Self-Evaluative and Other-Directed Emotional and Behavioral Responses to Gossip About the Self

Martinescu, Elena; Janssen, Onne; Nijstad, Bernard A.

doi:10.3389/fpsyg.2018.02603

ORIGINAL RESEARCH article

Front. Psychol., 04 January 2019

Sec. Organizational Psychology

Volume 9 - 2018 | https://doi.org/10.3389/fpsyg.2018.02603

This article is part of the Research Topic Why People Gossip and What It Brings About: Motives For, and Consequences Of, Informal Evaluative Information Exchange View all 7 articles

Self-Evaluative and Other-Directed Emotional and Behavioral Responses to Gossip About the Self

$\r\nElena Martinescu,*$ Elena Martinescu^1,2^*

Onne Janssen²

Bernard A. Nijstad²

¹King's Business School, King's College London, London, United Kingdom
²Department of Human Resource Management & Organizational Behaviour, University of Groningen, Groningen, Netherlands

Gossip, or informal talk about others who are not present, is omnipresent in daily interactions. As such, people who are targeted are likely to hear some gossip about themselves, which may have profound implications for their well-being. We investigated the emotions and behavioral intentions of people who hear performance-related gossip about themselves. Based on the affective events theory, we predicted that gossip incidents have strong emotional consequences for their targets and that these emotional responses trigger different behaviors. Two scenario studies (N₁ = 226, M_age = 21.76; N₂ = 204, M_age = 34.11) and a critical incident study (N = 240, M_age = 37.04) compared targets' responses to positive and negative gossip. Whereas, targets of positive gossip experienced positive self-conscious emotions (e.g., pride), targets of negative gossip experienced negative self-conscious emotions (e.g., guilt), especially when they had low core self-evaluations. In turn, these negative self-conscious emotions predicted repair intentions. Positive gossip also led to positive other-directed emotions (e.g., liking), which predicted intentions to affiliate with the gossiper. Negative gossip, however, also generated other-directed negative emotions (e.g., anger), especially for targets with high reputational concerns, which in turn predicted retaliation intentions against the gossiper. This pattern of emotional reactions to self-relevant gossip was found to be unique and different from emotional reactions to self-relevant feedback. These results show that gossip has self-evaluative and other-directed emotional consequences, which predict how people intend to behaviorally react after hearing gossip about themselves.

Introduction

People talk about others abundantly and such conversations about others play a pivotal role in the informal communication network in the workplace (Noon and Delbridge, 1993). Research suggests that up to 70% of our daily conversations contain positive or negative informal evaluations about someone who is not present (Dunbar et al., 1997), a type of interpersonal communication known as gossip (Foster, 2004). Accordingly, gossip has been portrayed as intrinsic to human nature and essential for group functioning (Dunbar, 2004). For example, the threat of becoming a gossip target has been found to deter self-serving behavior and to increase group-serving behavior in social dilemma situations (e.g., Beersma and van Kleef, 2011; Feinberg et al., 2012).

Although gossip is typically not intended to be heard by the gossip targets (i.e., people whom the gossip is about; Foster, 2004), gossip does occasionally leak out of the closed gossip circle and reaches its targets. As gossip is evaluative in nature and can damage or boost people's self-evaluation and reputation (Burt, 2008), it may have profound implications for its targets' well-being and is likely to induce specific emotions and behaviors. Negative gossip represents a social judgment about undesirable behaviors or characteristics, which may reduce target's self-confidence, as well as others' trust and willingness to cooperate with the target (e.g., Sommerfeld et al., 2007; Beersma and van Kleef, 2011). In contrast, positive gossip may benefit the target, because it promotes a positive view of the self and signals support and social inclusion (Merry, 1984; Burt, 2008; Ellwardt et al., 2012). Despite these profound consequences, empirical research has not yet clarified how people react when they do hear gossip about themselves. Thus, although we know that the threat of gossip may increase pro-group behavior, we do not know the consequences of actually being a gossip target.

