportunity
to meet challenges, discover new abilities, and
ultimately grow as a person. Individuals may also
experience growth post-dissolution through the
stress-relief pathway. This model suggests that
dissolution may not be upsetting because a person
is ending a low quality relationship that was
undesirable or damaging. As a result, a person may
experience growth through the subtraction of the
previously limiting relationship.
The modest amount of research that supports
the stress-relief pathway focuses predominantly
on the experience of relief following dissolution.
For example, Sprecher (1994) found that love and
relief were the two most commonly mentioned
positive emotions post-dissolution, and that the
reported means for both exceeded the scale’s
midpoint. In another study, a factor analysis
of emotional reactions following dissolution revealed
a factor comprised of joy and relief (Choo et al.,
1996). The experience of relief suggests that the
respondent’s present state compares favorably with
their previous relationship. However, it is not clear
whether the experience of a positive emotion such as
relief relates to the experience of growth. It would be
useful to determine more clearly if positive emotions
were related to the experience of growth post-
dissolution, as well as other determinants of the
growth experience.
Role of the self in post-dissolution growth
If growth is possible, identifying the key components
of the experience would constitute an important
first step toward promoting positive post-dissolution
experiences. Tashiro and Frazier’s (2003) findings
that growth involves the improvement of character-
istics, traits, and perspectives, suggest the self-
concept may be influenced by dissolution. Due to
its prominent role in everyday experience it would
seem likely that aspects of the self would be
influenced (Leary & Tangney, 2003).
Within romantic relationships, Aron and Aron’s
(1997) self-expansion model has been particularly
useful in understanding the role of the self.
According to the self-expansion model, people
are motivated to enhance and grow the self in areas
such as knowledge, identities, capabilities, and
experiences. Within this model, relationships are
the primary means of enhancing aspects of the
self, and experience that is positively associated
with relationship quality (Aron, Aron, & Norman,
2001).
Addition through subtraction 41
According to the stress-relief pathway of growth
following dissolution, ending a low quality relation-
ship is one way to experience growth following
dissolution (Tashiro et al., 2006). The association
between self-expansion and relationship quality
suggests that relationships with limited opportunities
for self-expansion would serve as a barrier to the
growth motivation (Aron et al., 2001). That is,
less expanding relationships, commonly referred to
as relationships that are ‘‘in a rut’’ or ‘‘stagnant,’’
prevent optimal expansion from occurring and result
in a lower quality relationship. Consequently,
relationships that offer low self-expansion would
be particularly unsatisfying, and could potentially
provide a sense of relief when ended. Consistent
with the stress-relief pathway, ending this type of
relationship would be less stressful due to new
opportunities that were not available within the
former relationship.
Potential mediators of the growth experience
In the context of ending a less expanding relation-
ship, growth seems especially likely to result from
avoiding a loss of self, becoming reacquainted with
aspects of the self that were neglected in the
relationship, experiencing positive emotions, and
from using coping strategies aimed at reinterpreting
and accepting the dissolution.
Loss of self. A relationship’s pre-dissolution level of
self-expansion should be related to post-dissolution
changes in the self. Specifically, if one ends a
high quality expanding relationship, a perceived
loss of self following dissolution is likely, and
may ultimately influence growth. One study found
that the degree of self-loss following dissolution
was positively associated with feelings of dysphoria
(Drew, Heesacker, Frost, & Oelke, 2004). Similarly,
a set of three studies has shown that greater loss of
self following dissolution results from ending a high-
quality, self-expanding relationship (Lewandowski,
Aron, Bassis, & Kunak, 2006). While this study did
not directly address the possibility of gains in the self
or growth, the correlational nature of the results
suggests that ending a relationship low in self-
expansion could encourage additions to the self,
and potentially growth. Loss of self seems antithetical
to the experience of growth such that the less loss of
self an individual experiences following dissolution,
the more growth would be possible.
Rediscovery of the self. If loss of self could potentially
inhibit growth, it is likely that rediscovery of the self
may promote growth. In this context, rediscovery of
the self involves focusing on parts of one’s identity
that may have been underdeveloped while in the
former relationship. Participants’ responses to open
ended questions about growth in the Tashiro
and Frazier (2003) study demonstrate experiences
consistent with rediscovery of the self. Responses
such as ‘‘I am more self-confident’’ (p. 120) suggest
self-enhancement, while responses such as ‘‘I rely on
my friends more. I forgot how important friends are
when I was with him’’ (p. 120) suggest rediscovering
aspects of the self that may have been neglected
during the relationship. These statements suggest
that rediscovering the self may be a key component
of the growth experience. Further, rediscovery of
the self seems especially likely following the
dissolution of a relationship that inhibits the self
(i.e., a relationship that did not provide high levels of
self-expansion).
Positive emotions. Consistent with a stress-relief
pathway to growth, ending a low quality relationship
should result in more positive emotions because it is
a less stressful experience overall (Tashiro et al.,
2006). Past research has found that relationship
dissolution produces feelings of relief (Choo et al.,
1996; Sprecher, 1994), and an association between
positive emotions and growth (Park et al., 1996).
However, in each case, this research did not
determine whether positive emotions were the
result of ending a low quality relationship. It is
plausible that ending a low quality relationship would
be associated with more positive emotions, and
that these emotions could then serve to energize
the individual and facilitate growth (e.g., Park et al.,
1996). In contrast, because the dissolution of a low
quality relationship produces a limited amount of
negative emotions, they would not be expected to
play a role in growth.
Coping strategies. Coping strategies are another
potential mediator of the relationship between
ending a low quality relationship and personal
growth. Previous research has found that coping
strategies involving positive reinterpretation and
acceptance were positively correlated with growth
following traumatic events (Park et al., 1996;
Schaefer & Moos, 1992). However, past research
has not determined if these strategies are particularly
useful following the end of a low quality relationship.
Since the dissolution of a low quality relationship is
likely to produce relief it may be easier to use
reinterpretation and acceptance coping strategies
(Tashiro et al., 2006). In contrast, coping strategies
such as social support, venting, denial, and mental
disengagement seem much more likely following the
dissolution of a high quality relationship, and likely
42 G. W. Lewandowski & N. M. Bizzoco
have a lesser role following the dissolution of a low
quality relationship.
The present study
A great deal of past research has focused on the
negative consequences of relationship dissolution.
However, the potential for positive outcomes
following dissolution seems likely in light of research
on traumatic events. Some research has begun to
study positive outcomes as well (e.g., Sprecher,
1994; Sprecher et al., 1998), but we could
only find one study to date that has specifically
focused on positive outcomes (growth) after
relationship dissolution (Tashiro & Frazier, 2003).
In fact, Tashiro et al. (2006) note that there are
no studies to date that have focused on the stress-
relief pathway in the context of premarital
dissolution. Thus, the present study would be the
first of its kind.
Using a methodology similar to Sprecher et al.
(1998), the present study seeks to extend previous
work by determining the extent to which partici-
pants reported growth following dissolution, as well
as investigating potential components of the experi-
ence. Past research has implied a link between
growth and variables such as self-expansion
(Lewandowski et al., 2006), loss of self (Drew
et al., 2004; Lewandowski et al., 2006), rediscovery
of the self (Tashiro & Frazier, 2003), and coping
strategies (Choo et al., 1996; Park et al., 1996).
