Você está na página 1de 7

10.5464.AMBPP.2011.210.

KNOW YOUR BOSS: THE INFLUENCE OF EMOTIONAL STABILITY ON THE AFFECTIVE AND ATTITUDINAL FEEDBACK REACTIONS OF POWERFUL LEADERS JANA NIEMANN Department of Social and Organizational Psychology University of Groningen Grote Kruisstraat 2/1 9712 TS Groningen The Netherlands BARBARA WISSE, DIANA RUS, NICO W. VAN YPEREN University of Groningen KAI SASSENBERG Eberhard Karls University Tuebingen ABSTRACT Given the prevalence of feedback at all organizational levels, one vexing question is how to prevent destructive reactions to negative feedback. In two experimental studies, we indicate that power moderates the effects of emotional stability on attitudinal feedback reactions, and that this relationship is mediated by other-directed emotional feedback reactions. Our findings suggest that it is crucial to take the interplay between the feedback receivers power and personality variables into account. INTRODUCTION Negative feedback is an essential organizational tool geared at preventing the recurrence of disruptive incidents and at guiding behavior towards organizationally-desirable paths (Bee & Bee, 1998). Unfortunately, it is also known to elicit negative emotions, such as anger and shame (Belschak & Den Hartog, 2009; Cupach & Carson, 2002), both of which have been shown to detrimentally impact work-related attitudes and performance (Fox & Spector, 1999; Thompson, Altmann, & Davidson, 2004). Consequently, particularly in this day and age when feedback is increasingly employed at all levels of the organizational hierarchy, with both leaders and subordinates being confronted with it (cf. van Dierendonck, Haynes, Borrill, & Stride, 2007), it is essential to identify factors that buffer against such adverse side effects of negative feedback. Emotional stability, an individual difference in the predisposition to experience negative emotions (Costa & McCrae, 1980), is likely to serve as an important factor in shielding individuals from detrimental reactions to stressful events such as negative feedback. Indeed, the more emotionally stable individuals are, the less likely they are to experience destructive emotions, such as anger and shame, in response to negative feedback (Larsen & Ketelaar, 1989). Interestingly, to date, we know little about how leaders emotional stability may influence their reactions to negative feedback. This is all the more interesting, given that leaders command at least a certain degree of power and power, defined as asymmetric control over valuable resources (Galinsky, Magee, Gruenfeld, Whitson, & Liljenquist, 2008), has been shown to

10.5464.AMBPP.2011.210.a

strengthen the influence of individuals internal states, traits and predispositions (Galinsky et al., 2008). Indeed, there is increasing evidence that power strengthens the correspondence between individual predispositions and behavior (Rus, van Knippenberg, & Wisse, 2010), and between individual predispositions and cognitions (Avey, Wernsing, & Luthans, 2008; Galinsky et al., 2008). Whereas to date, the effects of power on the convergence between individual differences and emotions have not been investigated, it is reasonable to assume that power will also increase the correspondence between individual predispositions, such as emotional stability, and the experience of emotions. Therefore, we posit that the effects of emotional stability on emotional feedback reactions will be particularly strong in high power feedback receivers. Hypothesis 1. With higher levels of emotional stability only high (vs. low) power individuals are less likely to experience anger and shame in response to negative feedback. In line with findings emphasizing the influence of emotions on attitudes (Avey et al., 2008; Belschak & Den Hartog, 2009), we further propose that emotional feedback reactions mediate the interactive effects of emotional stability and power on attitudinal feedback reactions (Study 2). That is, we posit that the negative emotional feedback reactions (particularly anger) experienced by powerful individuals with lower levels of emotional stability will detrimentally affect attitudes towards the feedback provider and the feedback itself. By increasing insights into the effects of power on reactions to negative feedback, our work is especially relevant to the study of leadership. Indeed, insights into leaders reactions to criticism are still scarce (van Dierendonck et al., 2007) while these reactions can engender large consequences for the feedback provider (Cupach & Carson, 2002; Smith & Fortunato, 2008). STUDY 1 Method Participants and design. Eighty-four participants were randomly assigned to one of two power conditions (high vs. low). Emotional stability was the continuous independent variable. Procedure. Participants were invited to participate in a computer-mediated study on task completion strategies and were seated in individual cubicles. Instructions were presented on computer screens and dependent measures were recorded by the program software. Emotional stability was assessed with Davis (1980; 1983) personal distress subscale of the Interpersonal Reactivity Index ( = .75). To manipulate power, participants were, ostensibly based on their responses to a questionnaire, assigned to either the high (supervisor) or the low power position (subordinate) in a dyad (cf. C. Anderson & Berdahl, 2002). Participants were asked to develop an advertising campaign for a pub. After waiting for about 3 minutes, participants received preprogrammed negative feedback on their campaign, ostensibly from their counterpart. They read that their rating was 2.8 on a 10-point scale, implying that their performance was highly insufficient. In the end, participants responded to the manipulation checks and dependent measures. To assess the success of the power manipulation, participants indicated whether they had been assigned to the supervisor or the subordinate role, and they responded to a 7-point scale (1 = not at all, 7 = a lot) 4-item power measure (cf. Galinsky, Gruenfeld, & Magee, 2003; = .75). To assess feelings of anger and shame, participants indicated on a 5-point scale (1 = very little, 5 = very