In this contribution, we used the affective events theory (AET, Weiss and Cropanzano, 1996) to examine how gossip targets feel and behave. Based on the AET, we predicted that gossip targets experience complex emotional and behavioral responses, that are shaped by the valence of the gossip incident (positive vs. negative) and its implications for self-evaluation and reputation. Specifically, we propose that gossip is an affective event, and employees who are confronted with negative gossip about themselves manifest the intention to repair their mistakes, due to self-conscious negative emotions (e.g., guilt), but also the intention to retaliate against the gossiper, due to other-directed negative emotions (e.g., anger). Thus, becoming the target of especially negative gossip may, on the one hand, stimulate group-serving behavior in the form of repair intentions, but it may also trigger more destructive behaviors such as retaliation. Those who encounter positive gossip about themselves might experience self-conscious positive emotions (e.g., pride) and/or other-directed positive emotions (e.g., liking), which may increase the development of affiliation intentions.

We examined two further issues. First, we investigated whether the predicted patterns of event-emotion-behavior (cf. AET, Weiss and Cropanzano, 1996) are contingent on the type of information people receive about themselves. Although (informal) gossip about the self and (formal) feedback may convey the same message about someone's behavior or abilities, they have a different intended audience and purpose: feedback is directed at the target person and meant to help the target, whereas gossip is not. Thus, we propose that being the target of gossip induces emotional reactions that are unique and different from the emotional reactions employees experience when they receive evaluative feedback about the self. Second, we investigated whether the emotions and behaviors individuals experience in response to gossip are contingent on their predispositions. We identified two dispositional traits that moderate the effect of gossip valence on emotions and behavioral intentions: core self-evaluations and concern for reputation. Because being the target of gossip may arouse both self-conscious and other-directed emotions, and because the behavioral consequences of those distinct emotions are quite different, it is important to understand when self-conscious vs. other-directed emotional responses are more likely.

Theoretical Framework

Because maintaining positive evaluations about the self and a positive reputation among group members are fundamental to well-being (Kunda, 1990; Sedikides and Strube, 1997; Anseel et al., 2007), people are emotionally sensitive to hearing evaluations about their attributes or performance (Frijda, 1988). As such, hearing gossip about the self is likely to cause strong affective responses and behavioral manifestations. Indeed, according to Weiss and Cropanzano's (1996) affective events theory (AET), incidents at work, such as gossip incidents, are likely to trigger emotional reactions. As a theory of workplace emotion, the AET offers a framework for explaining attitudes and behavior in the workplace, by incorporating within-person changes in affective states (Weiss and Beal, 2005; Cropanzano et al., 2017). Specifically, the AET conceptualizes events in the workplace as the cause of affective reactions, which have a direct influence on affect-based behaviors. Furthermore, the AET considers the role of individual predispositions in shaping behavior, by influencing affective reactions to events.

Consistent with the AET, we propose that receiving gossip about the self represents an affective event that generates complex affective experiences and behavioral manifestations. During critical events (such as hearing gossip about the self), emotions prioritize psychological experiences and behaviors, helping individuals respond to environmental challenges or opportunities and adjust their goals (Oatley and Johnson-Laird, 1987). Furthermore, emotions trigger certain behavioral responses (or behavioral intentions) that are aimed at achieving these goals (e.g., Frijda, 1988).

Self-Conscious Emotional Reactions and Behavioral Intentions

Gossip provides targets with information about their performance, abilities, moral behavior, or other self-relevant issues (Peters and Kashima, 2014), which may support or discredit their positive self-image (Argyle, 1969; Kunda, 1990; Sedikides and Strube, 1997). At face value, gossip about the self represents impromptu positive or negative feedback, which targets can use to evaluate their past behavior or their attributes and act more in line with the pressures and demands of their environment. Drawing from the AET, because gossip contains evaluative information about the self, it is a meaningful self-relevant event that is likely to evoke self-conscious emotional reactions and subsequent emotion-based behavior.

Self-conscious emotions arise when individuals examine their own traits or behavior and are pleased or displeased with themselves (Tracy and Robins, 2004). In the self-evaluation process, individuals compare characteristics of the self with certain social, normative standards, which helps them maintain their place in the social hierarchy. Because gossipers analyze targets' behavior in relation to social norms, gossip is, by definition, normative (e.g., Beersma and Van Kleef, 2012; Feinberg et al., 2012; Peters and Kashima, 2014) and is therefore likely to evoke self-conscious emotions for targets.