Specifically, this study extends this previous work
by focusing on the extent to which pre-dissolution
relationship self-expansion, post-dissolution loss of
self, rediscovery of self, and coping strategies relate
to growth.
Pre-dissolution self-expansion
Relationships that provide insufficient self-expansion
are likely to be unfulfilling, and may inhibit a
person’s opportunities for personal growth. The
general prediction of the stress-relief pathway
to growth is that ending a low quality relation-
ship results in new opportunities for growth
(Tashiro et al., 2006). In the present context,
we hypothesize that those who report less self-
expansion pre-dissolution will report more growth
post-dissolution. In addition, we hypothesized
that less pre-dissolution self-expansion would be
associated with more rediscovery of the self, less loss
of self, more positive emotions, and more
coping involving reinterpretation and acceptance
post-dissolution.
Mediation of the self-expansion growth relationship
In addition to predicting the general stress-relief
pathway in which less pre-dissolution self-expansion
would be associated with greater post-dissolution
growth, we also suggest several mediators of the path.
Rediscovery of the self. Self-expansion in a relationship
should serve to enhance the self-concept, and should
be associated with greater relationship quality
(Aron et al., 2001). Thus, a relationship that
provides less self-expansion could potentially serve
as a barrier to one’s desired growth. If this were
the case, ending a less expanding relationship
should be followed by a period of reclaiming under-
developed aspects of the self-concept, or rediscovery
of the self. Through this process, a greater amount
of post-dissolution growth could be expected.
We hypothesize that the association between pre-
dissolution self-expansion and post-dissolution
growth will be mediated by rediscovery of the self.
Loss of self. Following this logic, we expect that
ending a low quality relationship would be less likely
to produce negative outcomes such as feeling a loss
of self. This prediction is consistent with past
research that has established the link between loss
of self and negative experiences (Drew et al., 2004;
Lewandowski et al., 2006). Due to the cognitive
nature of growth (Tedeschi et al., 1998), we expect
that any influence of dissolution on the self
(even negative ones such as loss) to be related to
growth. Thus, to the extent that loss of self is
minimized, we expect the experience of growth to be
enhanced. We hypothesize that the association
between pre-dissolution self-expansion and post-
dissolution growth will be mediated by loss of self.
Positive emotions. When a low quality relationship
ends, a sense of relief can result (Tashiro et al.,
2006). Positive emotions have been found to facil-
itate the experience of growth (Park et al., 1996).
Taking these ideas together, we hypothesize that the
association between pre-dissolution self-expansion
and post-dissolution growth will be mediated by
positive emotions. In contrast, because ending a
low quality relationship is not likely to produce
a great deal of negative emotion, and due to
evidence that positive emotional experiences can be
independent of negative emotional experiences
(see Feldman-Barrett & Russell, 1999, for a
review), we do not expect negative emotions to
mediate the self-expansion to growth pathway.
Addition through subtraction 43
Coping strategies. Due to the less stressful nature of
ending a low quality relationship, it may be more
likely that more positive coping strategies such as
reinterpretation and acceptance can be utilized.
This would seem to be beneficial to growth based
on previous research (Park et al., 1996; Schaefer &
Moos, 1992). Thus, we hypothesize that the associa-
tion between pre-dissolution self-expansion and
post-dissolution growth will be mediated by both
positive reinterpretation and acceptance. In contrast,
coping strategies such as social support, denial,
mental disengagement, and venting focus more on
decreasing negative outcomes, rather than increasing
positive ones. For this reason we do not expect these
coping strategies to mediate the association between
self-expansion and growth.
Prediction of growth
The final purpose of this study is to identify
predictors of post-dissolution growth. Based on
previous research, we included the following sets
(in order): pre-dissolution self-expansion, post-
dissolution emotional experience (positive and nega-
tive), post-dissolution coping strategies (positive
reinterpretation and acceptance), and post-
dissolution impact on self (loss of self and redis-
covery of the self). As suggested by Tabachnick and
Fidell (2001), the order of entry was based on logical
considerations. Specifically, the order was based on
the likely chronological order of a person’s dissolu-
tion experience. That is, coping strategies are likely
enacted as a result of emotional experiences, the
success of which would then determine how the self
is ultimately influenced. We hypothesized that each
set of variables would add a significant increment to
the overall variance accounted for in growth.
Method
Participants
The study contained 155 participants (39 males,
116 females) from a private university in the
northeast USA.
1
The sample ranged in age from 18
to 32 years, with a mean age of 19 years. Of the 155
participants, a vast majority (93.5%) identified
themselves as Caucasian. The majority of the
participants were freshmen (62.6%), 18.7% were
sophomores, 12.9% were juniors, and 5.2% were
seniors. On average, participants had experienced
the break-up within the past 11 weeks (range ¼1–24
weeks). The average relationship length prior to
dissolution was approximately 77 weeks (range ¼
4–236 weeks). A majority had ended an exclusive
relationship (85.2%), while 27.9% were currently
involved in a new relationship. A total of 45.2%
reported the break-up was initiated by self, 25.8% by
the partner, and 29% by both. Data were collected
over the course of two semesters (spring and fall of
2004) from the participant pool and from the general
University population.
2
Materials
Groups of 1–4 participants completed a question-
naire packet that asked about their most recent
romantic relationship break-up, as well as a short
demographic questionnaire (gender, age, year in
school, ethnicity, time since dissolution, initiator
status, and current relationship status).
Pre-dissolution
Self-Expansion Questionnaire. The Self-Expansion
Questionnaire (SEQ; Lewandowski & Aron, 2002)
assesses the extent to which a relationship is
experienced as expanding the self. Example items
include ‘‘How much does your partner help to
expand your sense of the kind of person you are?’’
‘‘How much has knowing your partner made you a
better person?’’ and ‘‘How much do you see your
partner as a way to expand your own capabilities?’’
Responses are made on a 7-point Likert scale
(1 ¼not at all; 7 ¼very much). In two studies,
Lewandowski and Aron found this 14-item measure
to be unifactorial with Cronbach’s alphas of 0.87 and
0.89. In the present study, we modified instructions
to focus on ‘‘the time period right BEFORE your
former relationship began to break-up.’’ Cronbach’s
alpha in the present study was 0.92.
Post-dissolution
Loss of Self Scale. A 6 item loss of self was developed
for the purposes of this study to measure feelings of
loss in the context of the self-concept. Each item was
measured on a 7-point Likert scale ranging from
1 (not at all) to 7 (a great deal). Items included
‘‘I do not know who I am,’’ ‘‘I have lost my sense of
self,’’ ‘‘I feel as though I am missing a part of me,’’
‘‘I feel as though many of my good qualities have
been lost,’’ ‘‘I do not feel like myself anymore,’’ and
‘‘I feel incomplete.’’ Cronbach’s alpha ¼0.90.
Rediscovery of Self Scale. A 6 item rediscovery of self
scale was developed for the purposes of this study to
measure the extent to which participants felt they
had become reacquainted with aspects of the self.
