S TATUS DISTANCE : UBIQUITOUS AND HIGHLY CONSEQUENTIAL
A FIELD STUDY ON THE RELATIONSHIP BETWEEN STATUS DISTANCE AND HELPING : A MODERATED MEDIATION MODEL
Master thesis, MSc Human Resource Management University of Groningen, Faculty of Economics and Business
June 21, 2012
A NDRÉ VAN T OLY
Student number: 1999117 Het Hout 127 9723 LB Groningen Tel.: +31 (0)6-29108273 E-mail: a.h.van.toly@student.rug.nl
Supervisor
Drs. Y. Cantimur
2
S TATUS DISTANCE : UBIQUITOUS AND HIGHLY CONSEQUENTIAL
A FIELD STUDY ON THE RELATIONSHIP BETWEEN STATUS DISTANCE AND HELPING : A MODERATED MEDIATION MODEL
ABSTRACT
In this paper, a relationship between status distance – differences among individual team members in their level of influence, prominence, and respect – and helping behavior is tried to be found. Previous research relevant to this relationship is reviewed. The review finds that the effects of status distance on team functioning, processes and outcomes are mixed. In some cases status distance is found to have positive effects, sometimes negative effects are found. The relationship is therefore tested through three hypotheses, while mediated by intrateam conflict and moderated by psychological safety. Data were collected from 71 teams from 62 German and Dutch organizations. There was no moderated mediation of psychological safety found. Neither for intrateam conflict, nor for helping behavior. The proposed moderated mediation model did not hold. The results are discussed in light of the theoretical and practical implications.
Keywords: Status; teams; status distance; psychological safety; intrateam conflict; helping behavior.
Word count: 11.949
3
INTRODUCTION
The emphasis on status shifts away from individual and interteam status towards intrateam status. Correspondingly, research on teams has increased (Ilgen, Hollenbeck, Johnson, & Jundt, 2005). Therefore, it is important to consider the effects of status on the functioning of teams.
Status is defined as the amount of influence, prominence, and respect each team member enjoys in the eyes of others (Anderson, John, Keltner, & Kring, 2001). People can differ in the amount of status. When one or a few individuals possess high status and others have low status within a team, there exists a larger status distance among these team members. Status differences are omnipresent in human groups and emerge spontaneously from social interactions (Anderson &
Kilduff, 2009b; Ridgeway, 1987). This is interesting as the level of status distance affects team processes and team functioning. Research is not unambiguous in whether the effects of status distance are positive or negative. A larger status distance might have negative effects on relationships between individuals. These negative effects can manifest themselves in team processes or team performance. The allocation of status can shape directions and magnitude of influence and control in teams such that it decreases team satisfaction and team performance (Cohen & Zhou, 1991; Keltner, Van Kleef, Chen, & Kraus, 2008). Other research shows that larger status distance has potential value as it may facilitate the survival and success of teams (Halevy, Chou, & Galinsky, 2011). Due to the mixed findings on the effects of status distance, it is noted here that status distance does not necessarily need to be a problem, as the mere existence of differences in status might not be detrimental to team performance. Status distance may become a problem when team members do not feel psychologically safe (Edmondson, 1999). In a psychologically safe working environment, team members feel safe to take risks and open productive discussions take place (Edmondson, 1999), which in turn will enhance team performance.
Teams can be hotbeds of conflict (Jehn & Mannix, 2001). While teams become more central to organizations (e.g., many organizations shift towards team-based structures), problems of coordination, cooperation and conflict management rise (Boyett & Conn, 1991;
Chattopadhyay, Finn, & Ashkanasy, 2010; Gladstein, 1984; Jehn, 1995). Therefore, intrateam
conflict plays a large role in the relationship between status distance and helping behavior. With
organizations making more use of cooperative work practices (Ten Brummelhuis, Van der
Lippe, & Kluwer, 2010) helping behavior at work becomes increasingly important. Research has
shown that higher levels of helping behavior are found when the working climate in a team is
good (Burger, Messain, Patel, Del Prado, & Anderson, 2004; Sleebos, Ellemers, & De Gilder,
4
2007). Higher levels of helping behavior can possibly be found, even when one or few team members are more influential, prominent and respected than others (Anderson & Brown, 2010;
Halevy et al., 2011; Tyler, 2006). Key to a good working climate are low levels of intrateam conflict and a feeling of psychological safety among team members (Choi, 2006; Edmondson, 2003). “The relationship between status inequality and outcomes within teams is complex”
(Christie & Barling, 2010: 929). Thus, research is needed on the variables under consideration.
