CSU The Recommendations Regarding Employee Motivation Project

Journal of Applied Psychology
2014, Vol. 99, No. 6, 1278 ?1287
? 2014 American Psychological Association
0021-9010/14/$12.00 http://dx.doi.org/10.1037/a0038083
RESEARCH REPORT
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
How Fun Are Your Meetings? Investigating the Relationship Between
Humor Patterns in Team Interactions and Team Performance
Nale Lehmann-Willenbrock
Joseph A. Allen
VU University Amsterdam
University of Nebraska at Omaha
Research on humor in organizations has rarely considered the social context in which humor occurs. One
such social setting that most of us experience on a daily basis concerns the team context. Building on
recent theorizing about the humor?performance link in teams, this study seeks to increase our understanding of the function and effects of humor in team interaction settings. We examined behavioral
patterns of humor and laughter in real teams by videotaping and coding humor and laughter during 54
regular organizational team meetings. Performance ratings were obtained immediately following the
team meetings as well as at a later time point from the teams? supervisors. At the behavioral unit level
within the team interaction process, lag sequential analysis identified humor and laughter patterns
occurring above chance (e.g., a joke followed by laughter, followed by another joke). Moreover, humor
patterns triggered positive socioemotional communication, procedural structure, and new solutions. At
the team level, humor patterns (but not humor or laughter alone) positively related to team performance,
both immediately and 2 years later. Team-level job insecurity climate was identified as a boundary
condition: In low job insecurity climate conditions, humor patterns were positively related to performance, whereas in high job insecurity climate conditions, humor patterns did not relate to team
performance. The role of job insecurity as a boundary condition persisted at both time points. These
findings underscore the importance of studying team interactions for understanding the role of humor in
organizations and considering team-level boundary conditions over time.
Keywords: humor, team meetings, interaction patterns, team performance, lag sequential analysis
2006). A recent meta-analysis concluded that future research
should explicitly target the role of humor in organizational teams
and among coworkers (Mesmer-Magnus, Glew, & Viswesvaran,
2012). On a similar note, Romero and Cruthirds (2006) proposed
that understanding the role of humor could promote effective
management. However, our understanding of humor in team settings remains limited, as previous research has often neglected the
context in which humor is produced and reacted to (Westwood &
Johnston, 2013). Although humor becomes more relevant in contemporary organizations, where especially the younger generation
of employees expects work to be fun, creative, and collaborative
(Levine, 2005; Romero & Pescosolido, 2008), research on the role
of humor in real organizational settings remains sparse (cf. Lynch,
2002; Mesmer-Magnus et al., 2012; Romero & Pescosolido,
2008).
As humor and laughter are socially embedded within the context
of ongoing teams, we focus on team meetings as a specific context
for studying humor and laughter. An estimated 11 million meetings take place during a typical workday in the United States alone
(Newlund, 2012). In addition to being an increasingly frequent
activity at work, meetings can offer a window into team dynamics
within organizations (van Vree, 2011) and as such provide a rich
context for studying humor in teams. Additionally, the team meeting may be one of the few locations where all team members
interact with one another, thereby making it the ideal context for
Why do we joke and laugh at work? (And let us hope that we
do.) From an evolutionary perspective, humor and laughter have
likely evolved as group behaviors because they promote group
cohesion (Gervais & Wilson, 2005; Van Vugt & Kameda, 2013).
Humor functions as a social lubricant (Romero, 2005) and provides an important relationship maintenance tool for group members (Fine & de Soucey, 2005; Robinson & Smith-Lovin, 2001).
We most often experience humor in the context of others: People
are 30 times more likely to laugh in a group than in isolation,
suggesting a contagion pattern of humor in group settings (Johnson, 2007). One such group setting that most of us experience on
a daily basis concerns the team context (e.g., Kozlowski & Ilgen,
This article was published Online First October 13, 2014.
Nale Lehmann-Willenbrock, Department of Social and Organizational
Psychology, VU University Amsterdam; Joseph A. Allen, Department of
Psychology, University of Nebraska at Omaha.