Each item was measured on a 7-point Likert scale
ranging from 1 (not at all) to 7 (a great deal). Items
included ‘‘I have done the things I once enjoyed that
I could not do while I was in my relationship,’’
44 G. W. Lewandowski & N. M. Bizzoco
‘‘I have regained my identity,’’ ‘‘I have reclaimed lost
parts of myself that I could not express while with my
partner,’’ ‘‘I have focused more on my needs that
were neglected while with my partner,’’ ‘‘I have
become reacquainted with the person I was before
the relationship,’’ and ‘‘I have rediscovered who
I am.’’ Cronbach’s alpha ¼0.91.
Positive Emotions Scale. A 20 item positive emotions
scale adapted from a scale used by Sprecher et al.
(1998). Sixteen items (calm, comforted, competent,
confident, empowered, energized, free, fulfilled,
hopeful, optimistic, pleased, relaxed, strong, thank-
ful, thrilled, wise), based on pilot testing, were added
to the original items (contentment, satisfaction,
relief, happiness). Love, an item used in the original
scale, was omitted from this version.
3
Participants
were asked ‘‘Please indicate the extent to which you
have experienced the following since the break-up
of your relationship?’’ Each item was measured on a
7-point Likert scale ranging from 1 (not at all) to
7 (extremely). Cronbach’s alpha ¼0.96.
Negative Emotions Scale. As with the positive emo-
tions scales, an 18-item negative emotions scale was
adapted from a scale used by Sprecher et al. (1998).
Ten items (exhausted, traumatized, empty, anxious,
confused, rejected, betrayal, bored, indifferent,
dissatisfied), based on pilot testing, were added
to the original items (depression, anger, hate,
frustration, resentment, loneliness, jealousy, hurt).
Participants were asked ‘‘Please indicate the extent to
which you have experienced the following since the
break-up of your relationship?’’ Each item was
measured on a 7-point Likert scale ranging from
1 (not at all) to 7 (extremely). Cronbach’s
alpha ¼0.93.
COPE Scale. Due to space limitations, only seven of
the original 15 subscales from the COPE inventory
(Carver et al., 1989) based on their potential
relevance to relationship dissolution. Specifically,
the following subscales were used: seeking social
support for instrumental reasons (e.g., ‘‘I ask people
who have had similar experiences what they did,’’
Cronbach’s alpha ¼0.72), seeking support for
emotional reasons (e.g., ‘‘I talk to someone about
how I feel,’’ Cronbach’s alpha ¼0.82), positive
reinterpretation and growth (e.g., ‘‘I look for some-
thing good in what is happening,’’ Cronbach’s
alpha ¼0.87), acceptance (e.g., ‘‘I learn to live
with it,’’ Cronbach’s alpha ¼0.86), focus on and
venting of emotions (e.g., ‘‘I get upset and let my
emotions out,’’ Cronbach’s alpha ¼0.89), denial
(e.g., ‘‘I refuse to believe that it has happened,’’
Cronbach’s alpha ¼0.85), mental disengagement
(e.g., ‘‘I turn to work or other substitute activities
to take my mind off things,’’ Cronbach’s
alpha ¼0.59).
Growth Scale. Post-dissolution growth was measured
using a 15-item scale (adapted from Tashiro &
Frazier, 2003). Commonly occurring responses to
the open-ended growth item in the Tashiro and
Frazier (2003) were modified slightly to fit a Likert
scale format. Participants were asked to indicate the
extent to which they had experienced each instance
of growth on a 7-point Likert scale ranging from
1 (not at all) to 7 (extremely). Example items include
‘‘I have grown as a person,’’ ‘‘I am more goal
oriented,’’ and ‘‘I have learned a lot about myself.’’
Cronbach’s alpha ¼0.92.
Global impact of dissolution. To determine the overall
impact of relationship dissolution participants were
given a single item that asked ‘‘Overall, how would
you describe the break-up’s impact on you?’’
Responses were made on a 9-point scale on which
(4¼extremely negative, 0 ¼neutral, 4 ¼extremely
positive).
Results
Means and standard deviations for all variables are
shown in Table I.
Prevalence of positive outcomes
In order to determine the extent to which partici-
pants reported positive outcomes following dissolu-
tion, a series of frequencies were obtained. Based on
this analysis 58.1% of the sample reported a mean
level of positive emotions above the scale midpoint.
Similarly, 71% reported a mean level of growth
above the midpoint. In contrast, only 31% of the
sample reported a mean level of negative emotions
above the scale midpoint. Finally, on the global
impact of dissolution item 33% reported a negative
overall impact, 25.7% were neutral, and 41.3%
reported a positive overall impact (because this was
a single-item measure, it was only used in these
descriptive analyses). Taken together, these results
indicate the potential for positive outcomes following
relationship dissolution.
Pre-dissolution self-expansion
Correlations among key variables are shown in
Table I. As shown in the table, our hypotheses regard-
ing self-expansion were supported. As hypothesized,
Addition through subtraction 45
Table I. Descriptive statistics and correlations among key variables.
123 4 5678910111213
1. Growth 4.61 (1.17)
2. Self-expansion 0.17* 4.32 (1.09)
3. Rediscovery of self 0.65*** 0.33*** 3.95 (1.73)
4. Loss of self 0.25** 0.35*** 0.18* 2.37 (1.35)
5. Positive emotions 0.63*** 0.25** 0.55*** 0.56*** 4.23 (1.30)
6. Negative emotions 0.11 0.14
y
0.01 0.65*** 0.46*** 3.34 (1.40)
7. Positive reinterpretation 0.50*** 0.12 0.32*** 0.45*** 0.59*** 0.32*** 3.19 (0.79)
8. Acceptance 0.34*** 0.15 0.32*** 0.42*** 0.49*** 0.24** 0.47*** 3.14 (0.77)
9. Social support (instrumental) 0.09 0.10 0.06 17* 0.01 0.22** 0.00 0.12 2.70 (0.82)
10. Social support (emotional) 0.16* 0.08 0.06 0.14
y
0.03 0.19* 0.22** 0.02 0.53*** 2.95 (0.87)
11. Venting emotions 0.10 0.29*** 0.18* 0.53*** 0.33*** 0.50*** 0.21** 0.23** 0.48*** 0.51*** 2.48 (0.91)
12. Denial 0.09 0.26** 0.19* 0.49*** 0.32*** 0.40*** 0.30*** 0.38*** 0.21** 0.06 0.44*** 1.28 (0.51)
13. Mental disengagement 0.06 0.17* 0.13 0.35*** 0.21** 0.38*** 0.18* 0.14
y
0.19* 0.06 0.33*** 0.27** 2.09 (0.59)
Notes: n¼155. Higher scores indicate a greater magnitude of each variable. All analyses are two-tailed.
y
p< 0.10; *p< 0.05; **p< 0.01; ***p< 0.001.
46 G. W. Lewandowski & N. M. Bizzoco
pre-dissolution self-expansion was significantly cor-
related with growth (r¼0.17; p¼0.03), rediscovery
of self (r¼0.33; p< 0.001), loss of self (r¼0.35;
p< 0.001), positive emotion (r¼0.25; p¼0.002),
positive reinterpretation (r¼0.12; p¼0.14), and
acceptance (r¼0.15; p¼0.06). These results
indicate that ending a low quality or less expanding
relationship was associated with greater growth and
positive emotions following dissolution, as well
as less loss of self. Although in the predicted
direction, the hypothesized associations between
pre-dissolution self-expansion and coping through
positive reinterpretation and acceptance were not
supported.
