The impact of status hierarchy on individual behavior and team processes
Doornenbal, B.M.
2021
document version
Publisher's PDF, also known as Version of record
Link to publication in VU Research Portal
citation for published version (APA)
Doornenbal, B. M. (2021). The impact of status hierarchy on individual behavior and team processes.
General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain
• You may freely distribute the URL identifying the publication in the public portal ?
Take down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
E-mail address:
CHAPTER 2
STATUS HIERARCHIES IN TEAMS:
THE INTERACTION BETWEEN STEEPNESS AND SKEWNESS
ABSTRACT
Understanding the impact of status hierarchy can help to improve the performance of teams. Scholars proposed to increase the understanding of the status hierarchy – team performance relationship by reconsidering the conceptualization and measurement of hierarchy, and by unraveling mediators that explain how status hierarchy affects teams. In this chapter, we advance the conceptualization and measurement of status hierarchy and show that this advancement helps to understand team processes and – in turn – team performance. Our findings suggest that the impact of hierarchy depends on the shape of the hierarchy. We find that pyramid-shaped hierarchy ( ) inhibit team performance by reducing information elaboration. In reconsidering the conceptualization and measurement of hierarchy, we introduce status hierarchy skewness as a measure for hierarchy concentration. Hierarchy skewness overcomes a shortcoming of existing hierarchy concentration measures. Ultimately, this chapter advances the understanding of when and how status hierarchy affects team performance.
INTRODUCTION
A key feature in organizations is status – that is, the esteem, prestige, and respect held by individuals (Anderson et al., 2006). Status affects the behavior of individuals (Anderson et al., 2015). Moreover, the distribution of status in teams affects team processes and outcomes (Halevy et al., 2011; Magee & Galinsky, 2008). The hierarchy literature, which conceives the distribution of status as status hierarchy, demonstrated that status hierarchy emerges naturally and is ubiquitous to teams (Berger et al., 1972, 1977; Magee & Galinsky, 2008). Almost every team thus has a status hierarchy that affects the functioning of the team.
Although a large part of the hierarchy literature argues that status hierarchy has a positive impact on team processes (Anderson & Willer, 2014; Halevy et al., 2011), most empirical evidence suggests that status hierarchy has a negative impact (Greer et al., 2018). Scholars argue that the impact of status hierarchy might depend on contingencies and hurt some team processes more than others (Bunderson et al., 2016; Greer et al., 2018). However, it is still unclear when and how status hierarchy affects team performance (cf. Anderson & Brown, 2010; Bunderson, Van der Vegt, Cantimur, & Rink, 2015). This lack of clarity is problematic because it restricts our understanding of how to deal with status hierarchy in teams.
To advance the understanding of the relationship between status hierarchy and team performance, scholars proposed to reconsider the way in which hierarchy is conceptualized and measured (Bunderson et al., 2016). Status hierarchy is often conceived as disparity, a conceptualization in which the largest hierarchy has one team member with a maximum level of status while the rest of
the team has a minimum level of status (Harrison & Klein, 2007). However, studies often measure a single hierarchy property (Bunderson et al., 2016; Greer et al., 2018), mostly either how widely status levels are distributed (i.e. hierarchy steepness) or how asymmetrically status levels are distributed (i.e. hierarchy concentration). Often applied measurements of status hierarchy might thus yield in an incomplete understanding of the impact of status hierarchy. In reconsidering the conceptualization and measurement of hierarchy, scholars recommend to examine mediators that transmit the impact of hierarchy to team performance (Bunderson et al., 2015). Examining these mediators provides insight into the mechanisms through which status hierarchy affects teams (cf. Anderson & Brown, 2010; Bunderson, Van der Vegt, Cantimur, & Rink, 2015).
Building on these research recommendations, the aim of this chapter is to clarify when and how status hierarchy relates to team performance. First, we intend to clarify when status hierarchy affects team performance by reconsidering the conceptualizations and measurements of hierarchy. More specifically, we show that the conventional focus on either hierarchy steepness or hierarchy concentration should be moved towards focusing on the combination of these properties. Focusing on this combination is needed because it can help to measure the concept of disparity. While combining steepness and concentration, we introduce hierarchy skewness as a hierarchy concentration measure that overcomes some shortcomings of conventional concentration measures (e.g. Gini-coefficient). Second, we advance the understanding of how status hierarchy relates to team performance by examining the mediator information elaboration – that is, the exchange, discussion, and integration of individuals’ input (Van Knippenberg et al.,
2004). Although the domains of information elaboration (i.e. exchange, discussion, and integration) are theorized to transmit the impact of hierarchy to team
performance, little empirical evidence exists for this indirect relationship (Bunderson et al., 2015). A better understanding of the impact of status hierarchy on information elaboration is important because a free flow of information is vital to the success of teams in modern organizations – mostly knowledge intensive teams performing non-routine tasks (Mesmer-Magnus & DeChurch, 2009; Resick et al., 2014; Van Knippenberg et al., 2004). The conceptual model (see Figure 1) is tested on the basis of multi-source data obtained from 127 student teams.
