social psychology

THE JOURNAL OF

SOCIAL PSYCHOLOGY Volume 155, Number 5 September–October 2015 Special Issue: Investigating How Individuals Feel Ostracizing Others Guest Editors: James H. Wirth and Eric D. Wesselmann INTRODUCTION Commentary on Sources of Ostracism Research Jon E. Grahe

403

ARTICLES Ostracizing for a Reason: A Novel Source Paradigm for Examining the Nature and Consequences of Motivated Ostracism Sarah L. Gooley, Lisa Zadro, Lisa A. Williams, Elena Svetieva, and Karen Gonsalkorale

410

Ostracism in Everyday Life: The Effects of Ostracism on Those Who Ostracize John B. Nezlek, Eric D. Wesselmann, Ladd Wheeler, and Kipling D. Williams Who Would Do That? A Theory-Based Analysis of Narratives of Sources of Family Ostracism Joan R. Poulsen and Anna F. Carmon

432

452

Righting the Wrong: Reparative Coping After Going Along With Ostracism Nicole Legate, Cody DeHaan, and Richard Ryan

471

The Role of Burden and Deviation in Ostracizing Others Eric D. Wesselmann, James H. Wirth, John B. Pryor, Glenn D. Reeder, and Kipling D. Williams

483

Atimia: A New Paradigm for Investigating How Individuals Feel When Ostracizing Others James H. Wirth, Michael J. Bernstein, and Angie S. LeRoy

497

How Does the Source of Rejection Perceive Innocent Victims? Kai-Tak Poon and Zhansheng Chen

515

Insufficient Justification for Exclusion Prompts Compensatory Behavior Daryl R. Van Tongeren, Lindsey M. Root Luna, and Charlotte VanOyen Witvliet

527

This article was downloaded by: [Ohio State University Libraries] On: 20 August 2015, At: 08:46 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: 5 Howick Place, London, SW1P 1WG

The Journal of Social Psychology Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/vsoc20

Commentary on Sources of Ostracism Research a

Jon E. Grahe a

Pacific Lutheran University Published online: 12 Aug 2015.

Click for updates To cite this article: Jon E. Grahe (2015) Commentary on Sources of Ostracism Research, The Journal of Social Psychology, 155:5, 403-409, DOI: 10.1080/00224545.2015.1064069 To link to this article: http://dx.doi.org/10.1080/00224545.2015.1064069

PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &

Downloaded by [Ohio State University Libraries] at 08:46 20 August 2015

Conditions of access and use can be found at http://www.tandfonline.com/page/termsand-conditions

The Journal of Social Psychology, 155: 403–409, 2015 Copyright © Taylor & Francis Group, LLC ISSN: 0022-4545 print / 1940-1183 online DOI: 10.1080/00224545.2015.1064069


INTRODUCTION

Commentary on Sources of Ostracism Research JON E. GRAHE Pacific Lutheran University

ABSTRACT. This article briefly introduces and provides commentary on this special issue, “Investigating How Individuals Feel Ostracizing Others” in the Journal of Social Psychology. This commentary uses first-person recollections from early ostracism studies to help frame the special issue in the larger scope of ostracism research. Modern ostracism research started in the early 1990s and hundreds of studies have advanced our understanding of this phenomenon. However, the preponderance of research has focused on the target of ostracism (those being ostracized) and relied primarily on a single method in experimental studies. The present special issue includes articles that employed a variety of research approaches and focused exclusively on the sources (those who do the ostracism). Finally, this commentary invites researchers to more fully investigate this understudied aspect of a common phenomenon that people are likely to engage in regularly. Keywords: sources of ostracism

TO INTRODUCE THIS SPECIAL ISSUE, imagine this situation: Three young men are sitting in a room waiting for the experimenter to get some problem sorted out. One of them finds a racquetball ball in a pile of toys and throws it to one of the other participants who looks it over and then throws the ball to the third participant. The ball is subsequently bounced repeatedly between everyone. This enjoyable situation abruptly changes when two of them start throwing it exclusively only to each other. The third young man sits and watches for four minutes as the other two continue tossing the ball and smiling at each other while not making any eye contact or otherwise recognizing that he existed. After this painful experience finally ended, the group moved to another space where they listed uses of a knife for five minutes. Depending on condition, the experimenter explained that either their output would be counted as a group (collective condition) or individually (coactive condition) in order to induce social loafing (Karau & Williams, 1993). This situation was part of the “ostraloaf” study, as it was labeled during lab meetings, and I was Address correspondence to Jon E. Grahe, Pacific Lutheran University, Department of Psychology, 1010 S. 121st St., Tacoma, WA 98447, USA. E-mail: [email protected]


404

THE JOURNAL OF SOCIAL PSYCHOLOGY

one of those confederates. In the ostraloaf study, Williams and Sommer (1997) demonstrated that women, but not men, compensated after being ostracized by not loafing. However, all of the participants reacted negatively to the experience. Being a confederate in that study as a first-year graduate student introduced me to the systematic study of ostracism1 where I experienced the early development of Williams’ model of ostracism (Williams, 1997, 2001). My confederate partner and I started contributing to the ostraloaf study in the fall of 1992 and we were one pair of male confederates (18 confederates total) in a study conducted over a three-year period. Because of the intricate protocols, each trial required over 30 minutes, a substantial portion of lab time. That early study (the first ostracism experiment conducted by Kipling Williams) provided a powerful demonstration of the basic psychological experience of ostracism, but future research needed methods that did not require live confederates had control over the idiosyncrasies of specific confederates, and could be employed by different research labs to encourage large-scale replication and extensions. It is quite possible the reader already recognizes that Cyberball (Williams, Cheung, & Choi, 2000) replaced this face-to-face method. Cyberball allowed Williams and an ever-expanding group of researchers to examine targets of ostracism across a wide swath of conditions to better understand the effects and recovery from this ubiquitous life experience; there are well over 100 published studies using this method (a list of Cyberball publications is available at Kipling Williams’ website). Cyberball researchers documented many negative emotional and physiological responses and even some negative behavioral intentions directed toward those cruel computer animations (see van Beest, Hartgerink, Wicherts, & Williams, 2015, for review). However, until recently when Wesselmann, Wirth, Pryor, Reeder, and Williams (2013) altered the program, Cyberball research had largely ignored the sources of ostracism. Now with Cyberball redesigned, researchers can use this highly controlled situation to examine further research questions such as those investigated by three manuscripts in this special issue (Gooley, Zadro, Williams, Svetieva, & Gonsalkorale, 2015; Legate, DeHaan, & Ryan, 2015; Wesselmann, Wirth, Pryor, Reeder, & Williams, 2015). Cyberball ostracizers have it easy because a computer does not see, nor understand the effects on, the target and a computer does not experience empathy (not yet, anyway). Ostraloaf confederates acting as sources for Williams and Sommer (1997) had more opportunity to suffer. In that experiment, I regularly hoped it would be an inclusion condition. Three boys throwing a ball for 5 minutes is generally a pleasurable experience. However, when it was an ostracism condition, all the fun left the room. We did laugh and chuckle, sometimes at the expense of the target, but we also felt guilt and apprehension. If the ostraloaf experiment were to be replicated today, researchers should measure any changes in confederates’ feelings toward the target or effort at the subsequent loafing task. Bastian et al. (2013) used such a ball throwing paradigm and repeatedly demonstrated that sources “self-dehumanized” as a result of seeing their behavior as immoral. Similar feelings were a regular part of my confederate experience. Without research into different source motivations, it is not clear how to generalize these responses. Zadro and Gonsalkorale (2014) provided a breakthrough theoretical framework of sourceparadigm research to further document situational impacts on ostracism from a source perspective. Zadro and Gonsalkorale (2014) described recall and direct-, induced-, and motivated-source inductions2 . Following this framework, early attempts to examine the psychological experiences of sources used recall paradigms (Williams, Shore, & Grahe, 1998; Williams, Wheeler, & Harvey, 2001), direct-source inductions (Williams, Bernieri, Faulkner, Gada-Jain, & Grahe,


GRAHE

405

2000; Zadro, Williams, & Richardson, 2005), or induced-source paradigms (Ciarocco, Sommer, & Baumeister, 2001; Grahe, Williams, & Escamilla, 1998)3 . Zadro and Gonsalkorale (2014) thoroughly reviewed that early work and this special issue includes all but one of these methods. Zadro and Gonsalkorale’s framework augments existing ostracism theory (Williams, 1997, 2001, 2007, 2009) because the ostracism types (role-prescribed, punitive, defensive, or oblivious, which are defined by multiple articles in this issue) better describe target attributions about the experience rather than the sources’ motivations. Consider some examples; sources in a poor family relationship might engage in punitive or defensive ostracism (see Poulsen & Carmon, 2015). Other situations are yet to be studied. For instance, hypothetically, workers might be required to avoid a colleague for punitive reasons that they may or may not endorse (i.e., due to a restraining order or other legal proceedings). In this case, this type of ostracism is both punitive and role prescribed, even if the sources and target are friends. As such, future ostracism theory should disentangle source motive from the target’s perception of that motive. Briefly consider the current issue through the lens of Zadro and Gonsalkorale’s (2014) framework. First, the recall studies in the current issue (Nezlek et al., 2015; Poulsen & Carmon, 2015; Poon & Chen, 2015) reflect three distinct approaches to understanding the source experience. They include an intensive diary study (Nezlek et al., 2015), a deep recollection of a critical incident (Poulsen & Carmon, 2015), and the use of recall as a manipulation (Poon & Chen, 2015). The direct-source inductions (Legate et al., 2015; Tongeren et al., 2014) examine how sources respond to acts of ostracism or exclusion later in subsequent interactions. The one type that is not included here is induced-source manipulation. This might likely reflect the progress of ostracism research from artificial scenarios to a new, more ecologically valid approach where sources are motivated to ostracize others in sophisticated scenarios (Gooley, et al., 2015; Wesselmann, et al., 2015; Wirth et al., 2015). Altogether, these studies examine the psychological response to metaphorically delivering the scarlet letter “O” for ostracism to strangers, friends, family, and enemies. SCARLET LETTER INCIDENTS In 1997, the scarlet letter O was not metaphorical. On a given day, during a certain week, a member of the social psychology program in a medium-sized Midwestern University would find a red “O” hanging over their office door. This signaled everyone to ostracize this person. Williams et al. (2000) acted as participant observers documenting incidents and responses to ostracism. It was light-hearted ostracism because we knew the cause and the ostracism followed random chance. Yet, it still caused ambiguous attributions and distress for both targets and sources. My day to be target was Wednesday; I provoked ostracism regularly and I remember feeling nervous after seeing and hearing my dissertation advisor’s door slam after asking for feedback on a paper when it was not really necessary. Also, as I source, I also did not enjoy ostracizing him on a day that seemed long even though he left early. But with humor, I recall the day when we punitively ostracized the other professor, the protagonist of modern research in social ostracism. That day was more fun and the attribution process was clearer because it was his idea. The scarlet letter study was quite informal and qualitative, yet it revealed unconscious impacts of ostracism. Different motivations, whether justified or random, should yield distinct positive or negative outcomes. Multiple articles in this special issue (Gooley et al., 2015; Nezlek et al., 2015; Poulsen &


406


Carmon, 2015; Van Tongeren et al., 2015) continue to suggest that when sources feel justified, their responses are more positive than when they are directed or induced to ostracize (Bastian et al., 2013; Legate, DeHaan, Weinstein, & Ryan, 2013; Sommer & Yoon, 2013). Early theoretical models (Williams, 1997, 2001) examined ostracism effects on four basic needs (belongingness, control, meaningful existence, and self-esteem) that were intended to respond differentially depending on circumstance. However, most target research has failed to reveal response differences across these needs and recently they are regularly combined into a single needs threat construct with high reliability (vanBeest et al., 2015; Williams, 2009). However, multiple articles in this issue (Gooley et al., 2015; Nezlek et al., 2015) suggest that motivation type impacts belonging and control more than the other needs with positive effects when the motivation for ostracizing is justified. Further, Poulsen and Carmon (2015) suggested that self-esteem is related to the motives for ostracizing family members. In sum, while targets may experience ubiquitous needs threat when being ostracized, sources seem to experience differential benefits and costs depending on the type and target of ostracism. Researchers should consider employing these novel paradigms that motivate individuals to become sources by either introducing an unlikeable (Gooley et al., 2015; Sommer & Yoon, 2013) or a burdensome (Wesselmann et al., 2015; Wirth et al., 2015) group member.

IS CYBERBALL THE SAME AS CYBER-OSTRACISM? The bulk of articles in this issue using experimentally induced ostracism employed Cyberball (Legate et al., 2015; Gooley et al., 2015; Wesselmann et al., 2015). To help balance this Cyberball barrage, Wirth et al. (2015) demonstrated their own motivated sources with the Atimia paradigm (conceptually similar to the slow, burdensome computer player used by Wesselmann et al., 2015; see also Wesselmann et al., 2013). Finally, Van Tongeren et al. (2015) compared sources’ compensation behaviors as group leaders after excluding a group member when the participant was responsible for the ostracism or not. These studies provide strong experimental tests of their hypotheses, but there needs to be more face to face, in vivo experiments such as Legate et al. (2015). Although some in vivo experiments exist in the literature (Bastian et al., 2013; Ciarocco et al., 2001; Poulsen & Kashy, 2012; Sommer & Yoon, 2013), they are rare and employ direct source inducements. While Cyberball might be internally and externally valid, it is not all encompassing, and more naturalistic research is warranted. The presence of three studies that employed Cyberball in this special issue accurately reflects the status of this method in the larger milieu of social ostracism research. Cyberball was readily adopted partly because it reproduces the basic ostracism effect across laboratories and conceptual manipulations as those ghosts in the machine have masqueraded as friends, strangers, and enemies (van Beest et al., 2015). However, whereas Cyberball is a type of cyber-ostracism because it does occur on a computer or online, there are many other forms of cyber-ostracism that exist in our increasingly online and networked lives. While Wirth et al. (2015) provided another cyber-ostracism experience, it is still a game format. Some have studied other modalities. Examples include studies of target experiences of ostracism in chat rooms (Williams et al., 2002) or while texting (Smith & Williams, 2004) and most recently on Facebook (Tobin, Vanman, Verreynne, & Saeri, 2015). These studies did not provide compelling evidence to distinguish Cyberball (or game-based ostracism) from other types of online ostracism, in agreement with


GRAHE

407

what Kassner, Wesselmann, Law, and Williams (2012) concluded—that the target experience in Cyberball is similar to that of other ostracism modalities. However, this may not be true for sources. The collective findings from this special issue and other research on sources (Bastian et al., 2013; Zadro & Gonsalkorale, 2014) suggest that different motivations elicit different interpersonal responses. Therefore, ostracism should be further examined in the multiple modalities of cyberspace. In accordance with the call from Zadro and Gonsalkorale (2014), this research should further examine the situation from both the source and target perspective in order to verify the ecological validity of present conclusions. Such research would also establish any effects of nonindependence on ostracism that emerges from groups (Kenny & Judd, 1986). In this commentary, I have highlighted some personal recollections of being a source of ostracism to help place this special issue within the larger lens of social ostracism research that has focused overwhelmingly on the targets’ experience. In short, there have been hundreds of studies conducted since that early ostraloaf experiment (Williams & Sommer, 1997), but until recently only a few studies investigated the source. In addition to advancing our theoretical understanding of the experience of ostracizing others, this special issue highlights the current state of research on sources by introducing new protocols (Gooley et al., 2015; Wirth et al., 2015; VanTongren et al., 2015), further investigations into recently designed methods (Legate et al., 2015; Wesselmann et al., 2015; Poon & Chen, 2015), and adaptations to established protocols that answer new questions (Nezlek et al., 2015; Poulsen & Carmon, 2015). Yet, this special issue also identifies the need for further investigation. Ideally future studies should manipulate source induction (direct vs. motivated) under both online and in vivo conditions where both targets and sources are participants. Their behaviors should be measured in addition to their reported experiences of the basic needs and emotions. Finally, studies should continue to measure relevant individual differences. We invite readers to replicate the studies presented here and to work collaboratively to develop novel paradigms to further test the hypotheses inherent in our conclusions. To that end, several articles in this special issue have shared their materials on Open Science Framework (OSF) websites. Further, we have created an OSF page for sources of ostracism to provide a central location for readers to obtain these materials and to coordinate future efforts (Wesselmann, Wirth, & Grahe, 2015; https://osf.io/env5w/).

ACKNOWLEDGMENT I would like to thank Katie Coddington for providing comments on this article.

NOTES 1. We were careful in this special issue to delineate between ostracism, exclusion, and rejection (Williams, 2007). While some use these terms interchangeably and they often have similar negative outcomes, Williams (2009) has argued that there may be situations in which these different phenomena have important empirical nuances (see also Smart Richman & Leary, 2009). Thus, we feel that best practice is for researchers to pick the term that best fits a given manipulation and to be clear that they know there are differences but are simply being parsimonious because their theorizing does not require those distinctions. 2. Zadro and Gonsalkorale (2014) defined recall inductions as those where participants report on an ostracism incident that occurred previously, direct-source inductions as those where participations are instructed to ostracize another,

408


induced-source inductions as those where participants have a choice to either ostracize or include but tell the participant that they need ostracizers, and motivated-source inductions as those where participants experience a situation where participants naturally ostracize another because the target seems to deserve it. 3. We induced two participants in discussion groups to ostracize a third participant during a conversation about sensitive topics. When I changed institutions, promising effects were not replicating and we abandoned the study due to limited resources and a subject pool that had recurring evidence of cross-talk during debriefing. We offer our materials and poster slides on an Open Science Framework project page since the paper was never published (Grahe & Coddington, 2015).


AUTHOR NOTE Jon E. Grahe is Professor of Psychology at Pacific Lutheran University in Tacoma, Washington, President-Elect of Psi Chi, and Executive Editor of the Journal of Social Psychology.

REFERENCES Bastian, B., Jetten, J., Chen, H., Radke, H. R. M., Harding, J. F., & Fasoli, F. (2013). Losing our humanity: The self-dehumanizing consequences of social ostracism. Personality and Social Psychology Bulletin, 39, 156–169. doi:10.1177/0146167212471205 Ciarocco, N. J., Sommer, K. L., & Baumeister, R. F. (2001). Ostracism and ego depletion: The strains of silence. Personality and Social Psychology Bulletin, 27, 1156–1163. doi:10.1177/0146167201279008 Gooley, S. L., Zadro, L. A., Williams, L. A., Svetieva, E., & Gonsalkorale, K. (2015). Ostracizing for a reason: A novel source paradigm for examining the nature and consequences of motivated ostracism. Journal of Social Psychology, 155(5), 410–431. Grahe, J. E., & Coddington, K. (2015, June 3). Conversation paradigms and cyberostracism. Retrieved from osf.io/hwy8v Grahe, J. E., Williams, K. D., & Escamilla, N. A. (1998, May). Ostracism: A conversation paradigm to study sources and targets. Paper presented at Midwestern Psychology Association, Chicago, IL. Karau, S. J., & Williams, K. D. (1993). Social loafing: A meta-analytic review and theoretical integration. Journal of Personality and Social Psychology, 65, 681–706. doi:10.1037/0022-3514.65.4.681 Kassner, M. P., Wesselmann, E. D., Law, A. T., & Williams, K. D. (2012). Virtually ostracized: Studying ostracism in immersive virtual environments. Cyberpsychology, Behavior, and Social Networking, 15, 399–403. doi:10.1089/ cyber.2012.0113 Kenny, D. A., & Judd, C. M. (1986). Consequences of violating the independence assumption in analysis of variance. Psychological Bulletin, 99, 422–431. doi:10.1037/0033-2909.99.3.422 Legate, N., DeHaan, C., Weinstein, N., & Ryan, R. (2013). Hurting you hurts me too: The psychological costs of complying with ostracism. Psychological Science, 24, 583–588. doi:10.1177/0956797612457951 Legate, N., DeHaan, C., & Ryan, R. (2015). Righting the wrong: Reparative coping after going along with ostracism. Journal of Social Psychology, 155(5), 471–482. Nezlek, J. B., Wesselmann, E. D., Wheeler, L., & Williams, K. D. (2015). Ostracism in everyday life: The effects of ostracism on those who ostracize. Journal of Social Psychology, 155(5), 432–451. Poon, K.-T., & Chen, Z. (2015). How does the source of rejection perceive innocent victims? Journal of Social Psychology, 155(5), 515–526. Poulsen, J. R., & Carmon, A. F. (2015). Who would do that? A theory-based analysis of narratives of sources of family ostracism. Journal of Social Psychology, 155(5), 452–470. Poulsen, J. R., & Kashy, D. A. (2012). Two sides of the ostracism coin: How sources and targets of social exclusion perceive themselves and one another. Group Processes & Intergroup Relations, 15, 457–470. doi:10.1177/ 1368430211430517 Smart Richman, L., & Leary, M. (2009). Reactions to discrimination, stigmatization, ostracism, and other forms of interpersonal rejection: A dynamic, multi-motive model. Psychological Review, 116, 365–383. Smith, A., & Williams, K. D. (2004). R U there? Ostracism by cell phone text messages. Group Dynamics: Theory, Research, and Practice, 8, 291–301. doi:10.1037/1089-2699.8.4.291


GRAHE

409

Sommer, K., & Yoon, J. (2013). When silence is golden: Ego depletion following aversive social interactions. Journal of Personal and Social Relationships, 30, 901–919. doi:10.1177/0265407512473006 Tobin, S. J., Vanman, E. J., Verreynne, M., & Saeri, A. K. (2015). Threats to belonging on Facebook: Lurking and ostracism. Social Influence, 10, 31–42. van Beest, I., Hartgerink, C., Wicherts, J., & Williams, K. D. (2015). Ordinal effects of ostracism: A meta-analysis of Cyberball studies. PLoSONE, 10(5), e0127002. doi:10.1371/journal.pone.0127002 Van Tongeren, D. R., Root Luna, L. M., & VanOyen Witvliet, C. (2015). Insufficient justification for exclusion prompts compensatory behavior. Journal of Social Psychology, 155(5), 527–534. Wesselmann, E., Wirth, J., Pryor, J., Reeder, G., & Williams, K. (2013). When do we ostracize? Social Psychological and Personality Science, 4, 108–115. doi:10.1177/1948550612443386 Wesselmann, E., Wirth, J. H., & Grahe, J. E. (2015, June 3). “Sources of ostracism” hub. Retrieved from osf.io/env5w Wesselmann, E. D., Wirth, J. H., Pryor, J. B., Reeder, G. D., & Williams, K. D. (2015). The role of burden and deviation in ostracizing others. Journal of Social Psychology, 155(5), 483–496. Williams, K. D. (1997). Social ostracism. In R. M. Kowlalski (Ed.), Aversive interpersonal behaviors (pp. 133–170). New York, NY: Plenum. Williams, K. D. (2001). Ostracism: The power of silence. New York, NY: Guilford Press. Williams, K. D. (2007). Ostracism. Annual Review of Psychology, 58, 425–452. doi:10.1146/annurev.psych.58.110405. 085641 Williams, K. D. (2009). Ostracism: Effects of being excluded and ignored. In M. Zanna (Ed.), Advances in experimental social psychology (pp. 275–314). New York, NY: Academic Press. Williams, K. D., Bernieri, F. J., Faulkner, S. L., Gada-Jain, N., & Grahe, J. E. (2000). The scarlet letter study: Five days of social ostracism. Journal of Personal & Interpersonal Loss, 5, 19–63. doi:10.1080/10811440008407846 Williams, K. D., Cheung, C., & Choi, W. (2000). Cyberostracism: Effects of being ignored over the Internet. Journal of Personality and Social Psychology, 79, 748–762. doi:10.1037/0022-3514.79.5.748 Williams, K. D., Govan, C. L., Croker, V., Tynan, D., Cruickshank, M., & Lam, A. (2002). Investigations into differences between social-and cyberostracism. Group Dynamics: Theory, Research, and Practice, 6(1), 65–77. Williams, K. D., Shore, W. J., & Grahe, J. E. (1998). The silent treatment: Perceptions of its behaviors and associated feelings. Group Processes & Intergroup Relations, 1, 117–141. doi:10.1177/1368430298012002 Williams, K. D., & Sommer, K. L. (1997). Social ostracism by coworkers: Does rejection lead to loafing or compensation? Personality and Social Psychology Bulletin, 23, 693–706. doi:10.1177/0146167297237003 Williams, K. D., Wheeler, L., & Harvey, J. (2001). Inside the social mind of the ostracizer. In J. Forgas, K. D. Williams, & L. Wheeler (Eds.), The social mind: Cognitive and motivational aspects of interpersonal behavior (pp. 294–320). New York, NY: Cambridge University Press. Wirth, J. H., Bernstein, M. J., & LeRoy, A. S. (2015). Atimia: A new paradigm for investigating how individuals feel when ostracizing others. Journal of Social Psychology, 155(5), 497–514. Zadro, L., & Gonsalkorale, K. (2014). Sources of ostracism: The nature and consequences of excluding and ignoring others. Current Directions in Psychological Science, 23, 93–97. doi:10.1177/0963721413520321 Zadro, L., Williams, K. D., & Richardson, R. (2005). Riding the ‘O’ train: Comparing the effects of ostracism and verbal dispute on targets and sources. Group Processes & Intergroup Relations, 8, 125–143. doi:10.1177/1368430205051062

Received June 11, 2015 Accepted June 16, 2015



Ostracizing for a Reason: A Novel Source Paradigm for Examining the Nature and Consequences of Motivated Ostracism a

a

b

c

Sarah L. Gooley , Lisa Zadro , Lisa A. Williams , Elena Svetieva & a

Karen Gonsalkorale a

University of Sydney

b

UNSW Australia

c

Católica-Lisbon School of Business and Economics Published online: 12 Aug 2015.

Click for updates To cite this article: Sarah L. Gooley, Lisa Zadro, Lisa A. Williams, Elena Svetieva & Karen Gonsalkorale (2015) Ostracizing for a Reason: A Novel Source Paradigm for Examining the Nature and Consequences of Motivated Ostracism, The Journal of Social Psychology, 155:5, 410-431, DOI: 10.1080/00224545.2015.1060933 To link to this article: http://dx.doi.org/10.1080/00224545.2015.1060933






ARTICLES

Ostracizing for a Reason: A Novel Source Paradigm for Examining the Nature and Consequences of Motivated Ostracism SARAH L. GOOLEY and LISA ZADRO University of Sydney

LISA A. WILLIAMS UNSW Australia

ELENA SVETIEVA Católica-Lisbon School of Business and Economics

KAREN GONSALKORALE University of Sydney

ABSTRACT. Ostracism, a complex social phenomenon, involves both targets (ostracized individuals) and sources (ostracizers). The current experiment redressed a gap in the ostracism literature by devising a novel, three-phase paradigm to investigate motivated ostracizing. In the current study, 83 females were assigned to one of four conditions during a Cyberball game: motivated sources chose to ostracize an obnoxious fellow player, induced sources ostracized a fellow player at the behest of the experimenter, targets were ostracized, and included participants received the ball proportionately. Analysis of participants’ primary needs, emotions, ratings of their co-players, and behavior toward their co-players indicated that being the target of ostracism was a robustly aversive experience. Both motivated and induced sources reported fortified control. Moreover, a motive for ostracizing influenced source experience: induced sources experienced greater levels of negative moral emotion and behaved more prosocially toward their target than motivated sources. The flexibility and demonstrated impact of this novel paradigm adds to the toolkit available to researchers interested in expanding

Address correspondence to Lisa A. Williams, UNSW Australia, School of Psychology, Mathews Building, Sydney NSW 2052, Australia. E-mail: [email protected]

GOOLEY ET AL.

411

insight into the psychological processes underlying, and the motivational and behavioral outcomes of being, a source of ostracism.


Keywords: motivated sources, ostracism, primary needs, source paradigm

OSTRACISM OCCURS WHEN AN INDIVIDUAL is simultaneously excluded and ignored by one or more members of his or her social group (Williams, 2001). It is a phenomenon that is observed across cultures, generations, and species (Gruter & Masters, 1986), and is a daily experience for many (Nezlek, Wesselmann, Wheeler, & Williams, 2012). To date, empirical research on ostracism has focused on the targets of ostracism, including the changes in affect, cognition, and behavior that befall the ostracized individual. It is important to recognize, however, that ostracism is an inherently interpersonal behavior that requires not only a target, but also a source. In other words, for every individual who is being ostracized, there is at least one individual who is doing the excluding or ignoring (Williams, Wheeler, & Harvey, 2001). Most individuals will experience both perspectives across their life span (Williams, 2007). For instance, a person may be ignored by a colleague at work, but also decide to break ties with an undesirable friend. Two decades of work has focused on the negative outcomes of ostracism for targets, consistently demonstrating negative outcomes, including primary need threat (e.g., reduced belonging; Williams, 2001, 2009) and antisocial responses (van Beest, Carter-Sowell, van Dijk, & Williams, 2012; Warburton, Williams, & Cairns, 2006; Williams, Cheung, & Choi, 2000; for a review, see Williams, 2009). Yet, the nature, causes, and consequences of being a source of ostracism have, until recently, been understudied. Williams’s (1997) model of ostracism acknowledged sources by outlining potential antecedents and motives for ostracizing, but more recent iterations of the model are entirely target-centric (e.g., Williams, 2009). Moreover, although understanding of the outcomes of ostracism for targets has been primarily derived from experimental studies (for a review, see Williams, 2007), understanding of sources of ostracism has, until recently, been primarily derived from qualitative studies (e.g., Sommer, Williams, Ciarocco, & Baumeister, 2001; Zadro, 2004). The findings of such studies suggest that ostracizing is a complex phenomenon that varies on a number of dimensions. Sources vary in why, when, whom, how, and for how long they ostracize (Williams, 2001; Zadro, 2004), and these dimensions can lead to very different psychological consequences for sources. Some studies reveal that excluding and ignoring others is, for the most part, a positive and empowering experience: it is effective at terminating unwanted relationships, and it can be used to increase the source’s sense of power (control) over the target by dictating the type of interaction and when the conflict is resolved (Zadro & Gonsalkorale, 2014). Other findings, however, suggest that ostracizing can have negative consequences: ostracizing, particularly for prolonged periods of time, can be both cognitively and emotionally depleting (Williams, 2001; Zadro, 2004; Zadro, Godwin, & Gonsalkorale, 2013), and threatens belonging, self-esteem, and meaningful existence needs (Williams, Shore, & Grahe, 1998). Clearly, given the complexity of these findings, systematic experimental research on the nature and consequences of ostracizing is warranted. Yet, researchers interested in filling this gap have had few laboratory-based source paradigms at their disposal (for a review of source paradigms, see Zadro & Gonsalkorale, 2014). Broadly speaking, to induce a source experience, researchers can either ask participants to simply recall an ostracizing episode (i.e., write about an incident


412


where they excluded and ignored another person; e.g., Bastian et al., 2013; Poon & Chen, 2015; Sommer, Williams, Ciarocco, & Baumeister, 2001; Williams et al., 1998) or use an experimental paradigm whereby participants actively ostracize someone in the laboratory, whether it be a fellow participant (e.g., Poulsen & Kashy, 2012) or a confederate (e.g., Legate, DeHaan, Weinstein, & Ryan, 2013; Wesselmann, Williams, & Wirth, 2014; Wesselmann, Wirth, Pryor, Reeder, & Williams, 2013). Although more methodologically rigorous than recall paradigms (see Zadro & Gonsalkorale, 2014), experiential paradigms must overcome a number of methodological challenges in order to create a powerful and believable encounter in the laboratory that provokes participants to ostracize, not the least of which is the norm of social inclusiveness that typically characterizes novel social interactions (Zadro & Gonsalkorale, 2014). Of the experiential source paradigms developed to date, the most commonly used fall into three broad categories: direct-source inductions, induced-source inductions, and motivated-source inductions. In direct-source inductions (also known as simulations), participants are instructed to play the role of a source (e.g., during a role-play simulation, such as the O-Train paradigm; Zadro, Williams, & Richardson, 2005; ostracizing a naive participant during the experiment; Legate et al., 2013; Poulsen & Kashy, 2012; Sommer & Yoon, 2013; or being asked to write a rejection letter, for instance, to a job applicant or to a thesis student who has applied for a thesis prize; Chen, Poon, Bernstein, & Teng, 2014; Zhou, Zheng, Zhou, & Gou, 2009). In contrast, in induced-source inductions (Ciarocco, Sommer, & Baumeister, 2001), the experimenter gives participants the supposed choice of ostracizing or including another participant during an interaction. However, the experimenter mentions that, due to a large number of people previously choosing to include, it would be of great benefit to the study if the participant chose to ostracize. Inducedsource inductions are seen as a more valid source paradigm than direct-source inductions insofar as they encourage participants to take ownership of their actions on the grounds of having chosen to engage in ostracism rather than having been instructed to do so (Poulsen & Kashy, 2012). Although both direct-source and induced-source inductions are relatively easy to administer and result in ostracizing behavior, the extent to which either mirrors real-world ostracism situations is debatable. In both, participants are ostracizing at the behest of the experimenter, who is in a position of authority, even if they are given the illusion of choice (Zadro & Gonsalkorale, 2014). As such, findings from direct—and induced—source inductions may only apply to ostracism situations in which ostracism occurs solely at the behest of others. This is problematic, given that the vast majority of ostracism situations involve strong personal motives. Such motives vary from punishing the target for perceived or actual wrong-doing and/or correcting their behavior (i.e., punitive ostracism; Williams, 1997) to protecting oneself from potential harm (i.e., defensive ostracism; Williams, 1997; Zadro et al., in press). Such personal motives for ostracizing moderate the negative consequences of ostracizing others and influence the extent to which the target’s victimization is perceived to be justified (Williams, 2001; Wesselmann et al., 2013; 2014; Van Tongeren, Root Luna, & VanOyen Witvliet, 2015; Zadro et al., 2014). Thus, not surprisingly, simply following instructions to ostracize in direct-source inductions has been found to lead to psychological costs, such as reduced relatedness and autonomy needs, and increased guilt and shame (Legate et al., 2013). These psychological costs are thought to result from both violating the strong social norm of inclusion as well as ignoring a fellow participant who does not deserve this treatment (Wesselmann et al., 2013). To combat the potential limitations of induced-source paradigms, researchers have sought to devise motivated-source paradigms, whereby participants are provided with situational factors


GOOLEY ET AL.

413

that are designed to elicit personal motivations to exclude and ignore a target. These include eliciting punitive ostracism by giving the target negative personal characteristics (i.e., impatience, rudeness, self-aggrandizement, superiority; Sommer & Yoon, 2013; burdensome behavior, such as displaying poor skills in a ball-tossing task; Wesselmann et al., 2013; 2014; see also Wirth, Bernstein, & LeRoy, 2015); or eliciting defensive ostracism (e.g., by informing participants that the two other players will form an alliance against them; Benenson, Markovits, Thompson, & Wrangham, 2011). To date, however, pure motivated-source paradigms are rare. Some studies have combined personal motivations with induced paradigms (e.g., providing a punitive motive in the form of an undesirable target, but then inducing participants to ostracize or include; Sommer & Yoon, 2013; see also Ciarocco et al., 2001). In these studies, the majority of participants ostracize the target. In contrast, studies that have only used personal motivations to elicit ostracizing behavior have typically been hampered by strong norms of social inclusiveness. As a result, many participants are either not sufficiently motivated to ostracize or choose to only partially ostracize the target (e.g., Wesselmann et al., 2013). In real-world situations, there is considerable individual variation in the reasons for ostracizing, and these reasons can sometimes co-occur. Real-world sources state that multiple motivations to ostracize are typically more compelling than single motivations to ostracize (Zadro, 2004). Thus, current motivated-source paradigms may be hampered by their focus on eliciting a single ostracizing motivation (e.g., only punitive motives) which may not be sufficiently compelling to the entire sample.

THE CURRENT EXPERIMENT The aim of the current study was to create a motivated-source paradigm that would simultaneously provide participants with several compelling motivations to ostracize a target, thereby creating an environment conducive to ostracizing. The motivated-source paradigm was achieved by employing a three-phase methodology aimed to bring about multiple motivations shown in previous research to be effective in eliciting ostracism (see Figure 1). In the current study, the motivated-source paradigm was conducted by means of three computer-mediated activities with two other ‘participants’: Student 1 (a female) and Student 2 (a male; both computerized confederates). The motivated-source paradigm begins in Phase 1 by presenting participants with a punitive motivation for ostracizing. Punitive ostracism is the most common motive for ostracizing (Williams et al., 1998; Zadro, 2004) and is often directed against those who violate group norms (Wesselmann et al., 2013; Zadro, 2004). In the current study, Phase 1 was presented as a “Getting to Know You” task in which participants were asked to speak, via microphone, to Students 1 and 2 about their university experiences, and to listen to the (supposed) speeches of both students via headphones. To elicit punitive ostracism, we made a confederate, Student 2, a socially noxious character (see Sommer & Yoon, 2013) who exhibited intolerant and prejudiced traits, including racism and sexism. Given that these traits are in opposition to the egalitarianism and tolerance valued among the student populations from which we drew our sample, the inflammatory content of Student 2’s speech (the intended target of ostracism in the motivated-source condition) thus provided the initial motivation to ostracize.


414


FIGURE 1 Schematic of the motivated-source paradigm.

Phase 2 includes the provision of a defensive motivation to ostracize. In the current study, Phase 2 consisted of an instant-message task in which participants received messages from Students 1 and 2. During this task, participants received a message from Student 2 threatening to ostracize the participant during the subsequent task. This was designed to prompt participants to defensively ostracize Student 2 with the aim of preempting their own ostracism during the game. In order to maximize the effectiveness of raising defensive motivations in Phase 2, we chose to use a female-only sample in this experiment. Research has found that women are more likely than men to engage in defensive ostracism when they are threatened with being excluded and ignored (Benenson et al., 2011; see also, 2013). It is worth noting that the use of a femaleonly sample also augmented the impact of the punitive motivation used in this iteration of Phase 1 of the paradigm, not only with respect to the sexist remark but also the racist remarks made by Student 2. Specifically, studies suggest that women exhibit more egalitarian and less racist sociopolitical attitudes as compared to males (Ekehammar & Sidanius, 1982; Sidanius, Levin, Liu, & Pratto, 2000). Such sociopolitical attitudes have, in turn, been linked to gender differences in social dominance orientation (Pratto, Stallworth, & Sidanius, 1997). Given that the noxiousness of Student 2 in the current iteration of the paradigm stemmed from their highly social dominant behavior, using a female-only sample served to increase the punitive motivation to ostracize. Finally, in Phase 3, a socially sanctioned environment for the ostracizing event is provided. Given previous research that ostracizing is easier when socially sanctioned and enacted with a co-source (Zadro, 2004), in Phase 3 of the current study, the participant received a message from Student 1 suggesting that they both ostracize Student 2 together during the subsequent task (the ostracism opportunity, Cyberball—a triadic, virtual ball-tossing game; Williams et al., 2000), referencing Student 2’s obnoxious behavior. Student 1’s initiation of the ostracizing behavior


GOOLEY ET AL.

415

aimed to serve several purposes: to reduce potential negative emotions or self-image concerns that the participant may have experienced if they alone had initiated the ostracism of Student 2; to reassure the participant that negative reactions toward Student 2 were shared and justified; and to diffuse the responsibility for the ostracism act amongst both Student 1 and themselves (see Zadro et al., 2013). After Phase 3, the participant is given the opportunity to ostracize and their ostracizing behavior is observed and recorded. Although the motivated-source paradigm could be paired with almost all adapted experiential ostracism paradigms, the current experiment used the Cyberball paradigm (Williams et al., 2000; Williams & Jarvis, 2006) because it provides an effective, highly replicable paradigm that is used widely in the field. The newly designed motivated-source condition was compared to three other ostracism conditions that align with prior work: induced-source (e.g., Bastian et al., 2013; Ciarocco et al., 2001), target, and inclusion (e.g., Williams et al., 2000; Zadro, Williams, & Richardson, 2004). In these three conditions, Student 2 was portrayed as generally likable in Phase 1, and Student 1 made no suggestion of ostracizing in Phase 2. As such, in addition to introducing a new paradigm, the present experiment redresses a paucity of experimental research that compares the consequences of ostracism on targets and sources simultaneously (cf. Legate et al., 2013; Poulsen & Kashy, 2012; Zadro et al., 2005). Moreover, the current study is, to our knowledge, the first to systematically compare motivated and induced sources. To ensure the effectiveness of the motivated-source paradigm, participants’ perceptions of Student 2’s personality (e.g., trustworthiness, arrogance), behavior (i.e., the extent to which Student 2 would perform pro- or antisocial acts) and humanity (i.e., the extent to which Student 2 possessed traits indicative of human nature and human uniqueness; Bastian et al., 2013) were assessed. Participants in the motivated-source condition were expected to rate Student 2 more negatively on each of these dimensions than participants in the other conditions. Perceptions of Student 1 on these dimensions were also measured and were not expected to vary as a function of condition. The outcomes of the ostracism experience were captured via: (a) primary need threat (i.e., threats to belonging, control, self-esteem, and meaningful existence); (b) negative emotion; (c) self-dehumanization; and (d) behavior toward Students 1 and 2. The latter of these was operationalized using a taste-test task adapted from Chow, Tiedens, and Govan (2008), whereby participants allocated Students 1 and 2 a snack that they must taste in light of Student 1 and 2’s rankings of those snacks. In this task, giving a preferred snack reflects prosocial behavior whereas giving a disliked snack reflects antisocial behavior. Regarding primary needs, in accordance with previous research (see Williams, 2001; Zadro et al., 2013), we predicted that targets of ostracism would report greater need-threat than either type of sources and included participants. Predictions for the effects of ostracizing on primary needs were more equivocal, mirroring inconsistency in prior work. For instance, direct-source paradigms result in need fortification (e.g., Zadro et al., 2005) whereas induced-source paradigms result in self-reported need-threat (Daniels, 2011) and lowered behavioral control (e.g., Ciarocco et al., 2001). Given that motivated sources were likely to view their ostracizing behavior as more justified, we predicted that they would experience less need-threat than both induced sources and included participants. This was expected to be the case especially for belongingness, given that the participant ostracized alongside a co-source who encouraged the action (primarily as a means of defending the participant from being ostracized by Student 2 during the game, thereby


416


enhancing the bond between the participant and Student 1), and for control as a result of the participant choosing to ostracize Student 2 of their own free will (Poulsen & Kashy, 2012). With regard to negative emotion, and in line with past research (Gerber & Wheeler, 2009; Smart, Richman, & Leary, 2009), we hypothesized that targets of ostracism would experience more general negative affect than any of the other conditions. For sources, we hypothesized amplified guilt and shame, given their moral relevance. Guilt and shame arise in situations in which moral tenets have been violated (e.g., Tangney, Stuewig, & Mashek, 2007). Motivated sources were expected to experience lower levels of guilt and shame than induced sources as, given the noxious behavior of the target, they most likely perceived the enacted ostracism as warranted (Tangney et al., 2007). Participants’ perceptions of their own humanity (i.e., self-dehumanization; Bastian et al., 2013) were expected to be reduced among sources. Prior research has found that induced sources rate themselves as less human than included individuals (Bastian et al., 2013). Given that this self-dehumanization effect is causally explained by sources’ lowered sense of morality, and that ostracism—no matter the cause—represents a moral violation, we predicted that both induced and motivated sources would engage in self-dehumanization. However, following the logic for predictions for guilt and shame, we expected that self-dehumanization would be attenuated among motivated sources. With respect to actual behavior, there is evidence for both pro- and antisocial behavior by targets post-ostracism (see Zadro, 2011). For instance, women tend to show prosocial behavior (social compensation) as a means of regaining belongingness after being ostracized (Williams & Sommer, 1997), whereas targets show antisocial behavior if their sense of control post-ostracism is further thwarted (Warburton et al., 2006). However, consistent with previous research using the taste-test task (Chow et al., 2008), targets were predicted to behave relatively antisocially toward both Students 1 and 2 compared to included participants. Conversely, induced sources were predicted to behave equally as prosocially toward Student 2 as included participants, consistent with post-source affiliation motives (Bastian et al., 2013; Daniels, 2011). The novelty of the motivatedsource condition precluded specific predictions: Motivated sources could be expected to behave relatively prosocially, because they may see their ostracism as sufficient “punishment” for Student 2’s offensive behavior. Yet it is also possible that motivated sources would behave antisocially, carrying on Student 2’s punishment through to the taste-test task.

METHOD Participants and Design A total of 83 female undergraduate students at the University of Sydney (M age = 19.76, SD = 3.24) participated in return for course credit. Participants self-identified as Anglo or European (n = 45), Asian (n = 25), or other (n = 13). Participants were allocated to one of four conditions in a between-subjects design: inclusion condition, target of ostracism condition (hereafter, target condition), induced source of ostracism condition (hereafter induced-source condition), or motivated source of ostracism condition (hereafter motivated-source condition). We have made the materials and data set for this study publicly available at the Open Science Framework (OSF; project side: https://osf.io/cj3xd/). Below, we specify specific instances in which more detail is provided at OSF.

GOOLEY ET AL.