Mediation of the self-expansion growth relationship
We hypothesized that the relation of pre-dissolution
self-expansion to growth was mediated by several
variables, including rediscovery of the self, loss of
self, positive emotions, and coping strategies.
In order to test whether we had met the conditions
for mediation as delineated by Baron and Kenny
(1986), we conducted a series of three mediation
analyses for each potential mediator.
4
The first step in demonstrating mediation is to
establish the effect by demonstrating an association
between the independent variable (pre-dissolution
self-expansion) and the dependent variable (growth)
(Baron & Kenny, 1986). In this case, a regression
was run in which the hypothesized cause (pre-
dissolution self-expansion) was the predictor variable
and the hypothesized effect (growth) was the
dependent variable. The results for this analysis,
yielded a of 0.17, t(153) 2.17, p¼0.031.
This indicates that those who end a relationship with
less self-expansion are more likely to report growth
following dissolution. This step was consistent
through each of the following mediation analyses.
In addition, each mediation analysis, consistent
with Baron and Kenny (1986), contains a regression
in which the hypothesized cause (pre-dissolution self-
expansion) was the predictor, and the hypothesized
mediator (either rediscovery of self, loss of self,
positive emotions, or coping strategies) was the
dependent variable. Finally, each mediation analysis
contains a regression in which the hypothesized
cause (pre-dissolution self-expansion) and the
hypothesized mediator were entered simultaneously
as predictors, with the hypothesized effect (growth)
as the dependent variable.
Rediscovery of the self. We hypothesized that the
association between pre-dissolution self-expansion
and post-dissolution growth will be mediated by
rediscovery of the self. In the first regression, the
between self-expansion and rediscovery of the
self was 0.33, t(153) 4.26, p< 0.001, sr
2
¼0.33.
In the second regression, the between rediscovery
of the self and growth was 0.67, t(153) 10.22,
p< 0.001, sr
2
¼0.63. This is consistent with the
hypothesis that rediscovery of the self is a cause
of growth over and above any influence of self-
expansion on growth. Finally, the for self-
expansion in this regression was reduced to 0.04,
and was not significant, t(152) 0.67, p¼0.51,
sr
2
¼0.04. This represents a significant reduction
when compared to the of 0.17 in the unmediated
equation (Sobel’s test, Z¼3.92, p< 0.001). This
supports the hypothesis that rediscovery of the self
mediates pre-dissolution self-expansion’s effect on
growth. The pattern of the mediation is shown in
Figure 1.
Loss of self. We hypothesized that the association
between pre-dissolution self-expansion and post-
dissolution growth will be mediated by loss of self.
In the first regression, the between self-expansion
and loss of self was 0.35, t(153) 4.62, p< 0.001,
sr
2
¼0.35. In the second regression, the between
loss of self and growth was 0.22, t(153) 2.59,
p¼0.01, sr
2
¼0.20. This is consistent with the
hypothesis that loss of self is a cause of growth over
and above any influence of self-expansion on growth.
Finally, the for self-expansion in this regression
was reduced to 0.10, and was not significant,
t(152) 1.17, p¼0.24, sr
2
¼0.09. This represents
a significant reduction when compared to the of
Pre-dissolution
self-expansion
Pre-dissolution
self-expansion Growth
Growth
0.04
Rediscovery of
the self
−0.33*** 0.67***
Sobel’s test, Z = 3.92, p < 0.001
Pre-dissolution
self-expansion Growth
−0.10
Loss of self
0.35*** −0.22**
Sobel’s test, Z = 2.26, p = 0.02
Pre-dissolution
self-expansion Growth
−0.02
Positive emotions
−0.25** 0.62***
Sobel’s test, Z = −2.99, p < 0.01
−0.17*
Figure 1. Mediation of the self-expansion growth relationship by
rediscovery of the self, loss of self, and positive emotions.
Addition through subtraction 47
0.17 in the unmediated equation (Sobel’s test,
Z¼2.26, p¼0.02). This supports the hypothesis that
loss of self mediates pre-dissolution self-expansion’s
effect on growth. The pattern of the mediation is
shown in Figure 1.
Positive emotions. We hypothesized that the associa-
tion between pre-dissolution self-expansion and
post-dissolution growth will be mediated by positive
emotions. In the first regression, the between
self-expansion and positive emotions was 0.25,
t(153) 3.16, p¼0.002, sr
2
¼0.25. In the second
regression, the between positive emotions and
growth was 0.62, t(153) 9.58, p< 0.001, sr
2
¼0.60.
This is consistent with the hypothesis that positive
emotions are a cause of growth over and above any
influence of self-expansion on growth. Finally, the
for self-expansion in this regression was reduced
to 0.02, and was not significant, t(152) 0.29,
p¼0.77, sr
2
¼0.02. This represents a significant
reduction when compared to the of 0.17 in the
unmediated equation (Sobel’s test, Z¼2.99,
p¼0.002). This supports the hypothesis that positive
emotions mediate pre-dissolution self-expansion’s
effect on growth. The pattern of the mediation is
shown in Figure 1.
In contrast, we did not expect negative emotions to
mediate the self-expansion to growth pathway.
Consistent with this expectation, self-expansion was
not associated with negative emotions; ¼0.14,
t(153) 1.70, p¼0.09. The test of mediation was
also not significant, Sobel’s test, Z¼0.91, p¼0.36.
Coping strategies. We hypothesized that the associa-
tion between pre-dissolution self-expansion and
post-dissolution growth will be mediated by both
positive reinterpretation and acceptance. For positive
reinterpretation, in the first regression, the between
self-expansion and positive reinterpretation was
0.12, t(153) 1.49, p¼0.14, sr
2
¼0.12. In the
second regression, the between positive reinterpre-
tation and growth was 0.48, t(153) 6.85, p< 0.001,
sr
2
¼0.48. This is consistent with the hypothesis
that positive reinterpretation coping is a cause
of growth over and above any influence of self-
expansion on growth. Finally, the for self-
expansion in this regression was reduced to 0.12,
and was not significant, t(152) 1.64, p¼0.10,
sr
2
¼0.12. However, this does not represent a
significant reduction when compared to the of
0.17 in the unmediated equation (Sobel’s test,
Z¼1.45, p¼0.15). This fails to support the
hypothesis that positive reinterpretation mediates
pre-dissolution self-expansion’s effect on growth.
For acceptance, in the first regression, the
between self-expansion and acceptance was
0.15, t(153) 1.90, p¼0.06, sr
2
¼0.15. In the
second regression, the between acceptance and
growth was 0.32, t(153) 4.12, p< 0.001, sr
2
¼0.31.