The level of status distance affects team processes and team functioning (Anderson & Brown, 2010; Magee & Galinsky, 2008) and it is therefore posed that status distance possibly affects the level of intrateam conflict and helping behavior which are closely related to how effective a team functions (De Dreu & Weingart, 2003). This thesis investigates the effects of status distance and psychological safety on intrateam conflict and helping behavior.
This research intends to yield both scientific and practical implications. A scientific contribution lies in the fact that while there are mixed findings of status distance on team processes and team functioning (Anderson & Brown, 2010; Halevy et al., 2011; Magee &
Galinsky, 2008), a moderating variable is needed for better understanding. By suggesting psychological safety as a moderator, it is tried to reconcile the divergent views on the effects of status distance on team processes and outcomes. Furthermore, the research contributes scientifically to the investigation of how status distance influences team processes in general and how it influences intrateam conflict and helping behavior specifically. The moderating role of psychological safety on these processes is also explored. Status distance and more specific intrateam status distance receive increasing attention, as status distance is a rather new concept that needs additional research (Christie & Barling, 2010). While much attention is paid to the effects of status distance on team performance (cf. Anderson & Brown, 2010), and there are many theories about the effects of status distance on team processes (Anderson & Brown, 2010;
Anderson & Kilduff, 2009a; Halevy et al., 2011; Ridgeway & Walker, 1995), not much evidence is provided. This research contributes scientifically by providing evidence on the effects of status distance on team processes. Up to now, research has focused mainly on the antecedents and consequences of status (cf. Bendersky & Hays, 2010; Berger, Rosenholtz, & Zelditch, 1980;
Keltner, Gruenfeld, & Anderson, 2003), this research investigates differences in status within
teams. The within-team status configuration is at the core of this paper, as there is looked at the
status structures of teams. The practical contributions are clear as well. First, when psychological
safety is low, managers can change the status structure by lowering status distance. When
psychological safety among team members is high, the status distance among team members can
be larger. Second, when there is a large status distance, managers need to make sure that team
5
members dare to take risks and not feel embarrassed, rejected or punished when they speak up, through creating a psychologically safe working environment. Finally, unresolved conflicts within teams and a low level of experienced psychological safety in the past can have negative future effects on helping behavior (Blakar, 1984; Carron et al., 2003; Desiviliya & Eizen, 2005).
Research on this phenomenon and the possible influence of managers on creating psychologically safe environments can be of great value to organizations (Klaver, 2008).