417


Procedure Participants believed the study aimed to examine the effects of mental visualization on impression formation over the Internet. The experimenter informed participants that they would be interacting online with two fellow students (Student 1 and Student 2) who were purportedly in labs on other floors of the building. These other “students” were fabricated confederates; all online interactions were preprogrammed. Ostensibly for confidentiality reasons, participants were told that each student had been randomly assigned a student number and that they were “Student 3.” Participants indicated their snack preferences for a later taste-test and indicated their sex, age, and degree for their profile in the “Getting to Know You” task. Next, participants engaged in an orientation to Cyberball, the “Getting to Know You” task, and the instant-message task, and then played Cyberball. As will be detailed below, critical aspects of the experimental conditions were delivered during these four stages of the experiment. Finally, participants completed a series of measures. They first completed a questionnaire that included items assessing primary needs, efficacy of the manipulation, self-dehumanization, and experienced emotion. These items were presented in a random order and interspersed with distractor items not of central interest to the current research questions. Next, participants rated Student 1, and then Student 2, on items relating to personality and behavioral inferences and dehumanization. Then, participants engaged in the taste-test task.1 All participants were fully debriefed regarding the nature of the study and use of deception. Orientation to cyberball. The experimenter outlined the Cyberball task. Participants in the induced-source condition were informed that part of the goal of the study was to examine how mental visualization might be affected by being included or excluded during the game. Participants were told that they would choose whether to include or exclude Student 2. As per prior induced-source paradigms (Ciarocco et al., 2001), before soliciting a choice, the researcher informed the participant that everyone had, so far, elected to be in the inclusion condition and it would “help [the experimenter] out” if the participant chose to be in the exclusion condition. The experimenter made it clear that the choice was theirs. After receiving the participant’s agreement to exclude Student 2, the researcher next left the room under the guise of establishing Student 1’s agreement to exclude Student 2. Upon return, the researcher confirmed that Student 1 had agreed and consequently Student 2 would not receive the ball at all during Cyberball. In the other three conditions, the researcher did not mention inclusion or exclusion, nor did they give participants the choice to explicitly adopt a strategy during Cyberball. “Getting to Know You” task. Phase 1 of the motivated-source paradigm took the form of the “Getting to Know You” task, the purpose of which was to purportedly establish familiarity among participant groups to facilitate performance in Cyberball. All participants prepared answers to questions pertaining to their experience of university for 10 min. After listening to Student 1 and 2’s answers via headphones, participants delivered their own into a microphone. In the profiles that appeared on the screen during the prerecorded answers, Student 1 was presented as an 18year-old female enrolled in a bachelor of arts degree program and Student 2 as an 18-year-old male studying engineering. Participants in all four conditions heard the same prerecorded answer from Student 1, who appeared friendly and likable. Student 2’s response varied according to condition. In the

418



motivated-source condition, Student 2 conveyed a range of noxious traits including racism (toward Asians: “. . . how they even got their jobs and got into the country is beyond me. . . .”) and sexism (“. . . and like most girls, [the girls in Student 2’s course] are just dumb. . . .”; for audio stimuli, see OSF). These traits were chosen with the goal of creating a situation that would motivate punitive ostracism (Sommer & Yoon, 2013), in line with the aims of Phase 1 of the motivated-source paradigm. In the induced-source, target, and inclusion conditions, Student 2’s response conveyed a general likeability. Instant-message task. The instant-message task provided a context to bring about motivation to defensively ostracize and a social-sanctioned context in which ostracism might occur, thus satisficing Phases 2 and 3 of the motivated-source paradigm. Thus, in the motivated-source condition, participants first received a message sent from Student 2, addressed to both the participant and Student 1, that proposed the idea of excluding the participant during Cyberball (a section of the quote is as follows): “. . . how about we make things more interesting. I think ill [sic] just throw the ball to you student 1. . . .” (for the full message, see OSF). This message was designed to motivate defensive ostracism (Phase 2). Next, participants in the motivated-source condition received a message from Student 1, sent just to them, that was designed to socially sanction ostracizing Student 2 during the game (Phase 3). Thus, a section of Student 1’s message read: “. . . I can’t believe he’s talking about leaving you out of the game. Let’s give him a taste of his own medicine. I’ll just throw the ball to you and you just throw the ball to me for the whole game—what do you think? . . .” (for the full message, see OSF). The messages participants received from Student 1 and Student 2 in the induced-source, target, and inclusion conditions were identical across conditions. Student 1 wrote, “Sounds fun guys!” and Student 2 wrote, “Let’s do it!” These messages were ostensibly sent to the both of the other two group members. After having received messages from Students 1 and 2, participants in all conditions chose whether to send a message to Student 1, Student 2, or both. Messages were later coded by two condition-blind research assistants according to the following mutually exclusive categorical scheme: positive tone (e.g., “Let the games begin ”), agreeing to ostracize (e.g., “I SOOOO agree . . . lets [sic] do this!!!!!”), rationalizing ostracism (e.g., “I like that idea, his speech was so rude I couldn’t believe it”), guilty tone (e.g., “I feel bad for Student 2!”), or other. Initial ratings showed a high degree of agreement between raters (intraclass correlation coefficient = .87). For this and other coding detailed below, discrepancies in coded categories were resolved via discussion and analyses were carried out on the resolved coding. Cyberball Game Participants were led to believe that they were playing an Internet ball-toss game with Students 1 and 2 (Williams et al., 2000). The game consisted of 30 throws. Consistent with past research (Williams & Jarvis, 2006; Zadro et al., 2004), participants in the inclusion condition received approximately one third of the total throws, whereas participants in the target condition received the ball twice at the commencement of the game and then not again for the remainder of the game.

GOOLEY ET AL.

419


Although Cyberball is typically used to examine the consequences of being a target of ostracism or inclusion, the current experiment innovatively employed Cyberball as an opportunity to ostracize for motivated and induced sources. In both the induced-source and motivated-source conditions, participants were the recipient of every one of Student 1’s ball tosses. Participants chose to throw the ball back to Student 1 or to Student 2. Consistent with past work on induced ostracism (Bastian et al., 2013; Ciarocco et al., 2001), participants were considered to have ostracized if they threw all 15 throws back to Student 1, hence excluding Student 2 entirely from the game. As such, participants in the induced-source and motivated-source conditions received half of the total throws. Measures Unless otherwise noted, all measures were taken on 5-point scales anchored by not at all to very much so; no other options were labeled. Reported reliability estimates are based on Cronbach’s α. All measures were completed on a computer. Paradigm validation. Participants responded to two questions assessing their experience of Cyberball. Participants rated the degree to which they felt included during Cyberball and estimated the percent of total throws received (Zadro, Boland, & Richardson, 2006). Given that one of the primary aims of this research was to develop a paradigm that would motivate individuals to be the source of ostracism, a number of manipulation checks assessed the degree to which perceptions of the other players’ personality traits and behavior were affected by the experimental protocol. Participants rated Student 1 and Student 2 on eight traits, both positive (trustworthy, moral, friendly, attractive) and negative (irrational, apathetic, rude, and arrogant). Negative trait ratings were reverse-scored and averaged with positive trait ratings to form a perceived personality index (αStudent 1 = .80, αStudent 2 = .92). Participants also rated the extent to which they believed that Student 1 and Student 2 would engage in a variety of prosocial behaviors (i.e., engage in volunteer community work, return a lost wallet to its owner, donate money to charity, donate blood) and antisocial behaviors (i.e., steal, lie, commit fraud, cheat on a romantic partner). Antisocial items were reverse-scored and averaged with the prosocial items for each student separately to form indices of inferred behavior (αStudent 1 = .84, αStudent 2 = .92). Participants rated Student 1, Student 2, and themselves on the degree to which each person possessed human uniqueness and human nature traits (Bastian & Haslam, 2010; Bastian et al., 2013). In line with past research, human uniqueness (e.g., “I felt like I/Student 1/Student 2 was refined and cultured”) and human nature (e.g., “I felt like I/Student 1/Student 2 was mechanical and cold, like a robot”) scores were highly correlated (self: r(66) = .54, Student 1: r(66) = .72, Student 2: r(66) = .86, ps < .001) and hence combined into scores reflecting self humanity (α = .66), Student 1 humanity (α = .82), and Student 2 humanity (α = .92). In accordance with previous research, allocation of lower human nature and human uniqueness scores was inferred to reflect dehumanization (e.g., Bastian & Haslam, 2010; Bastian et al., 2013). Primary Needs Questionnaire (PNQ). Participants completed the Primary Needs Questionnaire (Zadro et al., 2006). The questionnaire comprises 12 items to assess four primary

420


needs: belonging (α = .83), self-esteem (α = .77), control (α = .75), and meaningful existence (α = .84).


Negative emotion. Participants rated how they felt during Cyberball on four items related to general negative affect (i.e., “I felt angry/sad/tense/happy[reversed]/relaxed[reversed],” α = .69) and two items assessing negative moral emotion (i.e., “I felt guilty/ashamed,” α = .62). Taste-test task. The taste-test task provided participants with the opportunity to engage in a behavior ranging from pro- to antisocial. This type of task reliably differentiates post-ostracism behaviors of targets and included participants (Chow et al., 2008). At the outset of the study, participants were asked to rank their own snack preferences from 1 (most preferred) to 6 (least preferred), ostensibly at the same time as the other two students. After the conclusion of Cyberball, participants were informed that they had been randomly assigned to select a snack for Student 1 and a snack for Student 2 and that Students 1 and 2 would have to taste whatever had been selected for them. Participants were then presented with the snack rankings for each student and were assured that their choices would remain anonymous. The most prosocial behavior in this task took the form of selecting Student 1 or 2’s first snack choice for them to eat, whereas the most antisocial behavior took the form of selecting Student 1 or 2’s sixth snack choice (see OSF for full description of the task). After making their selections for Student 1 and then Student 2, participants were asked to provide a reason for their choices. Responses were later coded by two independent, conditionblind research assistants according to four mutually exclusive categories reflecting the tone of the rationale: prosocial (e.g., “I thought they would like it”), retributive (e.g., “I didn’t like his attitude, I suppose you can see it as having a taste of your own medicine”), guilty (“Because I felt so bad about leaving him out of the ball game!!!”), or other. Initial ratings showed a high degree of agreement (intraclass correlation coefficient = .78, .89 for Student 1 and Student 2 rationales, respectively).

RESULTS Analytic Approach Continuous outcome measures were analyzed using a series of analyses of variance to examine the effect of condition on outcomes of interest. In all analyses, condition was treated as a between-subjects factor and Type III sums of squares was utilized. For outcomes that existed for both Student 1 and Student 2 (i.e., personality ratings, behavioral inferences, dehumanization, and snack choice), we chose to treat “person” as a within-subjects, repeated factor. Relatedly, for ratings for which there were theoretical subtypes (i.e., primary needs, negative emotion), we chose to treat subtype as a within-subjects, repeated factor. In cases where main effects or interactions emerged, post-hoc analyses on estimated marginal means were conducted. Bonferroni adjustment for multiple comparisons was utilized to control family-wise error rates. Reported p values for comparisons across conditions are adjusted values. For the categorical outcome measure of coded message content, we employed a Pearson chi-square test of independence. All analyses were carried out using SPSS v.22.

GOOLEY ET AL.

421


Paradigm Validation It is often the case with source paradigms—whether direct or indirect—that participants choose to defy ostracizing instructions (Legate et al., 2013), or—in the case of induced-source inductions— decide to include rather than exclude. Ostracizing in the current study was defined as excluding Student 2 completely from the game (i.e., throwing all 15 ball tosses to Student 1). Analysis of throwing behavior during Cyberball revealed that, despite agreeing to ostracize Student 2, one participant in the induced-source condition threw Student 2 the ball during the game, and hence data from this participant were excluded from the analyses reported below. Further, 16 participants in the motivated-source condition were partial ostracizers, throwing the ball to Student 2 at least once (despite seven of these participants identifying that they had excluded Student 2 for their noxious behavior and three stating that they complied with Student 1’s ostracism suggestion). Given the diversity of their behavior (throwing to Student 2 anywhere from one to ten times on anywhere from the second to the 18th throw), conclusions about their motivations and behaviors would be tenuous at best. Hence, we chose to exclude their data from the analyses reported below. Separate analyses of data from partial ostracizers appear in a supplement to this article available at the Open Science Framework (https://osf.io/cj3xd/). These exclusions resulted in a final sample of 66 participants (17 each in inclusion and induced-source and 16 each in target and motivated-source conditions).2 A number of measures served as checks on whether the ostracism paradigm was effective. These included perceived inclusion and throws during Cyberball and pregame messages, as well as inferences about the personality and behavior of Students 1 and 2 and humanity ratings of Student 1, Student 2, and themselves. Descriptive statistics for these variables, aside from coded message content, appear in Table 1. As expected, the degree to which participants felt included during the game varied as a function of condition, F(3,62) = 43.31, p < .001, ηp 2 = .68. Post-hoc comparisons between conditions indicated that, as would be expected, participants in the target TABLE 1 Descriptive Statistics of Paradigm Check Variables and Self-Reported Negative Emotion Variable Perceived inclusion Perceived percentage of received throws Personality Student 1 Student 2 Behavior Student 1 Student 2 Humanity Student 1 Student 2 Negative affect Negative moral emotion

Motivated-source

Induced-source

Target

Inclusion

4.69 (0.48) 57.19 (16.73)

4.12 (0.70) 53.88 (18.24)

1.75 (0.58) 9.88 (4.75)

3.88 (1.17) 28.94 (7.89)

4.34 (0.38) 1.92 (0.68)

4.08 (0.63) 3.97 (0.46)

3.88 (0.64) 3.72 (0.64)

4.04 (0.51) 3.79 (0.62)

4.13 (0.85) 2.12 (0.76)

4.04 (0.59) 3.98 (0.56)

4.00 (0.51) 3.72 (0.65)

3.88 (0.58) 3.54 (0.64)

35.88 17.38 1.95 1.91

(3.83) (5.48) (0.58) (0.97)

33.88 32.47 2.06 3.00

(4.27) (3.34) (0.62) (1.06)

Note. Standard deviations appear in parentheses next to corresponding means.

33.19 31.94 2.54 1.78

(4.74) (6.04) (0.77) (0.84)

32.53 31.06 1.64 1.24

(4.84) (6.04) (0.38) (0.40)


422


condition felt less included than participants in all three other conditions (ps < .001, ds ≥ 2.32). Interestingly, participants in the motivated-source condition reported higher levels of felt inclusion than participants in the inclusion condition (p < .03, d = 0.91). No other condition-wise differences were significant (ps > .24, ds ≤ 0.95). Similarly, estimates of ball-throws received varied as a function of condition, F(3,62) = 46.12, p < .001, ηp 2 = .69. Post-hoc comparisons between conditions indicated that participants in the motivated-source and induced-source conditions estimated having received more throws than included and excluded participants (ps < .001, ds ≥ 1.78), but did not differ from one another (p > .99, d = 0.19). Furthermore, included participants estimated that they received more throws than excluded participants (p = .001, d = 2.93). Efficacy of the motivated-source condition was expected to be confirmed in analysis of participants’ pregame messages, and perceptions of Students 1 and 2 (personality, behavior, and humanity). An analysis of pregame messages by two independent coders revealed that rates of rationalization varied by condition, χ 2 (3) = 32.57, p < .001, ϕ = .70. In the motivatedsource condition, 56.25% of participants wrote rationalizing content. Rationalizing content did not appear in any other condition. Personality ratings of Student 1 and Student 2 following the game were submitted to a (2) (person: Student 1 vs. Student 2) × 4 (condition) mixed-design ANOVA, which revealed main effects of person, F(1,62) = 135.68, p < .001, ηp 2 = .69, and condition, F(3,62) = 9.61, p < .001, ηp 2 = .32. These were qualified by a significant person by condition interaction, F(3,62) = 78.71, p < .001, ηp 2 = .79. Ratings of Student 1’s personality did not differ according to condition (ps ≥ .12, ds ≤ 0.88). Motivated sources rated Student 2’s personality as significantly more negative than participants in any of the other conditions (ps < .001, ds ≥ 2.86). No other condition-wise differences among ratings of Student 2’s personality were significant (ps > .99, ds ≤ 0.45). Inferences about the prosociality of Student 1 and Student 2’s behaviors followed a similar pattern. Main effects of person, F(1,62) = 58.88, p < .001, ηp 2 = .49, and condition, F(3,62) = 8.28, p < .001, ηp 2 = .29, were qualified by a significant person by condition interaction, F(3,62) = 25.75, p < .001, ηp 2 = .56. Student 1 was rated equally across conditions (ps > .99, ds ≤ 0.34). Motivated sources rated Student 2 as being less likely to engage in prosocial behaviors (and more likely to engage in antisocial behaviors) compared to participants in all three other conditions (ps < .001, ds ≥ 2.03), who did not differ from one another (ps ≥ .35, ds ≤ 0.72). Participants’ humanity ratings of Students 1 and 2 were subjected to a (2) (person: Student 1 vs. Student 2) × 4 (condition) mixed-design ANOVA. Main effects of person, F(1,62) = 87.41, p < .001, ηp 2 = .59, and condition, F(3,62) = 8.08, p < .001, ηp 2 = .28, were qualified by a significant interaction between person and condition, F(3,62) = 49.04, p < .001, ηp 2 = .70. Ratings of Student 1’s humanity did not differ according to condition (ps ≥ .21, ds ≤ 0.77). However, motivated sources rated Student 2 as lower in humanity compared to participants in any of the other three conditions (ps < .001, ds ≥ 2.37), who did not differ from one another (ps > .99, ds ≤ 0.29). Primary Needs A mixed-design ANOVA revealed main effects of primary need type, F(2.74,169.77) = 93.29, p < .001, ηp 2 = .60, and condition, F(3,62) = 31.79, p < .001, ηp 2 = .61.3 These were qualified,


GOOLEY ET AL.

423

FIGURE 2 Mean levels of primary need satisfaction according to condition. Error bars represent ±1 standard error of the mean.

as expected, by an interaction between primary need type and condition, F(8.22,169.77) = 4.21, p < .001, ηp 2 = .17 (see Figure 2). Post-hoc comparisons between conditions revealed that the pattern of means was largely the same across conditions for all four needs. Specifically, targets of ostracism reported lower levels of belongingness, self-esteem, meaningful existence, and control need satisfaction (hence higher need threat) compared to participants in the other three conditions (ps ≤ .01, ds ≥ 1.18), barring levels of control needs being equivalent among targets and included participants (p = .11, d = 1.09). For belonging, self-esteem, and meaningful existence, there were no significant differences between the other three conditions (ps ≥ .66, ds ≤ 0.65). However, motivated and induced sources reported higher levels of control need satisfaction compared to included participants (ps ≤ .007, ds ≥ 1.07), but did not differ from one another (p > .99, d = 0.29). Negative Emotion Self-rated emotions following Cyberball were submitted to a (2) (emotion type: negative affect vs. moral negative emotion) × 4 (condition) mixed-design ANOVA, which revealed a significant main effect of condition, F(3,62) = 8.95, p < .001, ηp 2 = .30, that was qualified by a significant emotion type by condition interaction, F(3,62) = 15.10, p < .001, ηp 2 = .42 (see Table 1 for descriptive statistics by condition). Post-hoc comparisons between conditions on negative affect revealed that targets reported higher levels of negative affect compared to motivated sources and included participants (ps ≤ .045, ds ≥ 0.86). With regard to negative moral emotion, induced sources reported the highest levels of guilt and shame compared to all three other conditions (ps ≤ .003, ds ≥ 1.08). No other condition-wise comparisons were significant for negative affect or for negative moral emotion (ps ≥ .17, ds ≤ 0.83).

424


Self-Dehumanization Contrary to hypotheses and extant evidence that induced sources demonstrate selfdehumanization (Bastian et al., 2013), participants’ ratings of their own humanity did not vary across conditions (M motivated-source = 26.63, SDmotivated-source = 4.98, M induced-source = 23.47, SDinduced-source = 6.08, M target = 25.88, SDtarget = 4.27, M inclusion = 26.88, SDinclusion = 4.53), F(3,62) = 1.62, p = .20, ηp 2 = .07.


Taste-Test Selections Taste-test selections for Students 1 and 2 were submitted to a (2) (person: Student 1 vs. Student 2) × 4 (condition) mixed-design ANOVA. In this analysis, higher numbers reflect selecting a lower-ranked snack for the other person. Main effects of person, F(1,62) = 7.68, p = .007, ηp 2 = .11, and condition, F(3,62) = 3.95, p = .01, ηp 2 = .16, were qualified by a person by condition interaction F(3,62) = 15.19, p < .001, ηp 2 = .42. Post-hoc comparisons between conditions for allocations to Student 1 revealed no differences in behavior (Ms = 1.31–1.94, SDs = 0.86–1.56, ps ≥ .83, ds ≤ 0.48). Allocations to Student 2, however, varied according to condition. Motivated sources (M = 4.06, SD = 2.29) acted more antisocially, allocating a lesser-favorite snack toward Student 2 compared to participants in all three other conditions (M induced-source = 1.41, SDinduced-source = 0.62, M target = 1.56, SDtarget = 0.81, M inclusion = 1.71, SDinclusion = 1.31, ps < .001, ds ≥ 1.26), who did not differ from one another (ps > .99, ds ≤ 0.29). Investigation of the frequency with which the most antisocial choice was made for Student 2 further illustrates the nature of this finding. Whereas 50% of participants in the motivated-source condition selected Student 2’s least favorite snack, only one other participant across all three other conditions did so (in the inclusion condition). Instead, 62.5–64.7% of participants in the inclusion, target, and induced-source conditions selected Student 2’s first choice. Participants’ motives for their snack allocations were coded as to whether they were prosocial, retributive, or guilty. Cell counts were too low to conduct non-parametric frequency analyses (i.e., less than five in some cells), but a clear pattern emerged for the rationale given for snack allocations to Student 2 (see Figure 3). Whereas the majority of included participants, induced sources, and targets reported prosocial motivations, a small proportion of motivated sources mentioned prosocial reasons. Instead, over half of motivated sources reported retributive motivations. Note also that a small proportion of targets reported retributive motivations. Finally, only two participants mentioned guilty motivations: one in the motivated-source condition and the other in the induced-source condition. For comparison, prosocial motivations for Student 1 allocations were robust across the four conditions.

DISCUSSION The aim of the current experiment was to develop a source paradigm that motivates participants to ostracize of their own volition (i.e., motivated ostracism) by inducing multiple motives and providing a socially sanctioned ostracizing environment. We used this new, three-phase paradigm not only to examine the effects of ostracism on both targets and sources, but also to compare


GOOLEY ET AL.

425

FIGURE 3 Percentages of motivation types given by participants as rationale for their snack choices to Student 1 (left) and Student 2 (right).

two different types of source experiences—motivated and induced sources—which is a first in the field. Examining ostracism from multiple perspectives revealed new insights into the intrapersonal effects of ostracizing another. Being a source of ostracism, whatever the reason, evinced benefits in terms of greater control need fortification. Our new paradigm revealed that the emotional and behavioral concomitants of being a motivated source differed from those of induced sources: motivated sources reported lower levels of guilt and shame and behaved more antisocially toward their target than induced sources. A key innovation of the motivated-source paradigm is creating a situation that elicits punitive, defensive, and socially sanctioned ostracism. In the motivated-source condition, the potential target was obnoxious and threatened to ostracize the participant. Further, the other person present became a co-source, suggesting and engaging in ostracism of the potential target. In line with this design, participants in the motivated-source condition rated Student 2’s personality and behavior more negatively, and perceived him to be lower in humanity, than participants in the other conditions. Importantly, the goal to engender these motivations in the absence of negative emotions for engaging in motivated ostracism was successful: motivated sources did not differ from included participants on levels of negative affect or negative moral emotion. Thus, the three-pronged approach to motivating ostracism appears to be effective. The success of the motivated-source paradigm is particularly notable, given the strong norm of social inclusiveness that typically dictates behavior in novel social interactions (Poulsen & Kashy, 2012; Wesselmann et al., 2013). That said, inclusiveness norms may explain why some participants in the motivated-source condition chose not to ostracize fully, and underscores the methodological challenges of creating a source experience in the laboratory. Rates of ostracizing in the motivated-source condition (50%) were lower than in the induced-source condition in the current experiment (94%) and in previous studies (89–100%, Bastian et al., 2013; Ciarocco et al., 2001). We wish to note that there was evidence of partial ostracism among the other 50% of participants in the motivated-source condition, but that the diversity of their behavior precluded


426


sensible analysis of their data. As noted above, analyses of data from partial ostracizers is available on OSF. Reduced rates of complete ostracism were counteracted by gains in ecological validity over induced-source paradigms (e.g., Ciarocco et al., 2001; Legate et al., 2013; Poulsen & Kashy, 2012). Indeed, many real-life ostracism sources are motivated in their behavior rather than instructed or pressured into it, engage in ostracism to varying degrees (from mild to severe), and do so for a variety of motivations (Zadro, 2004; Zadro et al., 2013). It is our sense that the motivated-source paradigm we developed effectively captures these characteristics of ostracizing. With respect to observed effects of being a source on need satisfaction, emotion, and behavior toward targets, sources reported greater control need satisfaction than both targets and included participants, which is consistent with previous findings that sources perceive ostracism as a means of attaining power and control over the target (Sommer et al., 2001) and that sources of ostracism report greater control satisfaction compared to sources of argument (Zadro et al., 2005). Given that they ostracized along with a co-source, we predicted that motivated sources would report higher levels of belonging than induced sources. Although belongingness needs did not differ as a function of source type, participants in the motivated-source condition reported higher levels of felt inclusion during Cyberball than participants in the induced-source condition. Given that the Primary Needs Questionnaire (Zadro et al., 2006) was designed to measure need-threat among targets, it may lack the sensitivity to effectively capture need-fortification among sources and, further, to discern differences between distinct types of sources. As such, developing a source-relevant measure of needs to capture the dynamics of the ostracizing experience remains a challenge for future research. We observed differentiation between motivated and induced sources with respect to negative moral emotion. Specifically, induced sources reported higher levels of guilt and shame than motivated sources, even though they behaved in an identical manner during Cyberball (i.e., they did not throw the ball to Student 2 at all). This finding is consistent with research showing that people feel sympathy toward targets when the ostracism is unjustified (Wesselmann et al., 2013); in the induced-source condition, the target had done nothing to warrant ostracizing them. In many ways, the attenuated guilt and shame among motivated sources could be a result of moral justification: ostracism represents a violation of social and moral norms, and hence elicits negative moral emotion (Tangney et al., 2007), but the violation is justifiable given the obnoxious behavior of the target and thus elicits less guilt and shame. Future research should explore the situational contexts that amplify or reduce moral justification and ensuing negative moral emotions, as experienced negative moral emotion dictates post-transgression behaviors (e.g., de Hooge, Zeelenberg, & Breugelmans, 2007; Sheikh & Janoff-Bulman, 2010). A related focus of future research stands in relation to the role of emotions as mechanisms of the effects of source motivations on ostracism and subsequent pro- and antisocial behavior. Motivated and induced sources also differed in their behavior toward their targets. Motivated sources allocated a lesser-favorite snack to Student 2 compared to induced sources, with many reporting retributive motives for their decision. Thus, it appears that motivated sources chose to punish Student 2 for his offensive behavior both during Cyberball and the taste-test task. In contrast, the majority of participants in the induced-source condition allocated Student 2 his favorite snack, and the vast majority reported prosocial motivations for their snack allocations. The prosocial behavior of induced sources is consistent with previous findings that people will include undeserving targets more than others who have not been ostracized (Wesselmann et al., 2013; for a discussion on compensatory behavior by sources toward rejected targets, see Van Tongeren, Root Luna, & VanOyen Witvliet, 2015).


GOOLEY ET AL.

427

The current experiment revealed that ostracism is more aversive to targets than sources. Targets reported lower levels of need satisfaction and highest levels of negative affect than sources (and than included participants, barring control needs). Unexpectedly, however, targets did not engage in antisocial behavior toward their sources compared to included participants, which contradicts the findings of Chow et al. (2008).4 A potential reason for these discrepant findings may stem from differences in the design of the taste-test task. Chow et al. (2008) extrapolated antisocial behavior based on pilot-study rankings of snacks (e.g., prunes were unappealing and hence antisocial). In the current experiment, participants were made explicitly aware of Student 1 and 2’s snack choices and made their snack choices while viewing those rankings. Hence, allocating a less-favored snack represented strong antisocial intent, perhaps stronger than that assessed by Chow et al. (2008). The lack of antisocial behavior is not terribly surprising given prior research on targets’ desires to regain affiliation among sources (usually as a means of regaining belongingness; see Williams, 2001; Zadro, 2004, 2011). Consistent with affiliative aims, targets reported prosocial motivations for their snack choices, and, in fact, refrained from derogating the personality, behavior, and humanity of their sources (Students 1 and 2). Future research is needed to identify key moderators of antisocial versus affiliative behavior among targets. The current iteration of the motivated-source paradigm was optimized to result in a female sample voluntarily choosing to ostracize a male target. It is worth noting that the motivated-source paradigm is highly flexible and can be enacted in a number of ways simply by varying the nature of each of the three components of the paradigm (e.g., the characteristics of the sample, the operationalization of the punitive and defensive motivations, the characteristics of the target and the co-source, the type of ostracism paradigm used to observe ostracizing behavior, and the ostracism opportunity medium [e.g., online or face-to-face]). Future research can modify aspects of the motivated-source paradigm to enable researchers to study populations of interest. For instance, the target may be portrayed as racist or intolerant in order to elicit punitive ostracism in a sample consisting of a particular ethnic minority or religious group, or the target’s speech may contain competitive overtones to elicit ostracism in male samples (see Benenson et al., 2008). Overall, the gender of participants is not at the crux of this paradigm—rather, the flexible nature of the motivated-source paradigm makes it a strong tool for studying how members of different social groups (e.g., religious groups, ethnic or racial groups, groups based on political or sexual orientation) respond to the opportunity to ostracize others. Overall, we expect that the basic mechanisms, processes and outcomes that we uncovered using the iteration of the paradigm deployed in the current study are likely to be observed across other samples and other adaptations of the paradigm. A potential limitation of the study is our failure to replicate prior findings of inducedsource self-dehumanization (Bastian et al., 2013). One possibility is that the current study was underpowered to replicate this effect. However, a post-hoc power analysis conducted in G∗ Power (v3.1) indicated that, given our sample size, we observed a power level above .99 to detect the effect sizes of the difference between induced sources and included participants observed by Bastian et al. (Study 2: d = 2.26, Study 4: d = 1.93; 2013). Comparison between the induced-source paradigm used by Bastian and colleagues (Studies 2 and 4; 2013) and the paradigm we used reveals more theoretically-oriented insight as to why this may have occurred. Whereas Bastian and colleagues’ participants were induced to ostracize another person face-to-face (i.e., exclude them from a physical, in person ball-toss game [Study 2] or give them the silent treatment in person [Study 4]), our induced sources ostracized their targets during Cyberball and did not, in fact, meet their target face-to-face. Although this could be seen as a weakness, this aspect of our paradigm has clear extensions into cyberbullying (Dooley, Pyzalski,


428


& Cross, 2009). We expect that psychological distance between an ostracizer and their target might moderate the degree to which sources self-dehumanize, although this possibility remains to be examined in future research. Despite these limitations, our motivated-source paradigm carries the benefit of capturing real-life aspects of ostracism. Although it is true that individuals are sometimes induced to ostracize someone nice at the behest of another (such as in the induced-source condition), there are other times when individuals are motivated to ostracize an objectionable person (such as in the motivated-source condition). Future research is needed to identify whether all three aspects of our three-pronged approach to motivating sources are required. For example, participants’ decision to ostracize may have been influenced primarily by Student 2 being an unlikable character, his threat to ostracize the participant, Student 1’s encouragement to ostracize Student 2, or some combination of these factors. In sum, the present experiment presents a novel paradigm that motivates individuals into deciding to ostracize another. Our findings demonstrate that the effects of ostracism vary according to one’s role in the ostracism episode. Being the target of ostracism is a universally aversive experience. Ostracizing, no matter the reason, fortifies control needs. Moreover, the concomitants of being the source of ostracism vary according to whether sources were induced or motivated to ostracize. Whereas being an induced source raises negative moral emotion, being a motivated source occurs without guilt or shame. Induced sources subsequently behave more prosocially toward their target than motivated sources. Clearly, then, not all source experiences are equal, nor do they all result in the same consequences. It is our hope that this paradigm can be deployed in future research to systematically test the motives offered by Williams’s (1997) original model of ostracism, and, in so doing, prompt deeper insight into the full human ostracism experience.

ACKNOWLEDGMENT We would like to thank Alexandra Godwin for her feedback and assistance.

FUNDING This project was supported by an Australian Research Council Discovery Grant (DP110105195).

ORCID Lisa A. Williams http://orcid.org/0000-0002-0021-5613 Elena Svetieva http://orcid.org/0000-0001-9632-4728

GOOLEY ET AL.

429


NOTES 1. After the taste-test task, but before reporting demographics, participants completed a delayed Primary Needs Questionnaire (PNQ) interspersed with filler items. This questionnaire consisted of the same items used in the immediately following Cyberball; however, the instructions asked them to answer the questions according to how they felt “right now.” Analysis of self-esteem, belonging, meaningful existence, and control need levels as a function of condition revealed main effects of need type, F(2.85,176.40) = 117.87, p < .001, ηp 2 = .66, and condition, F(3,62) = 6.92, p < .001, ηp 2 = .25. Bonferroni-adjusted post-hoc tests revealed that levels of primary needs all differed significantly from one another (ps < .001, ds ≥ 0.50), with meaningful existence needs more satisfied than belonging needs, which were more satisfied than self-esteem needs, which in turn were more satisfied than control needs. Further, total need satisfaction (collapsing across need type) was lower among targets compared to participants in all other conditions (ps ≤ .02, ds ≥ 0.77), who did not differ from one another (ps > .99, ds ≥ 0.27). The interaction between primary need type and condition was not significant, F(8.54,176.40) = 0.58, p = .80, ηp 2 = .03. 2. In the current study, two participants in the motivated-source condition received a slight variation of the paradigm (i.e., Student 2’s speech was less derisive, but still obnoxious). Both of these participants engaged in complete ostracism of Student 2. We have included their data in the analyses reported here. Exclusion of their data from reported analyses resulted in no substantive changes to the nature of the reported findings, barring one post-hoc comparison: participants in the motivated-source condition reported only marginally higher levels of felt inclusion than participants in the inclusion condition (p = .06). 3. Mauchly’s test of sphericity revealed that sphericity was violated for primary needs ratings, χ 2 (5) = 18.10, p = .002. Hence, degrees of freedom were corrected using the Huynh-Feldt estimate of sphericity (ε = .91). 4. Chow et al. (2008) documented that increased anger among targets causally explained increased antisociality in the taste-test task. In our data, exploratory analyses of just the self-reported anger variable revealed that, whereas anger was elevated among targets compared included participants, t(31) = 2.86, p = .008, d = 0.98, anger did not correlate with antisocial behavior toward Students 1 or 2 in these conditions, Student 1: r(33) = −.22, p = .22; Student 2: r(33) = −.11, p = .55. There are many situational constraints on anger-induced aggression (e.g., Denson, 2013), and it remains a question for future research to determine the circumstances in which ostracism-induced anger guides antisocial behavior toward sources.

AUTHOR NOTES Sarah L. Gooley and Lisa Zadro are affiliated with the University of Sydney. Lisa A. Williams is affiliated with UNSW Australia. Elena Svetieva is affiliated with Católica-Lisbon School of Business and Economics. Karen Gonsalkorale is affiliated with the University of Sydney.

REFERENCES Bastian, B., & Haslam, N. (2010). Excluded from humanity: The dehumanizing effects of social ostracism. Journal of Experimental Social Psychology, 46, 107–113. doi:10.1016/j.jesp.2009.06.022 Bastian, B., Jetten, J., Chen, H., Radke, H., Harding, J., & Fasoli, F. (2013). Losing our humanity: The selfdehumanizing consequences of social ostracism. Personality and Social Psychology Bulletin, 39(2), 156–169. doi:10.1177/0146167212471205 Benenson, J., Hodgson, L., Heath, S., & Welch, P. (2008). Human sexual differences in the use of social ostracism as a competitive tactic. International Journal of Primatology, 29, 1019–1035. doi:10.1007/s10764-008-9283-4 Benenson, J. F., Markovits, H., Hultgren, B., Nguyen, T., Bullock, G., & Wrangham, R. (2013). Social exclusion: More important to human females than males. PLoS ONE, 8(2), e55851. doi:10.1371/journal.pone.0055851 Benenson, J., Markovits, H., Thompson, M., & Wrangham, R. (2011). Under threat of social exclusion, females exclude more than males. Psychological Science, 22, 538–544. doi:10.1177/0956797611402511 Chen, Z., Poon, K.-T., Bernstein, M. J., & Teng, F. (2014). Rejecting another pains the self: The impact of perceived future rejection. Journal of Experimental Social Psychology, 50, 225–233. doi:10.1016/j.jesp.2013.10.007 Chow, R. M., Tiedens, L. Z., & Govan, C. L. (2008). Excluded emotions: The role of anger in antisocial responses to ostracism. Journal of Experimental Social Psychology, 44, 896–903. doi:10.1016/j.jesp.2007.09.004


430


Ciarocco, N. J., Sommer, K. L., & Baumeister, R. F. (2001). Ostracism and ego depletion: The strains of silence. Personality and Social Psychology Bulletin, 27(9), 1156–1163. doi:10.1177/0146167201279008 Daniels, J. R. (2011). Two sides of the coin: Breaking social units and the effects of ostracism on the source of ostracism (Doctoral dissertation). Retrieved from http://digitalcommons.uconn.edu/dissertations/AAI3492074 de Hooge, I. E., Zeelenberg, M., & Breugelmans, S. M. (2007). Moral sentiments and cooperation: Differential influences of shame and guilt. Cognition & Emotion, 21(5), 1025–1042. doi:10.1080/02699930600980874 Denson, T. F. (2013). The multiple systems model of angry rumination. Personality and Social Psychology Review, 17(2), 103–123. doi:10.1177/1088868312467086 Dooley, J. J., Pyzalski, J., & Cross, D. S. (2009). Cyberbullying versus face-to-face bullying: A theoretical and conceptual review. Journal of Psychology, 217(4), 182–188. doi:10.1027/0044-3409.217.4.182 Ekehammar, B., & Sidanius, J. (1982). Sex differences in sociopolitical attitudes: A replication and extension. British Journal of Social Psychology, 21, 249–257. doi:10.1111/j.2044-8309.1982.tb00546.x Gerber, J., & Wheeler, L. (2009). On being rejected: A meta-analysis of experimental research on rejection. Perspectives on Psychological Science, 4(5), 468–488. doi:10.1111/j.1745-6924.2009.01158.x Gruter, M., & Masters, R. D. (1986). Ostracism as a social and biological phenomenon: An introduction. Ethology and Sociobiology, 7, 149–158. doi:10.1016/0162-3095(86)90043-9 Legate, N., DeHaan, C., Weinstein, N., & Ryan, R. (2013). Hurting you hurts me too: The psychological costs of complying with ostracism. Psychological Science, 24(4), 583–588. doi:10.1177/0956797612457951 Nezlek, J. B., Wesselmann, E. D., Wheeler, L., & Williams, K. D. (2012). Ostracism in everyday life. Group Dynamics: Theory, Research, and Practice, 16, 91–104. doi:10.1037/a0028029 Poon, K.-T., & Chen, Z. (2015). How does the source of rejection perceive innocent victims? Journal of Social Psychology, 155(5), 515–526. Poulsen, J. R., & Kashy, D. A. (2012). Two sides of the ostracism coin: How sources and targets of social exclusion perceive themselves and one another. Group Processes & Intergroup Relations, 15(4), 457–470. doi:10.1177/1368430211430517 Pratto, F., Stallworth, L., & Sidanius, J. (1997). The gender gap: Differences in political attitudes and social dominance orientation. British Journal of Social Psychology, 36, 49–68. doi:10.1111/j.2044-8309.1997.tb01118.x Sheikh, S., & Janoff-Bulman, R. (2010). The “shoulds” and “should nots” of moral emotions: A self-regulatory perspective on shame and guilt. Personality and Social Psychology Bulletin, 36(2), 213–224. doi:10.1177/0146167209356788 Sidanius, J., Levin, S., Liu, J. H., & Pratto, F. (2000). Social dominance orientation, anti-egalitarianism and the political psychology of gender: An extension and cross-cultural replication. European Journal of Social Psychology, 30, 41–67. doi:10.1002/(SICI)1099-0992(200001/02)30:13.0.CO;2-O Smart, R. L., & Leary, M. R. (2009). Reactions to discrimination, stigmatization, ostracism, and other forms of interpersonal rejection: A multimotive model. Psychological Review, 116, 365–383. doi:10.1037/a0015250 Sommer, K. L., Williams, K. D., Ciarocco, N. J., & Baumeister, R. F. (2001). When silence speaks louder than words: Explorations into the intrapsychic and interpersonal consequences of social ostracism. Basic and Applied Social Psychology, 23(4), 225–243. doi:10.1207/S15324834BASP2304_1 Sommer, K., & Yoon, J. (2013). When silence is golden: Ostracism as resource conservation during aversive interactions. Journal of Social and Personal Relationships, 30(7), 901–919. doi:10.1177/0265407512473006 Tangney, J. P., Stuewig, J., & Mashek, D. J. (2007). What’s moral about the self-conscious emotions. In J. L. Tracy, R. W. Robins, & J. P. Tangney (Eds.), The self-conscious emotions: Theory and research (pp. 21–37). New York, NY: Guildford Press. van Beest, I., Carter-Sowell, A. R., van Dijk, E., & Williams, K. D. (2012). Groups being ostracized by groups: Is the pain shared, is recovery quicker, and are groups more likely to be aggressive? Group Dynamic: Theory, Research, and Practice, 16(4), 241–254. doi:10.1037/a0030104 Van Tongeren, D. R., Root Luna, L. M., & VanOyen Witvliet, C. (2015). Insufficient justification for rejection elicits compensatory behavior. Journal of Social Psychology, 155(5), 527–534. Warburton, W. A., Williams, K. D., & Cairns, D. R. (2006). When ostracism leads to aggression: The moderating effects of control deprivation. Journal of Experimental Social Psychology, 42, 213–220. doi:10.1016/j.jesp.2005.03.005 Wesselmann, E., Williams, K. D., & Wirth, J. (2014). Ostracizing group members who can (or cannot) control being burdensome. Human Ethology Bulletin, 29, 82–103. Wesselmann, E., Wirth, J., Pryor, J., Reeder, G., & Williams, K. (2013). When do we ostracize? Social Psychological and Personality Science, 4, 108–115. doi:10.1177/1948550612443386


GOOLEY ET AL.

431

Williams, K. D. (1997). Social ostracism. In R. M. Kowalski (Ed.), Aversive interpersonal behaviors (pp. 133–170). New York, NY: Plenum. Williams, K. D. (2001). The power of silence. New York, NY: Guilford Press. Williams, K. D. (2007). Ostracism. Annual Review of Psychology, 58, 425–452. doi:10.1146/annurev.psych.58.110405. 085641 Williams, K. D. (2009). Ostracism: Effects of being excluded and ignored. In M. Zanna (Ed.), Advances in experimental social psychology (pp. 275–314). New York, NY: Academic Press. Williams, K. D., Cheung, C. K. T., & Choi, W. (2000). Cyberostracism: Effects of being ignored over the Internet. Journal of Personality and Social Psychology, 79, 748–762. doi:10.1037/0022-3514.79.5.748 Williams, K. D., & Jarvis, B. (2006). Cyberball: A program for use in research on interpersonal ostracism and acceptance. Behavior Research Methods, 38(1), 174–180. doi:10.3758/BF03192765 Williams, K. D., Shore, W. J., & Grahe, J. E. (1998). The silent treatment: Perceptions of its behaviors and associated feelings. Group Processes & Intergroup Relations, 1(2), 117–141. doi:10.1177/1368430298012002 Williams, K. D., & Sommer, K. L. (1997). Social ostracism by coworkers: Does rejection lead to loafing or compensation? Personality and Social Psychology Bulletin, 23(7), 693–706. doi:10.1177/0146167297237003 Williams, K. D., Wheeler, L., & Harvey, J. (2001). Inside the social mind of the ostracizer. In J. Forgas, K. D. Williams, & L. Wheeler (Eds.), The social mind: Cognitive and motivational aspects of interpersonal behavior (pp. 294–320). New York, NY: Cambridge University Press. Wirth, J. H., Bernstein, M. J., & LeRoy, A. S. (2015). Atimia: A new paradigm for investigating how individuals feel when ostracizing others. Journal of Social Psychology, 155(5), 497–514. Zadro, L. (2004). Ostracism: Empirical studies inspired by real-world experiences of silence and exclusion. Unpublished doctoral dissertation. School of Psychology, University of New South Wales, Sydney, Australia. Zadro, L. (2011). Silent Rage: When being ostracized leads to aggression. In J. P. Forgas, A. Kruglanski, & K. D. Williams (Eds.), The psychology of social conflict and aggression (pp. 201–216). New York, NY: Psychology Press. Zadro, L., Boland, C., & Richardson, R. (2006). How long does it last? The persistence of the effects of ostracism in the socially anxious. Journal of Experimental Social Psychology, 42(5), 692–697. doi:10.1016/j.jesp.2005.10.007 Zadro, L., Godwin, A., & Gonsalkorale, K. (2013). “It is better to give than to receive”: The role of motivation & self control in determining the consequences of ostracism for targets and sources. In J. P. Forgas, & E. Harmon-Jones (Eds.), The control within: Motivation and its regulation (pp. 351–366). New York, NY: Psychology Press. Zadro, L., Godwin, A., Svetieva, E., Sethi, N., Iannuzzelli, R., & Gonsalkorale, K. (in press). Creating the silence: Ostracism from the perspective of the source. In K. D. Williams, & S. A. Nida (Eds.), Ostracism, social exclusion and rejection. Frontiers of social psychology handbook series. New York, NY: Psychology Press. Zadro, L., & Gonsalkorale, K. (2014). Sources of ostracism: The nature and consequences of excluding and ignoring others. Current Directions in Psychological Science, 23(2), 93–97. doi:10.1177/0963721413520321 Zadro, L., Williams, K. D., & Richardson, R. (2004). How low can you go? Ostracism by a computer is sufficient to lower self-reported levels of belonging, control, self-esteem and meaningful existence. Journal of Experimental Social Psychology, 40, 560–567. doi:10.1016/j.jesp.2003.11.006 Zadro, L., Williams, K. D., & Richardson, R. (2005). Riding the “O” train: Comparing the effects of ostracism and verbal dispute on targets and sources. Group Processes & Intergroup Relations, 8(2), 125–143. doi:10.1177/1368430205051062 Zhou, X., Zheng, L., Zhou, L., & Guo, N. (2009). The act of rejecting reduces the desire to reconnect: Evidence for a cognitive dissonance account. Journal of Experimental Social Psychology, 45, 44–50. doi:10.1016/j.jesp.2008.08.011

Received December 1, 2014 Accepted June 3, 2015



Ostracism in Everyday Life: The Effects of Ostracism on Those Who Ostracize a

b

c

John B. Nezlek , Eric D. Wesselmann , Ladd Wheeler & Kipling D. Williams

d

a

College of William & Mary University of Social Sciences and Humanities, Poznan b

Illinois State University

c

Macquarie University

d

Purdue University Published online: 12 Aug 2015.

