Negative Self-Schemas and Social Rejection Decisions

Gili Freedman1 and Justin Dainer-Best2 1St. Mary’s College of Maryland 2Bard College

How do self-schemas and their consequences guide social rejection decisions? When making rejection decisions, individuals consider the difficulty and emotional consequences of rejecting, and both of these considerations are likely to involve self-schemas. In three preregistered stud- ies, we examine the roles of self-esteem, symptoms of depression and anxiety, and rejection sensitivity in rejection decisions. In an initial set of studies (N1a = 214, N1b = 264), partic- ipants forecast their willingness to reject and their emotional responses in friendship (Study 1a) and romantic (Study 1a–1b) vignettes. In Study 2 (N2 = 259), participants who recently rejected rated that experience on the same measures. Correlates of negative self-schema were associated with negative emotions. General distress, self-esteem, and rejection sensitivity were associated with forecasted difficulty rejecting, but only anxiety and general distress were asso- ciated with retrospectively-reported increased difficulty. Taken together, psychological distress may decrease willingness to reject in a way that participants cannot predict.

Keywords: social rejection, depression, anxiety, self-esteem, rejection sensitivity, self-schema

Introduction in depression and anxiety disorders (Kendler et al., 2003; Nolan et al., 2003; Slavich et al., 2010; Slavich et al., Self-schemas, representations of the self that result from 2009). They are also theorized to directly cause negative past reactions and experience (Beck, 1967; Segal, 1988) play self-schema (Beck & Bredemeier, 2016). However, while an important role in the experience of being rejected, and these corollaries of negative self-schemas have been impli- negative self-schemas may be particularly influential. One cated in more elaborative decision-making about negative aspect of a negative self-schema is lower feelings of self- stimuli (e.g., Dainer-Best et al., 2017), it is unclear whether worth (i.e., low self-esteem), and such negative self-schemas they directly impact individuals’ active use of social rejec- may have downstream associations with constructs associ- tion. Such rejection may take various forms, from minor ated with rejection such as anxious concerns about interper- (turning down an invitation to get coffee) to major (ending sonal acceptance (i.e., high rejection sensitivity) and memo- a friendship or romantic relationship). How will an indi- ries of ended relationships. Self-esteem can be construed as vidual’s feelings about engaging in rejection be impacted by a measure of one’s level of interpersonal acceptance (Leary their self-schema and self-concept? Will those anticipated et al., 1995), and being rejected reduces self-esteem (e.g., feelings come to fruition? Leary, 1990; Williams, 2007). Similarly, individuals with higher levels of rejection sensitivity anxiously expect—and Social rejection occurs in a wide range of contexts and strongly react to—experiences of rejection (Downey & Feld- with many types of relationships including romantic rela- man, 1996). Furthermore, rejection experiences can be an tionships as well as friendships. Yet, despite the ubiquity impetus for forming negative views of the self, implicated of social rejection, engaging in rejection can be a difficult endeavor for the rejector. First, engaging in social rejec- tion is generally perceived as effortful (Ciarocco et al., 2001; Williams et al., 2000). Individuals are often reluctant to re- St. Mary’s College of Maryland, 18952 E. Fisher Rd., St. Mary’s ject romantic suitors even when they are not interested in pur- City, MD 20686, USA. Bard College, 30 Campus Rd, Annandale- suing a romantic relationship with them (Joel et al., 2014). on-Hudson, NY 12504, USA Gili Freedman; Justin Dainer-Best Second, engaging in social rejection can lead to unpleas- Authors’ contributions: GF and JDB collaboratively designed the ant emotional experiences, including guilt and negative af- study, analyzed the data, and drafted the manuscript. GF managed fect (e.g., Baumeister et al., 1993; Legate et al., 2013; Peril- data collection and privacy. loux & Buss, 2008; Poulsen & Kashy, 2012) and negative af- Availability of data, code, and material: Data, code (in SPSS and fect. The effortfulness of rejecting coupled with the potential R syntax), and materials are available on the Open Science Frame- for negative emotional consequences make for a potentially work (OSF). difficult decision for a would-be-rejector. How a would-be- 2 FREEDMAN & DAINER-BEST rejector thinks about the self, particularly in relation to oth- For example, an ambiguous situation will be interpreted neg- ers, is likely to play an important role in that decision-making atively if one is feeling depressed or anxious (Watkins et al., process. Thus, in the present research we examine the role of 1992), and therefore moments of potential rejection are pro- four constructs closely connected to negative self-schema— cessed through past experience of being rejected and engag- low self-esteem, anxious and depressive symptoms, and re- ing in rejection oneself. It is clear that ending a relationship jection sensitivity—in vignette-based affective forecasts and is depressing—for example, in a meta-analysis of 21 stud- actual experiences of social rejection decisions. We argue ies, the dissolution of relationships was positively associated that each will be invoked in the decision-making process as with depressive symptoms (Mirsu-Paun & Oliver, 2017). It individuals consider rejecting others, and when they actually is also evident that rejecting is anxiety-provoking, especially do so. within the context of a relationship (Rizvi et al., 2021). Fur- thermore, although evidence is mixed about whether recip- Low Self-Esteem ients or initiators experience more distress and internalizing Self-esteem is a central component of self-schema and is symptoms (e.g., Boelen & Reijntjes, 2009; Davis et al., 2003; directly linked to social rejection. According to the Sociome- Field et al., 2009; Kansky & Allen, 2018; Tashiro & Frazier, ter Theory, self-esteem functions as a monitor of the degree 2003), it is clear that both depressive and anxiety symptoms to which others value an individual: high self-esteem signals are associated with the experience of romantic rejection (e.g., interpersonal acceptance, and low self-esteem signals social Field et al., 2009; Kansky & Allen, 2018). Current mood exclusion (Baumeister & Leary, 1995; Leary et al., 1995). or anxiety may shape whether and how an individual rejects Thus, being rejected is likely to reduce one’s self-esteem. Al- a partner or friend, as well as their experience of being re- though there is some controversy surrounding the degree to jected. which social rejection paradigms affect self-esteem (Black- Additionally, the general feelings of negative affect and hart et al., 2009; Gerber & Wheeler, 2009), research that has diffuse symptoms of mood and anxiety are markers of disentangled the effects of different paradigms provides ev- what can sometimes be rephrased as “general distress”: idence that specific episodes of rejection do in fact threaten the commonly-experienced nonspecific symptoms of depres- self-esteem (Bernstein & Claypool, 2012). sion, anxiety, and other disorders (Clark & Watson, 1991; Research on the relationship between engaging in rejec- D. Watson et al., 1995). Such experiences of general dis- tion and self-esteem has produced somewhat mixed findings. tress may be compounded by other factors such as low self- For example, in one daily diary study, when individuals de- esteem. Based on this research, we hypothesized that higher scribed experiences of ostracizing others, they reported feel- symptoms of anxiety, depression, and general distress are ing higher self-esteem (Nezlek et al., 2015). However, in likely to be associated with increased difficulty of rejecting a recalled experiences study, the relationship between self- someone else and with more negative emotions