This is consistent with the hypothesis that acceptance
coping is a cause of growth over and above any
influence of self-expansion on growth. Finally, the
for self-expansion in this regression was reduced to
0.13, and was not significant, t(152) 1.63,
p¼0.11, sr
2
¼0.12. However, this does not
represent a significant reduction when compared
to the of 0.17 in the unmediated equation
(Sobel’s test, Z¼1.72, p¼0.08). This fails to
support the hypothesis that acceptance mediates
pre-dissolution self-expansion’s effect on growth.
In contrast, we did not expect social support,
denial, mental disengagement, and venting to
mediate the association between self-expansion and
growth. In each case, these forms of coping did not
mediate the relationship based on Sobel’s test:
social support–instrumental (Z¼0.91, p¼0.36),
social support–emotional (Z¼0.89, p¼0.37),
denial (Z¼0.55, p¼0.59), mental disengagement
(Z¼1.01, p¼0.31), and venting (Z¼0.62,
p¼0.53). In sum, the hypothesized mediations of
positive reinterpretation and acceptance were in the
predicted direction and in the case of acceptance,
approached significance (p¼0.08). However, as
expected, the other types of coping were not
significant mediators.
Prediction of growth
We conducted a hierarchical regression by sets in
which the dependent variable was growth. The first
set focused on the pre-dissolution relationship’s
amount of self-expansion. The second set, focused
on emotional reactions, and included two variables,
positive and negative emotions.
5
The third set,
focused on coping strategies, and included two
variables, positive reinterpretation, and acceptance.
The fourth and final set, focused on post-dissolution
impact on the self, and included two variables,
loss of self, and rediscovery of the self. Results of the
analysis are shown in Table II.
As can be seen from Table II, the contribution
of the first set (pre-dissolution self-expansion) to
predicting the dependent variable, though signifi-
cant, was small (R
2
¼0.03; p¼0.03). Adding the
second set (emotional reactions) significantly
increased the variance accounted for (R
2
¼0.41;
p< 0.001). Adding the third set (coping strategies)
also contributed a significant increment to the
variance accounted for (R
2
¼0.03; p¼0.02).
Adding the final set (impact on self) also contributed
a significant increment to the variance accounted for
(R
2
¼0.10; p< 0.001). Thus, as hypothesized, each
48 G. W. Lewandowski & N. M. Bizzoco
set of variables significantly contributed to predicting
growth following dissolution. Overall, the variables in
this analysis accounted for 57% of the variance in
growth.
Exploratory analyses
Although not hypothesized, and not the focus of
the present study, past research suggests a few
potential correlates of positive outcomes such as
initiator status, current relationship status, and time
factors.
6
Discussion
This study sought to determine the prevalence of
positive outcomes such as growth and positive
emotions following dissolution as well as potential
factors associated with the experience of post-
dissolution growth. The findings generally support
the hypotheses and are consistent with previous
research. According to the stress-relief pathway to
growth, ending a low quality relationship increases
the likelihood of post-dissolution growth (Tashiro
et al., 2006). Previous work on this pathway has
been done exclusively in the context of divorce.
The present study, however, represents the first time
the stress-relief pathway has been examined in the
context of premarital dissolution. As hypothesized in
the present study, ending a relationship that provided
less self-expansion was associated with greater
growth post-dissolution. This relationship was
mediated by rediscovery of the self, loss of self, and
positive emotions. Finally, as hypothesized several
sets of variables significantly accounted for variance
in post-dissolution growth.
Prevalence of positive outcomes
The present study focuses on the positive side
of relationship dissolution. This approach is consis-
tent with previous work on stress-related growth
(e.g., Park et al., 1996; Schaefer & Moos, 1992;
Tedeschi & Calhoun, 1996), as well as positive
psychology more generally (cf. Seligman &
Csikszentmihalyi, 2000). Similar to work done by
Tashiro and Frazier (2003), the present results
indicate that growth and positive emotions may be
a larger part of the relationship dissolution experi-
ence than previously thought. In the current sample,
a majority of participants reported mean levels of
growth (71%) and positive emotions (58%) above
the scale midpoint, compared to only 31% for
negative emotions. The prevalence of positive
outcomes is consistent with previous work demon-
strating positive life changes in the context of
bereavement (e.g., Frazier et al., 2004; Park &
Cohen, 1993). The present results also offer a
complimentary perspective to existing research on
distress following dissolution, and extend existing
work demonstrating positive outcomes.
It is important to note that the nature of this study
can not establish whether participants’ positive views
of their dissolution experience represent their true
experience or their perspective on the events that
transpired. This is especially true in light of Murray,
Holmes, & Griffins (1996) work on how positive
illusions operate within relationships. However, their
work also suggests that possibility that the true
experience may be less important, and that poten-
tially biased views can be beneficial to psychological
functioning.
Pre-dissolution self-expansion
According to the self-expansion model (Aron &
Aron, 1997), relationships that provide opportunities
to enhance the self are particularly rewarding and are
associated with high relationship quality. Losing
this type of relationship has been shown to be
detrimental, particularly to the self (Lewandowski
et al., 2006). The present study examined the
opposite possibility that ending a less expanding
relationship results in positive outcomes such as
growth. This prediction is consistent with the stress-
relief pathway to growth which suggests that ending a
Table II. Hierarchical multiple regression by sets predicting growth from pre-dissolution and post-dissolution self variables,
coping strategies, and emotional reactions.
sr
2
R
2
change Fchange Overall R
2
Step 1 Self-expansion 0.17* 0.17 0.03* 4.72* 0.03
Step 2 Negative emotions 0.23*** 0.26
Positive emotions 0.73*** 0.64 0.41*** 54.53*** 0.43
Step 3 Positive reinterpretation 0.22** 0.22
Acceptance 0.01 0.01 0.03* 4.12* 0.45
Step 4 Loss of self 0.03 0.03
Rediscovery of self 0.42*** 0.43 0.10*** 17.03*** 0.57
Notes: n¼155. p< 0.10; *p< 0.05; **p< 0.01; ***p< 0.001.
Addition through subtraction 49
low quality relationship results in less stress in
addition to new opportunities for growth (Tashiro
et al., 2006). In essence, a person can accomplish
additions to one’s life through subtraction of a
lesser quality relationship. As hypothesized, ending
a relationship containing less self-expansion was
associated with greater amounts of growth post-
dissolution. Further, lower pre-dissolution self-
expansion was associated with more positive
emotions, rediscovery of the self, and less loss of self.
Taken together, the present findings related to
pre-dissolution self-expansion represent a novel
application of the self-expansion model. Previous
research in this model has focused on self-expansion
as a motivation for entering relationships (e.g., Aron,
Paris, & Aron, 1995), and as a means of improving
existing relationships (e.g., Aron, Norman, Aron,
& Lewandowski, 2003). This study broadens the
self-expansion model by demonstrating its ability to
predict experiences following the dissolution of a
relationship.
Mediation of the self-expansion growth relationship
In addition to establishing the basic stress-relief
pathway between pre-dissolution self-expansion and
post-dissolution growth, we also found support for
several hypothesized mediators. As hypothesized,
rediscovery of the self mediated the self-expansion
to growth pathway. This finding is consistent
with Tashiro and Frazier (2003) who found that
rediscovery of the self was associated with positive
emotions and growth. The fact that rediscovery
of the self mediates the self-expansion to growth
pathway suggests that the previous relationship may
have been serving as a barrier to growth, such that
when it ended it provided an opportunity to regain
lost elements that ultimately facilitates growth.