Click for updates To cite this article: John B. Nezlek, Eric D. Wesselmann, Ladd Wheeler & Kipling D. Williams (2015) Ostracism in Everyday Life: The Effects of Ostracism on Those Who Ostracize, The Journal of Social Psychology, 155:5, 432-451, DOI: 10.1080/00224545.2015.1062351 To link to this article: http://dx.doi.org/10.1080/00224545.2015.1062351





Ostracism in Everyday Life: The Effects of Ostracism on Those Who Ostracize


JOHN B. NEZLEK College of William & Mary University of Social Sciences and Humanities, Poznan

ERIC D. WESSELMANN Illinois State University

LADD WHEELER Macquarie University

KIPLING D. WILLIAMS Purdue University

ABSTRACT. Ostracism is a negative interpersonal experience that has been studied primarily in laboratory settings. Moreover, these studies have focused primarily on how people feel when they have been ostracized. The present study extended this research by investigating ostracism as it occurs in daily life, focusing on how people feel about ostracizing someone. Using a method modeled after the Rochester Interaction Record (RIR), for two weeks, 64 participants (adults residing in the community) described what happened each time they ostracized someone. The questions in the diary were based on Williams’s (2001) need–threat model of ostracism. Most ostracism episodes were directed toward people of equal status, and participants reported lower levels of belonging but higher levels of control after ostracizing someone. Punitive ostracism was associated with more positive outcomes for the source than when people ostracized someone for other reasons. Keywords: event-contingent diary, social exclusion, sources of ostracism

OSTRACISM—THE ACT OF EXCLUDING AND IGNORING another person—occurs in both human and nonhuman social animals (Williams, 2009; Williams & Nida, 2011). Further, there is evidence that it has occurred throughout history, across various social contexts and cultures (Williams, 2001). It also appears that ostracism is fairly common in contemporary daily life. A recent study suggests that people may be ostracized on a regular, almost daily basis (Nezlek, Wesselmann, Wheeler, & Williams, 2012). Address correspondence to John B. Nezlek, College of William & Mary, Department of Psychology, Box 8795, Williamsburg, VA 23187-8795, USA. E-mail: [email protected]


NEZLEK ET AL.

433

Despite the considerable attention that ostracism has received, the impact/influence of ostracism on the source of ostracism (i.e., the person who is ostracizing someone else) has received relatively little attention compared to the attention effects on the target (i.e., the person being ostracized) has received. Moreover, to our knowledge, no study has examined the influence of ostracizing on the sources of ostracism within a naturalistic, nonlaboratory setting, and the present study was designed to examine ostracism from the perspective of the source as it occurs in a naturalistic, nonlaboratory setting. This study complements Nezlek et al. (2012), which was (to our knowledge) the first study of naturally occurring ostracism from the perspective of the target. A common finding is that ostracism is painful, both psychologically and physically, to those who experience it (Williams, 2009). Consistent with this, research suggests that people recognize that ostracism/social rejection harms others and that people are often hesitant to inflict such pain on others (Joel, Teper, & MacDonald, 2014; Legate, DeHaan, & Ryan, 2015; Wesselmann, Wirth, Pryor, Reeder, & Williams, 2013). Assuming that ostracism is common and people recognize it is as harmful, why do individuals ostracize each other willingly and with such regularity in daily life? The regular use of ostracism may reflect the fact that it can serve valuable purposes. It appears that ostracism has served a useful function perhaps since humans’ evolutionary past (e.g., Gruter & Masters, 1986; Wesselmann, Nairne, & Williams, 2012). Specifically, ostracism can be used to control group members who pose a threat to the collective by their deviate behavior or stigmatized status (Kurzban & Leary, 2001; Williams, 2009). Ostracism, when used for punitive reasons, focuses on motivating individual group members to change undesirable or harmful behavior and on protecting the group from harm (Wesselmann, Williams, & Wirth, 2014; Wesselmann, Wirth, Pryor, Reeder, & Williams, 2015; Williams, 1997; Wirth, Bernstein, & LeRoy, 2015). Consistent with this, individuals who use ostracism regularly in their interpersonal relationships indicate that it can change undesirable behavior in others (Buss, Gomes, Higgins, & Lauterbach, 1987; Zadro, Williams, & Arriaga, 2008). Williams (1997) described other types of ostracism that are not directly related to evolutionary (i.e., survival-based) goals but may occur on a regular basis. Role-prescribed ostracism occurs when the social situation requires (or at least allows for) ostracism to occur between individuals, e.g., ignoring waitstaff who refill water glasses or remove dishes. This behavior may be tacitly approved by the patrons and waitstaff. Defensive ostracism occurs when individuals ostracize another person to defend against being ostracized or otherwise negatively treated themselves. Individuals may ostracize another person because they are afraid that simply being associated with that person may cause them to be ostracized by proxy (e.g., courtesy stigma; Goffman, 1963), or because they expect that someone else may ostracize them first (Williams, 2001). Oblivious ostracism occurs when one person simply does not notice or acknowledge the existence of another. This type of ostracism may occur when there is a status difference between the source and target, with the source believing that the target is simply unworthy of attention. The final type of ostracism, not-ostracism, occurs when people feel ostracized when in reality they have not been ostracized. For example, an individual greets a passerby, but the passerby does not hear or see the greeter. As such, the passerby did not intentionally ignore the greeter, but the greeter could still (mistakenly) assume he or she had been intentionally ignored (ostracized).

434



STUDYING THE USES OF OSTRACISM IN LABORATORY RESEARCH To study the effects of ostracism on sources of ostracism, several researchers have manipulated ostracism in face-to-face interactions. Each of these studies has relied upon role-playing in which participants have been randomly assigned to either include or ostracize another person (Ciarocco, Sommer, & Baumeister, 2001; Poulsen & Kashy, 2012; Williams, Bernieri, Faulkner, Gada-Jain, & Grahe, 2000; Zadro, Williams, & Richardson, 2005). Although the ostracizers were following instructions, in each of these studies they experienced some negative psychological effects (e.g., threat to belonging needs, negative affect, guilty feelings). These effects have been replicated in online interactions as well (Bastian et al., 2013; Legate, DeHaan, Weinstein, & Ryan, 2013). It is likely that sources of ostracism need a stronger justification than simply obeying an authority figure (e.g., an experimenter) to mitigate the negative effects ostracizing another person can have on one’s self. For example, Sommer and Yoon (2013) were able to reverse the cognitive depletion effects found by Ciarocco et al. (2001) by manipulating the characteristics of the person people were asked to ostracize. When participants obeyed the experimenter and ostracized a polite confederate they experienced emotional distress and cognitive depletion, but when they ostracized a rude confederate they did not show the same distress—presumably because the confederate’s inhospitable disposition gave them additional justification for ostracizing this person. Although previous studies have been informative, ostracism does not typically occur because one is asked to do it. People typically engage in ostracism because they are motivated to do so. Using a version of Cyberball (Williams, Cheung, & Choi, 2000), an online game of ball toss in which participants interact with two other players (actually computer-programmed confederates), and ostracism is defined as not “throwing” the ball to someone, Wesselmann, Wirth et al. (2013) manipulated the time each player (confederate) took to decide to throw every ball toss. They created a burdensome group member by programming a confederate to take considerably longer than other players (16 sec. vs. 3–5 sec.) for each toss. Participants were more likely to ostracize (i.e., give fewer tosses to) a slow group member than to a group member who threw at a normal speed. Within Williams’ model, this type of ostracism may best be described as punitive in nature, and people may feel justified in ostracizing someone who is violating the norms of the group (i.e., going too slowly). These laboratory studies have provided a useful starting point for studying the sources of ostracism. Nevertheless, in their daily lives, people ostracize others for various reasons and probably have different reactions to these different experiences. The feelings of unpleasantness, self-degradation, and fatigue found in laboratory research may be less likely to occur outside of the lab when an individual feels justified in ostracizing another or has a reason to do so. Thus, it is important to study ostracism as it occurs naturally in the ebb and flow of daily life.

STUDYING THE USES OF OSTRACISM IN EVERYDAY LIFE Research suggests that ostracism is a common occurrence in everyday life. One study surveyed over 2,000 Americans; 67% reported having used ostracism and 75% reported having been ostracized during their lifetime (Faulkner, Williams, Sherman, & Williams, 1997). The results of the diary study conducted by Nezlek et al. (2012) suggest that people may experience ostracism on a daily basis, and it appears that people are more frequently ostracized by acquaintances or strangers than by friends or family members.


NEZLEK ET AL.

435

In other research on naturally occurring ostracism, people have recalled autobiographical events in which they ostracized another person (Chen, Poon, Bernstein, & Teng, 2013; Sommer, Williams, Ciarocco, & Baumeister, 2001; Williams, Shore, & Grahe, 1998). In these studies, people indicated that they had ostracized another person as a form of punishment or a manipulation tactic during interpersonal conflict. Some, particularly those lower in self-esteem, reported that they often used ostracism as a defense against criticism or rejection (Sommer et al., 2001). Research also suggests that ostracism—an effective influence tactic in the short term—may be toxic to relationships in the long term. Ostracized partners may develop feelings of disconnection and resentment and may ultimately withdraw from the relationship entirely (Gottman & Krokoff, 1989; Sommer et al., 2001; Zadro et al., 2008). The present study investigated the uses of ostracism as it occurs in daily life. Participants were adults living in Sydney, Australia, who described what happened each time they ostracized someone over a two-week period. These descriptions included the nature and target of the ostracism and how they reacted to ostracizing another person. These data allowed us to examine the uses of ostracism as ostracism occurs in vivo, including examining the differences in peoples’ reactions to using ostracism as a function of their motives and the characteristics of their targets. We used an “event-contingent” diary method, which provides several advantages over studies using only single assessments requiring retrospection over an extended period of time (Reis & Gable, 2000; Wheeler & Reis, 1991). In such studies, participants are instructed to recognize specific events in their daily lives (in this case, when they ostracized another person) and to describe each event shortly after it occurs. This method reduces the influence that recalling single events may have on global retrospective assessments, minimizes the influence of the difficulty in recalling distant events accurately, and because it involved repeated measures, provides more reliable measures. See Nezlek (2012) for a discussion of the rationale for using diary methods.

RESEARCH QUESTIONS Although the existing research on sources of ostracism did not provide a basis for clear predictions for our study, it did provide a basis for some expectations. Assuming that people are motivated to ostracize someone (for whatever reasons), ostracism represents goal achievement of some kind, and as such, it should result in some positive changes in the ostracizer. We expected that one outcome of this should be an increased sense of control. When people ostracize someone they are controlling the social interaction because they are determining an essential aspect of an interaction—whether the interaction occurs or continues. Another consequence of ostracizing someone should be a decreased sense of belonging. By definition, ostracism consists of ignoring or not acknowledging some type of relationship—however transient and ephemeral that relationship might be. So, when people ostracize someone they should experience a decreased sense of belonging relative to the target of their ostracism. Previous research has not found consistent or clear patterns of ostracism affecting sources’ self-esteem or meaningful existence needs. As such, we do not have specific predictions about these two needs. Although we collected numerous descriptions of the structural characteristics of ostracism (e.g., function, type, and relationship with target), the available research and theory did not provide a basis for clear predictions about how reactions to ostracizing someone might vary as a

436


function of such characteristics. Somewhat tentatively, we thought that punitive ostracism would have the greatest influence on the ostracizer because among the types of ostracism, it is the most focused, most specific, and most directly intended to influence someone’s behavior. Other types of ostracism (e.g., defensive, role prescribed) may be less straightforward in terms of their motives, and so their influences on the ostracizer may not be as uniform across situations and contexts.


METHOD Participants The sample consisted of 64 people (25 men and 39 women; age, M = 26.9, SD = 1.61) who answered an advertisement in a Sydney, Australia, newspaper and were paid 40AUD for their participation. Participants recorded a total of 1,005 ostracism episodes (M = 15.70, SD = 9.01) across an average of 10.2 days (SD = 3.82). No participant was dropped from the analyses for failure to comply with instructions. All data and materials are available at osf.io/fx8d2. Materials and Procedure The procedure and instructions for the study were modeled after those introduced by Wheeler and Nezlek (1977) and were similar to those used in Nezlek et al. (2012). In small groups of less than 20, participants were informed about the study. They were told that the study was about ostracism and that it was specifically about when they ostracized someone, which was described as ignoring or excluding someone. We told participants that even though there might be times when ostracizing someone was not important, we wanted them to record these instances. We provided conceptual definitions and examples of each different type of ostracism and asked participants to generate examples from their own life experience. We also provided written scenarios representing each ostracism type, and participants were taught how to use the scales to describe these examples. They were given specific definitions of each measured concept and given the opportunity to ask clarification questions about the measures and procedure. The training sessions took about 90 min. Participants were told to describe when each event occurred (time, date). Participants also described their relationship with the target: stranger, acquaintance, ordinary friend, close friend, partner, or relative, and the relative social status of the target: inferior, equal, or superior. We did not provide participants with specific definitions of these two aspects but instead let them choose the terms they thought most appropriate for each ostracism episode. Participants included how the ostracism was accomplished: socially, in the presence of others; physical separation, ignoring others or physically removing oneself, and what we called cyber ostracism: ignoring someone over the telephone, mail, e-mail, chat rooms, and so forth. Participants also indicated whether they ostracized the target on their own or if there were others involved (and if so, how many and who they were).

NEZLEK ET AL.

437

Participants also described the type of ostracism:


1. Not ostracism: their behavior might have mistakenly been interpreted by the other as ostracism. 2. Role prescribed: the norms and roles within the situation dictated the ostracism. 3. Punitive: the ostracism occurred to punish or to indicate disapproval. 4. Defensive: the ostracism occurred for self-protection of the ostracizer. 5. Oblivious: the ostracism occurred because the source did not notice the target. The clarity of the reasons for the ostracism was rated on a 5-point scale, with endpoints labeled Unclear (“You give no announcement and don’t imply/infer the reason behind your ostracism”) and Clear (“You are explicit—e.g., ‘I’m mad at you so I’m not going to talk with you’”). Degree/strength of ostracism was also measured on a 5-point scale with endpoints labeled Barely (“Your behavior has barely changed from what it would be during normal interaction”) and Complete (“Your behavior has completely changed from what it would be during normal interaction”). Participants indicated how they felt about ostracizing the target by responding to a series of questions that began with the stem: “Compared to how you felt prior to ostracizing, how have your feelings changed as a result of ostracizing?” They provided these ratings on −3 to +3 scales, for which −3 represented lower or less, the midpoint (0) represented no change, and +3 represented higher or more. The six ratings were belonging, control, self-esteem, meaningful existence, angry, and apologetic. Participants were told to maintain the diary for two weeks, and they were told to complete forms as soon as possible after they had ostracized someone. We asked them to return their completed records every two to three days. After they completed the diary, using 7-point scales they answered a series of questions about how they had maintained the diary. Their responses indicated that it was not difficult to keep the diary (M = 2.89, SD = 1.26) and they thought their diaries were accurate (M = 5.68, SD = .92).

RESULTS The present data formed a multilevel data structure in which observations at one level of analysis (ostracizing events) were nested within another level of analysis (people). Accordingly, the data were analyzed with a series of multilevel models (MLM) using the program HLM (Raudenbush et al., 2011). Using MLM to analyze event-contingent data is discussed in Nezlek (2012).1 ,2 We should note that preliminary findings from this study were discussed in Williams, Wheeler, and Harvey (2001), a presentation and description that were intended to provide a general overview of the study. These preliminary findings did not take into account the nested structure of the data, which is critical in diary studies in which observations are nested within persons (e.g., Nezlek, 2001). For example, although there is some overlap in the variables presented in Table 2 and in Table 14.8 in Williams et al. (2001), the results presented in Table 2 were taken from multilevel models. The results presented in the other tables were not presented in Williams et al. (2001).

438


Descriptive Statistics: An Overview of Ostracizing Others in Everyday Life


To provide a context for evaluating our results, we begin by presenting descriptive statistics for our measures. The presentation is organized around the nature of the measure, either continuous or categorical, because the two types of measures required slightly different analyses and provide slightly different summary statistics. Continuous measures were analyzed series of unconditional models (i.e., no predictors at either level of analysis): Level 1 (within − person) : yij = b0j + rij Level 2 (between − person) : b0j = g00 + u0j In the level 1 model, yij is a measure for person j for event i, β0j is a random coefficient representing the mean of y for person j (collapsed across the i events each person described), rij represents the error associated with each measure (deviation from each person’s mean), and the variance of rij constitutes the within-person (level 1) variance. At level 2, γ00 represents the mean of the β0j , and the variance of u0j represents the between-person (level 2) variance. The results of these analyses are summarized in Table 1. In terms of clarity, on average, participants reported that their ostracizing another person was between the scale points of “pretty unclear” (2) and “moderately clear” (3), suggesting some lack of clarity/directness. In terms of the degree of ostracism, the mean was just below the midpoint of the scale (3, labeled “moderate”), indicating some change in the interaction, but not great or dramatic changes. It is also worth noting that for clarity, most of the variance (82%) was between persons, suggesting that this aspect of ostracism reflects individual differences between people rather than situational characteristics (within-person variance). In contrast, for degree of ostracism and the ratings presented below, the majority of the variance was within persons (i.e., situational). For ratings of belonging, control, self-esteem, meaningful existence, angry, and apologetic, the midpoint of the scale was 0, representing no change. Negative numbers represent diminished

TABLE 1 Descriptive Statistics for Continuous Measures

Variable

Mean

Clear [unambiguous] Degree/change Belong Control Self-esteem Meaning Anger Apology

2.43 2.80 −.46 .39 .09 .10 .01 −.13

∗∗∗ significantly

t ratio

6.37∗∗∗ 4.06∗∗∗ 1.05 1.42 .35), and so the significant relationship between clarity and belongingness in the analysis with both predictors represents a suppressor effect. Clarity was positively related to belongingness only after degree of ostracism was controlled. Strong and clear ostracism is a two-edged sword. The clarity may help people understand where they stand in the relationship, but the strength of the treatment may alienate them (Buss et al., 1987; Gottman & Krokoff, 1989). One of the strongest findings of the study (in terms of consistency of results across measures) was the finding that punitive ostracism brought about the most changes in ostracizers compared to other types. These changes were generally positive, and they suggest that punitive ostracism is the most rewarding form of ostracism. Post-hoc tests also found that it was the clearest form of ostracism and that it resulted in the most change in the situation compared to other types (ps < .05). Previous research relying upon single assessments of uses of ostracism suggests that punitive ostracism is the most common type of ostracism (Sommer et al., 2001; Williams et al., 1998). Largely consistent with this, we found that unclear/ambiguous ostracism was the most common type of ostracism, and punitive ostracism was the second most common. Moreover, post-hoc pairwise comparisons found that the reasons for punitive ostracism were much more likely to be something about the target than about the ostracizer or the situation (ps < .01). Such an attribution is consistent with the fact that punitive ostracism led to increased anger and decreased feelings of apology. Clearly, participants felt justified when they used punitive ostracism. Such a possibility is consistent with the results of Wesselmann et al. (2014), who found that people ostracized a burdensome group member regardless of whether or not that member was responsible for being burdensome. Although people recognized that the burdensome behaviors were not under the target’s control, they still described this person in negative terms that suggested elements of victim blame. Indeed, attributing responsibility for using ostracism to something about the target may be a necessary component for providing sources sufficient justification for knowingly subjecting this person to social pain (see Chen & Poon, 2015; Gooley, Zadro, Williams, Svetieva, & Gonsalkorale, 2015; Van Tongeren, Root Luna, & Van Oyen Witvliet, 2015).

Correspondence Between Reports of Sources and Targets of Ostracism The present study is a companion to the study describing targets’ experiences of ostracism in daily life (Nezlek et al., 2012). Although the samples and focus of the studies are different, both relied upon the same model of ostracism and the measures of ostracism collected in the studies


446


were similar. These similarities provided a basis to compare and contrast the pictures of ostracism provided by the two studies. The reports of sources and targets converged in various ways, suggesting that participants in the two studies were describing the same class of events. In both studies, the majority of ostracizing events were between people of the same status (source: 77%; target, 78%); the majority of ostracizing events were in person (source: 61%; target, 71%); and close others (close friends, partners, and relatives) were ostracized less often than others (source: 23%; target, 22%). Moreover, although the distributions of type of ostracism were not identical, there were not pronounced differences between the two studies. In terms of reactions to ostracizing, both sources and targets felt a decreased sense of belonging after ostracism (source: −.46; target, −.83), and for both studies, most of the variance in reactions was at the event level rather than at the person level. Reactions to events varied more from event to event than from person to person. The reports of sources and targets also diverged in various ways. Targets of ostracism thought that the ostracism they experienced was due to something about them only 12% of the time, whereas sources thought that the target was responsible about 30% of the time. Reactions to ostracism also varied considerably across the two points of view. On average, sources felt more in control after ostracizing someone, but reported no other changes in terms of the needs specified by Williams (2009). In contrast, after being ostracized, targets felt less in control and felt less positively in terms of all of the needs specified by Williams (2009). Differences between sources’ and targets’ reactions to ostracism suggest that being ostracized is much more impactful than ostracizing someone. Although ostracizing someone can provide some rewards, these rewards are not nearly as great as the costs incurred by the targets. This discrepancy in importance is also reflected in the fact that targets’ reactions varied more as a function of the structural characteristics of the ostracism (type and relationship) than sources’ reactions. The specifics of ostracizing events were more important to targets than they were to sources. Situational cues should generally be crucial to targets because such cues will help them understand why the ostracism originally occurred and, perhaps, how to avoid it in the future (Williams, 2009). Implications for Future Research The present results suggest ostracizing someone can be rewarding, particularly in the case of punitive ostracism. These rewards accrue despite the fact that people recognize that ostracism has negative effects on targets (Wesselmann, Hales, & Williams, 2013; Wesselmann et al., 2015), and therefore, ostracizing someone means intentionally harming that person. Within this context, ostracizing can be, and we believe should be, considered as a form of aggression. Previous research suggests that hostile cognitions mediate the relationship between exclusion and aggressive behavior (DeWall, Twenge, Gitter, & Baumeister, 2009). Hostile cognitions can lead individuals to behave aggressively because hostile cognitions cause people to perceive ambiguous acts as hostile and to expect aggression from others (Anderson & Bushman, 2002). Thus, it is likely that experiencing repeated ostracism would increase someone’s hostile cognitive biases, and it is possible that continually using ostracism would also increase a source’s hostile cognitive biases. If this is the case, ostracism could be part of cycles of negativity and aggression in relationship (for examples, see interviews in Williams, 2001), and such cycles tend to undermine relationship functioning (Rusbult, Johnson, & Morrow, 1986).


NEZLEK ET AL.

447

Moreover, some data suggest that some individuals use ostracism on a regular basis in their interactions with others. Many of these individuals recognize how using ostracism can become addictive and how difficult it can be to stop using ostracism once it becomes a habit (William, 2009). Although ostracism is an effective manipulation tactic (e.g., Buss et al., 1987), it can quickly initiate a vicious cycle of attack and counterattack. Nevertheless, the immediate success of ostracism may make it a primary choice for future conflict resolution, even though long-term use of ostracism reduces relationship quality (Williams, 2009; Zadro et al., 2008). Based on the negative effects ostracism has on targets, it is reasonable to assume that individuals who are regularly ostracized by their partners would have less satisfying relationships than individuals who are not regularly ostracized. Moreover, relationship satisfaction is an important predictor of overall relationship commitment and longevity (Le & Agnew, 2003), and so regular use of ostracism in a relationship may increase the likelihood that the relationship ends. It is not clear, however, if people believe that ostracism is more beneficial for their relationships than verbal arguments and do not realize the harmful effects ostracism may have. If this is the case— if sources were made aware of these negative outcomes—they might use ostracism less often (Bushman, Baumeister, & Phillips, 2001). It is also possible that individuals who frequently use ostracism in one relationship may generalize this behavior to other relationships, resulting in greater dispositional aggressiveness and poorer relationship functioning generally (Patterson & Maccoby, 1980). Such possibilities need to be examined using research designed explicitly for these purposes. Finally, we found some gender differences in some aspects of how and why people use ostracism. There is little research specifically on gender differences in using ostracism, compared to other forms of social aggression (Archer & Coyne, 2005; Crick & Grotpeter, 1995). Nevertheless, the existing research suggests that women are more likely to use ostracism in interpersonal relationships as both a competitive and defensive social influence tactic (Benenson, Hodgson, Heath, & Welch, 2008; Benenson et al., 2013; Benenson, Markovits, Thompson, & Wrangham, 2011). Our results support the conclusion of these previous studies using a different method, and suggest potential gender differences in the psychological effect of using ostracism on others. Some Limitations and Concluding Remarks Although we think that the present study provides important insights into the nature of ostracism, the study has some important limitations. First, these data were collected in 1999 before the dramatic increase in the use of social media. Although it appears that the processes underlying online social contacts are similar to those underlying face-to-face contact (e.g., Kovas & Nezlek, 2014), we cannot be certain that this similarity includes ostracism. There is also the possibility that when a group ostracizes someone, the ostracizers derive benefits that are unique to ostracism versus other types of group action (e.g., Zadro et al., 2005). The nature of our design did not allow us to examine this possibility. We also did not collect any data describing what the ostracism was intended to accomplish (if anything). For example, in terms of emotional outcomes, different types of ostracism might be used to make someone angry vs. jealous vs. guilty and so forth. There is also the issue of the possible influence of the instructional set we used. To provide some level of comparability across participants, they were told what types of behaviors constituted ostracism, which may have biased their perceptions. Although plausible, the existing research suggests that the procedure we used probably did not bias people’s reports (Nezlek,


448


2012; pp. 7–8). Moreover, given that our analyses focused primarily on how people reacted to ostracizing another (not on how often they did this), it is not clear exactly what type of bias our instructional set may have created. Nevertheless, it is possible that our instructional set biased perceptions in ways that have not yet been found in diary research. We also defined ostracism conceptually for participants as being ignored or excluded, whereas the typical conceptual definition involves both components (Williams, 2009). Although such difference may be important for constructing operational definitions and experimental manipulations of ostracism, they may not matter for participants’ subjective experiences of being ignored (vs. being excluded or rejected). Nevertheless, our data cannot speak to such nuances. Along these same lines, our instructions may have biased participants’ reports of the types of ostracism that laypeople may not commonly identify as ostracism (i.e., not-ostracism, oblivious, and role-prescribed). If this was the case, our results concerning these types of ostracism need to be considered in the context of these influences, whatever they might be. Future research should investigate these three types using different research methods (e.g., experimental manipulations; autobiographical recall without using loaded terms such as ostracism) to see if these patterns replicate across paradigms. Nevertheless, although these limitations are important, they do not call into question the validity of what we found, and we hope that this study serves as a useful point of departure for future work. Summary We investigated participants’ reports of using ostracism in their daily lives. We employed an event-contingent diary method and found that participants reported using ostracism in various social situations for diverse reasons. Participants reported that they used ostracism mostly on people of equal status, and they reported experiencing lower levels of belonging but higher levels of control after using ostracism. Participants also reported ostracizing strangers and acquaintances more often than they ostracized individuals with whom they had closer relationships. Punitive ostracism was associated with more positive outcomes than other types of ostracism (e.g., defensive, role-based). We believe that diary methods will be useful ways of studying other aspects of using ostracism as this research area further develops.

FUNDING Preparation of this article was funded in part by a Senior Research Fellowship from the Council for International Exchange of Scholars to John B. Nezlek. Data collection was funded by an Australian Large Research Council Grant (A79800071) and an ARC small grant (“Testing the Visibility Dimension of Ostracism Using an Event-Contingent Self-Recording Method”) awarded to Kipling D. Williams.

NEZLEK ET AL.

449


NOTES 1. In this article we discuss only results of analyses of participants’ descriptions of ostracizing someone. We do not consider the frequency with which they ostracized, e.g., the number of times they ostracized each day. We did not analyze how often people ostracized because there was some ambiguity about this measure. On average, participants returned their completed forms every 2–3 days. Unfortunately, when they did this, they were not asked why there were some days for which they described no events. A missing day could mean that they had not ostracized that day or that they had not maintained the record. Regardless of the assumptions made about missing days, our data suggest that ostracizing someone is a regular occurrence. If days for which no event was recorded are assumed to be days that the record was not maintained, the average number of ostracizing events each day was 1.54 (SD = .88). Depending upon the assumptions made about why days were missing, the mean per day is lower than 1.54, but it is always close to 1.0 or more than 1.0. See Nezlek (2012) for a discussion of estimating the frequency of events in event-contingent diary studies. 2. Williams et al. wrote their chapter assuming that a published article in a refereed journal would follow and supplant it. Given this, we do not compare and contrast the analyses we present and the findings Williams et al. (2001) presented.

AUTHOR NOTES John B. Nezlek is affiliated with the Department of Psychology, College of William & Mary, and with the University of Social Sciences and Humanities, Poznan. Eric D. Wesselmann is affiliated with the Department of Psychology, Illinois State University. Ladd Wheeler is affiliated with the Department of Psychology, Macquarie University. Kipling D. Williams is affiliated with the Department of Psychological Sciences, Purdue University.

REFERENCES Anderson, C. A., & Bushman, B. J. (2002). Human aggression. Annual Review of Psychology, 53, 27–51. doi:10.1146/ annurev.psych.53.100901.135231 Archer, J., & Coyne, S. M. (2005). An integrated review of indirect, relational, and social aggression. Personality and Social Psychology Review, 9, 212–230. doi:10.1207/pspr.2005.9.issue-3 Arriaga, X. B., Slaughterbeck, E. S., Capezza, N. M., & Hmurovic, J. L. (2007). From bad to worse: Relationship commitment and vulnerability to partner imperfections. Personal Relationships, 14, 389–409. doi:10.1111/ pere.2007.14.issue-3 Bastian, B., Jetten, J., Chen, H., Radke, H. R. M., Harding, J. F., & Fasoli, F. (2013). Losing our humanity: The self-dehumanizing consequences of social ostracism. Personality and Social Psychology Bulletin, 39, 156–169. doi:10.1177/0146167212471205 Benenson, J. F., Hodgson, L., Heath, S., & Welch, P. J. (2008). Human sexual differences in the use of social ostracism as a competitive tactic. International Journal of Primatology, 29, 1019–1035. doi:10.1007/s10764-008-9283-4 Benenson, J. F., Markovits, H., Hultgren, B., Nguyen, T., Bullock, G., & Wrangham, R. (2013). Social exclusion: More important to human females than males. PLOS ONE, 8(2), e55851. Benenson, J. F., Markovits, H., Thompson, M. E., & Wrangham, R. W. (2011). Under threat of social exclusion, females exclude more than males. Psychological Science, 22(4), 538–544. doi:10.1177/0956797611402511 Bushman, B. J., Baumeister, R. F., & Phillips, C. M. (2001). Do people aggress to improve their mood? Catharsis beliefs, affect regulation opportunity, and aggressive responding. Journal of Personality and Social Psychology, 81, 17–32. doi:10.1037/0022-3514.81.1.17 Buss, D. M., Gomes, M., Higgins, D. S., & Lauterbach, K. (1987). Tactics of manipulation. Journal of Personality and Social Psychology, 52, 1219–1229. doi:10.1037/0022-3514.52.6.1219 Chen, Z., Poon, K., Bernstein, M. J., & Teng, F. (2013). Rejecting another pains the self: The impact of perceived future rejection. Journal of Experimental Social Psychology, 50, 225–233. doi:10.1016/j.jesp.2013.10.007 Ciarocco, N. J., Sommer, K. L., & Baumeister, R. F. (2001). Ostracism and ego depletion: The strains of silence. Personality and Social Psychology Bulletin, 27, 1156–1163. doi:10.1177/0146167201279008


450


Crick, N. R., & Grotpeter, J. K. (1995). Relational aggression, gender, and social-psychological adjustment. Child Development, 66, 710–722. DeWall, C. N., Twenge, J. M., Gitter, S. A., & Baumeister, R. F. (2009). It’s the thought that counts: The role of hostile cognition in shaping aggressive responses to social exclusion. Journal of Personality and Social Psychology, 96, 45–59. doi:10.1037/a0013196 Faulkner, S., Williams, K., Sherman, B., & Williams, E. (1997, May). The “silent treatment”: Its incidence and impact. Paper presented at the meeting of the Midwestern Psychological Association, Chicago, IL. Goffman, E. (1963). Stigma: Notes on the management of spoiled identity. Englewood Cliffs, NJ: Prentice-Hall. Gooley, S. L., Zadro, L., Williams, L. A., Svetieva, E., & Gonsalkorale, K. (2015). Ostracizing for a reason: A novel source paradigm for examining the nature and consequences of motivated ostracism. Journal of Social Psychology, 155(5), 410–431. Gottman, J. M., & Krokoff, L. J. (1989). Marital interaction and satisfaction: A longitudinal view. Journal of Consulting and Clinical Psychology, 57, 47–52. doi:10.1037/0022-006X.57.1.47 Gruter, M., & Masters, R. D. (Eds.). (1986). Ostracism as a social and biological phenomenon: An introduction. Ethology and Sociobiology, 7, 149–158. doi:10.1016/0162-3095(86)90043-9 Holtzworth-Munroe, A., & Jacobson, N. S. (1985). Causal attributions of married couples: When do they search for causes? What do they conclude when they do? Journal of Personality and Social Psychology, 48, 1398–1412. doi:10.1037/0022-3514.48.6.1398 Joel, S., Teper, R., & MacDonald, G. (2014). People overestimate their willingness to reject potential romantic partners by overlooking their concern for other people. Psychological Science, 25, 2233–2240. doi:10.1177/0956797614552828 Kovas, D., & Nezlek, J. B. (2014). Social compensation and social enhancement in everyday online conversations. Unpublished manuscript, College of William & Mary, Williamsburg, VA. Kurzban, R., & Leary, M. R. (2001). Evolutionary origins of stigmatization: The functions of social exclusion. Psychological Bulletin, 127, 187–208. doi:10.1037/0033-2909.127.2.187 Le, B., & Agnew, C. R. (2003). Commitment and its theorized determinants: A meta-analysis of the investment model. Personal Relationships, 10, 37–57. doi:10.1111/pere.2003.10.issue-1 Legate, N., DeHaan, C., & Ryan, R. (2015). Righting the wrong: Reparative coping after going along with ostracism. Journal of Social Psychology, 155(5), 471–482. Legate, N., DeHaan, C. R., Weinstein, N., & Ryan, R. M. (2013). Hurting you hurts me too: The psychological costs of complying with ostracism. Psychological Science, 24, 583–588. doi:10.1177/0956797612457951 Nezlek, J. B. (2001). Multilevel random coefficient analyses of event and interval contingent data in social and personality psychology research. Personality and Social Psychology Bulletin, 27, 771–785. doi:10.1177/0146167201277001 Nezlek, J. B. (2012). Diary methods for social and personality psychology. In J. B. Nezlek (Ed.), The Sage library in social and personality psychology methods. London, UK: Sage. Nezlek, J. B., Wesselmann, E. D., Wheeler, L., & Williams, K. D. (2012). Ostracism in everyday life. Group Dynamics: Theory, Research, & Practice, 16, 91–104. doi:10.1037/a0028029 Patterson, G. R., & Maccoby, E. E. (1980). Mothers: The unacknowledged victims. Monographs of the Society for Research in Child Development, 45, 1–64. Poon, K.-T., & Chen, Z. (2015). How does the source of rejection perceive innocent victims? Journal of Social Psychology, 155(5), 515–526. Poulsen, J. R., & Carmon, A. F. (2015). Who would do that? A theory-based analysis of narratives of sources of family ostracism. Journal of Social Psychology, 155(5), 452–470. Poulsen, J. R., & Kashy, D. A. (2012). Two sides of the ostracism coin: How sources and targets of social exclusion perceive themselves and one another. Group Processes & Intergroup Relations, 15, 457–470. doi:10.1177/ 1368430211430517 Raudenbush, S. W., Bryk, A. S., Cheong, A. S., Fai, Y. F., Congdon, R. T., & Du Toit, M. (2011). HLM 7: Hierarchical linear and nonlinear modeling. Lincolnwood, IL: Scientific Software International. Reis, H. T., & Gable, S. L. (2000). Event-sampling and other methods for studying everyday experience. In H. T. Reis & C. M. Judd (Eds.), Handbook of research methods in social and personality psychology (pp. 190–221). Cambridge, UK: Cambridge University Press. Rusbult, C. E., Johnson, D. J., & Morrow, G. D. (1986). Impact of couple patterns of problem solving on distress and nondistress in dating relationships. Journal of Personality and Social Psychology, 50, 744–753. doi:10.1037/ 0022-3514.50.4.744 Rusbult, C. E., & Martz, J. M. (1995). Remaining in an abusive relationship: An investment model analysis of nonvoluntary dependence. Personality and Social Psychology Bulletin, 21, 558–571. doi:10.1177/0146167295216002


NEZLEK ET AL.

451

Sommer, K. L., Williams, K. D., Ciarocco, N. J., & Baumeister, R. F. (2001). When silence speaks louder than words: Explorations into the intrapsychic and interpersonal consequences of social ostracism. Basic and Applied Social Psychology, 23, 225–243. doi:10.1207/S15324834BASP2304_1 Sommer, K. L., & Yoon, J. (2013). When silence is golden: Ostracism as resource conservation during aversive interactions. Journal of Social and Personal Relationships, 30, 901–919. doi:10.1177/0265407512473006 Van Lange, P. A. M., Rusbult, C. E., Drigotas, S. M., Arriaga, X. B., Witcher, B. S., & Cox, C. L. (1997). Willingness to sacrifice in close relationships. Journal of Personality and Social Psychology, 72, 1373–1395. doi:10.1037/ 0022-3514.72.6.1373 Van Tongeren, D. R., Root Luna, L. M., & VanOyen Witvliet, C. (2015). Insufficient justification for rejection elicits compensatory behavior. Journal of Social Psychology, 155(5), 527–534. Wesselmann, E. D. (2011). Rehabilitation and protection: Beneficial uses of ostracism in groups (Unpublished doctoral dissertation). Purdue University, West Lafayette, IN. Wesselmann, E. D., Nairne, J. S., & Williams, K. D. (2012). An evolutionary social psychological approach to studying the effects of ostracism. Journal of Social, Evolutionary, and Cultural Psychology, 6, 309–328. doi:10.1037/h0099249 Wesselmann, E. D., Williams, K. D., & Hales, A. H. (2013). Vicarious ostracism. Frontiers in Human Neuroscience, 7, 153. doi:10.3389/fnhum.2013.00153 Wesselmann, E. D., Williams, K. D., & Wirth, J. H. (2014). Ostracizing group members who can (or cannot) control being burdensome. Human Ethology Bulletin, 29, 82–103. Wesselmann, E. D., Wirth, J. H., Pryor, J. B., Reeder, G. D., & Williams, K. D. (2013). When do we ostracize? Social Psychological and Personality Science, 4, 108–115. doi:10.1177/1948550612443386 Wesselmann, E. D., Wirth, J. H., Pryor, J. B., Reeder, G. D., & Williams, K. D. (2015). The role of burden and deviation in ostracizing others. Journal of Social Psychology, 155(5), 483–496. Wheeler, L., & Nezlek, J. (1977). Sex differences in social participation. Journal of Personality and Social Psychology, 35, 742–754. doi:10.1037/0022-3514.35.10.742 Wheeler, L., & Reis, H. T. (1991). Self-recording of everyday life events: Origins, types, and uses. Journal of Personality, 59, 339–354. doi:10.1111/j.1467-6494.1991.tb00252.x Williams, K. D. (1997). Social ostracism. In R. M. Kowalski (Ed.), Aversive interpersonal behaviors (pp. 133–170). New York, NY: Plenum Press. Williams, K. D. (2001). Ostracism: The power of silence. New York, NY: Guilford Press. Williams, K. D. (2009). Ostracism: Effects of being excluded and ignored. In M. Zanna (Ed.), Advances in experimental social psychology (pp. 275–314). New York, NY: Academic Press. Williams, K. D., Bernieri, F. J., Faulkner, S. L., Gada-Jain, N., & Grahe, J. E. (2000). The scarlet letter study: Five days of social ostracism. Journal of Personal & Interpersonal Loss, 5, 19–63. doi:10.1080/10811440008407846 Williams, K. D., Cheung, C. K. T., & Choi, W. (2000). CyberOstracism: Effects of being ignored over the Internet. Journal of Personality and Social Psychology, 79, 748–762. doi:10.1037/0022-3514.79.5.748 Williams, K. D., & Nida, S. A. (2011). Ostracism: Consequences and coping. Current Directions in Psychological Science, 20, 71–75. doi:10.1177/0963721411402480 Williams, K. D., Shore, W. J., & Grahe, J. E. (1998). The silent treatment: Perceptions of its behaviors and associated feelings. Group Processes & Intergroup Relations, 1, 117–141. doi:10.1177/1368430298012002 Williams, K. D., Wheeler, L., & Harvey, J. (2001). Inside the social mind of the ostracizer. In J. Forgas, K. Williams, & L. Wheeler (Eds.), The social mind: Cognitive and motivational aspects of interpersonal behaviour (pp. 294–320). New York, NY: Cambridge University Press. Wirth, J. H., Bernstein, M. J., & LeRoy, A. S. (2015). Atimia: A new paradigm for investigating how individuals feel when ostracizing others. Journal of Social Psychology, 155(5), 497–514. Zadro, L., Williams, K. D., & Arriaga, X. B. (2008). Relational ostracism. In J. P. Forgas, & J. Fitness (Eds.), Social relationships: Cognitive, affective and motivational processes (pp. 305–319). New York, NY: Psychology Press. Zadro, L., Williams, K. D., & Richardson, R. (2005). Riding the ‘O’ train: Comparing the effects of ostracism and verbal dispute on targets and sources. Group Processes & Intergroup Relations, 8, 125–143. doi:10.1177/1368430205051062




Who Would Do That? A Theory-Based Analysis of Narratives of Sources of Family Ostracism a

Joan R. Poulsen & Anna F. Carmon

a

a

Indiana University–Purdue University Columbus Published online: 12 Aug 2015.

Click for updates To cite this article: Joan R. Poulsen & Anna F. Carmon (2015) Who Would Do That? A Theory-Based Analysis of Narratives of Sources of Family Ostracism, The Journal of Social Psychology, 155:5, 452-470, DOI: 10.1080/00224545.2015.1064347 To link to this article: http://dx.doi.org/10.1080/00224545.2015.1064347





Who Would Do That? A Theory-Based Analysis of Narratives of Sources of Family Ostracism JOAN R. POULSEN and ANNA F. CARMON Downloaded by [Ohio State University Libraries] at 08:49 20 August 2015

Indiana University–Purdue University Columbus

ABSTRACT. There are many benefits derived from families, but not all family members are loving and accepting. Family members may act as sources of ostracism (people or groups who ostracize another person/group). We suggest sources engage in family ostracism for extended periods, their motives fit with prior theoretical models, and trait-level forgiveness may help understand source behavior. We analyzed data from 63 narratives and questionnaires to investigate the motives, power dynamics, and psychological correlates of sources of family ostracism. We found sources of ostracism are often of equal status to the targets of ostracism, and termination often occurs informally or is prompted by major changes in the family (e.g., birth, move). Also, sources of ostracism are often targets themselves suggesting family ostracism may be reciprocal in nature. Our findings support existing theory, but suggest ostracism in families has unique dynamics not captured in laboratory designs. Keywords: exclusion, family, ostracism, sources of ostracism

MANY CONSIDER FAMILY RELATIONSHIPS TO BE among the most central, defining relationships in one’s life. Families can serve as groups that help form one’s identity and serve as a strong support system (Zimet, Dahlem, Zimet, & Farley, 1988). However, not every family member is always loving, supportive, or accepting of each other. Just like any other group, family members can be unpleasant, critical, or downright abusive. Among one of the array of aversive interpersonal behaviors possible in a family is social ostracism, the act of an individual or group ignoring and excluding other individuals or groups (Williams, 1997, 2001). The process of social ostracism has been well studied in laboratory settings with a strong focus on targets (those who are ostracized), but less work has focused on understanding everyday groups, with even less research focusing on the sources of ostracism (those who ostracize another person/group). We lay a foundation for understanding the basic processes and phenomenon of sources in a specific context: family ostracism. We describe aspects of the phenomenon of family ostracism and the motivations and power dynamics in families. Additionally, we offer a comparison between participants who claim to have never been a source of family ostracism with those who have, and we explore how sources of family ostracism are impacted.

Address correspondence to Joan R. Poulsen, Indiana University–Purdue University Columbus, Department of Science/Psychology, 4601 Central Ave., Columbus, IN 47203, USA. E-mail: [email protected]

POULSEN AND CARMON

453


FAMILY OSTRACISM Zadro, Arriaga, and Williams (2008) and Fitness (2005) specifically examined ostracism and interpersonal rejection within a family context. Zadro et al. (2008) looked at real-world ostracism broadly by exploring real-world ostracism using in-depth interviews with a variety of sources and targets of ostracism. Of those interviewed, they found that the vast majority (both targets and sources) discussed ostracism by a close family member. In particular, for targets, 60% noted experiencing ostracism from a spouse and 28% noted experiencing ostracism from a mother/mother-in-law. These results suggest ostracism can commonly occur in families. They also found being a target of ostracism from family or friends led to negative outcomes, such as somatic symptoms (i.e., nausea, upset stomach), self-destructive thoughts, health issues, strains on their social lives, and an inability to form new bonds due to insecurity. Ostracism from closeothers can have a severe impact on targets, particularly when considering the opportunity for ostracism to last for an extended period in these contexts. Fitness (2005) also examined how family dynamics relate to ostracism. Specifically, she investigated how families reject “black sheep”—family members considered troublemakers or misfits. Fitness (2005) suggested that black sheep often violate serious family norms, such as abandoning dependents, engaging in criminal behavior, and/or committing sexual abuse or taboos. Black sheep experience various types of rejection such as pointed criticism, sarcasm, hostility, and outright coldness; this rejection is likely punitive in nature. Thus, even though Fitness (2005) did not explicitly look at family ostracism, it is possible that black sheep may be ostracized as punishment for perceived transgressions. Family Systems Theory Researchers have not extensively examined family ostracism in an empirical context, but clinicians in family therapy have acknowledged the presence of ostracism and exclusion in families, particularly in the family systems approach (Bowen, 1978). Family systems theory posits that various forms of ostracism, exclusion, and rejection occur regularly within families. For instance, when an adult child moves away from parents or when family members choose to avoid discussing sensitive topics, family systems theory refers to such behaviors as part of an emotional cutoff. Social psychologists may consider such behaviors as forms of ostracism (Agllias, 2013). Agllias (2013) reported that estrangement among family members, particularly between parents and a child/offspring, can have effects similar to those of ostracism. Specifically, Agllias (2013) noted that immediate effects for targets include numbness and feeling hurt, similar to what Blackhart, Nelson, Knowles, and Baumeister (2009) found. The long-term impact can include ongoing anxiety and stress (Agllias, 2013), similar to Zadro’s (2004) findings in her interviews with targets. Finally, family systems theorists suggest no difference in impact on sources or targets of ostracism in situations of cutoff or estrangement, rather that both parties experience negative effects.

