Back to Journal Cover Page


Volume10, Number 5
Submitted: September 21, 2004
First Revision: September 27, 2004

Second Revision: November 2, 2004
Accepted: November 2, 2004
Publication Date: November 2, 2004


Birgit Schyns
Tilburg University, The Netherlands

Karin Sanders
University of Twente, The Netherlands


A recently published study in Current Research in Social Psychology (Schyns & Sanders, 2003) focused on the relationship between mood and the perception of leadership. Although this experimental study showed a relationship between mood and the perception of management-by-exception passive, most of the hypotheses could not be confirmed. The present study tries to overcome the most important restrictions of that prior study and seeks to examine the same hypotheses using an employee sample and a different assessment of mood. Results indicate that mood and the perception of leadership are indeed connected, especially in the case of less active leadership styles. Controlling for effects of contact with the leader did not alter these results.



The effect of mood on evaluation of events has been a topic of research by Schwarz and Clore (1983). Recently, Schyns and Sanders (2003) transferred the affect-as-information framework to the perception of leadership. The authors examined the effect of mood on the perception of leadership styles such as transformational, transactional, and laissez-faire leadership. These styles are part of the leadership model put forward by Bass (1985), which has attracted a lot of attention in leadership research and can be regarded as one of the most studied models in this research area.

In short, Bass' (1985) full-range of leadership model comprises leadership styles from exceptional (transformational leadership) to effective (transactional leadership style) to laissez-faire. Transformational leadership is concerned with visions, individual consideration, stimulation, and motivation. Transactional leadership refers to an exchange view of leadership: the leader makes clear to followers what they will get for their performance. This style consists of several subscales, namely, contingent reward, management-by-exception active, and management-by-exception passive. Some authors refer to management-by-exception passive as being similar to Laissez-faire leadership (Den Hartog, Van Muijen, and Koopman, 1997). Laissez-faire means that the leader refrains from showing leadership behavior of any kind. Only recently, Antonakis, Avolio, and Sivasubraminiam (2003, p 265) labeled laissez-faire leadership as "generally considered the most passive and ineffective form of leadership" (see Lowe, Kroek, & Sivasubraminiam, 1996, for a comparison of the effectiveness of transformational and transactional leadership styles). In this paper, we label these leadership styles as more active (transformational leadership, contingent reward, and management-by-exception active) and less active (management-by-exception passive and laissez-faire).

Schyns and Sanders (2003) assumed that employees' perception of the style of leadership is related to their present mood. The hypotheses read: "The worse the mood, the more likely the perception of the leadership style as management-by-exception passive or laissez faire (H1a and b)." The hypotheses were restricted to less active leadership styles (management-by-exception passive and laissez-faire leadership) as they only found evidence in literature of an effect of negative mood on less positive evaluations (e.g., Fried, Levi, Ben-David, Tiegs, & Avital, 2000, on performance ratings). Still, they additionally examined the effect of bad mood on more active leadership styles such as transformational, and part of transactional leadership (contingent reward, management-by-exception active).

Their results only partly supported their hypotheses. Whereas the relationship between mood and management-by-exception passive was in the expected direction, the effect for laissez-faire was not significant although, again, in the right direction. Examining the effect of negative mood on transformational leadership and management-by-exception active did not lead to significant results.


However, the above-mentioned study on mood and leadership had some shortcomings. First of all, the participants of this study were students rather than employees. Consequently, they were not asked to rate real leaders but descriptions of leaders. Second, the authors conducted a manipulation by giving the students two different descriptions of leaders (transformational and transactional leaders). This manipulation, however, did not work out as well as desired, as the description did not have a significant effect on the perception of the different leadership styles. Third, the assessment of mood was not very explicit. The authors only assessed bad mood and even this measurement should have been more pronounced.

This study aims at overcoming these shortcomings. We wanted to examine the hypotheses set up by Schyns and Sanders (2003) in an organizational setting, thus questioning employees rather than students. In addition, the participants should rate real leaders rather than descriptions. Finally, we used a different assessment of mood.


Although we do not intend to re-outline all arguments presented in the prior study by Schyns and Sanders, we will give some background information on our hypotheses.