Past research suggests that dissolution can result in
loss of self (Lewandowski et al., 2006) and that loss
of self is associated with negative emotions (Drew
et al., 2004). If loss of self is a negative experience,
the extent to which it could be minimized
would seem to increase the possibility of growth.
In addition, loss of self would be less likely following
the dissolution of a low quality relationship. Thus,
as our present results support, loss of self seems
ideally suited as a mediator of the self-expansion to
growth path.
Previous work in the area of posttraumatic growth
suggests that positive emotions and growth are
positively correlated (Park et al., 1996). However,
this is the first study that examines growth solely in
context of dissolution rather than traumatic events
in general. This is a particularly important contribu-
tion because some may not consider a premarital
break-up to be overly traumatic. While this study
confirms that growth is possible following a poten-
tially traumatic event, it extends the basic finding by
placing the variables in the larger context of the
stress-relief pathway. Specifically, as hypothesized,
positive emotions mediated the self-expansion to
growth path which suggests that the relationship
between positive emotions and growth is particularly
important following the dissolution of a low quality
relationship. Further, as expected, negative emotions
did not mediate the self-expansion to growth path
which suggests that the experience of growth is
more closely related to the presence of positive
emotions, rather than the absence of negative
emotions. This finding lends additional support
for conceptualizing positive and negative emotions
as independent experiences (Feldman-Barrett &
Russell, 1999).
Although in the predicted direction, our hypoth-
eses regarding the mediating effects of positive
reinterpretation and acceptance coping on the self-
expansion to growth pathway were not supported.
One possible explanation is that ending a relationship
that provided little self-expansion produces such low
levels of stress and that positive feelings are already
so prevalent that enlisting these techniques becomes
superfluous. It is also possible that these types of
coping took place while the relationship was still
intact as a means of mentally disengaging and
preparing for the relationship’s eventual demise
(Duck & Lea, 1983). However, as expected, coping
strategies such as social support, denial, venting, and
mental disengagement that aim more specifically at
decreasing negative emotions did not mediate the
self-expansion to growth path.
It should also be noted that, consistent with
previous research (Park et al., 1996; Schaefer &
Moos, 1992), positive reinterpretation and accep-
tance were positively correlated with growth.
Consistent with the findings of Tashiro and Frazier
(2003), we found a small association between
emotional social support coping and growth.
However, instrumental social support was not
associated with growth. Taken together, these find-
ings suggest that all forms of social support may not
be beneficial for growth. In fact, the small effect that
emotional social support has on growth suggests that,
perhaps due to the more cognitive nature of growth
(Tedeschi et al., 1998), other types of coping may be
better suited for promoting growth.
Initiator status
Consistent with the stress-relief pathway, those
who were in a less expanding (i.e., low quality)
relationship were more likely to initiate dissolution
50 G. W. Lewandowski & N. M. Bizzoco
(Tashiro et al., 2006). In this situation, it is likely that
dissolution was initiated as a means of removing
one’s self from a limiting relationship. Also,
consistent with a stress-relief pathway, initiators
experienced more relief in the form of rediscovery
of the self, less loss of self, and increased positive
emotions. However, initiator status was not
significantly related to the experience of growth.
This finding contrasts some previous work
(Helgeson, 1994), but it supports other work that
specifically focused on growth in the context of
relationship dissolution (Tashiro & Frazier, 2003).
Prediction of growth
Finally, this study sought to identify predictors of
post-dissolution growth. A regression analysis by
sets revealed that each set (pre-dissolution self-
expansion, post-dissolution emotional experience,
coping strategies, and self change) accounted for
a significant amount of the variance in growth.
The overall model accounted for nearly 57% of
the variance in growth. Within this analysis, the
set involving impact on the self contributed an
additional 10% of variance in the last step of the
analysis. This suggests that in order to fully account
for a person’s post-dissolution growth experience,
several different types of variables need to be
assessed, but that impact on the self may be
especially important to examine.
Strengths and limitations
The main strength of this study was the novel
approach to understanding post-dissolution experi-
ences. In addition, this study brings together several
theoretical perspectives that inform the pursuit of
factors related to positive outcomes. Although
correlational in nature, this study represents an
important step toward obtaining a more complete
picture of dissolution. A few limitations of the study
should also be noted. First, the use of a college
sample may not allow the results to generalize to
other populations such as those dealing with divorce.
In fact, the additional stressors (e.g., combined
finances, children, etc.) associated with divorce may
make positive outcomes less prevalent than indicated
in the present sample. Second, the present sample
was predominantly female which may influence
mean levels of positive vs. negative outcomes.
However, research has shown that males and
females report similar distress levels post-dissolution
(e.g., Attridge, Berscheid, & Simposon, 1995;
Frazier & Cook, 1993; Sprecher, 1994, Sprecher
et al., 1998).
Finally, although fairly common in dissolution
research (e.g., Frazier & Cook, 1993), and in
previous research dealing with positive outcomes
(e.g., Sprecher et al., 1998; Tashiro & Frazier, 2003),
this study did contain a retrospective component and
is thus prone to the typical problems associated with
retrospective data. For example, people may be
biased to view a particularly negative event such
as dissolution more positively in order to preserve
self-esteem. However, if this were exclusively the
case, other retrospective research on dissolution
would have had difficulty findings reports of distress
(e.g., Frazier & Cook, 1993; Simpson, 1987;
Sprecher et al., 1998). To help combat some of
this potential bias, all analyses were run to control for
the influence of length of relationship, time since
dissolution, and current relationship status. In spite
of these efforts, retrospective data such as this should
be seen as an initial step toward understanding the
wide range of outcomes post-dissolution, and serve
as the foundation for more scientifically rigorous
examinations.
Future directions and implications
Due to the novelty of looking at the positive side of
dissolution, possibilities for future research are
abundant. New potential areas for exploration
include examining other pre- and post-dissolution
correlates of positive outcomes such as love, satisfac-
tion, commitment, and attributions for the dissolu-
tion. In addition, other potential positive outcomes
could be explored such as general happiness,
increases in social networks, and self-esteem.
Future research in these areas should focus on
measuring participants at several time points,
before and after dissolution, to more clearly establish
the directionality of the findings.
The present findings also have implications for
therapeutic contexts. Specifically, it is possible that
findings linking growth and positive emotions to
rediscovering the self, positive reinterpretation,
and acceptance coping can be used to develop
strategies to help people experience more positive
outcomes following dissolution. In addition, the
general finding that a large percentage of people
experience positive outcomes following dissolution
may be beneficial to individuals currently in bad
relationships. That is, rather than focusing on the
negative consequences of dissolution as a reason to
stay in a bad relationship, people could use the
present results as motivation for leaving the bad
relationship. In fact, the present results suggest that
leaving a bad relationship is likely to result in
personal growth and positive emotions.