The initial data collection for this study was partially supported by a
grant from the German Research Foundation. We appreciate the support by
Simone Kauffeld.
Correspondence concerning this article should be addressed to Nale
Lehmann-Willenbrock, Department of Social and Organizational Psychology, VU University Amsterdam, Van der Boechorststraat 1, Amsterdam
1081 BT, the Netherlands. E-mail: [email protected]
1278
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
HUMOR PATTERNS
studying humor as well as many other team interaction processes
(Kauffeld & Lehmann-Willenbrock, 2012; Lehmann-Willenbrock,
Meyers, Kauffeld, Neininger, & Henschel, 2011). Moreover, humor in the meeting context is embedded in the temporal flow of
verbal interactions by different team members. As such, team
meetings provide an opportunity for examining what happens after
humor, in terms of the team interaction processes that get triggered
by humor at the microlevel of conversational moments over time.
Joking is often referential (i.e., team members understand a
particular joke within the group?s established culture; Fine & de
Soucey, 2005). Almost all groups exhibit some form of repeated
humor involving joking references (Holmes & Marra, 2002).
Moreover, positive, successful humor requires not only an attempt
to be funny but also a reaction by the audience (i.e., laughter).
Presumably, humor as an expression of positive affect carries the
potential to trigger team interaction patterns (cf. LehmannWillenbrock et al., 2011). Indeed, previous research suggests that
humor occurs in specific sequences of behavior, or behavioral
patterns, in which humor begets laughter and/or more humor
(Consalvo, 1989; Glenn, 1989). When a team member makes a
humorous statement or tells a joke, he or she invites others to
participate. The team member who told the joke may laugh first; if
others join, shared laughter results (Glenn, 1989). Initial humorous
phrases or jokes are often brought up again later, sparking another
laughing episode (Consalvo, 1989). Not only laughter but also
humor tends to occur in temporally contiguous bursts (Scogin &
Pollio, 1980). Ullian (1976) found that employees joked with
others just as often as they were joked with. Joking remarks
seemed to be followed by similar joking statements. Similarly,
Robinson and Smith-Lovin (2001) reported contagion effects of
humor, such that humorous statements beget more humorous statements within team interaction processes. These previous findings
suggest that humor should be examined in terms of behavioral
patterns of humor and laughter that develop within (team) interaction settings, rather than individual experiences.
In addition to identifying how humor patterns develop within
team interactions, the current study investigates how humor patterns relate to important team outcomes. Although some previous
research indicates that humor in team interactions can promote
positive team outcomes (Robinson & Smith-Lovin, 2001), we
currently do not understand how and when humor can benefit
performance in real teams. In this paper, we take first steps in this
direction by exploring how humor unfolds in team interaction
patterns and how those humor patterns relate to team performance
over time. Building on previous theorizing about humor and team
effectiveness (Romero & Pescosolido, 2008), this study seeks to
increase our understanding of how humor works in team interaction settings. As such, this study offers the following contributions.
First, we examine the role of humor in the context of organizational teams, a largely unexplored phenomenon. Second, we shed
light on humor patterns by examining real team interaction behaviors over time. Third, we identify both immediate and longitudinalrelationships between the frequency of humor patterns
and supervisors? ratings of team performance. Fourth, we examine team-level job insecurity perceptions as a potential
boundary condition for these effects. Finally, we discuss managerial implications for reaping the benefits of humor in teams.
