ArticlePDF Available

Can We Be (and Stay) Friends? Remaining Friends After Dissolution of a Romantic Relationship



Although many individuals report being friends with their ex-romantic partners (Wilmot, Carbaugh, & Baxter, 198514. Wilmot , W. W. , Carbaugh , D. A. and Baxter , L. A. 1985. Communicative strategies used to terminate romantic relationships. Western Journal of Speech Communication, 49: 204–216. [Taylor & Francis Online]View all references), the literature regarding post-romantic friendships is very limited. We investigated whether satisfaction in the dissolved romantic relationship could predict post-romantic friendships and friendship maintenance. We found that the more satisfied individuals were during the dissolved romance, the more likely they were to remain friends and the more likely they were to engage in friendship maintenance behaviors. We also found that friendship maintenance fully mediated the association between past romantic satisfaction and current friendship satisfaction.
The Journal of Social Psychology, 2011, 151(5), 662–666
Copyright © Taylor & Francis Group, LLC
Can We Be (and Stay) Friends? Remaining
Friends After Dissolution of a Romantic
Saint Louis University
Murray State University
Saint Louis University
Ashland University
ABSTRACT. Although many individuals report being friends with their ex-romantic
partners (Wilmot, Carbaugh, & Baxter, 1985), the literature regarding post-romantic
friendships is very limited. We investigated whether satisfaction in the dissolved romantic
relationship could predict post-romantic friendships and friendship maintenance. We found
that the more satisfied individuals were during the dissolved romance, the more likely they
were to remain friends and the more likely they were to engage in friendship mainte-
nance behaviors. We also found that friendship maintenance fully mediated the association
between past romantic satisfaction and current friendship satisfaction.
Keywords: friendship, relationship maintenance, relationship quality
INDIVIDUALS ARE GETTING MARRIED later in life (Popenoe & Whitehead,
2004), and opportunities for non-marital romantic relationships are increasing
(Hebert & Popadiuk, 2008). However, the extant non-marital dissolution litera-
ture is limited. Research shows that ex-partners are more likely to be friends if
they were friends before the romance (Metts, Cupach, & Bejlovec, 1989), if the
breakup was mutual (Hackathorn, Clark, Mattingly, Bullock, & Weaver, 2008),
Address correspondence to Melinda Bullock, Saint Louis University, Department of Psy-
chology, 221 N. Grand Blvd., Shannon Hall, St. Louis, MO 63128 USA;
Downloaded by [Saint Louis University] at 06:41 30 August 2011
Bullock et al. 663
or if the friendship is socially supported (Busboom, Collins, Givertz, & Levin,
Like any relationship, ex-partners must maintain their friendship if they want
it to last. The Friendship Maintenance Scale (FMS; Oswald, Clark, & Kelly, 2004)
assesses maintenance by measuring a variety of behaviors (e.g., “How often do
you and your friend compliment each other?”) using four subscales: positivity
(makes the friendship more rewarding), supportiveness (friend/friendship sup-
port), openness (self-disclosure), and interaction (joint activities).
While a clear reciprocal association exists between maintenance behaviors
and relationship satisfaction (e.g., Canary & Stafford, 2001), the association
between satisfaction from the dissolved romance and the post-romantic friendship
has not been examined. We expected past romantic satisfaction to be positively
associated with current friendship status and friendship maintenance behaviors
with an ex-partner. Because relationally satisfied individuals engage in mainte-
nance behaviors (e.g., Canary & Stafford, 2001) and satisfaction is a relational
outcome of maintenance (e.g., Stafford & Canary, 1991), we expected that
friendship maintenance would mediate the association between past romantic
satisfaction and current friendship satisfaction.
Participants were 131 undergraduates who were relatively young (M=19.11
years, SD =1.36), predominantly female (81%) and Caucasian (75%), who
dated their most recent ex-romantic partner for an average of 13.60 months
(Mdn =8 months), and were broken-up for an average of 12.30 months
(Mdn =7 months).