THEORETICAL FRAMEWORK Status distance, psychological safety, and intrateam conflict
Status is defined by many authors. Anderson et al. (2001) define status as the amount of influence, prominence, and respect one enjoys in the eyes of others. Status also concerns the possession over and allocation of resources and responsibilities (Blieszner & Adams, 1992;
Anderson & Brown, 2010; French & Raven, 1959). Status fulfils a number of important roles within teams, especially when individual team members differ in status. High-status individuals can control team interactions and give verbal assignments to team members, while low-status individuals can exert less influence, are expected to obey the orders given and need to keep their opinions to themselves (Berger et al., 1980; Goffman, 1967; Keltner et al., 2003). Benefits of being in a high-status position include higher compensation, having greater influence and valuable exchange partners, having access to information and resources contributing to individuals’ performance, and receiving more positive evaluations than those with low status (Belliveau, O’Reilly, & Wade, 1996; Bendersky & Hays, 2010; Bendersky & Shah, 2010;
Berger et al., 1980; Foschi, 2000; Friedkin, 1999; Thye, 2000). It can be beneficial to attain a higher status level than others within a team. Within-team status difference, or status distance, is an interesting phenomenon. Most definitions of status distance hold aspects in it such as
“differences between persons with regard to status”, and “inequality with respect to the status people hold” (Blau, 1977; Christie & Barling, 2010; McPherson & Smith-Lovin, 1987). Status distance is also called status inequality (e.g., Bottero & Prandy, 2003; Prandy, 1999; Remøe, 1982), status difference (e.g., Anderson & Kilduff, 2009a; Bobo & Zubrinsky, 1996; Vorauer &
Sakomoto; 2008), status dispersion (e.g., Greer & Van Kleef, 2010) or steepness of a hierarchy
(e.g., Anderson & Brown, 2010). These concepts are all treated here as status distance – the
situation in which status is concentrated in one, or only few people and not in others. When two
individuals are exactly the same in status, there is zero status distance. In case there is a large
status distance within a team, the actors with higher status possess more resources (Ellis, 1994)
6
and enjoy higher levels of influence, prominence, and respect than actors with lower status (Anderson, Beer, Srivastava, Spataro, & Chatman, 2006; Keltner et al., 2003).
Research has shown mixed effects of status distance on team processes (Anderson &
Brown, 2010; Magee & Galinsky, 2008; Tarakci & Groenen, 2011). Large status distance can facilitate team survival and team success whereas it possibly creates a psychologically rewarding environment, it may support coordination, and it can reduce conflict and therefore might enhance voluntary cooperation (Anderson & Brown, 2010; Halevy et al., 2011; Sondak & Bazerman, 1991). Furthermore, a larger status distance facilitates clear division of labor in which team members know each others’ responsibilities, thereby supporting smooth and efficient interactions, preventing intrateam conflict and facilitating conflict resolution (Magee &
Galinsky, 2008; Sondak & Bazerman, 1991). A negative effect of a low status distance is that teams can experience inefficient decision making due to the fact that too many team members want to have a say in the decision making process (Berger, Cohen, & Zelditch, 1972; Berger et al., 1980). This occurs when too many team members strive to enhance their status positions. In contrast, large status distance can have a negative effect on the quality of relationships among team members. Due to the fact that status shapes interpersonal interactions, (Byrne, 1971;
Ridgeway & Walker, 1995) large status distance is ought to be an obstacle for developing high- quality relationships (Blau, 1977; Kalmijn, 1991; Lincoln & Miller, 1979; McPherson & Smith- Lovin, 1987; Wuthnow, 2003). Larger status distances produce a competitive environment due to the fact that team members compete for salary, voice, promotion, and recognition (Greer & Van Kleef, 2010; Halevy et al., 2011). Large status distances might also impair conflict resolution (Mannix, 1993; Wolfe & McGinn, 2005) due to feelings of inequality and injustice among team members (Henderson & Frederickson, 2001; Lawler & Proell, 2009). Finally, research has shown that larger status distance within a team increases the team members’ motivation to achieve higher status while the disparity in resources becomes more salient when status distance increases (Christie & Barling, 2010). When team members strive for higher levels of status, the emphasis is placed on individual advancement, “often irrespective of collective interests”, which possibly leads to higher levels of intrateam conflict (Christie & Barling, 2010: 921).
The existence of teams can thus be conducive to conflict (Jehn & Mannix, 2001). Some researchers even go as far as stating that conflict is inevitable in teams (e.g., Jehn, 1995).
Intrateam conflict is defined as a disagreement about issues, or a difference in opinion about the
best way to perform, which results from incompatible goals and interests (Amason & Sapienza,
1997; Jehn, 1995; Jehn, Northcraft, & Neale, 1999). Conflict within teams is also defined as the
awareness by employees that discrepancies, incompatible wishes or desires exist among them
7
(Boulding, 1963). Conflict is often categorized in three distinct types: relationship, task, and process conflict. This categorization lies beyond the scope of this research; intrateam conflict will therefore be treated as a general construct. As stated, asymmetric status relationships (i.e., large status differences between team members) lead to lower-quality relationships (Phillips, Rothbard, & Dumas, 2009). Lower-quality relationships within teams report higher levels of conflict (Druskat & Wolff, 1999). Especially when there is a large status distance and individuals within the team feel that psychological safety is low, levels of conflict are expected to increase.