experienced esteem and ostracism varied depending on the reason for os- during such a rejection. tracizing, with people higher in self-esteem being more likely to use ostracism to end a relationship, and people lower in Rejection Sensitivity self-esteem being more likely to use ostracism to avoid get- Like self-esteem, rejection sensitivity is closely connected ting rejected themselves (Sommer et al., 2001). Furthermore, to experiences of social rejection. Rejection sensitivity is the in a laboratory experiment involving instructed ostracism, tendency to anxiously expect rejection and to react partic- participants felt worse about themselves when engaging in ularly negatively to rejection (Downey & Feldman, 1996). ostracism compared to inclusion, but they also felt more su- Thus, individuals who experience higher levels of rejection perior to others in that same context (Zadro et al., 2005). sensitivity tend to have stronger emotional and interpersonal Taken together, social rejection and self-esteem are clearly reactions to being rejected. For example, rejection sensitive linked, yet there is some ambiguity regarding the direction individuals are more likely to respond to rejection with anger, of the relationship. However, based on Sociometer Theory hostility, and aggression (e.g., Ayduk et al., 1999; Downey et and the finding that people may feel worse about themselves al., 2000; Romero-Canyas et al., 2010). Furthermore, rejec- after rejecting (Zadro et al., 2005), we hypothesized that in- tion sensitivity has been linked to several other constructs as- dividuals with lower self-esteem would view engaging in re- sociated with negative self-schema including anxiety symp- jection as more difficult and more likely to induce negative toms, depressive symptoms, and low self-esteem (e.g., Ay- emotions, and they would indicate less inclination to reject. duk et al., 2001; Downey & Feldman, 1996; P. Gilbert et al., 2006; London et al., 2007; J. Watson & Nesdale, 2012; Symptoms of Depression and Anxiety Zhou et al., 2018). For instance, women with higher levels Theories of self-schema suggest that individuals process of rejection sensitivity experience higher levels of depressive incoming information or situations through the lens of their symptoms after they are broken up with (Ayduk et al., 2001). schema, which impacts symptoms of anxiety and depression. Similarly, rejection sensitivity has been shown to mediate the SELF-SCHEMAS AND SOCIAL REJECTION 3 relationship between low self-esteem and depressive symp- et al., 2013; Williams et al., 2000). When individuals pre- toms in an adolescent sample (Zhou et al., 2018). dict their willingness to reject, they do not always consider Yet, much of the research on rejection sensitivity has fo- the emotional consequences for themselves and for the tar- cused on expectations toward and reactions to being rejected, get of rejection in the same way that they do when actually and less is known about the role of rejection sensitivity in engaging in rejection (Joel et al., 2014). Thus, the potential expectations toward and reactions to engaging in rejection. disconnect between forecasting the difficulty and emotions Given the links between rejection sensitivity, mood symp- associated with rejecting compared to the true level of diffi- toms, and self-esteem, it is possible that rejection sensitive culty and emotion is potentially important for understanding individuals may forecast feeling worse about engaging in when and how individuals choose to reject others. rejection and may also experience engaging in rejection as more difficult and emotionally negative. However, in a study Present Research on women’s experiences of romantic rejection, rejection sen- sitivity was only associated with depressive symptoms for In the present research, we test whether individuals with those who experienced a partner-initiated breakup. In other more negative views of themselves and of the world around words, rejection sensitivity did not predict negative emotions them 1) are likely to predict rejection would be difficult and following rejection for those who engaged in rejection (Ay- unpleasant to accomplish and that they would be unwilling duk et al., 2001). Thus, in the present research, we specif- to do so, and 2) will recall experiences of engaging in re- ically examine the role of rejection sensitivity in both fore- jection as difficult and emotionally negative. These hypothe- casting and experiencing the act of engaging in rejection. ses are tested in three studies utilizing two different meth- ods. The first set of studies (Studies 1a and 1b) use vignette Forecasting paradigms to assess forecasts about likelihood of rejecting, method of hypothetical rejection, perceived difficulty, and To understand the role of self-schema and its correlates perceptions about negative emotions that would be felt after in social rejection, it is important to consider both the actual rejecting. The final study uses a recall paradigm in which experience of rejecting as well as expectations about that ex- participants who had engaged in social rejection within the perience. Affective forecasting is the idea that individuals try last two to three weeks wrote about their rejection experience to predict how they will feel in the future (Wilson & Gilbert, and then rated the difficulty and negative emotions they ex- 2003)—that is, that one can read a vignette or description of perienced. In all studies, we measure self-esteem, rejection an experience and place oneself into that situation, imagin- sensitivity, anxiety symptoms, and depressive symptoms. We ing how it might feel. Such forecasting tends to be guided by report all manipulations, measures, and exclusions in these one’s self-schema (Liberman et al., 2002), suggesting that it studies. The hypotheses, materials, analyses, and stopping may respond to mood states, and be most accurate when a plans for all studies were pre-registered on the Open Science respondent has had similar past experiences. However, fore- Framework. We have reported all pre-registered analyses in casting is not always accurate. Early between-participants the main body of the manuscript and supplemental materi- studies of forecasting about rejections showed that people als, and clearly marked any deviations from the preregistered believe a breakup will make them feel worse than it actu- analysis plan in the main body of the paper. ally does (D. T. Gilbert et al., 1998), and similar longitudi- nal studies show that this is in fact the case (Eastwick et al., 2008). Study 1 The experience of engaging in rejection appears to elicit similar forecasting discrepancies. For example, when would- For Studies 1a and 1b, we measured the associations be- be rejectors imagine how likely they would be to engage in tween constructs connected to negative self-schemas (i.e., romantic rejection, they express a higher likelihood of re- depressive and anxiety symptoms, self-esteem, rejection sen- jecting than when they are faced with the real potential rejec- sitivity) and rejection decision forecasts. In Study 1a, we tion situation (Joel et al., 2014). Similarly, individuals may studied both friendships and romantic relationships in a sam- not accurately predict the negative emotional consequences ple of undergraduate students; in Study 1b, we used an on- of engaging in rejection. Although being rejected is clearly line sample of similar age but focused solely on romantic perceived as emotionally negative and difficult, would-be- relationships. rejectors also face a negative interpersonal process. For ex- ample, when individuals engage in rejection, they report that Study 1a it is difficult and emotionally taxing, and they often do not know the best way to approach the situation (e.g., Baumeis- Study 1a focused on self-reported willingness to engage ter et al., 1993; Ciarocco et al., 2001; Freedman et al., 2016; in two different types of social rejection. The study was ap- Legate et al., 2013; Poulsen & Kashy, 2012; Wesselmann proved by the authors’ Institutional Review Boards. 4 FREEDMAN & DAINER-BEST