Addition through subtraction 51
Conclusion
The purpose of this study was to explore correlates
of positive the possibility that relationship dissolution
could result in outcomes such as growth and
positive emotions. Results indicate that a substantial
number of participants report these positive
outcomes post-dissolution. Consistent with predic-
tions, growth was significantly associated with
pre-dissolution self-expansion, and this relationship
was mediated by rediscovery of the self, loss of self,
and positive emotions. These results demonstrate
that dissolution is not an entirely negative experi-
ence. Rather, this research suggests that positive
outcomes are possible following dissolution, changes
to the self play a large part in that experience,
and that growth is likely to occur after the subtraction
of a low quality relationship.
Acknowledgements
We would like to thank the following individuals for
their help on this project: Robert Ackerman, Lynne
Canberg, Katherine Evans, Jennifer Krops, Christina
Messina, Natalie Nardone, Alanna Raines, and
Krystle Serago.
Notes
1. Forty participants were omitted for failing to meet
the inclusion criteria of the study (i.e., break-up
occurred more than 6 months ago), or for
completing the study packet in an unreasonably
short amount of time (i.e., finishing an 8 page
packet in under 10 minutes).
2. Participants from the general University sample
received a chance to win a US$50 mall gift
certificate for their participation. All other
participants received course credit.
3. In light of research on unrequited love (e.g.,
Baumeister & Wotman, 1992), love was omitted
due to concerns over whether it could be
considered a purely positive emotional experi-
ence. If anything, this omission would work
against our effect due to previous findings
showing love to be the most frequently mentioned
post-dissolution positive emotion (Sprecher,
1994).
4. All of the mediation analyses (as well as the basic
associations between self-expansion and key
variables) were also run controlling for initiator
status, current relationship status, time since
relationship dissolution, and pre-dissolution rela-
tionship length. In each case the results parallel
those reported, with all associations remaining
significant.
5. The reported order of entry for sets was based on
logical considerations (Tabachnick & Fidell,
2001). We also ran this analysis several other
times with each possible order of post-dissolution
sets. In each analysis using an alternate order, the
addition of each set contributed a significant
increase in the overall amount of variance
accounted for.
6. Initiator status. A one-way analysis of variance
comparing initiator status (Self, Partner, Mutual)
on key variables was conducted. As seen in
Table III, although initiator status was not
significantly related to growth, it was significantly
related to pre-dissolution self-expansion, redis-
covery of the self, loss of self, positive emotion,
Table III. The influence of initiator status on key variables.
Initiator status
Dependent variables Self Partner Mutual F
Growth 4.75 4.33 4.65 1.65
Self-expansion 4.07
a
4.64
a
4.45 4.09*
Rediscovery of self 4.27
a
3.29
a
4.06 4.34*
Loss of self 2.02
a
3.05
ab
2.31
b
8.17***
Positive emotions 4.60
a
3.42
ab
4.37
b
12.43***
Negative emotions 2.80
a
4.37
ab
3.27
b
20.62***
Positive reinterpretation 3.30
a
2.92
a
3.26 3.27*
Acceptance 3.17 2.99 3.23 1.18
Social support (instrumental) 2.69 2.97
a
2.47
a
4.07*
Social support (emotional) 2.87 3.17 2.89 1.66
Venting emotions 2.28
a
3.01
ab
2.32
b
10.19***
Denial 1.20
a
1.48
ab
1.22
b
4.50*
Mental disengagement 2.03 2.15 2.12 0.52
Notes: n¼155. Within each dependent variable, means with superscripts in common are significantly different at the
p< 0.05 level based on a post hoc (Bonferroni) analysis. *p< 0.05; ***p< 0.001.
52 G. W. Lewandowski & N. M. Bizzoco
negative emotions, positive reinterpretation,
social support (instrumental), venting, and
denial. In a majority of the cases, there were
significant differences between self initiation and
partner initiation based on post hoc (Bonferroni)
analyses such that those who reported being
the initiator experienced less pre-dissolution
self-expansion, more rediscovery of the self, less
loss of self, more positive emotions, less negative
emotions, more positive reinterpretation, less
venting, and less denial. Compared to partner
initiation, those who reported mutual initiation
had the same experiences. The only exceptions
were for self-expansion, rediscovery of the self,
and positive reinterpretation where no differences
between partner and mutual initiation were
found. In addition, those who reported their
partner initiated the dissolution reported coping
through instrumental social support more than
those who reported that the dissolution was
mutual.
Due to the association between initiator status,
self-expansion, and growth, an additional analysis
was conducted to determine if initiator status
moderated the effect of pre-dissolution self-
expansion on growth. To test this, we conducted
a simultaneous multiple regression analysis with
self-expansion and initiator status entered as the
independent variables in step one, the interaction
term entered in step two, and growth as the
dependent variable. Results of the analysis
indicate that self-expansion was the only signifi-
cant predictor in step one, ¼0.17, t(152)
2.09, p¼0.04, and that the influence of initiator
status was not significant, ¼0.02, t(152)
0.27, p¼0.79. In addition, the interaction
term was not significant, ¼0.27, t(151) 0.66,
p¼0.51, indicating that initiator status does not
moderate the association between self-expansion
and growth. This analysis was also run comparing
self initiators with partner initiators (excluding
those who reported the break-up was mutually
initiated). The results were virtually identical.
Current relationship status. At-test was com-
puted comparing those in a new relationship with
those not in a new relationship on all key
variables. Two significant differences (self-expan-
sion and rediscovery of the self) emerged from the
13 comparisons. For self-expansion, those in a
new relationship (M¼3.98, SD ¼1.11) were
significantly different than those not in a new
relationship (M¼4.45, SD ¼1.06);
t(152) ¼2.42, p¼0.02 (two-tailed), effect size
(d)¼0.19. For rediscovery of the self, those in
a new relationship (M¼4.55, SD ¼1.79)
were significantly different than those not in a
new relationship (M¼3.72, SD ¼1.66);
t(152) ¼2.75, p¼0.01 (two-tailed), effect size
(d)¼0.22. These findings suggest that those who
are in a new relationship had lower self-expansion
in the previous relationship, and have rediscov-
ered the self to a greater extent.
Due to the association between current rela-
tionship status and self-expansion, an additional
analysis was conducted to determine if current
relationship status moderated the effect of pre-
dissolution self-expansion on growth. To test this,
we conducted a simultaneous multiple regression
analysis with self-expansion and current relation-
ship status entered as the independent variables in
step one, the interaction term entered in step two,
and growth as the dependent variable. Results of
the analysis indicate that self-expansion was the
only significant predictor in step one, ¼0.17,
t(152) 2.03, p¼0.04, and that the influence of
current relationship status was not significant,
¼0.04, t(152) 0.49, p¼0.62. In addition,
the interaction term was not significant, ¼0.45,
t(151) 0.93, p¼0.35, indicating that current
relationship status does not moderate the associa-
tion between self-expansion and growth.
Time factors. A series of correlations were run to
determine if amount of time since dissolution,
or length of the dissolved relationship were
related to the experience of growth and positive
emotions. Time since dissolution, was not
significantly associated with any of the key
variables. The only exception was that time
since dissolution’s association with growth appro-
ached significance (r¼0.15, p¼0.06). Similarly,
length of the dissolved relationship was largely
unrelated to the key variables. The only excep-
tions were that length of dissolved relationship
was significantly associated with loss of self
(r¼0.34, p< 0.001), negative emotions (r¼0.26,
p¼0.001), and coping through venting (r¼0.17,
p¼0.04).