Figure 1: Conceptual model.
THEORY Status hierarchies
Individuals ascribe status to each other (Anderson et al., 2015). More status is ascribed to those who are perceived to have more instrumental social value (Anderson et al., 2015), regardless of whether they actually possess it (Correll & Ridgeway, 2006). Because individuals naturally ascribe each other status, status hierarchies emerge naturally and are ubiquitous to teams (Magee & Galinsky, 2008). A well-established stream of literature has shown that status hierarchy affect
individuals (cf. Anderson et al., 2015). Beyond the effects on individuals, status hierarchy is important for the functioning of teams (Anderson & Willer, 2014; Magee & Galinsky, 2008).
Status hierarchy is important for team functioning as it affects the use of team resources (Anderson & Willer, 2014; Halevy et al., 2011). More specifically, status hierarchy results in more attention for higher-status individuals (Anderson & Willer, 2014; Bunderson et al., 2016) – those who are perceived to have more instrumental social value (Anderson et al., 2015). Scholars argue that that a maximum hierarchy has one team member with a maximum level of status while the rest of the team has a minimum level of status (Harrison & Klein, 2007). Individuals have – on average – the greatest status disadvantage in such teams with maximum disparity (Harrison & Klein, 2007).
Status hierarchy as the combination of steepness and skewness
Although scholars elaborately proposed a concept of hierarchy (as
disparity), often applied hierarchy measures are deficient in measuring the concept of disparity. Disparity is conceived as “differences in concentration of valued social assets or resources such as pay and status among unit members – vertical differences that, at their extreme, privilege a few over many.” (Harrison & Klein, 2007, p. 1200) Disparity is greater when the distribution of valued social assets or resources is positively skewed, with one member at highest endpoint of the continuum of the valued social assets or resources and others at lowest (see configuration at the top right of Figure 2). Hierarchy measurements typically focus on either hierarchy steepness – that is, how widely status levels are distributed
among team members (Anderson & Brown, 2010) – or hierarchy concentration – that is, how asymmetrically status levels are distributed among team members (Bunderson et al., 2016). Steepness and concentration measurements are incomplete measurements of the concept of disparity, at least on their own. Hierarchy steepness, which focuses on the average difference in status held by individuals, is greater when the gap between low-status positions and high-status positions is larger (Anderson & Brown, 2010). Steepness thus neglects the portion of low-status team member. Hierarchy concentration, which focuses on the extent to which status is asymmetrically distributed within the team, is greater when the concentration is located at the bottom of the hierarchy – when there are more low-status team members (Bunderson et al., 2016). Concentration thus neglects the degree to which low-status individuals are disadvantaged in status.
In addition to the shortcoming overlap between the concept and measurement of hierarchy, conventional hierarchy concentration measurements have an important shortcoming. As shown in Table 1, conventional concentration measures (i.e. the Gini-coefficient and the Freeman-coefficient) do not accurately account for the location of concentration. The concentration can be at the top of a hierarchy (i.e. hierarchies shaped like inverted pyramids ), when there are more high-status team members, or at the bottom, when there are more low-status team members (i.e. pyramid-shaped hierarchies ). Unless concentration is at its maximum value (Dawson, 2011), a higher value according to conventional concentration indices does not suggest that the hierarchy is more concentrated at the bottom. In view of this deficiency, we introduce hierarchy skewness as concentration measure. A more positive skewness suggests a concentration at the
bottom of the hierarchy, whereas a more negative skewness suggests a concentration at the top. In contrast to conventional concentration measures (Bunderson et al., 2016), skewness is statistically distinct from steepness. More specifically, teams can have a different skewness while being similar in steepness, and can have a different steepness while being identical in skewness (see Figure 2).
Figure 2: Status hierarchy configurations.
Note. Hierarchy concentration measured as hierarchy skewness. A larger
concentration is depicted by a more positive skewness. The numbers in the circles denote individual’s status level.