Method my part to end it”; α = .90), negative emotion (“I would feel...” distressed, upset, guilty, ashamed; α = .92), and pos- Participants. Two hundred and ninety-two participants itive emotion (“I would feel?” proud, strong, determined, were recruited from two college psychology participant relieved; α = .94). pools from February to December of 2019. Participants re- To assess past experiences with rejection, participants ceived course credit for participating in this study. We con- were asked about their number of previous romantic rela- ducted a power analysis in R for a small correlation (r = .2) tionships. Romantic relationship was defined liberally: “an with 95% power and an alpha of .05, which indicated a target intimate relationship with an individual that lasts for more sample size of 194. We decided to oversample to account for than two dates.” They were then asked how many of those exclusions. The pre-registered stopping rule was that data relationships they had ended, how many the other person collection would end after collecting data from 300 partici- had ended, and how many were ended mutually. Percent of pants or by the end of December 2019; it ended at the end relationships ended by the participant was calculated as the of December with data collected from 292 participants. The number the participant ended divided by the number of total final sample consisted of 214 participants (see Table1 for relationships. demographics and exclusions). A sensitivity analysis per- To measure anxiety and depression, participants com- formed in R for a correlation with a two-tailed test and an pleted the short-form of the MASQ (Clark & Watson, 1991; alpha of .05 indicated that our sample provides 95% power Wardenaar et al., 2010), a 30-item inventory with three sub- to detect a correlation of r = .24 and 80% power to detect a scales: General Distress (α = .91), Anxious Arousal (α = correlation of r = .19. .87), and Anhedonic Depression (α = .90). The General Dis- Procedure. After consenting to participate, participants tress subscale is thought of as measuring negative affect and were directed to a survey presented on the Qualtrics platform. incorporating both depressive and anxiety symptoms (Clark They read a set of 10 brief vignettes describing a friendship & Watson, 1991). or romantic situation in which there were possible grounds To measure self-esteem, participants completed the RSE for ending the relationship (see Materials on OSF). For each (Rosenberg, 1965), which consists of ten items (e.g., “On the situation, participants were asked a set of questions about whole, I am satisfied with myself”) rated on a 1 (strongly engaging in rejection in that situation. The vignettes were disagree) to 4 (strongly agree) scale (α = .91) with higher presented in a randomized order. After completing the vi- numbers indicating higher levels of self-esteem. gnette measures, participants completed a 30-item version of Rejection sensitivity was included as an exploratory mea- the Mood and Anxiety Symptom Questionnaire (MASQ) the sure in Study 1a. For the RSQ-8 (Downey & Feldman, 1996), Rosenberg Self-Esteem Scale (RSE), and the 8-item Rejec- participants read eight scenarios and for each scenario indi- tion Sensitivity Questionnaire (RSQ) in a randomized order cated how likely they would be to experience acceptance on (see Supplemental Table A1 for intercorrelations). Finally, a 1 (very unlikely) to 6 (very likely) scale and how concerned participants completed questions about past experiences with they would be about being rejected on a 1 (very unconcerned) rejection, demographic questions, an attention check, and to 6 (very concerned)1 scale (α = .78). The overall score were asked if we should use their data. reflects anxious expectation and concern about rejection. Measures. For the vignette responses, participants were Analyses. Analyses were conducted using a pre- first asked how likely they would be to end the friendship registered alpha criterion of p = .05 and were performed or relationship on a 1 (extremely unlikely) to 5 (extremely in SPSS and R (R Core Team, 2021). likely) scale (see Table2 for means and standard deviations). Next, they were told to imagine that they had decided to end Results the friendship or romantic relationship and were asked how General Distress and Rejection. As predicted, there likely they would be to use explicit rejection (“How likely was a positive correlation between general distress and fore- would you be to explicitly reject the person [i.e., tell them casted difficulty of engaging in rejection: r(212) = .317, that the friendship is over]?”) and passive rejection (“How p < .001. Contrary to predictions, there was no association likely would you be to passively reject the person [i.e., ghost between general distress and the overall likelihood of reject- them, ignore them until they get the message]?”), in a ran- ing in the vignettes, r(207) = .017, p = .808. An exploratory domized order. Finally, they were told, “Imagine you de- linear mixed effects model showed, however, that there was cided to end the relationship. Regardless of whether you’ve an interaction between relationship type and general distress done so passively or explicitly, indicate how much you agree in the prediction of likelihood of rejecting, t(206) = −3.669, with the following statements about ending the relationship” and were asked to rate a set of ten statements on a 1 (strongly 1The labeling on the concern scale was inadvertently reversed disagree) to 7 (strongly agree) scale. The ten statements as- in the survey such that 1 corresponded to “very concerned” and 6 sessed difficulty of engaging in rejection (“It would be dif- corresponded to “very unconcerned.” Therefore, the concern items ficult for me to end it” and “It would take a lot of effort on were reverse-scored before the composite was calculated. SELF-SCHEMAS AND SOCIAL REJECTION 5