10.5464.AMBPP.2011.210.a

much) the extent to what they felt angry, irritated and annoyed (anger; = .78) and the extent to what they felt ashamed, embarrassed and regretful (shame; = .78). Results In all hierarchical regression analyses main effect terms for power and emotional stability were included in Step 1. The power by emotional stability interaction term was included in Step 2. A chi-square test revealed that 97.62% of the participants accurately reported whether they were assigned the supervisor role or the subordinate role, (1, N = 84) = 76.15, p < .001. Furthermore, a hierarchical regression analysis on the power index as the dependent variable only revealed a main effect of power (b = 1.48, SE = 0.12, p < .001; Mhigh power = 4.99, SD = 1.14; Mlow power = 2.03, SD = 0.90). In line with our hypothesis, hierarchical regression analyses revealed a significant power x emotional stability interaction on both anger and shame. Simple slope analyses showed that only for high power individuals emotional stability yielded a strong negative relationship to anger (1 SD above the mean, b = -0.71, SE = 0.17, p < .001) and to shame (1 SD above the mean, b = -0.34, SE = 0.14, p < .05), but not for low power individuals. Discussion Study 1 In Study 1, we argued and showed that with higher levels of emotional stability high power individuals, but not low power individuals, were less likely to experience anger and shame in response to negative feedback. In doing so, our findings underscore the importance of systematically addressing power differentials in studies that focus on buffering detrimental responses to criticism and advice during interpersonal interaction. Two issues, however, remain open for further exploration. First, recent evidence suggests that anger can be directed towards the self or others (Ellsworth & Tong, 2006). Because primarily the latter evokes deviant and aggressive behavior (Berkowitz, Lepinski, & Angulo, 1969; Geen, 1990), it is important to clarify whom the feedback-induced anger is directed at. Contrary to self-directed anger, a scarce emotion experienced if no other individuals are involved, individuals experience other-directed anger if they hold others responsible for events or feel unfairly treated (Ellsworth & Tong, 2006). Feedback receivers may blame the feedback provider for delivering the negative feedback or perceive the feedback as unjustified. Therefore, we expect negative feedback to elicit other-directed anger, more so than self-directed anger. Second, it seems particularly interesting to explore the downstream effects of feedbackinduced negative emotions. Indeed, attitudinal feedback reactions such as devaluing the feedback provider or rejecting the feedback can harm the relationship between the feedback provider and the receiver as well as impede learning processes crucial for feedback effectiveness. Research on the influence of emotions on attitudes suggests that cognitions and behaviors directed at the environment are more likely to predict how individuals interact with their environment, rather than how they treat themselves, whereas self-directed emotions are more likely to predict selfdirected cognitions and behavior (Jang & Johnson, 2003, Aseltine, Gore, & Gordon, 2000; Mazerolle & Piquero, 1998). Based on this, we posit that the other-directed emotion of anger, but not the self-directed emotion of shame, will mediate the interactive effects of power and emotional stability on attitudes towards the feedback provider and the feedback content.