References
Aron, A., & Aron, E. N. (1997). Self-expansion motivation
and including other in the self. In S. Duck (Ed.), Handbook of
personal relationships: Theory, research and interventions (2nd ed.,
pp. 251–270). New York: John Wiley & Sons Ltd.
Aron, A., Aron, E. N., & Norman, C. (2001). Self-expansion
model of motivation and cognition in close relationships
and beyond. In M. Clark & G. Fletcher (Eds.), Blackwell’s
handbook of social psychology, Vol. 2: Interpersonal processes
(pp. 478–501). Oxford: Blackwell Publishers.
Aron, A., Norman, C., Aron, E. N., & Lewandowski Jr, G. W.
(2003). Shared participation in self-expanding activities:
Positive effects on experienced marital quality. In P. Noller &
J. Feeney (Eds.), Marital interaction (pp. 177–196). Cambridge:
Cambridge University Press.
Addition through subtraction 53
Aron, A., Paris, M., & Aron, E. N. (1995). Falling in love:
Prospective studies of self-concept change. Journal of Personality
and Social Psychology,69, 1102–1112.
Attridge, M., Berscheid, E., & Simpson, J. A. (1995). Predicting
relationship stability from both partners versus one. Journal of
Personality and Social Psychology,69, 254–268.
Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator
variable distinction in social psychological research:
Conceptual, strategic, and statistical considerations. Journal of
Personality and Social Psychology,51, 1173–1182.
Baumeister, R. F., & Wotman, S. R. (1992). Breaking hearts:
The two sides of unrequited love. New York: Guilford.
Calhoun, L. G., Cann, A., Tedeschi, R. G., & McMillan, J.
(2000). A correlational test of the relationship between
posttraumatic growth, religion, and cognitive processing.
Journal of Traumatic Stress,13, 521–527.
Caplan, G. (1964). Principles of preventative psychiatry. New York:
Basic Books.
Carver, C., Scheier, M., & Weintraub, J. (1989). Assessing coping
strategies: A theoretically based approach. Journal of Personality
and Social Psychology,56, 267–283.
Choo, P., Levine, T., & Hatfield, E. (1996). Gender, love
schemas, and reactions to romantic break-ups. Journal of
Social Behavior & Personality,11, 143–160.
Drew, S. S., Heesacker, M., Frost, H. M., & Oelke, L. E. (2004).
The role of relationship loss and self-loss in women’s and
men’s dysphoria. Journal of Social and Personal Relationships,21,
381–397.
Duck, S. W., & Lea, M. (1983). Breakdown of relationships as a
threat to personal identity. In G. M. Breakwell (Ed.), Threatened
identities (pp. 53–73). New York: Wiley.
Feldman-Barrett, L., & Russell, J. A. (1999). The structure
of current affect: Controversies and emerging consensus.
Current Directions in Psychological Science,8, 10–14.
Fine, M. A., & Sacher, J. A. (1997). Predictors of distress
following relationship termination among dating couples.
Journal of Social and Clinical Psychology,16, 381–388.
Frazier, P., & Cook, S. W. (1993). Correlates of distress following
heterosexual relationship dissolution. Journal of Social and
Personal Relationships,10, 55–67.
Frazier, P., Steward, J., & Mortensen, H. (2004). Perceived
control and adjustment to trauma: A comparison across events.
Journal of Social and Clinical Psychology,23, 303–324.
Helgeson, V. S. (1994). Long-distance romantic relationships: Sex
differences in adjustment to break-up. Personality and Social
Psychology Bulletin,20, 254–265.
Holmes, T. H., & Rahe, R. H. (1967). The Social Readjustment
Rating Scale. Journal of Psychosomatic Research,11, 213–218.
Leary, M. R., & Tangney, J. P. (2003). The self as an organizing
construct in the behavioral and social sciences. In M. R. Leary
& J. P. Tangney (Eds.), Handbook of self and identity (pp. 3–14).
New York: Guilford Press.
Lewandowski, G. W., Aron, A. P., Bassis, S., & Kunak, J. (2006).
Losing a self-expanding relationship: Implications for the
self-concept. Personal Relationships,13, 317–331.
McMillen, J. C., & Fisher, R. H. (1998). The perceived benefit
scales: Measuring perceived positive life changes after negative
events. Social Work Research,22, 173–186.
Murray, S. L., Holmes, J. G., & Griffin, D. W. (1996). The
self-fulfilling nature of positive illusions in romantic
relationships: Love is not blind, but prescient. Journal of
Personality and Social Psychology,71, 1155–1180.
Park, C. L., & Cohen, L. H. (1993). Religious and non-religious
coping with the death of a friend. Cognitive Therapy and
Research,17, 561–577.
Park, C. L., Cohen, L. H., & Murch, R. L. (1996). Assessment
and prediction of stress-related growth. Journal of Personality,
64, 71–105.
Ruvolo, A. P., Fabin, L. A., & Ruvolo, C. M. (2001). Relationship
experiences and change in attachment characteristics of
young adults: The role of relationship breakups and conflict
avoidance. Personal Relationships,8, 265–281.
Schaefer, J., & Moos, R. (1992). Life crises and personal growth.
In B. Carpenter (Ed.), Personal coping: Theory, research, and
application (pp. 149–170). Westport, CT: Praeger.
Seligman, M. E. P., & Csikszentmihalyi, M. (2000). Positive
psychology: An introduction. American Psychologist,55, 5–14.
Simpson, J. A. (1987). The dissolution of romantic
relationships: Factors involved in relationship stability and
emotional distress. Journal of Personality and Social Psychology,
53, 683–692.
Sprecher, S. (1994). Two sides to the breakup of a dating
relationship. Personal Relationships,1, 199–222.
Sprecher, S., Felmlee, D., Metts, S., Fehr, B., & Vanni, D.
(1998). Factors associated with distress following the breakup
of a close relationship. Journal of Social and Personal
Relationships,15, 791–809.
Tabachnick, B. G., & Fidell, L. S. (2001). Using multivariate
statistics (4th ed.) Boston, MA: Allyn and Bacon.
Tashiro, T., & Frazier, P. (2003). ‘‘I’ll never be in a relationship
like that again’’: Personal growth following romantic relation-
ship breakups. Personal Relationships,10, 113–128.
Tashiro, T., Frazier, P., & Berman, M. (2006). Stress-related
growth following divorce and relationship dissolution.
In M. A. Fine & J. H. Harvey (Eds.), Handbook of divorce and
relationship dissolution (pp. 361–384). Mahwah, NJ: Lawrence
Erlbaum and Associates.
Tedeschi, R. G., & Calhoun, L. G. (1996). The posttraumatic
growth inventory: Measuring the positive legacy of trauma.
Journal of Traumatic Stress,9, 455–471.
Tedeschi, R. G., Park, C. L., & Calhoun, L. G. (Eds.). (1998).
Posttraumatic growth:Positive changes in the aftermath of crisis.
Mahwah, NJ: Lawrence Erlbaum.
54 G. W. Lewandowski & N. M. Bizzoco