Table 1

Demographics and Exclusions for all studies. Parts (a) and (b) show sample demographics; part (c) shows participants dropped due to exclusion criteria. Gender (a) n Age, M (SD) No. Womena (%) No. Mena (%) Non-binary gender responses Study 1a 292 19.27 (1.65) 146 (50%) 63 (21%) 5: 2 gender fluid, 2 non-binary, 1 demigirl Study 1b 264 21.36 (1.99) 125 (47%) 133 (50%) 6: 3 non-binary, 2 agender, 1 genderqueer Study 2 259 21.55 (2.02) 137 (53%) 88 (34%) 14: 13 non-binary, 1 agender

Race/Ethnicity (b) White Black Asian/Asian Hispanic/ American Indian Multiracial Another American Latinx or Alaska Native race/ethnicity Study 1a 162 (75%) 15 (7%) 12 (6%) 8 (4%) 2 (1%) 11 (5%) 4 (2%) Study 1b 153 (58%) 26 (10%) 32 (12%) 27 (10%) 2 (1%) 23 (9%) 1 Study 2 147 (57%) 17 (7%) 31 (12%) 26 (10%) 3 (1%) 15 (6%) 0

Exclusion criteriab (c) Recruited Failed attention Incomplete Left vignettes Had not No description Asked not to Final check MASQ blank rejected of rejection use data sample Study 1a 292 62 34 24 N/AN/A 38 214 Study 1b 309 25 21 N/AN/AN/A 17 264 Study 2 329 40 18 N/A 16 2 12 259 Note. aParticipants reported free response gender; cis- and transgender respondents are grouped together. bSome participants were excluded due to multiple criteria.

Table 2 ended by the participant, r(156) = .127, p = .110. A lin- ear mixed effects model with predictors of general distress Means and standard deviations for overall likelihood of and whether rejection was explicit or passive found no effect engaging in rejection, likelihood of explicit rejection, and of general distress on method of rejection: t(211) = 0.784, likelihood of ghosting in response to the vignettes in Study 1a β = 0.006, p = .434. As predicted in the secondary hypothe- and Study 1b. ses, general distress was positively correlated with negative Study 1a Study 1a Study 1b emotions in response to thinking about rejecting in the hypo- Friendship Romantic Romantic thetical scenarios: r(212) = .391, p < .001. Anxiety and Rejection. Contrary to hypotheses, anxi- Overall likelihood 3.44 3.76 3.82 ety was not associated with forecasted difficulty of engag- of rejecting (0.65) (0.63) (0.64) ing in rejection (r[212] = .096, p = .161), likelihood of re- Explicit rejection 3.21 3.98 4.17 jecting in rejection vignettes (r[207] = .015, p = .834), or (0.86) (0.66) (0.65) the percent of romantic relationships ended by the partici- Ghosting 3.36 2.48 2.34 pant (r[156] = .094, p = .239). The hypothesis that indi- (0.97) (1.03) (0.97) viduals with more anxiety symptoms would be more likely Note. Responses were on a 1 (extremely unlikely) to 5 (extremely to engage in ghosting than explicit rejection was likewise likely) Likert scale. not supported: t(211) = 0.968, β = 0.008, p = .334. As predicted in secondary hypotheses, anxiety symptoms were positively correlated with negative emotions in re- β = −0.020, p < .001, such that, for friendships, general sponse to thinking about rejecting in the hypothetical sce- distress was positively correlated with increased likelihood narios: r(212) = .249, p < .001. rejecting but for romantic relationships it was negatively cor- Self-esteem and Rejection. Self-esteem was negatively related (see Table3 for correlations). correlated with forecasted difficulty of engaging in rejection Contrary to predictions, there was no association between as predicted: r(209) = −.325, p < .001. Similarly, self- general distress and the percent of romantic relationships esteem was negatively correlated with forecasting negative 6 FREEDMAN & DAINER-BEST emotions in response to thinking about hypothetical rejec- mantic vignettes were modified to describe a shorter period tion: r(209) = −.399, p < .001. However, contrary to hy- of dating. Furthermore, as it is possible that the term “ghost- potheses, self-esteem was not associated with likelihood of ing” has negative connotations (Freedman et al., 2019) and engaging in rejection in the vignettes: r(204) = .049, p = that participants were reluctant to admit they would consider .485. Self-esteem was also not associated with percent of engaging in ghosting, the expression “ghost them” was re- relationships participants had ended: r(155) = −.145, p = moved from the survey. .073. Anhedonic Depression. Analyses examined the rela- Method tionship between the Anhedonic Depression subscale and the rejection variables. Because this subscale was unintention- Participants. Participants between the ages of 18 and ally excluded from the pre-registration, it might be consid- 24 who were living in the United States and spoke English ered exploratory; we report the analyses as parallel to the as their first language were recruited via Prolific. We aimed other sub-scales of the MASQ. There was a positive correla- to collect data from a similar number of participants as in tion between anhedonic depressive symptoms and forecasted Study 1a. Thus, the pre-registered stopping rule for data col- difficulty of engaging in rejection: r(212) = .166, p = .015. lection was when 300 participants completed the study. Data However, there was no association between anhedonic de- were collected on February 24 and 25, 2020. The final sam- pressive symptoms and the likelihood of rejecting in the vi- ple had 264 participants (see Table1 for demographics and gnettes, r(207) = −.035, p = .614, or the percent of romantic exclusions). Participants were paid $2.17 for their participa- relationships ended by the participant, r(156) = .053, p = tion, which took a median of 14 minutes and 15 seconds. A .510. Similarly, a linear mixed effects model with predic- sensitivity analysis performed in R for a correlation with a tors of anhedonic depressive symptoms and whether rejec- two-tailed test and an alpha of .005 indicated that our sample tion was explicit or passive found no effect of anhedonic de- provides 95% power to detect a correlation of r = .27 and pressive symptoms on method of rejection: t(211) = 0.624, 80% power to detect a correlation of r = .22. β = 0.006, p = .533. Finally, anhedonic depressive symp- Procedure. After consenting to participate, participants toms were positively correlated with negative emotions in read the five romantic vignettes from Study 1a with the modi- response to thinking about rejecting in the hypothetical sce- fication that all relationships were described as lasting a “few narios: r(212) = .271, p < .001. weeks” instead of a “few months.” For each vignette, partic- Exploratory analyses. A set of analyses was conducted ipants were asked a set of questions about engaging in re- to examine the relationship between rejection sensitivity and jection in that situation. The vignettes were presented in a forecasts about engaging in rejection. Rejection sensitivity randomized order. After completing the vignette measures, was significantly positively correlated with forecasted diffi- participants completed identical questionnaires to Study 1a culty of engaging in rejection, r(210) = .240, p < .001, and (see Supplemental Table A2 for intercorrelations). Analy- forecasted negative emotions in response to thinking about ses were conducted using a preregistered alpha criterion of hypothetical rejection, r(210) = .202, p = .003. Rejection p = .005, chosen to highlight findings from Study 1a that sensitivity was not associated with likelihood of engaging in replicated unequivocally. rejection: r(205) = −.017, p = .811. A set of exploratory analyses was conducted to examine Measures. For the vignette responses, participants were whether the correlations between depression, anxiety, self- first asked how likely they would be to end the relationship, esteem and the vignette responses differed based on type of as in Study 1a. Next, they were told to imagine that they relationship scenario (i.e., friendship versus romantic). Over- had decided to end the relationship and were asked how all, the patterns were similar across friendship and romantic likely they would be to use explicit rejection and passive scenarios (see Table3) with somewhat stronger correlations rejection/ghosting, in a randomized order. (Exact wording for romantic scenarios. However, there were weak and op- of questions is available on OSF.) To allow for an examina- posing correlations between general distress and likelihood tion of how participants would feel about each rejection op- of engaging in rejection for romantic versus friend relation- tion (i.e., explicit rejection and ghosting), after each of those ships such that negative affect was weakly, positively cor- questions, participants were instructed to imagine they had related with likelihood of ending a friendship, but weakly, ended it in that way and were asked to rate the ten state- negatively correlated with likelihood of ending a romantic ments from Study 1a. The ten statements assessed difficulty relationship (see Table3). of engaging in rejection (difficulty α = .90; negative emotion α = .93; positive emotion α = .91). Participants also com- Study 1b pleted the MASQ (AD α = .92, AA α = .87, GD α = .93), the RSE (α = .93), the RSQ (α = .77), and the same ques- Study 1b was designed as a replication of Study 1a with tions from Study 1a about past experiences ending relation- only romantic relationship vignettes. Additionally, the ro- ships. SELF-SCHEMAS AND SOCIAL REJECTION 7