To measure the concept of disparity, we propose a combination of steepness and skewness measurements. As illustrated in Figure 2, hierarchies with an identical level of steepness can vary considerably in their skewness, ranging from high status positions held by a small team subset (i.e. positive skewness), by half of the team (i.e. no skewness), or by the majority of the team members (i.e. negative skewness). Likewise, hierarchies with an identical skewness can vary
considerably in their steepness, ranging from a small status gap (i.e. low
steepness) between low-status team members and high-status team members to a wide status gap (i.e. high steepness) between them. However, the combination of steepness and skewness suggests the degree of the status differences among individuals and the extent to which a smaller part of the team holds most of the status. Therefore, the combination of steepness and concentration measurements can help to measure the concept of disparity. Specifically, greater disparity has a larger steepness and a more positive skewness.
Towards a conceptual model of the impact of status hierarchy
Examining the combination of different hierarchy properties is in line with previous efforts. Scholars have studied contextual moderators of the impact of hierarchy, such as team size (Greer et al., 2018), task interdependence (Ronay et al., 2012), and task complexity (Bunderson et al., 2016). More recently, scholars have studied the extent to which characteristics of hierarchy themselves can interact to predict team performance (Anicich et al., 2016). Instead of focusing on combining different hierarchical bases (i.e. power hierarchy and status hierarchy), we focus on combining different hierarchy measurements.
In studying the impact of the combination of steepness and skewness, we focus on the effect on team performance through information elaboration. Previously, studies demonstrated that the impact of hierarchy is mediated through processes such as conflict and coordination (Anderson & Willer, 2014; Bunderson et al., 2016). In different studies, scholars argued that the separate domains of information elaboration – i.e. the exchange (Bunderson et al., 2016), the
discussion, and the integration of information (Greer et al., 2018) – mediate the influence of status hierarchy on team performance. Consistent with these
suggestions, we will subsequently introduce information elaboration as a mediator that transmits the impact of the combination between steepness and skewness on performance.
In previous efforts to increase the understanding of the relationship between status hierarchy and team performance, studies have identified
moderators and mediators separately. These studies suggest that moderators and mediators explain when and how status hierarchy affects team performance. In this chapter, we examine when and how status hierarchy affects team performance using a moderated mediated model. Within this model, we expand on the moderators and mediators that have been studied thus far.
The joint impact of hierarchy steepness and skewness on information elaboration We propose that the interaction between steepness and skewness affects teams through information elaboration. Previous research demonstrates that lower-status team members are getting less speaking time during the exchange of information (Bales et al., 1951) and that their input both receives less attention during team discussions (Buzaglo & Wheelan, 1999) and is less often utilized in the integration of information (Anderson & Kennedy, 2012; Wittenbaum & Bowman, 2005). When the steepness is greater and the hierarchy is more positively skewed (i.e. pyramid-shaped ), teams have more low-status members who are more disadvantaged in status. If lower-status members are more ignored during team
processes, less information elaboration can be expected when the hierarchy is greater.
Less information elaboration can be expected in hierarchies with both a higher steepness and a more positive skewness for several reasons. First, larger steepness motivates individuals to neglect the input of the low-status members because these members are perceived to be less competent (Barton & Bunderson, 2014; Correll & Ridgeway, 2006). When the hierarchy is more positively skewed, more team members have a lower status level and are thus perceived as less competent. Second, larger steepness withholds low-status individuals from critically discussing the input of high-status individuals to prevent a status competition (Anderson et al., 2006; Edmondson, 2002). When the hierarchy is more positively skewed, more team members will refrain from criticizing the input of high-status members because more members will try to prevent a status competition. Third, larger steepness leads to less input from the low-status team members because their lower status is associated with a lower level of commitment and responsibility (Kennedy & Anderson, 2017; Willer, 2009). When the hierarchy is more positively skewed, fewer individuals will actively provide input because more members will feel less committed and less accountable.
In contrast to this negative impact, hierarchies with a lower steepness and/or a more negative skewness (i.e. shaped like inverted pyramids ) are less likely to hinder information elaboration. First, lower steepness is less likely to hinder information elaboration because the lower steepness implies that the input of more team members is perceived as equally valuable (Barton & Bunderson, 2014; Correll & Ridgeway, 2006). When the hierarchy is more negatively skewed, the
input of more team members is considered valuable. Second, lower steepness leads to fewer team conflicts (Bunderson et al., 2016; Cantimur et al., 2016), which are argued to hinder information elaboration (Van Knippenberg et al., 2004). When the hierarchy is more negatively skewed, we expect fewer conflicts that hinder information elaboration because fewer members are disadvantaged in status (Harrison & Klein, 2007). Third, we argue that lower steepness is better for information elaboration because lower steepness is associated with solidarity and closeness among team members (Bottero & Prandy, 2003; Locke, 2003). When the hierarchy is more negatively skewed, the status differences that are present are compressed in such a way that most are about equally high (and not
disadvantaged) in their status-level. Hence, we hypothesize the following relationship.