This might become a problem, as research has shown that conflict within teams is often detrimental for performance (De Dreu & Weingart, 2003; De Wit, Greer, & Jehn, 2012).
As research does not show a clear, unidirectional effect of status distance on team processes, it is theorized that status distance may not be a problem in itself. The mere existence of status distance does not necessarily have large negative implications (e.g., increased intrateam conflict). It might, however, influence relations and the level of conflict among team members when team members do not feel psychologically safe. Psychological safety is defined as “a shared belief that the team is safe for interpersonal risk taking” (Edmondson, 1999: 354).
Interpersonal risk taking is the “sense of confidence that others will not embarrass, reject or punish someone for speaking up” (Edmondson, 1999: 354). This clearly indicates that psychological safety has to do with the environment and context people are in and that trust is a part of psychological safety; but psychological safety goes beyond trust, also including a
“comfortable environment” (Edmondson, 2003: 241). The definition Kahn (1990: 708) gives of psychological safety also emphasizes this environment in which it is accepted and possible “to show and employ one’s self without fear of negative consequences”. Psychological safety is a requisite for individuals to feel secure (Schein & Bennis, 1965). Nevertheless, this safe and comfortable environment does not imply that team members are necessarily close friends, nor does it imply that there are no problems or that there is no pressure (Edmondson, 2003). Rather, as team members feel that they do not have to protect themselves, but can speak openly, “it describes a climate in which the focus lies on productive discussion, enabling early prevention of problems and accomplishment of shared goals” (Edmondson, 2003: 242). In a climate in which team members feel psychologically safe, there seems to be no or little room for intrateam conflict as concerns are expressed and people who make mistakes are not punished by others, but helped in preventing it to happen again.
Research has shown that psychological safety leads to open and productive
communication (Edmondson, 1999; 2003), which in turn leads to higher levels of conflict
resolution or even prevents the existence of unsatisfied team members (Simons & Peterson,
8
2000). High-status individuals are influential and prominent team members, compared to low- status team members (Anderson & Brown, 2010). Due to higher levels of psychological safety, low-status team members are expected to engage in open communication with both high- and low-status team members through which team members’ perceived feelings of inequality and injustice will decrease (Halevy et al., 2011). The strive for status in teams with large status distance is expected to decrease when team members feel psychologically safe as in a psychologically safe environment, team members do not need resources or prominence before they are given voice and allowed to speak up (Edmondson, 1999; Liang, Farh, & Farh, 2012).
Therefore, when there is a psychologically safe working environment, and team members feel safe enough to express their concerns and take risks, fewer conflicts are expected within the team, even when a large status distance exists (Christie & Barling, 2010). The following hypothesis is posed:
Hypothesis 1: Psychological safety moderates the relationship between status distance and intrateam conflict, such that status distance is more negatively related to intrateam conflict when psychological safety is higher.