SOURCES OF OSTRACISM Research about the effects of ostracism on the source is scarce; however, existing research suggests some distinctions from the experiences of targets. Nezlek, Wesselman, Wheeler, and


454


Williams (2012) conducted a diary study over a 2-week period examining ostracism. This study provided findings about ostracism in daily experiences; including that ostracism from a variety of sources is prevalent in daily life. Findings from Nezlek et al. (2012) suggest most real-world ostracism is done by strangers, acquaintances, or friends, and only 22% of ostracism is performed by close others (i.e., close friend, partner, or relative), suggesting that ostracism is somewhat rare. Thus, the present study will use a retrospective method to investigate the prevalence of family ostracism ever occurring, since ostracism specific to the family context may be difficult to capture in a given 2-week period. Various accounts suggest that sources use ostracism as a means of obtaining power over a target. Nezlek et al. (2012) addressed the role social power can play in ostracism and how it often encouraged conformity within groups (see also Wesselmann, Wirth, Pryor, Reeder, & Williams, 2013). For instance, several sources explained they use ostracism as a way to get the upper hand in a conflict (Buss, Gomes, Higgins, & Lauterbach, 1987; Sommer, Williams, Ciarocco, & Baumeister, 2001; Williams, Shore, & Grahe, 1998; Zadro et al., 2008). Nezlek et al. (2012) reported that targets were mainly ostracized by people of equal status (78%), demonstrating that in everyday ostracism, sources are likely to be peers. Peers may use ostracism to gain power, whereas those with more or less power than a target refrain from ostracism due to existing status differences. In the present study, we seek to determine if there is a similar pattern in accounts of ostracism specific to the family context, given that power in families can be complex. For the sake of simplicity, we define equal status to be family members from the same generation, higher status to be family members of an older generation, and lower status to be family members in a younger generation.

MOTIVES OF OSTRACISM In addition to power dynamics, Nezlek et al. (2012) examined motives of ostracism in everyday life using Williams’ (1997, 2007) model of five motives for ostracism: role-prescribed (situational role prescribes ostracism), punitive (deliberate, aversive, punishment), (ego-) defensive (preemptive in anticipation of social threat), oblivious (target is unworthy of attention such as lower status caste/class), and not ostracism (unintentional). Little work to date has addressed how these motives play into the dynamics of ostracism (but see Nezlek, Wesselmann, Wheeler, & Williams, 2015, for an exception). Nezlek et al. (2012) found that the prevalence of motives of ostracism were fairly equally distributed between the five motives, although defensive ostracism was least commonly reported (Nezlek et al., 2012; Williams, 1997, 2001). Nezlek et al. (2012) found that the motive of ostracism perceived by targets did influence outcomes for the target, such as selfesteem. Specifically, when examining outcomes of self-esteem, control, and meaning existence, targets who experienced punitive or oblivious ostracism were more negatively impacted relative to other motives. Researchers examined sources of ostracism using laboratory studies. In a study by Poulsen and Kashy (2012), they randomly assigned participants to the role of a source or target of ostracism, provided no motive for ostracizing the target, and then had the participants perform a 10-minute brainstorming task. After the task, sources noted liking other sources more than they liked the target and more than the target liked sources. Additionally, sources reported emotions that were more positive and fewer negative emotions than targets, except for guilt. Further, Wesselmann

POULSEN AND CARMON

455

et al. (2013) found people are more reluctant to ostracize when they are not provided a motive, thus if instructed to ostracize without a motive, sources may experience increased feelings of guilt.


DESCRIBING FAMILY OSTRACISM Given the limited scope of research focused specifically on sources of family ostracism, the present study seeks to discover and describe this specific form of ostracism and to explore some of the outcomes of family ostracism for sources. One basic aspect of the phenomenon we seek to describe specifically for families is duration of ostracism. In most laboratory studies, ostracism lasts just a few minutes, yet in prior work, researchers have documented some individuals experience daily ostracism lasting for several decades (Zadro, 2004). Although some cases of decades-long ostracism exist, we expect that these are rare, and that most instances of family ostracism may last only a short time. As Nezlek et al. (2012) stated, some forms of ostracism are poignant and memorable, but not frequent in occurrence.

SOURCES AND SELF-ESTEEM Self-esteem, which indicates how positively individuals evaluate themselves (Baumeister, 1998), tends to be lower in targets of ostracism. Little work, however, has examined how self-esteem may be linked to ostracizing another person, especially a close family member. Bullying research suggests sources of bullying often suffer from low self-esteem (O’Moore & Kirkham, 2001). Research on romantic couples finds that when problems occur, partners with lower self-esteem were more likely to derogate their partner than partners with higher self-esteem—an ironic behavior given that partners with low self-esteem typically also had chronic acceptance needs (Murray, Rose, Bellavia, Holmes, & Kusche, 2002). These correlational data converge with experimental data comparing sources and targets of ostracism. Specifically, researchers using induced ostracism paradigms found links between using ostracism and self-esteem. Zadro, Williams, and Richardson (2005) found that sources experienced a decrease in state self-esteem, although not nearly as low as targets (c.f., Poulsen & Kashy, 2012). Taken together, the research on trait and state self-esteem suggests ostracism lowers individuals’ self-evaluations.

SOURCES AND DEPRESSION Due to the possibility for family ostracism to disrupt a critical social bond, and to do so for an extended period, it seems likely that sources of ostracism may suffer greatly. We begin to explore this notion by examining if sources of ostracism feel higher levels of depression, a mental disorder involving sadness, lack of interest, and feelings of worthlessness, among other symptoms (Kazdin, 2000). As both Leary (1990) and Williams (2007) stated, targets of ostracism are susceptible to depression; yet the consequences of initiating the disruption of a social bond may also signal loss and trigger a sense of depression for sources of ostracism. Related work about bullying found that both targets and sources of adolescent bullying reported higher levels of

456


depression than those who were neither sources nor targets of this behavior (Seals & Young, 2003). Similarly, after a break-up, feelings of depression occur similarly both for the person who initiated the break-up, as well as the person who did not initiate the break-up (Boelen & Reijntjes, 2009). Research about bullying and break-ups suggests the initiator of unpleasant behavior still suffers emotionally; therefore, we examine if sources of family ostracism also might experience depression.


SOURCES AND FORGIVENESS Thompson et al. (2005) considered dispositional forgiveness a trait, rather than an isolated action. Forgiveness involves reframing actions from a transgressor to be neutral or positive; as such, people who are more forgiving override a more automatic reaction to seek retribution. In enduring relationships (i.e., friendships, romantic partnerships, families), conflicts are nearly a certainty. Forgiveness is one way of moving beyond relational transgressions and to help refocus away from transgressions (Fincham, Beach, & Davila, 2004, 2007). Clinical research demonstrates forgiveness benefits victims of abuse and recovering alcoholics (Webb, Hill, & Brewer, 2012), but little work has explored if forgiveness might moderate sources’ motives for and experiences of ostracizing others. Of note, one documented motive for ostracism is to punish (Williams, 1997). When a family member perceives a transgression, they have many options in how to respond, including ostracism. Fitness (2005) noted that families often use extreme forms of social rejection on members who commit severe and unforgiveable transgressions. Thus, dispositional forgiveness may influence how willing or hesitant someone is to use ostracism on a family member. In the present study, we seek to understand how forgiveness may be related to different outcomes for sources.

HYPOTHESES After reviewing existing work on sources and family ostracism, we focused our study on a few key areas: description of the phenomenon, associated constructs, and dispositional factors. First, we measured prevalence, duration, and source to describe family ostracism. Based on past work about ostracism in close relationships (Zadro et al., 2005), we expected that 50%–80% of our sample would report ostracizing a family member at some point in their lives. We assessed how long ostracism lasts according to the sources. We used this exploratory question to determine what a common experience of ostracism is like: Does it tend to last hours, days, weeks, or years? Additionally, who tends to ostracize whom in familial ostracism? Based on past work demonstrating that much of everyday ostracism involves peers (Nezlek et al., 2012), we expected that sources in families will be more likely to ostracize equal- (i.e., same generation) or lower-status (i.e., younger generation) members rather than higher-status (i.e., older generation) members. We gathered additional descriptive information including motives for ostracism. Williams (1997, 2001) suggested that there are five possible motives for ostracism: not ostracism, role prescribed, punitive, defensive, and oblivious. It seems based on prior work (Nezlek et al., 2012), punitive ostracism may be the most common in families. We also conducted an exploratory analysis looking at what prompts ostracism to end. In addition to describing ostracism, this study sought to examine how family ostracism is associated with self-esteem and depression symptoms. We predicted that results would mirror

POULSEN AND CARMON

457

those of Zadro et al. (2005, 2008), which demonstrated a decrease in self-esteem for sources compared to nonsources, and that long-term ostracism was linked with depressive symptoms. Finally, we explored if more forgiving people are less likely to report having ever ostracized a family member, if they tend to have shorter instances of ostracism, and if they will be less likely to have punitive motives than less forgiving people.


METHOD Participants Participants in this study initially included 279 individuals (195 females, 76 males, 8 who chose not to report their biological sex) who completed an online survey. We recruited participants from psychology and communication studies classes and more broadly via social media. In this study, we randomly assigned1 participants to write about either being a target, source, or witness of family ostracism. In total, 69 participants were randomly assigned to write about an experience as a source (52 females, 16 males, 1 not reporting), and 68 participants provided a narrative about their experiences being a source of family ostracism. Of these, 5 participants did not provide sufficient detail for any analysis, and 59 participants provided enough information for complete content analyses2 . The remainder of the sample did not answer questions about being a source. Because some analyses involve the full sample, and others involve only sources, we provide descriptive statistics for both groups. For the full sample, the average age was M = 24.81, SD = 8.24, with a range from 18 to 59 years of age; for sources only it was M = 25.12, SD = 9.16, with a range from 18 to 54 years of age. In terms of ethnicity, 92.1% of the full sample was Caucasian/White (91.4% for sources only), with all other ethnic categories having six or fewer persons per category (true also for subsample of sources). In terms of family demographics, 64.9% of the full sample were single/never married (61.4% for source sample), 26.5% were married (30% of sources), and 7.2% were divorced/widowed (5.7% of sources), with 1.4% not reporting (2.9% of sources). In the full sample, 35.8% were parents (38.6% of sources), 62.7% were not (58.6% sources), with 1.4% not reporting (2.9% of sources). Of our parents in the full sample (n = 100), 38 had one child, 34 had two children, 19 had three children, and 10 reported having four or more children; of parents in the sample of sources (n = 27), 10 had one child, six participants had two children, six participants had three children, and five participants had four or more children. Procedures Participants completed an online survey via Surveymonkey.com lasting 20–30 minutes. We randomly assigned participants to either answer questions about experiences of ostracism as a target, source, or witness of family ostracism. The focus of the present study is about sources of ostracism and data about targets and witnesses of family ostracism were used in other studies. Acknowledging that perhaps not every participant had experienced family ostracism in the way we randomly assigned them, we then screened participants for eligibility again by asking if they had excluded another family member before. We automatically took participants who responded they had not excluded any family member to the demographic questions of the study and did not have them respond to the open-ended questions. We directed those who affirmed that they had

458


been a source to more detailed questions about their experiences as a source. Four independent raters coded narratives and open-ended responses from sources. After coding independently, the four coders met to discuss their ratings, and came to consensus after discussion.


Measures Family ostracism. To measure overall perceptions of family ostracism, we used a revised version of Carter-Sowell’s (2010) eight-item Ostracism Experience Scale. All items were retained, but the wording of questions was changed from “in general” to “in my family” in order to have participants consider their families specifically. The scale was administered using endpoints of 1 (hardly ever) to 5 (almost always), with higher values indicating more severe ostracism experiences. For this scale, M = 1.32, SD = .57 (M = 1.43, SD = .59 for source sample), and Cronbach’s α = .91 (α = .90 for source sample). Sample items include, “In general, my family treats me like I’m in solitary confinement,” and “In general, my family ignores me during conversation.” Depression symptoms. To assess depression symptoms, we used the Beck Depression Inventory (Beck, Steer, & Brown, 1996). Participants responded using a 3-point scale, with higher values indicating higher levels of depressive symptoms. For this scale, raw scores were used with, M = 10.55, SD = 9.04 (M = 12.00, SD = 9.52 for source sample), and Cronbach’s α = .92 (α = .92 for source sample). An example item with responses is: “Concentration difficulty—0: I can concentrate as well as ever, 1: I can’t concentrate as well as usual, 2: It is hard to keep my mind on anything for too long, and 3: I find I can’t concentrate on anything.” Self-esteem. To measure trait-level self-esteem, we used the Rosenberg (1965) Self-Esteem Scale. The 10-item scale was administered on a 5-point scale ranging from 1 (strongly disagree) to 5 (strongly agree), with higher values on the scale indicating higher levels of self-esteem. For this scale M = 2.33, SD = .80 (M = 3.20, SD = .29 for source sample), and Cronbach’s α = .86 (α = .78 for source sample). Example items include: “On the whole, I am satisfied with myself,” and the reverse-scored item, “All in all, I am inclined to feel that I am a failure.” Forgiveness. The Heartland Forgiveness Scale (HFS; Thompson et al., 2005) is an 18-item scale that measures how forgiving a person is in general. It contains three subscales with six items each to assess forgiveness of different perceived sources of transgressions: forgiveness of oneself, others, and the situation. We summed these subscales to create an overall measure of forgiveness. The self-forgiveness scale contains items such as “I hold grudges against myself for negative things I’ve done” (reversed), and has M = 4.81, SD = 1.06 (M = 4.60, SD = 1.08 for source sample), and Cronbach’s α = .76 (α = .77 for source sample). The other forgiveness scale contains items such as “I continue to be hard on others who have hurt me” (reversed), and has M = 4.87, SD = 1.08 (M = 4.80, SD = 1.13 for source sample), and Cronbach’s α = .83 (α = .84 for source sample). The situational forgiveness scale contains items such as, “With time I can be understanding of bad circumstances in my life,” and has M = 4.77, SD = 1.05 (M = 4.77, SD = 1.00 for source sample), and Cronbach’s α = .79 (α = .78 for source sample). The total forgiveness scale of all 18 items has M = 4.82, SD = .88 (M = 4.72, SD = .86 for source sample), and Cronbach’s α = .88 (α = .88 for source sample).


POULSEN AND CARMON

459

Narratives of family ostracism. We asked participants to respond to a series of open-ended questions about their experiences as a source of ostracism. We elected to use the term “exclusion” as part of our prompts, rather than ostracism, because we found it was more understandable to laypeople. We obtained descriptions of sources of family ostracism experiences by asking participants to respond to the open-ended statement: “Please explain who you excluded, what the exclusion entailed, and how long it went on.” We inquired about motive(s) of ostracism by asking participants to respond to the open-ended question: “What do you think caused you to exclude this person?” In trying to understand processes that end ostracism, we asked participants the open-ended question: “What do you think caused the exclusion to end? Please explain if and how you and the family member(s) reconciled. If the exclusion is going on right now, please note this.” Finally, to address potential issues of memory biases, we asked people to explain: “When did the exclusion happen—how long ago?” Coding procedures. We gave four independent coders the open-ended responses of the sources, unedited, and then coders completed a code sheet about all sources. Both authors served as coders and the other two coders were upper-level undergraduates. All coders read Williams’ (1997, 2007) theoretical papers on ostracism, particularly to grasp differences in motives of ostracism, in order to be able to effectively code the data. The code sheets used required each rater to code eight details about the narratives: length of ostracism in days, identity of the source, generation of the source (older, same, younger), motive (not ostracism, role prescribed, punitive, defensive, oblivious), if ostracism is ongoing or ended; cause of ostracism to end, how long ago it occurred, the coder’s perceived severity of ostracism. After coding all items independently, the four coders met and discussed discrepancies. When there was disagreement, all coders explained their rationale and then each coder marked their judgment again on a paper code sheet privately to reduce conformity or authority influence. We acknowledge the possibility for bias because both authors were coders. However, we argue that having greater expertise with both coding and relevant theoretical perspectives motivated us to code in order to explore any unanticipated themes in the data. Most of our a priori hypotheses were more questions than directional hypotheses, reducing the impact of potential biases. Finally, because we are from different disciplines, we often found ourselves engaging in more discussion to explain our coding clearly. In order to determine interrater reliability, we calculated two different statistics for both the first round of coding and after a discussion aimed at coming to a consensus. Because more than two raters coded interval/ratio data, we used intraclass correlation (ICC) to calculate interrater reliability for the constructs of length and severity of ostracism (Hallgren, 2012). We used Light’s kappa to determine interrater reliability for all other items. Light’s kappa assessed interrater reliability for categorical data in the case of more than two raters (Hallgren, 2012). Length or duration of family ostracism proved to be difficult to summarize because of the extreme variety of responses. We asked participants how long the ostracism experience lasted, and we transformed these responses into days as units of analysis. Thus, the response of 1 month became 30 days. ICCs revealed relatively good interrater reliability for initial (ICC = .99) and second (ICC = 1.00) rounds of coding. The next question was about who in a family tends to ostracize whom. These data were gleaned from open-ended responses and were used by independent coders to determine if the source was of an older generation (i.e., parent, grandparent, aunt/uncle, etc.), same generation (i.e., sibling, cousin, etc.), or younger generation (i.e., child, grandchild, niece/nephew, etc.). Light’s kappa revealed relatively good agreement in the first round of coding (κ = .84), with improvement after discussion (κ = .97).


460


Participants were asked why they thought the ostracism started (motive). We coded the answers using Williams’ (1997) theoretical model of ostracism. We coded motives as 1 = not ostracism, unintentional/accidental; 2 = role-prescribed, that a role temporarily sanctions ostracism such as not including a restaurant server into your conversation; 3 = punitive, act that is deliberate and aversive, intended to punish; 4 = defensive, an act that is pre-emptive in anticipation of negative interaction, protects the person from the target; 5 = oblivious, the targets is unworthy of attention; or 6 = other, for cases which did not fit in any of the five categories. We used Light’s kappa to test interrater reliability, and found initial agreement to be poor (κ = .21),3 with improvement after discussion (κ = .97). We explored what caused the termination of ostracism. With little theory to guide the coding of these responses, we were guided by reading responses from Zadro’s (2004) four themes, which suggested several reasons for ostracism to end: (a) passage of time, (b) spiritual or religious reasons, (c) apology by the target, (d) apology by the source, (e) apology not specified, (f) family intervention, (g) unknown/not stated, (h) other category, and (i) ostracism has not yet ended. We coded many responses as the “other” category, which often involved a change in the family situation, and some categories had few or no narratives that fit. Because of this difficulty, initial interrater reliability was quite low (Light’s κ = .22), but after discussion and referring to past work by Williams (2007), reliability improved substantially (Light’s κ = .97). Coders assessed if ostracism was still ongoing or not. They also assessed how long ago the ostracism incident occurred, in years. On average, narratives were written about ostracism done a little over a year ago (M = 1.27, SD = 2.82). Given the variance in the data, we explored median and mode as more accurate measures of central tendency. When responding to the question “When did this happen—how long ago?,” subjects reported how many years ago the ostracism incident had occurred and both revealed that most responses were either ongoing or ended less than a year ago (Mdn = 0 years; Mode = 0 years). Interrater reliability on whether ostracism was ongoing was acceptable initially (Light’s κ = .88), with improvement after discussion (Light’s κ = .98). In our sample, 42 narratives involved ongoing ostracism or ostracism ending less than 1 year ago. Finally, the coders provided their overall perception of severity of the ostracism incident. As one way to examine validity of this assessment, the correlation between coded severity and days of ostracism was assessed (r = .54, p < .001), suggesting that although length of ostracism is associated with severity (as one would expect), it is not interchangeable. Interrater reliability of this construct was assessed using ICC, with good reliability initially (ICC = .88), and substantial improvement after discussion (ICC = 1.00).

RESULTS We divided the results into three main sections: first, descriptive information about sources of family ostracism; then associated constructs (self-esteem and depression symptoms); finally, we address the dispositional factor of trait-level forgiveness. Describing Family Ostracism We addressed the five features describing sources of family ostracism: prevalence, duration, role of the source in the family, motives for ostracizing, and reasons for terminating ostracism.


POULSEN AND CARMON

461

We asked all 279 participants if they had ever been a source of family ostracism, using the question, “Have you ever excluded another family member?” We found that 39.1% (or n = 109) of participants reported being a source of ostracism at some point. Relatedly, of the 279 sampled, 42.3% reported being a target of family ostracism, and 65.9% indicated that they had witnessed family ostracism at some point. To examine the post hoc possibility that ostracism is a pattern in families—that is, those who report being sources are likely to also be targets—a chi-square analysis revealed that those who reported being targets were also more likely to report being sources than expected, χ 2 = 38.25, p < .001, φ = .37. The duration of family ostracism, as reported by sources was M = 963.13 days, SD = 1,546.37. Due to the large variance in the sample, we consider the median to be a better measure of central tendency describing family ostracism. The median number of days sources reported ostracizing a family member was 182.5, or about six months. The modal response was 1 day, with nine people reporting ostracism lasting a day in length. Data in our sample range from reports of under 1 day of ostracism to 6,570 days (18 years). In examining who the source of ostracism is in a family, power dynamics were addressed by examining the generation of the source relative to the target: younger, same generation, or older. Of the 59 cases with enough information to analyze, most reported that the source was either the same generation as the target (40.6%) or a younger generation than the target (47.4%), with only 10.2%, reporting the source was of an older generation than the target. Regarding motives of ostracism, independent coders found that in the 63 narratives on sources of family ostracism (that could be coded), sources indicated that they engaged in ostracism largely for punitive reasons (63.5%), with the second-most common motive being defensive (25.4%). Of the remaining, three responses were “not ostracism,” two were role-prescribed, and two were oblivious. Because some motives were quite rare in our data, there was not sufficient power to conduct precise comparisons between all five motives as we had intended. Instead, we focused later analyses on the two most frequently occurring motives, punitive and defensive. Investigating what sources perceived caused the termination of the ostracism, initial coding suggested six possible ways that ostracism could end. Of the 59 narratives with enough information to code, coders found it was most common that the narrative was about an ongoing ostracism experience (n = 26). Another common theme was that ostracism was ended by the passage of time (n = 12), characterized by responses such as, “I just forgot to be angry with her.” Six respondents did not provide a reason why the ostracism ended. This left 11 responses that did not match our initial coding. Within these responses, a common thread was that something major about the family situation changed such as birth, death, or moving away. For instance, one participant stopped ostracizing her father because: “I gave birth to my son and I wanted him to have a normal family.” In another case, ostracism against a grandparent ended because “Grandpa passed away.” Finally, another participant stopped ostracizing her sister because “she moved out and we eventually grew closer.” Self-Esteem and Depression Symptoms in Sources of Family Ostracism Turning to the relation between constructs of self-esteem, depressive symptoms, and other measured constructs (severity of ostracism, duration, Family Ostracism Scale), the results in Table 1 demonstrate that self-esteem and depression symptoms were unrelated to any of the three aspects of ostracism coded from the narratives (i.e., duration, how long ago it occurred, and severity).


462


Predictably, there was an inverse relationship between self-esteem and depression. Self-esteem was negatively associated with higher scores on the Family Ostracism Scale. Depression was positively correlated to more experiences with family ostracism. In considering if there are differences between people who reported being a source of family ostracism from those who report not being a source, we used a MANOVA to conduct an omnibus test comparing sources and nonsources on several factors. We elected to use a MANOVA to reduce Type 1 error and, from our theoretical perspective, to acknowledge the potential for criterion variables in these analyses to be related (e.g., self-esteem may be inversely linked to depression): a MANOVA will statistically account for these issues. As shown in Table 2, there were no differences between participants who reported being a source of ostracism and those who reported never being a source of ostracism in self-esteem or depression symptoms. One difference emerged as an exploratory analysis: those who reported being a source also had higher scores on the Family Ostracism Scale than those who reported never using ostracism on family members. In addition, comparing the punitive and defensive motives of ostracism (the two most common), again using a MANOVA, sources who used ostracism for punitive reasons had higher self-esteem than those using it for defensive motives. The inferential and descriptive statistics are located in Table 3. Forgiveness We explored dispositional forgiveness as it pertained to sources of family ostracism in three ways. First, we conducted analyses to examine associations between forgiveness and other variables in the study. As shown in Table 1, there were consistent patterns showing that as people increased in their level of forgiveness, they tended to experience less family ostracism, had less severe depression symptoms, and had higher self-esteem. When relating dispositional forgiveness with coded variables, we found several marginally significant correlations (see Table 1) including positive associations between duration of ostracism and both total dispositional forgiveness (r = .27, p = .053) and forgiveness of others (r = .26, p = .064), and an association between forgiveness of oneself with severity of ostracism (r = .22, p = .088). Next, we were particularly interested if forgiveness was associated with being a source (or not), and if it was linked with different motives of ostracism. To conduct multiple comparisons with forgiveness as the criterion variable, we used a MANOVA to organize our analyses and account for Type 1 error. The MANOVA indicated no differences between sources and nonsources in dispositional forgiveness (see Table 2). Using a MANOVA to compare the two most-common motives for ostracism in our sample, punitive and defensive, we found that sources differed on how forgiving they were of others (not themselves, the situation, or in general), such that those who ostracized for punitive reasons were less forgiving of others than those who ostracized for defensive motives (see Table 3).

DISCUSSION This study provided descriptive information about being a source of ostracism in families, examined associated constructs of self-esteem and depression, and explored how forgiveness might influence sources. We gathered several pieces of information to describe what family ostracism is

463

.27† .22 .26† .18

−.05 −.10 .10 −.13

−.01 .54∗ .02 −.06 .08 .15 .22† .03 .11

How long ago Coded: Duration Family Ostracism Scale Self-esteem BDI Forgiveness Total Self Other Situation − −.48∗ .30∗ .35∗ .03 .35∗

−.39∗ −.39∗ −.22† −.31∗

Self-esteem

− −.24† .33∗

Family Ostracism Scale

−.54∗ −.53∗ −.18 −.58∗

−

BDI

− .76∗ .77∗ .86∗

Total

Note. BDI = Beck Depression Inventory. Analyses were conducted using only participants who began the narrative; n = 51–63. † p values between .06 and .10. ∗ Denotes significance at p < .05 level.

− −.20 −.01 −.07

− −.08 .07 −.23 .10

Severity

Variable

Duration

How long ago

TABLE 1 Correlations Between Coded Variables and Self-Reported Constructs


− .28∗ .52∗

Self

Forgiveness

− .55∗

Other

464


TABLE 2 MANOVA Results Comparing Differences Between Participants Who Reported Ever Being a Source of Ostracism in Their Family With Those Who Reported Never Being a Source of Ostracism in Their Family Never a source (n = 91)


Variable Self-esteem Depression Family ostracism Forgive Total Self Other Situation

Source (n = 70)

M

SD

M

SD

F (1, 159) MANOVA

3.16 10.22 1.24

0.28 9.26 0.50

3.20 12.00 1.43

0.29 9.52 0.59

F = 0.86, p = .355, partial η2 = .01 F = 1.43, p = .234, partial η2 = .01 F = 5.04, p = .026, partial η2 = .03

4.87 4.88 4.93 4.79

0.91 1.10 0.97 1.09

4.72 4.60 4.80 4.77

0.86 1.08 1.13 1.00

F = 1.03, p = .311, partial η2 F = 2.56, p = .112, partial η2 F = 0.65, p = .421, partial η2 F = 0.02, p = .899, partial η2

= .01 = .02 < .01 < .01

Note. Higher numbers indicate higher scores on the measure reported, thereby indicating stronger endorsement of the construct.

TABLE 3 Comparing Sources Based on Motive of Ostracism Punitive motive (n = 40) Variable Self-esteem Depression Family ostracism Forgive Total Self Other Situation

Defensive motive (n = 16)

M

SD

M

SD

F (1, 159) MANOVA

3.28 11.75 1.42

0.26 9.53 0.59

3.10 12.37 1.46

0.32 10.54 0.59

F = 4.64, p = .036, partial η2 = .08 F = 0.05, p = .830, partial η2 < .01 F = 0.06, p = .810, partial η2 < .01

4.64 4.62 4.62 4.67

0.83 1.03 1.13 0.92

4.98 4.61 5.40 4.93

0.89 1.25 0.89 1.23

F = 1.87, p = .177, partial η2 F = 0.00, p = .995, partial η2 F = 6.10, p = .017, partial η2 F = 0.74, p = .393, partial η2

= .03 < .01 = .10 = .01

Note. Higher numbers indicate higher scores on the measure reported, thereby indicating stronger endorsement of the construct.

like from a source’s perspective. Among the most basic areas related to sources of ostracism we explored was an examination of prevalence. Compared with past work, in which targets reported being ostracized by romantic partners or family about 22%–24% of the time (Nezlek et al., 2012), our study found 39.1% of participants acknowledged that they had engaged in being a source of family ostracism at some point. Clearly, this may reflect the different approaches in measurement. Past work used experience-sampling methods from the target perspective, and it is quite possible that either family ostracism did not occur often during the sampling window or that targets may


POULSEN AND CARMON

465

report these events differently than sources. Our results suggest family ostracism is an experience many have, particularly when using a longer period to capture the event. Relatedly, we noted that a majority of participants (65.9%) reported witnessing ostracism among family members. As Wesselmann, Williams, and Hales (2013) reviewed, such forms of vicarious ostracism can have negative effects on the witness, particularly if the witness has a close relationship with the target. Given that familial relationships tend to be close and enduring, the effects of vicarious family ostracism could be even more profound than that of ostracism among coworkers or classmates. Future research should investigate this possibility. Based on our findings that those who report being a source are more likely also to be a target, it seems ostracism is a two-way street in families. These findings replicate Williams’ (2001) study, which found several cases in which sources were also targets. More specifically, our results relate to Zadro (2004, Zadro et al., 2008) who found similar results among those in close relationships. Thus, it is less common only to be a victim or perpetrator in a family; more often an individual is both. This pattern relates to an established set of findings in couples that show that once negative behaviors are reciprocated, the negative behaviors can result in a toxic cycle (Rusbult, Johnson, & Morrow, 1986)—a pattern that becomes more ingrained the more it is enacted. This mirrors bullying research, where sources may be people who are only bullies or only victims of bullying, but about 10% are bully–victims—people who are both sources and targets of bullying (O’Moore & Kirkham, 2001). When family members engage in estrangement and cutoff, the family systems approach suggests there may be a family member who begins this process, but such processes involve family members being both sources and targets of exclusionary processes. Future work should examine these dynamics more carefully and examine different constructs related to the experience of sources. Our study replicated Nezlek et al. (2012) who sampled broadly. Similarly, we found someone of equal status, with targets and sources being from the same generation, was often a source of ostracism. Ostracism is a phenomenon which occurs most commonly within those of the same status, a finding which Williams (1997) posited nearly 20 years ago. We posit two possible explanations for the performance of ostracism. First, social norms, which guide behaviors between people of different statuses, may include ostracism as a normative behavior as individuals cannot detect it as readily. For instance, if an executive in a company does not socially engage with every employee, employees may not consider this ostracism, rather as normative behavior. It seems possible that ostracism between individuals of different statuses will generally be less noticeable because it is expected, which may be why ostracism among individuals of the same status may be more noticeable, because it is less expected. Secondly, individuals may explain this phenomenon more simply by considering that people generally spend more time with people of equal status than different statuses. Certainly, these ideas are something future research should examine more thoroughly. Unlike Nezlek et al. (2012), who found that motives of ostracism were equally distributed when reported by the targets, or Nezlek et al. (2015) who also found motives were nearly equally distributed when reported by sources, we found two motives were more commonly used in family ostracism: punitive and defensive. Of note, Nezlek et al. (2015) found that sources of ostracism in everyday life reported having a defensive motive less often than other forms, whereas we found a defensive motive to be more prevalent in the family-specific context of ostracism. When sources engage in ostracism, they risk damage to important social bonds which feed source’s own needs for belonging (Baumeister & Leary, 1995). Individuals may be willing to take these risks in


466


order to punish a family member or defend themselves. Wesselmann, Wirth et al. (2013) recently demonstrated that people are willing to ostracize a burdensome group member for punitive reasons. Little theory is currently available to guide our findings about the termination of ostracism. Contrary to our expectations that sources or targets of ostracism would apologize, we found that ostracism seemed to dwindle over time, rather than having a marked ending. Family members eventually reconnected. However, a major change (e.g., birth, death, marriage, and career change) in the family was another common way that ended ostracism. One possible explanation may be that these are all events where family norms are disrupted and ripe for change. After examining sources of family ostracism, we explored how different factors, such as length of ostracism or the motive for ostracism might be linked to self-esteem. Although there were no differences in trait self-esteem for sources versus nonsources, sources who used ostracism for punitive reasons had higher self-esteem than those using ostracism for defensive reasons. Perhaps those with higher self-esteem have greater confidence in their social bonds and their abilities to wield ostracism as a form of punishment without fear of permanently ending these bonds. This both supports and contradicts previous work suggesting those who bully have low self-esteem (O’Moore & Kirkham, 2001), but also that bullying has been linked to narcissism (Baumeister, Bushman, & Campbell, 2000). People with higher self-esteem may create downward social comparisons and use ostracism to distance themselves from those they do not want to emulate. Thus, results remain conflicted about the role of self-esteem in aversive interpersonal behaviors, but certainly, there is a nuanced relationship, which warrants additional research. Turning our attention to the construct of depression, our results indicated that people who reported being a source had no difference in depression symptoms than those who report never being a source. Similarly, there were no differences in depression symptoms for sources who had a punitive motive compared with those with a defensive motive. However, correlational results indicated that participants who reported higher levels of family ostracism on the modified Experiences with Ostracism scale (revised to ask about family ostracism) tended to report more depression symptoms. Because these results are correlational in nature, one must not conclude a causal relationship. It seems plausible that being ostracized by family might play a role in increasing a family member’s feelings of depression due to the threat of losing an important social bond. However, a family member exhibiting depression symptoms such as irritability or withdrawal from daily activities could pose a burden on the family, conceivably prompting ostracism from family (Wesselmann, Wirth et al., 2013). Relatedly, clinical work based in family systems theory has found that depressive symptoms can be part of the marital dynamic in which behavior in the couple triggers and maintains depressive symptoms (Feldman, 1976). Perhaps family/spousal ostracism is one behavior in this dynamic. Finally, we examined the role of trait forgiveness for sources of family ostracism. We found that dispositional forgiveness is unrelated to engaging in family ostracism or the duration of ostracism. Another finding was that people who were less forgiving ostracized for punitive reasons. Those who are less forgiving may hold a grudge against a family member, leading to using a variety of punishments against them, including ostracism. Further, forgiveness was a trait associated with higher self-esteem, lower depression, and fewer experiences of family ostracism. These exploratory findings suggest that although forgiveness may not have been critical to understanding many aspects of sources of family ostracism, it may be worth exploring in the future how a source’s forgiveness pertains to reflective stage responses to ostracism in targets.

POULSEN AND CARMON

467


Limitations This study examined narratives of family ostracism and combined this approach with quantitative analyses. The goal was to create a study capitalizing on the strengths of both designs; however, we must note several limitations. First, sampling was a challenge in that participants who began the survey did not always provide a complete or descriptive narrative to code. Several participants quit the study early and did not provide ample data for qualitative analysis. We are unable to know if this attrition was random, or something more systematic like selection effects. It may be that only those who had vivid or strong memories of ostracism completed the narrative. It may also be possible that if the memory was too strong, people chose to discontinue participation. An additional issue is that of memory bias. As Loftus and Pickrell (1995) suggested, individuals can alter their memories or can even create new ones from scratch. Different errors in memory may have occurred within our study; in particular, that individuals tend to recall more extreme cases of a behavior, rather than a typical instance (Schwarz, Groves, & Schuman, 1998). We aimed to help account for this by asking participants to report on the most recent incident of ostracism as well as asking how long ago ostracism occurred, and the average response indicated either the event was ongoing or had happened a year or less ago. Although we cannot dismiss memory bias, we argue that because most accounts are recent, there is less chance for profound memory biases to occur. Three additional limitations warrant consideration here. First, although we feel we met the aims of our study, these aims did not include determining causality. We do not draw causal conclusions from this study. Second, our lack of including targets in our sample means we lack the ability to make direct comparisons with them, even though other work on induced ostracism has made direct comparisons before (e.g., Poulsen & Kashy, 2012; Zadro et al., 2008). We are hopeful that future work on family ostracism will incorporate information on targets of family ostracism. Third, our operationalization of ostracism may have affected our participants’ responses. We asked participants to recall a time they excluded a family member, but by asking them to recall a time of exclusion, we did not necessarily cause them to recall a time then they were excluded and ignored an individual, which is ostracism by definition. Conclusions Family ostracism is a unique experience, and this study has begun to shed light on this phenomenon. Our work replicated and extended Williams’ (1997, 2007) theory of ostracism to the family context. For instance, our findings suggest it is relatively common to ostracize a family member at some point in life. Sources of family ostracism report ostracizing for months at a time on average, a finding that shows how severe ostracism in families can be. Individuals are often both targets and sources of ostracism in families, a pattern similar to some in romantic relationships and among bullies, and something we hope future research can explore. We also highlight that punitive motives drive sources of family ostracism, a finding similar to Nezlek et al. (2015). We are excited that these findings are in congruence with other current research (e.g., Nezlek et al., 2012) and begin to paint a picture of sources of ostracism within families. Finally, we find that sources, who are less forgiving of others in general, tend to engage in more punitive than defensive ostracism. Given its prevalence, and potential for damaging some of our cherished, important bonds of belonging and attachment, family ostracism certainly warrants more attention.

468



NOTES 1. We achieved random assignment by assigning participants to a different writing prompt based on the day of the month they were born. Those reporting being born the 1st–10th, 11th–20th, and 21st–31st were each assigned to write about their experience as target, source, or witness, respectively. Target and witness data were not the focus of this article, thus were not included in analyses discussed in this study. 2. Five accounts were considered to have insufficient detail for analysis if no coders could discern who ostracized whom, for how long, what the ostracism entailed, or motive. For instance, one such narrative stated, “I’m sure in my life I have done things without all of the members of my family,” another “just for small things going on nothing huge.” Other times certain questions were left blank by participants, making it impossible to know who the source was, how long ago ostracism occurred, or why it began or ended. These five cases were not used in any analyses with coded data. An additional four respondents provided partial information for coding but lacked complete details, we included them in analyses where they provided sufficient data. 3. During discussions of coding for motives, we became aware of confusion about how different motives may manifest in families. For example, we discussed if ostracizing a younger sibling was role-prescribed or punitive. Another example was someone ostracized an adult sibling for substance use with a vague reference to protecting her children, and we discussed if this was punitive or defensive. After discussing, and reviewing Williams (1997) again, we recoded this construct with better agreement.

AUTHOR NOTES Joan R. Poulsen is an Associate Professor in Psychology at Indiana University-Purdue University Columbus. Anna F. Carmon is an Assistant Professor in Communication Studies at Indiana University-Purdue University Columbus.

REFERENCES Agllias, K. (2013). The gendered experience of family estrangement in later life. Journal of Women and Social Work, 28, 309–312. doi:10.1177/0886109913495727 Baumeister, R. F. (1998). The self. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), Handbook of social psychology (pp. 143–179). New York, NY: Oxford University Press. Baumeister, R. F., Bushman, B. J., & Campbell, W. K. (2000). Self-esteem, narcissism, and aggression: Does violence result from low self-esteem or from threatened egotism? Current Directions in Psychological Science, 9, 26–29. doi:10.1111/1467-8721.00053 Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonal attachments as a fundamental human motivation. Psychological Bulletin, 117, 497–529. doi:10.1037/0033-2909.117.3.497 Beck, A. T., Steer, R. A., & Brown, G. K. (1996). Manual for the Beck Depression Inventory-II. San Antonio, TX: Psychological Corporation. Blackhart, G. C., Nelson, B. C., Knowles, M. L., & Baumeister, R. F. (2009). Rejection elicits emotional reactions but neither causes immediate distress nor lowers self-esteem: A meta-analytic review of 192 studies on social exclusion. Personality and Social Psychology Review, 13, 269–309. doi:10.1177/1088868309346065 Boelen, P. A., & Reijntjes, A. (2009). Negative cognitions in emotional problems following romantic relationship breakups. Stress & Health, 25, 11–19. doi:10.1002/smi.1219 Bowen, M. (1978). Family therapy in clinical practice. New York, NY: Jason Aronson. Buss, D. M., Gomes, M., Higgins, D. S., & Lauterbach, K. (1987). Tactics of manipulation. Journal of Personality and Social Psychology, 52, 1219–1229. doi:10.1037/0022-3514.52.6.1219 Carter-Sowell, A. (2010). Salting a wound, building a callous, or throwing in the towel? The measurement and effects of chronic ostracism experiences (Unpublished doctoral thesis). Purdue University, West Lafayette, IN. Feldman, L. B. (1976). Depression and marital interaction. Family Process, 15, 389–395. doi:10.1111/ j.1545-5300.1976.00389.x Fincham, F. D., Beach, S. R. H., & Davila, J. (2004). Forgiveness and conflict resolution in marriage. Journal of Family Psychology, 18, 72–81. doi:10.1037/0893-3200.18.1.72


POULSEN AND CARMON

469

Fincham, F. D., Beach, S. R. H., & Davila, J. (2007). Longitudinal relations between forgiveness and conflict resolution in marriage. Journal of Family Psychology, 21, 542–545. doi:10.1037/0893-3200.21.3.542 Fitness, J. (2005). Bye bye, black sheep: The causes and consequences of rejection in family relationships. In K. Williams, J. Forgas, & W. von Hippel (Eds.), The social outcast: Ostracism, social exclusion, rejection, and bullying (pp. 445–468). Hove, UK: Psychology Press. Hallgren, K. A. (2012). Computing inter-rater reliability for observational data: An overview and tutorial. Tutorials in Quantitative Methods of Psychology, 8, 23–34. Kazdin, A. E. (Ed.). (2000). Encyclopedia of psychology. Washington, DC: American Psychological Association. Leary, M. R. (1990). Responses to social exclusion: Social anxiety, jealousy, loneliness, depression, and low self-esteem. Journal of Social and Clinical Psychology, 9, 468–488. doi:10.1521/jscp.1990.9.2.221 Loftus, E. F., & Pickrell, J. E. (1995). The formation of false memories. Psychiatric Annals, 25, 720–725. doi:10.3928/ 0048-5713-19951201-07 Murray, S. L., Rose, P., Bellavia, G. M., Holmes, J. G., & Kusche, A. G. (2002). When rejection stings: How selfesteem constrains relationship-enhancement processes. Journal of Personality and Social Psychology, 83, 556–573. doi:10.1037/0022-3514.83.3.556 Nezlek, J. B., Wesselmann, E. D., Wheeler, L., & Williams, K. D. (2012). Ostracism in everyday life. Group Dynamics: Theory, Research, and Practice, 16, 91–104. doi:10.1037/a0028029 Nezlek, J. B., Wesselmann, E. D., Wheeler, L., & Williams, K. D. (2015). Ostracism in everyday life: The effects of ostracism on those who ostracize. Journal of Social Psychology, 155(5), 432–451. O’Moore, M., & Kirkham, C. (2001). Self-esteem and its relationship to bullying behaviour. Aggressive Behavior, 27, 269–283. doi:10.1002/ab.1010 Poulsen, J. R., & Kashy, D. A. (2012). Two sides of the ostracism coin: How sources and targets of social exclusion perceive themselves and one another. Group Processes and Intergroup Relations, 15, 457–470. doi:10.1177/ 1368430211430517 Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton University Press. Rusbult, C. E., Johnson, D. J., & Morrow, G. D. (1986). Impact of couple patterns of problem solving on distress and nondistress in dating relationships. Journal of Personality and Social Psychology, 50, 744–753. Schwarz, N., Groves, R. M., & Schuman, H. (1998). Survey methods. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), Handbook of social psychology (pp. 143–179). New York, NY: Oxford University Press. Seals, D., & Young, J. (2003). Bullying and victimization: Prevalence and relationship to gender, grade level, ethnicity, self-esteem, and depression. Adolescence, 38, 735–747. Sommer, K. L., Williams, K. D., Ciarocco, N. J., & Baumeister, R. F. (2001). When silence speaks louder than words: Explorations into the intrapsychic and interpersonal consequences of social ostracism. Basic and Applied Social Psychology, 23, 225–243. doi:10.1207/S15324834BASP2304_1 Thompson, L. Y., Snyder, C. R., Hoffman, L., Michael, S. T., Rasmussen, H. N., Billings, L. S., . . . Roberts, D. E. (2005). Dispositional forgiveness of self, others, and situations. Journal of Personality, 73, 313–360. doi:10.1111/ jopy.2005.73.issue-2 Webb, J. R., Hill, S. K., & Brewer, K. G. (2012). Dimensions of social support as mediators of the forgiveness-alcohol outcome relationship. Addictive Disorders & Their Treatment, 11, 64–75. doi:10.1097/ADT.0b013e31822dcd56 Wesselmann, E. D., Williams, K. D., & Hales, A. H. (2013). Vicarious ostracism. Frontiers in Neuroscience, 7. doi:10.3389/fnhum.2013.00153 Wesselmann, E. D., Wirth, J. H., Pryor, J. B., Reeder, G. D., & Williams, K. D. (2013). When do we ostracize? Social Psychological and Personality Science, 4, 108–115. doi:10-1177/1948550612443386 Williams, K. D. (1997). Social ostracism. In R. M. Kowalski (Ed.), Aversive interpersonal behavior (pp. 133–170). New York, NY: Plenum Press. Williams, K. D. (2001). Ostracism: The power of silence. New York, NY: Guilford Press. Williams, K. D. (2007). Ostracism. Annual Review of Psychology, 58, 25–452. doi:10.1146/annurev.psych.58.110405. 085641 Williams, K. D., Shore, W. J., & Grahe, J. E. (1998). The silent treatment: Perceptions of its behaviors and associated feelings. Group Processes & Intergroup Relations, 1, 117–141. doi:10.1177/1368430298012002 Zadro, L. (2004). Ostracism: What the real world can teach us about silence (Unpublished doctoral thesis). University of New South Wales, Sydney, Australia.

470


Zadro, L., Arriaga, X. B., & Williams, K. D. (2008). Relational ostracism. In J. P. Forgas, & J. Fitness (Eds.), Social relationships: Cognitive, affective, and motivational processes (pp. 305–320). New York, NY: Psychology Press. Zadro, L., Williams, K., & Richardson, R. (2005). Riding the ‘O’ train: Comparing the effects of ostracism and verbal dispute on targets and sources. Group Processes & Intergroup Relations, 8, 125–143. doi:10.1177/1368430205051062 Zimet, G. D., Dahlem, N. W., Zimet, S. G., & Farley, G. K. (1988). The multidimensional scale of perceived social support. Journal of Personality Assessment, 52, 30–41. doi:10.1207/s15327752jpa5201_2





Righting the Wrong: Reparative Coping After Going Along With Ostracism a

b

Nicole Legate , Cody DeHaan & Richard Ryan a

Illinois Institute of Technology

b

University of Rochester

c

c

Australian Catholic University Published online: 12 Aug 2015.

