Humor is a basic element of human interaction, closely linked to well-being and affection, witnessed in the context of different social entities. Given that this construct has effects on an individual’s cognitive, affective, and behavioral responses, it may be suggested that it also influences the quality of workplace interactions (WISSE; RIETZSCHEL, 2014).

Research on humor at work was, until the 1990s, relatively limited due to the emphasis on other behavioral phenomena (WESTWOOD; JOHNSTON, 2013). In the 1990s, research indicated that humor at work can relieve stress, build group cohesion, improve communication (MEYER, 1997), stimulate creativity (BROTHERTON, 1996) and therefore promote the effectiveness of leadership (DECKER; ROTONDO, 2001), increase satisfaction (DECKER, 1987) and employee performance (AVOLIO; HOWELL; SOSIK, 1999). However, this theme remains considerably unexplored, with little theoretical and empirical basis.

There is a history of research related to the attempt to define the humor construct, However, this objective is always hampered due to the complexity of constructing the concept (MARTIN, 2001; MARTIN; PUHLIK-DORIS; LARSEN; GRAY; WEIR, 2003; ROBERT; YAN, 2007; WARNARS-KLEVERLAAN; OPPENHEIMER, 1996). For Gray and Watson (2001), humor represents a summary of a person’s affective state. Nevertheless, other researchers have defined humor as communication that has the power to evoke laughter or produce a cognitive or affective response (ROBERT; YAN, 2007).

Although this diversity of conceptions exists, most researchers agree that humor is a mechanism of communication and social interaction, producer of cognitions and emotions in an individual, group or organization, whether positive or negative (MARTIN, 1996; MARTIN; PUHLIK-DORIS; LARSEN; GRAY; WEIR, 2003; ROMERO; CRUTHIRDS, 2006; THORSON; POWELL, 1993a). Another challenge in defining humor refers to the multidimensionality of this concept, since there are quantitative differences in the way researchers presented different humor styles.

We developed this research developed based on the multidimensional conception of humor by Martin et al. (2003), in which four humor styles are defined by the intersection of two primary dimensions. The first dimension of this model involves intrapsychic styles, in which humor is used for one's own benefit, and interpersonal styles, in which the use of humor is associated with relationships with other individuals. The second dimension of the model determines the positive styles in which the use of humor is relatively benign, and negative styles, which are potentially harmful to both oneself and their interaction with others. Figure 1 indicates that different combinations of the two dimensions give rise to the four humor styles: affiliative, self-promoting, aggressive and self-deprecating.

Figure 1.
Humor Styles
Source: Martin et al. (2003). Adapted by the author.

Affiliative humor, determined by combining the dimensions of positive and interpersonal styles, is found in individuals who tend to amuse others by telling jokes and funny stories and engaging in spontaneous play, in order to facilitate relations and reduce interpersonal tensions (LEFCOURT, 2001).

Self-promoting humor, in turn, comes from the combination of the dimensions of positive and intrapsychic styles, and involves individuals with a humorous outlook on life who tend to have fun even in the face of adversity (MARTIN; KUIPER; OLINGER; DANCE, 1993). It is a defense mechanism avoiding negative emotions such as stress (MARTIN; PUHLIK-DORIS; LARSEN; GRAY; WEIR, 2003).

The combination of negative and interpersonal style dimensions gives rise to aggressive humor, which is related to the use of sarcasm, irony and teasing (ZILLMAN, 1983) without considering the potential negative impact on others, in addition to the compulsive use of humorous expressions to manipulate people through acts of offending, belittling or ridiculing (JANES; OLSEN, 2000).

Finally, self-deprecating humor is determined by combining the dimensions of negative and intrapsychic styles and, involves those individuals who tend to amuse others at their expense, make people laugh at their own weaknesses and even ridicule themselves, in order to get approval from other people (MARTIN; PUHLIK-DORIS; LARSEN; GRAY; WEIR, 2003). Although individuals with this style are seen as entertaining, there is an element of emotional neediness and low self-esteem in them that justifies the use of humor as a way of hiding negative feelings (FABRIZI; POLLIO, 1987).

Satisfaction with management is a difficult subject in the organizational behavior literature, since several authors approach it as one of the dimensions that compose job satisfaction. Locke (1976), defining job satisfaction as a positive emotional state determined by work-related agents and events, indicated that relationship and satisfaction with the manager is one of the requirements for job satisfaction.