Participants indicated on an 8-point scale (0 =not friends, 7=best
friends) their current friendship status with their most recent ex-romantic partner
(M=3.49, SD =1.96). A continuous scale was used because it allowed partici-
pants to indicate: 1) if a friendship existed at all (i.e., with a response of “0”) and
2) the relative closeness of the friendship. Only 11.5% of participants answered
“0” to this question; 7.6%. answered “1”; 13.0%,answered “2”; 10.0% answered
“3”; 22.1% answered “4”; 21.4%,answered “5”; 11.5% answered “6”; and 3.1%
answered “7”.
Participants also completed the satisfaction subscale of the Investment
Model Scale (Rusbult, Martz, & Agnew, 1998) regarding their current friend-
ship (M=3.28, SD =1.97) and a modified version of the satisfaction subscale
regarding their dissolved romance (M=4.50, SD =2.14) with their ex-partner.
Satisfaction responses were averaged for the dissolved romance (α=.90) and
for the current friendship (α=.96). Finally, participants completed the 37-item
FMS (M=5.01, SD =2.10). Each subscale demonstrated adequate reliability
among the current sample: positivity, α=.73; supportiveness, α=.90; openness,
α=.82; and interaction, α=.92.
Downloaded by [Saint Louis University] at 06:41 30 August 2011
664 The Journal of Social Psychology
As hypothesized, past romantic satisfaction and friendship status were sig-
nificantly correlated, r=.48, p<.001, indicating that those who were more
satisfied during the romantic relationship were more likely to remain friends.
Dichotomizing friendship status (participants who answered “0” to the 8-point
item were classified as not friends, whereas everyone else was classified as
friends), and conducting a logistic regression yielded similar results, OR =1.64,
p<.001. Additionally, among participants who reported a friendship with
their ex-partner, satisfaction during the dissolved relationship was significantly
correlated with each subscale of the FMS (rs ranging from .33 to .46, ps<.001).
Because the FMS subscales were highly intercorrelated (rs ranged from .60 to
.89), and there was evidence of multicollinearity when predicting current friend-
ship satisfaction (tolerance ranged from .18 to .54; Tabachnik & Fidell, 2007), we
created a composite FMS score (α=.94), which was significantly correlated with
past romantic satisfaction (r=.51, p<.001). Also as hypothesized, friendship
maintenance fully mediated the association between past romantic satisfaction
and current friendship satisfaction (Baron & Kenny, 1986): Past romantic satisfac-
tion significantly predicted satisfaction with the friendship (β=.17, p=.05) and
friendship maintenance (β=.51, p<.001), friendship maintenance significantly
predicted satisfaction with the friendship after controlling for past romantic satis-
faction (β=.61, p<.001), and the association between past romantic satisfaction
and current friendship satisfaction was significantly reduced when friendship
maintenance was included in the model, suggesting full mediation (β=–.13,
p>.05), Z =4.91, p<.001. Furthermore, this pattern remained when controlling
for time since breakup.
The current study indicates that individuals who were satisfied during the dis-
solved romantic relationship are more likely to be friends with their ex-romantic
partner and are more likely to use friendship maintenance behaviors. Importantly,
friendship maintenance fully mediated the association between past romantic
satisfaction and current friendship satisfaction, suggesting that if a romantic rela-
tionship was satisfying, it is more likely that a friendship will emerge and be
maintained, which in turn makes the friendship more satisfying.
Many of the current participants reported being friends with their ex-romantic
partner. However, this could be because they are likely to live in close physi-
cal proximity (i.e., campus housing) and are more likely to share friends with
an ex-romantic partner than with other individuals, thus limiting generalizabil-
ity. Additionally, our mostly Caucasian and female sample limits generalizability.
Finally, current friendship satisfaction could inhibit accurate recall of the past
Downloaded by [Saint Louis University] at 06:41 30 August 2011
Bullock et al. 665
romantic satisfaction. Research indicates negative affect fades faster than positive
affect (Ritchie & Skowronski, 2008), and within the context of marriages, past
relationship quality resembles current relationship quality (Karney & Coombs,
2000). Therefore, the relationship between current friendship and past romantic
satisfaction could be biased.
The current study adds to the post-romance friendship literature by demon-
strating that past romantic satisfaction influences whether a friendship will emerge
and be maintained. Although the romantic relationship may not have worked,
individuals can turn a satisfying, albeit dissolved, romance into a meaningful and
rewarding friendship.
