چگونه نمایش شادی و ناراحتی رهبری بر عملکرد پیروان تأثیر می گذارد :عملکرد سرایت احساسی و خلاقانه در مقابل عملکرد تحلیلی

Previous studies have found mixed results regarding the influence of positive and negative leader affect on follower performance. We propose that both leader happiness and leader sadness can be beneficial for follower performance contingent on whether the task concerns creative or analytical performance. This proposition was put to the test in two experiments in which leader affective display was manipulated and the performance of (student) participants was assessed. The results supported our hypothesis that a leader's displays of happiness enhance follower creative performance, whereas a leader's displays of sadness enhance follower analytical performance. Contrasting these findings with evidence for a subjective rating of leadership effectiveness, in line with an implicit leadership theory interpretation, leaders were perceived as more effective when displaying happiness rather than sadness irrespective of task type. The second study showed that the effects of leader affective displays on followers' creative performance and perceived leadership effectiveness are mediated by follower positive affect, indicating that emotional contagion partly underlies these effects.

Inevitably, people in leadership positions display their feelings—facially, vocally, and in more subtle nonverbal communication. Such affective displays may play a role in leadership effectiveness that research has only recently started to address. An important question that emerges from these recent research efforts concerns the contingencies of the effectiveness of leader displays of positive affect (e.g., a team leader in a happy mood) as compared with negative affect (e.g., a team leader in a sad mood). This is the issue that we address in the current study. In doing so, we focus both on the performance effects of leader affective displays and on their influence on subjective perceptions of leadership. We develop the propositions that the creative versus analytical nature of the performance task moderates whether the display of happiness (creative performance) or sadness (analytical performance) is more conducive to follower performance, whereas subjective ratings of leadership effectiveness are more favorable following happy than following sad displays regardless of the nature of the task. We provide experimental evidence for these propositions as well as partial evidence for the hypothesis that these effects are mediated by emotional contagion. Leadership effectiveness has been a core topic in leadership research (Bass, 2008). Leadership, by definition, implies that a leader influences one or more followers (Yukl & Van Fleet, 1992), and leader affect (i.e., moods and emotions) may be a key issue in understanding how leaders influence their followers and why leaders with equal skills and competences sometimes succeed and sometimes fail (George & Bettenhausen, 1990). The effects of leader affect on their followers are not fully uncovered yet, but critical to understand (Sy, Côté, & Saavedra, 2005). Humphrey (2002) has argued that a key leadership function is to manage the affect of followers, and that this is one of the main ways in which leaders influence performance. Thus, affect is a core issue within leadership, but unfortunately also one where our understanding is least developed. The most important criterion for leadership effectiveness is typically understood to be follower performance (Kaiser, Hogan, & Craig, 2008), and our goal in the current study is to contribute to the development of our understanding of the role of affect in leadership effectiveness by zooming in on what arguably is a key issue here: the nature of the task. We advance and test the hypotheses that leader displays of positive versus negative affect influence follower performance differently on creative versus analytical tasks, and that this effect is mediated by emotional contagion. Previous studies have shown that leader affect influences leadership effectiveness (Bono and Ilies, 2006 and Gaddis et al., 2004). However, the specific direction of this influence remains unclear. Both positive and negative leader affect have been shown to increase and decrease leadership effectiveness. We propose that this ambiguity is due to the fact that the effectiveness of leader affective displays is contingent on the kind of task that has to be performed by the followers. Our studies integrate different lines of research, and test relationships that have been unaddressed in previous studies, with the aim to contribute valuable new insights on leader affect and leadership effectiveness to the existing literature. Another aim of the present studies is to test our prediction that, despite being used interchangeably in previous research, objective (i.e., performance) and subjective (i.e., perceptions) leadership effectiveness measures may not correspond in terms of how they are influenced by leader affect.

ANOVA yielded an effect of leader affective displays on perceived leader happiness, F(2,160) = 561.08, p < .001, η2 = .88. Participants rated a leader who displayed happiness (M = 6.03, SD = 0.73) as happier than a leader who displayed sadness (M = 1.25, SD = 0.40), t(160) = 32.21, p < .001, or had a neutral display (M = 2.45, SD = 1.05, t(160) = 24.02, p < .001), t(160) = 24.02, p < .001. ANOVA also yielded an effect of leader affective displays on perceived leader sadness, F(2,160) = 334.89, p < .001, η2 = .81. Participants rated a leader who displayed sadness (M = 6.38, SD = 0.69) as sadder than a leader who displayed happiness (M = 1.50, SD = 0.76), t(160) = 25.82, p < .001, or had a neutral display (M = 3.66, SD = 1.36), t(160) = 14.36, p < .001.For an overview of the results see Table 2. ANOVA yielded a significant interaction of leader affective display × task type on performance (see Fig. 4). No other effects were found, and pairwise comparisons confirmed our predictions. Participants scored higher on the creative than the analytical task when their leader displayed happiness. Furthermore, participants scored higher on the analytical than the creative task when their leader displayed sadness. Participants' creative and analytical scores did not differ within the neutral leader display condition.ANOVA yielded a significant effect of leader affective displays on perceived leadership effectiveness. Participants with a happy leader rated their leader as most effective, followed by participants with a neutral leader, and participants with a sad leader rated their leader as least effective. In line with our hypothesis, a leader with a happy display was perceived as more effective than a leader with a sad, t(160) = 8.27, p < .001, or neutral display, t(160) = 2.96, p = .004. Moreover, a leader with a neutral display was perceived as more effective than a leader with a sad display, t(160) = 5,26, p < .001.The effect of leader affective displays on follower performance is moderated by the task type. Therefore, to test whether this effect is mediated by follower affect, we are testing a moderated mediation model (see Fig. 5). The most recent recommendation to test moderated mediation models is to compute the direct, indirect, and total effect across different levels of the moderator variable (Edwards and Lambert, 2007 and Preacher et al., 2007). Moreover, recent advancements have resulted in the advice to test indirect effects with a bootstrapping procedure, because bootstrapping does not require assumptions regarding the underlying sampling distribution (Preacher and Hayes, 2004 and Shrout and Bolger, 2002). Bootstrapping is a non-parametric test, which estimates the sampling distribution of the indirect effect and randomly samples observations with replacement from the data-set to create a larger sample from the original data. Our independent variable (i.e., leader affective displays) of three levels (i.e., happy, sad, and neutral) was dummy coded.To conduct bootstrapping analyses, we used the SPSS modmed macro provided by Preacher, Curran, and Bauer (2003, September) on their website. The moderated mediation results for the two analyses with follower happiness as a mediator are shown in Table 3. As can be seen in the mediator variable model results, a happy leader display yields significantly happier follower affective states than a sad or a neutral leader display. Second, the dependent variable model shows a significant follower happiness × task type interaction on follower performance. Third, the conditional indirect effects show moderated mediation in the creative task condition, but not in the analytical task condition. In short, a happy compared to a sad or a neutral leader affective display increases followers' happy affective states, which in turn increases followers' creative performance. Moderated mediation analyses with follower sadness as a mediator did not yield significant indirect effects.2