Hypothesis 1: The negative association between hierarchy steepness and information elaboration is moderated by hierarchy skewness, such that the negative association is stronger when the hierarchy is more positively skewed.
Information elaboration and team performance
We expect that the impact of status hierarchy on information elaboration to be transmitted to team performance. Team performance suffers from impaired information elaboration when teams have a wider information distribution (Mesmer-Magnus & DeChurch, 2009), conduct more complex tasks (Van Knippenberg et al., 2004), and operate in more turbulent environments (Resick et al., 2014). These teams suffer from impaired information elaboration because they
rely on their ability to combine the input from the team members (Mesmer-Magnus & DeChurch, 2009; Resick et al., 2014; Van Knippenberg et al., 2004). By
elaborating on the input of the team members, the most valuable resources can be used for the tasks at hand. When information elaboration is impaired, teams are less likely to find the most valuable resources. Therefore, we expect that the interaction between steepness and skewness has a negative influence on team performance through lower levels of information elaboration. Hence, we hypothesize the following relationship.
Hypothesis 2. The negative relationship between hierarchy steepness and team performance is mediated by information elaboration and moderated by hierarchy skewness, such that steepness impairs information elaboration and – in turn – team performance more when the hierarchy is more positively skewed.
METHODS Data and sample
Data for this survey study were gathered among 181 student teams consisting of 716 first-year bachelor students enrolled in a 10-week Organizational Behavior course at a Dutch university. We received 681 (95.1%) surveys of which we deleted both the incomplete teams and the teams in which team members did not grant permission to use their response for scientific research. Furthermore, we deleted one team that consisting of two individuals because these teams cannot vary in hierarchy concentration. As a result, the conceptual model was tested among 508 students (M age = 18.6, SD age = 1.2, women = 65.6%) working in 127
teams (M = 4.0, SD = 0.2). To allow the teams to negotiate their hierarchy positions, surveys were distributed three weeks before the deadline of their team assignment.
In examining the conceptual model, the present study context was chosen for three reasons. First, information elaboration was a central aspect of the team assignment. The assignment existed of both an individual and a team part. In the individual part, each team member was asked to analyze a social situation of a television show. In the subsequent team part, each team was asked to discuss the differences and similarities in their individual analyses, integrate these differences and similarities into an overall analysis, and conduct an additional team analysis. The teams received a collective grade for the team part. Second, power hierarchy – that is, vertical differences between team members’ control of valued resources (French & Raven, 1959) – is unlikely to be present in student teams because they commonly have no clear power markers (e.g., job titles and office sizes) or differences in formal power (e.g., reward power, legitimate power). The negligible power hierarchy is beneficial in studying the impact of status hierarchy because this impact can be moderated by power hierarchy (Anicich et al., 2016). Third, the teams are likely to have a salient status hierarchy. Given that students were randomly assigned to teams and data were gathered during the first course of a Bachelor program, it is likely that students did not know each other before participating in this study. This situation is beneficial in studying status hierarchy because teams are likely to develop prominent status hierarchies in the early team phases (Greer et al., 2018).
Measures
To measure individual-level status, we used a round-robin measure that consisted of 3-items developed by Bendersky and Shah (2013). Each team member was asked to rate the amount of prestige, respect, and esteem they assigned to their fellow team members ( = .90) on a 7-point scale (1 = “very little”, 7 = “very much”). Individuals’ status levels were obtained by mean-aggregation. We assessed whether this aggregation was justified by computing rwg (James et al.,
1984) and interclass correlation coefficient (ICC) scores (Bliese, 2000). These assessments indicated a median rwg of .99 (M = .98, SD = .04) and ICC(1) scores .45 and .77 ICC(2) (F = 4.32, p < .01), which justifies the aggregation.
In line with a convention in measuring the steepness of a hierarchy (Bunderson et al., 2015), we measured hierarchy steepness as the within-team standard deviation of individuals’ status. To compute steepness, the following formula was used:
where is the status scores of the individual members in the team, is the mean status scores within the team, and n is the number of team members in the team.
Although conventional concentration measures are the Gini coefficient (e.g., Christie & Barling, 2010; He & Huang, 2011) and the Freeman index (Bunderson et al., 2016), these measures suffer from two main problems. First, these measures are statistically highly related to conventional steepness measures (Bunderson et al., 2016), resulting in a low discriminant validity between concentration and steepness (Campbell & Fiske, 1959).
Bunderson et al. (2016) reported a minimal correlation of .63 between
concentration and steepness measures, and in our sample the correlation between these measures is even .98. Second, conventional concentration measures fail to accurately capture both how asymmetrical status is distributed among team members and whether the concentration is located at the bottom or at the top of the hierarchy. As illustrated in Table 1, teams that have a similar concentration according to the Gini coefficient and Freeman coefficient can have a very different asymmetry in their status distribution, ranging from very positively skewed (i.e. concentration at the bottom) to very negatively skewed (i.e. concentration at the top).