Melinda Bullock is a Doctoral Candidate in social psychology at Saint Louis
University. Her research interests include close relationships and health persua-
sion. Jana Hackathorn is an Assistant Professor of psychology at Murray State
University. Her research interests lie in romantic relationships, attraction, and infi-
delity. Eddie M. Clark is a Professor of psychology at Saint Louis University.
He is a Social Psychologist with interests in close relationships and heath
attitudes/persuasion. Brent A. Mattingly is an Assistant Professor of psychol-
ogy at Ashland University, and his research primarily focuses on relationships,
motivation, and the self.
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.
Busboom, A. L., Collins, D. M., Givertz, M. D., & Levin, L. A. (2002). Can we still
be friends? Resources and barriers to friendship quality after romantic relationship
dissolution. Personal Relationships,9, 215–223.
Canary, D. J., & Stafford, L. (2001). Equity in the preservation of personal relationships.
In J. Harvey & A. Wenzel (Eds.), Close Romantic Relationships: Maintenance and
Enhancement (pp. 133–151). Mahwah, NJ: Lawrence Erlbaum.
Hackathorn, J., Clark, E. M., Mattingly, B. A., Bullock, M., & Weaver, K. (2008, May).
Breaking up might be hard to do: Interpersonal predictors of post-romantic friendships.
Poster presented at the annual meeting of the Midwestern Psychological Association,
Chicago, IL.
Hebert, S., & Popadiuk, N. (2008). University students’ experiences of nonmarital
breakups: A grounded theory. Journal of College Student Development,49, 1–14.
Karney, B. R., & Coombs, R. H. (2000). Memory bias in long-term close relation-
ships: Consistency or improvement? Personality and Social Psychology Bulletin,26,
Downloaded by [Saint Louis University] at 06:41 30 August 2011
666 The Journal of Social Psychology
Metts, S., Cupach, W. R., & Bejlovec, R. A. (1989). “I love you too much to ever start liking
you”: Redefining romantic relationships. Journal of Social and Personal Relationships,
6, 259–274.
Oswald, D. L., Clark, E. M., & Kelly, C. M. (2004). Friendship maintenance: An anal-
ysis of individual and dyad behaviors. Journal of Social and Clinical Psychology,23,
Popenoe, D., & Whitehead, B. D. (2004). The state of our unions, 2004.Piscataway, NJ:
The National Marriage Project.
Ritchie, T. D., & Skowronski, J. J. (2008). Perceived change in the affect associated with
dreams: The fading affect bias and its moderators. Dreaming,18, 27–43.
Rusbult, C. E., Martz, J. M., & Agnew, C. R. (1998). The Investment Model scale:
Measuring commitment level, satisfaction level, quality of alternatives, and investment
size. Personal Relationships,5, 357–391.
Stafford, L., & Canary, D. J. (1991). Maintenance strategies and romantic relationship type,
gender and relational characteristics. Journal of Social and Personal Relationships,8,
Tabachnick, B. G., & Fidell, L. S. (2007). Using multivariate statistics (5th ed.). Boston,
MA: Pearson.
Wilmot, W. W., Carbaugh, D. A., & Baxter, L. A. (1985). Communicative strategies used
to terminate romantic relationships. Western Journal of Speech Communication,49,
Received January 14, 2010
Accepted July 8, 2010
Downloaded by [Saint Louis University] at 06:41 30 August 2011
... For example, Wilmot et al. (1985) found that the majority of people in their study (61%) reported being friends with a former romantic partner. Similarly, Bullock, Hackathorn, Clark, and Mattingly (2011) found that about 60% of their participants reported being friends with an ex-partner. The fact that PDFs are so common emphasizes the need to better understand them, their predictors, and outcomes. ...
... Furthermore, the authors suggested people can draw on that established schema to defuse any uncertainty in the new postdating friendship. Bullock et al. (2011) studied romantic relationship quality using the Investment Model (Rusbult, Drigotas, & Verette, 1994) as a potential predictor of PDFs. They found that satisfaction in the romantic relationship predicted postdating friendship as well as friendship maintenance behaviors, such as joint activities and self-disclosure. ...