In their recent retrospect on their well-known and classic experiment on the effect of mood on the perception of life satisfaction, Schwarz and Clore (2003) report being in a very good mood when hearing about their 1983 paper being called "a modern classic". Their vivid description of the time when they did their study is full of fun memories and no talk of obstacles. They wonder themselves if they see the time more "rosy" than it actually was because of their - previously induced - good mood.

Two effects play a role with respect to the impact of mood on evaluations: First, individuals may use it as an information source on the subject to be evaluated (Schwarz & Clore, 1983). As Schwarz (1990) argues, individuals in a positive mood may recall more positive information, which leads to a positive judgment. In addition, positive mood may lead to less elaborate information processing (Schwarz, 1990, based on Frijda, 1988). The phenomenon is known: As Schwarz and Clore (2003) suggest, a positive mood may lead to a less systematic processing of information (e.g., Schwarz, Bless, & Bohner, 1991) because the situation is more likely to be seen as being "positive" and it is the "problematic" or unpleasant situation that calls for further cognitive exploration. Bless, Clore, Schwarz, Golisano, Rabe, and Wölk (1996) infer from three reported experiments that happy mood leads to a higher reliance on general knowledge structures than sad mood.

This means that happy people tend first of all to make more positive judgments and second to process information heuristically, as they deduce from their mood that the present situation is unproblematic and does not need further analysis (see also Bless et al., 1996). In an experiment in the same tradition, Gasper (2004) found that happy individuals process information more globally. This was especially true for ambiguous tasks. Of course, we can't directly transfer these experimental results to the perception of leadership, still we assume that the rating of somebody else's behavior is a rather ambiguous task, a fact that may add to our participants processing information more globally and recalling only the positive aspects of their leader's behavior.


Taking into account that being asked to rate one's supervisor makes it necessary for the participants to remember actions and behaviors of their supervisors, it is interesting to have a look at the effects mood has on their memory. Research by Levine and Burgess (1997) implies that mood does have an effect on the memory of - in this case - stories. They found that happiness enhances recall. We have to keep in mind that, in their study, the participants were already in a specific mood when they heard the story. Our case is different: We do not know what mood our participants were in when interacting with their leader (that is, during the encoding period). Our interest is in the recall period, that is, the mood the participants are in when recalling their supervisor's behavior in order to rate it on our questionnaire.

We have seen that people in a happy mood will tend to give more positive judgments, as they (a) use their mood as information on their leader and (b) are less likely to seek for more than superficial information.

A superficial judgment in the context of leadership is probably based on the prototypes people have of leaders. This makes it likely that our participants will apply their prototypes when rating their leaders to a greater extent when they are in a positive mood than when they are in a negative mood. It seems that most prototypes found in research are rather positive towards leaders (see e.g., the notion of Romance of Leadership, Meindl, 1993). This leads us to hypothesize that people in a positive mood rate their leader as more active. Consequently our first hypothesis reads: (H1) the more positive the mood of our participants, the more active leadership styles they perceive.

Additionally, because the results reported by Schyns and Sanders (2003) imply that individuals see leaders as less active when in a negative mood, we hypothesize that (H2) the more positive the mood of our participants, the less less active leadership styles they perceive.



Participants were 160 employees in a Dutch accountancy company. The mean age was 34 years (SD = 10 years; based on data from 139 participants) [1], the mean organizational tenure was 8.5 years (SD = 8.3 years; based on data from 143 participants). Sixty-five of the participants were male, 78 female (N = 143). Most of the participants had a medium level education. Two-thirds (68%) had never worked as a manager (N = 142).


The data reported here constitute one part of a larger organizational study on the perception of leadership. We report only the data relevant to this particular study. Participants indicated their mood and their perception of their actual leader. Equivalent to Schyns and Sanders (2003), we did not induce mood but worked with the naturally occurring mood of our participants.



Perception of Leadership. Participants rated their actual leaders on the transformational, transactional, and laissez-faire subscales of the Multifactor Leadership Questionnaire (MLQ; Bass and Avolio, 1990; Dutch translation by Vinkenburg & van Engen, 2002 [2]). In order to make our results comparable to those of Schyns and Sanders (2003), we used the same factor structure they used. For transformational leadership, a one-factor solution was used [3]. The internal consistency (Cronbach's alpha) for this scale was Alpha = .93 for twenty items. For contingent reward, the internal consistency was .67 for four items. For management-by-exception active, we found an internal consistency of .62 for four items. For management-by-exception passive, the internal consistency was .74 for four items. An internal consistency of .81emerged for laissez-faire leadership (four items). The answer categories ranged from 1 (almost always) to 5 (never). This means that lower values indicate a higher degree of the respective leadership style. All scores we used in our analyses were sum-scores divided by number of items.