1279
Humor Patterns in Team Interactions
Organizational scholars largely agree that humor is a basic
element of interaction (for an overview, see Romero & Cruthirds,
2006). Humor can be defined as ?any communicative instance
which is perceived as humorous? (Martineau, 1972, p. 114). This
definition implies that the humor is successful, in terms of being
perceived as amusing rather than offensive (for a discussion of
negative humor, see Malone, 1980; Meyer, 2000). In this paper, we
explicitly refer to positive and successful humor. More specifically, in the organizational setting, (successful) humor ?consists of
amusing communications that produce positive emotions and cognitions in the individual, group, or organization? (Romero &
Cruthirds, 2006, p. 59). Positive humor is distinct from mean
humor (humorous statements that are intentionally negative) or
put-down humor (i.e., sarcastic or mean comments). The latter
may still produce laughter but is not likely to produce positive
emotions. In fact, disparaging or sarcastic humor in team interactions, aimed at criticizing others, has shown negative relationships
to team productivity (Kauffeld & Lehmann-Willenbrock, 2012). In
the current study, we therefore focus exclusively on the phenomenon of positive, well-intentioned humor in teams. We specifically
refer to humor in team contexts, in line with definitions of humor
as a communicative element as well as recent theorizing about
humor as a positive team resource (Martineau, 1972; Romero &
Pescosolido, 2008).
To account for the results of humor, we need to consider the
interaction context surrounding the humorous remarks, in terms of
the sequence of behaviors during which humor occurs (Fine, 1984;
Robinson & Smith-Lovin, 2001). Understanding the sequence of
behaviors that constitute humor patterns or episodes can provide
valuable information on how team members respond to humor
attempts. Humor may also occur in the absence of patterns, such as
a humor statement without subsequent laughter (i.e., some jokes
are not funny); thus, we focus on those humor instances that create
patterns of interaction. Moreover, such patterns should be identified in order to understand how and why humor relates to relevant
team performance outcomes.
Taken together, previous findings suggest that humor and laughter occur as temporally contiguous patterns of interaction in teams
(e.g., Consalvo, 1989; Glenn, 1989; Robinson & Smith-Lovin,
2001). However, to the best of our knowledge, no previous research has actually captured the behavioral patterns assumed to
exist when humor occurs in real teams embedded in organizational
contexts. These patterns comprise sequences of behavior (humor
and laughter) that occur significantly above chance (see LehmannWillenbrock et al., 2011; Stachowski, Kaplan, & Waller, 2009, for
examples). We expect humor and laughter to form specific behavioral sequences, or patterns, within the team interaction flow.
Humor patterns could be built out of behavioral sequences of
humor and laughter or out of several humorous remarks in a row.
We hypothesize the occurrence of humor patterns as follows:
Hypothesis 1: Within team interaction processes, sequential
humor patterns develop.
Team Interaction Processes After Humor Patterns
Romero and Pescosolido (2008) proposed that successful organizational humor can enhance team communication processes. For
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
1280
LEHMANN-WILLENBROCK AND ALLEN
example, humor can enhance listeners? attentiveness and facilitate
persuasion (e.g., Gruner, 1976; Lyttle, 2001). This may be especially helpful in the context of team meetings, where team members need to build on each other?s contributions and take initiative
to develop and implement new ideas (Kauffeld & LehmannWillenbrock, 2012). Moreover, previous findings from communication research suggest that humor as a microlevel process can
help facilitate meeting interaction (Beck, Littlefield, & Weber,
2012).
Previous research suggests that positive socioemotional behavior?such as humor? enhances team interaction processes (Keyton & Beck, 2009). That is, Lehmann-Willenbrock, Allen, and
Kauffeld (2013) found that positive socioemotional behaviors, in
this case supportive statements, in team meetings sustained effective procedural communication in team meetings that is necessary
for positive meeting outcomes. Moreover, theoretical assumptions
about the benefits of humor for team communication (Romero &
Pescosolido, 2008) should hold true at the micro level of temporally sequenced team interaction behaviors as well. As such, we
expect that humor, as a type of socioemotional behavior in meetings, can serve as a trigger to subsequent functional communication. We are particularly interested in the effects of humor within
the team interaction process (i.e., in the communication instances
or lags that immediately follow humor patterns within temporal
team interaction processes). Focusing on what follows humor
patterns within such team interactions, we hypothesize
Hypothesis 2: Within temporal team interaction processes,
humor patterns trigger functional communication at subsequent conversational moments.