According to Tamayo (2000), job satisfaction is influenced by external factors related to work itself, such as compensation, nature of activity, promotion possibilities, team interaction, organizational support, job stability, personal development and supervisory behavior. Martins (1984) reinforces that it is an affective variable, resulting from the degree of contentment of the individual in relation to the dimensions of their work. From a more recent conceptual perspective, Valle (2007) emphasizes that to assess job satisfaction, it is necessary to evaluate work-related cognitive aspects and their determinants, such as: rewards, distribution and centralization of power, peer-leader relationships, and individual differences. Finally, Siqueira (1995) concludes that job satisfaction is a multidimensional concept, composed of variables sensitive to workers' perceptions of their work environment, the nature of work and management policies and practices, which involve five theoretical dimensions: 1) satisfaction with colleagues; 2) satisfaction with the salary; 3) satisfaction with management; 4) satisfaction with the nature of work; and, 5) satisfaction with promotions.

The most recent definitions of job satisfaction presented their dimensions in isolation, which allows specific investigations of this variable in the context of organizations (COELHO JR., 2009). It is noteworthy that there are two usual ways to evaluate job satisfaction: the first refers to an overall measure of satisfaction, obtained by the average degree of satisfaction of the individual in relation to the dimensions evaluated; and the second refers to a measure determined by factors that compose the individual’s satisfaction with their work (SIQUEIRA, 1995; ALBUQUERQUE, 2004; MARTINS; SANTOS, 2006). As the first construct evaluated in this research refers to leaders’ humor styles, we chose to analyze just one of the facets of job satisfaction proposed by Siqueira (1995): satisfaction with management. This is the employee’s degree of contentment with their leaders regarding the organization, professional capacity and interest in the work of subordinates.

Increased competition due to globalization and therefore, the need to adapt to market demands has highlighted the need for individual performance management. In organizational studies, performance is associated with the behavior manifested by the individual in the exercise of his duties (STARBUCK, 2005), the quality and quantity of work done, as well as the cost and time spent. For Armstrong and Baron (1998), performance management is an integrated approach, development of individual capacities to achieve organizational results.

It turns out that individual performance is a variable investigated by heterogeneous and multicriteria measures, sometimes quantitative or objective, sometimes qualitative or subjective (SELLITTO; BORCHARDT; PEREIRA, 2006). There are empirical reports that refer to performance self-assessment measures (BABIN; BOLES, 1996; VISWESVARAN; ONES; SCHMIDT, 1996) and measures of performance hetero-evaluation, mainly of 360º (WRIGHT; COPRANZANO, 2000). Gerhart et al. (2000) indicate that the use of these performance perception measures allows investigating situational factors that would not normally be captured in the unilateral leader-employee assessment. It is precisely in this context that the importance of 360º evaluation is evidenced, which is performance feedback from all those who participate in the employee routine.

Previous studies have linked work humor with individual performance: humor can make routine activities more interesting, in addition to relieving pressure from the organizational environment, which favors communication with leaders and, consequently, facilitates clarification of what is expected of followers in terms of performance (AVOLIO; HOWELL; SOSIK, 1999; ROMERO; CRUTHIRDS, 2006; and VECCHIO; JUSTIN; PEARCE, 2009). Employees led by humorous individuals are shown to have higher levels of well-being in the workplace and, as a result, they present higher levels of performance (RYAN; DECI, 2000). However, there is little empirical evidence about the belief that the leader's posture and humor impact the performance of their followers. The lack of research on this topic may be related to the difficulty of measuring employee performance, which in many researches is seen as the perception of the subject regarding their own performance. Therefore, for the employee performance analysis of this research, we use the 360º performance evaluation applied in the company under analysis.

In any attempt to explain the effects of humor on the leader-follower relationship, it is important to recognize the fact that individuals can use or express humor in different ways (WISSE; RIETZSCHEL, 2014). Although empirical evidence on the effects of humor on organizations is still scarce, a growing body of research demonstrates the important role that humor can play in the context of work, especially in the interaction between leader and follower (AVOLIO; HOWELL; SOSIK, 1999; COOPER, 2008; DECKER; ROTONDO, 2001; ROMERO; CRUTHIRDS, 2006).