Mood. The participants indicated their present mood on the affective well-being instrument by Warr (1990, 1994; Dutch translation: Schalk, Keunen, & Meijer). The twelve items that make up this instrument are tense, optimistic, contented, gloomy, worried, miserable, calm, enthusiastic, relaxed, depressed, cheerful, and uneasy. All items ranged from 1 = does not apply at all to 5 = applies completely. We recoded the items in a way that the higher the value the more positive the mood [4]. A factor analysis indicated a one-factor solution (screeplot criterion). Cronbach's alpha was .86. For the analyses, we again used sum-scores divided by the number of items.

Treatment of Missing Data

We excluded participants with missing data from our analysis. For the correlations, we used a pairwise deletion of cases, leading to Ns between 135 and 154. We deleted all missing data for the scales for the hierarchical linear modeling, leading to an N of 106, and the number of groups was N = 31.

Table 1 contains the means, standard deviation, internal consistencies of the scales and the intercorrelations of the scales.

Table 1: Means, Standard Deviation, Internal Consistencies, and Intercorrelations of the Scales










1. TL





2. CR






3. MBA







4. MBP








5. Laissez-faire









6. Mood









Note: TL = Transformational leadership; CR = Contingent reward; MBA = Management-by-exception active; MBP = Management-by-exception passive; *Correlation is significant at the .05-level (2-tailed); **Correlation is significant at the .01 level (2-tailed).



Preliminary Analyses

We conducted t-tests to examine whether or not men and women differed with respect to any of the scales used in this study. As we conducted six tests at the same time, an adjustment of the alpha-level was necessary (0.05/6 = 0.008). The only significant effect we found was for the mood scale, indicating that our female participants were in a better mood than our male participants (M female = 3.94 and M male = 3.67, t(136) = -2.953, p = .004).

We found no significant difference for participants who currently worked or had worked as managers and those who did or had not. No significant correlation emerged for age and any of the scales considering the adjusted alpha-level.

The absolute values of mood indicate that our participants were in a rather good mood.

Test of Hypotheses and Examination of Assumptions

The results of a correlational analysis indicate that mood is only significantly related to the perception of management-by-exception passive and Laissez-faire leadership style, which is to say that the better the mood, the less management-by-exception passive and the less Laissez-faire leadership style our participants perceive. This is in line with our expectations although we had also expected significant relationships with the transformational leadership, contingent rewards, and management-by-exception active.

In order to control for the effect that rating the same leader may have, we used hierarchical linear modeling (HLM; Raubenbush & Bryk, 1986; Snijders & Bosker, 1999), with the perception of the leadership styles of the respondent as dependent variables, and the mood of the respondents as independent variable. HLM is a statistical model for hierarchically structured data that takes into account within-group variability as well as between-group variability. It is similar to a regression model but it also includes random effects to represent the unexplained differences between groups. Using an ordinary least squares regression analysis would lead to unreliable results because employees within the same team have common influences, so that the assumption of independent observations, required for ordinary regression analyses, would be violated (Snijders & Bosker, 1999). The model fit is determined by the Chi square statistic. In addition, the final model is compared to a null model. The significant deviance from this null model indicates a good model fit and, thus, allows interpretation of the effects. The effects are unstandardized coefficients. These coefficients are significant if they are at least two times the size of the standard error.

The group size ranged from 1 to 11 members. We controlled for degree of contact with the supervisor (M = 29.6 minutes, SD = 59.51, N = 106) because we expect that the more contact employees have with their supervisor, the less they are biased by their mood when rating him/her.


Table 2 shows the results of analyses with perception of the leadership style as dependent variables and the mood of the respondents as independent variable, with the degree of contact as control variable.

Table 2. Results (Unstandardized Coefficients) of the Hierarchical Linear Models with Perception of the Leadership Style as Dependent Variable




















Model fit (Chi square)






Deviance from null model












Note: TL = Transformational leadership; CR = Contingent reward; MBA = Management-by-exception active; MBP = Management-by-exception passive; *Correlation is significant at the .05-level (2-tailed); **Correlation is significant at the .01 level (2-tailed).