Status distance, psychological safety, and helping behavior
Increasingly, organizations work with collective goals for several individual organizational members. This requires information and idea sharing, the integration of perspectives and the coordination of tasks (Edmondson, 2003). In achieving these collective goals, collaboration and cooperation are essential (Chattopadhyay et al., 2010). Research has shown that higher status team members receive more helping than those with lower status (Ellis, 1994). Helping behavior consists of collectively eliminating workload distribution problems through sharing the workload in order to get the work done (Choi, 2009; Ilgen et al., 2005; Porter, Hollenbeck, Ilgen, Ellis, &
West, 2003). In previous research, helping behavior is described as either proactive or reactive
(Grant, Parker, & Collins, 2009), based upon a lack of ability or a lack of effort (LePine,
Hollenbeck, Ilgen, Colquitt, & Ellis, 2002) and resulting from an explicit or implicit request for
help (Bierhoff, 2002; Guégen, Martin, & Meineri, 2011). Status distance can have various effects
on helping behavior. Low-status individuals exhibit less helping behavior when status distance
increases (Caverley, Cunningham, & MacGregor, 2007). Other researchers have also shown a
negative relationship between the two variables. It is found that when there exists a large status
distance, team members feel more detached from each other, which will lead to lower levels of
helping behavior (Chattopadhyay, 1999; Tsui & O’Reilly, 1989). People are more likely to work
9
together with those equal in status (McPherson & Smith-Lovin, 1987). Research also shows that helping behavior positively affects other team members and team performance (Flynn, 2006;
Mossholder, Richardson, & Settoon, 2011; Podsakoff, Ahearne, & MacKenzie, 1997; Podsakoff, MacKenzie, Paine, & Bachrach, 2000; Settoon & Mossholder, 2002; Van Dyne & LePine, 1998). It is found that differences in status may be seen as contributing to both outcome and procedural fairness (Halevy et al., 2011), which in turn promotes voluntary cooperation (Tyler, 2006). It is suggested that in steep, well established hierarchies (e.g., large status distance) higher levels of helping behavior among team members are found than in teams in the opposite fashion (Chattopadhyay et al., 2010). When there exists a large status distance within a team, the low- status individuals possibly cooperate with the higher-status team members, as this might serve their own interests whereas the latter possesses influence and resources the low-status team members desire and employing non-cooperative behavior will not lead to these resources (Gruenfeld & Tiedens, 2010; Halevy et al., 2011).
From this it can be derived that status distance and helping behavior affect each other, although the direction and causality between the two variables might still considered to be unclear. Status shapes interpersonal interactions and large distance in status among team members is found to be an obstacle for developing high-quality relationships. Due to the existence of status distances in virtually all teams, this might lead to problems concerning the helping behavior among team members. Despite the fact that a larger status distance was shown to lead to less helping behavior due to the fact that team members feel more detached from each other (Chattopadhyay, 1999; Tsui & O’Reilly, 1989), it is expected that in a team with large status distance among team members, higher levels of helping behavior are found when psychological safety is higher as research has already shown this effect among students who worked together (Strange & Banning, 2001). Sharing knowledge, workload and accomplishing shared goals (Edmondson, 2003), as well as asking questions or seeking help are expected to occur when team members are in a comfortable context. Through psychological safety, team members will be transparent, trusting, honest, and open towards each other and their high-status team members (Eggers, 2011). Furthermore, higher levels of psychological safety create a feeling of inclusion (Foldy, Rivard, & Buckley, 2009). Therefore, even in a situation with large status distance among team members, collective workload sharing is expected to happen when team members feel psychologically safe. Psychological safety also decreases concerns about being viewed as incompetent when asking for feedback and assistance on the job (Eggers, 2011).
Hence, in case psychological safety is high, team members are expected to be more willing to
10
ask each other for help and help each other in achieving collective goals despite a larger status distance, compared to the situation in which psychological safety is low.
It is thus important for organizations to create a supportive working environment, in which individuals feel safe to take risks and team members can employ themselves while they do not fear negative consequences as in such a psychologically safe working environment helping behavior is likely to increase (Klaver, 2008). Hence, the second hypothesis reads as follows:
Hypothesis 2: Psychological safety moderates the relationship between status distance and helping behavior, such that status distance is more positively related to helping behavior when psychological safety is higher.
The mediating role of intrateam conflict in the relationship between status distance, psychological safety, and helping behavior
The final part of this research investigates the mediating role of intrateam conflict on the interaction of status distance and psychological safety and helping behavior. The necessity for low-status individuals to comply with demands of high-status individuals increases as the likelihood of sanctions increases in a team with larger status distance (Halevy et al., 2011).