Table 3

Correlations between depressive symptoms, anxiety symptoms, self-esteem, rejection sensitivity and negative emotion and difficulty in Studies 1a, 1b, and 2. Study 1a Study 1b Study 2 Friend Romantic Explicit Rejection Ghosting All experiences DifficultyNegative DifficultyNegative DifficultyNegative DifficultyNegative DifficultyNegative Emotion Emotion Emotion Emotion Emotion Anhedonic depression .295** .179* .413** .117 .111 .018 .032 .083 -.04 .14* Anxious arousal .040 .181* .131 .271** .116 .168* .151* .105 .15* .30** General distress .222* .295** .341** .413** .302** .271** .242** .236** .20** .40** Self-esteem -.279** -.331** -.296** -.391** -.304** -.227** -.190** -.214** .00 -.22** Rejection sensitivity .236** .177* .189* .190* .245** .157* .186* .170* .04 .22** Note. ** p < .005, * p < .05

Results symptoms were also not associated with the percent of rela- tionships that participants had ended: r(195) = −.022, p = As predicted, increased general distress and lower self- .760. Finally, there was no statistical support for the hypoth- esteem were associated with forecasting higher levels of dif- esis that individuals with more anhedonic depressive symp- ficulty of engaging in rejection for both explicit rejection toms were more likely to engage in ghosting than explicit and ghosting (see Table3). In addition, as predicted, in- rejection, t(262) = −1.07, β = −0.009, p = .283. creased general distress and lower self-esteem were asso- As in Study 1a, rejection sensitivity was significantly pos- ciated with the forecast that engaging in rejection would itively associated with perceived difficulty of ending rela- lead to more negative emotions for both explicit rejection tionships via explicit rejection, r(262) = .245, p < .001, and ghosting (see Table3). Contrary to predictions, anxi- and ghosting, r(262) = .186, p = .002. Based on the pre- ety symptoms were not significantly associated at the pre- registered criterion level of p < .005, rejection sensitivity registered criterion level with forecasting negative emotions was not significantly correlated with overall likelihood of for explicit rejection (r[262] = .168, p = .006) or ghost- ending the relationships, r(262) = −.169, p = .006. Nor was ing (r[262] = .105, p = .089). The hypothesis that indi- it significantly correlated with negative emotions in response viduals with more anxiety symptoms would be more likely to explicitly rejecting, r(262) = .157, p = .011, or ghosting, to engage in ghosting than explicit rejection was not sig- r(262) = .170, p = .006. Scores on the RSQ were also not nificant at the pre-registered criterion level, t(262) = 2.52, associated with the percent of relationships that participants β = 0.02, p = .012. There was a main effect of preference to had ended, r(195) = −.096, p = .179. explicitly reject, t(262) = −11.95, β = −2.27, p < .001. As in Study 1a, general distress, anxiety symptoms, and Discussion self-esteem were not significantly correlated with the percent of relationships that participants had ended (general distress: Studies 1a and 1b found that general distress and lower r[195] = −.017, p = .813; anxiety symptoms: r[195] = self-esteem were associated with forecasts about the diffi- −.047, p = .511; self-esteem: r[193] = .058, p = .423), culty of rejecting and the negative emotions that would re- the reported likelihood of ending the relationships described sult from rejecting. Specifically, general distress, lower self- in the vignettes (general distress: r[262] = −.034, p = .583; esteem, and rejection sensitivity were associated with fore- anxiety symptoms: r[262] = −.026, p = .678; self-esteem: casts of greater difficulty and more negative emotion. Re- r[259] = .066, p = .289), or the reported likelihood of end- jection sensitivity was also associated with forecasting rejec- ing the relationships described in the vignettes via explicit tion to be more difficulty and negatively emotional, though rejection or ghosting (see Table3). the correlations did not reach the preregistered alpha crite- Exploratory Analyses. Again, the Anhedonic Depres- rion value in Study 1b. Anxious arousal was not consistently sion subscale from the MASQ was not included in the pre- associated with difficulty forecasts but was associated with registration; we report the analyses here. Contrary to the forecasting more negative emotions after rejecting in both general prediction, anhedonic depressive symptoms were not studies, but in Study 1b this did not meet the preregistered significantly associated with forecasting of negative emo- criterion. tions for explicit rejection (r[262] = .018, p = .777) or There were not, however, significant differences in how ghosting (r[262] = .083, p = .180). Anhedonic depressive these correlates of negative self-schema related to predicted 8 FREEDMAN & DAINER-BEST willingness to engage in different types of rejection (i.e., ex- Procedure plicit rejection vs. ghosting). There was some indication that higher levels of anxiety might be associated with an unwill- Prescreening. After consenting to participate in the pre- ingness to explicitly reject in Study 1b, but the interaction screening survey, participants were asked to check off the did not meet the pre-registered criterion level of p < .005. things they had done within the last two weeks from a list We therefore chose to investigate this possible relationship of five items that included the rejection question (see Mea- in Study 2. sures) and four filler items. Then, participants completed three items from the Adult Rejection Sensitivity Question- naire (ARSQ; Berenson et al., 2009) and were debriefed. The Study 2 ARSQ was used in this study as the questions are more ap- In Study 2, we broadened the scope of how we defined plicable to a broader sample than the RSQ. rejection and used a different methodology to assess feelings Main Survey. Participants who indicated that they had about the emotional consequences and difficulty of rejection rejected someone in the previous two weeks were sent an episodes. Although Studies 1a and 1b provide converging ev- invitation to participate in the main survey. After providing idence about the roles of depressive symptoms, self-esteem, consent, they read the following definition of social rejection: and (to a lesser degree) rejection sensitivity in how likely in- [Social rejection is] a time when you told a dividuals are to perceive the process of engaging in rejection, friend, romantic partner, colleague, acquain- both studies relied on vignette methodology. These vignettes tance, or family member that you did not want in Study 1 provided useful forecasts; however, in Study 2 to engage in a social interaction with them (e.g., we focused on how people report having felt in real situa- turning down a date, saying no to a lunch re- tions where they rejected someone else—and how this expe- quest, telling someone you didn’t want to come rience connects to the consequences and components of self- over to their place) OR a time when you ig- schema discussed in Study 1. We thus used a recall paradigm nored someone’s request for an interaction (e.g., to assess how depression, anxiety, self-esteem, and rejection ghosting, not responding to text messages) OR sensitivity relate to a previous, retrospectively reported re- a time when you ended a relationship (e.g., a ro- jection experience. We predicted similar patterns to those mantic breakup, ending a friendship). observed in the forecasts of individuals described in Study 1. After reading the definition, participants were asked when Method they had last socially rejected someone and if they answered that it had occurred within the last three weeks, they were Participants directed to a set of questions about the experience and then a set of questionnaires (see Supplemental Table A3 for in- Participants between the ages of 18 and 24 who were liv- tercorrelations): the MASQ, the RSE, the ARSQ, attention ing in the United States and spoke English as their first lan- check, data usage, and demographics. If the participants had guage were recruited via Prolific Academic. To determine not rejected anyone in the last three weeks, they were sent eligibility for the main study, a prescreening questionnaire directly to the questionnaires, and were compensated but not was administered. The eligibility criterion was that the par- included in analyses. Three weeks was used to allow for time ticipants had socially rejected someone in the last two weeks. that had elapsed since participants completed the prescreen- Participants were paid $0.30 for their participation in the pre- ing. (However, only 14 participants responded about a rejec- screening survey, which took a median of 78 seconds. For the tion that had occurred more than 2 weeks prior to completing prescreening questionnaire, 1472 participated and 400 indi- the survey.) cated that they had rejected someone in the last two weeks. Those 400 participants were then invited to participate in Measures the main survey and 329 elected to participate. Based on a power analysis conducted in R with 95% power, p = .05, and Prescreening. Respondents were asked to indicate if r = .222 (Study 1b, correlation between MASQ-GD and per- they had “done each of the following within the past two ceived difficulty in rejecting a friend), data collection contin- weeks” for a series of five items presented in randomized ued until 258 participants completed the study and qualified order, four of which were filler items. The filler items in- according to pre-registered exclusion criteria. The final sam- cluded questions such as, “Have you engaged in your favorite ple had 259 participants (see Table1 for demographics and hobby? (e.g., have you gardened, read a book, played a sport, exclusions). Participants were paid $1.60 for their participa- or done yoga?)”, going out to eat, going on a date, and argu- tion in the main survey, which took a median of 6 minutes 54 ing with someone. The item of interest asked “Have you seconds. Data were collected between June 4, 2021 and June told a friend or romantic interest that you cannot do some- 27, 2021. thing together, ignored them when they tried to talk to you, SELF-SCHEMAS AND SOCIAL REJECTION 9 or indicated that you wanted to end the relationship? (i.e., emotions to having rejected someone. Although not pre- have you rejected anyone?).” Participants who said that they registered due to the results from Studies 1a and 1b, anxiety had done so were invited to participate in the main survey. was also negatively associated with how difficult participants (Additionally, participants responded to questions 5, 7, and 9 found it to reject others, r(256) = .15, 95% CI [.03, .27], from the Adult RSQ.) p = .013 (see the top of Figure1). Main Survey. Following their response to the rejection A logistic regression did not find a significant effect of item described in Procedures, above, participants were asked anxiety symptoms predicting the method participants had about their method of rejection—whether it had been passive used to reject, β = −0.02, 95% CI [-0.06, 0.01], z = (“I used an indirect approach [e.g., ghosting, ignoring, not −1.47, p = .143. responding]”), active (“I used a direct approach [e.g., telling them face-to-face, via text, via email, via a phone call]”), or Self-esteem and Rejection Sensitivity a different approach. Participants were also asked to spend at Contrary to predictions, self-esteem was not significantly least one minute writing about how the rejection occurred. correlated with difficulty rejecting, r(252) = −.003, 95% CI Following this question, participants rated the ten state- [-.13, .12], p = .959. Self-esteem was, however, negatively ments from Study 1. These assessed difficulty of engaging correlated with negative emotions experienced while reject- in rejection (difficulty α = .87; negative emotion α = .87; ing, r(252) = −.22, 95% CI [-.34, -.10], p < .001. positive emotion α = .76). Participants also completed the Contrary to predictions, rejection sensitivity was not sig- MASQ (AD α = .92, AA α = .88, GD α = .93), the RSE nificantly correlated with difficulty rejecting, r(252) = .04, (α = .93), the ARSQ (α = .81), and provided demographics. 95% CI [-.08, .16], p = .539. Rejection sensitivity was, The MASQ was modified to ask about the past two weeks however, positively correlated with negative emotions expe- rather than one. rienced while rejecting, r(252) = .22, 95% CI [.10, .34], p < .001. A logistic regression did not find a signifi- Results cant effect of adult RSQ score predicting the method par- ticipants had used to reject, β = −0.04, 95% CI [-0.09, 0.00], General Distress and Rejection z = −1.82, p = .069. As predicted, general distress was positively correlated with difficulty experienced rejecting others, r(256) = .20, Discussion 95% CI [.08, .32], p = .001; increased scores on the MASQ- In these data, individuals with elevated levels of gen- GD were associated with increased difficulty rejecting. Gen- eral distress and anxious arousal reported increased difficulty eral distress was also positively correlated with negative choosing to reject another person, as well as increased neg- emotions experienced while rejecting, r(256) = .40, 95% CI ative emotions after doing so. The other measurements re- [.29, .50], p < .001; increased scores on the MASQ-GD were lating to negative self-schema (depressive symptoms, self- associated with more negative emotions about the experience esteem, and rejection sensitivity) were also associated with (see the lower portion of Figure1). negative emotions experienced after rejecting, but were not Anhedonic Depression and Rejection significantly associated with difficulty doing so. In this sam- ple, there was no indication that any of these symptoms mod- Anhedonic depression was not correlated with difficulty erated the method of rejection individuals used, be it passive rejecting others, r(256) = −.04, 95% CI [-.17, .08], p = .479. (ghosting) or active (direct). However, the types of rejec- It was, however, positively correlated with negative emotions tions described by participants were highly varied (includ- experienced while rejecting, r(256) = .14, 95% CI [.01, ing, e.g., a breakup, or ignoring a text), and it remains possi- .25], p = .029, such that increased scores on MASQ-AD ble that negative self-schema may moderate rejection method were associated with more negative emotions about having for only more serious rejections. Regardless, the results from rejected someone. A logistic regression did not find a sig- Study 2 demonstrate that mood and anxiety symptoms, as nificant effect of MASQ-AD predicting the method partici- well as low self-esteem and heightened rejection sensitivity, pants had used to reject, β = −0.02, 95% CI [-0.05, 0.01], may increase the negative experience of having to turn some- z = −1.15, p = .249. one down or otherwise reject them.