... In the current studies, we focused on the aftermath of romantic relationship breakups, specifically the friendships people form with their ex-romantic partners, referred to as PDFs (Tan et al., 2015). PDFs are a common phenomenon that characterizes the majority of people who have experienced romantic relationships (Bullock et al., 2011;Schneider & Kenny, 2000). PDFs are thought to be different from other friendships and serve different functions (Schneider & Kenny, 2000). ...
Full-text available
In two studies, we identified predictors of staying friends with an ex-partner, developed and validated a measure for reasons to stay friends with exes, and examined the outcomes of staying friends. We identified four reasons to stay friends with exes: Security, Practical, Civility, and Unresolved Romantic Desires. Sex, sexual orientation, attachment styles, personality traits, time since breakup, and reasons for breakup all predicted staying friends. Staying friends due to Unresolved Romantic Desires resulted in negative outcomes, whereas staying friends due to Security and Practical reasons resulted in more positive outcomes. Staying friends due to Practical or Civility reasons resulted in the friendship being less likely to last in the long term. Implications for future research and clinical work are discussed.
... Scholars have pointed out that friendship maintenance may change the status of a relationship from romantic to friendship (Oswald, Clark, Kelly, & Oswald, 2004). Bullock, Hackathorn, Clark, and Mattingly (2011) found that individuals who are satisfied during the dissolved romance are more likely to remain friends with their ex-partners. In the marketing field, scholars have argued that the equivalent to remaining friends with a former brand partner would be to feel appreciative of the brand, even though the consumer no longer uses the brand after switching to another one (Appiah et al., 2019). ...
The phenomenon of consumers developing relationships with brands is well documented. However, few researchers have addressed the brand relationship breakup process. Since consumers choose to break up with brands for several reasons, the purpose of this study is to better understand brand relationship breakups and reconciliation through the development and validation of a measure of the valence of the brand relationship breakup (i.e., how positively, negatively, or neutrally the consumer perceives the brand relationship breakup). The findings indicate that a more positive brand relationship breakup makes consumers more prone to reconcile with the brand and less likely to spread negative word of mouth. In addition, consumer desire for vengeance mediates this relationship. Finally, this paper explores how the valence of a brand relationship breakup influences the type of recovery strategies a company should use to reestablish the brand relationship with the consumer.
... They found, for example, that being the rejected party is more functionally costly than being the rejecter, that women are more likely than men to report experiencing ''costly'' stalking by their ex-partners, and that rejecters experienced social costs such as being perceived as mean. In addition to this research, several groups have focused on the predictors of friendship after relationship dissolution (e.g., Bullock, Hackathorn, Clark, & Mattingly, 2011;Metts, Cupach, & Bejlovec, 1989;Tan, Agnew, Vanderdrift, & Harvey, 2014). Most recently, Mogilski and Welling (in press) examined the reasons people provided for retaining friendships with ex-partners. ...
Full-text available
We argue that mate rejection and ex-partner relationships are important, multifaceted topics that have been underresearched in social and evolutionary psychology. Mate rejection and relationship dissolution are ubiquitous and form integral parts of the human experience. Both also carry with them potential risks and benefits to our fitness and survival. Hence, we expect that mate rejection would have given rise to evolved behavioral and psychological adaptations. Herein, we outline some of the many unanswered questions in evolutionary psychology on these topics, at each step presenting novel hypotheses about how men and women should behave when rejecting a mate or potential mate or in response to rejection. We intend these hypotheses and suggestions for future research to be used as a basis for enriching our understanding of human mating from an evolutionary perspective.
The formation and functioning of close relationships can alter individuals’ self-concepts in such a manner that the self-concepts are cognitively linked with the partner; however, relationship dissolution directly threatens the loss of this intertwined self-concept. In this chapter, we first discuss the degree to which prior relationship-induced self-concept change predicts, and in some cases inoculates against, dissolution. Second, we discuss the extent to which relationship dissolution leads to subsequent self-concept changes within individuals (e.g., loss of self-concept content, increased self-concept confusion), as well as the mechanisms underlying whether these post-dissolution self-concept changes are deemed harmful versus beneficial. Third, we explain how individuals may recover from post-dissolution self-concept changes by seeking to repair or restore the damaged self-concept. Finally, we briefly identify avenues for future research that scholars may consider pursuing.