Therefore, larger status distance might motivate, or force, low-status team members to show
more cooperative behavior in the short term, but with the risk of increasing dissatisfaction among
team members, possibly leading to intrateam conflict and lower levels of helping behavior in the
long run. Successful team action almost always requires cooperation (Halevy et al., 2011). Thus,
team members operating in a larger status distance are expected not to show higher levels of
helping behavior due to higher levels of intrateam conflict. Intrateam conflict potentially harms
team processes, such as cooperation, as well as performance outcomes, such as goal
accomplishment (De Dreu & Weingart, 2003; Langfred, 2007; Peterson & Behfar, 2003). Team
members often hold past and future relationships with other team members in mind when
evaluating current team membership (Jetten, Branscombe, Spears, & McKimmie, 2003). This
suggests that unresolved conflicts within the team negatively affects concern for others and sorts
negative effects on the future willingness of team members to help others (Blakar, 1984; Carron
et al., 2003; Desiviliya & Eizen, 2005). Other research shows that conflict within teams can lead
to ineffective teamwork and other negative outcomes when the conflicts are poorly managed
(Barki & Hartwick, 2001; McGrath, 1991; Robey, Smith, & Vijayasarathy, 1993). Several
researchers have reported higher levels of mental and physical withdrawal from the job as a
result of conflicts within teams (Jehn, 1995; Jehn & Mannix, 2001; Medina, Munduate, Dorado,
11
Martínez, & Guerra, 2005; Surra & Longstreeth, 1990). Withdrawal from the job reduces helping behavior. This suggests that organizations must make sure that intrateam conflict is actively managed to create a supportive working environment in which team members feel psychologically safe. It is shown that when there is less intrateam conflict, there is more helping behavior (Carron et al., 2003; Desiviliya & Eizen, 2005). Furthermore, it is suggested that low levels of intrateam conflict lead to a healthy disagreement in which team members create an environment that enhances team innovation and cooperative problem-solving (Gobeli, Koenig, &
Bechinger, 1998). Low levels of conflict are considered to be good as they encourage debate and discussion over divergent ideas, issues, and important decisions (Johnson, 2008). The key to maintain and exploit this low level of conflict is a psychologically safe context in which open energetic discussions can take place to exchange ideas (Johnson, 2008; Kahn, 1990). In addition, higher levels of psychological safety are desired whereas the increased possibility of sanctions due to larger status distances among team members leads to lower levels of helping behavior (De Dreu & Weingart, 2003; Halevy et al., 2011). Team members must be made more willing to take risks and communicate openly, without the fear of negative consequences (e.g., being punished, embarrassed, or rejected by their team members). The following hypothesis results:
Hypothesis 3: Intrateam conflict mediates the relationship between the interaction of status distance and psychological safety and helping behavior
The conceptual model is illustrated in Figure 1.
FIGURE 1
Conceptual model of the relationship between status distance, psychological safety, intrateam conflict and helping behavior
Intrateam conflict
Status distance Helping behavior
Psychological safety
12
METHOD Participants
The sample comprised 71 teams from 62 German and Dutch organizations, which in total amounted to 433 individuals. The non-governmental and governmental organizations operated in different industries and were located in different regions in Germany and the Netherlands. A team is defined as composed of more than two individuals, existing to perform tasks relevant to the organization, conducting interdependent tasks, socially interacting and sharing responsibility and accountability for collective common goals (Hollenbeck, Ilgen, Sego, Hedlund, Major, &
Phillips, 1995; Ilgen et al., 2005; Kozlowski & Bell, 2001) The average team contained 6.69 employees (s.d. = 2.17), with a minimum of 5 team members and a maximum of 12 team members, including the team leaders. Overall, of the 475 individuals who were asked to participate, 433 agreed to participate. The response rate was 91%.
There was no large distinction in the number of females (55%) and males participating.
Most participants were well educated at or above the higher vocational educational level (65.6%). The participants averaged 38.85 years of age (s.d. = 11.71), ranging from 17 to 62 years old. The organizational tenure was 10.08 years (s.d. = 9.17). Team tenure was 4.93 years (s.d. = 5.40).