10.5464.AMBPP.2011.210.a

Hypothesis 2. Anger towards the feedback provider mediates the effects of emotional stability and power of the feedback receiver on perceived ability and liking of the feedback provider and feedback acceptance (see Figure 1). -----------------------Figure 1 about here -----------------------Importantly, by incorporating both, attitudes towards the feedback provider and the feedback content, we assess all three basic attitudinal components (Breckler, 1984), the cognitive (perceived ability), the affective (likability) and the behavioral component (feedback acceptance). STUDY 2 Method Participants and design. Forty-seven participants were randomly assigned to one of two power conditions (high vs. low). Emotional stability was the continuous independent variable. Procedure. The procedure was similar to Study 1 with the following exceptions. We used a different power manipulation and the feedback was not number-based but verbal. In addition to that, we assessed other-directed anger as well as self-directed anger. First, emotional stability was assessed with the emotional stability subscale of the Big 8 personality questionnaire (De Raad & Barelds, 2008; = .86). The power manipulation was introduced after task completion, in order not to influence task performance. The task in Study 2 was similar to Study 1 with the exception that all participants were assigned the supervisor role and the alleged other was assigned the subordinate role. Power was manipulated via an experiential priming procedure that asks participants in the high (low) power condition to vividly recall an incident when they had high (low) power (Galinsky et al., 2003). Participants received an email message from their alleged subordinate containing preprogrammed negative feedback concerning their presentation. They read inter alia: the information campaign will not be successful with your presentation. As in Study 1, the manipulation check consisted of a 7-point scale 4-item power measure (cf. Galinsky, et al., 2003; = .91). Affective reactions were assessed with expanded scales from Study 1. Moreover, we specified whom the anger was directed at (e.g., angry with the subordinate; = .92, or angry with myself; = .96). Attitudinal reactions were assessed on a 7point scale by three or four items each. Sample items are I think that my subordinate is able to give useful comments (perceived ability; = .92), My subordinate is a nice person (liking; = .94) and My subordinate rightfully criticized me (feedback acceptance = .72). Results In all hierarchical regression analyses main effect terms for power and emotional stability were included in Step 1. The power by emotional stability interaction term was included in Step 2. A hierarchical regression analysis on the power index revealed only a main effect of power (b = 2.37, SE = 0.27, p <.001, Mhigh power = 5.17, SD = 0.64; Mlow power = 2.85, SD = 1.11). In line with expectations, we did find interaction effects between power and emotional stability on anger towards the subordinate and shame but not on anger towards oneself. Simple