Negative gossip raises targets' awareness of their inadequate attributes or behavior (e.g., substandard performance, task failure, avoiding responsibilities). Targets of negative gossip may feel responsible for not meeting certain standards or for harming or disappointing others and are likely to experience self-conscious negative emotions, such as guilt and shame (Tracy and Robins, 2004). Guilt comes from transgressing a social or moral standard, while shame stems from a failure to live up to an ego ideal (e.g., Frijda et al., 1989). Accordingly, it is likely that negative gossip leads to higher self-conscious negative emotions than positive gossip. Furthermore, self-conscious negative emotions like guilt and shame induce tension, regret and a tendency to alter the situation (Solomon, 1993). This makes individuals less likely to repeat the negative behavior and more likely to repair the harm they have caused (Frijda et al., 1989). Indeed, research suggests that individuals who have violated norms or underperformed at work engage in compensatory behavior to mitigate their negative emotions and repair their social relations (Nelissen and Zeelenberg, 2009). Therefore, and in accordance with the social control function of gossip (Beersma and van Kleef, 2011; Beersma and Van Kleef, 2012; Feinberg et al., 2012), we propose the following:

Negative (relative to positive) gossip triggers repair intentions and this effect is mediated by self-conscious negative emotions (hypothesis 1).

Positive gossip entails a favorable evaluation (Foster, 2004). Because people need positive evaluations from others to function well (Sedikides and Strube, 1997), this is likely to please its targets. Because people are motivated to believe that the favorable evaluation is due to their own merit (Kunda, 1990), positive gossip is likely to result in positive self-conscious emotions (such as pride; see Lazarus, 1991; Solomon, 1993; Tracy and Robins, 2004; Tracy et al., 2010). Pride makes people feel that they deserve high status and group acceptance and also signals to others that they have high personal value (Tracy et al., 2010). However, expressions of pride are likely to manifest subtly in an interpersonal interaction. Lazarus (1991) notes that it is difficult to specify a clear action tendency for pride, because self-praise is socially undesirable. Therefore, we do not predict specific behavioral intentions for positive gossip targets who feel positive self-conscious emotion, but we predict that:

Positive gossip generates higher self-conscious positive emotion than negative gossip (hypothesis 2).

Other-Directed Emotional Reactions and Behavioral Intentions

Self-conscious emotions may, however, not be the only emotional reaction to gossip. Gossip communicates information and evaluations about targets to third parties, and shapes targets' reputation among group members in a positive or negative way (e.g., Burt, 2008; Beersma and Van Kleef, 2012; Feinberg et al., 2012). Individuals' reputation influences the extent to which they are socially accepted and trusted, or ostracized from groups (Feinberg et al., 2014), which is fundamental for well-being. Based on the AET, we propose that individuals are likely to react affectively to events that shape their reputation and to act to protect or promote their reputation. Because gossipers are the agents that influence targets' reputation, gossip is also likely to generate other-directed positive or negative emotions.

Negative gossip informs targets that the gossipers criticized them and exposed their weaknesses to others. Because it takes place in their absence, targets may experience negative gossip about themselves as offensive and unfair and may hold the gossiper responsible for harming them (Lazarus, 1991; Solomon, 1993). Gossipers may indeed abusively self-enhance at the expense of gossip targets (Wert and Salovey, 2004), by damaging their reputation (Burt, 2008), restraining their power (Ogasawara, 1998), or decreasing their sexual attractiveness (Massar et al., 2012). Targets of negative gossip are likely to feel that the gossiper has violated their interests and to experience other-directed negative emotion (e.g., anger, Lazarus, 1991). Because people strive for equitable relationships with co-workers (Adams, 1965), experiencing unfair treatment causes distress and desire for revenge (Lazarus, 1991). Anger facilitates punishing gossipers and justifies counter-attacks against offenders (Lazarus, 1991). Thus, we expect that:

Negative (relative to positive) gossip leads to retaliation intentions against the gossiper and this relationship is mediated by other-directed blame (hypothesis 3).