Click for updates To cite this article: Nicole Legate, Cody DeHaan & Richard Ryan (2015) Righting the Wrong: Reparative Coping After Going Along With Ostracism, The Journal of Social Psychology, 155:5, 471-482, DOI: 10.1080/00224545.2015.1062352 To link to this article: http://dx.doi.org/10.1080/00224545.2015.1062352





Righting the Wrong: Reparative Coping After Going Along With Ostracism NICOLE LEGATE Downloaded by [Ohio State University Libraries] at 08:50 20 August 2015

Illinois Institute of Technology

CODY DEHAAN University of Rochester

RICHARD RYAN Australian Catholic University

ABSTRACT. Most of the focus within the ostracism literature concerns the negative effects on the ostracized and how they cope following ostracism. Research is now beginning to illuminate negative psychological effects for ostracizers, yet no studies to date have examined their coping responses. This study continues this line of inquiry focusing on experiences of going along with ostracism, both by employing a face-to-face interaction and by exploring prosocial versus antisocial coping reactions in ostracizers. Results reveal that compared to those in a neutral condition, compliant ostracizers suffered because ostracizing someone else frustrated their psychological needs for autonomy and relatedness. Further, when given the chance, ostracizers were more inclusive of the person they previously ostracized. Discussion considers important avenues for future research as well as implications of results. Keywords: coping strategies, motivation, self-determination theory, sources of ostracism

HUMANS ARE SOCIAL CREATURES, and as a large body of evidence shows, it hurts to be ostracized1 (see Williams, 2007b, for a review). Being ostracized is painful (e.g., Eisenberger, Lieberman, & Williams, 2003), and produces strong feelings of sadness and anger (Williams, 2009). Although there are factors that exacerbate the pain and distress of ostracism (Bernstein, Sacco, Young, Hugenberg, & Cook, 2010; Sacco, Bernstein, Young, & Hugenberg, 2014), nothing dampens the immediate pain of being ostracized. However, longer-term emotional adjustment after ostracism depends on coping (Williams & Nida, 2011). Research on coping among the ostracized shows that people react in ways that restore the needs that were threatened, such as increasing chances for future inclusion (e.g., being cooperative) to restore a sense of belonging, or with aggressive and antisocial behavior (e.g., Catanese & Tice, 2005) as a way to establish control that was previously lost. Address correspondence to Nicole Legate, Illinois Institute of Technology, Department of Psychology, 3105 S. Dearborn St., Chicago, IL 60616, USA. E-mail: [email protected]


472


As social creatures, people tend to go along with others even when it can do harm (e.g., Milgram, 1963), especially when the consequences of not doing so will likely result in ostracism (e.g., Ouwerkerk, Kerr, Gallucci, & Van Lange, 2005). This may help explain why youth frequently go along with a dominant peer in hurting or ostracizing someone in elementary school years and adolescence (Berndt, 2002; Crick & Grotpeter, 1995; Dodge, Coie, & Lynam, 2008). Ostracism continues in adulthood as well. For example, one study reported that 78% of adults have experienced some form of ostracism at work (O’Reilly, Robinson, Berdahl, & Banki, 2014). This same study found that ostracism was more common than harassment, and also more detrimental. Other research shows that ostracizing others is as prevalent as being ostracized (Faulkner, Williams, Sherman, & Williams, 1997) and that people admit to ostracizing others frequently in their daily lives (Nezlek, Wesselmann, Wheeler, & Williams, 2015). A few recent studies (Crick, Ostrov, & Werner, 2006; Legate, DeHaan, Weinstein, & Ryan, 2013; Poulsen & Kashy, 2012) have revealed that ostracizing others can be as psychologically harmful as being ostracized. Specifically, prior research (Legate et al., 2013) has found that ostracizing hurts both parties for the same reason: it thwarts people’s psychological needs for autonomy and relatedness. Targets of ostracism felt severed connections that thwarted relatedness (see also Williams, 2009). Yet persons complying in ostracism also suffered lowered relatedness, as well as thwarted autonomy, because they felt controlled or compelled to go along with something that they would not typically choose to do. Insofar as compliance with ostracism is associated with lowered autonomy and relatedness, of interest is examining how compliant ostracizers cope after they ostracize, and particularly whether they try and make amends for the potential harm they have done. SELF-DETERMINATION THEORY FRAMEWORK To understand why people suffer when they ostracize, as well as give insight into how they might cope, we use a self-determination theory (SDT; Ryan & Deci, 2000) framework. SDT concerns itself with the conditions that lead to growth and human flourishing, as well as those that lead to suffering, aggression, and antisocial behaviors (see Vansteenkiste & Ryan, 2013). Specifically, people thrive when their basic psychological needs for autonomy, competence, and relatedness are met, and they suffer and react defensively when these needs are thwarted. Although the specification of these basic needs arises from a separate tradition than the needs outlined within the needs–threat model of ostracism (Williams, 2009), they overlap in many ways. Relatedness is the need to experience close and caring connections with others and clearly maps onto the need to belong (Baumeister & Leary, 1995); it is experienced when one cares for or is cared for by others, and is thwarted when one experiences isolation or disconnection. Autonomy is the need to experience one’s own behavior as volitional, valued, and self-endorsed; it is supported under conditions that allow for meaningful choice, and is thwarted under conditions that are pressuring or coercive. However, autonomy is independent from the need for control (e.g., Seligman, 1975) as one can have control over outcomes without feeling autonomous about the actions entailed (Deci & Ryan, 2000). Competence is the need to feel capable and effective at important life activities, is supported when an individual encounters appropriate challenges with informational feedback, and can contribute to self-esteem (e.g., Steele, 1988), as can the other two needs within the SDT tradition. Similarly, although meaningful existence (Solomon, Greenberg, & Pyszczynski, 1991) is not

LEGATE, DEHAAN, AND RYAN

473

a need recognized within SDT, satisfaction of autonomy, competence, and relatedness promote a sense of meaning in one’s life (Ryan & Deci, 2004).


OSTRACIZING THWARTS NEEDS Compliance with ostracizing others can thwart autonomy and relatedness, as people are often pressured to do it (e.g., Dodge et al., 2008), and it functions to disconnect people. Indeed, prior research has demonstrated this (Legate et al., 2013). Importantly, this effect was not due to the act of compliance per se, as compliance with requests to be inclusive resulted in smaller costs to autonomy, and did not result in costs to relatedness or mood. Conversely, compliance with ostracism produced significantly larger costs to autonomy, as well as decrements in relatedness and mood. In the present research we will again test whether compliance with ostracizing someone else is need thwarting, and whether need thwarting has emotional costs. We included the need for competence in the model, but had no a priori assumptions about whether or not it would also be thwarted following ostracism. We also extend this previous research by examining coping responses in ostracizers, and particularly propensities toward repairing relations with those they have wronged, an untouched area of research.

HELPING SATISFIES NEEDS According to SDT, people are generally intrinsically motivated to help others, and they derive need satisfactions directly from helping behaviors (Ryan & Hawley, in press; Weinstein & Ryan, 2010). Specifically, when acting in helpful ways people generally feel enhanced competence (for being effective), enhanced relatedness (connecting with others), and enhanced autonomy (because most prosocial behavior is volitional and self-valued). Further, need satisfaction explains, in turn, why helping enhances positive affect and reduces negative affect (Weinstein & Ryan, 2010). In contrast, in complying with ostracism, one is doing social harm rather than helping, diminishing both autonomy and relatedness satisfactions, as Legate and colleagues (2013) demonstrated. Accordingly, we expect that when compliant ostracizers are no longer constrained, they will attempt to “right the wrong” by showing more inclusive behaviors. Indeed, prior SDT research suggests that under need-satisfying conditions, people tend to be inclusive (Legault, Gutsell, & Inzlicht, 2011) and support the basic needs of others (Deci, La Guardia, Moller, Schiener, & Ryan, 2006). Following from this, we expect when the pressure to ostracize is removed, ostracizers will cope by showing more inclusive behaviors that may also restore their thwarted needs. Specifically, we expect that when given the chance to interact with the person they previously ostracized, individuals will go out of their way to include, as opposed to continuing to ostracize.

PROSOCIAL AND REPARATIVE COPING Just as those who have been ostracized or hurt must cope with the pain of ostracism (Williams & Nida, 2011), so do the individuals who have hurt others. However, very little work has examined


474


how people cope when they do harm. Generally, research has found that “doing good” in the form of prosocial behavior helps people when they are feeling bad (e.g., Grant & Sonnentag, 2010; Schaller & Cialdini, 1988) and has been associated with better coping (Midlarsky, 1991), and with reduced depression over time (Musick & Wilson, 2003). When examining prosocial and antisocial coping responses side by side, prosocial coping is related to adjustment, whereas antisocial coping predicts depressive symptoms (Monnier, Cameron, Hobfall, & Gribble, 2000). People also tend to want to correct past wrongdoings. At the group-level, perceived responsibility motivates people to make reparations for past wrongdoing (e.g., Doosje, Branscombe, Spears, & Manstead, 1998; Dumont & Waldzus, 2014). One study of employees found that helping others was an effective way to cope after hurting others, because helping attenuated the emotional costs associated with doing harm to others (Grant & Campbell, 2007). Closely related to the present work, recent research has revealed that people tend to compensate those unfairly ostracized. Specifically, people tended to compensate an individual who they saw being unjustly ostracized by others (Wesselmann, Wirth, Pryor, Reeder, & Williams, 2013) or after they ostracized someone without sufficient justification to do so (Van Tongeren, Root Luna, & Witvliet, 2015). The current work adds to research on compensation after ostracism by observing what happens when people are given a second chance to interact with people they have personally wronged, whether or not they perceive their own ostracizing behavior as unjust.

THE CURRENT RESEARCH In the current study, we investigated the phenomenon of compliance with ostracism. Extending past research, we created an ostracism experience using a face-to-face interaction paradigm. Following the period of ostracism, we then gave people a second opportunity to include or continue to ostracize the person they had ignored in the face-to-face encounter using a Cyberball (Williams, Cheung, & Choi, 2000) game interaction. In line with prior work (Legate et al., 2013), we expected that complying with ostracizing would yield emotional costs (negative affect) because it thwarts needs for autonomy and relatedness. We chose to examine this meditational pathway because it was demonstrated in prior work (Legate et al., 2013), and is consistent with our theorizing and other research in SDT (e.g., Ryan, Bernstein, & Brown, 2010; Weinstein & Ryan, 2010). We also expected ostracizers to cope in a prosocial manner. Specifically, we predicted that when given the opportunity, ostracizers would include—rather than continue to ostracize—the individual they had previously ignored, even though the other group confederate continued to ostracize.

METHOD Fifty-eight undergraduates (age: M = 20; gender: 46 women, 12 men; race: 58% Caucasian, 25% Asian, 17% other) participated. Each participant was met in the lab by an experimenter and two confederates. The participant and two confederates were seated in separate lab spaces, ostensibly to complete surveys before a “getting to know you” conversation task. Participants completed a baseline measure of the Positive and Negative Affect Schedule (PANAS; Watson, Clark, & Tellegen, 1988) to assess mood. Items were rated on a scale from 1 (very slightly or not at all) to



475

5 (extremely). Our focus, following previous findings of Legate et al. (2013) was on the negative affect subscale (α = .80). We used the Basic Psychological Needs scale (BPNS; Ilardi, Leone, Kasser, & Ryan, 1993) to assess state autonomy (sample item: “Right now I feel like I am free to decide for myself how to act”; α = .81), competence (sample item: “Right now I do not feel very capable”, reverse coded; α = .82), and relatedness (“Right now I really like the people I’m interacting with”; α = .82). These items were rated on a scale from 1 (not at all true) to 7 (very true). Following this, the experimenter randomly assigned participants to one of two conditions, either the ostracizer condition or the neutral condition, and silently informed one confederate whether they would be ostracizing or including the target confederate; in all cases the target confederate was naïve to condition. Participants were told they would spend six minutes in a free-form conversation, during which their task was to get to know each other. They picked a name out of a bowl to ostensibly pick one of the other two participants, but both names in the bowl were the target confederate’s. In the ostracizer condition, participants were instructed to exclude from the conversation the person whose name they drew. Exclusion involved not directing comments, conversation, or gestures toward the excluded party. In the neutral condition, the experimenter explained that the person whose name was drawn would start the conversation. The informed confederate either went along with ostracism (in the ostracizer condition) or engaged everyone in conversation (in the neutral condition). Although naïve to condition initially (until he or she noticed being included or ostracized), the target confederate was instructed to always start the conversation. The same two confederates were used for all participants—one male and one female—and their roles were counterbalanced across condition. Following the conversation, the participant and confederates were taken back to their separate lab spaces, and participants completed the PANAS and BPNS in terms of how they felt during the conversation (negative affect subscale α = .89; autonomy α = .86; competence α = .82; relatedness α = .86). They also completed manipulation check questions asking about their autonomous and controlled reasons for including and excluding (sample item for controlled motivation: “I included [excluded] a participant from the conversation because I felt I should”; sample item for autonomous motivation: “I included [excluded] a participant from the conversation because I valued doing so”), using a modified version of the Self-Regulation Questionnaire (Ryan & Connell, 1989). Items were rated on a scale from 1 (not at all true) to 7 (very true). To serve as the manipulation check, we collapsed items about inclusion and exclusion into two composites (inclusion: α = .95; exclusion: α = .80), and in later analyses separated them into autonomous and controlled motives (autonomous inclusion: α = .95; controlled inclusion: α = 87; autonomous exclusion: α = .72; controlled exclusion: α = .75). Finally, after the initial phase of the experiment was completed, participants were ostensibly given a chance to interact with the same two people with whom they conversed through a virtual ball-tossing game, Cyberball (Williams et al., 2000). Participants were instructed to visualize the situation, themselves, and the other players, but no instructions were given about how to interact with the other players. Participants typed their names, and saw their names and the confederates’ names above the computerized images of the players. For those in the ostracizer condition, the game was programmed such that the confederate ostracizing during the conversation continued to ostracize the other confederate; he or she threw the ball twice to the other confederate, but then stopped doing so for the remainder of the game. In other words, this was testing whether the participant would choose to continue to go along with ostracism, or instead include the previously

476


ostracized confederate. In the neutral condition, both confederates threw the ball equally to the other players (i.e., there was no programmed exclusion). Finally, participants were debriefed, and no participant was able to identify the manipulation or purpose of the study when asked. For materials, please see https://osf.io/knxcg/ on Open Science Framework (OSF), and to access the data please contact the first author.


RESULTS As a manipulation check, we compared groups on questions about their reasons for inclusion versus exclusion. Ostracizers (coded 1) highly endorsed being motivated to exclude (r = .80, p < .001) and those in the neutral group (coded −1) were motivated to include (r = −.81, p < .001). Interestingly, ostracizers tended to endorse both controlled (r = .84, p < .001) and to a lesser degree, autonomous (r = .28, p = .030) reasons for exclusion, while those in the neutral condition endorsed both autonomous (r = −.82, p < .001) and controlled (r = −.69, p < .001) reasons for inclusion. Not surprisingly, ostracizers reported significantly more controlled versus autonomous reasons for exclusion (z = 4.90, p < .001); the neutral group reported marginally more autonomous vs. controlled reasons for inclusion (z = 1.57, p = .058). Repeated-measures analysis of variance tested Time 1 and Time 2 indicators as a function of condition, controlling for gender in all analyses as gender differences emerged in Time 1 means for relatedness and competence (i.e., females reported greater need satisfaction than males). Preand post-manipulation means by condition are presented in Table 1. Results showed a significant interaction of time by condition for negative affect: F(1, 54) = 17.83, p < .001, ηp 2 = .25, 95% confidence interval (CI) of the effect size (.07, .42). Simple slopes revealed that negative affect did not change in the neutral group (F(1, 23) = .02, p = .875, ηp 2 = .001 CI [.00, .04]), while it

TABLE 1 Means and Standard Deviations at Time 1 and Time 2 as a Function of Condition: Results of Repeated-Measures Analysis of Covariance (Controlling for Gender), Effect Sizes and Their Confidence Intervals (CI) for Moderation of Condition by Time and Simple Effects by Condition Outcome Neg. affect Neutral Ostracizer Autonomy Neutral Ostracizer Relatedness Neutral Ostracizer Competence Neutral Ostracizer ∗∗∗ p

< .001.

Time 1

Time 2

1.30(0.33) 1.47(0.46)

1.22(0.33) 2.12(0.79)

5.31(0.92) 5.33(0.85)

5.86(0.97) 3.31(1.17)

5.80(0.76) 5.59(0.81)

5.46(1.05) 4.80(1.09)

5.20(0.81) 5.15(1.13)

5.92(1.50) 5.52(1.56)

F

ηp 2

CI

17.83∗∗∗ 0.02 31.50∗∗∗ 70.57∗∗∗ 1.32 67.71∗∗∗ 2.16 0.22 18.48∗∗∗ 1.16 1.63 0.16

.25 .00 .51 .56 .05 .69 .04 .01 .37 .02 .07 .01

.07, .42 .00, .04 .24, .67 .38, .68 .00, .28 .47, .79 .00, .17 .00, .19 .11, .56 .00, .14 .00, .29 .00, .14



477

significantly increased for ostracizers, F(1, 30) = 31.50, p < .001, ηp 2 = .51 CI (.24, .67). There was also a significant time by condition interaction for autonomy, F(1, 55) = 70.57, p < .001, ηp 2 = .56, CI (.38, .68) with simple slopes showing no change for those in the neutral condition, F(1, 23) = 1.32, p = .263, ηp 2 = .05 CI (.00, .28), while autonomy decreased for ostracizers after the manipulation, F(1, 31) = 67.77, p < .001, ηp 2 = .69 CI (.47, .79). There was a nonsignificant trend by condition in how relatedness scores changed before and after the manipulation, F(1, 55) = 2.16, p = .14, ηp 2 = .04, CI (.00, .17). We continued to look at simple slopes because we expected to see a difference by condition; they revealed that relatedness significantly decreased for ostracizers, F(1, 31) = 18.48, p < .001, ηp 2 = .37, CI (.11, .56) whereas it did not change in the neutral condition, F(1, 23) = .22, p = .642, ηp 2 = .01 CI (.00, .19). Finally, there was no change in competence by condition, F(1, 55) = 1.16, p = .286, ηp 2 = .02 CI (.00, .14) and to be sure we examined simple slopes, neither of which were significant (F(1, 23) = 1.63, p = .215, ηp 2 = .07 CI [.00, .29] and F(1, 31) = .16, p = .693, ηp 2 = .005 CI [.00, .14]). See Figure 1 for mean scores by condition. There were no gender differences in changes for affect or needs across the two time points. Next, we tested our expectation that ostracizing leads to more negative affect because it thwarts needs for autonomy and relatedness (aggregated together). This mediation hypothesis was tested using the Process procedure (Hayes, 2013) to obtain bias-corrected bootstrapped estimates based on 10,000 bootstrapping samples. Analyses revealed that ostracizing increased negative affect through autonomy and relatedness need thwarting (indirect effect = 0.32, 95% CI [0.18, 0.46]). In other words, going along with ostracism yielded greater negative affect because it thwarted needs for autonomy and relatedness.2 Finally we predicted efforts to repair relations, such that persons in the ostracism condition were expected to show a behavioral compensation after ostracizing another person, manifest in 7

Neutral Ostracizer

6 5 4 3 2 1 0 Negative affect Autonomy

Relatedness Competence

Throw ratio

FIGURE 1 Mean scores on measures of post-conversation negative affect, psychological needs, and throws to previously ostracized confederate as a function of condition (ostracizer or neutral). Note. Bars represent standard errors of the means. Negative affect used a scale of 1 to 5, and needs used a scale of 1 to 7. Throw ratio refers to the proportion of throws to the formerly ostracized target confederate compared to the ostracizing confederate (or comparable targets in the neutral group), where 1 indicates an equal proportion of throws to both confederates and a score greater than 1 indicates more throws to the formerly ostracized confederate (or comparable target in the neutral group).


478


differential ball throwing to the previously ostracized player (the target) in the Cyberball game. Following Wesselmann et al. (2013), we defined ball toss behavior as a ratio of throws to the target player compared to throws to the other confederate to account for the unequal number of participant throws across groups (total participant throws by group: M neutral = 20, SD = 0.20; M ostracizer = 27, SD = 5.05). There was a significant effect of condition: F(1, 53) = 8.68, p = .005, ηp 2 = .14, CI (.01, .31). Interestingly, those in the neutral condition threw the ball equally (throw ratio M = 1.04, SD = .18) to the other two players, while ostracizers threw the ball significantly more to the person they previously ostracized (throw ratio M = 1.40, SD = .69). In line with our prediction, it appears that, on average, ostracizers made a concerted effort to include the formerly ostracized confederate even when the other confederate continued to ostracize. Having identified the effect that ostracizers attempted to repair the situation with the person they ostracized, a speculative research question concerned the motivation behind it. Thus in a fully post hoc analysis, we extracted the “guilt” item from the negative affect subscale of the PANAS to see if guilt could explain the effect of inclusion after ostracism. We specifically explored guilt as a potential mediator because it is an action-orienting emotion that can drive people to correct past transgressions (Niedenthal, Tangney, & Gavanski, 1994; Tangney & Fischer, 1995). Using the same mediation procedure, we found no indirect effect of ostracizers being more inclusive of the previously ostracized individual compared to those in the neutral condition because of guilt (indirect effect = .003, CI [−.18, .28]). Although ostracizers felt more guilt than those in the neutral condition (B = 1.21, p < .001, CI [.92, 1.51]), guilt failed to explain inclusion after ostracism (see Figure 1). Related to this, we explored our manipulation check items concerning motivation for exclusion in another fully post hoc analysis in an effort to better understand the motivation driving differential throwing behavior. In a simultaneous mediation model using the same procedure, we tested for indirect effects of both autonomous and controlled reasons for exclusion, and found that excluding for controlled reasons mediated the effect of being an ostracizer on throw behavior (controlled motives indirect effect = −.34 CI [−.60, −.08]; autonomous motives indirect effect = .01 CI [−.04, .06]). In other words, those who ostracized tended to do so out of a sense of pressure and coercion (B = 1.14, p < .001, CI [.93, 1.36]), but the more they felt pressured and coerced, the less they threw the ball to the person they previously ostracized (B = −.30, p = .002, CI [−.48, −.12]).3

DISCUSSION A handful of studies have found that ostracism is not just painful for the target—it can hurt those compliant in it too (e.g., Poulsen & Kashy, 2012). In the current study, we replicated and extended Legate et al. (2013) showing that compliance in ostracizing another person leads to negative affect, mediated by lowered experiences of autonomy and relatedness. We found this pattern using a different and highly palpable face-to-face ostracism paradigm, in which participants had to actively ignore the target person in front of them, thus increasing the generalizability of these findings. Second, beyond testing affective outcomes, we also used Cyberball behavior as a dependent variable to see whether ostracizers made reparations to their “victims.” Results confirmed that they did: ostracizers threw the ball more to the person they previously ostracized compared



479

to the other player, whereas those in the neutral condition threw the ball equally to the other two players. It seems then that people who complied in ostracizing another made concerted efforts to include the person they ostracized when given the opportunity. In line with self-determination theory, this supports the notion that ostracizing others, and more generally causing social harm, is something people do for mostly controlled reasons (a notion these data supported); and when people do ostracize, it frustrates needs for autonomy and relatedness. Moreover, the current data suggest a desire to repair harms done. Although people may comply with ostracism, doing so thwarts their psychological needs. When this pressure is removed, they then act in ways that tend to satisfy autonomy and relatedness needs, potentially as a way to regain homeostasis. In other words, inclusion, as opposed to continued ostracizing, may help to restore emotional balance. Exploratory analyses suggested that this reparation effect was not explained by guilt, or by feelings of pressure. When people felt pressure to ostracize, they actually threw the ball less to the person they previously ostracized, further suggesting that people are not making attempts to repair wrongdoing out of pressure, defensiveness, or because they feel like they should. Interpreted through cognitive dissonance theory (Festinger, 1962), findings may also suggest that people who ostracized out of pressure are motivated to move away from the source of their discomfort—the individual they ostracized—to reduce dissonance. However, because we cannot know people’s motives for repairing relations without directly measuring it—a limitation of this work—these remain speculative interpretations. In the absence of guilt or feelings of pressure being able to account for this reparation effect, understanding their behavior as coping aimed at need restoration seems plausible. However, without data showing that inclusion restored people’s needs for autonomy and relatedness—another limitation of the work—this interpretation is also speculative. As such, future research should measure basic needs satisfaction before and after an opportunity to repair relations with the ostracized confederate. Understanding motivational dynamics underlying reparative coping reactions in ostracizers thus represents a critical area of future research. This research tested consequences and coping among those who complied, or went along with ostracism, as opposed to cases of ostracism that occur in the absence of coercion and pressure. It is likely that the motivational dynamics differ between internally motivated ostracizers and the compliant ostracizers we are studying. We began to examine this issue by measuring people’s motives for exclusion, finding that people mostly ostracized others because they felt pressured or coerced, linking to Zadro and Gonsalkorale’s (2014) category of “induced sources” of ostracism. Future research should therefore explore ostracism that occurs more out of the participant’s own volition. Specifically, it would be informative to compare those who initiate versus those who are induced to ostracize on resultant mood, basic needs, and coping responses. Additionally, future research should compare the two needs traditions—the needs–threat model of ostracism and the needs within SDT—to determine overlap as well as unique contributions of each when studying both targets and sources of ostracism. Despite limitations, this work is the first to examine coping reactions among compliant ostracizers, and provides initial evidence that people compliant in ostracizing others are motivated to “right the wrong” when given the opportunity to do so. Findings extend our insights into the psychological and relational dynamics of ostracism, and point to important opportunities to intervene with people who go along with harming others. Focusing on the fact that people who

480



harm must cope too, this work suggests that people are motivated to reconcile and correct past wrongdoings when they are removed from the source of pressure to ostracize.

NOTES 1. There is a debate about the use of the terms ostracism, exclusion, and rejection, and throughout this paper we only use the term ostracism for parsimony. The definition of ostracism in Williams (2007a) is most consistent with the phenomenon we are studying here. 2. Because tests of mediation cannot determine causality, we explored two alternative models, which can be found on the study’s OSF page (https://osf.io/z82hm/). 3. To see additional post hoc tests of mediation we ran, please see the OSF page, https://osf.io/z82hm/

AUTHOR NOTES Nicole Legate is an Assistant Professor in the Department of Psychology at the Illinois Institute of Technology. Cody DeHaan is a doctoral candidate in the Department of Clinical and Social Sciences in Psychology at the University of Rochester. Richard Ryan is a Professor in the Institute for Positive Psychology and Education at Australian Catholic University.

REFERENCES Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonal attachments as a fundamental human motivation. Psychological Bulletin, 117, 497–529. doi:10.1037/0033-2909.117.3.497 Berndt, T. J. (2002). Friendship quality and social development. Current Directions in Psychological Sciences, 11, 7–10. doi:10.1111/1467-8721.00157 Bernstein, M. J., Sacco, D. F., Young, S. G., Hugenberg, K., & Cook, E. (2010). Being “in” with the in-crowd: The effects of social exclusion and inclusion are enhanced by the perceived essentialism of ingroups and outgroups. Personality and Social Psychology Bulletin, 36, 999–1009. doi:10.1177/0146167210376059 Catanese, K. R., & Tice, D. M. (2005). The effect of rejection on antisocial behaviors: Social exclusion produces aggressive behaviors. In K. D. Williams, J. P. Forgas, & W. Von Hippel (Eds.), The social outcast: Ostracism, social exclusion, rejection, and bullying (pp. 297–306). New York, NY: Psychology Press. Crick, N. R., & Grotpeter, J. K. (1995). Relational aggression, gender, and social psychological adjustment. Child Development, 66, 710–722. doi:10.1111/j.1467-8624.1995.tb00900.x Crick, N. R., Ostrov, J. M., & Werner, N. E. (2006). A longitudinal study of relational aggression, physical aggression, and children’s social-psychological adjustment. Journal of Abnormal Child Psychology, 34, 127–138. doi:10.1007/ s10802-005-9009-4 Deci, E. L., La Guardia, J. G., Moller, A. C., Scheiner, M. J., & Ryan, R. M. (2006). On the benefits of giving as well as receiving autonomy support: Mutuality in close friendships. Personality and Social Psychology Bulletin, 32, 313–327. doi:10.1177/0146167205282148 Deci, E. L., & Ryan, R. M. (2000). The “what” and “why” of goal pursuits: Human needs and the self-determination of behavior. Psychological Inquiry, 11, 227–268. doi:10.1207/S15327965PLI1104_01



481

Dodge, K. A., Coie, J. D., & Lynam, D. (2008). Aggression and antisocial behavior in youth. In W. Damon & R. M. Lerner (Eds.), Child and adolescent development: An advanced course (pp. 437–472). Hoboken, NJ: John Wiley & Sons, Inc. Doosje, B., Branscombe, N. R., Spears, R., & Manstead, A. S. R. (1998). Guilty by association: When one’s group has a negative history. Journal of Personality and Social Psychology, 75, 872–886. doi:10.1037/0022-3514.75.4.872 Dumont, K., & Waldzus, S. (2014). Group-based guilt and reparation in the context of social change. Journal of Applied Social Psychology, 44, 331–341. doi:10.1111/jasp.12224 Eisenberger, N. I., Lieberman, M. D., & Williams, K. D. (2003). Does rejection hurt? An fMRI study of social exclusion. Science, 302, 290–292. doi:10.1126/science.1089134 Faulkner, S., Williams, K., Sherman, B., & Williams, E. (1997, May). The “silent treatment”: Its incidence and impact. Paper presented at the meeting of the Midwestern Psychological Association, Chicago, IL. Festinger, L. (1962). A theory of cognitive dissonance (Vol. 2). Redwood City, CA: Stanford University Press. Grant, A. M., & Campbell, E. M. (2007). Doing good, doing harm, being well and burning out: The interactions of perceived prosocial and antisocial impact in service work. Journal of Occupational and Organizational Psychology, 80, 665–691. doi:10.1348/096317906X169553 Grant, A. M., & Sonnentag, S. (2010). Doing good buffers against feeling bad: Prosocial impact compensates for negative task and self-evaluations. Organizational Behavior and Human Decision Processes, 111, 13–22. doi:10.1016/ j.obhdp.2009.07.003 Hayes, A. F. (2013). An introduction to mediation, moderation, and conditional process analysis: A regression-based approach. New York, NY: Guilford Press. Ilardi, B. C., Leone, D., Kasser, R., & Ryan, R. M. (1993). Employee and supervisor ratings of motivation: Main effects and discrepancies associated with job satisfaction and adjustment in a factory setting. Journal of Applied Social Psychology, 23, 1789–1805. doi:10.1111/j.1559-1816.1993.tb01066.x Legate, N., DeHaan, C. R., Weinstein, N., & Ryan, R. M. (2013). Hurting you hurts me too: The psychological costs of complying with ostracism. Psychological Science, 24, 583–588. doi:10.1177/0956797612457951 Legault, L., Gutsell, J. N., & Inzlicht, M. (2011). Ironic effects of antiprejudice messages: How motivational interventions can reduce (but also increase) prejudice. Psychological Science, 22, 1472–1477. doi:10.1177/0956797611427918 Midlarsky, E. (1991). Helping as coping. In M. S. Clark (Ed.), Prosocial behavior (pp. 238–264). Thousand Oaks, CA: Sage. Milgram, S. (1963). Behavioral study of obedience. The Journal of Abnormal and Social Psychology, 67, 371–378. doi:10.1037/h0040525 Monnier, J., Cameron, R. P., Hobfall, S. E., & Gribble, J. R. (2000). Direct and crossover effects of prosocial and antisocial coping behaviors. Journal of Family Psychology, 14, 570–584. doi:10.1037/0893-3200.14.4.570 Musick, M. A., & Wilson, J. (2003). Volunteering and depression: The role of psychological and social resources in different age groups. Social Science & Medicine, 56, 259–269. doi:10.1016/S0277-9536(02)00025-4 Nezlek, J. B., Wesselmann, E. D., Wheeler, L., & Williams, K. D. (2015). Ostracism in everyday life: The effects of ostracism on those who ostracize. Journal of Social Psychology, 155(5), 432–451. Niedenthal, P. M., Tangney, J. P., & Gavanski, I. (1994). “If only I weren’t” versus “If only I hadn’t”: Distinguishing shame and guilt in counterfactual thinking. Journal of Personality and Social Psychology, 67, 585–595. doi:10.1037/ 0022-3514.67.4.585 O’Reilly, J., Robinson, S. L., Berdahl, J. L., & Banki, S. (2014). Is negative attention better than no attention? The comparative effects of ostracism and harassment at work. Organization Science. doi:10.1287/orsc.2014.0900 Ouwerkerk, J. W., Kerr, N. L., Gallucci, M., & Van Lange, P. A. (2005). Avoiding the social death penalty: Ostracism and cooperation in social dilemmas. In K. D. Williams, J. P. Forgas, & W. Von Hippel (Eds.), The social outcast: Ostracism, social exclusion, rejection, and bullying (pp. 321–332). New York, NY: Psychology Press. Poulsen, J. R., & Kashy, D. A. (2012). Two sides of the ostracism coin: How sources and targets of social exclusion perceive themselves and one another. Group Processes & Intergroup Relations, 15, 457–470. doi:10.1177/ 1368430211430517 Ryan, R. M., Bernstein, J. H., & Brown, K. W. (2010). Weekends, work, and well-being: Psychological need satisfactions and day of the week effects on mood, vitality, and physical symptoms. Journal of Social and Clinical Psychology, 29, 95–122. doi:10.1521/jscp.2010.29.1.95 Ryan, R. M., & Connell, J. P. (1989). Perceived locus of causality and internalization: Examining reasons for acting in two domains. Journal of Personality and Social Psychology, 57, 749–761. doi:10.1037/0022-3514.57.5.749


482


Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55, 68–78. doi:10.1037/0003-066X.55.1.68 Ryan, R. M., & Deci, E. L. (2004). Avoiding death and engaging life as accounts of meaning and culture: A comment on Pyszczynski, Greenberg, Solomon, Arndt, and Schimel (2004). Psychological Bulletin, 130, 473–477. doi:10.1037/ 0033-2909.130.3.473 Ryan, R. M., & Hawley, P. (in press). Naturally good? Basic psychological needs and the proximal and evolutionary bases of human benevolence. In M. Leary, & K. W. Brown (Eds.), The psychology of hypoegoic behavior. New York: Guildford Press. Sacco, D. F., Bernstein, M. J., Young, S. G., & Hugenberg, K. (2014). Reactions to social inclusion and ostracism as a function of perceived in-group similarity. Group Dynamics: Theory, Research, and Practice, 18, 129–137. doi:10.1037/gdn0000002 Schaller, M., & Cialdini, R. B. (1988). The economics of empathic helping: Support for a mood management motive. Journal of Experimental Social Psychology, 24, 163–181. doi:10.1016/0022-1031(88)90019-4 Seligman, M. E. P. (1975). Helplessness: On depression, development, and death. San Francisco, CA: Freeman. Solomon, S., Greenberg, J., & Pyszczynski, T. (1991). A terror management theory of self-esteem and its role in social behavior. In M. Zanna (Ed.), Advances in experimental social psychology (pp. 93–159). New York, NY: Academic Press. Steele, C. M. (1988). The psychology of self-affirmation: Sustaining the integrity of the self. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 21, pp. 261–302). New York, NY: Academic Press. Tangney, J. P., & Fischer, K. W. (1995). Self-conscious emotions: The psychology of shame, guilt, embarrassment and pride. New York, NY: Guilford Press. Vansteenkiste, M., & Ryan, R. M. (2013). On psychological growth and vulnerability: Basic psychological need satisfaction and need frustration as a unifying principle. Journal of Psychotherapy Integration, 23, 263–280. doi:10.1037/ a0032359 Van Tongeren, D. R., Root Luna, L. M., & VanOyen Witvliet, C. (2015). Insufficient justification for rejection elicits compensatory behavior. Journal of Social Psychology, 155(5), 527–534. Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54, 1063–1070. doi:10.1037/ 0022-3514.54.6.1063 Weinstein, N., & Ryan, R. M. (2010). When helping helps: Autonomous motivation for prosocial behavior and its influence on well-being for the helper and recipient. Journal of Personality and Social Psychology, 98, 222–244. doi:10.1037/a0016984 Wesselmann, E. D., Wirth, J. H., Pryor, J. B., Reeder, G. D., & Williams, K. D. (2013). When do we ostracize? Social Psychological and Personality Science, 4, 108–115. doi:10.1177/1948550612443386 Williams, K. D. (2007a). Ostracism. Annual Review of Psychology, 58, 425–452. doi:10.1146/annurev.psych.58.110405. 085641 Williams, K. D. (2007b). Ostracism: The kiss of social death. Social and Personality Psychology Compass, 1, 236–247. doi:10.1111/j.1751-9004.2007.00004.x Williams, K. D. (2009). Ostracism: A temporal need-threat model. In M. Zanna (Ed.), Advances in experimental social psychology (Vol. 41, pp. 279–314). New York, NY: Academic Press. Williams, K. D., Cheung, C. K. T., & Choi, W. (2000). CyberOstracism: Effects of being ignored over the Internet. Journal of Personality and Social Psychology, 79, 748–762. doi:10.1037/0022-3514.79.5.748 Williams, K. D., & Nida, S. A. (2011). Ostracism: Consequences and coping. Current Directions in Psychological Science, 20, 71–75. doi:10.1177/0963721411402480 Zadro, L., & Gonsalkorale, K. (2014). Sources of ostracism: The nature and consequences of excluding and ignoring others. Current Directions in Psychological Science, 23, 93–97. doi:10.1177/0963721413520321




The Role of Burden and Deviation in Ostracizing Others a

b

a

Eric D. Wesselmann , James H. Wirth , John B. Pryor , Glenn D. a

Reeder & Kipling D. Williams

c

a


b

The Ohio State University at Newark

c

Purdue University Published online: 12 Aug 2015.

Click for updates To cite this article: Eric D. Wesselmann, James H. Wirth, John B. Pryor, Glenn D. Reeder & Kipling D. Williams (2015) The Role of Burden and Deviation in Ostracizing Others, The Journal of Social Psychology, 155:5, 483-496, DOI: 10.1080/00224545.2015.1060935 To link to this article: http://dx.doi.org/10.1080/00224545.2015.1060935





The Role of Burden and Deviation in Ostracizing Others ERIC D. WESSELMANN



JAMES H. WIRTH The Ohio State University at Newark

JOHN B. PRYOR and GLENN D. REEDER Illinois State University

KIPLING D. WILLIAMS Purdue University

ABSTRACT. Ostracism (being excluded and ignored) is a painful experience, so why do individuals ostracize others? Previous research suggests individuals often ostracize those who are deviate, but not always. We posit that there may be two types of deviation, burdensome and non-burdensome, and the former is most likely to be ostracized. Study 1 manipulated burdensome deviation by programming a group member to perform more slowly (8 or 16 sec.) than others (4 sec.) in a virtual ball-toss game. Participants perceived slower players as more burdensome and deviate than normal speed players. Additionally, participants ostracized (gave fewer ball tosses to) the slowest player. Study 2 examined participant responses to both burdensome deviation (8- and 16-sec. players) and non-burdensome deviation (goth appearance). Participants again perceived the slower players to be burdensome and deviate, and ostracized them. They perceived the goth player to be deviate but not burdensome and did not ostracize this player. Keywords: burden, deviate, deviation, social control, social influence, sources of ostracism

OSTRACISM (BEING EXCLUDED AND IGNORED) is a painful social experience that occurs in both humans and other social animals (Williams, 2009). Individuals recognize that ostracism hurts the target and individuals can even experience the ostracized individual’s pain vicariously (Wesselmann, Williams, & Hales, 2013). Given the acknowledgement that ostracism harms others, we ask the general question: Whom do individuals choose to ostracize? Researchers initially contended that deviation in attitudes, behaviors, or characteristics (Archer, 1985) leads to punishment and social rejection/ostracism (Wahrman & Pugh, 1972).1 Address correspondence to Eric D. Wesselmann, Illinois State University, Department of Psychology, Campus Box 4620, Normal, IL 61790-4620, USA. E-mail: [email protected] Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/vsoc.


484


Punitive ostracism is an effective social influence tactic, serving both to motivate deviate individuals to conform and also to warn others against future deviation (Williams, 2009). Classic research conducted by Schachter (1951) found that eventually group members stopped talking to a deviate group member when he refused to comply with the group’s norm. Further, group members did not choose him for future interactions (and some actively requested he not be included). These findings were replicated across several studies (Berkowitz & Howard, 1959; Emerson, 1954; Schachter et al., 1954; Tata et al., 1996; Wesselmann, Williams, Pryor et al., 2014). However, deviate individuals are not always treated negatively—sometimes they are perceived as helpful and even revered (Eisinger & Mills, 1968; Moscovici, 1976). Deviate individuals help groups by enhancing creativity and innovation (Nemeth & Staw, 1989; Nemeth & Wachtler, 1983), inducing positive change via criticism (Hornsey & Imani, 2004; Hornsey, Oppes, & Svensson, 2002), and introducing diversity that can combat group members all thinking alike (groupthink; Postmes, Spears, & Cihangir, 2001; Stasser, Stewart, & Wittenbaum, 1995). This leads us to ask: How do we understand which deviates are ostracized and which are not?

BURDEN AS A MOTIVATOR FOR OSTRACISM Wesselmann and colleagues (Wesselmann, Williams, & Wirth, 2014; Wesselmann, Wirth, Pryor, Reeder, & Williams, 2013) examined how individuals use ostracism from an evolutionary psychological perspective, which argues that groups primarily ostracize members who are poor social exchange partners, whose social costs outweigh their benefits, or who do not meet the criteria for being a valuable group member (Kurzban & Leary, 2001). Basically, groups ostracize those who encumber a group from achieving its purpose (Wirth, Turchan, Zimmerman, & Bernstein, 2014). Wesselmann and colleagues (Wesselmann, Williams, & Wirth, 2014; Wesselmann, Wirth et al., 2013) operationalized this by creating an online group interaction in which one confederate performed poorly, thus hindering the group’s ability to engage in the task. Participants played a game called Cyberball (Williams, Cheung, & Choi, 2000), ostensibly online with other players. In reality, the other players were computer-controlled confederates. Wesselmann, Wirth et al. (2013) told participants the goal of the game was mental visualization (e.g., imagine what the weather and the other players were like) and that they should use the ball-tossing to facilitate visualization. The researchers controlled the speed that a particular confederate tossed the ball (4 or 16 sec. between each toss). Participants perceived the slower confederate as burdensome; subsequently, participants ostracized this confederate (i.e., allocated them less-than-equal tosses) during the interaction. Wesselmann, Williams, and Wirth (2014) increased the manipulation salience by explicitly telling participants the goal of Cyberball was to visualize the situation while finishing the game as quickly as possible. Participants again perceived the slow player as burdensome, allocated that person fewer tosses, and wished to avoid the confederate in any future interactions. We believe that the burden-ostracism perspective can help elucidate the conflicting literature on when deviation is lamented or lauded. In both Schachter’s (1951) and Wesselmann and colleagues (Wesselmann, Williams, & Wirth, 2014; Wesselmann, Wirth et al., 2013) research, the paradigms yoked burden and deviation together. Whether a confederate simply refused to change his or her deviate opinion or behaved more slowly than other groups members, in both cases


WESSELMANN ET AL.

485

the person was different from group norms and was hindering group efforts. However, contextual factors may influence whether the deviation is perceived as burdensome. For example, Berkowitz and Howard (1959) found that groups low in interdependence were less likely to reject a deviate member than high interdependence groups. For these groups, a deviate opinion would not be as threatening to their group’s perceived unanimity and thus would not be as burdensome to their functioning (Festinger, 1950). We posit that there may be two types of deviation—burdensome and non-burdensome deviation—and the former is most likely to be rejected/ostracized. We present two studies that replicate and extend previous research investigating ostracism of a burdensome group member. Study 1 establishes that a slow player in Cyberball is perceived as both burdensome and deviate. Study 2 investigates whether deviate individuals are always ostracized or if burdensome deviation is the most likely to be ostracized. Further, Study 2 assesses if participants recognize that their lack of tosses to slow players make those players feel ostracized and whether playing with slow players makes participants feel ostracized.

STUDY 1 We hypothesized that participants would perceive slower group members (8 and 16 sec.) as increasingly more burdensome, more deviate, and participants would be more likely to ostracize the group member, compared to a normal speed (4-sec.) group member. Support for these hypotheses would establish that deviation and burden coincide to facilitate ostracism.

METHOD We report all data exclusions and manipulations (data and measures available at osf.io/nrf72).2 Participants and Design Seventy students participated to earn credit in an elementary psychology course. We removed any participants who previously played or learned about Cyberball (N = 6) or experienced technological failures (N = 1), which left a final sample of 63 participants (46 males, 17 females; Mage = 19.57, SDage = 1.19; 76.20% Caucasian). In a between-participants design, we randomly assigned participants to play Cyberball with a confederate delaying 4, 8, or 16 sec. before throwing (normal delay is 4 sec.; Wesselmann, Williams, & Wirth, 2014; Wesselmann, Wirth et al., 2013). Procedure We instructed participants to mentally visualize all aspects of an online ball-toss scenario (i.e., Cyberball; Williams et al., 2000) and that their ball-toss performance was unimportant. Additionally, we told participants the game would last 45 tosses and to use the 1 and 2 key to throw to the left or right player (respectively). Participants played with two computer-controlled confederates. We manipulated the speed of one confederate player, labeled the target player. The target player took 4, 8, or 16 sec. between

486


each ball toss. Across all conditions, the non-target confederate always took 4 sec. We counterbalanced the target player’s position and the confederate tossed the ball to the non-target confederate and the participant equally. Dependent Variables


Ostracizing behavior. We examined participants’ ostracism behavior by calculating the percent of total throws they gave to the target player. Target player evaluations. Participants answered questions about both confederate players; we randomized which player was asked about first and the order of the questions. Participants completed questions about one confederate before immediately answering the same questions about the other confederate. Participants answered all questions on a 7-point scale ranging from 1 (not at all) to 7 (very much so) unless otherwise noted. Each question included a screen shot of the Cyberball game with the relevant player highlighted. As a manipulation check, participants indicated how slowly the confederate played. We created mean scores for each of the other measures; we report Spearman-Brown coefficients for reliability of two-item measures (Eisinga, Te Grotenhuis, & Pelzer, 2013). Participants reported the degree to which the player was perceived as deviate (“The left/right player was different from us,” “The left/right player was deviant from the group”; Spearman-Brown coefficient = .80), and burdensome (“The left/right player was costing me time,” “The left/right player was wasting my time,” and “How much of a burden was the left/right player to the group”; Cronbach’s α = .88). Two questions assessed participants’ desired future interaction with each player (how interested they would be in doing “another task” and “playing Cyberball again” with each player; SpearmanBrown coefficient = .88). Table 1 lists descriptive statistics for each measure.