To overcome these shortcomings, we measure concentration using the skewness coefficient. To compute skewness, the following formula was used:
where is the status scores of the individual members in the team, is the mean status scores within the team, s is the standard deviation of the status scores within the team, and n is the number of team members in the team. This measure has two main advantages over conventional measures. First, as both previously described based on Figure 2 and as indicated by a correlation of -.39 in our sample, skewness is empirically distinct from steepness. Second, as a (more) positive skewness indicates that there is a large(r) amount of low-status team members whereas a (more) negative skewness indicates that whereas there is a large(r) amount of low-status team members, skewness captures more accurately where
the concentration is located. In our sample, the hierarchy skewness ranged from -2.0 to +-2.0.
Table 1: Status skewness across teams with similar concentration. Coefficient Team Individuals’
status level
SD Gini Freeman Skewness
1 3 – 3 – 3 – 5 1.0 .07 2 + 2.00
2 3 – 6 – 6 – 6 1.5 .07 1 – 2.00
3 1 – 7 – 7 – 7 3.0 .14 2 – 2.00
4 1 – 1 – 1 – 7 3.0 .30 6 + 2.00
Note. Indices calculated based on 4-person teams. The standard deviation (SD) refers to the hierarchy steepness, whereas the Gini coefficient, the Freeman coefficient and the Skewness Coefficient refers to hierarchy concentration. Formula 15 from Biemann and Kearney (2010) is used to calculate the Gini coefficient. The Freeman coefficient indicates the Freeman’s degree centralization index (Freeman, 1979). The formula reported by Sinha et al. (2016) was used to calculate the Skewness coefficient.
We assessed information elaboration using a seven-item scale adapted from Van Knippenberg, Van Ginkel, and Giessner (2014). This scale integrates items previously used (e.g., van Dick et al., 2008), and ensures content validity by tapping into underlying domains of the exchange (e.g., “we often exchange our ideas and perspectives on the team assignment”), discussion
(e.g., “we often discuss the content of the team assignment”), and integration (e.g., “we often integrate viewpoints of multiple team members in making decisions”) of information ( = .90). Respondents were asked to rate these items on a 7-point scale (1 = “very little”, 7 = “very much”). Both within-team agreement and the ICC scores were sufficient to justify mean-aggregation (rwg: Mdn = .92, M = .92, SD =
.19; ICC(1) = .16, ICC(2) = .42; F = 1.74, p < .01).
We measured team performance using the grades that teams received for their team assignment. These grades could range from 1.0 (low) to 10.0 (high). These team assignments were assessed on a variety of indicators, such as the completeness of their analysis and the extent to which they gave accurate examples of the analyzed team behavior (e.g. stereotyping). The assignments were assessed by one of five professors or teaching assistants involved in the course. No significant grading differences were found among the assessors.
In examining our research model, we identified two factors that relate to the focal variables in our study. More specifically, both the team size and the average status scores are statistically related to our dispersion measures. First, dispersion is less biased when the team size is larger (Biemann & Kearney, 2010). Second, a more extreme mean-score and a smaller team size reduces the statistically possible dispersion (Lindell & Brandt, 2000). The effects of our dispersion measures could thus be affected by team size and average status. In line with suggestions of Bernerth and Aguinis (2015), we therefore control for these variables.
RESULTS
Table 2 describes the means, standard deviations, and the correlations among the variables. We tested the hypotheses using a hierarchical multiple regression analysis. We first regressed information elaboration on the control variables team size and mean status, then added the lower order terms, and finally added the cross-product term. Although researchers commonly mean-center lower order terms prior to the creation of cross-product term to facilitate interpretability (Dalal & Zickar, 2012), doing so would harm the interpretability of the influence of skewness. Specifically, in our case, by mean-centering skewness, the nature of the skewness sign would no longer indicate anymore whether most of the team members hold low-status or high-status positions. Therefore, we refrained from mean-centering. Note that this decision has no impact on the direction and significance of the interaction (Dalal & Zickar, 2012).
Table 2: Means, standard deviations, and correlations among the variables
Variables Mean SD 1 2 3 4 5 1. Team size 3.99 0.15 2. Status mean 5.15 0.48 .21* 3. Hierarchy steepness 0.48 0.32 -.13 -.50** 4. Hierarchy skewness -0.23 1.20 .01 .22* -.39** 5. Information elaboration 5.54 0.51 .18* .51** -.24** .08 6. Team performance 7.55 1.02 -.03 .06 .09 .05 .18* Note. N = 127 teams. * p < .05. ** p < .01.