The literature mentions that humor at work is related to many organizational behaviors, among them, increased commitment and job satisfaction (BURFORD, 1987; COOPER, 2002; DECKER, 1987; DECKER; ROTONDO, 2001). Burford (1987) indicated that subordinates 'perceptions of their supervisors’ humors are related to the subordinate’s job satisfaction. Recent studies have found that a leader’s positive humor style is positively correlated with interpersonal relationships with employees (DECKER; YAO; CALO, 2011), affiliative commitment and follower identification (HUGHES; AVEY, 2009).

More specifically, Romero and Arendt (2011) pointed out that the employees’ affiliative humor style is positively related to the satisfaction with the co-workers and, Wisse and Rietzschel (2014) complement by clarifying that, particularly the affiliative humor, positively correlates with indicators of interpersonal relationships, for example, relationship satisfaction and social life. Therefore, individuals with this style are appreciated by their peers, since the use of this non-hostile humor can ease interpersonal tensions and create a healthy work environment (MARTIN; PUHLIK-DORIS; LARSEN; GRAY; WEIR, 2003), besides reducing social distance by generating the perception of the leader as a member of the group (ROMERO; CRUTHIRDS, 2006). Given this, we propose that:

Hypothesis 1: A leader’s humor is positively related to satisfaction with management;

Self-promoting humor is seen as a coping mechanism for coping with stress and usually, individuals who have such a style have a positive and humorous outlook on life (ROMERO; CRUTHIRDS, 2006). The same authors also point out that when this type of humor is used in organizations, the individual’s intention is to improve their image towards other group members.

In the research by Wisse and Rietzschel (2014) no statistically significant relationships were found regarding the perceived quality of the leader-follower relationship for self-promoting and self-deprecating humor styles. According to the authors, it is still unclear why only some of the styles affect the quality of interpersonal relationships and warn that this may signal fundamental differences between humor styles. Among the positive styles, the literature indicates that individuals who use affiliative humor are more adored and admired than those who use self-promoting humor, which has already been associated with reduced employee proactivity towards the leader (MARTIN, 2007; MARTIN; LASTUK; SCHERMER; JEFFERY; VERNON; VESELKA, 2012; ROMERO; ARENDT, 2011; WISSE; RIETZSCHEL, 2014; YIP; MARTIN, 2006). Therefore, we propose that:

Hypothesis 2: The self-promoting humor of the leader is negatively related to satisfaction with management;

Individuals who use aggressive humor generally aim to manipulate others through a threat (JANES; OLSEN, 2000) or ridicule and belittle someone to feel superior (ZILLMAN, 1983). Martin et al., 2003 suggest that aggressive humor is negatively related to affability and positively related to anxiety, depression and burnout.

According to Bippus (2003) and Decker and Rotondo (2001), interacting with individuals who use aggressive humor can have a particularly negative effect on the quality of interpersonal relationships. However, Cooper (2008) warned against an overly simplified perspective on the effects of workplace humor styles; the author advocated the idea that not everyone reacts similarly to different forms of humor and emphasized that negative styles do not necessarily weaken the bond between individuals. Romero and Cruthirds (2006) corroborate by indicating that mild aggressive humor can have positive functions, since seeing others being ridiculed can lead to appropriate behaviors, something constructive in cohesive teams (JANES; OLSEN, 2000). Therefore, when manifested as satire or provocation, mild aggressive humor may represent a strong scolding message with a positive tone (MEYER, 1997) or express disagreement and conflict without negative effect provided the message is delivered in a playful manner.

The literature highlights that aggressive humor is negatively correlated with interpersonal relationships and satisfaction with co-workers, and positively related to hostility towards others (CANN; ZAPATA; DAVIS, 2011; MARTIN, 2007; MARTIN; LASTUK; SCHERMER; JEFFERY; VERNON; VESELKA, 2012; ROMERO; ARENDT, 2011; YIP; MARTIN, 2006). We note that affiliative and aggressive humor styles predict the perceived quality of the relationship with the leader (WISSE; RIETZSCHEL, 2014). Therefore, we propose that:

Hypothesis 3: The aggressive humor of the leader is negatively related to satisfaction with management;

Finally, self-deprecating humor is associated with reduced social distance between leader and follower (SMITH; POWELL, 1988), once individuals ridicule themselves in trying to seek acceptance from others (MARTIN; PUHLIK-DORIS; LARSEN; GRAY; WEIR, 2003). Romero and Cruthirds (2006) indicate that leaders who moderately use self-deprecating humor often want to lower their status and become more accessible, and this can be beneficial for interpersonal relationships, because they can leave their followers at ease in the workplace. However, over time, there is a risk that what was initially regarded as funny might give the impression of lack and low self-esteem (MARTIN; PUHLIK-DORIS; LARSEN; GRAY; WEIR, 2003; ROMERO; ARENDT, 2011). Therefore, we propose that:

Hypothesis 4: The self-deprecating humor of the leader is negatively related to satisfaction with management;

Petty et al. (1984), tested the hypothesis that the more satisfied with their work, the higher would be the individual performance levels. In addition, the same authors pointed out that the variables job satisfaction and performance have been significantly related, although small in the organizational context. Therefore, it is necessary to empirically verify such predictive relationship with the addition of other variables (PETTY; MCGEE; CAVENDER, 1984).