Cambridge Core - Social Psychology - On-Again, Off-Again Relationships - by René M. Dailey
This study tested a moderated mediation model that commitment prior to an unrequited love episode will be related to higher levels of friendship maintenance behaviors after the episode and that this relationship will be mediated by the individual’s motivations to remain friends with the rejecter. We predicted that rejection distress would weaken the mediational model among those reporting high distress. Participants wrote about an unrequited love experience as a pursuer and completed measures of pre-unrequited love commitment, rejection distress, motivations to remain friends, and friendship maintenance behaviors. Our results confirmed the moderated mediation model when the motivations to remain friends measure overall score, the interpersonal connection motivation, or the social connections motivation was the mediator. When covariates were added, only the model with the overall motivations score as mediator was supported. These data can help laypersons and mental health professionals understand and potentially repair friendships following an unrequited love episode.
Full-text available
The study queried whether the relational characteristics that influence individuals to remain face-to-face friends with former romantic partners following a break up also impact the decisions to remain Facebook “friends” with former romantic partners. The sample included over 300 young adults who met two criteria: They maintained an active Facebook account and reported a pre-marital, romantic break-up. The results revealed that the variables that impact post-dissolution friendship decisions of former romantic partners in the face-to-face context (quantity of relational investments, relational satisfaction, and relational disengagement strategies) do not impact former romantic partners' decisions to maintain or dissolve Facebook friendships. These results provide evidence that romantic partners may experience different relational motivations and dynamics in online versus off-line venues.
Prior research examining maintenance in cross-sex friendships focuses heavily on platonic (i.e., nonsexually active) friendships with limited research examining sexually involved cross-sex friendships (i.e., “friends with benefits relationships”). In this study, we investigated differences in relational maintenance behaviors between sexually and nonsexually active cross-sex friendships types. In an online survey, 531 emerging adult participants from large southwestern and southeastern universities identified either a friends with benefits or platonic opposite sex friendship and then completed items asking them to report the frequency with which they enacted each of 36 relationship maintenance behaviors. Overall, participants involved in casual sex friendships engaged in the least, and those who transitioned from a friends with benefits relationship to a romantic relationship engaged in the most frequent relationship maintenance. Platonic friendships employed more frequent maintenance than casual sex friendships but less than either true friends with benefits or participants who transitioned to a romantic relationship.
Romantic dissolution is a common experience throughout the life course, particularly during emerging adulthood (ages 18-29). The purpose of this review was to summarize and critique theoretical approaches and empirical findings of the aftermath of dating relationship dissolution. Article searches were conducted within PsycINFO. We utilized terms related to romantic relationship dissolution (i.e., breakup, romantic breakup, relationship termination, relationship dissolution, romantic dissolution, romantic termination, post-dissolution) in a search for key words. We narrowed the results further by limiting the search to include participants between the ages of 18 and 29. Experiencing romantic dissolution can result in both positive and negative emotional reactions and behaviors, including personal growth and self-expansion as well as experiencing physical and emotional abuse from ex-partners. Furthermore, former romantic partners commonly form other types of friendships and casual sexual relationships after the termination of committed romantic relationships. Many theoretical frameworks are used to guide these investigations, and some articles lack a theoretical framework. Developmental Systems Theory might be a theoretical framework that best shapes our investigations of romantic dissolution in dating relationships that occur in emerging adulthood. Relationship education programs would be enhanced by discussing the developmental needs that are important for young people and the ways in which their romantic experiences can or cannot meet those needs. In addition to learning about how to have healthy romantic relationships, young people can also benefit from learning how to identify when romantic relationships should end, and how to end them successfully.