Hypothesis 1 predicted that hierarchy steepness has a negative influence on information elaboration when this hierarchy is more positively skewed. In support of this hypothesis, our analysis shows (see Table 2) a significant negative interaction effect between steepness and skewness on information elaboration (B = -0.30, SE = 0.12, p < .05). To further examine the pattern of this effect, we plotted the influence of steepness on information elaboration for teams with larger portion of high-status team members and teams with smaller portion of high-status team members (± 1 SD hierarchy skewness). This plot (see Figure 3) suggests that hierarchy steepness is negatively related to information elaboration when hierarchy is pyramid shaped ( ; + 1 SD hierarchy skewness) and is weakly related to information elaboration when hierarchy is shaped like an inverted pyramid ( ; -1 SD hierarchy skewness). To test both for which skewness values steepness had a significant influence on information elaboration and how strong this influence is, we plotted the regions of significance (Preacher et al., 2006). This plot (see Figure 3) indicates that steepness has a significant (p < .05) negative influence on information elaboration when hierarchy is more pyramid shaped ( ), and that this influence is non-significant when hierarchy is shaped like an inverted pyramid ( ). Figure 3 furthermore indicates that the impact of steepness ranges from non-significant, for teams with a skewness smaller than 0.54, to -0.97 (B, SE = 0.42, p < .05), for teams that have a maximum skewness of 2.0. We thus find support for Hypothesis 1.
Figure 3: Interaction and regression of significance plots: Information elaboration and the steepness-skewness interaction.
Note. Regions of significance (on the right side of confidence band) are based on non-centralized variables (95% Confidence Interval).
1 1.5 2 2.5
Low hierarchy
steepness High hierarchysteepness
In fo rm at io n el ab or at io n Negative hierarchy skewness Positive hierarchy skewness
Table 3: Results of regression analysis
M: Information Elaboration DV: Team
Performance
Step 1 Step 2 Step 3 Step 4
Variables Team size 0.24 0.26 0.24 0.26 0.30 0.16 -0.37 0.51 Status mean 0.53*** 0.08 0.54*** 0.10 0.49*** 0.10 0.08 0.27 Hierarchy steepness 0.02 0.15 -0.36 0.21 0.49 0.34 Hierarchy skewness -0.01 0.04 0.11 0.06 Hierarchy steepness x Hierarchy skewness -0.30* 0.12 Information elaboration 0.41* 0.18 R2 .26 .24 .31 .05 F 22.62*** 11.19*** 9.95*** 1.94
Note. N = 127 teams. Unstandardized regression coefficients (i.e. B) were used. M = Mediator. DV = Dependent Variable.
p <.10. * p < .05. ** p < .01. *** p < .001.
Hypothesis 2 predicted that teams with a steeper status hierarchy perform worse through impaired information elaboration when this hierarchy is more positively skewed. Following Hayes’ (2015) recommendations, we tested this hypothesis by computing the index of moderated mediation – i.e. a quantified association between the indirect effect and the moderator (Hayes, 2015). In analyzing the value of this index, we first examined the indirect effect of steepness on team performance (see step 4 in Table 3). Secondly, we assessed the difference in this indirect effect for different skewness values by conducting a 95% bootstrap estimation. This analysis indicated an index of -.13, with a confidence interval excluding zero [-.33, -.02], suggesting that there is a stronger negative indirect effect when the skewness is more positive.
Complementing this analysis, we analyzed the conditional indirect effects of steepness of team performance for both teams with a pyramid-shaped hierarchy ( ; + 1 SD hierarchy skewness) and teams with a hierarchy shaped like an inverted pyramid ( ; -1 SD hierarchy skewness). In teams with hierarchies shaped like an inverted pyramid ( ), we did not find support for an indirect influence of steepness on team performance; we thus found support for Hypothesis 2.
DISCUSSION
This chapter focuses on explaining when and how status hierarchy relates to team performance, which is important because status hierarchies are prevalent in organizations (Anderson & Brown, 2010; Magee & Galinsky, 2008). Our findings suggest that the impact of hierarchy depends on the hierarchy shape. To capture the shape of the hierarchy, we advanced the conceptualization and measurement
of status hierarchy and introduced status hierarchy skewness as a measure for hierarchy concentration that overcomes a shortcoming of existing hierarchy concentration measures. We demonstrated that pyramid-shaped hierarchy ( ) inhibits team performance by reducing information elaboration.