The results show that the model fit of all models improved significant by adding mood to the analyses. Significant effects of mood are found only for Management-by-exception passive and for Laissez Faire: the better the mood of the respondents, the less respondents perceive their supervisor as management-by-exception passive and as laissez faire.

Although not presented in Table 2 we add the amount of contact between the respondents and the supervisor to the analyses. Because the effects of this variable were not significant and it did not change any of the effects of mood we did not present these effects in the table.


In a replication of a prior study (Schyns & Sanders, 2003), we tried to overcome some of the problems connected with that study. Most important, we used an employee sample to examine the relationship between mood and the perception of (real) leadership instead of a student sample rating descriptions of leaders.

On an individual level, we found that mood is related to the perception of management-by-exception passive and laissez-faire leadership. The better the mood of our participants, the less management-by-exception passive and laissez-faire leadership they perceive. We found a similar result when analyzing leadership in a hierarchical linear modeling. Participants saw leaders as less passively managing by exception and less laissez-faire when they were in a good mood. This seems to be in line with a kind of rose-colored-glasses-effect, that is, people see less active leadership when they are in a good mood. However, we could not find effects of mood on the more active leadership styles (transformational leadership and active transactional leadership styles). These results are in line with the findings obtained by Schyns and Sanders (2003) who also only found an effect of mood on less active leadership styles.


We did not find any significant results for the amount of daily contact leader and follower have in our analyses. We may expect that followers who have more contact with their leader are less biased by their mood, as they know the leader better. We could not confirm that assumption here. However, we do not know the quality of contact. One can imagine that a one-minute contact with the perfect leader improves the followers' mood for the day. This needs to be examined in a future study.

We also do not know at what point in time the last contact to the leader was. We could imagine that a positive contact to the leader just minutes before the questionnaire was filled in may well have affected the results of our study.

Our analyses cannot rule out that the causality of the relationship we found between mood and the perception of leadership is in the reversed direction than the one we assumed, namely, that the leader has an impact on the mood and not the mood on the perception of leadership. However, we would then expect to find an effect of the amount of contact with the leader, with more contact leading to a better mood. We did not find such a relationship in our study.

Although we could question approximately 160 participants, we could only track down the ratings for thirty-one supervisors. It is clearly an advantage of our study that we were able to control for participants rating the same leader, however, a future study should examine a larger number of leaders.

Similar to a prior study (Schyns & Sanders, 2003), we used the actual moods of our participants in order to examine our hypotheses. While this allows drawing conclusions on naturally occurring moods, we cannot account for extreme moods. In addition, the mood of our participants was quite positive.

However, our results do indicate a relationship between mood and the perception of leadership. While further research is needed in this area, companies should be aware of possible biases in followers' ratings of their leaders.


1. For some participants, no demographic information was available.

2. Credit line = Research Edition Translation performed by Claartje Vinkenburg and Marloes van Engen date June 27, 2002. Translated and reproduced by special permission of the publisher, MIND GARDEN, Inc., Redwood City, CA, 94061, USA from Multifactor Leadership Questionnaire, Copyright 1995, 2000 by Bernard Bass and Bruce Avolio. All rights reserved. Further reproduction is prohibited without the Publisher's written consent.

3. Although we aimed at using the same factor structure as Schyns and Sanders (2003), we examined the factor structure of all scales. The results confirmed the structure as used in the prior study.


4. One may argue that positive and negative mood are not different poles of the same dimension, however, a Maximum likelihood factor analysis with a varimax-rotation and a forced two-factor solution did not support this notion in our data. The two-factor solution did not yield an interpretable result.


Antonakis, J., Avolio, B. J. & Sivasuvramaniam, N. 2003. Context and leadership: An examination of the nine-factor full-range leadership theory using the Multifactor Leadership Questionnaire. Leadership Quarterly, 14: 261-295.

Bass, B. M. 1985. Leadership and performance beyond expectations. New York: The Free Press.

Bass, B. M., & Avolio, B. 1990. Transformational leadership development: Manual for the Multifactor Leadership Questionnaire. Palo Alto, CA: Consulting Psychologist Press.