Although there is no direct and significant relationship, the satisfaction variable is fundamental for the reduction of negative occurrences related to work performance, for example, turnover, absenteeism and stress, which would influence, even if indirectly, individual and organizational productivity (WRIGHT; KIM, 2004). Babin and Boles (1996) detected a positive correlation between satisfaction with management and performance, so that the individuals most satisfied with the manager perceived their performance more positively than the less satisfied. Thus, the last hypothesis presents satisfaction with leadership as an antecedent of job performance:

Hypothesis 5: Job satisfaction is positively related to job performance.

Figure 2
Layout of the hypotheses

AFH: Affiliative Humor; SPH: Self-Promoting Humor; AGH: Aggressive Humor;

SDH: Self Deprecating Humor; SMH: Satisfaction with Management and; PER: Individual Performance

Source: Author’s own elaboration

We applied a structured questionnaire, made available on a website, whose link was sent via email to several groups of Vale S.A employees, Brazilian multinational corporation in the mining sector. We used the non-probabilistic convenience sampling technique and the data collection period was 30 days, in which 301 responses were obtained. After checking missing values and outliers, 234 responses were considered valid, this indicated that the sample could be considered of adequate size for the research purposes, according to the empirical rule of five to ten cases per variable (HAIR JR; ANDERSON; TATHAM; BLACK, 2005), given that the collection instrument is composed of 36 assertions.

Humor Styles. To evaluate leadership humor styles, we used the Humor Styles Questionnaire by Martin et al. (2003). We requested the original scale via email to the author himself, who promptly sent the original English version and the validated Portuguese translation. Based on a review by three independent judges, the instrument was adapted to Brazilian Portuguese. It is noteworthy that the original questionnaire contains 32 items designed to measure the four humor styles (affiliative, self-promoting, aggressive and self-deprecating) and for each statement participants would rate how much they agree on their humor on a 7-point Likert scale. Given that the first construct of this research is the leader's humor, we were not able to use two statements on self-promoting humor (“If I feel upset or sad, I often try to think of something funny about the situation to make me feel better” and “If I'm alone and sad, I make an effort to think of something funny and cheer myself up”). Thus, the questionnaire consisted of only 30 assertions.

Satisfaction with management. To measure satisfaction with the manager, we chose to use the Job Satisfaction Scale, elaborated by Siqueira (1995) and newly validated by Albuquerque (2004), given its consistency and factorial invariance relative to more recent revalidations. It is a multidimensional measure, composed of 25 affirmative and intended to evaluate the degree of contentment of the worker facing five dimensions of work: satisfaction with management, colleagues, salary, promotions and nature of work. However, given that the independent variable of the survey is the humor of the leader, it was convenient to use only the five statements about satisfaction with the leadership, also classified on a 7-point Likert scale.

Individual performance. The final score of the 360 º evaluation internally applied at Vale S.A. measured the third variable, related to employee performance. In this evaluation, they are scored from 1 to 5, where 1=significant gaps and 5= internal reference, seven items: adaptability/flexibility; interpersonal skills; technical competence; problem analysis and solution; quality; focus on results and; health and safety. The final score comes from the mean of the scores of these seven items.

It is worth mentioning that three demographic questions were inserted in the questionnaire, related to gender, age and position, in order to detect the profile of the respondents. Other data such as location, company time and sector were not requested to preserve the anonymity of the research.

We coded the data and the reverse questions had their answers recoded for further statistical analysis of the psychometric properties of reliability, convergent and discriminant validity, which guarantee the quality of the model for the sample used. We performed data analysis by the Structural Equation Modeling method, in order to test the hypotheses presented, based on the PLS estimation (Partial Least Squares – Path Modeling), through SmartPLS 2.0 software.