Humor Patterns and Team Performance
We focus on team performance as an expression of team effectiveness for two reasons. First, team productivity is often the most
salient feature of team effectiveness (Romero & Pescosolido,
2008). Second, our focus on team performance can help recover
humor from its current status as an ?unsung hero in peoples? day
to day affective [organizational] lives? (Robert & Wilbanks, 2012,
p. 1093) by highlighting the potential benefits of humor behaviors
in the workplace.
Only a few studies have examined a potential relationship
between humor and performance outcomes. Avolio, Howell, and
Sosik (1999) found a positive connection between employees?
ratings of supervisors? use of humor and managerial performance.
Similarly, Priest and Swain (2002) asked employees to recall good
or bad leaders and rate their use of humor. They found that good
leaders were reported to use significantly more humor. Although
these previous findings refer to leaders rather than teams and do
not account for the interaction context in which humor and laughter occur, they suggest that humor could foster team effectiveness.
Because humor is an essentially social, context-driven phenomenon (e.g., Robinson & Smith-Lovin, 2001), an examination of the
potential relationship between humor and team performance
should be based on observations of humor and laughter in team
interaction contexts, rather than reports of individual experiences.
This is consistent with recent research and theory concerning
interactive team cognition (see Cooke, Gorman, Myers, & Duran,
2013), which emphasizes team interaction as the key ingredient for
team processes necessary for successful performance. Similarly,
the team interaction context surrounding a humor occurrence can
play an important role. For example, sometimes when a person
tells a joke or tries to be funny, the humor attempt falls flat and no
laughing occurs (i.e., failed humor; see Romero & Pescosolido,
2008). In the absence of a laughter response, the likelihood of
continued humor attempts on the part of the individual diminishes.
This likely has a different impact on team performance than when
the humor attempt is reinforced (by laughter or additional humor),
such that a humor pattern develops. Thus, any humor?performance
relationship in team settings should be due to humor?laughter
patterns within the team interaction process (Cooke et al., 2013),
rather than the frequency of individual humor. Thus, we presume
Hypothesis 3: Humor patterns are positively related to team
performance.
The Role of Job Insecurity Climate
Although humor holds the promise of alleviating stress, reducing conflict, and promoting team performance, there may be
boundary conditions for the positive effect of humor in team
settings (see also Romero & Pescosolido, 2008). Team-level perceptions of job insecurity present one rather salient boundary
condition or moderating factor, especially during challenging economic situations for organizations. Perceptions of job insecurity
have become particularly salient in recent years with organizational downsizing (Kivim?ki, Vahtera, Elovainio, Pentti, & Virtanen, 2003), economic struggles (Irwin, 2013), and other sources
of concern for employees? job security in the long term (Sverke,
Hellgren, & N?swall, 2002). In essence, it may be particularly
difficult for teams to enjoy the benefits of humor in their team
interactions when they are constantly aware of the tenuous nature
of their employment situation.
Perceived job insecurity has been defined as employees? ?concern about the future permanence of the job? (van Vuuren &
Klandermans, 1990, p. 133). Employees who are concerned about
the future of their job are often preoccupied with this concern such
that deteriorating psychological health, job withdrawal, and negative employee attitudes typically follow (e.g., Debus, Probst,
K?nig, & Kleinmann, 2012; Dekker & Schaufeli, 1995; Huang,
Zhao, Niu, Ashford, & Lee, 2013; Mohr, 2000; Sverke et al.,
2002). Meta-analytic findings also show that job insecurity negatively relates to performance outcomes (Cheng & Chan, 2008).