Anxiety and Rejection General Discussion As predicted, anxiety symptoms on the MASQ were pos- Across three studies, the present research shows that con- itively correlated with the negative emotions experienced structs associated with negative self-schemas are associated while rejecting, r(256) = .30, 95% CI [.19, .41], p < with forecasts and actual experiences of engaging in rejec- .001—increased anxiety was associated with more negative tion. First, depressive symptoms, including those of gen- 10 FREEDMAN & DAINER-BEST

Figure 1

Anxious arousal (top) and general distress (bottom) are associated with difficulty (left) and negative emotions (right) of rejecting in Study 2.

Note. Points show individual responses with a .3/.1 x/y jitter; blue trend lines show the linear trend, with 95% confidence intervals in gray. eral distress, are associated with predictions that engaging tion were associated with self-esteem and rejection sensitiv- in rejection will be more difficult and will lead to more neg- ity. In other words, although the forecasts of how difficult ative emotions. Furthermore, these predictions were accu- rejection would be were associated with self-esteem and re- rate: in Study 2, individuals with higher levels of depressive jection sensitivity, the difficulty of the actual experience of symptoms were more likely to recount their most recent ex- rejection difficulty was not. Prior research on rejection sensi- perience of engaging in rejection as difficult and emotionally tivity has focused on how individuals with differing levels of negative. This result is in line with what participants in Study rejection sensitivity respond to being rejected (e.g., Downey 1 predicted. et al., 1998; London et al., 2007; Mendoza-Denton et al., Second, anxious symptoms were associated with forecasts 2002), but less research has considered the role of rejection of negative emotion, though not consistently at the preregis- sensitivity on engaging in rejection. The present research be- tered criterion levels below .05 in Study 1b. However, anx- gins to address this gap and provides evidence that rejection ious symptoms were significantly associated with both per- sensitivity is important for understanding both sides of the ceived difficulty and negative emotions for the recalled rejec- rejection process. tion experience. This suggests that while anxiety may in fact Taken together, the results indicate that negative self- play a role in whether a rejection is carried out, participants schemas, and the constructs that stem out of them, may play are not necessarily able to predict this when forecasting. an important role in both how individuals forecast rejection Finally, self-esteem is negatively associated—and rejec- experiences as well as the actual experiences of rejection. tion sensitivity is positively associated—with forecasting re- They also have clinical implications: they provide evidence jection to be difficult and emotionally negative. However, that young adults experiencing elevated depressive and anx- when participants recounted their most recent rejection ex- iety symptoms may be less willing to consider engaging in perience, only the negative emotional consequences of rejec- social rejection, as they believe that doing so will increase SELF-SCHEMAS AND SOCIAL REJECTION 11 their negative emotions. And, in the retrospective report of ships. Across the three studies in the present research, there other young adults, it appears that these beliefs are correct: were a number of different contexts. In Study 1a, the vi- rejection does increase their negative emotions. This raises gnettes depicted several reasons that one might engage in so- the possibility that such internalizing symptoms may have cial rejection within both friendships and romantic relation- measurable external consequences. For example, heightened ships. In Study 1b, the vignettes were only within the con- anxiety may lead a young adult to not end a friendship that is text of short-term romantic relationships. Finally, in Study causing harm because of the forecasted emotional repercus- 2, individuals wrote about a variety of experiences ranging sions of doing so, when in fact ending that friendship might from not responding to a text to a breakup. Thus, it will be be a psychologically healthy decision. important in future research to examine how negative self- These studies also revealed two potential forecasting er- schemas operate within different rejection contexts includ- rors when individuals think about how engaging in rejec- ing the type of relationship in which the rejection occurs and tion will feel. First, higher levels of rejection sensitivity and the magnitude of the rejection (e.g., is the rejection ending a lower self-esteem were associated with forecasting rejection relationship or is it turning down an offer to get coffee?). as more difficult and more emotionally negative, yet rejection sensitivity and self-esteem were only associated with nega- Limitations and Future Directions tive emotions and not difficulty when participants recounted A strength of the present research was the replication of their most recent rejection. That is, individuals with higher findings from Study 1a in a generalized (but similarly-aged) levels of rejection sensitivity and lower self-esteem thought sample in Study 1b; further, the lowered alpha criterion in engaging in rejection would be more difficult, but when par- Study 1b highlighted results that replicated unequivocally. ticipants in Study 2 recalled their most recent rejection ex- Additionally, Study 2 provided evidence about how individ- perience, difficulty was not associated with these correlates uals actually feel after having engaged in rejection, allowing of negative self-schema. Second, anxious symptoms were us to consider forecasting errors. However, all three studies not associated with forecasted difficulty, but when partici- relied on non-clinical samples, and findings may be differ- pants recounted their most recent rejection experience, anx- ent in patients diagnosed with anxiety or mood disorders, or ious symptoms were positively associated with how difficult those currently seeking psychological treatment. Findings it felt to reject as well as negative emotional responses. Thus, are based on a short version of the Mood and Anxiety Symp- it seems that negative self-schema and its consequences are toms Questionnaire (Clark & Watson, 1991), which is in- at times associated with forecasts of rejection difficulty and tended for non-clinical samples; nonetheless, the use of such at other times with actual perceived difficulty, but the as- a scale means that we cannot make direct connections to di- sociations between forecasts and experiences do not always agnoses. However, correlations between the 30-item MASQ match. and traditional measures of depression such as the Beck De- These findings build on prior work indicating that per- pression Inventory (BDI-II; Beck et al., 1996) show good forming social rejection is an effortful task (Ciarocco et al., convergent validity (Wardenaar et al., 2010). 2001; Williams et al., 2000) and that initiators of social re- Additionally, there are several constraints on generaliz- jection experience a range of negative emotions (Baumeis- ability. First, the samples in the present research were age- ter et al., 1993; Davis et al., 2003; Kansky & Allen, 2018; restricted. Study 1a was conducted in college participant Legate et al., 2013; Perilloux & Buss, 2008; Poulsen & pools and Studies 1b and 2 were restricted to ages 18 to 24 to Kashy, 2012). Importantly, the present research adds to match the initial sample. Thus, it will be important to repli- this literature by examining the ways in which existing self- cate the current research with participants of other ages to perceptions may change how individuals perceive the pro- understand how negative self-schemas relate to the forecasts cess of social rejection. That is, self-concept, especially the and experiences of engaging in rejection across the lifespan. negative measures described here, was associated with the In addition, all three samples were majority White (1a: 75%, perception that engaging in social rejection is more difficult 1b: 58%; 2: 57%), although the samples for Studies 1b and and more likely to lead to further negative emotion. Fur- 2 were more diverse in terms of race/ethnicity than Study 1a. thermore, these constructs were also associated with actual Finally, all three studies were conducted with participants difficulty of rejecting (for depressive symptoms and anxi- from the United States, and there may be important differ- ety symptoms) and with the experience of negative emotion ences in self-schema and in rejection experiences in other when rejecting (for depressive symptoms, anxiety symptoms, countries. For example, there is some evidence for cross- rejection sensitivity, and self-esteem). It will be important for cultural differences in levels of rejection sensitivity and in future work to longitudinally examine how a given individ- the affective experience of being rejected (e.g., Garris et al., ual’s forecasts of rejection relate to future rejection decisions. 2010; Sato et al., 2014). It is also important to consider the context, as rejections Finally, although Study 2 allowed for an examination of occur across a wide range of social scenarios and relation- how negative self-schemas relate to actual rejection expe- 12 FREEDMAN & DAINER-BEST riences, it used a retrospective method, and memory errors Beck, A. T., & Bredemeier, K. (2016). A unified model and biases may occur when individuals recount even recent of depression. Clinical Psychological Science, experiences. To minimize errors, we specifically recruited 4(4), 596–619. https : // doi . org / 10 . 1177 / individuals who had engaged in rejection within the past two 2167702616628523 (Cited on p.1) weeks, but future research would benefit from longitudinal Beck, A. T., Steer, R. A., & Brown, G. K. (1996). Manual studies with prospective designs that allow for more immedi- for the Beck Depression Inventory-II. San Antonio, ate recounting of rejection experiences, e.g., using ecological TX: Psychological Corporation. (Cited on p. 11). momentary assessment. Future work could also differentiate Berenson, K. R., Gyurak, A., Ayduk, Ö., Downey, G., Gar- between experiences of rejection compared to wanting to re- ner, M. J., Mogg, K., Bradley, B. P., & Pine, D. S. ject, but not doing so. (2009). Rejection sensitivity and disruption of at- tention by social threat cues. Journal of Research in Conclusion Personality, 43(6), 1064–1072. https://doi.org/10. The present research points to the importance of consider- 1016/j.jrp.2009.07.007 (Cited on p.8) ing the roles of negative self-schemas and their consequences Bernstein, M. J., & Claypool, H. M. (2012). Not all social when examining forecasts and experiences of engaging in re- exclusions are created equal: Emotional distress fol- jection. Markers associated with negative self-schema were lowing social exclusion is moderated by exclusion associated with experiencing more negative emotions in re- paradigm. Social Influence, 7(2), 113–130. https:// sponse to engaging in rejection and some were also associ- doi.org/10.1080/15534510.2012.664326 (Cited on ated with forecasting and experiencing rejection as more dif- p.2) ficult. These results suggest that when individuals with psy- Blackhart, G. C., Nelson, B. C., Knowles, M. L., & Baumeis- chological distress are faced with the question of whether to ter, R. F. (2009). Rejection elicits emotional reac- turn a friend or romantic partner down, they anticipate neg- tions but neither causes immediate distress nor low- ative emotions, but believe that they will be able to make ers self-esteem: A meta-analytic review of 192 stud- that rejection. And in situations where individuals do reject, ies on social exclusion. Personality and Social Psy- those same markers of distress are in fact associated with the chology Review, 13(4), 269–309. https://doi.org/10. experience of negative emotions and finding rejection more 1177/1088868309346065 (Cited on p.2) difficult. Boelen, P. A., & Reijntjes, A. (2009). Negative cognitions in emotional problems following romantic relation- References ship break-ups. Stress and Health, 25(1), 11–19. https://doi.org/10.1002/smi.1219 (Cited on p.2) Ayduk, Ö., Downey, G., & Kim, M. (2001). Rejection sensi- Ciarocco, N. J., Sommer, K. L., & Baumeister, R. F. (2001). tivity and depressive symptoms in women. Person- Ostracism and ego depletion: The strains of si- ality and Social Psychology Bulletin, 27(7), 868– lence. Personality and Social Psychology Bulletin, 877. https://doi.org/10.1177/0146167201277009 27(9), 1156–1163. https : // doi . org / 10 . 1177 / (Cited on pp.2,3) 0146167201279008 (Cited on pp.1,3, 11) Ayduk, Ö., Downey, G., Testa, A., Yen, Y., & Shoda, Y. Clark, L. A., & Watson, D. (1991). Tripartite model of anxi- (1999). Does rejection elicit hostility in rejection ety and depression: Psychometric evidence and tax- sensitive women? Social Cognition, 17(2), 245– onomic implications. Journal of Abnormal Psychol- 271. https://doi.org/10.1521/soco.1999.17.2.245 ogy, 100(3), 316–336. https : // doi . org / 10 . 1037 / (Cited on p.2) 0021-843x.100.3.316 (Cited on pp.2,4, 11) Baumeister, R. F., & Leary, M. R. (1995). The need to Dainer-Best, J., Trujillo, L. T., Schnyer, D. M., & Beevers, belong: Desire for interpersonal attachments as a C. G. (2017). Sustained engagement of attention fundamental human motivation. Psychological Bul- is associated with increased negative self-referent letin, 117(3), 497–529. https://doi.org/10.1037/ processing in major depressive disorder. Biological 0033-2909.117.3.497 (Cited on p.2) Psychology, 129, 231–241. https://doi.org/10.1016/ Baumeister, R. F., Wotman, S. R., & Stillwell, A. M. (1993). j.biopsycho.2017.09.005 (Cited on p.1) Unrequited love: On heartbreak, anger, guilt, script- Davis, D., Shaver, P. R., & Vernon, M. L. (2003). Physi- lessness, and humiliation. Journal of Personality cal, emotional, and behavioral reactions to breaking and Social Psychology, 64(3), 377–394. https://doi. up: The roles of gender, age, environmental involve- org/10.1037/0022-3514.64.3.377 (Cited on pp.1, ment, and attachment style. Personality and Social 3, 11) Psychology Bulletin, 29(7), 871–884. https:// doi. Beck, A. T. (1967). Depression: Clinical, experimental, org/10.1177/0146167203252884 (Cited on pp.2, and theoretical aspects. University of Pennsylvania 11) Press. (Cited on p.1). SELF-SCHEMAS AND SOCIAL REJECTION 13