Implications for the hierarchy literature
This chapter contributes to the hierarchy literature in a number of ways. First, this chapter helps to explain when status hierarchy affects team performance. A stream of hierarchy research argues for a positive impact of status hierarchy on team processes (Anderson & Willer, 2014; Halevy et al., 2011), but most empirical evidence points towards a negative impact (Greer et al., 2018). Scholars proposed that the negative impact of status hierarchy is partly due to deficiencies in the measurement of hierarchy (Bunderson et al., 2016; Hays & Bendersky, 2015). In support of this proposition, we found that improving the validity of existing measures can help to explain when hierarchy affects team outcomes. Specifically, we found that status hierarchy only hurt information elaboration and team performance when the hierarchy is pyramid-shaped ( ). Possibly, previous findings would have been more nuanced about the impact of hierarchy when hierarchy was measured as the steepness-skewness interaction – such as in this chapter. Before rejecting the proposed functionalist impact of status hierarchy (Anderson & Willer, 2014; Halevy et al., 2011), based on the support for a negative impact (Greer et al., 2018), scholars should consider using the hierarchy
measurement introduced in this chapter.
A second contribution of this chapter is the introduction of hierarchy skewness as an improved measure for concentration. We demonstrated an important deficiency of conventional concentration measures. Even though hierarchy is argued to be larger when the concentration is located at the bottom of the hierarchy (Harrison & Klein, 2007), we demonstrated that conventional concentration measures do not always accurately represent the location of the concentration. Previous work furthermore argued that conventional measures of concentration and steepness are statistically too related for conceptually distinct concepts (Bunderson et al., 2016). To help overcome these deficiencies, we introduced skewness as a concentration measure. Skewness is statistically unrelated to steepness and accurately reflects the location of the concentration in the hierarchy.
In addition to contributing to the hierarchy literature by addressing the conceptualization and measurement of hierarchy, this chapter contributes to the hierarchy literature by expanding on the mechanisms that transmit the impact of hierarchy to team performance. More specifically, we provided empirical evidence for information elaboration as a mediator of the status hierarchy – team
performance relationship. Although scholars have explicitly suggested that the separate dimensions of information elaboration (i.e. exchange, discussion, and integration) mediate the impact of status hierarchy (Bunderson et al., 2015; Greer et al., 2015), we do not know of any other studies that demonstrate this mediating role of information elaboration. By providing empirical evidence for the mediating role of information elaboration, we clarify how status hierarchy relates to team performance.
Implications for the within-team dispersion literature
In addition to contributing to the hierarchy literature, this chapter has important implications for the broader within-team dispersion literature –which focuses of variety (e.g. ethnic diversity), separation (e.g. personality diversity) and disparity (e.g. status hierarchy) (Harrison & Klein, 2007). In particular, we argue that the within-team dispersion literature can benefit from the way in which we studied the combined impact of steepness and skewness, which are team properties that more generally represent respectively the degree of dispersion and the shape of dispersion. After decades of research interest in team phenomena at mean-level (Mathieu et al., 2014; Rousseau et al., 2006), scholars have studied team
phenomena captured as the dispersion degree (Bell, 2007; Van Dijk, Van Engen, & Van Knippenberg, 2012) and as the dispersion shape (DeRue et al., 2010; Sinha et al., 2016). Scholars have demonstrated that team performance can be predicted by the interaction between mean-level phenomena and dispersion properties (e.g., mean and dispersion degree; De Jong & Dirks, 2012; Liden et al., 2006) and by the direct influence of multiple dispersion properties (e.g., dispersion degree and dispersion shape; Li & Liao, 2014). We revealed that also the combination of steepness (i.e. dispersion degree) and skewness (i.e. dispersion shape) can predict team performance. In contrast to previous efforts to capture this combined influence by artificially classifying teams into a limited number of categories (González-Romá & Hernández, 2014), which results in a loss of variation across hierarchies (MacCallum et al., 2002), our approach accounts more accurately for differences across configurations of within-team dispersion. By demonstrating that the interaction between steepness (i.e. dispersion degree) and skewness (i.e.
dispersion shape) can affect team outcomes, this chapter is an example for the broader within-team dispersion literature.
Practical Implications
This chapter demonstrates that status hierarchy can impair the performance of teams by hindering the exchange, discussion, and integration of input. This information elaboration is especially important for knowledge intensive teams. It is therefore crucial for these teams, even if they do not have a formal hierarchy, to mind the status hierarchy. When practitioners are confronted with low levels of information elaboration, this may be the result of status hierarchy. In the current chapter, status hierarchy had a negative impact only when the hierarchy was pyramid-shaped ( ). To measure (the shape of) the status hierarchy, practitioners need to measure the status of individuals. Similar to our approach, individuals’ status level can be measured by asking each team members how much status they think that their fellow team members hold, but could also be estimated on the basis of the characteristics that are status indicators in that specific context (Correll & Ridgeway, 2006). To subsequently intervene with the status hierarchy and, in turn, alter information elaboration, practitioners can change status level of individuals by stressing the (ir)relevance of the characteristics of team members (Wagner & Berger, 1982; Webster Jr & Driskell Jr, 1978). For example, the social value of a low-status members can be stressed in order to increase the status of this person. Possibly, creating awareness of the team’s status hierarchy and its consequences may be sufficient to motivate team members to be more inclusive during information elaboration.