Full-text available
In this article, we attempt to distinguish between the properties of moderator and mediator variables at a number of levels. First, we seek to make theorists and researchers aware of the importance of not using the terms moderator and mediator interchangeably by carefully elaborating, both conceptually and strategically, the many ways in which moderators and mediators differ. We then go beyond this largely pedagogical function and delineate the conceptual and strategic implications of making use of such distinctions with regard to a wide range of phenomena, including control and stress, attitudes, and personality traits. We also provide a specific compendium of analytic procedures appropriate for making the most effective use of the moderator and mediator distinction, both separately and in terms of a broader causal system that includes both moderators and mediators. (46 ref) (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Full-text available
In three studies, friendship maintenance behaviors were investigated. In the first study, apsychometrically valid measure of friendship maintenance behaviors was developed via exploratory and confirmatory factor analyses. Friendship maintenance behaviors of positivity, supportiveness, openness, and interaction were identified as key factors. Furthermore, maintenance behaviors differed by friendship status (best, close, or casual friends) and gender. Best friends engaged in more maintenance behaviors than close or casual friends. Gender differences in maintenance behaviors were found, such that female same-sex friendships reported using more supportiveness, openness, and interaction than male same-sex friendships. In the second study, a confirmatory factor analysis revalidated the four-factor scale structure. In the third study, the interdependence of maintenance behaviors among friendship dyads was examined. Results indicated that friendships are equitable, display self-other agreement, and perceive equity of maintenance behaviors. Furthermore, both individuals' self-reported behaviors predicted dyad level satisfaction and commitment. Areas for future research are identified and clinical implications are discussed.
The purpose of this study was to determine the factors affecting the likelihood that a disengaged romantic relationship will be redefined as a friendship rather than completely terminated. College students were surveyed about a recalled romantic relationship that either broke up or evolved into a friendship. Respondents answered questions regarding characteristics of the relationship prior to its decline, strategies used to disengage the relationship, and feelings during the disengaging period. Results indicated that being friends prior to romantic involvement was a significant predictor of friendship, both for people who initiated the disengagement and for those who were recipients of partner's desire for disengagement. In addition, for those who initiated disengagement, withdrawal strategies and feelings of being taken advantage of in the relationship were negatively related to post-disengagement friendship. For persons who were recipients of partner's desire for disengagement, positive tone strategies were positively related to post-disengagement friendship while manipulation strategies were negatively related to post-disengagement friendship.
This study examines the manner in which perceptions of relational maintenance strategies used in romantic dyads vary according to relationship type (married, engaged, seriously dating and dating) and gender. Additionally, this study investigates how perceptions of partners' maintenance behaviors differentially affect the relational characteristics of control mutuality, commitment, liking and satisfaction. Research assumptions were cast within a developmental framework. Five maintenance strategies were derived through factor analyses: positivity, assurances, openness, sharing tasks and social networks. Results indicate that relationship type moderately affected perceptions of partner maintenance strategies and gender weakly affected perceptions of maintenance behaviors. The findings also reveal that positivity, assurances and sharing tasks were consistent and strong predictors of control mutuality, commitment, liking and satisfaction.
How do partners in long-term relationships construct memories of the past? The current study examined 20 years of retrospective and longitudinal data from a sample of wives to evaluate two possible answers to this question. Findings indicate that wives invoke different theories of the past at different stages of life. At 10 years into the study, wives’ memories of the past were negatively biased, such that present ratings seemed a significant improvement. At 20 years into the study, wives’ memories of the past continued to be negatively biased; however, at this stage, recollections of the past resembled current perceptions. Across both intervals, concurrent ratings in fact declined significantly. Longitudinal analyses revealed that the degree of bias in wives’ memories at Time 2 predicted the course of their marital satisfaction over the subsequent 10 years. These results support the view that memory bias may be a mechanism of maintaining satisfaction in long-term relationships.
Prior nonmarital breakup research has been focused on negative outcomes, rarely examining the personal growth aspects of this experience. In this study, we used a qualitative grounded theory methodology to explore the changes that university students reported experiencing as a result of a heterosexual nonmarital breakup and how those changes arose. From the participant accounts, we constructed a grounded theoretical model that conceptualizes and explains change and personal growth following nonmarital breakups. In this model, we assert that students may experience stress-related growth after a breakup and we provide a framework for explaining these changes. Implications for counseling are discussed.
Most data‐based research on relationship dissolution has been limited to hypothetical relationships or those in the distant past. This longitudinal study focused on long‐distance romantic relationships that terminated within a 15‐month time frame. Subjects provided reports of the termination strategies they and their partners used. We found (1) symmetrical use of verbal indirectness, (2) directness/indirectness as a central dimension of strategy use, (3) dissolution choices uninfluenced by who initiated the dissolution, (4) mutual terminations producing fewer negative emotional reactions than one‐sided initiations, and (5) verbal directness used to terminate expressed regrets about the choice. The findings are discussed in terms of extant theory and research on the termination process.