Downey, G., & Feldman, S. I. (1996). Implications of rejec- Gilbert, P., Irons, C., Olsen, K., Gilbert, J., & McEwan, tion sensitivity for intimate relationships. Journal K. (2006). Interpersonal sensitivities: Their links of Personality and Social Psychology, 70(6), 1327– to mood, anger and gender. Psychology and Psy- 1343. https://doi.org/10.1037/0022-3514.70.6.1327 chotherapy: Theory, Research and Practice, 79(1), (Cited on pp.1,2,4) 37–51. https://doi.org/10.1348/147608305x43856 Downey, G., Feldman, S. I., & Ayduk, Ö. (2000). Rejection (Cited on p.2) sensitivity and male violence in romantic relation- Joel, S., Teper, R., & MacDonald, G. (2014). People over- ships. Personal Relationships, 7(1), 45–61. https : estimate their willingness to reject potential roman- //doi.org/10.1111/j.1475- 6811.2000.tb00003.x tic partners by overlooking their concern for other (Cited on p.2) people. Psychological Science, 25(12), 2233–2240. Downey, G., Freitas, A. L., Michaelis, B., & Khouri, H. https://doi.org/10.1177/0956797614552828 (Cited (1998). The self-fulfilling prophecy in close rela- on pp.1,3) tionships: Rejection sensitivity and rejection by ro- Kansky, J., & Allen, J. P. (2018). Making sense and moving mantic partners. Journal of Personality and Social on: The potential for individual and interpersonal Psychology, 75(2), 545–560. https : // doi . org / 10 . growth following emerging adult breakups. Emerg- 1037/0022-3514.75.2.545 (Cited on p. 10) ing Adulthood, 6(3), 172–190. https://doi.org/10. Eastwick, P. W., Finkel, E. J., Krishnamurti, T., & Loewen- 1177/2167696817711766 (Cited on pp.2, 11) stein, G. (2008). Mispredicting distress following Kendler, K. S., Hettema, J. M., Butera, F., Gardner, C. O., romantic breakup: Revealing the time course of the & Prescott, C. A. (2003). Life event dimensions of affective forecasting error. Journal of Experimental loss, humiliation, entrapment, and danger in the pre- Social Psychology, 44(3), 800–807. https://doi.org/ diction of onsets of major depression and general- 10.1016/j.jesp.2007.07.001 (Cited on p.3) ized anxiety. Archives of General Psychiatry, 60(8), Field, T., Diego, M., Pelaez, M., Deeds, O., & Delgado, 789. https://doi.org/10.1001/archpsyc.60.8.789 J. (2009). Breakup distress in university students. (Cited on p.1) Adolescence, 44(176), 705–727 (Cited on p.2). Leary, M. R. (1990). Responses to social exclusion: Social Freedman, G., Powell, D. N., Le, B., & Williams, K. D. anxiety, jealousy, loneliness, depression, and self- (2019). Ghosting and destiny: Implicit theories of esteem. Journal of Social and Clinical Psychology, relationships predict beliefs about ghosting. Journal 9(2), 221–229. https://doi.org/10.1521/jscp.1990.9. of Social and Personal Relationships, 36(3), 905– 2.221 (Cited on p.1) 924. https://doi.org/10.1177/0265407517748791 Leary, M. R., Tambor, E. S., Terdal, S. K., & Downs, D. L. (Cited on p.6) (1995). Self-esteem as an interpersonal monitor: Freedman, G., Williams, K. D., & Beer, J. S. (2016). Soft- The sociometer hypothesis. Journal of Personality ening the blow of social exclusion: The responsive and Social Psychology, 68(3), 518–530. https://doi. theory of social exclusion. Frontiers in Psychology, org/10.1037/0022-3514.68.3.518 (Cited on pp.1, 7. https://doi.org/10.3389/fpsyg.2016.01570 (Cited 2) on p.3) Legate, N., DeHaan, C. R., Weinstein, N., & Ryan, R. M. Garris, C. P., Ohbuchi, K.-i., Oikawa, H., & Harris, (2013). Hurting you hurts me too. Psychological M. J. (2010). Consequences of interpersonal rejec- Science, 24(4), 583–588. https://doi.org/10.1177/ tion. Journal of Cross-Cultural Psychology, 42(6), 0956797612457951 (Cited on pp.1,3, 11) 1066–1083. https : // doi . org / 10 . 1177 / Liberman, N., Sagristano, M. D., & Trope, Y. (2002). The 0022022110381428 (Cited on p. 11) effect of temporal distance on level of mental con- Gerber, J., & Wheeler, L. (2009). On being rejected: A meta- strual. Journal of Experimental Social Psychology, analysis of experimental research on rejection. Per- 38(6), 523–534. https://doi.org/10.1016/s0022- spectives on Psychological Science, 4(5), 468–488. 1031(02)00535-8 (Cited on p.3) https://doi.org/10.1111/j.1745-6924.2009.01158.x London, B., Downey, G., Bonica, C., & Paltin, I. (2007). (Cited on p.2) Social causes and consequences of rejection sensi- Gilbert, D. T., Pinel, E. C., Wilson, T. D., Blumberg, S. J., & tivity. Journal of Research on Adolescence, 17(3), Wheatley, T. P. (1998). Immune neglect: A source 481–506. https://doi.org/10.1111/j.1532- 7795. of durability bias in affective forecasting. Journal of 2007.00531.x (Cited on pp.2, 10) Personality and Social Psychology, 75(3), 617–638. Mendoza-Denton, R., Downey, G., Purdie, V. J., Davis, A., https://doi.org/10.1037/0022-3514.75.3.617 (Cited & Pietrzak, J. (2002). Sensitivity to status-based re- on p.3) jection: Implications for african american students’ college experience. Journal of Personality and So- 14 FREEDMAN & DAINER-BEST