Previous research, however, has focused almost entirely on job
insecurity at the individual level (see Sora, De Cuyper, Caballer,
Peir?, & De Witte, 2013, for an exception), and some previous
(individual-level) findings suggest that job insecurity limits the
benefits of positive resources. K?nig, Debus, H?usler, Lendenmann, and Kleinmann (2010) found that employees? perceived
communication quality was more strongly related to self-rated
performance when job insecurity was low rather than high. Because humor can be considered a positive resource (e.g., Robert &
Wilbanks, 2012), we argue that job insecurity can function as a
boundary condition in the humor?performance relationship as
well. That is, under conditions of high job insecurity, humor will
no longer positively relate to performance, because the insecurity
becomes a job demand that absorbs the resource benefits of positive workplace interactions in the form of humor in meetings.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
HUMOR PATTERNS
When operationalized as team job insecurity climate (Sora et al.,
2013) and building upon organizational support theory (Rhoades
& Eisenberger, 2002), job insecurity is believed to be a good
indicator of the lack of feelings of support and would likely inhibit
the positive benefits from humor on performance behavior. People
individually and in teams would filter their behavior more readily
in order to feel more in control and avoid making interpersonal
mistakes that could justify termination from the job. Specifically,
if a team was constantly aware of the fact that its members? jobs
were tenuous at best (i.e., high job insecurity), jokes in that context
would likely take on a different meaning. Instead of facilitating
team unity and performance, they may heighten awareness of
organizational concerns relative to the economy, thus further distracting team members from task performance. Thus, we would
expect that in a secure job climate, humor is beneficial to team
performance, whereas in an insecure job climate, humor may not
relate to team performance as strongly. Our final hypothesis posits
Hypothesis 4: Job insecurity climate moderates the relationship between humor patterns and team performance, such that
the positive relationship between humor patterns and performance is stronger when job insecurity climate is low (i.e.,
when teams feel secure in their jobs).
Method
Data were gathered during a large longitudinal field research project. Participating teams were situated either in the manufacturing
departments or in the assembly line process of two industrial organizations in Germany. Although the research project was fully endorsed
by the union councils and company management, participation was
voluntary. Participants were guaranteed that their data would remain
confidential at all times during and after the project.
Sample
A total of 54 teams (N ? 352 employees) participated in this study.
Participants? age ranged from 18 to 59 years (M ? 34.99, SD ?
10.85). Ninety percent of the sample was male, which is characteristic
of these industrial settings. Organizational tenure ranged from 0 to 44
years (M ? 10.70, SD ? 9.14). The majority of the participants had
completed an apprenticeship (79.3%). 4.9% had a college or university degree, and 4.1% had no vocational training.
Coding Humor and Laughter in Meeting Interaction
To obtain behavioral data on the use and effects of humor, we
videotaped regular team meetings (one meeting per team). These
meetings took place as part of the continuous improvement process
(CIP; e.g., Liker, 2006) implemented in both organizations. As part
of the CIP, teams regularly meet at least once a month for approximately one hour to discuss their work process, identify any
problems or obstacles, and come up with ideas for improvement.
These meetings are organized by the teams themselves. Supervisors are generally not present during the meetings. We asked
participants to ignore the camera, which was placed at the end of
their rectangular meeting tables in order to cause as little distraction as possible. Observations such as negative remarks about
(absent) supervisors or participants leaving the room during the
meeting indicate that the videotaping was largely ignored by
1281
participants. The length of the meetings ranged from 20 to 65
minutes (M ? 47.41, SD ? 10.31). Team meeting interactions
were coded with the act4teams team interaction scheme and
INTERACT software (Mangold, 2010). Within the act4teams
scheme, we focused our analysis on codes for problem-solving
behavior, positive procedural behavior, and positive socioemotional statements. Humor is a distinct behavioral category situated
in the positive socioemotional facet of team communication. Negative humor such as put-down humor or sarcasm is coded with a
different behavioral code (?criticizing?; see Kauffeld & LehmannWillenbrock, 2012, for more details on the coding scheme). A
subset of our data was coded twice by separate coders, in order to
establish interrater reliability (? ? .81).
Survey Measures
We asked the teams? supervisors to rate team performance immediately after the meeting (t1) and again approximately two years later
(t2), using the following items adapted from Kirkman and Rosen
(1999): ?The team reaches their (quantitative) targets?; ?The team
exceeds their qualitative targets?; and ?The team continuously improves their efficiency? (Cronbach?s ? ? .63 at t1 and ? ? .68 at t2).