Demographic data indicate that most participants belong to the age group of 31 to 35 years old, with 28.63% (67 respondents), followed by the age group of 25 to 30 years, with 24.79% (58 respondents). Of the 234 participants, the male gender predominated, with 65.81% (154 respondents), and only 34.19% (80 respondents) female. Although the sample presents employees of different hierarchical levels, most of the research participants act as analyst (52.14%) and engineer (25.64%). The demographic profile of the sample is shown in Table 1.

Table 1:
Demographic data

Source: Author’s own elaboration based on the data collected.

The evaluation of the measurement model consists of the analysis of reliability, convergent validity and discriminant validity. Chin (1998) clarifies that the index called composite reliability denotes the reliability of the model and must be greater than 0.7 for this property to be present. Table 2 shows the presence of the composite reliability for each latent variable: self-deprecating humor (0.81), affiliative humor (0.89), aggressive humor (0.81), self-promoting humor (0.85), and satisfaction with management (0.93).

Table 2:
Descriptive Statistics and Correlation between Constructs

Source: Author’s own elaboration based on the data collected.Note: Square root values of extracted average variance are bold (diagonally).

Hair Jr. et al. (2005) point out that the values of the average variance extracted greater than 50% denote the convergent validity of the model. In this study, all latent variables fully met this parameter: self-deprecating humor (0.62), affiliative humor (0.58), aggressive humor (0.59), self-promoting humor (0.59), and satisfaction with management (0.74)).

Finally, to evaluate the discriminant validity it should be observed whether the assertions presented have higher correlations in their respective constructs when compared to the other ones (CHIN, 1998). Table 3 shows that this parameter was also met. Another way to assess discriminant validity is to observe if the correlations between the constructs are smaller than the square root of the average variance extracted (CHIN, 1998) and, by this criterion, the discriminant validity of the model was also confirmed.

Table 3:
Cross loadings Matrix

Source: Authors’ own elaborationNote Items with the suffix “_rev” had their scales reversed (1→7, 2→6, 3→5, 4→4, 5→3, 6→2 e 7→1);

The assertive HAD1, HAD2, HAD3, HAD4, HAD7, HAD8, HAF1, HAF3, HAG3, HAG4, HAG5, HAG6, HAG8, HAP3 and HAP4 presented a low factor loading with their respective constructs and were eliminated, interspersed by repeated data processing, until the measurement model meets the necessary requirements for the measurements to be considered satisfactory for the hypothesis testing.

This model assumes that humor styles predict satisfaction with management, that employee performance is a consequence of satisfaction with management. Figure 3 shows the path indices between humor styles, satisfaction with management and job performance, to analyze the hypotheses formulated.

Figure 3
Structural Measurement Model

AFH: Affiliative Humor; SPH: Self-Promoting Humor; AGH: Aggressive Humor;

SDH: Self Deprecating Humor; SMH: Satisfaction with management and; PER: Individual performance.

*** p Value <0,001, ** p Value <0,01 and * p Value <0,05

Source: Author’s own elaboration

As shown in Figure 3, we identified a positive and significant relationship between affiliative humor and satisfaction with management (p<0.05), which does not allow us to reject hypothesis 1 (“Affiliative humor is positively related to satisfaction with management”). Contrary to expectations, self-promoting humor also had a positive and significant relationship with satisfaction with management (p<0.01), which leads to the rejection of hypothesis 2 (“self-promoting humor is negatively related to satisfaction with management”).

The relationship between aggressive humor and satisfaction with the boss was negative and highly significant (p<0.001), fact indicating the non-rejection of hypothesis 3 (“aggressive humor is negatively related to satisfaction with management”). It is noteworthy that the coefficient of paths between the constructs “aggressive humor” and “satisfaction with management “was -0.41, and that means 16.81% (R2=0.1681) of dissatisfaction with leadership was associated with the use of aggressive humor.

The result of the relationship between self-deprecating humor and satisfaction with management was negative and significant (p<0.05), which does not allow to reject hypothesis 4 (“self-deprecating humor is negatively related to satisfaction with management”). It is also worth noting that the path coefficients previously presented between the constructs indicate the contribution (R2) of each style of humor for satisfaction with management: 1.44% of affiliative humor, 8.41% of self-promoting humor, 16.81% of aggressive humor and 1.69% of self-deprecating humor.

Finally, hypothesis 5 (“satisfaction with management is positively related to job performance”) could not be rejected either, given the positive and significant relationship between satisfaction with management and job performance (p<0.05).