In contrast, positive gossip may generate positive emotions directed at the gossiper (i.e., interpersonal liking¹), because targets will recognize the active role gossipers play in elevating their reputation (Sommerfeld et al., 2007; Burt, 2008). Positive gossip creates “cost free social rewards” (Merry, 1984) and people like others with whom they associate rewards and positive outcomes with (Sternberg, 1987). Moreover, behaviors that imply warmth and affiliation, such as positive gossip, tend to elicit reciprocal responses from others (Kiesler, 1983). Indeed, a meta-analysis showed that indirect ingratiating behaviors, such as other-enhancement, have strong effects on interpersonal attraction and liking (Gordon, 1996). As such, targets of positive gossip may experience higher other-directed happiness than targets of negative gossip. Furthermore, interpersonal liking helps individuals develop trust and establish collaborations due to a feeling of shared social identity (Hogg and Turner, 1985). Because people who are liked are attributed with favorable motives (Nicholson et al., 2001), positive gossip targets might think that they have shared values with the gossiper and will find a trustworthy ally in this person. Thus, we predict that:

Positive (relative to negative) gossip predicts targets' affiliation intentions with the gossiper due to increased other-directed positive emotions (hypothesis 4).

Gossip vs. Feedback

As gossip provides targets with evaluative information about themselves, it may serve as informal feedback. This raises the question whether targets' emotional and behavioral reactions to gossip are unique and can be distinguished from reactions to (formal) feedback, or whether those reactions to gossip are general response tendencies to positive and negative evaluations about the self.

Although gossip and formal feedback both contain self-relevant information for targets, they have a different audience and purpose and may therefore have different effects on targets' emotions and subsequent behavior. Feedback is addressed directly to its targets and is meant to help them improve (Kluger and DeNisi, 1996) and is usually embedded in organizational procedures and policies. However, gossip is spread in the targets' absence, with an obscure purpose [e.g., to gain information, protect others, harm targets, or for enjoyment, (Beersma and Van Kleef, 2012)], usually in spontaneous interactions between gossipers. Because feedback is addressed directly and usually obtained through formal channels, with the intention to help, feedback is likely to be perceived as more legitimate by targets than gossip. Therefore, we expect that:

Feedback leads to stronger self-conscious negative (hypothesis 5a) and positive emotions than gossip (hypothesis 5b).

Furthermore, because gossip is likely to be more threatening, more socially undesirable and more intrusive than feedback (Dunbar, 2004; Foster, 2004), we predict that:

Gossip leads to higher other-directed negative emotions (hypothesis 6a) and to lower other-directed positive emotions than feedback (hypothesis 6b).

Predispositions

The AET contends that people's predispositions influence their attention to and processing of social and emotional stimuli and thus their emotional and behavioral responses to affective work events (Weiss and Cropanzano, 1996). Whereas, positive self-conscious (pride) and other-directed emotions (liking) can easily co-occur and may be universally experienced by targets of positive gossip, while individuals may react differently to negative gossip. Threats and losses trigger more specific responses than gains (Baumeister et al., 2001) and these responses may depend on dispositional factors (Mischel and Shoda, 1995). Specifically, we propose that responses to negative gossip will depend on individuals' core self-evaluations [CSE, (Judge et al., 1997)] and concern for reputation [CR, (De Cremer and Tyler, 2005)].

Core Self-Evaluations

People's sensitivity to negative stimuli may vary depending on their core self-evaluations [CSE, (Judge et al., 1997)]. CSE is a higher order construct that reflects individuals' fundamental appraisals about their self-worth and capabilities. Individuals with low CSE evaluate their identities and capabilities negatively and have low self-esteem, self-efficacy, internal locus of control, and emotional stability (Judge et al., 1997, 2003). Due to their weak coping and self-regulation skills, people with low CSE seek little social support, develop few social ties, are vulnerable to threats (Chang et al., 2012) and more intensely experience stressful work events, such as negative gossip (Grant and Sonnentag, 2010). Because targets of negative gossip with low CSE see themselves as unworthy individuals, unable to cope with challenges, they may think that indeed they are responsible for their negative evaluations. Thus, we hypothesize that:

The effect of gossip valence on self-conscious negative emotions is stronger for individuals with low rather than high CSE (hypothesis 7a).

The indirect effect of gossip valence on repair intentions through self-conscious negative emotions is stronger for individuals with low rather than high CSE (hypothesis 7b).