Limitations and Future Research Directions
Although our study has several important implications, it also has limitations that warrant consideration. One limitation is that our non-experimental research design prevents us from drawing inferences about the causality of our findings. Even though status and information elaboration were measured preceding the measurement of performance, they were measured at one moment in time. As such, we cannot rule out reverse causality between status hierarchy and information elaboration. To test the causality underlying our findings, we encourage scholars to test the conceptual model experimentally.
The study findings should be appreciated in light of their context. First of all, the teams in our sample were operating in a learning environment in which information elaboration was critical for teams to perform well. Although this context is ideally-suited for examining information elaboration as a mediator of the status hierarchy – team performance relationship, information elaboration may actually drain time (Resick et al., 2014) which may negatively affect to the performance of teams executing more routine and predictable tasks (e.g., relay race teams, fire fighter teams, manufacturing teams) (Bigley & Roberts, 2001; Bunderson et al., 2016). In these settings, teams may suffer from steeper hierarchies that have a more negative skewness (i.e. hierarchies shaped like inverted pyramids ) rather than a more positive skewness (i.e. pyramid-shaped hierarchies ). Secondly, although the study context was ideally suited for isolating the effect of status hierarchy, in organizational settings power hierarchy could moderate the effects of hierarchy such that more conflict is expected when low-status individuals hold more power (Anicich et al., 2016). Therefore, in future
studies on the status hierarchy – team performance relationship, we encourage scholars to examine the conceptual model both in settings in which hierarchy operates through mediators other than information elaboration (e.g., conflict) and in which power hierarchy is more likely to be present (e.g., professional business settings).
The limitations of our study open up a number of avenues for future research. First, even though we argue that the study sample helps in controlling for power hierarchy, the interaction between steepness and skewness may produce similar results for other social hierarchies such as power hierarchy. Status hierarchy and power hierarchy operate independently of one another and can have dissimilar outcomes (Bunderson & Reagans, 2011; Hays & Bendersky, 2015; Magee & Galinsky, 2008). However, just as in research on status hierarchy, power hierarchy research tends to focus on either steepness or concentration (Bunderson et al., 2016). Furthermore, there is also a tendency to defer more to higher-ranked individuals in power hierarchy (Anderson & Brion, 2014). We therefore expect that the impact of power hierarchy is also an outcome of the interaction between steepness and skewness. Hence, we encourage scholars to explore whether our findings are generalizable to other types of social hierarchy.
Second, one surprising result of the study, which included mean-level status as a control variable, is the positive association between mean-level status and information elaboration (r = .51, p < .05). We argue that future research should consider studying the role of mean-level status in teams. Although the individual-level status literature has commonly studied status as a between-unit phenomenon (Anderson et al., 2015), the team-level status literature has commonly studied
status as a within-unit phenomenon (Anderson & Brown, 2010; Halevy et al., 2011; Magee & Galinsky, 2008). More specifically, in the team literature, status is often conceived as a hierarchy that evolves from within-team comparisons (Anderson & Brown, 2010; Halevy et al., 2011; Magee & Galinsky, 2008). When assigning status to others, however, individuals may also make between-team comparisons (e.g. based on interaction across teams or previous experience). As a result, there may be meaningful differences in the mean-level status across teams with a similar status hierarchy. As people tend to defer to others with a higher status (Berger et al., 1972, 1977; Correll & Ridgeway, 2006), higher levels of information elaboration are likely to be present in teams with higher average status scores. Hence, we encourage scholars to further explore the role of average status levels in the study of the impact of status hierarchy.
CONCLUSION
In conceptualizing and measuring team status hierarchy, it is important to account for both status hierarchy steepness and status hierarchy concentration. Hierarchy skewness, a measure that we introduced to accurately capture concentration, moderates the indirect influence of steepness on team
performance. We demonstrated that the negative impact of hierarchy steepness on team performance is stronger when the hierarchy skewness is more positively skewed – that is, for pyramid-shaped hierarchy ( ). Pyramid-shaped hierarchy is detrimental to team performance because it hinders information elaboration – the exchange, discussion, and integration of information. This chapter thus shows that
status hierarchy hurt team performance more when the hierarchy is pyramid-shaped ( ), through lower levels of information elaboration.