RESULTS For the 4-sec. condition, both confederates performed similarly so we randomly selected one player to be the target player for analysis purposes. We conducted a multivariate analysis of

TABLE 1 Study 1: Means and Standard Deviations for Dependent Variables 4-sec. (N = 20) Manipulation check Percent tosses to target player Desired future interaction Target player deviation Target player burden

2.85 .47 3.50 2.72 2.78

(1.53)a (.16)a (1.87)a (1.32)a (1.29)a

8-sec. (N = 27) 5.67 .46 2.04 4.39 4.96

(1.54)b (.19)a (1.47)b (1.61)b (1.57)b

16-sec. (N = 16) 6.56 .28 2.03 5.09 5.58

(.73)c (.14)b (1.04)b (1.83)b (1.92)b

Note. Standard deviations appear in the parentheses. Means in the same row with different subscripts are significantly different (p < .020). All post-hoc tests were Tukey’s HSD except for the manipulation check, which violated Levene’s Test of Equality of Error Variances (p = .004). For this variable we used the Dunnet T3 adjustment, which is a conservative adjustment regarding Type 1 errors (Field, 2013). Differences are significant p < .050.


WESSELMANN ET AL.

487

variance on the five dependent variables with experimental condition as a between-subjects factor. We calculated confidence intervals for each effect size using Wuensch’s SPSS macros and recommendations for ηp2 and d-scores (Wuensch, 2012, 2014). We found an overall multivariate effect of condition, Wilks’ Lambda = .35, F(10, 112) = 7.79, p < .001, ηp2 =.41 (90% C.I. .25−.46), affecting participants’ tosses to the target (F(2,60) = 6.88, p = .002, ηp2 =.19, 90% C.I. .05−.31), perceived target slowness (F(2,60) = 37.61, p < .001, ηp2 =.56, 90% C.I. .40−.64), perceived burden (F(2,60) = 20.42, p < .001, ηp2 =.40, 90% C.I. .23−.52), perceived deviation (F(2,60) = 11.09, p = .001, ηp2 =.27, 90% C.I. .11−.39), and desired future interaction (F(2,60) = 6.34, p = .003, ηp2 =.17, 90% C.I. .04−.30). We conducted post-hoc tests on each of these dependent variables across speed conditions (see Table 1 for p values). Participants perceived the 16-sec. player as taking longer than the 8- and 4sec. players (ds = 0.81 and 3.21, respectively, 95% C.I.s .16−1.45 and 2.14−4.26) and the 8-sec. player longer than the 4-sec. player (d = 1.83, 95% C.I. 1.13−2.51), suggesting our manipulation was effective. Participants allocated fewer ball tosses to the 16-sec. target player compared to 8and 4-sec. players (ds = 1.00 and 1.26, 95% C.I.s 0.34−1.65 and 0.53−1.98), but there was no significant difference in tosses between 8- and 4-sec. players (d = 0.09, 95% C.I. −0.49−.66). Participants perceived both the 16- and 8-sec. players as more deviate than the 4-sec. player (ds = 1.51 and 1.11, 95% C.I.s 0.75−2.25 and 0.49−1.73), but there was no significant difference between 16- and 8-sec. players (d = 0.42, 95% C.I. −0.21−1.04). Participants also perceived both the 16- and 8-sec. players as more burdensome than the 4-sec. player (ds = 1.89 and 1.63, 95% C.I.s 1.09−2.68 and 0.96−2.29), but there was no significant difference between 16- and 8-sec. players (d = 0.33, 95% C.I. −0.29−0.95). Finally, participants desired less future interaction with the 16- and 8-sec. players compared to the 4-sec. players (ds = 1.00 and 0.89, 95% C.I.s 0.29−1.70 and 0.28−1.49), but there was no significant difference between 16- and 8-sec. players (d < 0.01, 95% C.I. −0.18−0.19).

DISCUSSION This study replicated past findings that participants ostracize (i.e., allocate fewer tosses to) a group member who impedes the group (i.e., delayed game completion by playing slowly; Wesselmann Williams, & Wirth, 2014; Wesselmann, Wirth et al., 2013). Participants had uniformly negative reactions to the 16-sec. player, but their reactions to the 8-sec. player were less consistent. Participants perceived that the 8-sec. player played less slowly than the 16-sec. player, but perceived these two slower players similarly in terms of deviation and burden. We note, however, that the effect sizes were in the small-medium range (Cohen, 1988) and our small sample size suggests we may be underpowered to detect significant differences between these two conditions on burden and deviation. Wesselmann, Williams, and Wirth (2014) found a significant difference on perceived burden between 8- and 16-sec. and had a similar effect size to the current study but had twice the number of participants. Further, participants had paradoxical responses with the two ostracism-relevant measures when comparing the 8- and 4-sec. players. Participants included these two players at statistically similar levels, but desired less future social interaction with the 8-sec. player. These

488



findings suggest participants may tolerate a small amount of burdensome deviation during the game but when given the option to avoid this person in the future they want to avoid them. We note that participants’ similar tossing behavior between these two conditions conflicts with Wesselmann, Williams, and Wirth (2014). Given the small effect size we could conclude that it may be due to lower statistical power, but the effect sizes between these two studies are different in size (in Wesselmann, Williams, & Wirth, 2014, d = .74). However, this is only the second study to use an 8-sec. condition and future replications should use this moderately slow condition to elucidate participants’ conflicting reactions.

STUDY 2 We aimed to replicate the burden effects found in Study 1 and separate reactions to burdensome versus non-burdensome deviation. We used a manipulation of deviation—a target player who either has a goth or non-goth appearance. Members of the goth subculture are often perceived as deviate by members of mainstream society, particularly in terms of musical and visual aesthetics (Goodlad & Bibby, 2007; Haenfler, 2009). We selected two female confederates (for replication purposes) and created pictures of each of them having either a non-goth or goth appearance for our deviation manipulation. We hypothesized that the goth player would be perceived as more deviate than the non-goth player, but there should be no difference for perceived burden or ostracism. We hypothesized slow players (8 and 16 sec.) would be seen as more burdensome, deviate, and more likely to be ostracized than a normal speed (4-sec.) player. We also examined the possibility that participants may feel ostracized themselves when playing with a slow player. Koudenburg and colleagues (2011) found that individuals experience uncomfortable silences during discussions as subtle forms of ostracism; it is possible that participants may interpret pauses between ball tosses similarly and they are ostracizing the slow player reciprocally.

METHOD We report all data exclusions and manipulations (data and measures available at osf.io/nrf72).

PARTICIPANTS AND DESIGN Elementary psychology students (N = 178) participated to earn partial course credit. We removed participants with missing data due to technological malfunctions (N = 6), those who requested their data discarded (N = 3), and those who indicated previous knowledge of Cyberball (N = 10). This left 159 participants (109 males; 50 females, Mage = 19.45, SDage = 1.37; 73.6% Caucasian) whom we assigned randomly to a 3 (Target player speed: 4/8/16 sec.) × 2 (Appearance deviation: goth vs. non-goth) between-participants design.

WESSELMANN ET AL.

489


PROCEDURE At the beginning of the study, the experimenter took digital photographs of each participant and explained that the pictures would facilitate mental visualization. Further, they would see their group members’ photographs during Cyberball but they would not be able to see their own; the other group members would see their photograph, however. Participants then filled out personality measures while the experimenter ostensibly uploaded the photographs. Before Cyberball began, participants saw their group members’ photographs on the screen; during Cyberball they saw smaller versions next to the Cyberball players (Figure 1). We used the player photographs to manipulate appearance deviation: participants always saw a non-target player who was non-goth and a target player who was either goth or non-goth.3 We used the same player speed manipulation and game instructions from Study 1.

Not-goth appearance

Goth appearance

Cyberball game

Right Player

Left Player

You FIGURE 1 Photographs used to manipulate appearance deviation in Study 2.

490



DEPENDENT MEASURES We used percent tosses again as our ostracism measure. We administered the questions about both confederate players in the same way as Study 1, but the image accompanying each question was the appropriate group member’s picture. We asked the same manipulation check and desired future interaction (Spearman-Brown coefficient = .89) items from Study 1. We added more items to Study 1’s deviation and burden items: we had seven deviation items (e.g., The left/right player was atypical, . . . was unusual; Cronbach’s α = .90) and eight burden items (e.g., The left/right player was hindering me; . . . was handicapping me; Cronbach’s α = .996). Finally, we asked participants the degree to which both the participant (Spearman-Brown coefficient = .80; 1— not at all, 5—extremely) and the target player felt “ignored” and “excluded” (Spearman-Brown coefficient = .91; 1—not at all, 7—extremely).

RESULTS We randomly selected the target player when the confederate players were equivalent (i.e., 4-sec. and non-goth appearance). We conducted a multivariate analysis of variance on the seven dependent variables with both experimental conditions as a between-participants factor. We found an overall multivariate effect of player speed, Wilks’ Lambda = .45, F(14, 294) = 10.15, p < .001, ηp2 =.33 (90% C.I. .22−.36), affecting participants’ tosses to the target (F(2,153) = 17.76, p < .001, ηp2 =.19, 90% C.I. .10−.27), perceived target slowness (F(2,153) = 74.90, p < .001, ηp2 =.49, 90% C.I. .40−.56), perceived burden (F(2,153) = 39.52, p < .001, ηp2 =.34, 90% C.I. .24−.42), perceived deviation (F(2,153) = 18.18, p < .001, ηp2 =.19, 90% C.I. .10−.27), desired future interaction (F(2,153) = 17.08, p < .001, ηp2 =.18, 90% C.I. .09−.26), and perceptions of the target’s ostracized feelings (F(2,153) = 21.00, p < .001, ηp2 =.21, 90% C.I. .12−.30). It did not affect participants’ feeling of being ostracized (F(2,153) = .46, p = .634, ηp2 =.01, 90% C.I. .00−.03). We conducted post-hoc tests on each of these dependent variables across the target player speed conditions (see Table 2 for descriptive statistics and p values). Participants perceived the 16-sec. player as taking longer than the 8- and 4-sec. players (ds = 1.15 & 3.04, respectively; 95% C.I.s 0.73−1.57 and 2.47−3.61) and the 8-sec. player longer than the 4-sec. player (d = 1.02, 95% C.I. 0.61−1.43), replicating Study 1. Again, participants allocated fewer ball tosses to the 16-sec. target player compared to 8- and 4-sec. players (ds = 0.73 and 1.02, 95% C.I.s 0.34−1.12 and 0.61−1.43), but there was no significant difference in tosses between 8and 4-sec. players (d = 0.38, 95% C.I. −0.01−0.76). Participants perceived the 16-sec. player as more deviate than the 8- and 4-sec. players (ds = 0.65 and 1.18, 95% C.I.s .26−1.04 and 0.76 − 1.59) and the 8-sec. more deviate than the 4-sec. player (d = 0.47, 95% C.I. 0.08−0.86). Participants also perceived the 16-sec. player as more burdensome than the 8- and 4-sec. players (ds = 0.97 and 1.83, 95% C.I.s 0.57−1.37 and 1.37−2.28) and the 8-sec. more burdensome than the 4-sec. player (d = 0.67, 95% C.I. 0.27−1.06). Participants desired less future interaction with the 16-sec. target player compared to 8- and 4-sec. players (ds = 0.86 and 1.17, 95% C.I.s 0.46−1.26 and 0.76 − 1.59), but there was no significant difference between 8- and 4-sec. players (d = 0.24, 95% C.I. −0.15−0.62); this pattern only partially replicated Study 1.

491

Not-Goth Goth Total Not-Goth Goth Total Not-Goth Goth Total Not-Goth Goth Total

.47 (.07) .54 (.12) .50 (.10)a .47 (.07) .44 (.14) .46 (.11)a .34 (.17) .35 (.20) .35 (.19)b .43 (.13) .44 (.17) .44 (.15)

1.96 2.28 2.11 3.52 2.77 3.16 4.82 4.89 4.85 3.45 3.31 3.39

(1.18) (1.44) (1.31)a (1.93) (1.61) (1.81)b (1.82) (1.51) (1.67)c (2.04) (1.88) (1.96)

Perceived burden 2.33 3.36 2.81 3.41 3.59 3.50 4.19 4.76 4.45 3.32 3.90 3.59

(.88) (1.60) (1.36)a (1.56) (1.49) (1.52)b (1.54) (1.24) (1.43)c (1.56) (1.56) (1.58)

Perceived deviation 2.18 2.56 2.36 4.48 4.08 4.29 6.55 6.08 6.33 4.43 4.24 4.34

(1.28) (1.66) (1.47)a (2.24) (2.23) (2.23)b (.78) (1.38) (1.12)c (2.37) (2.29) (2.32)

Manipulation check 3.70 3.48 3.59 2.93 3.48 3.19 1.72 2.12 1.91 2.77 3.03 2.89

(1.67) (1.64) (1.65)a (1.77) (1.70) (1.74)a (1.00) (1.36) (1.18)b (1.71) (1.68) (1.69)

Desired future interaction 1.84 2.10 1.96 2.04 1.94 1.99 1.65 2.00 1.81 1.84 2.01 1.92

(.95) (.92) (.94)a (.94) (.96) (.94)a (.95) (.97) (.96)a (.95) (.94) (.95)

Participant’s ostracized feelings

1.84 1.48 1.67 2.50 2.24 2.37 2.64 4.26 3.39 2.33 2.66 2.48

(1.07) (.57) (.88)a (1.54) (1.29) (1.42)b (1.74) (1.93) (1.99)c (1.51) (1.80) (1.66)

Target’s ostracized feelings

28 25 53 27 25 52 29 25 54 84 75 159

N

Note. Standard deviations appear in parentheses. For main effect of target player speed, means in each column with different subscripts are significantly different (p < .040). All post-hoc tests were Tukey’s HSD.

Total

16-sec.

8-sec.

4-sec.

Target player speed

Appearance deviation

Percent tosses to target player

TABLE 2 Study 2: Means and Standard Deviations for Dependent Variables



492


Participants perceived the 16-sec. player as feeling more ostracized than the 8- and 4-sec. players (ds = 0.59 and 1.12, 95% C.I.s 0.20−0.98 and 0.57−1.54) and the 8-sec. player more ostracized than the 4-sec. player (d = 0.59, 95% C.I. 0.20−0.99). We found an overall multivariate effect of player appearance (i.e., goth vs. not-goth), Wilks’ Lambda = .87, F(7, 147) = 3.10, p = .004, ηp2 =.13 (90% C.I. .02−.18), but the manipulation only affected deviation ratings (F(1,153) = 7.11, p = .008, ηp2 =.04, 90% C.I. .01−.11). As expected, participants perceived the goth player as more deviate than the non-goth player. This manipulation did not significantly affect any other variable (Fs < 2.15, ps > .145, ηp2 s < .01, 90% C.I.s .00−.06). Interestingly, we found a two-way interaction between player speed and appearance, Wilks’ Lambda = .80, F(14, 294) = 2.54, p = .002, ηp2 =.11 (90% C.I. .02−.12), but only for participants’ perceptions of how ostracized the target player felt (F(2,153) = 8.05, p < .001, ηp2 = .09, 90% C.I. .03−.17; other dependent variables, Fs < 1.72, ps > .209, ηp2 s < .01, 90% C.I.s .00−.06). We decomposed the significant interaction using simple effects with Bonferroni adjustments. The pattern suggests that the 16-sec. goth player felt more ostracized than the 16sec. normal player (F(2,153) = 17.14, p < .001, d = 0.88, 95% C.I. 0.32−1.44); the player’s appearance did not matter for the other two speeds (Fs < .83, ps > .364, ds < 0.42, 95% C.I.s −0.36−.95).

DISCUSSION Study 2 replicated most findings from Study 1 regarding negative reactions to slow players. Further, participants recognized that the slow player felt ostracized during the game (which had to be because of the participant; we programmed the other confederate to include each player equally). Also, participants did not feel ostracized themselves when playing with a slow player, ruling out the possibility they were only ostracizing the slow player reciprocally. Although participants perceived a goth player as more deviate than a non-goth player, they did not perceive this player as more burdensome or ostracize them. These data suggest that, at least in Cyberball, slow speed is perceived as a burdensome deviation but a goth appearance is perceived as a non-burdensome deviation; one is ostracized and the other is not. Unexpectedly, participants perceived that the slow goth player felt more ostracized than the normal-appearing slow player even though behaviorally this did not occur. We interpret this finding with caution and suggest future replications and extensions. Additionally, participants again had inconsistent reactions to the 8-sec. player. Participants perceived the 8-sec. player as more deviate and burdensome than the 4-sec. player, but they did not treat the 8-sec. player differently on either ostracism-related measure. The effect sizes for both of these measures were in the small-medium range (Cohen, 1988), and although the lack of tossing behavior still conflicts with Wesselmann, Williams, and Wirth (2014), the effect size is at least closer to previous research than Study 1’s effect. Further, the effect size C.I.s for both ostracism-related measures (as well as tosses in Study 1) have a wide range with the upper limit corresponding to the medium-large range. This suggests that there is considerable variability in participant reactions to the 8-sec. player, which may be explained by individual differences or situational moderators.

WESSELMANN ET AL.

493


GENERAL DISCUSSION We sought to understand when deviation facilitates ostracism. Previous research shows mixed reactions to deviate individuals: sometimes they were ostracized (e.g., Schachter, 1951; Tata et al., 1996) and other times they were valued group members (e.g., Eisinger & Mills, 1968; Moscovici, 1976; Hornsey & Imani, 2004). Our research takes an evolutionary psychology approach (e.g., Kurzban & Leary, 2001) and suggests that the degree of ostracism deviate individuals experience may be contingent upon how burdensome that deviation is perceived to be. In Study 1, when a group member delayed before throwing the ball (i.e., 8 or 16 sec.), the member was perceived as being deviate and burdensome. The slowest (16-sec.) player was ostracized. This study demonstrated how being burdensome was also perceived as being deviate, similar to previous research (e.g., Schachter, 1951; Wesselmann, Williams, & Wirth, 2014; Wesselmann, Wirth et al., 2013) in which burden and deviation coincided. In Study 2, we sought to understand the effects of burdensome versus non-burdensome deviation. Study 2 demonstrated that non-burdensome deviation (i.e., goth appearance) did not prompt ostracism. Rather, when a deviate individual was burdensome—playing slowly—she was ostracized. Our results do present limitations worth noting and raise questions for future investigation. First, we could not design a manipulation to disentangle burden and deviation orthogonally. A slow player was both burdensome and deviate, while a goth player was only deviate. However, creating a burden manipulation that does not also manipulate deviation may not be realistic; burdensome individuals deviate from group norms by nature (Kurzban & Leary, 2001). As an additional limitation, our study examined the links between deviation, burden, and ostracism within a single group context—Cyberball. Cyberball is a social interaction with minimal information, although it is based on a face-to-face interaction paradigm (Williams & Sommer, 1997). We similarly limited the generalizability of our results by manipulating only one type of burden: confederate player speed. A group member may be burdensome for various reasons (e.g., ability; Wirth, Bernstein, & LeRoy, 2015; Wirth, LeRoy, & Bernstein, 2014). Further, appearance deviation may not be perceived as burdensome in Cyberball but may be perceived as burdensome in other contexts, such as professional work groups or situations in which being associated with deviate individuals may reflect poorly on oneself (e.g., courtesy stigma; Goffman, 1963). Finally, we defined ostracism as participants allocating fewer tosses to a target player than equality (Wesselmann, Wirth et al., 2013); we should note that most participants gave a few tosses to slower target players even if this amount was significantly lower than equal inclusion. This behavior may be more accurately called partial ostracism (Gooley, Zadro, Williams, Svetieva, & Gonsalkorale, 2015; Williams, 2009) because participants are not completely shutting a slow target out of the game. To our knowledge, the only paradigms in which participants completely ostracize a target player involve instructions explicitly requesting that participants ostracize the target (e.g., Legate, DeHaan, & Ryan, 2015; Legate, DeHaan, Weinstein, & Ryan, 2013). We believe that our participants’ hesitancy to completely ostracize a slower target player lends further evidence that individuals recognize ostracism is a powerful social weapon that necessitates strong motivation for usage. Our study raises interesting new questions. Previous theory suggests groups ostracize burdensome individuals, but we want to be cautious not to declare that burdensome individuals will always be ostracized. In Study 1, data suggested participants may tolerate some burden during the game, but this tolerance may not extend to future interactions. This pattern partially replicated

494



in Study 2. Future research should examine both individual differences and situational factors that may influence reactions to moderate/ambiguous burden. Another limitation is that the non-target confederate always included the target, regardless of speed or appearance. Wesselmann, Wirth et al. (2013) found that participants ostracized a slow player regardless of how other confederates behaved, but they only used 4- and 16-sec. targets. Future studies should examine whether group norms facilitate ostracism for moderately burdensome deviates (e.g., 8-sec. player) or nonburdensome deviates (e.g., goth appearance). In ambiguous situations, individuals may be more willing to conform to group ostracism. Regardless, our research offers an initial starting-point for understanding reactions to burdensome and non-burdensome deviation during group interactions.

ACKNOWLEDGMENTS We thank Jeremy Rohrer for modifying Cyberball to include manipulating confederate player speed. We also thank Katie Wirth and Amanda Wesselmann for being our deviation models.

FUNDING This material is based upon work supported by the National Science Foundation under Grant No. 0519209 award to KDW.

NOTES 1. The literature on negative reactions to deviate group members sometimes uses the terms deviant or deviance. Deviance implies a negative evaluation but deviation is not always viewed negatively (e.g., Moscovici, 1976). It is unclear if laypeople make these conceptual distinctions, but for consistency and theoretical clarity we use the neutral terms deviate and deviation (e.g., Schachter, 1951). 2. We had other measures besides the one reported in this manuscript; some were filler items/personality measures to bolster the cover story and others were pilot measures (e.g., emotions toward the players; self-reported motives for ball-tossing). These measures were not germane to our hypotheses so we did not analyze them and do not included them here. These data can be found at osf.io/nrf72. Additionally, these data come from a larger data set in which we were piloting instruction manipulations. The instructions for the study reported told participants that the game will last 45 tosses; the other condition instructions specified 5 minutes. This time/tosses instruction manipulation had no significant effects on our dependent variables; therefore we excluded participants in the conditions specifying that the game would last 5 minutes. These data and measures are available from the first author by request. 3. We created the deviate appearance stimuli by taking two sets of pictures using two female confederates. We photographed each confederate in normal appearance and in goth appearance. We asked a pilot sample (N = 71) to rate each of these pictures on a 7-point scale ranging from 1 (not at all deviant) to 7 (extremely deviant). We informed participants: “Deviate means to differ from some commonly accepted standard of appearance, beliefs, or behaviors.

WESSELMANN ET AL.

495

In the United States, how much does this individual deviate from the norm?” We neglected to change the scale anchors from “deviant” to “deviate,” but as previously noted it is unclear if laypeople make conceptual distinctions between these two terms. In both pairs, participants rated the goth appearance picture as more deviate than the non-goth appearance picture (ts > 3.50, p < .01, ds > 0.38). Our stimuli successfully manipulated deviation, but the effect was small (Cohen, 1988) and the photograph means were both below the midpoint (2.21, SD = 1.41 and 3.38, SD = 1.62) so we decided to bolster the images by adding heaver makeup to our confederates, resulting in the images we used in the current study.


AUTHOR NOTES Eric D. Wesselmann is affiliated with Illinois State University. James H. Wirth is affiliated with The Ohio State University at Newark. John B. Pryor and Glenn D. Reeder are affiliated with Illinois State University. Kipling D. Williams is affiliated with Purdue University.

REFERENCES Archer, D. (1985). Social deviance. In G. Lindzey, & E. Aronson (Eds.), Handbook of social psychology (3rd ed., Vol. 2, pp. 743–804). New York, NY: Random House. Berkowitz, L., & Howard, R. C. (1959). Reactions to opinion deviates as affected by affiliation need (n) and group member interdependence. Sociometry, 22, 81–91. Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates. Eisinga, R., Te Grotenhuis, M., & Pelzer, B. (2013). The reliability of a two-item scale: Pearson, Cronbach, or SpearmanBrown? International Journal of Public Health, 58, 637–642. doi:10.1007/s00038-012-0416-3 Eisinger, R., & Mills, J. (1968). Perception of the sincerity and competence of a communicator as a function of the extremity of his position. Journal of Experimental Social Psychology, 4, 224–232. doi:10.1016/0022-1031(68) 90043-7 Emerson, R. M. (1954). Deviation and rejection: An experimental replication. American Sociological Review, 19, 688–693. Festinger, L. (1950). Informal social communication. Psychological Review, 57, 271–282. doi:10.1037/h0056932 Field, A. (2013). Discovering statistics using SPSS (4th ed.). Los Angeles, CA: Sage. Goffman, E. (1963). Stigma: Notes on the management of spoiled identity. Englewood Cliffs, NJ: Prentice-Hall, Inc. Goodlad, L. M. E., & Bibby, M. (Eds.). (2007). Goth: Undead subculture. Durham, UK: Duke University Press. Gooley, S. L., Zadro, L., Williams, L. A., Svetieva, E., & Gonsalkorale, K. (2015). Ostracizing for a reason: A novel source paradigm for examining the nature and consequences of motivated ostracism. Journal of Social Psychology, 155(5), 140–432. Haenfler, P. (2009). Goths, gamers, and grrrls: Deviance and youth subculture. Oxford, UK: Oxford University Press. Hornsey, M. J., & Imani, A. (2004). Criticizing groups from the inside and the outside: An identity perspective on the intergroup sensitivity effect. Personality and Social Psychology Bulletin, 30, 365–383. Hornsey, M. J., Oppes, T., & Svensson, A. (2002). “It’s OK if we say it, but you can’t”: Responses to intergroup and intragroup criticism. European Journal of Social Psychology, 32, 293–307. Koudenburg, N., Postmes, T., & Gordijn, E. H. (2011). Disrupting the flow: How brief silences in group conversations affect social needs. Journal of Experimental Social Psychology, 47, 512–515. doi:10.1016/j.jesp.2010.12.006 Kurzban, R., & Leary, M. R. (2001). Evolutionary origins of stigmatization: The functions of social exclusion. Psychological Bulletin, 127, 187–208. doi:10.1037/0033-2909.127.2.187 Legate, N., DeHaan, C., & Ryan, R. (2015). Righting the wrong: Reparative coping after going along with ostracism. Journal of Social Psychology, 155(5), 471–482. Legate, N., DeHaan, C. R., Weinstein, N., & Ryan, R. M. (2013). Hurting you hurts me too: The psychological costs of complying with ostracism. Psychological Science, 24, 583–588. doi:10.1177/0956797612457951 Moscovici, S. (1976). Social influence and social change. London, UK: Academic Press.


496


Nemeth, C., & Staw, B. M. (1989). The tradeoffs of social control and innovation in groups and organizations. In L. Berkowitz (Ed.), Advances in experimental social psychology: (Vol. 22, pp. 175–210). San Diego, CA: Academic Press. Nemeth, C., & Wachtler, J. (1983). Creative problem solving as a result of majority vs. minority influence. European Journal of Social Psychology, 13, 45–55. doi:10.1002/(ISSN)1099-0992 Postmes, T., Spears, R., & Cihangir, S. (2001). Quality of decision making and group norms. Journal of Personality and Social Psychology, 80, 918–930. doi:10.1037/0022-3514.80.6.918 Schachter, S. (1951). Deviation, rejection, and communication. The Journal of Abnormal and Social Psychology, 46, 190–207. doi:10.1037/h0062326 Schachter, S., Nuttin, J., De Monchaux, C., Maucorps, P. H., Osmer, D., & Duijker, H., . . . Israel, J. (1954). Cross-cultural experiments on threat and rejection. Human Relations, 7, 403–439. doi:10.1177/001872675400700401 Stasser, G., Stewart, D. D., & Wittenbaum, G. (1995). Expert roles and information exchange during discussion: The importance of knowing who knows what. Journal of Experimental Social Psychology, 31, 244–265. doi:10.1006/jesp.1995.1012 Tata, J., Anthony, T., Lin, H., Newman, B., Tang, S., Millson, M., & Sivakumar, K. (1996). Proportionate group size and rejection of the deviate: A meta-analytic integration. Journal of Social Behavior and Personality, 11, 739–752. Wahrman, R., & Pugh, M. D. (1972). Competence and conformity: Another look at Hollander’s study. Sociometry, 35, 376–386. Wesselmann, E. D., Williams, K. D., & Hales, A. H. (2013). Vicarious ostracism. Frontiers in Human Neuroscience, 7. doi:10.3389/fnhum.2013.00153. Wesselmann, E. D., Williams, K. D., Pryor, J. B., Eichler, F. A., Gill, D. M., & Hogue, J. D. (2014). Revisiting Schachter’s research on rejection, deviance, and communication (1951). Social Psychology, 45, 164–169. doi:10.1027/1864-9335/a000180 Wesselmann, E. D., Williams, K. D., & Wirth, J. H. (2014). Ostracizing group members who can (or cannot) control being burdensome. Human Ethology Bulletin, 29, 82–103. Wesselmann, E. D., Wirth, J. H., Pryor, J. B., Reeder, G. D., & Williams, K. D. (2013). When do we ostracize? Social Psychological and Personality Science, 4, 108–115. doi:10.1177/1948550612443386 Williams, K. D. (2009). Ostracism: Effects of being excluded and ignored. In M. Zanna (Ed.), Advances in experimental social psychology (pp. 275–314). New York, NY: Academic Press. Williams, K. D., Cheung, C. K. T., & Choi, W. (2000). Cyberostracism: Effects of being ignored over the Internet. Journal of Personality and Social Psychology, 79, 748–762. Williams, K. D., & Sommer, K. L. (1997). Social ostracism by one’s coworkers: Does rejection lead to loafing or compensation? Personality and Social Psychology Bulletin, 23, 693–706. Wirth, J. H., Bernstein, M. J., & LeRoy, A. S. (2015). Atimia: A new paradigm for investigating how individuals feel when ostracizing others. Journal of Social Psychology, 155(5), 497–514. Wirth, J. H., LeRoy, A. S., & Bernstein, M. J. (2014). “You’re such a pain!”: Pain motivates the ostracism of a burdensome group member. Manuscript submitted for publication. Wirth, J. H., Turchan, P. J., Zimmerman, A. G., & Bernstein, M. J. (2014). Atimia: A novel group-based paradigm for manipulating ostracism and group members’ performance. Group Dynamics: Theory, Research, and Practice, 18, 251–266. doi:10.1037/gdn0000006 Wuensch, K. L. (2012). Using SPSS to obtain a confidence interval for Cohen’s d. Retrieved from http://core.ecu.edu/ psyc/wuenschk/SPSS/CI-d-SPSS.pdf Wuensch, K. L. (2014). Putting confidence intervals on R2 or R. Retrieved from http://core.ecu.edu/psyc/wuenschk/ StatHelp/CI-R2.htm

Received January 12, 2015 Accepted June 3, 2015

This article was downloaded by: [98.31.2.248] On: 20 August 2015, At: 05:33 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: 5 Howick Place, London, SW1P 1WG


Atimia: A New Paradigm for Investigating How Individuals Feel When Ostracizing Others a

b

c

James H. Wirth , Michael J. Bernstein & Angie S. LeRoy a

The Ohio State University at Newark

b

Penn State University–Abington

c

University of Houston Published online: 12 Aug 2015.

Click for updates To cite this article: James H. Wirth, Michael J. Bernstein & Angie S. LeRoy (2015) Atimia: A New Paradigm for Investigating How Individuals Feel When Ostracizing Others, The Journal of Social Psychology, 155:5, 497-514, DOI: 10.1080/00224545.2015.1060934 To link to this article: http://dx.doi.org/10.1080/00224545.2015.1060934


Downloaded by [98.31.2.248] at 05:33 20 August 2015



Atimia: A New Paradigm for Investigating How Individuals Feel When Ostracizing Others JAMES H. WIRTH The Ohio State University at Newark

MICHAEL J. BERNSTEIN Downloaded by [98.31.2.248] at 05:33 20 August 2015

Penn State University–Abington

ANGIE S. LEROY University of Houston

ABSTRACT. To date, researchers studying ostracism (being excluded and ignored) focused on examining the consequences of ostracism. However, researchers have not yet systematically investigated why individuals ostracize others. One impediment to this research is lacking multiple means to successfully induce individuals to be sources, those who ostracize others. Using Cyberball, researchers found participants ostracized a player delaying the game. To aid in systematic research on sources, we developed the game Atimia. In Atimia, players took turns solving remote associate word items and we varied the performance of a computer-controlled player to perform equal to or significantly worse than the group. Without prompting, participants ostracized a poor- versus equal-performing player more and found the poor-performing player burdensome, less likable, and less desirable to work with on a future task. Study outcomes suggest Atimia is a viable paradigm for investigating sources of ostracism. Keywords: burden, exclusion, motives, ostracism, rejection, sources of ostracism

WHEN SOCIAL CONNECTIONS ARE BROKEN, individuals are not fulfilling their basic need to belong, which prompts a number of negative consequences. Ostracized (excluded and ignored) individuals experience thwarted satisfaction of fundamental basic needs (e.g., belonging, selfesteem) and worsened mood (e.g., Wesselmann, Wirth, Mroczek, & Williams, 2012; Williams, Cheung, & Choi, 2000; Wirth, Turchan, Zimmerman, & Bernstein, 2014). At a physiological level, rejected, compared to included, individuals have an increase in their blood pressure, experience elevated levels of stress hormones (Stroud, Tanofsky-Kraff, Wilfley, & Salovey, 2000), and regions of the brain associated with physical pain are activated (Eisenberger & Lieberman, 2004; Eisenberger, Lieberman, & Williams, 2003). Previous research on ostracism (see Williams, 2007; Williams & Nida, 2011) investigated the experiences victims of ostracism Address correspondence to James H. Wirth, The Ohio State University at Newark, Department of Psychology, 1179 University Drive, Newark, OH 43055, USA. E-mail: [email protected]

498



(targets) share. Focusing on the victims of ostracism occurred in part because of the ease of ostracizing participants using Cyberball (Williams, Cheung et al., 2000). In Cyberball, participants toss a virtual ball with two other players online, who are actually computer-controlled agents. Researchers use this paradigm to manipulate the ostracism of participants, such that participants are either included (receiving the ball about a third of the time in a three-player game) or ostracized (only receiving the ball once from each player at the beginning and then never again; e.g., Wesselmann et al., 2012; Williams, Cheung et al., 2000; Wirth & Williams, 2009; Zadro, Williams, & Richardson, 2004). Cyberball has been employed widely, with researchers utilizing the paradigm to conduct studies with over 5,000 participants (Williams & Nida, 2011). Cyberball helped researchers understand how individuals feel when they are ostracized, but paradigms also need to be developed to understand how individuals feel when they ostracize others.

OSTRACISM SOURCE PARADIGMS What is known about how it feels to be a source of ostracism comes from studies using diverse methods. Williams, Bernieri, Faulkner, Gada-Jain, and Grahe (2000) used a daily diary method over the course of five days; each author took a turn being ostracized for a day. Over the five days participants reported how it felt to be ostracized and how it felt to ostracize a fellow colleague. The authors found being a source of ostracism was unpleasant, cognitively draining, and often reported difficulty when ostracizing others (Williams, Bernieri et al., 2000). This was especially true when the relationship was close, as demonstrated by Baumeister, Wotman, and Stillwell (1993), who found that individuals who turned down an offer of affection reported the experience as being particularly unpleasant and stressful. These results are corroborated by the unpleasantness reported by confederates in an experiment when they were asked to ostracize participants by not including them in an in-person ball-toss game (Williams & Sommer, 1997) and by participants reporting that ostracizing others was a more difficult task than conversing with them (Ciarocco, Sommer, & Baumeister, 2001). To evaluate the consequences of being a source of ostracism, several researchers encouraged participants to reject another individual through suggestion or explicitly in the experiment instructions. For example, when researchers asked participants to assist by being the source (Ciarocco et al., 2001) of ostracism, participants experienced self-control depletion. Specifically, participants were less persistent on an unsolvable anagram task and held a handgrip for a shorter amount of time than those allowed to converse with a confederate. Individuals who were instructed to reject others, by either rejecting a job applicant, rejecting an author’s research report, or sending a negative evaluation over instant messenger, felt less of a desire to socially connect with others (Zhou, Zheng, Zhou, & Guo, 2009). In another study, participants had to reject an applicant in a competition for an outstanding thesis prize. Following this, participants reported experiencing pain which, according to the authors, was due to concerns that by rejecting another individual, they themselves (as the source of the rejection) might conversely become a future target of rejection (Chen, Poon, Bernstein, & Teng, 2014). Additional research paradigms on sources include instructing participants to ostracize others in a group context. For instance, Poulsen and Kashy (2012) set up a group interaction in which


WIRTH, BERNSTEIN, AND LEROY

499

three group members were either instructed explicitly to ostracize a fourth during a group task or were given no instructions regarding how to treat any of the group members. When participants were told to ostracize others, the researchers found that sources of ostracism felt guilty for having done so. In a similar group context, researchers (Zadro, Williams, & Richardson, 2005) had participants ride the O-train, a role-playing scenario in which two participants (sources) either ostracized a third participant (a target) or argued and insulted the target who sat between the sources on the train. The researchers found sources of ostracism experienced greater need satisfaction than sources who argued. In yet another example, Legate, DeHaan, Weinstein, and Ryan (2013) examined sources of ostracism in virtual group interactions. In their first study, participants were asked to ostracize a player in Cyberball, were not given ball-toss instructions, or were instructed to include both players equally. Those instructed to ostracize reported lower levels of autonomy and relatedness and greater amounts of negative affect compared to the control groups. In a follow-up study (Legate et al., 2013), participants who were asked to ostracize others felt more guilt and shame, and lower autonomy, compared to participants who were ostracized from the group. Bastian et al. (2013; Study 2) used a similar Cyberball manipulation and found sources of ostracism felt less human than those who included the Cyberball player. Bastian et al. replicated this effect across several source paradigms.

LIMITATIONS OF PREVIOUS OSTRACISM SOURCE PARADIGMS All of this research indicates that being the source of ostracism or rejection is an aversive experience and that it is possible to create situations where people do indeed ostracize or reject others, but these examples all have a noteworthy limitation; in each case, participants were explicitly asked to ostracize or reject a group member rather than doing it of their own accord. Although this may seem trivial, work by Sommer and Yoon (2013) suggests that the motivation for why people ostracize another individual matters. In their study, participants were less successful selfregulating when they had to ignore, versus converse with a likeable person, but the opposite effect occurred when ignoring, versus conversing with an unlikable person. This study suggests participants dislike treating others negatively without just cause, thus suggesting that motivations for ostracizing and perceptions of the target matter. Further, asking participants to ostracize an individual could be likened to Milgram’s (1963, 1974) work on obedience. In this seminal work, an experimenter requested that participants continue to shock an individual failing to perform well on a learning task. Many participants obeyed the experimenter and shocked the victim (a confederate who was not actually being shocked) all the way to the point at which participants had reason to believe they had severely harmed confederate (potentially fatally). Importantly, their obedience to authority created a situation of blame shifting or of rationalizing their behavior (e.g., “I have to do it because he is making me”). It is possible that when experimenters or experimental instructions indicate to participants that they should ostracize another participant, this may be creating a situation of obedience to an authority at the very moment when they are requesting participants to act in an antisocial fashion. Exemplifying this point, in a study involving confederates ostracizing participants, Williams and Jarvis (2006) indicated, “the task of ignoring and excluding is an arduous one that confederates approach first with concern and sympathy, but later, seemingly, with sadistic glee” (p. 174). Part of

500


what researchers may find when asking participants to be sources of ostracism is the source’s negative response to recognizing that they are causing an individual pain (see Wesselmann, Williams, & Hales, 2013). Considering that participants report feeling guilty (Poulsen & Kashy, 2012), less self-control (Ciarocco et al., 2001), and less autonomy (Legate et al., 2013) in response to ostracizing another person, these responses could capture an individual feeling compelled to comply with an antisocial request. Recent research paradigms address issues related to asking individuals to ostracize rather than individuals ostracizing of their own accord.


PARADIGMS FOR PARTICIPANTS ACTING AS SOURCES INDEPENDENT OF INSTRUCTION To best assess how individuals feel when they ostracize others, researchers need a multimethod approach to study when individuals ostracize others of their own volition. Zadro and Gonsalkorale (2014) explained the current source paradigm limitations and need for alternative paradigms when they said, “Existing source paradigms focus on situations in which sources model the ostracizing behavior of others or are induced to ostracize. Thus, what is missing is a paradigm that motivates participants to engage in ostracism of their own accord” (p. 96). Thus far, few paradigms can cause participants to engage in ostracism of their own choice. One method used previously to understand when people will ostracize or reject others is to have participants recall a time they ostracized or rejected others. For instance, Sommer, Williams, Ciarocco, and Baumeister (2001) asked participants to write a story about when the participant used the silent treatment. Other researchers (Chen et al., 2014) asked participants to recall a time they rejected another person. The approach of having participants recall a time they ostracized or rejected others is beneficial because researchers were not asking participants to ostracize or reject an individual, thereby examining ostracizing or rejecting others autonomously (acting of their own accord). However, recall paradigms can be limiting. Specifically, researchers (Schwarz, Groves, & Schuman, 1998) found that individuals tend to recall instances of an event that are extreme or highly emotional, meaning the experience of ostracizing or rejecting others that participants recalled may not be representative of typical experiences being a source of ostracism or rejection. Cyberball as a Source Paradigm As another approach to investigating sources of ostracism, researchers adapted Cyberball to aid in investigating how individuals feel when they ostracize of their own accord. For instance, Gooley, Zadro, Gonsalkorale, Williams, and Sethi (2013) had participants interact with an objectionable partner during a get-to-know-you game before playing Cyberball. In Cyberball, participants ostracized the obnoxious partner without instruction to do so, thereby successfully prompting individuals to be sources of ostracism. Prior to this study, researchers (Wesselmann, Wirth, Pryor, Reeder, & Williams, 2013) also used Cyberball to induce participants to ostracize a group member. In a slow-player game of Cyberball, participants tossed the ball with a player manipulated to hold the ball briefly (3–5 seconds), similar to the participant, or to delay throwing for a considerable amount of time (16 seconds). Participants found the slow-throwing player to be burdensome and were less likely to include this player compared to a normal-throwing player. The researchers


501

(Wesselmann, Williams, & Wirth, 2014) replicated these results while further investigating the motives for ostracizing. Cyberball may be one means to investigate systematically why individuals ostracize and how they feel when ostracizing. A slow-player version of Cyberball can be applied systematically because it can be administered consistently, allowing for comparisons across findings. However, an additional method for investigating sources is needed in order to create a multimethod approach to investigating sources (see also Gooley, Zadro, Williams, Svetieva, & Gonsalkorale, 2015).


ATIMIA In the current research, we aimed to develop a paradigm that would achieve three central goals: (a) to replicate and extend previous work demonstrating that individuals ostracize burdensome group members and to investigate perceptions of the burdensome group member, (b) to induce individuals to ostracize others without explicitly instructing them to do so, and (c) to create another method for systematically investigating how individuals feel as sources of ostracism, thus creating a multimethod approach. To achieve these three goals, we developed a computer game called Atimia.1 In this game, participants take turns playing with computer-controlled agents solving Remote Associates Test (RAT) items (Bowden & Jung-Beeman, 2003; Mednick, 1968). During an RAT trial, participants are given three words and asked to find a fourth word that is related to the three other disparate words (e.g., “play,” “fold,” and “duck” are all related to the fourth word, “bill,” i.e., playbill, billfold, duckbill). During a previous study using Atimia (Wirth et al., 2014), we manipulated the performance of the computer agents to create players that performed either similar or worse than the participant and either ostracized or included the participant. The poor-performing players burdened the group due to impeding the group from achieving its goal. In the current work, we tested if a computer agent performing poorly was perceived as burdening the group and whether this motivated the participants to ostracize the burdensome player. Demonstrating this outcome would mean Atimia can be used as an additional means for studying sources of ostracism. Specifically, we hypothesized that a poor- versus equal-performing group member will be ostracized more, perceived as more burdensome, and liked less. To further demonstrate the averseness of a poor-performing group member, we also hypothesized that a poor-performing group member will be ostracized more, perceived as more burdensome, and liked less compared to a fellow group member. We expect no differences between an equal-performing group member and the fellow group member.

METHOD Participants and Design A total of 56 individuals initially participated in this study, but 10 participants’ data were removed due to computer error. This left a final sample consisting of 46 participants (63% female) with a mean age of 20.65 years (SD = 3.36). The sample was 56.5% Caucasian, 13% African American, 13% Asian, and 13% indicating their race was different than any of the listed options. The

502


majority of participants (93.5%) indicated they primarily spoke English. The data and materials are available at osf.io/6fgnq. Participants were randomly assigned to a 2 (manipulated player performance: poorperformance vs. equal-performance) × 2 (group member: nontarget [Alex] vs. target [Pat]) mixed design2 . Target player performance was a between-participants factor and group member was a within-participants factor.


Procedure In groups of up to four people, participants came into the research lab and were instructed that they would play a word game, Atimia (see Figure 1). Participants were told they were playing with other students in another lab space through the campus network. This was not actually the case because participants were playing with two computer-controlled agents. They were then given instructions, including pictures, which provided information on (a) how to log into the game, (b) how the game display functioned, (c) the object of the game, (d) how to submit their answers, and (e) how to pick the next player. To log into the game, participants were instructed that they would enter in a desired screen name (that did not uniquely identify them) and then click the login button. Participants were told where in the Atimia game they would see the other players, the game information and statistics (i.e., group and participant accuracy), and the area for submitting their answer to the trials and selecting the next player. As part of these instructions,

FIGURE 1 A screenshot of Atimia. The participant is currently solving the task, as denoted by the opaque avatar and light-colored name and accuracy score. Pat is playing poorly in this example.