Reputational Concerns

Reputation is the extent to which someone is known to be trustworthy (Burt, 2008) and represents shared opinions about a focal person above and beyond directly observable behavior (Sommerfeld et al., 2007; Burt, 2008). People need a favorable reputation to ensure future collaborations with others (Burt, 2008), but their reputation is vulnerable to gossip (Foster, 2004). Concern for reputation (CR) reflects the importance people give to their public image of being trustworthy (De Cremer and Tyler, 2005). Gossip targets who are concerned about their reputation are highly aware of the losses associated with a bad reputation (Burt, 2008). Consequently, targets with high CR may perceive others who spread negative gossip about them as offensive and unfair, causing them harm. We therefore expect that:

The effect of gossip valence on other-directed negative emotions is stronger for people with high rather than low CR (hypothesis 8a).

The indirect effect of gossip valence on retaliation intentions through other-directed negative emotions is stronger for people with high rather than low CR (hypothesis 8b).

Overview of Studies

To test our hypotheses, we conducted three studies in which we manipulated the valence of gossip overheard by targets. We focused on performance-related gossip, which often occurs in the workplace or similar achievement environments. In Study 1, hypothesized mediation relationships between gossip valence, emotional reactions, and behavioral intentions (H1-4) were tested in a scenario experiment with a sample of university students. In Study 2, a scenario study among employees aimed to replicate these findings and additionally tested whether gossip and feedback differ in the emotional and behavioral reactions they elicit for gossip targets (H5-6). Study 3 was a critical incidents study that examined the hypothesized moderating effects of core self-evaluation and concern for reputation in targets' responses to positive and negative gossip (H7-8). Figure 1 presents the overview of our model.

FIGURE 1

Figure 1. Theoretical model. Solid lines represent relationships tested in all three studies. Dashed lines represent relationships tested in Study 2. Dotted lines represent relationships tested in Study 3.

Study 1

Method

Design and Participants

Two hundred and twenty six economics and business undergraduates at a Dutch university (108 female, M_age = 21.76, SD = 3.21), participated in exchange for a course credit or for 4 Euros. Participants were randomly assigned to a positive (N = 116) or negative (N = 110) gossip condition. The study was approved by a research ethics committee, and participants gave written informed consent.

Procedure

Upon arrival at the research laboratory, participants were seated in separate cubicles and informed that they would participate in research about students' reactions to informal evaluative talk about themselves. Participants read a scenario and were asked to imagine that they had written it.

Gossip Valence Manipulation

Participants imagined working on a group assignment and overhearing two classmates talk about them behind their back. Participants were targets of either positive or negative performance-related gossip: “A said to B that you are/are not a good group member and likes/dislikes working with you, because your contribution to the assignment is remarkable/disappointing. You seem to be very hardworking/lazy, since you always/never come prepared to the meetings and choose to do the most difficult/easiest tasks.” Next, participants filled in manipulation checks and dependent variable measures, were debriefed, compensated and thanked for participation².

Measures

Manipulation Checks

Participants were asked to indicate whether the overheard gossip was positive or negative. They could choose between (1) The information classmate A told classmate B about me is positive and (2) The information classmate A told classmate B about me is negative. Next, participants were asked to summarize the information they overheard.

Dependent Measures

For all dependent variables we used a 7-point Likert response with scales ranging from 1 (strongly disagree) to 7 (strongly agree).

Emotions

Emotions were measured using items from the PANAS-X scale (Watson and Clark, 1994). Specifically, participants indicated to what extent the overheard gossip information made them feel “guilty,” “ashamed,” “blameworthy,” “angry at self” (self-conscious negative emotions, α = 0.96), “proud,” “strong,” “bold” (self-conscious positive emotions, α = 0.91), “hostile,” “irritable,” and “angry” (other-directed negative emotions, α = 0.93). Other-directed positive emotions were measured with 3 items adapted from Wojciszke et al. (2009): “I like classmate A,” “I have warm feelings about classmate A,” and “I feel close to classmate A” (α = 0.93).

Behavioral Intentions

Repair intentions were measured with 3 items developed by Martinescu et al. (2014). Following the introductory phrase “The things classmate A said about me to classmate B would help me,” the specific items were: “Understand how to improve my contribution to the group assignment,” “Improve my performance in the group assignment,” and “Understand that I can do better in the group assignment” (α = 0.86). Behavioral intentions regarding the gossiper were measured following the introductory phrase: “How likely would it be for you to do the following things…” using two items for retaliation intentions: “Talk badly about classmate A” and “Punish classmate A if I can” (α = 0.74), and two items for affiliation intentions: “Team-up with classmate A in the future,” “Try to become friends with classmate A” (α = 0.76).