Bless, H., Clore, G. L., Schwarz, N., Golisano, V., Rabe, C., & Wölk, M. 1996. Mood and the use of scripts: Does happy mood really lead to mindlessness? Journal of Personality and Social Psychology, 71: 665-679.

Den Hartog, D., van Muijen, J. J., & Koopman, P. L. 1997. Transactional versus transformational leadership: An analysis of the MLQ. Journal of Occupational and Organizational Psychology, 70: 19-34.

Fried, Y., Levi, A. S., Ben-David, H. A., Tiegs, R. B. & Avital, N. 2000. Rater positive and negative mood predispositions as predictors of performance rating of ratee in simulated and real organization settings: Evidence from US and Israel samples. Journal of Occupational and Organizational Psychology, 73: 373-378.

Frijda, N. H. 1988. The laws of emotion. American Psychologist, 43: 349-358.

Gasper, K. 2004. Do you see what I see? Affect and visual information processing. Cognition and Emotion, 18: 405-421.

Levine, L. J., & Burgess, S. L. 1997. Beyond general arousal: Effect of specific emotions on memory. Social Cognition, 15: 157-181.

Lowe, K. B., Kroek, K. G. & Sivasubramaniam, N. 1996. Effectiveness correlates of transformational and transactional leadership: A meta-analytic review of the MLQ literature. Leadership Quarterly, 7: 385-425.

Meindl, J. R. 1993. Reinventing leadership: A radical, social psychological approach. In: J. K. Murnighan (Ed.), Social Psychology in Organizations (p. 89-118), Englewood Cliffs, NJ: Prentice Hall.


Raubenbush, S.W., & Bryk, A.S. (1986). A hierarchical model for studying school effects. Sociology of Education, 59, 1-17.

Schalk, R., Keunen, A., & Meijer, T. 1995. Warr’s schalen voor welzijn en mentale gezondheid: factorstructuur en betrouwbaarheid (Warr’s scales for well-being and mental health: Factor structure and reliability). Gedrag en organisatie, 8: 116-127.

Schwarz, N. 1990. Feelings as information: Informational and motivational functions of affective states. In: E. T. Higgins & R. M. Sorrentino (Eds.), Handbook of Motivation and Cognition. Foundations of Social Behavior (p.527-562). New York: Guilford.

Schwarz, N., Bless, H., & Bohner, G. (1991). Mood and persuasion: Affective states influence the processing of persuasive communication. In: M. Zanna (Ed.), Advances in Experimental Social Psychology (Vol. 24, p. 161-197). New York: Academic Press.

Schwarz, N., & Clore, G. L. 1983. Mood, misattribution, and judgments of well-being, Informative and directive functions of affective states. Journal of Personality and Social Psychology, 45: 513-523.

Schwarz, N. & Clore, G. L. 2003. Mood as information: 20 years later. Psychological Inquiry, 14: 296-303.

Schyns, B. & Sanders, K. 2003. Mood and the evaluation of leaders. Current Research in Social Psychology, 9: 50-59 [].

Snijders, T.A.B., & Bosker, R.J. (1999). Multi-level analysis. London: Sage.

Vinkenburg, C. & van Engen, M. L. 2002. Translation of the Multifactor Leadership Questionnaire. In: Engen, M. L. van, Johannesen-Schmidt, M. J., & Vinkenburg, C. J. (2003). Transformational versus Transactional leadership as a Route to Career Advancement, Doing the Right Thing differs for Male and Female leaders. Paper presented at the European Congress of Work and Organizational Psychology (EAWOP), Lisbon, May 14-17.

Warr, P. 1990. The measurement of well-being and other aspects of mental health. Journal of Occupational Psychology, 63: 1993-210.

Warr, P. 1994. A conceptual framework for the study of work and mental health. Work and Stress, 8: 84-97.


The authors wish to thank Eveline Glansbeek for her help in gathering the data and Jule Epp for her proofreading.



Birgit Schyns ( is assistant professor at Department of Human Resource Studies, Tilburg University. Her main research interests are leadership with a special focus on followers' perception, self-fulfilling prophecies in leadership interaction, as well as self-efficacy, and organizational change processes.

Karin Sanders ( is full professor at the department of Work and Organizational Psychology, University of Twente. Her main research interests are social relations, especially solidarity behavior and its effect on organizational outcomes.