10.5464.AMBPP.2011.210.a

slope analysis revealed that only for high power individuals emotional stability yielded strong negative relationships with anger towards the subordinate (1 SD above the mean, b = -0.77, SE = 0.35, p = .034) and shame (1 SD above the mean, b = -1.03, SE = 0.30, p = .001). We also did find interaction effects between power and emotional stability on perceived subordinate ability, liking and feedback acceptance. Simple slope analyses showed that only for high power individuals emotional stability yielded a strong positive relationship to perceived subordinate ability (1 SD above the mean, b = 1.76, SE = 0.64, p = .008), subordinate liking (1 SD above the mean, b = 1.74, SE = 0.54, p = .002) and feedback acceptance (1 SD above the mean, b = 1.31, SE = 0.47, p = .008). For low power individuals these relationships were non-significant (see Figure 2). -----------------------Figure 2 about here -----------------------We tested our moderated mediation hypothesis, that anger towards the subordinate would mediate the relationship between the interaction of power and emotional stability on attitudes with a procedure suggested by Preacher, Rucker, and Hayes (2007). First, a regression analysis tests whether the interaction of power and emotional stability influences the mediator. Second, a regression analysis tests whether the independent variable, the moderator, their interaction and the mediator separately influence the dependent variable. Third, conditional indirect effects test the influence of emotional stability via the mediator on the dependent variable at different levels of power. As mentioned before, the interaction term of power and emotional stability influenced anger towards the subordinate. In step two, anger towards the subordinate significantly influenced perceived subordinate ability (b = -0.90, SE = 0.24, p < .001). In step three, bootstrapping confirmed that the positive effect of emotional stability via anger towards the subordinate on perceived subordinate ability was only present when power was high (estimate: 0.73; BCa CI: 1.30 to 0.23) but absent when power was low. Anger towards the subordinate also influenced the dependent variable subordinate liking (b = -0.64, SE = 0.22, p = .005). Bootstrapping showed that the positive effect of emotional stability via anger towards the subordinate on subordinate liking was only present when power was high (estimate: 0.50; BCa CI: 0.97 to 0.13) but absent when power was low. Anger towards the subordinate also influenced the dependent variable feedback acceptance (b = -0.53, SE = 0.19, p = .008). Bootstrapping revealed that the positive effect of emotional stability via anger towards the subordinate on feedback acceptance was only present when power was high (estimate: 0.43; BCa CI: 0.81 to 0.09) but not when power was low. To conclude, anger towards the subordinate mediates the effects of emotional stability on perceived subordinate ability, subordinate liking and feedback acceptance for high power individuals but not for low power individuals. Additional exploratory analyses did not reveal evidence for backwards mediation. Neither anger towards oneself nor shame functioned as mediators. GENERAL DISCUSSION In two studies, we have consistently shown that power moderates the effects of emotional stability on individuals affective reactions to negative feedback. Specifically, we have found that, with higher levels of emotional stability, high power individuals, in contrast to low power individuals, are less likely to experience negative emotions, such as anger and shame, in

10.5464.AMBPP.2011.210.a

response to negative feedback (Study 1 and Study 2). In line with findings emphasizing the influence of emotions on attitudes (Belschak & Den Hartog, 2009; Clore & Schnall, 2005), we have also found evidence for the hypothesis that specific emotional responses to feedback mediate the interactive effects of emotional stability and power on attitudinal feedback reactions. Anger directed towards the feedback provider (but not shame) experienced by high power individuals with lower levels of emotional stability resulted in less liking of and lower perceived ability of the feedback provider as well as in lower feedback acceptance (Study 2). The growing use of feedback at all organizational levels implies that leaders and subordinates with varying levels of power are increasingly confronted with negative feedback. Therefore, it is surprising that previous research did not acknowledge the pivotal role of the feedback receivers power on feedback reactions. By showing that leaders reactions differ from those of subordinates (Study 1), and that high power leaders reactions differ from those of low power leaders (Study 2), our findings suggest that conclusions from feedback studies with subordinates cannot indiscriminately be used to make inferences about leaders. We hope that our work stimulates researchers exploring the effects of organizational feedback processes as well as practitioners implementing feedback procedures to take the hierarchical level of the feedback receiver into consideration. ENDNOTES 1. A more elaborate version of this manuscript has been submitted for publication. Niemann, J., Wisse, B. M., Rus, D., Van Yperen, N. W., & Sassenberg, K. 2011. Know your boss: The influence of emotional stability on the affective and attitudinal feedback reactions of powerful leaders. Manuscript submitted for publication. REFERENCES AVAILABLE FROM THE AUTHORS Figure 1. Mediated Moderation Hypothesis Emotional Stability Power Figure 2. Schematic Reproduction of the Interaction Between Power and Emotional Stability on Perceived Ability, Liking and Feedback Acceptance in Study 2, N = 47. Other-directed Anger Perceived Ability, Liking, Feedback Acceptance

Copyright of Academy of Management Annual Meeting Proceedings is the property of Academy of Management and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.

Você também pode gostar