Results

Manipulation Checks

One participant in the positive gossip condition and three in the negative condition incorrectly indicated the valence of overheard information. The summaries of overheard information were coded by the first author blind to the condition and a second coder, blind to the gossip condition and the hypotheses, coded a subset of 72 stories (37 in the positive condition). All stories (100%) were coded as matching the condition by both coders, thus the gossip valence manipulation was successful.

Descriptive Statistics

Table 1 presents means, standard deviations, and correlations for variables in Study 1. Gossip valence was strongly correlated with all other variables (see zero-order correlations in Table 1, below the diagonal), which may generate spurious correlations among emotions and behavioral intentions. Therefore, Table 1 also contains partial correlations among these variables, controlling for valence (above the diagonal). Partial correlations showed that self-conscious and other-directed negative emotions were unrelated (r = −0.05, ns), whereas self-conscious and other-directed positive emotions were positively related (r = 0.25, p < 0.01).

TABLE 1

Table 1. Means, standard deviations, and zero-order Pearson correlations for variables in Study 1.

Hypotheses Testing

To test our hypotheses, we conducted multiple regressions with the Preacher et al.'s (2007) procedure for mediation analysis. To assess the indirect effect of gossip valence on each of the three behavioral intentions we employed 5,000 bootstrap samples; all four types of emotional responses were included as simultaneous mediators and we examined the unique indirect effects for each emotion type (see Table 2).

TABLE 2

Table 2. Indirect effect analyses in Study 1.

Targets of negative gossip experienced higher self-conscious negative emotions (M = 4.35; SD = 1.54) than targets of positive gossip (M = 1.35; SD = 0.73), b = −1.48, p < 0.001. Self-conscious negative emotions were positively related to repair intentions, b = 0.36, p < 0.001, and mediated the effect of gossip valence on repair intentions, indirect effect = −0.53, 95% CI [−0.82; −0.28], supporting hypothesis 1. As predicted by hypothesis 2, positive gossip targets experienced higher self-conscious positive emotions (M = 5.44; SD = 0.92) than negative gossip targets (M = 2.05; SD = 1.15), b = 1.69, p < 0.001.

Furthermore, negative gossip targets experienced higher other-directed negative emotions (M = 4.49; SD = 1.41) than positive gossip targets (M = 1.44; SD = 0.71), b = −1.52, p < 0.001. Other-directed negative emotions positively predicted retaliation intentions, b = 0.44, p < 0.001, and mediated the effect of gossip valence on retaliation intentions, indirect effect = −0.68, 95% CI [−0.95; −0.43], supporting hypothesis 3. Finally, positive gossip targets experienced higher positive other-directed emotions (M = 4.51; SD = 1.28) than negative gossip targets (M = 2.07; SD = 1.02), b = 1.21, p < 0.001. Other-directed positive emotions predicted affiliation intentions, b = 0.60, p < 0.001, and mediated the effect of gossip valence on affiliation intentions, indirect effect = 0.73, 95% CI [0.53;0.94], supporting hypothesis 4.

In addition to the hypothesized indirect effects, the analyses reported in Table 2 showed that other-directed positive emotion also predicted repair intentions and mediated the effect of gossip valence on repair intentions. Furthermore, self-conscious positive emotion positively predicted retaliation intentions, and mediated the effect of gossip valence on retaliation intentions. Although these indirect effects were unexpected, they were weaker than the hypothesized effects.

Discussion

Results of Study 1 supported our expectations that gossip targets experience distinct emotions which predict specific behavioral intentions. Specifically, self-conscious negative emotions were higher for targets of negative than for targets of positive gossip, and generated repair intentions. Negative compared to positive gossip also evoked stronger other-directed negative emotions, which increased retaliation intentions against gossipers. Compared to negative gossip, positive gossip made targets feel stronger positive self-conscious emotions, as well as stronger other-directed positive emotions, which increased affiliation intentions. In Study 1 we thus found good support for our predictions that gossip about the self, evokes positive and negative emotions and behavioral intentions. In a second scenario study we investigated whether people have different reactions to gossip and feedback.