503

participants were told that a player’s avatar would turn dark blue when it was their turn to solve the problem. Following this, participants were given a clear explanation of the objective of the game. Participants were told, “The objective of Atimia is to enter the word that links or relates the three words in the word list.” Specifically, in Atimia, participants attempted to complete Remote Associates Test (RAT) items (Bowden & Jung-Beeman, 2003; Mednick, 1968). In this task, participants identify a word that links three separate words together, such as the solution word “girl” can be added to the words “flower,” “friend,” and “scout” (i.e., flower girl, girlfriend, girl scout). Participants went through several examples and we noted that the solution word could come before or after the three given words. Participants were instructed that during the game they could put their solution in the response box and hit submit to see if they were correct. After each response the game statistics updated (i.e., round number, participant accuracy, group accuracy, and question/question total). Last, participants learned that they should click on boxes marked with the names of the other players to select the next player to take a turn. During the game we varied the performance of one of the computer agent players to manipulate the presence of a burdensome group member. Participants played with two computer-controlled players named Alex and Pat. Pat was the manipulated player whose performance we manipulated to be either approximately equal to the group (equal-performing Pat), getting 75% of the answers correct (nonburdensome) or performing worse than the group (poor-performing Pat), getting approximately 13% correct (burdensome). Alex was the nonmanipulated player whose performance was designed to always be comparable to the participant’s performance. Pilot testing found participants were correct approximately 75% of the time. Therefore we set Alex, the nonmanipulated player, to perform similarly to how we expected participants to perform given the pilot data. Throughout the game, each computer-controlled player equally selected the participant or the other computer-controlled player in a random fashion. Participants played Atimia until the group answered 20 RAT items correctly and then participants went on to answer several questions about the other players and their game-playing experience. Dependent Measures For all of the self-report measures, participants answered questions about both of the computercontrolled players (Pat and Alex) separately. For each measure, participants answered all questions about one of the players before going on to the other player. Throughout the separate measures, we counterbalanced whether participants answered questions first about Pat or Alex and within each measure we randomized the order items were presented.

Ostracism of the target player. We evaluated how much participants ostracized Pat, the manipulated player, in three ways. First, as a behavioral measure of ostracism, we assessed how much participants ostracized Pat during the game, by analyzing the log files of Atimia, which recorded how often the participant selected each computer-controlled player. This method of recording behavioral ostracism is similar to the approach used with Cyberball (Wesselmann, Wirth et al., 2013; Wesselmann et al., 2014) where ostracism was operationalized as the number of the participant’s tosses to the target player. Similarly, we evaluated what percent of the time the participant selected Pat.

504



Second, participants self-reported how much they ignored and excluded (i.e., ostracized) each player during the Atimia game on a 1 (Not at all) to 7 (Very much so) scale (Spearman-Brown Coefficients ≥ .74; Eisinga, Te Grotenhuis, & Pelzer, 2013).3 As a third measure of ostracism, we asked participants how much they would like to work with Alex or Pat on a future task on a 1 (Not at all) to 7 (Very much so) scale (Wesselmann et al., 2014). This measure was intended to demonstrate the participant’s future ostracism intentions within the confines of the experiment interaction. Perception of group member burdensomeness. To assess how much participants found each player to be burdensome, we asked participants to make ratings on five items: “I felt like the presence of Alex/Pat threatened the group’s success,” “Alex/Pat was burdensome to the group,” “Alex/Pat kept the group from achieving its goals,” “Alex/Pat failed the group,” and “Alex/Pat contributed to the group,” (reverse scored). Participants completed their ratings on a 1 (Not at all) to 7 (Very much so) scale (αs ≥ .85).4 Liking of the players. We also asked participants how much they liked each of the players on a 1 (Not at all) to 7 (Very much so) scale. This item was intended to assess global feelings of dislike of a target player. Manipulation Check As a manipulation check, to see how well participants perceived each players’ performance, participants responded to the statement “Alex/Pat made mistakes,” on a 1 (Not at all) to 7 (Very much so) scale.

RESULTS With the exception of analyzing the behavioral ostracism, we conducted the analyses by using a 2 (Manipulated Player Performance: poor-performance vs. equal-performance) × 2 (group member: nontarget [Alex] vs. target [Pat]) mixed ANOVA with group members as the repeated factor and manipulated player performance as a between-participant factor. To evaluate our hypotheses directly, we used an independent-samples t test, to examine if poor-performing Pat (a burdensome group member) caused more aversive responses than equally performing Pat (a nonburdensome group member). We also made a series of comparisons (paired t tests) between Pat and Alex. To further demonstrate the averseness of a poor-performing Pat, a burdensome group member, we conducted a paired-samples t test comparing poor-preforming Pat to Alex (the nonmanipulated group member). We also conducted a paired-samples t test to demonstrate that equal-performing Pat, a nonburdensome group member, was evaluated similarly to Alex. Comparisons between Pat and Alex were made in their respective within-participants condition. Manipulation Check We evaluated the manipulation check using a mixed ANOVA. Overall, participants reported that Pat made more mistakes than Alex, F(1,44) = 5.01, p = .030, ηp 2 = .10. This main effect was


505

qualified by the expected significant interaction between the manipulated player performance and group member, F(1,44) = 26.62, p < .001, ηp 2 = .39. Specifically, participants indicated that Pat made more mistakes when performing poorly compared to performing equal with the group, t(44) = 5.79, p < .001, d = 1.71, 95% CI = [1.58, 3.27].5 Also by design, participants reported that poor-performing Pat made more mistakes than Alex (t(20) = −4.91, p < .001, d = 1.58, 95% CI = [−2.92, −1.18]). Although not our intention, participants also indicated that Alex made more mistakes compared to an equal-performing Pat (t(24) = 2.40, p = .024, d = 0.53, 95% CI = [0.12, 1.56]. There were no remaining significant differences between the players, ts ≤ −1.04, ps ≥ .31, ds ≤ 0.31.


Ostracism of the Target Player

Behavioral ostracism. We found, using an independent samples t-test, that participants were significantly less likely to select a poor-performing Pat (M = 27.18%, SD = 21.47) compared to an equally performing Pat (M = 45.66%, SD = 14.24), Levene’s Test of Equality of Variance, F = 6.86, p = .012; tadjusted (33.68) = −3.37, p = .002, d = 1.02, 95% CI = [−0.30, −0.07]. We used an independent samples t test, rather than a mixed ANOVA, because the number of times the nontarget was selected is a function of the number of times that the target player was selected, therefore making the results nonindependent of each other. Self-reported ostracism. Using a mixed ANOVA, we found in general that participants self-reported that they ostracized Pat more than Alex, F(1,44) = 22.10, p < .001, ηp 2 = .33. Central to our analysis, we found this main effect was again qualified by a hypothesized significant manipulated player performance × group member interaction for how much participants reported ostracizing the players, F(1,44) = 10.53, p = .002, ηp 2 = .19. Supporting our hypothesis, we found participants self-reported ostracizing poor-performing Pat more than equally performing Pat, tadjusted (37.30) = 2.53, p = .016, d = .77, 95% CI = [0.33, 3.00] (See Table 1 for descriptive statistics and Figure 2 for interaction graphs). Further demonstrating the ostracism of poor-performing Pat, participants also self-reported ostracizing poor-performing Pat more than Alex (the nonmanipulated player), (t[20] = −4.83, p < .001, d = 1.50, 95% CI = [−3.75, −1.49]). Lastly, there was no significant difference in self-reported ostracism of equal-performing Pat compared to Alex (t[24] = −1.21, p = .239, d = 0.32, 95% CI = [−1.30, −0.34]). Future ostracism intentions. Based on a mixed ANOVA, we found that in general participants had less of an intention to work on a future task with Pat compared to Alex, F(1,44) = 6.42, p = .015, ηp 2 = .13. This main effect was qualified by our hypothesized interaction between manipulated player performance and group member, F(1,44) = 32.34, p < .001, ηp 2 = .42. Breaking down this interaction, we found support for our hypothesis such that participants were less intent on working with poor-performing Pat than an equally performing Pat, t(44) = −2.43, p = .019, d = 0.72, 95% CI = [−2.63, −0.24]. Further, participants had less of an intent to work with poor-performing Pat than Alex (t[20] = 4.65, p < .001, d = 1.22, 95% CI = [1.21, 3.17]). We also found that participants were less interested in working with Alex compared to an equally performing Pat (t[24] = −2.93, p = .007, d = 0.46, 95% CI = [−1.43, −0.25]).

506 1.57 0.96 1.79 1.32 1.59

M

4.32

1.52 3.88 2.39 4.12

Variable

Manipulation check—Player made mistakes Self-reported ostracism Future ostracism intention Perceptions of burdensomeness Liking of the players

SD

Equalperforming Pat (n = 25)

Manipulated player’s performance

1.05 5.48 1.23 5.19

3.86

M

0.15 1.40 0.36 1.29

1.42

SD

Poorperforming Pat (n = 21)

Alex

1.30 4.61 1.86 4.61

4.11

M

Total

0.75 1.80 1.16 1.54

1.51

SD

2.00 4.72 2.08 4.64

3.48

M

1.90 1.88 1.68 1.82

1.61

SD

Equalperforming Pat (n = 25)

Group Member

TABLE 1 Means and Standard Deviations for the Dependent Variables


3.67 3.29 4.38 3.48

5.90

M

2.46 2.13 2.12 1.72

1.14

SD

Poorperforming Pat (n = 21)

Pat

2.76 4.07 3.13 4.11

4.59

M

Total

2.31 2.10 2.20 1.85

1.86

SD



507

FIGURE 2 Manipulated player performance × group member interactions for the four primary dependent variables. Manipulated player performance was a between-participants factor and group member was a within-participants factor. Higher values indicate increasing amounts.

Perception of Group Member Burdensomeness When we analyzed how much participants perceived each player as being burdensome, we found a pattern of results similar to the self-reported ostracism findings. Overall, participants rated Pat as more burdensome than Alex (F(1,44) = 24.29, p < .001, ηp 2 = .36), but similar to before this result was qualified by a hypothesized significant manipulated player performance × group member interaction (F(1,44) = 36.13, p < .001, ηp 2 = .45). When we further examine this interaction, we again find support for our hypothesis; participants reported that a poor-performing Pat was more burdensome than an equal-performing Pat, tadjusted (37.87) = 4.02, p < .001, d = 1.21, 95% CI = [1.14, 3.46]. Further, participants found poor-performing Pat as more burdensome than Alex (t[20] = −6.87, p < .001, d = 2.07, 95% CI = [−4.11, −2.20]). There was no significant difference between equal-performing Pat and Alex on how burdensome participants perceived each group member (t[24] = 0.86, p = .396, d = 0.21, 95% CI = [−0.43, 1.06]). Liking of the Players When we evaluated how much participants liked each of the players, we found overall that participants liked Pat less than Alex (F(1,44) = 4.41, p = .042, ηp 2 = .09). Central to supporting our hypothesis, we found the main effect of liking was qualified by a hypothesized significant manipulated player performance × group member interaction (F(1,44) = 15.42, p < .001, ηp 2 = .26).

508


Further specific analyses supported our hypothesis. Participants liked poor-performing Pat less than an equally performing Pat, (t[44] = −2.21, p = .032, d = 0.65, 95% CI = [−2.22, −0.10]). Participants also liked poor-performing Pat less than Alex (t[20] = 3.58, p = .002, d = 1.27, 95% CI = [0.72, 2.71]). Last, we found no difference in liking between equal-performing Pat and Alex (t[24] = −1.57, p = .131, d = 0.30, 95% CI = [−1.21, 0.17]). Exploratory Analyses


Perceptions of Alex. When Alex played with a poor- versus well-performing Pat, participants reported including Alex more, wanted to work with Alex more on a future task, perceived Alex as less burdensome, and liked Alex more (ps ≤ .023, ds ≥ 0.66). Interaction with participant gender. Contrary to initial investigations (Williams, Wheeler, & Harvey, 2001), we found no significant interaction between manipulated player performance and participant gender on behavioral ostracism, F(1,42) = 0.14, p = .713, ηp 2 < .01. Mediation analyses. We found no consistent mediation of the relationship between manipulated player’s performance and ostracism outcomes by the potential mediators of perceived burdensomeness and liking. This may be due to all measures being assessed after playing Atimia.

DISCUSSION To fully understand the experience of ostracism, researchers need to consider not only the impact of being ostracized, but also the experience of ostracizing others. To increase our understanding of sources of ostracism, we had three principle goals: (a) to replicate and extend previous research where participants ostracized a burdensome group member and to understand the perceptions of the burdensome group member, (b) to create a situation where participants autonomously ostracized a group member, and (c) to generate a multimethod approach to systematically investigate sources of ostracism. Our results suggest we achieved all three goals. Replicate and Extend Previous Research Our current results replicated previous research and extended research on sources by investigating ostracism of a burdensome player and how participants perceived a burdensome group member. Similar to previous research (i.e., Wesselmann, Wirth et al., 2013; Wesselmann et al., 2014), participants ostracized the burdensome group member by selecting that group member less than a nonburdensome group member. Similar to the current results, participants acknowledged ostracizing the burdensome group member through self-report. Further supporting previously demonstrated effects (Wesselmann et al., 2014), we found that the ostracism of the burdensome group member extended beyond the confines of the Atimia game; participants also had less of an intent on doing an immediate new task with the burdensome player. We also found that in addition to finding a poor-performing group member to be more burdensome (see also Wesselmann, Wirth et al., 2013; Wesselmann et al., 2014; Wirth et al., 2014), participants also liked this group member less than an group member who performed equally as well as the participant. We were


509

able to replicate and extend previous findings while inducing participants to ostracize without explicit instructions to do so.


Autonomously Ostracizing a Group Member Part of our intent for developing Atimia was to create a paradigm where participants are induced to ostracize, rather than being told explicitly to do so or have ostracism behavior modeled for them. This paradigm answers the call Zadro and Gonsalkorale (2014) put forth and it allows researchers to investigate what motivates ostracism of an individual and to investigate autonomous ostracism, an everyday experience of ostracizing others through one’s own volition. Up to this point, researchers could only observe sources of ostracism using Cyberball. In one case, participants had a negative interaction with an individual, before ostracizing this person during a subsequent game of Cyberball (Gooley et al., 2013) and in another case, participants ostracized a Cyberball player modified to take a long amount of time before deciding to throw the ball (Wesselmann, Wirth et al., 2013; Wesselmann et al., 2014). Similarly, our paradigm showed individuals will also ostracize an individual without being given any instructions about whom to choose in the game; participants autonomously ostracized a player performing poorly on the task. In fact, if we compare the effect sizes of behavioral ostracism (d = 1.02) and perceptions of burdensomeness (d = 1.21) as a result of playing with a poor performer in Atimia, we see that they are large effect sizes, just like the results of slow-player (16-second delay) Cyberball studies of Wesselmann, Wirth et al. (2013; d = 1.27, d = 2.18) and Wesselmann et al. (2014; d = 1.77, d = 1.55). This suggests that utilizing Atimia will help create a multimethod approach that is an ideal standard for understanding individual’s behavior. LIMITATIONS The conclusions we can draw from the current research are limited by examining only one type of group task, one motivation for ostracism, and an unexpected finding with the manipulation. Participants engaged in a conjunctive task, where group performance is dependent on the weakest member (Steiner, 1966). Tasks that rely less on the weakest link (i.e., disjunctive, complementary) may not replicate the current effects, but additive or compensatory tasks might. In addition to the group task, the motivation for ostracizing is important. Williams (1997) outlines several motivations for ostracism: punitive, not ostracism, role prescribed, defensive, or oblivious. We may have most closely examined punitive ostracism, “acts of ignoring that are perceived to be or intended to be deliberate and aversive” (Williams, 1997, p. 145). Individuals may ostracize for a host of reasons as outlined by Williams or for additional reasons such as revenge or to establish power within a group. Unexpectedly, participants indicate differences in the number of mistakes between Pat and Alex (the manipulation) in the control condition where no differences should have emerged. This led to differences in future ostracism intentions, but no differences in behavioral ostracism (t[24] = 1.52, p = .140, 95% CI = [−0.03, 0.20]), self-reported ostracism, or perceptions of burden. This outcome suggests small amounts of burden could be tolerated within the immediate interaction (Wesselmann, Wirth, Pryor, Reeder, & Williams, 2015). The current research has these limitations that could be addressed in future research.

510


FUTURE DIRECTIONS: ATIMIA AS A METHOD FOR SYSTEMATICALLY INVESTIGATING SOURCES OF OSTRACISM The Atimia paradigm may generate numerous research opportunities to systematically investigate sources of ostracism. Researchers could apply Atimia to investigate the experience of ostracizing another individual, individual differences influencing when one becomes a source of ostracism, and finally, situations that may influence when an individual becomes a source of ostracism.


The Experience of Ostracizing Another Individual We believe Atimia can be used to continue the investigation of outcomes associated with ostracizing others. For instance, Atimia could be used to examine whether there is variation in the extent to which individuals are aware that they are harming the target of the ostracism. Wesselmann, Williams et al. (2013) found that individuals feel the same effects of being ostracized when they are merely watching an individual be ostracized, especially if they take the perspective of the ostracized individual. This suggests sources of ostracism are aware that they are harming the target of their ostracism and the degree to which participants are empathetic may affect an individual acting as a source of ostracism. Researchers might also examine how the perceptions of the group change from the source’s perspective. Gruter and Masters (1986) contend that the group becomes closer when it ostracizes a group member, in part because of the group having to coordinate their efforts to remove the group member. Along similar lines of future work, researchers could examine how sources feel when they ostracize alone or with other group members. Based on social impact theory (Latané, 1981), sharing the role of ostracizing with others may diminish the aversive impact of harming the target. Atimia could be implemented to investigate these questions and how individual differences could impact being a source of ostracism. Individual Differences Individual characteristics may influence when one becomes a source of ostracism. As with many social interactions, the five factor model (Costa & McCrae, 2009) may also play an important role. In the case of ostracizing others, highly agreeable individuals (i.e., those who want to minimize conflict; Jensen-Campbell & Graziano, 2001), may be less likely to create conflict by ostracizing an individual. Similarly, conscientious individuals may be especially aware of the hurt they are causing the ostracized individual and therefore be hesitant to do so. Conversely, personality traits consistent with Machiavellianism, part of the Dark Triad (which also includes narcissism; Paulhus & Williams, 2002), may lead to increased use of ostracism to manipulate an individual, similar to how groups enforce norms through ostracism (e.g., Dijker & Koomen, 2007). Additionally, individuals high in psychopathy may also ostracize others. These individuals are characterized by being manipulative, egocentric, impulsive, deceitful, irresponsible, arrogant, and also having a superficial charm and shallow affect (Cleckley, 1941; Hare, 1991). Because individuals high in psychopathy are impulsive and have low empathy and anxiety, they may be quick to ostracize individuals due to lack of concern for harming the target. Last, while exploratory analyses did not find that gender affected the source’s ostracism ball-toss behavior, a study designed specifically to investigate this question may be warranted. Researchers could use Atimia


511

to present scenarios where there is motivation to ostracize a group member and investigate if these individuals differences play a moderating role.


Situational Factors Finally, we believe Atimia could be used to investigate situational factors that could increase or decrease the likelihood of sources ostracizing a group member. A basic situational factor to investigate is whether male or female burdensome targets are more likely to be ostracized. Similarly, researchers could investigate if a target from a minority social group (ethnic/racial) is more likely to be ostracized than a target from a majority social group. Individuals from a minority social group generally have a greater likelihood of social exclusion than majority group members (e.g., Kurzban & Leary, 2001). A second situational factor could involve manipulations designed to increase burden tolerance, therefore causing individuals to be more likely to include the burdensome group member. One manipulation might be increasing one’s compassion for others through cognitive-based compassion training (CBCT; Mascaro, Rilling, Tenzin Negi, & Raison, 2013; Ramsey & Jones, 2015). Researchers could investigate the differences for sources when they ostracize others without prompting compared to being asked to ostracize a particular group member (e.g., Legate et al., 2013). Based on the research by Sommer and Yoon (2013), sources may feel less distress ostracizing others when the target deserves the treatment compared to when it is undeserved. Features of Atimia (see Wirth et al., 2014) may be used effectively to look at any one of these situational factors and potential others.

CONCLUSION Ostracism researchers are now taking a full perspective in their investigation of an ostracism episode by not only looking at how it feels to be ostracized, but also what it feels like to be the one doing the ostracizing. Researchers demonstrated that individuals who are ostracizing others are not free from feeling harmed, but rather can also feel similarly negative as the individual being ostracized. Given that ostracism occurs on nearly a daily basis (Nezlek, Wesselmann, Wheeler, & Williams, 2012) and both the individual being ostracized and the individual doing the ostracism feel negative, it is imperative to understand why individuals ostracize in an effort to reduce how badly both parties involved in ostracism feel. By creating Atimia, a paradigm that can be easily implemented in a systematic way, we hope to increase the amount of research being conducted investigating sources of ostracism.

512


ACKNOWLEDGMENTS Researchers can request the Atimia program, without any cost, or ask any questions by contacting James H. Wirth at [email protected]. We would like to thank Paul Turchan for his input on this research.


NOTES 1. The name for the game, Atimia, is derived from the ancient Greek practice of atimia, which was the penalization of the atimos, individuals who were in debt to the city of Athens (Allen, 2000). 2. The original experiment design included an additional between-participants manipulation where we manipulated whether players could freely choose which player would go next or the computer assigned what player would go next in a round-robin format. There were no significant interactions with this factor, so we focused on the performance manipulation and omitted participants in the round-robin conditions. 3. We calculated separate Spearman-Brown coefficients for the target (Pat) and nontarget group players (Alex). 4. For each of the scales, we created a separate Cronbach’s alpha for questions about Pat and Alex. 5. Confidence intervals are calculated based on mean differences between conditions.

AUTHOR NOTES James H. Wirth s an Assistant Professor in the Department of Psychology at The Ohio State University at Newark. Michael J. Bernstein is an Associate Professor in the Psychological and Social Science Program at Penn State University–Abington. Angie S. LeRoy is a PhD student in the Department of Psychology at the University of Houston.

REFERENCES Allen, D. S. (2000). The world of Prometheus: The politics of punishing in democratic Athens. Princeton, NJ: Princeton University Press. Bastian, B., Jetten, J., Chen, H., Radke, H. R., Harding, J. F., & Fasoli, F. (2013). Losing our humanity: The self-dehumanizing consequences of social ostracism. Personality and Social Psychology Bulletin, 39, 156–169. doi:10.1177/0146167212471205 Baumeister, R. F., Wotman, S. R., & Stillwell, A. M. (1993). Unrequited love: On heartbreak, anger, guilt, scriptlessness, and humiliation. Journal of Personality and Social Psychology, 64, 377–394. doi:10.1037/0022-3514.64.3.377 Bowden, E. M., & Jung-Beeman, M. (2003). Normative data for 144 compound remote associate problems. Behavior Research Methods, Instruments, & Computers, 35, 634–639. doi:10.3758/BF03195543 Chen, Z., Poon, K.-T., Bernstein, M. J., & Teng, F. (2014). Rejecting another pains the self: The impact of perceived future rejection. Journal of Experimental Social Psychology, 50, 225–233. doi:10.1016/j.jesp.2013.10.007 Ciarocco, N. J., Sommer, K. L., & Baumeister, R. F. (2001). Ostracism and ego depletion: The strains of silence. Personality and Social Psychology Bulletin, 27, 1156–1163. doi:10.1177/0146167201279008 Cleckley, H. (1941). The mask of sanity. St. Louis, MO: C.V. Mosby. Costa, P. T., Jr., & McCrae, R. R. (2009). The five-factor model and the NEO inventories. In J. N. Butcher (Ed.), Oxford handbook of personality assessment (pp. 299–322). New York, NY: Oxford University Press. Dijker, A. J. M., & Koomen, W. (2007). Stigmatization, tolerance and repair: An integrative psychological analysis of responses to deviance. New York, NY: Cambridge University Press. Eisenberger, N. I., & Lieberman, M. D. (2004). Why rejection hurts: A common neural alarm system for physical and social pain. Trends in Cognitive Sciences, 8, 294–300. doi:10.1016/j.tics.2004.05.010 Eisenberger, N. I., Lieberman, M. D., & Williams, K. D. (2003). Does rejection hurt? An fMRI study of social exclusion. Science, 302, 290–292. doi:10.1126/science.1089134



513

Eisinga, R., Te Grotenhuis, M., & Pelzer, B. (2013). The reliability of a two-item scale: Pearson, Cronbach, or SpearmanBrown? International Journal of Public Health, 58, 637–642. doi:10.1007/s00038-012-0416-3 Gooley, S., Zadro, L., Gonsalkorale, K., Williams, L., & Sethi, N. (2013). Motivating participants to ostracize: A new sources paradigm. Unpublished manuscript, School of Psychology, University of Sydney, Australia. Gooley, S. L., Zadro, L., Williams, L. A., Svetieva, E., & Gonsalkorale, K. (2015). Ostracizing for a reason: A novel source paradigm for examining the nature and consequences of motivated ostracism. Journal of Social Psychology, 155(5), 410–431. Gruter, M., & Masters, R. D. (Eds.). (1986). Ostracism as a social and biological phenomenon: An introduction. Ethology and Sociobiology, 7, 149–158. doi:10.1016/0162-3095(86)90043-9 Hare, R. D. (1991). The hare psychopathy checklist-revised. Toronto, Ontario: Multi-Health Systems. Jensen-Campbell, L. A., & Graziano, W. G. (2001). Agreeableness as a moderator of interpersonal conflict. Journal of Personality, 69(2), 323–362. doi:10.1111/1467-6494.00148 Kurzban, R., & Leary, M. R. (2001). Evolutionary origins of stigmatization: The functions of social exclusion. Psychological Bulletin, 127, 187–208. doi:10.1037/0033-2909.127.2.187 Latané, B. (1981). The psychology of social impact. American Psychologist, 36, 343–356. doi:10.1037/0003-066X. 36.4.343 Legate, N., DeHaan, C. R., Weinstein, N., & Ryan, R. M. (2013). Hurting you hurts me too: The psychological costs of complying with ostracism. Psychological Science, 24, 583–588. doi:10.1177/0956797612457951 Mascaro, J. S., Rilling, J. K., Tenzin Negi, L., & Raison, C. L. (2013). Compassion meditation enhances empathic accuracy and related neural activity. Social Cognitive & Affective Neuroscience, 8, 48–55. doi:10.1093/scan/nss095 Mednick, S. A. (1968). The remote associates test. The Journal of Creative Behavior, 2, 213–214. doi:10.1002/j.2162-6057.1968.tb00104.x Milgram, S. (1963). Behavioral study of obedience. Journal of Abnormal & Social Psychology, 70, 207–217. doi:10.1037/h0040525 Milgram, S. (1974). Obedience to authority. New York, NY: Harper & Row. Nezlek, J. B., Wesselmann, E. D., Wheeler, L., & Williams, K. D. (2012). Ostracism in everyday life. Group Dynamics: Theory, Research, and Practice, 16, 91–104. doi:10.1037/a0028029 Paulhus, D. L., & Williams, K. M. (2002). The dark triad of personality: Narcissism, Machiavellianism, and psychopathy. Journal of Research in Personality, 36, 556–563. doi:10.1016/S0092-6566(02)00505-6 Poulsen, J. R., & Kashy, D. A. (2012). Two sides of the ostracism coin: How sources and targets of social exclusion perceive themselves and one another. Group Processes & Intergroup Relations, 15, 457–470. doi:10.1177/1368430211430517 Ramsey, A. T., & Jones, E. E. (2015). Minding the interpersonal gap: Mindfulness-based interventions in the prevention of ostracism. Consciousness and Cognition, 31, 24–34. doi:10.1016/j.concog.2014.10.003 Schwarz, N., Groves, R. M., & Schuman, H. (1998). Survey methods. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), Handbook of social psychology (pp. 143–179). New York, NY: Oxford University Press. Sommer, K. L., Williams, K. D., Ciarocco, N. J., & Baumeister, R. F. (2001). When silence speaks louder than words: Explorations into the intrapsychic and interpersonal consequences of social ostracism. Basic and Applied Social Psychology, 23, 225–243. doi:10.1207/S15324834BASP2304_1 Sommer, K. L., & Yoon, J. (2013). When silence is golden: Ostracism as resource conservation during aversive interactions. Journal of Social and Personal Relationships, 30, 901–919. doi:10.1177/0265407512473006 Steiner, I. (1966). Models for inferring relationships between group size and potential group productivity. Behavioral Science, 11, 273–283. doi:10.1002/bs.3830110404 Stroud, L. R., Tanofsky-Kraff, M., Wilfley, D. E., & Salovey, P. (2000). The Yale Interpersonal Stressor (YIPS): Affective, physiological, and behavioral responses to a novel interpersonal rejection paradigm. Annals of Behavioral Medicine, 22, 204–213. doi:10.1007/BF02895115 Wesselmann, E. D., Williams, K. D., & Hales, A. H. (2013). Vicarious ostracism. Frontiers in Human Neuroscience, 7, 153. doi:10.3389/fnhum.2013.00153 Wesselmann, E. D., Williams, K. D., & Wirth, J. H. (2014). Ostracizing group member who can (or cannot) control being burdensome. Human Ethology Bulletin, Proceedings of the IV ISHE Summer Institute, 82–103. Wesselmann, E. D., Wirth, J. H., Mroczek, D. K., & Williams, K. D. (2012). Dial a feeling: Detecting moderation of affect decline during ostracism. Personality and Individual Differences, 53, 580–586. doi:10.1016/j.paid.2012.04.039


514


Wesselmann, E. D., Wirth, J. H., Pryor, J. B., Reeder, G. D., & Williams, K. D. (2013). When do we ostracize? Social Psychological and Personality Science, 4, 108–115. doi:10.1177/1948550612443386 Wesselmann, E. D., Wirth, J. H., Pryor, J. B., Reeder, G. D., & Williams, K. D. (2015). The role of burden and deviation in ostracizing others. Journal of Social Psychology, 155(5), 434–496. Williams, K. D. (1997). Social ostracism. In R. Kowalski (Ed.), Aversive interpersonal behaviors (pp. 133–170). New York, NY: Plenum. Williams, K. D. (2007). Ostracism. Annual Review of Psychology, 58, 425–452. doi:10.1146/annurev.psych.58.110405. 085641 Williams, K. D., Bernieri, F., Faulkner, S., Gada-Jain, N., & Grahe, J. (2000). The scarlet letter study: Five days of social ostracism. Journal of Personal and Interpersonal Loss, 5, 19–63. doi:10.1080/10811440008407846 Williams, K. D., Cheung, C. K. T., & Choi, W. (2000). Cyberostracism: Effects of being ignored over the Internet. Journal of Personality and Social Psychology, 79, 748–762. doi:10.1037/0022-3514.79.5.748 Williams, K. D., & Jarvis, B. (2006). Cyberball: A program for use in research on interpersonal ostracism and acceptance. Behavior Research Methods, 38, 174–180. doi:10.3758/BF03192765 Williams, K. D., & Nida, S. A. (2011). Ostracism: Consequences and coping. Current Directions in Psychological Science, 20, 71–75. doi:10.1177/0963721411402480 Williams, K. D., & Sommer, K. L. (1997). Social ostracism by coworkers: Does rejection lead to loafing or compensation? Personality and Social Psychology Bulletin, 23, 693–706. doi:10.1177/0146167297237003 Williams, K. D., Wheeler, L., & Harvey, J. (2001). Inside the social mind of the ostracizer. In J. P. Forgas, K. D. Williams, & L. Wheeler (Eds.), The social mind: Cognitive and motivational aspects of interpersonal behavior (pp. 294–320). New York, NY: Cambridge University Press. Wirth, J. H., Turchan, P. J., Zimmerman, A. G., & Bernstein, M. J. (2014). Atimia: A novel group-based paradigm for manipulating ostracism and group members’ performance. Group Dynamics: Theory, Research, and Practice, 18, 251–266. doi:10.1037/gdn0000006 Wirth, J. H., & Williams, K. D. (2009). “They Don’t Like Our Kind”: Consequences of being ostracized while possessing a group membership. Group Processes & Intergroup Relations, 12, 111–127. doi:10.1177/1368430208098780 Zadro, L., & Gonsalkorale, K. (2014). Sources of ostracism: The nature and consequences of excluding and ignoring others. Current Directions in Psychological Science, 23, 93–97. doi:10.1177/0963721413520321 Zadro, L., Williams, K. D., & Richardson, R. (2004). How low can you go? Ostracism by a computer is sufficient to lower self-reported levels of belonging, control, self-esteem, and meaningful existence. Journal of Experimental Social Psychology, 40, 560–567. doi:10.1016/j.jesp.2003.11.006 Zadro, L., Williams, K. D., & Richardson, R. (2005). Riding the “O” train: Comparing the effects of ostracism and verbal dispute on targets and sources. Group Processes & Intergroup Relations, 8, 125–143. doi:10.1177/1368430205051062 Zhou, X., Zheng, L., Zhou, L., & Guo, N. (2009). The act of rejecting reduces the desire to reconnect: Evidence for a cognitive dissonance account. Journal of Experimental Psychology, 45, 44–50. doi:10.1016/j.jesp.2008.08.011

Received December 1, 2014 Accepted May 26, 2015



How Does the Source of Rejection Perceive Innocent Victims? a

b

Kai-Tak Poon & Zhansheng Chen a

Hong Kong Institute of Education

b

University of Hong Kong Published online: 12 Aug 2015.

Click for updates To cite this article: Kai-Tak Poon & Zhansheng Chen (2015) How Does the Source of Rejection Perceive Innocent Victims?, The Journal of Social Psychology, 155:5, 515-526, DOI: 10.1080/00224545.2015.1061972 To link to this article: http://dx.doi.org/10.1080/00224545.2015.1061972





How Does the Source of Rejection Perceive Innocent Victims? KAI-TAK POON Downloaded by [Ohio State University Libraries] at 08:55 20 August 2015

Hong Kong Institute of Education

ZHANSHENG CHEN University of Hong Kong

ABSTRACT. Prior studies have consistently revealed a strong positive association between general just-world beliefs and victim blaming. The present research aims to extend the literature by testing whether an act of social rejection overrides the influence of general just-world beliefs on victim blaming. Building upon the theory of moral compensation that people are more prosocial after behaving undesirably, we predicted that people should be less likely to blame an innocent victim after rejecting another person and that general just-world beliefs were not associated with victim blaming among the sources of rejection. To test these predictions, participants first completed a measure of general just-world beliefs and then recalled a past incident in which they rejected or accepted another person. They then read a scenario about school bullying and made attributions for the victim’s suffering. The results supported our predictions and the implications of the study are discussed. Keywords: just-world beliefs, rejection, social exclusion, sources of rejection, victim blaming

THE NEED TO GAIN SOCIAL ACCEPTANCE and avoid social rejection is one of the most fundamental human motivations (Baumeister & Leary, 1995). Without a doubt, social rejection1 can bring various negative consequences to the targets of rejection (i.e., people who are socially rejected). For example, people experience intense social distress when they are socially rejected (e.g., Wesselmann, Cardoso, Slater, & Williams, 2012; Williams, Cheung, & Choi, 2000; see also Williams, 2007, 2009 for reviews), and thus people are motivated to restore these threatened feelings of belonging following social rejection through various channels (e.g., Epley, Akalis, Waytz, & Cacioppo, 2008; Maner, DeWall, Baumeister, & Schaller, 2007; Poon, Teng, Chow, & Chen, 2015). In addition, recent brain imagining studies have shown that social rejection can activate brain areas that are associated with the detection and experience of physical pain (e.g., Eisenberger, Lieberman, & Williams, 2003; Kross, Berman, Mischel, Smith, & Wager, 2011), further highlighting the fact that social rejection is a detrimental and painful experience for targets. Worse still, a considerable number of laboratory experiments have shown that socially rejected individuals have a higher tendency to engage in different forms of antisocial behaviors, Address correspondence to Kai-Tak Poon, Hong Kong Institute of Education, Department of Psychological Studies and Centre for Psychosocial Health, 10 Lo Ping Road, Tai Po HK, Hong Kong. E-mail: [email protected]


516


such as aggression and dishonesty (e.g., DeWall, Twenge, Gitter, & Baumeister, 2009; Kouchaki & Wareham, 2015; Poon, Chen, & DeWall, 2013; Twenge, Baumeister, Tice, & Stucke, 2001; Wesselmann, Butler, Williams, & Pickett, 2010). The literature has also documented that social rejection can bring various negative outcomes to the sources of rejection (i.e., people who reject another person). In particular, prior studies have consistently shown that people experience negative emotions, distress, and painful feelings after rejecting another person (e.g., Baumeister, Wotman, & Stillwell, 1993; Chen, Poon, Bernstein, & Teng, 2014; Legate, DeHaan, Weinstein, & Ryan, 2013; Poulsen & Kashy, 2012). Moreover, rejecting another person is difficult, which depletes people’s physical stamina and ego resources (e.g., Ciarocco, Sommer, & Baumeister, 2001; Williams, Bernieri, Faulkner, Grahe, & Gada-Jain, 2000). Although these studies advanced our understanding about the immediate negative impact social rejection has on the source of rejection, research has ignored how sources of rejection might view events happening in the social world differently, specifically, perceptions of victims in other social situations. The present research aims to fill this knowledge gap by examining whether the sources of rejection are less susceptible to the influence of general just-world beliefs because people are less likely to blame innocent victims after they have rejected another person. People have a need to believe that they live in a just world (Lerner, 1980; Lerner & Simmons, 1966). When just-world believers observe other people suffering without sufficient justifications, they tend to blame these innocent victims by making internal attributions for their suffering and misfortune to restore their feelings of justice (see Dalbert, 2009; Hafer & Bègue, 2005 for reviews). We argue that an act of social rejection should threaten moral images of the sources of rejection because rejectors understand that their action may inflict severe social pain on others. Building upon the theory of moral compensation that people are more prosocial after transgressions (Zhong, Liljenquist, & Cain, 2009), people should be motivated to engage in moral compensatory behaviors after rejecting another person. As a consequence, people who have rejected another person should be less likely to blame innocent victims than people who have accepted another person to restore their moral self-image. Moreover, the act of rejecting another person should override the influence of general just-world beliefs, such that these beliefs should only be positively associated with victim blaming among the source of social acceptance. In contrast, general just-world beliefs should not have reliable influence on victim blaming among the source of social rejection. JUST-WORLD BELIEFS AND VICTIM BLAMING General just-world beliefs are defined as one’s beliefs that the world is generally just and fair (Dalbert, 1999, 2009). Such a belief in a just world is conceptualized as a stable personality trait that has considerable individual differences (Furnham, 2003), and it often influences how people perceive and construct the social environment. For example, prior research has shown that strong just-world believers experience less anger upon provocations (e.g., Nesbit, Blankenship, & Murray, 2012). In addition, people with strong just-world beliefs tend to experience less distress in challenging (Tomaka & Blascovich, 1994) and painful situations (McParland & Knussen, 2010). When just-world beliefs are threatened by potential injustice situations, people are generally motivated to defend the beliefs. Observing an individual suffering without sufficient justifications is a potential injustice situation, which may increase one’s feelings of distress. People are


POON AND CHEN

517

motivated to defend the just-world beliefs and reduce the distress by cognitively reevaluating the situation. Such a motivation often makes people blame innocent victims and believe that victims are responsible for their misfortune and suffering (see Dalbert, 2009; Hafer & Bègue, 2005). For example, an early study found that people would devalue and reject a female peer who received electric shocks for mistakes in a learning task when they could not stop her suffering (Lerner & Simmons, 1966). By making internal attributions to victims’ suffering, people’s just-world beliefs can be defended. The threat triggered by injustices should be more prominent among people with strong justworld beliefs. According to the cognitive dissonance theory (Festinger, 1957), these people may be particularly motivated to cope with the threat by cognitively reevaluating potential injustice situations. However, people with weak just-world beliefs may not be threatened by injustice, and hence they may not be motivated to defend the beliefs. Indeed, the literature has accumulated substantial amounts of evidence showing a robust positive relationship between general justworld beliefs and victim derogation in various situations. For example, when strong just-world believers witness other people encounter various forms of suffering (e.g., sexually transmitted diseases, assault, and illnesses) with no justified reasons, they tend to believe that the victims deserve their suffering (e.g., Correia, Vala, & Aguiar, 2001; Hafer, 2000; see also Dalbert, 2009; Hafer & Bègue, 2005). Compared with their counterparts with weak just-world beliefs, people with strong just-world beliefs also make stronger internal attributions and weaker external attributions for their own suffering and negative experiences (e.g., Hafer & Correy, 1999; Poon & Chen, 2014). Although prior studies have shown that people with high just-world beliefs tend to blame innocent victims, it is still unclear whether this relationship can also be observed among people who have rejected another person. In the next section, we will explain why we predict that rejecting another person should moderate the relationship between just-world beliefs and victim blaming.

REJECTING ANOTHER PERSON AND VICTIM BLAMING People’s current behavioral decisions are often influenced by their past behaviors. People often feel psychologically licensed to engage in ethically questionable deeds without discrediting themselves after they have behaved morally or desirably. Prior studies have provided substantial amounts of evidence showing the phenomenon of moral licensing in various social domains (Miller & Effron, 2010; Zhong et al., 2009). In particular, people are more likely to express potentially prejudiced attitudes when their past behavior has provided them credentials (Monin & Miller, 2001). Moreover, people are more likely to behave selfishly and unethically after being primed with positive images about themselves through writing themselves with positive traits (Sachdeva, Iliev, & Medin, 2009) or buying environmental friendly products (Mazar & Zhong, 2010). In contrast, people generally engage in compensatory behaviors when their moral self-image is threatened by their previous ethically questionable or undesirable behaviors (Zhong et al., 2009). Such compensations may help transgressors experience less negative emotions and threats associated with their transgressions. For example, people who have previously inflicted physical pain on another person are more likely to comply with other people’s requests for help (Carlsmith


518


& Gross, 1969). Overindulgent consumers tend to make more subsequent utilitarian consumption decisions (Ramanathan & Williams, 2007). Moreover, Tetlock, Kristel, Elson, Green, and Lerner (2000) found that people engaged in moral cleansing that could reaffirm their values when they violate their own values. To reduce their feelings of moral violations, when people recall a past transgression, they are more likely to hurt themselves (Bastian, Jetten, & Fasoli, 2011) and engage in physical cleansing (Zhong & Liljenquist, 2006). In sum, these findings suggest that transgressors tend to engage in various actual and symbolic compensatory behaviors to repair their moral self-image. People have a fundamental need for social acceptance (Baumeister & Leary, 1995). Social rejection is an unpleasant and painful experience because it threatens one’s fundamental need to belong (e.g., Wesselmann et al., 2012; Williams, Cheung et al., 2000). Moreover, people are aware that the targets of rejection would experience intense negative emotion and merely observing another person being rejected can threaten their feelings of social satisfaction (Wesselmann, Bagg, & Williams, 2009; Wesselmann, Williams, & Hales, 2013). As a consequence, an act of rejecting another person may potentially threaten one’s moral self-image because the source of rejection understands that that their action may inflict severe social pain on the target of rejection and that it may violate the norm of reciprocal social exchange. Indeed, prior studies have shown that people experience intense distress after rejecting another person (e.g., Baumeister et al., 1993; Chen et al., 2014; Legate et al., 2013; Poulsen & Kashy, 2012). For example, although rejectors depicted themselves as morally innocent, they still felt guilty about hurting another person (Baumeister et al., 1993). Moreover, a recent investigation showed that people’s social needs are threatened when they comply with a request to reject another person (Legate et al., 2013). Worse still, people would experience distress after recalling a past incident because they were concerned that their behavior could lead to future rejection (Chen et al., 2014). As a result, we predict that the source of rejection should be more motivated to engage in compensatory behaviors to restore their thwarted moral image. Specifically, we predict that the source of rejection should be less likely to blame an innocent victim than the source of acceptance because such blaming may further threaten one’s moral self-image. As noted in the previous section, people with high just-world beliefs tend to blame innocent victims (see Dalbert, 2009; Furnham, 2003; Hafer & Bègue, 2005 for reviews). However, an act of social rejection may make people less likely to blame the victim because they are motivated to compensate for their past undesirable deeds. Such a motivation should override the influence of general just-world beliefs on victim blaming. Taken together, we predict that an act of social rejection should moderate the relationship between general just-world beliefs and victim blaming. Specifically, general just-world beliefs should be positively correlated with victim blaming among people who have accepted another person. However, such a positive relationship should not be observed among people who have rejected another person.

METHOD Participants and Design Ninety-three individuals (28 men; mean age = 36.84; SD = 13.11) in the United States completed this research study online via the Amazon’s Mechanical Tuck, which is a platform that aids

POON AND CHEN

519

researchers in collecting reliable and representative data (see Buhrmester, Kwang, & Gosling, 2011). We set our worker approval criteria at 80% and provided keywords (e.g., personality, survey, questionnaire) to allow workers to search for our HIT. Participants received a payment of US$0.3 in exchange for their participation. Participants were randomly assigned to one of the two experimental conditions (accepting condition vs. rejecting condition).


Procedures and Measures Participants were told that the study consisted of several unrelated parts. After providing their consent, they first completed a scale to measure their general just-world beliefs (e.g., “I think basically the world is a just place” and “I am confident that justice always prevails over injustice”; Dalbert, 1999). Participants indicated the extent to which they agree with each of the six statements on a six-point scale (1 = strongly disagree; 6 = strongly agree). The scores were averaged to index participants’ general beliefs in a just world (α = .79), in which higher scores indicated higher endorsement of general just-world beliefs. Next, participants were exposed to our experimental manipulation, which was adopted from previous studies (e.g. Chen et al., 2014; Williams, Shore, & Grahe, 1998). Specifically, participants in the rejecting condition recalled and wrote down a past incident in which they rejected another person2 ; whereas participants in the accepting condition recalled and wrote down a past incident in which they accepted another person. Participants were instructed to describe the incident with as many details as possible, and they could use as much time as they needed. Participants then indicated the extent to which they agreed with two statements, “I rejected another person” and “someone was rejected by me” (1 = not at all; 7 = extremely). The scores were averaged to check the effectiveness of the manipulation (Spearman-Brown Coefficient = .86). We reported Spearman-Brown Coefficient because recent research has suggested that it provides a more appropriate reliability assessment for a two-item measure (compared to Cronbach’s alpha or Pearson’s r; Eisinga, Te Grotenhuis, & Pelzer, 2013). Next, participants were asked to read a short scenario about school bullying, which was modified from previous research (Nordgren, Banas, & MacDonald, 2011). Specifically, participants read that Anna was one of the least popular girls in school, who had little contact with her classmates. Worst still, she was often bullied by other classmates, such as being teased and physically attacked. Participants then indicated the extent to which they agreed with six statements, which assessed their attributions of other’s suffering. The items were “Her situation was as a result of the type of person she is,” “Her character/personality was responsible for her situation,” “Her character/personality was to blame for her situation,” “Her situation was as a result of her behavior,” “Her behavior was responsible for her situation,” and “Her behavior was to blame for her situation” (α = .96). Similar items have been commonly adopted in prior studies to capture one’s tendency to blame innocent victims (e.g., Correia et al., 2001; Hafer, 2000; 2002; see also Hafer & Bègue, 2005; for a review). The scores were averaged to form a composite score of victim blaming, in which higher scores indicated a higher tendency to make internal attributions for the victim’s suffering. Participants completed a few simple unrelated filler tasks to disguise the study’s purpose. Finally, participants received a debriefing. No participants correctly guessed the research hypotheses.