cial Psychology, 83(4), 896–918. https://doi.org/10. Segal, Z. V. (1988). Appraisal of the self-schema construct in 1037/0022-3514.83.4.896 (Cited on p. 10) cognitive models of depression. Psychological Bul- Mirsu-Paun, A., & Oliver, J. A. (2017). How much does love letin, 103(2), 147–162 (Cited on p.1). really hurt? a meta-analysis of the association be- Slavich, G. M., O’Donovan, A., Epel, E. S., & Kemeny, M. E. tween romantic relationship quality, breakups and (2010). Black sheep get the blues: A psychobiolog- mental health outcomes in adolescents and young ical model of social rejection and depression. Neu- adults meta-analysis of the association between ro- roscience & Biobehavioral Reviews, 35(1), 39–45. mantic relationship quality, breakups and mental https://doi.org/10.1016/j.neubiorev.2010.01.003 health outcomes in adolescents and young adults. (Cited on p.1) Journal of Relationships Research, 8. https://doi. Slavich, G. M., Thornton, T., Torres, L. D., Monroe, S. M., & org/10.1017/jrr.2017.6 (Cited on p.2) Gotlib, I. H. (2009). Targeted rejection predicts has- Nezlek, J. B., Wesselmann, E. D., Wheeler, L., & Williams, tened onset of major depression. Journal of Social K. D. (2015). Ostracism in everyday life: The ef- and Clinical Psychology, 28(2), 223–243. https:// fects of ostracism on those who ostracize. The Jour- doi.org/10.1521/jscp.2009.28.2.223 (Cited on p.1) nal of Social Psychology, 155(5), 432–451. https: Sommer, K. L., Williams, K. D., Ciarocco, N. J., & Baumeis- //doi.org/10.1080/00224545.2015.1062351 (Cited ter, R. F. (2001). When silence speaks louder than on p.2) words: Explorations into the intrapsychic and in- Nolan, S. A., Flynn, C., & Garber, J. (2003). Prospective re- terpersonal consequences of social ostracism. Ba- lations between rejection and depression in young sic and Applied Social Psychology, 23(4), 225–243. adolescents. Journal of Personality and Social Psy- https : // doi . org / 10 . 1207 / s15324834basp2304 _ 1 chology, 85(4), 745–755. https://doi.org/10.1037/ (Cited on p.2) 0022-3514.85.4.745 (Cited on p.1) Tashiro, T., & Frazier, P. (2003). “I’ll never be in a rela- Perilloux, C., & Buss, D. M. (2008). Breaking up ro- tionship like that again”: Personal growth following mantic relationships: Costs experienced and cop- romantic relationship breakups. Personal Relation- ing strategies deployed. Evolutionary Psychology, ships, 10(1), 113–128. https : // doi . org / 10 . 1111 / 6(1), 147470490800600. https://doi.org/10.1177/ 1475-6811.00039 (Cited on p.2) 147470490800600119 (Cited on pp.1, 11) Wardenaar, K. J., van Veen, T., Giltay, E. J., de Beurs, E., Poulsen, J. R., & Kashy, D. A. (2012). Two sides of the os- Penninx, B. W., & Zitman, F. G. (2010). Develop- tracism coin: How sources and targets of social ex- ment and validation of a 30-item short adaptation clusion perceive themselves and one another. Group of the Mood and Anxiety Symptoms Questionnaire Processes & Intergroup Relations, 15(4), 457–470. (MASQ). Psychiatry Research, 179(1), 101–106. https://doi.org/10.1177/1368430211430517 (Cited https://doi.org/10.1016/j.psychres.2009.03.005 on pp.1,3, 11) (Cited on pp.4, 11) Rizvi, M. S., Tram, J. M., & Bergström, B. (2021). Fear of Watkins, P. C., Mathews, A., Williamson, D. A., & Fuller, rejecting others: An overlooked construct in dating R. D. (1992). Mood-congruent memory in depres- anxiety. The Family Journal, 106648072110219. sion: Emotional priming or elaboration? Journal of https : // doi . org / 10 . 1177 / 10664807211021969 Abnormal Psychology, 101(3), 581–586. https : // (Cited on p.2) doi . org / 10 . 1037 / 0021 - 843X . 101 . 3 . 581 (Cited Romero-Canyas, R., Downey, G., Berenson, K., Ayduk, Ö., on p.2) & Kang, N. J. (2010). Rejection sensitivity and Watson, D., Clark, L. A., Weber, K., Assenheimer, J. S., the rejection-hostility link in romantic relationships. Strauss, M. E., & McCormick, R. A. (1995). Testing Journal of Personality, 78(1), 119–148. https://doi. a tripartite model: II. exploring the symptom struc- org/10.1111/j.1467-6494.2009.00611.x (Cited on ture of anxiety and depression in student, adult, and p.2) patient samples. Journal of Abnormal Psychology, Rosenberg, M. (1965). Society and the adolescent self- 104(1), 15–25. https : // doi . org / 10 . 1037 / 0021 - image. Princeton University Press. (Cited on p.4). 843x.104.1.15 (Cited on p.2) Sato, K., Yuki, M., & Norasakkunkit, V. (2014). A socio- Watson, J., & Nesdale, D. (2012). Rejection sensitivity, so- ecological approach to cross-cultural differences in cial withdrawal, and loneliness in young adults. the sensitivity to social rejection. Journal of Cross- Journal of Applied Social Psychology, 42(8), 1984– Cultural Psychology, 45(10), 1549–1560. https : // 2005. https://doi.org/10.1111/j.1559-1816.2012. doi . org / 10 . 1177 / 0022022114544320 (Cited on 00927.x (Cited on p.2) p. 11) Wesselmann, E. D., Wirth, J. H., Pryor, J. B., Reeder, G. D., & Williams, K. D. (2013). When do we os- SELF-SCHEMAS AND SOCIAL REJECTION 15

tracize? Social Psychological and Personality Sci- ence, 4(1), 108–115. https : // doi . org / 10 . 1177 / 1948550612443386 (Cited on p.3) Williams, K. D. (2007). Ostracism. Annual Review of Psy- chology, 58(1), 425–452. https://doi.org/10.1146/ annurev.psych.58.110405.085641 (Cited on p.1) 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 and Interpersonal Loss, 5(1), 19–63. https : // doi . org/10.1080/10811440008407846 (Cited on pp.1, 3, 11) Wilson, T. D., & Gilbert, D. T. (2003). Affective forecast- ing. Advances in experimental social psychology (pp. 345–411). Academic Press. https : // doi . org / 10.1016/s0065-2601(03)01006-2. (Cited on p.3) 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. https://doi.org/10.1177/1368430205051062 (Cited on p.2) Zhou, J., Li, X., Tian, L., & Huebner, E. S. (2018). Lon- gitudinal association between low self-esteem and depression in early adolescents: The role of rejec- tion sensitivity and loneliness. Psychology and Psy- chotherapy: Theory, Research and Practice, 93(1), 54–71. https://doi.org/10.1111/papt.12207 (Cited on pp.2,3) 16 FREEDMAN & DAINER-BEST

Appendix Study 1b: Preregistered Gender Analyses

Study 1a: Exploratory Gender Analyses Contrary to the findings of Study 1a, participant gender did not play a role in perceived difficulty for engaging The role of gender on perceived difficulty of re- in explicit rejection (t[253] = −0.756, p = .450, d = 0.10) jecting and likelihood of rejecting was examined. Men re- or ghosting (t[253] = −0.402, p = .688, d = 0.05). Par- ported finding rejecting less difficult (M = 4.47, SD = 1.15) ticipant gender also did not play a statistically significant than women (M = 4.81, SD = 0.96); t(205) = −2.20, p = role in participants’ willingness to engage in explicit re- .029, d = 0.31, 95% CI on mean difference [-0.65, -0.04]), jection (t[253] = −0.304, p = .761, d = 0.04), ghosting but men (M = 3.41, SD = 0.52) were also less likely to re- (t[253] = −1.748, p = .082, d = 0.22), or in their overall port being willing to end the relationships than women (M = likelihood of ending the relationship (t[253] = −1.860, p = 3.70, SD = .48); t(200) = −3.78, p < .005, d = 0.53, 95% CI .064, d = 0.23). [-.43, -.14]). There were no gender differences in predicted willingness to engage in the specific behaviors of explicit re- jection (t[205] = −0.46, p = .649, d = 0.06, 95% CI [-0.23, Supplemental Tables 0.15]) or ghosting (t[205] = −1.46, p = .146, d = 0.20, 95% CI [-0.44, 0.07]). See next page. SELF-SCHEMAS AND SOCIAL REJECTION 17

Table A1

Intercorrelations between variables in Study 1a Variable 1 2 3 4 1. General Distress 2. Anhedonic Depression .459** - 3. Anxious Arousal .639** .250** - 4. Self-esteem -.716** -.555** -.532** - 5. Rejection Sensitivity .332** .225** .187* -.410** Note. * p < .05, ** p < .005

Table A2

Intercorrelations between variables in Study 1b Variable 1 2 3 4 1. General Distress 2. Anhedonic Depression .422** - 3. Anxious Arousal .588** .072 - 4. Self-esteem -.710** -.594** -.405** - 5. Rejection Sensitivity .439** .311** .281* -.500** Note. * p < .05, ** p < .005

Table A3

Intercorrelations between variables in Study 2 Variable 1 2 3 4 1. General Distress 2. Anhedonic Depression .472** - 3. Anxious Arousal .523** .041 - 4. Self-esteem -.764** -.668** -.318** - 5. Rejection Sensitivity .493** .434** .254* -.563** Note. * p < .05, ** p < .005