Job insecurity was measured in a reduced sample from one of the two
organizations (N ? 29 teams) after the meeting with three items
focusing on perceptions of the likelihood of losing one?s job (Borg &
Elizur, 1992; see also Staufenbiel & K?nig, 2011). A sample item was
?Thinking of losing my job makes me worry? (Cronbach?s ? ? .93;
rwg ? .82 across all teams). For all survey items, the answering format
ranged from 1 (completely disagree) to 7 (completely agree).
Control Variables
We controlled for demographic characteristics of the teams
(number of women in each team, age, and team members?
average organizational tenure) as well as the team size in the
meeting and the length of each team meeting. Moreover, we
controlled for the organization (coded as 0 or 1) in all analyses.
Analysis Strategy
Upon coding the videotaped meetings, we performed a lag
sequential analysis to identify potential humor patterns, using
INTERACT software. Lag sequential analysis analyzes behavioral
interdependencies and temporal patterns in sequentially recorded
events of groups or individuals (e.g., Bakeman & Quera, 2011; see
also Lehmann-Willenbrock et al., 2011, 2013, for applications of this
method). We generated interaction sequence matrices to determine
how often one behavior was followed by another. Next, we calculated
transition frequencies for each pair of behaviors. Lag1 transitions
occur when one behavior directly follows the previous one. Lag2
transitions occur when a behavior is followed by the next-but-one
behavior. Based on the frequency matrix of these transitions at lag1
and lag2, we derived transition probabilities that indicate the probability that a specific behavior B occurs after a particular given behavior A within the interaction process (Benes, Gutkin, & Kramer, 1995).
In other words, they describe the likelihood that behavior B is triggered by A. To test whether any transition probability differed significantly from the unconditional probability for the event that followed, we calculated z values for lag1 and lag2. At either lag, z values
LEHMANN-WILLENBROCK AND ALLEN
1282
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
larger than 1.96 or smaller than ?1.96 indicate that the respective
sequence is statistically significant. In general, positive z values imply
that a behavior promotes the likelihood of the subsequent other
behavior, and negative z values imply an inhibitory effect (from one
behavior to the next at lag1; from one behavior to the next but one at
lag2). Afterward, we calculated the overall frequency of humor patterns per team and tested longitudinal relationships with team performance via regression analysis. Finally, we calculated an interaction
term between the number of humor patterns per team and the aggregated measure of perceived job insecurity to test for moderating
effects. All analyses were performed at the team level.
Results
Lag Sequential Analysis
Across all team meetings, we identified statistically significant lag1 sequences: humor?laughter (z ? 77.83), laughter?
humor (z ? 26.87), and humor? humor (z ? 17.58, p ? .01,
respectively). Moreover, the lag2 sequence humor? . . . ? humor
was statistically significant (z ? 23.39, p ? .01). Taken together, these findings represent the following humor patterns:
humor?laughter? humor as well as humor? humor? humor. Hypothesis 1 was supported.
Lag0
Upon establishing the hypothesized humor patterns in the data, we
recoded our pooled data set (across all team meetings) such that
humor patterns represented a single behavioral event. Next, we ran a
lag sequential analysis to explore the effects of humor patterns on
subsequent behaviors within the team interaction process. As depicted
in Figure 1, humor patterns significantly triggered several important
problem-solving behaviors both at lag1 and at lag2. After humor
patterns, procedural behaviors such as procedural suggestions (z ?
4.53), goal orientation (z ? 2.71 and z ? 3.71), or summarizing (z ?
4.60) were significantly more likely. Positive socioemotional behaviors were also triggered by humor patterns at lag1 (z ? 3.66 for
offering praise; z ? 3.66 for encouraging participation). Of importance, at lag2, humor patterns also promoted statements about new
ideas or solutions (z ? 3.18) as well as questions (z ? 2.66). Although
these findings are exploratory in nature, they indicate that humor
patterns indeed increased functional communication behaviors within
the team interaction process, thus lending support to Hypothesis 2.
Regression Analyses
Table 1 shows the means, standard deviations, and intercorrelations of all variables. In order to connect the sequential
analysis results to our outcomes of interest, we c