520


RESULTS Manipulation Check


Participants in the rejecting condition (M = 4.99; SD = 1.98) agreed more with the statements that they had rejected another person than participants in the accepting condition did (M = 2.44; SD = 1.89), t(91) = 6.33, p < .001, Cohen’s d = 1.32, 95% C.I. = 1.75 to 3.35. Therefore, our experimental manipulation was effective in inducing the feelings of accepting versus rejecting another person. Victim Blaming Participants in the rejecting condition (M = 2.49; SD = 1.36) made less internal attributions for another person’s suffering than participants in the accepting condition did (M = 3.09; SD = 1.54), t(91) = −2.02, p = .047,Cohen’s d = .35, 95% C.I. = −1.21 to −0.01. Therefore, participants who recalled a past incident of rejecting another person were less likely to blame the innocent victim than participants who recalled a past incident of accepting another person were3 . A hierarchical regression analysis was conducted (Aiken & West, 1991) to examine whether general just-world beliefs interacted with the experimental condition in predicting one’s tendency to blame the victim. The experimental condition was coded as 1 (rejecting condition) and −1 (accepting condition), and general just-world-belief scores were centered. The results revealed a significant main effect of the experimental condition, in which participants in the rejecting condition made less internal attributions for another person’s suffering than participants in the accepting condition did (b = −.32, se = .14, t(89) = −2.24, p = .028, 95% C.I. = −.61 to −.04). The main effect of general just-world beliefs was also statistically significant, such that participants with high general just-world beliefs tended to make more internal attributions for another person’s suffering (b = .41, se = .15, t(89) = 2.67, p = .009, 95% C.I. = .11 to .71). More important, the expected moderation effect emerged (b = −.32, se = .15, t(89) = −2.09, p = .040, 95% C.I. = −.62 to −.02) (see Figure 1). The R-square of the regression model was .15. Further analyses showed that among participants who recalled a past accepting experience, general just-world beliefs were positively associated with internal attributions for another person’s sufferings (b = .73, se = .22, t(89) = 3.37, p = .001, 95% C.I. = .30 to 1.16). In contrast, among participants who recalled a past rejecting experience, general just-world beliefs were statistically not associated with internal attributions (b = .09, se = .22, t(89) = .41, p = .680, 95% C.I. = −.34 to .52).

DISCUSSION Recent research demonstrates that rejecting another person is an unpleasant and painful experience for both the target (e.g., Williams, 2009) and the source (e.g., Chen et al., 2014; Legate et al., 2013). Our research builds on this by demonstrating that rejection also impacts the source’s perceptions of innocent victims of aversive events. Specifically, general just-world beliefs predicted participants’ tendency to blame an innocent victim only when they recalled a time in which


POON AND CHEN

521

FIGURE 1 Victim blaming as a function of rejecting condition and general just-world beliefs.

they accepted another person. This finding replicates prior findings that people with high general just-world beliefs tend to blame innocent victims by making internal attributions for their sufferings (e.g., Correia et al., 2001; Hafer, 2000; Lerner, 1980). When participants recalled a time in which they rejected another person, however, general just-world beliefs no longer predicted victim blaming. We argue that because individuals recognize that rejection is harmful to targets (Wesselmann, Williams et al., 2013), this likely makes the source of rejection feel uncomfortable because they think they have harmed another person. These findings are consistent with the moral compensation theory, which states that people typically engage in moral compensatory behavior after engaging in ethically questionable behavior (Zhong et al., 2009). Rejecting another person potentially threatens one’s moral self-image because it inflicts severe social pain on another person and violates the norm of social acceptance (Wesselmann, Wirth, Pryor, Reeder, & Williams, 2013). Consequently, the sources of rejection are less likely to blame innocent victims. Such a prosocial initiation may help restore one’s moral image. Implications and Future Directions The present findings carry important implications and open new research areas. First, we speculate that people are less likely to blame innocent victims after rejecting another person because they want to restore their threatened self-image and the negative affect (e.g., shame and guilt) associated with their act of social rejection. Future research should be conducted to provide empirical evidence for these proposed mechanisms underlying the present observed effect. Although the present findings provide empirical support for our moral compensation account, it is still likely that the act of social rejection may change people’s perspective when learning about others’ sufferings. Prior research has shown that people have a heightened concern for potential future suffering following an act of social rejection (Chen et al., 2014). Given that people who have concern or experience of suffering usually have a better understanding of suffering


522


and higher empathy toward victims (e.g., Nordgren et al., 2011), the increased concern of future suffering may make the source of rejection take the victims’ perspective in understanding potential causes of their suffering. As a consequence, they may be less likely to blame the victims. Moreover, future research may also examine whether the source of rejection would engage in other compensatory behaviors (e.g., behaving less selfishly and behaving more prosocially), and whether these behaviors would influence various psychological outcomes (e.g., moral self-image and negative affect) among the source of social rejection. Addressing these issues may help to advance our understanding about the impacts of social rejection on the source. Second, we used the recall paradigm to induce feelings of rejecting another person, which is adopted from prior research (e.g., Chen et al., 2014; Williams et al., 1998). However, this manipulation does not allow researchers to control for some potential confounding variables, such as the targets, intensity, and significance of rejecting experiences. Future research may use an in-vivo paradigm to examine the impacts of concurrent rejecting experience (e.g., Gooley, Zadro, Williams, Svetieva, & Gonsalkorale, 2015; Wirth, Bernstein, & LeRoy, 2015). Third, we did not ask participants to indicate the extent to which people perceived their act of rejection as justified or not, which serves as one limitation of the present study. From an evolutionary perspective, social rejection evolved partly because it may potentially facilitate individual and group survival by allowing groups and individuals to reject people who can threaten group and individual well-being (Kurzban & Leary, 2001). Indeed, some people may choose to reject burdensome group members (Wesselmann, Wirth et al., 2013), and such rejection may be perceived as justified. Future research should also examine whether people will only be less likely to blame innocent victims when they believe their act of rejection is unjustified, and whether they will be more likely to blame innocent victims when they believe their act of rejection is justified and the targets of rejection deserve the rejection experience. Fourth, social rejection has many forms, such as romantic rejection (e.g., Baumeister et al., 1993), informational rejection (e.g., Jones, Carter-Sowell, Kelly, & Williams, 2009), induced rejection (e.g., Legate et al., 2013), and linguistic rejection (e.g., Dotan-Eliaz, Sommer, & Rubin, 2009). These forms may lead to different psychological and behavioral outcomes for the sources of rejection. Moreover, people may reject another person for different motives and reasons (see Nezlek, Wesselmann, Wheeler, & Williams, 2012; Williams, 2001). In particular, people may ignore waiters who are serving them in a restaurant (i.e., role-prescribed rejection) or they may reject a stranger whom they do not recognize (i.e., oblivious rejection). People may also reject deviants who have done something inappropriate (i.e., punitive rejection). Future research may also examine whether the forms and motives of social rejection would influence the relationship between general just-world beliefs and victim blaming after rejecting another person. Finally, the current research measured how sources of rejection attribute a female stranger’s hurtful experience. The results showed that the sources of rejection were less likely to make internal attributions for a girl’s suffering, and general just-world beliefs did not predict victim blaming among participants who recalled a past rejecting experience. Future research may examine whether similar findings will be observed if the victim is male.

POON AND CHEN

523

CONCLUSION


Our research shows that an act of social rejection overrides the influence of general just-world beliefs by making people less likely to blame an innocent victim of bullying. By providing the first experimental evidence for how the sources of rejection make sense of another person’s suffering, the present research advances our understanding of the aftermath of social rejection on the sources and creates new areas of future research.

ORCID Zhansheng Chen

http://orcid.org/0000-0003-0873-3391

NOTES 1. In this article, we used social rejection to describe the experience of rejecting, excluding, ignoring, and ostracizing other person(s) in social interaction. Although prior studies have started to reveal that these constructs could lead to different outcomes (e.g., Bernstein & Claypool, 2012a, 2012b; Molden, Lucas, Gardner, Dean, & Knowles, 2009; see also Smart Richman & Leary, 2009 for a review), we did not intend to examine potential differences between them. For the sake of brevity, we treated social rejection, social exclusion, and ostracism as synonyms and used them interchangeably in this article. 2. The relationship type for the social rejection experience was coded. Participants in the rejecting condition reported that they had rejected their friends (16.3%), families (4.1%), classmates/colleagues (22.4%), romantic admirers/partners (26.5%), and strangers (10.2%). In all, 20.4% participants did not specify their targets of rejection or reported that they rejected targets that could not be classified in the above categories (e.g., roommate, personal trainer). 3. A 2 (experimental condition: rejecting vs. accepting) × 2 (gender: male vs. female) was conducted to examine whether gender interacted with the experimental condition in predicting victim blaming. The results revealed only a significant main effect of gender, such that female participants made fewer internal attributions for the victim’s suffering than male participants, F(1, 89) = 4.76, p = .032, ηp 2 = .05. The interaction effect was not statistically significant, F(1, 89) = 2.41, p = .118, ηp 2 = .03. The main effect of the experimental condition on victim blaming was approaching significant after controlling for the effect of gender, F(1, 90) = 3.51, p = .064, ηp 2 = .04. The main purpose of the present research was to examine whether the experimental manipulation would moderate the positive association between general just-world beliefs and victim blaming. Controlling for the effect of gender did not substantially alter our observed moderation effect. The present research did not aim to examine the potential impacts of gender and we did not have specific predictions about the role of gender. For the sake of brevity, we reported the main regression analysis without controlling for the effect of gender in the main text.

AUTHOR NOTES Kai-Tak Poon is affiliated with the Department of Psychological Studies and Centre for Psychosocial Health, Hong Kong Institute of Education. Zhansheng Chen is affiliated with the Department of Psychology, University of Hong Kong.

REFERENCES Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Newbury Park, CA: Sage. Bastian, B., Jetten, J., & Fasoli, F. (2011). Cleansing the soul by hurting the flesh: The guilt-reducing effect of pain. Psychological Science, 22, 334–335. doi:10.1177/0956797610397058


524


Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonal attachments as a fundamental human motivation. Psychological Bulletin, 117, 497–529. doi:10.1037/0033-2909.117.3.497 Baumeister, R. F., Wotman, S. R., & Stillwell, A. M. (1993). Unrequited love: On heartbreak, anger, guilt, scriptlessness, and humiliation. Journal of Personality and Social Psychology, 64, 377–394. doi:10.1037/0022-3514.64.3.377 Bernstein, M. J., & Claypool, H. M. (2012a). Not all social exclusions are created equal: Emotional distress following social exclusion is moderated by exclusion paradigm. Social Influence, 7, 113–130. doi:10.1080/ 15534510.2012.664326 Bernstein, M. J., & Claypool, H. M. (2012b). Social exclusion and pain sensitivity: Why exclusion sometimes hurts and sometimes numbs. Personality and Social Psychology Bulletin, 38, 185–196. doi:10.1177/0146167211422449 Buhrmester, M. D., Kwang, T., & Gosling, S. D. (2011). Amazon’s mechanical turk: A new source of inexpensive, yet high-quality data? Perspectives on Psychological Science, 6, 3–5. doi:10.1177/1745691610393980 Carlsmith, J. M., & Gross, A. E. (1969). Some effects of guilt on compliance. Journal of Personality and Social Psychology, 11, 232–239. doi:10.1037/h0027039 Chen, Z., Poon, K.-T., Bernstein, M. J., & Teng, F. (2014). Rejecting another pains the self: The impact of perceived future rejection. Journal of Experimental Social Psychology, 50, 225–233. doi:10.1016/j.jesp.2013.10.007 Ciarocco, N. J., Sommer, K. L., & Baumeister, R. F. (2001). Ostracism and ego depletion: The strains of silence. Personality and Social Psychology Bulletin, 27, 1156–1163. doi:10.1177/0146167201279008 Correia, I., Vala, J., & Aguiar, P. (2001). The effects of belief in a just world and victim’s innocence on secondary victimization, judgements of justice and deservingness. Social Justice Research, 14, 327–342. doi:10.1023/ A:1014324125095 Dalbert, C. (1999). The world is more just for me than generally: About the personal belief in a just world scales validity. Social Justice Research, 12, 79–98. doi:10.1023/A:1022091609047 Dalbert, C. (2009). Belief in a just world. In M. R. Leary & R. H. Hoyle (Eds.), Handbook of individual differences in social behavior (pp. 288–297). New York, NY: Guilford Publications. DeWall, C. N., Twenge, J. M., Gitter, S. A., & Baumeister, R. F. (2009). It’s the thought that counts: The role of hostile cognition in shaping aggressive responses to social exclusion. Journal of Personality and Social Psychology, 96, 45–59. doi:10.1037/a0013196 Dotan-Eliaz, O., Sommer, K. L., & Rubin, Y. S. (2009). Multilingual groups: Effects of linguistic ostracism on felt rejection and anger, coworker attraction, perceived team potency, and creative performance. Basic and Applied Social Psychology, 31, 363–375. doi:10.1080/01973530903317177 Eisenberger, N. I., Lieberman, M. D., & Williams, K. D. (2003). Does rejection hurt? An fMRI study of social exclusion. Science, 302, 290–292. doi:10.1126/science.1089134 Eisinga, R., Te Grotenhuis, M., & Pelzer, B. (2013). The reliability of a two-item scale: Pearson, Cronbach, or SpearmanBrown? International Journal of Public Health, 58, 637–642. doi:10.1007/s00038-012-0416-3 Epley, N., Akalis, S., Waytz, A., & Cacioppo, J. T. (2008). Creating social connection through inferential reproduction: Loneliness and perceived agency in gadgets, gods, and greyhounds. Psychological Science, 19, 114–120. doi:10.1111/ psci.2008.19.issue-2 Festinger, L. (1957). A theory of cognitive dissonance. Evanston, IL: Row, Peterson. Furnham, A. (2003). Belief in a just world: Research progress over the past decade. Personality and Individual Differences, 34, 795–817. doi:10.1016/S0191-8869(02)00072-7 Gooley, S. L., Zadro, L., Williams, L. A., Svetieva, E., & Gonsalkorale, K. (2015). Ostracizing for a reason: A novel source paradigm for examining the nature and consequences of motivated ostracism. Journal of Social Psychology, 155(5), 410–431. Hafer, C. L. (2000). Investment in long-term goals and commitment to just means drive the need to believe in a just world. Personality and Social Psychology Bulletin, 26, 1059–1073. doi:10.1177/01461672002611004 Hafer, C. L. (2002). Why we reject innocent victims. In M. Ross & D. T. Miller (Eds.), The justice motive in everyday life (pp. 109–126). New York, NY: Cambridge University Press. Hafer, C. L., & Bègue, L. (2005). Experimental research on just-world theory: Problems, developments, and future challenges. Psychological Bulletin, 131, 128–167. doi:10.1037/0033-2909.131.1.128 Hafer, C. L., & Correy, B. L. (1999). Mediators of the relation between beliefs in a just world and emotional responses to negative outcomes. Social Justice Research, 12, 189–204. doi:10.1023/A:1022144317302 Jones, E. E., Carter-Sowell, A., Kelly, J. R., & Williams, K. D. (2009). ‘I’m out of the loop’: Ostracism through information exclusion. Group Processes & Intergroup Relations, 12, 157–174. doi:10.1177/1368430208101054


POON AND CHEN

525

Kouchaki, M., & Wareham, J. (2015). Excluded and behaving unethically: Social exclusion, physiological responses, and unethical behavior. Journal of Applied Psychology, 100, 547–556. doi:10.1037/a0038034 Kross, E., Berman, M., Mischel, W., Smith, E. E., & Wager, T. (2011). Social rejection shares somatosensory representations with physical pain. Proceedings of the National Academy of Sciences, 108, 6270–6275. doi:10.1073/ pnas.1102693108 Kurzban, R., & Leary, M. R. (2001). Evolutionary origins of stigmatization: The functions of social exclusion. Psychological Bulletin, 127, 187–208. doi:10.1037/0033-2909.127.2.187 Legate, N., DeHaan, C. R., Weinstein, N., & Ryan, R. M. (2013). Hurting you hurts me too: The psychological costs of complying with ostracism. Psychological Science, 24, 583–588. doi:10.1177/0956797612457951 Lerner, M. J. (1980). The belief in a just world: A fundamental delusion. New York, NY: Plenum Press. Lerner, M. J., & Simmons, C. H. (1966). The observer’s reaction to the “innocent victim”: Compassion or rejection? Journal of Personality and Social Psychology, 4, 203–210. doi:10.1037/h0023562 Maner, J. K., DeWall, C. N., Baumeister, R. F., & Schaller, M. (2007). Does social exclusion motivate interpersonal reconnection? Resolving the “porcupine problem.” Journal of Personality and Social Psychology, 92, 42–55. doi:10.1037/ 0022-3514.92.1.42 Mazar, N., & Zhong, C. (2010). Do green products make us better people? Psychological Science, 21, 494–498. doi:10.1177/0956797610363538 McParland, J. L., & Knussen, C. (2010). Just world beliefs moderate the relationship of pain intensity and disability with psychological distress in chronic pain support group members. European Journal of Pain, 14, 71–76. doi:10.1016/ j.ejpain.2008.11.016 Miller, D. T., & Effron, D. A. (2010). Psychological license: When it is needed and how it functions. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 43, pp. 115–155). San Diego, CA: Academic Press. Molden, D. C., Lucas, G. M., Gardner, W. L., Dean, K., & Knowles, M. (2009). Motivations for prevention or promotion following social exclusion: Being rejected versus being ignored. Journal of Personality and Social Psychology, 96, 415–431. doi:10.1037/a0012958 Monin, B., & Miller, D. T. (2001). Moral credentials and the expression of prejudice. Journal of Personality and Social Psychology, 81, 33–43. doi:10.1037/0022-3514.81.1.33 Nesbit, S. M., Blankenship, K. L., & Murray, R. A. (2012). The influence of just-world beliefs on driving anger and aggressive driving intentions. Aggressive Behavior, 38, 389–402. doi:10.1002/ab.2012.38.issue-5 Nezlek, J. B., Wesselmann, E. D., Wheeler, L., & Williams, K. D. (2012). Ostracism in everyday life. Group Dynamics: Theory, Research, and Practice, 16, 91–104. doi:10.1037/a0028029 Nordgren, L. F., Banas, K., & MacDonald, G. (2011). Empathy gaps for social pain: Why people underestimate the pain of social suffering. Journal of Personality and Social Psychology, 100, 120–128. doi:10.1037/a0020938 Poon, K.-T., & Chen, Z. (2014). When justice surrenders: The effect of just-world beliefs on aggression following ostracism. Journal of Experimental Social Psychology, 52, 101–112. doi:10.1016/j.jesp.2014.01.006 Poon, K.-T., Chen, Z., & DeWall, C. N. (2013). Feeling entitled to more: Ostracism increases dishonest behavior. Personality and Social Psychology Bulletin, 39, 1227–1239. doi:10.1177/0146167213493187 Poon, K.-T., Teng, F., Chow, J. T., & Chen, Z. (2015). Desiring to connect to nature: The effect of ostracism on ecological behavior. Journal of Environmental Psychology, 42, 116–122. doi:10.1016/j.jenvp.2015.03.003 Poulsen, J. R., & Kashy, D. A. (2012). Two sides of the ostracism coin: How sources and targets of social exclusion perceive themselves and one another. Group Processes & Intergroup Relations, 15, 457–470. doi:10.1177/ 1368430211430517 Ramanathan, S., & Williams, P. (2007). Immediate and delayed emotional consequences of indulgence: The moderating influence of personality type on mixed emotions. Journal of Consumer Research, 34, 212–223. doi:10.1086/518524 Sachdeva, S., Iliev, R., & Medin, D. L. (2009). Sinning saints and saintly sinners: The paradox of moral self-regulation. Psychological Science, 20, 523–528. doi:10.1111/psci.2009.20.issue-4 Smart Richman, L., & Leary, M. R. (2009). Reactions to discrimination, stigmatization, ostracism, and other forms of interpersonal rejection: A multimotive model. Psychological Review, 116, 365–383. doi:10.1037/a0015250 Tetlock, P. E., Kristel, O. V., Elson, S. B., Green, M. C., & Lerner, J. S. (2000). The psychology of the unthinkable: Taboo trade-offs, forbidden base rates, and heretical counterfactuals. Journal of Personality and Social Psychology, 78, 853–870. doi:10.1037/0022-3514.78.5.853


526


Tomaka, J., & Blascovich, J. (1994). Effects of justice beliefs on cognitive appraisal of and subjective physiological and behavioral responses to potential stress. Journal of Personality and Social Psychology, 67, 732–740. doi:10.1037/ 0022-3514.67.4.732 Twenge, J. M., Baumeister, R. F., Tice, D. M., & Stucke, T. S. (2001). If you can’t join them, beat them: Effects of social exclusion on aggressive behavior. Journal of Personality and Social Psychology, 81, 1058–1069. doi:10.1037/ 0022-3514.81.6.1058 Wesselmann, E. D., Bagg, D., & Williams, K. D. (2009). “I feel your pain”: The effects of observing ostracism on the ostracism detection system. Journal of Experimental Social Psychology, 45, 1308–1311. doi:10.1016/ j.jesp.2009.08.003 Wesselmann, E. D., Butler, F. A., Williams, K. D., & Pickett, C. L. (2010). Adding injury to insult: Unexpected rejection leads to more aggressive responses. Aggressive Behavior, 35, 1–6. Wesselmann, E. D., Cardoso, F. D., Slater, S., & Williams, K. D. (2012). To be looked at as though air: Civil attention matters. Psychological Science, 23, 166–168. doi:10.1177/0956797611427921 Wesselmann, E. D., Williams, K. D., & Hales, A. H. (2013). Vicarious ostracism. Frontiers in Human Neuroscience, 7, 153. doi:10.3389/fnhum.2013.00153 Wesselmann, E. D., Wirth, J. H., Pryor, J. B., Reeder, G. D., & Williams, K. D. (2013). When do we ostracize? Social and Psychological and Personality Science, 4, 108–115. doi:10.1177/1948550612443386 Williams, K. D. (2001). Ostracism: The power of silence. New York, NY: Guilford Publications. Williams, K. D. (2007). Ostracism. Annual Review of Psychology, 58, 425–452. doi:10.1146/annurev.psych.58. 110405.085641 Williams, K. D. (2009). Ostracism: Effects of being excluded and ignored. Advances in Experimental Social Psychology, 41, 275–314. Williams, K. D., Bernieri, F., Faulkner, S., Grahe, J., & Gada-Jain, N. (2000). The scarlet letter study: Five days of social ostracism. Journal of Personal and Interpersonal Loss, 5, 19–63. doi:10.1080/10811440008407846 Williams, K. D., Cheung, C. K. T., & Choi, W. (2000). Cyberostracism: Effects of being ignored over the Internet. Journal of Personality and Social Psychology, 79, 748–762. doi:10.1037/0022-3514.79.5.748 Williams, K. D., Shore, W. J., & Grahe, J. E. (1998). The silent treatment: Perceptions of its behaviors and associated feelings. Group Processes & Intergroup Relations, 1, 117–141. doi:10.1177/1368430298012002 Wirth, J. H., Bernstein, M. J., & LeRoy, A. S. (2015). Atimia: A new paradigm for investigating how individuals feel when ostracizing others. Journal of Social Psychology, 155(5), 497–514. Zhong, C., & Liljenquist, K. (2006). Washing away your sins: Threatened morality and physical cleansing. Science, 313, 1451–1452. doi:10.1126/science.1130726 Zhong, C., Liljenquist, K., & Cain, D. M. (2009). Moral self-regulation: Licensing and compensation. In D. De Cremer (Ed.), Psychological perspectives on ethical behavior and decision making. Charlotte, NC: Information Age Publishing.




Insufficient Justification for Exclusion Prompts Compensatory Behavior a

a

Daryl R. Van Tongeren , Lindsey M. Root Luna & Charlotte VanOyen Witvliet

a

a

Hope College Published online: 12 Aug 2015.

Click for updates To cite this article: Daryl R. Van Tongeren, Lindsey M. Root Luna & Charlotte VanOyen Witvliet (2015) Insufficient Justification for Exclusion Prompts Compensatory Behavior, The Journal of Social Psychology, 155:5, 527-534, DOI: 10.1080/00224545.2015.1060936 To link to this article: http://dx.doi.org/10.1080/00224545.2015.1060936






Insufficient Justification for Exclusion Prompts Compensatory Behavior DARYL R. VAN TONGEREN LINDSEY M. ROOT LUNA CHARLOTTE VANOYEN WITVLIET Hope College

ABSTRACT. This experiment examined participants’ compensatory behavior toward an excluded stranger. Participants engaged in a four-person social introductions task and rank-ordered each other member of the group; the lowest ranked person was excluded from a subsequent game. Using a 2 × 2 design, participants were randomly assigned to a justification condition (insufficient vs. sufficient) and to an exclusion responsibility condition (responsible for exclusion vs. exclusion by random selection). Results revealed that after limited introductions (i.e., insufficient justification for one’s ranking decision), being responsible for the exclusion prompted compensatory behavior toward the excluded stranger. However, after extended introductions (i.e., sufficient justification of one’s ranking decision), participants did not compensate the excluded person. These results suggest that insufficient justification for exclusion may lead to compensatory behavior, when one is responsible for the exclusion. Keywords: compensatory behavior, responsibility, restitution, sources of exclusion

BUILDING AND MAINTAINING CLOSE RELATIONSHIPS is a fundamental feature of social life. Humans demonstrate a strong need to belong, presumably driven, in part, by evolutionary instincts for survival and reproduction (Baumeister & Leary, 1995). When individuals are excluded or rejected from a group, they typically experience a variety of negative reactions, such as decreased self-esteem (Leary, Tambor, Terdal, & Downs, 1995), emotional numbness (DeWall, Baumeister, & Masicampo, 2009), reduced prosocial behavior (Twenge, Baumeister, DeWall, Ciarocco, & Bartels, 2007), and even a loss of meaning in life (Stillman et al., 2009; Twenge, Catanese, & Baumeister, 2003). Given the deleterious effects of social exclusion, research which identifies factors that predict compensatory behaviors is valuable. Research on exclusion, rejection, and ostracism has a long history.1 Schachter’s (1951) original “deviate” study demonstrated how rejection may be used as a social punishment or a form of social control for a dissenting group member, and this work has recently been replicated (Wesselmann et al., 2014). However, sufficient justification is necessary for exclusion. Excluding a group member without sufficient justification likely violates an “inclusion norm” and can lead Address correspondence to Daryl R. Van Tongeren, Hope College, Department of Psychology, Schaap Science Center, 35 E. 12th Street, Holland, MI 49422-9000, USA. E-mail: [email protected]

528



to greater sympathy for the excluded individual (Wesselmann, Wirth, Pryor, Reeder, & Williams, 2013). However, with proper justification, such as when a group member is considered burdensome, exclusion is a useful strategy for social control within the group (Williams, 2009). Thus, individuals likely need sufficient justification for their exclusion decisions, and without it, they might be motivated to compensate the excluded individual for their exclusion decision. We sought to expand on this basic premise by conducting an experiment to examine the effects of justification on compensatory behavior toward the excluded individual.

MOTIVATING COMPENSATORY BEHAVIOR What situational factors may motivate an individual to behave in ways that compensate for excluding another person? We suspect that the first key dimension in eliciting conciliatory behaviors is sufficient justification for the exclusion decision. Longer and more extensive social encounters provide additional opportunities to gather social information, which establishes justification for future decisions compared to shorter, less extensive encounters. Both balance theory (Heider, 1958) and cognitive dissonance theory (Festinger, 1957) underscore that individuals are motivated to maintain cognitive consistency among their cognitions and behaviors. Indeed, it has been postulated that cognitive consistency is a core social motivation (Gawronski, 2012; Gawronski & Strack, 2012), and recently this theorizing has been extended to work on exclusion (Critcher & Zayas, 2014). Accordingly, individuals may behave in ways that affirm their previously established cognitions. Thus, in shorter, less extensive encounters, individuals likely lack the social information necessary to justify an inclusion/exclusion decision, resulting in a greater degree of dissonance. Such dissonance might motivate future compensatory responses. On the other hand, a longer, more extensive encounter may result in greater justification for one’s future inclusion/exclusion decision. Put differently, following an irrevocable decision that might exclude an individual one has just met, one way of reducing the dissonance aroused by excluding another may be to strengthen one’s certainty that committing the offense was justifiable. Thus, we may expect to see behavioral compensation occurring when an individual lacks the time and information to sufficiently justify an exclusion decision. A second feature that may affect how an individual responds toward an excluded other is the individual’s perception that he or she is responsible for the exclusion. Without the implication of culpability, there would be no reason to engage in conciliatory behaviors toward the victim. Feeling responsible for an offense can increase guilt and feelings of responsibility regarding the behavior, which can motivate compensatory behaviors as part of the relational repair process (Fisher & Exline, 2006; Riek, Root Luna, & Schnabelrauch, 2014; Struthers, Eaton, Mendoza, Santelli, & Shirvani, 2010). Thus, we expected to only see participants engage in compensatory reactions when they were implicated as responsible for the exclusion decision.

OVERVIEW AND HYPOTHESIS To systematically investigate the effects of justification and responsibility for exclusion on compensatory behavior, we conducted a laboratory study to create the illusion of excluding a stranger within the lab. Using a 2 × 2 design, participants were randomly assigned to a justification


VAN TONGEREN, ROOT LUNA, AND VANOYEN WITVLIET

529

condition (insufficient vs. sufficient) and then again to an exclusion responsibility condition (responsible for exclusion vs. exclusion by random selection). Following this, they had the opportunity to engage in compensatory behavior toward the excluded stranger. Given that exclusion requires sufficient justification to avoid violating social norms (Wesselmann et al., 2013), we predicted that in conditions in which participants believed they had excluded a stranger without sufficient justification, they would be more likely to engage in compensatory behavior toward the stranger. However, we suspected this would only be the case when the participant believed that he or she was responsible for excluding the stranger (rather than the exclusion occurring because of random selection). Thus, we predicted an interaction effect between justification and exclusion responsibility on compensatory behavior toward the excluded individual. More specifically, we predicted that whereas there would be no effect of responsibility in the sufficient justification condition, participants with insufficient justification who believed they were responsible for exclusion would compensate by assigning fewer boring math problems to the excluded participant than when they had insufficient justification but were not responsible for exclusion.

METHOD Participants Participants were 128 (94 female) undergraduates enrolled in an introductory psychology course at a small, private college in the Midwest. Data from seven participants were dropped because (a) software failure terminated the study prior to the performance allocation measure for two participants, (b) four participants failed to follow instructions, and (c) one participant was a statistical outlier on the allocation task.2 This resulted in a final sample of 121 (87 female) participants in 33 groups, and they ranged in age from 17–26 years (M = 18.82, SD = 1.21). Procedure Participants were recruited to the lab in groups of four and were told that they would be completing several tasks as a group. They would first spend time getting to know one another, and then they would complete a variety of tasks either together as a group or individually, including the possibility of being assigned to a boring peg-turning task. To convey the unpleasant nature of the peg-turning task, the experimenter demonstrated, monotonously, how to turn the pegs on a small pegboard (cf. Festinger & Carlsmith, 1959). First, participants were randomly assigned to a social introductions task condition as a manipulation of justification for a subsequent ranking of people in the group. For those in the insufficient justification condition (i.e., limited introductions), participants wore participant numbers (ranging from 1–4) and spent 5 minutes getting to know each other, whereas for those in the sufficient justification condition (i.e., extended introductions), participants wore name tags, spent 10 minutes getting to know information about each other, and—once alone in an individual room within the laboratory suite—wrote down characteristics they valued in each other group member. This would allow participants to gain more social information about each individual that could be used to justify a potential exclusion decision. After the group interaction, participants completed all


530


remaining activities on a computer in separate experiment rooms. Next, they ranked how much they wanted to work with each other participant on future tasks: first, second, or last. Participants were also randomly assigned to the exclusion responsibility condition; they were all informed (via computer) that the group member they ranked last was selected to complete the boring peg-turning task (i.e., was excluded) while the remaining members (including the participant) played a video game in which they would collaborate via computers to earn a combined score. Participants in the random exclusion condition were told that assignment of one group member to the boring task was based on random selection, whereas those in the responsible for exclusion condition were told that assignment of the excluded group member to the boring task was because the participant had ranked him or her last. All participants completed this protocol and no participants were actually excluded nor required to complete the boring peg-turning task. Following this, in order to enhance the cover story of the experiment, participants were told that they and their other group members would be playing a video game on the computer for 5 minutes, in which they had to successfully navigate a character around obstacles, and their goal was to get the highest collaborative score (compared to other participant groups in the study). After this, as a manipulation check, they rated (on 1–5 scales) how much guilt they felt for playing the game (while another participant was purportedly turning the pegboard) and how responsible they felt for the other group member’s exclusion. Finally, participants were told that they had been randomly selected to be the leader for their team for the final task: assign 700 mathematical division problems among the four members of the group (including themselves).3 Participants were told the division problems were described by previous participants as “boring, dull, and repetitive” and they would need to complete those problems as the final part of the study (although the study ended after the allocation decision). Participants were assured that their allocation decisions were anonymous (i.e., no other participant knew who made the allocation decisions) and final (i.e., no one could contest the decision). This allocation task served as the behavioral measure of compensation for the offense of rejecting the last-ranked team member.

RESULTS Manipulation Check We sought to ensure that our responsibility manipulation was effective in eliciting feelings of responsibility. Participants in the exclusion responsibility condition reported feeling greater guilt (M = 3.39, SD = 1.23) than participants in the random exclusion condition (M = 2.64, SD = 1.28), t(120) = 3.32, p = .001, d = .60 (mean difference = .76, 95% CI[0.31, 1.21]), and greater feelings of responsibility (M = 3.59, SD = 1.19) for the participant who was excluded from the game than those in the no responsibility condition (M = 1.87, SD = 1.27), t(120) = 7.73, p < .001, d = 1.40 (mean difference = 1.71, 95% CI[1.26, 2.15]). Primary Analyses We tested the interactive effects of justification and responsibility on a compensatory response, as measured by the work allocation task. A 2 (justification: insufficient vs. sufficient) × 2


531

TABLE 1 Percentage of Division Problems (and Standard Deviation) Assigned to the Excluded Team Member out of the Total Number Assigned to Both the Participant and the Excluded Team Member


Random exclusion Responsible for exclusion Overall

Insufficient justification

Sufficient justification

Overall

62.02% (23.41%) 48.14% (10.50%) 55.08% (19.30%)

55.10% (16.94%) 55.10% (17.25%) 55.10% (16.95%)

58.68% (20.67%) 51.56% (14.53%)

(responsible for exclusion vs. exclusion by random selection) ANOVA on compensatory behavior, measured as the ratio of division problems assigned to the excluded team member divided by the total number assigned to the participant and the excluded team member (see Table 1). Although there was no main effect for justification condition, F(1, 113) = .00, p = .996, partial η2 < .001 (mean difference = 0.00%, 95% CI[−6.5%, 6.4%]), there was a significant main effect for responsibility, F(1, 113) = 4.52, p = .036, partial η2 = .038, (mean difference = 6.9%, 95% CI[.005%, 13.4%]) in which being responsible for the exclusion evoked assigning a significantly lower percentage of work to the excluded participant (presumably as a signal of compensation). However, this was subsumed by a significant interaction effect, F(1, 113) = 4.53, p = .036, partial η2 = .039.4 Follow-up comparisons (with Bonferroni corrections) revealed that, as predicted, being responsible for exclusion did not alter compensatory behavior among those with sufficient justification for exclusion (mean difference = 0.00%, 95% CI[−9.3%, 9.3%], p = .999). Moreover, as predicted, the critical comparison was significant: for those with insufficient justification for exclusion, being responsible for exclusion prompted allocations of a significantly lower percentage of problems to the excluded participant compared to those who were told the last-ranked group member was randomly excluded, F(1, 113) = 9.29, p = .003, partial η2 = .076 (mean difference = 13.88%, 95% CI[4.9%, 22.9%]). This critical comparison provides support for our primary hypothesis: participants whose low ranking of a group member was based only on a brief, limited encounter (and thus lacked sufficient justification for that ranking), and who perceived that they were responsible for the exclusion, responded with compensatory behavior toward the excluded individual relative to those who did not feel responsible for exclusion. However, for those participants who justified their judgments through an extended social encounter, being responsible for exclusion did not prompt compensatory behaviors toward the excluded individual.

DISCUSSION Based on previous research (Wesselmann et al., 2013), we predicted that insufficient justification for an exclusion decision would prompt compensatory behaviors toward the excluded individual. However, this only occurred when individuals were responsible for the exclusion of another group member based on insufficient justification for ranking the excluded person last. When participants excluded a stranger without enough social information to provide a sufficient justification


532


for doing so, they compensated by assigning this excluded individual fewer boring division problems on a subsequent task. However, when participants had sufficient justification to rationalize their exclusion decision (through a more thorough encounter prior to exclusion), their compensatory behaviors mirrored those who were told the participant they ranked last was excluded from the game due to random selection. Thus, in true exclusion situations (when one is actually responsible for the exclusion), insufficient justification can lead to compensatory behaviors toward the excluded individual. Because exclusion can be a form of social control for managing members of a group (Williams, 2009) and excluding an individual without proper justification can lead to positive feelings toward the excluded person (Wesselmann et al., 2013), compensatory behaviors (i.e., reducing unfair workloads in future group tasks) toward the excluded individual may signal an effort to make restitution for one’s offense toward the excluded individual and recover her or his place in the group, thus overcoming the previous violation of the inclusion norm. This study had several strengths. First, we used a live interaction method, rather than a computer-simulation exclusion paradigm (e.g., Cyberball). Although such methodological approaches have merit and are important for advancing research on exclusion, our procedure was high in mundane realism. Second, we used a behavioral measure of compensation following exclusion, rather than relying solely on self-report of one’s feelings or behavioral intentions. Third, this work advances prior work on exclusion by clarifying that the degree of justification for excluding another is an important consideration in the compensatory behaviors for the exclusion. That is, individuals with sufficient justification for exclusion, as well as those who did not feel responsible for the exclusion decision, were not motivated to attempt to repair the relationship through compensatory behavior. Previous work has not directly addressed the issue of justification for exclusion (e.g., Ciarocco, Sommer, & Baumeister, 2001; Legate, DeHaan, Weinstein, & Ryan, 2013; Poulsen & Kashy, 2012), yet future work might benefit from the consideration of justification (cf. Sommer & Yoon, 2013). Finally, this work highlights the interactive effect of perceived responsibility for exclusion (rather than exclusion through random selection) and justification in compensatory behaviors. Nonetheless, our work had some limitations; for example, we used self-reported feelings of guilt and responsibility as our primary manipulation checks. We acknowledge that explicitly asking participants how justified their ranking and the exclusion were would have provided a more explicit and face valid measurement. Future work could explore whether cognitive judgments (e.g., justification) and affective responses (e.g., guilt) exert differential effects on compensation. Similarly, future research could advance this work by examining actual interactions following an exclusion episode (not just beforehand) and coding for compensatory inclusion behaviors toward the excluded individual under conditions of insufficient justification. Moreover, the identification of mediating mechanisms (e.g., empathy) or contextual moderators (e.g., similarity) may be useful in advancing an understanding of compensatory behavior following exclusion. Our results suggest that individuals make compensatory efforts toward others they are responsible for excluding when they have insufficient justification for their exclusion. Given the fundamental need to belong and the social utility of effectively navigating group interactions, research examining how individuals respond following excluding another is important for reducing the deleterious effects of exclusion and better understanding ways of promoting relational repair that restores social harmony.


533


NOTES 1. Although highly similar, we acknowledge the difference between ostracism, rejection, and exclusion. Williams (2009) contends that ostracism involves both excluding and ignoring others, rejection involves a clear pronouncement that one is not wanted, and exclusion is the process of being separated or held apart from others. For the sake of clarity, and because our method most clearly aligns with exclusion, we will use the term exclusion for the remainder of the paper, though it is possible that some of the processes we describe apply to the other terms as well. 2. Inclusion or removal of the outlier did not substantively change the results; the key interaction was still significant if the outlier was left in the sample. 3. We pilot-tested this work allocation task on a separate sample of 59 undergraduate students. After imagining interacting with strangers for 5–10 minutes, participants allocated work based on rankings. As predicted, a repeated measures t test revealed that participants allocated themselves fewer division problems (M = 151.95, SD = 77.60) than those they ranked last (M = 200.42, SD = 93.75), t(58) = 2.51, p = .015, d = .56 (mean difference = 48.47, 95% CI [9.83, 87.12]). 4. We also analyzed the data by examining the number of problems assigned to the last-ranked team member as the primary dependent variable, and although there were no significant main effects, the same predicted interaction emerged as significant, F(1, 115) = 6.09, p = .015, partial η2 = .050, as was the key simple main effect of responsibility in the insufficient justification condition, F(1, 115) = 7.37, p = .008, partial η2 = .060 (mean difference = 26.44 problems, 95% CI[7.15, 45.74 problems]), demonstrating a similar pattern as the results we report.

AUTHOR NOTES Daryl R. Van Tongeren, Lindsey M. Root Luna, and Charlotte VanOyen Witvliet are affiliated with the Department of Psychology at Hope College.

REFERENCES Baumeister, R. F., & Leary, M. (1995). The need to belong: Desire for interpersonal attachments as a fundamental human motivation. Psychological Bulletin, 117, 497–529. doi:10.1037/0033-2909.117.3.497 Ciarocco, N. J., Sommer, K. L., & Baumeister, R. F. (2001). Ostracism and ego depletion: The strains of silence. Personality and Social Psychology Bulletin, 27, 1156–1163. doi:10.1177/0146167201279008 Critcher, C. R., & Zayas, V. (2014). The involuntary excluder effect: Those included by an excluder are seen as exclusive themselves. Journal of Personality and Social Psychology, 107, 454–474. doi:10.1037/a0036951 DeWall, C. N., Baumeister, R. F., & Masicampo, E. J. (2009). Feeling rejected but not much else: Resolving the paradox of emotional numbness following social exclusion. In A. L. Vangelisti (Ed.), Feeling hurt in close relationships (pp. 123–142). Cambridge: New York, NY. Festinger, L. (1957). A theory of cognitive dissonance. Stanford, CA: Stanford University Press. Festinger, L., & Carlsmith, J. M. (1959). Cognitive consequences of forced compliance. The Journal of Abnormal and Social Psychology, 58(2), 203–210. Fisher, M. L., & Exline, J. J. (2006). Self-forgiveness versus excusing: The roles of remorse, effort, and acceptance of responsibility. Self and Identity, 5, 127–146. doi:10.1080/15298860600586123 Gawronski, B. (2012). Back to the future of dissonance theory: Cognitive consistency as a core motive. Social Cognition, 30, 652–668. doi:10.1521/soco.2012.30.6.652 Gawronski, B., & Strack, F. (Eds.). (2012). Cognitive consistency: A fundamental principle in social cognition. New York, NY: Guilford. Heider, F. (1958). The psychology of interpersonal relations. New York, NY: John Wiley & Sons. Leary, M. R., Tambor, E. S., Terdal, S. K., & Downs, D. L. (1995). Self-esteem as an interpersonal monitor: The sociometer hypothesis. Journal of Personality and Social Psychology, 68, 518–530. Legate, N., DeHaan, C. R., Weinstein, N., & Ryan, R. M. (2013). Hurting you hurts me too: The psychological costs of complying with ostracism. Psychological Science, 24, 583–588. doi:10.1177/0956797612457951


534


Poulsen, J. R., & Kashy, D. A. (2012). Two sides of the ostracism coin: How sources and targets of social exclusion perceive themselves and one another. Group Processes & Intergroup Relations, 15, 457–470. doi:10.1177/ 1368430211430517 Riek, B. M., Root Luna, L. M., & Schnabelrauch, C. A. (2014). Transgressors’ guilt and shame: A longitudinal examination of forgiveness seeking. Journal of Social and Personal Relationships, 31, 751–772. doi:10.1177/ 0265407513503595 Schachter, S. (1951). Deviation, rejection, and communication. Journal of Abnormal and Social Psychology, 46, 190–208. Sommer, K., & Yoon, J. (2013). When silence is golden: Ego depletion following aversive social interactions. Journal of Personal and Social Relationships, 30, 901–919. doi:10.1177/0265407512473006 Stillman, T. F., Baumeister, R. F., Lambert, N. M., Crescioni, A. W., DeWall, C. N., & Fincham, F. D. (2009). Alone and without purpose: Life loses meaning following social exclusion. Journal of Experimental Social Psychology, 45, 686–694. doi:10.1016/j.jesp.2009.03.007 Struthers, C. W., Eaton, J., Mendoza, R., Santelli, A. G., & Shirvani, N. (2010). Interrelationship among injured parties’ attributions of responsibility, appraisal of appropriateness to forgive the transgressor, forgiveness, and repentance. Journal of Applied Social Psychology, 40, 970–1002. doi:10.1111/jasp.2010.40.issue-4 Twenge, J. M., Baumeister, R. F., DeWall, C. N., Ciarocco, N. J., & Bartels, J. M. (2007). Social exclusion decreases prosocial behavior. Journal of Personality and Social Psychology, 92, 56–66. doi:10.1037/0022-3514.92.1.56 Twenge, J. M., Catanese, K. R., & Baumeister, R. F. (2003). Social exclusion and the deconstructed state: Time perception, meaninglessness, lethargy, lack of emotion, and self-awareness. Journal of Personality and Social Psychology, 85, 409–423. doi:10.1037/0022-3514.85.3.409 Wesselmann, E., Williams, K., Pryor, J., Eichler, F., Gill, D., & Hogue, J. (2014). Revisiting Schachter’s research on rejection, deviance, and communication (1951). Social Psychology, 45, 164–169. doi:10.1027/1864-9335/a000180 Wesselmann, E., Wirth, J., Pryor, J., Reeder, G., & Williams, K. (2013). When do we ostracize? Social Psychological and Personality Science, 4, 108–115. doi:10.1177/1948550612443386 Williams, K. D. (2009). Ostracism: Effects of being excluded and ignored. In M. Zanna (Ed.), Advances in experimental social psychology (pp. 275–314). New York, NY: Academic Press.

Received March 15, 2015 Accepted June 3, 2015