Design
The research was conducted on the team level. Students from the pre-master Human Resource Management 2010-2011 conducted the surveys for the course “Research Paper for Pre-MSc Human Resource Management”.
After having agreed to participate, each team member was instructed about the procedure of the study. Team members, blind to the purposes of the research, were asked to fill out two structured questionnaires concerning team functioning, the functioning of individual team members, including themselves, and the functioning of the team leader. The team leaders were asked to fill out one questionnaire on the performance of the team members. The team leaders’
questionnaire and the first questionnaire from the team members were filled out in the first
meeting with the researcher. The researcher collected the first questionnaires in closed envelopes
to guarantee anonymity, after which the second questionnaires were handed out in envelopes to
the team members. Each individual was assigned a unique code, which was only known by the
researcher, to enable a match between the first and second questionnaires. The researcher
returned after seven to ten days to collect the second questionnaires. A short introductory text on
13
each questionnaire explicitly stated that the confidentiality of the participants’ and their responses was guaranteed and that their names and the corresponding unique codes were only used to link their responses from the two different questionnaires. The respondents were not rewarded for their participation and participated voluntarily. The measures in the questionnaires were translated to Dutch, using a double-blind back-translation procedure. Only the Dutch version of the questionnaire was offered to the participants.
Measures
Status distance. In order to measure status distance, I first needed to measure the status of each individual team member. Therefore, I have used a round-robin (peer-rating) design. Each individual team member rated and was rated by all team members. This design prevented the outcomes to be affected by common method variance. The team members also indicated their own (perceived) level of status according to the definition of Anderson et al., (2001). The item measured was: “To what extent is this person / are the following team members influential, prominent, and respected within the team?” (see appendix A). A 7-point Likert scale (Likert, 1932) was used, anchored by 1 = “not at all” and 7 = “to a great extent”. These ratings were equally weighted and averaged to form an overall single measure of the specific status level within the team. The validity and reliability of the Likert scales have been demonstrated in previous studies (Ary, Jacobs, Sorensen, & Razavieh, 2002).
I have operationalized status distance as the within-team standard deviation of the team members’ level of status (cf. Chan, 2008). Research has shown that when looking at the interaction of dispersion, standard deviation is the best measure to be used (Roberson, Sturman,
& Simons, 2007). Through using the standard deviation of all individual members averaged status scores within the same team, I could derive the level of status distance (Aiken & West, 1991). Higher values of standard deviation indicated larger status distance.
Psychological safety. Psychological safety was measured with 7 items developed by Edmondson (1999), of which three items were reverse-coded. Items were measured on a response scale ranging from 1 (= “completely disagree”) to 7 (= “completely agree”), see appendix B. Two sample items used were: “People on this team sometimes reject others for being different” and “It is safe to take a risk in this team”. As a final step, all items were averaged and aggregated to the team level. The items demonstrated a reliable representation of psychological safety, Cronbach’s alpha was .750.
Intrateam conflict. Intrateam conflict was measured through nine items by the existing
Intragroup Conflict Scale of Jehn (1995) and the process conflict items from Shah and Jehn
14
(1993). The scale ranged from 1 (= “never”) to 7 (= “always”). For specific items, see appendix C. Two sample items were: “To what extent are personality conflicts evident in your work group?” and “How frequently are there conflicts about ideas in your work group?”. All items were averaged and then aggregated to the team level. The items were reliable with a demonstrated Cronbach’s alpha for intrateam conflict of .897.
Helping behavior. Helping behavior was measured through six items on a seven-point response scale (Settoon & Mossholder, 2002), anchored by 1 = “never” and 7 = “always” (see appendix D). Two sample items used were: “To what extent do you assist co-workers with heavy workloads?” and “To what extent do you help co-workers with work when they have been absent?”. This variable is also averaged and then aggregated to the team level. The Cronbach’s alpha was .854.
Control variables. Three control variables were used: mean status, team tenure and team size. Status distance is measured through the dispersion of the standard deviation of status.
According to Harrison and Klein (2007) it is important to first statistically control for the within- team mean of an attribute in testing a relationship between separation or disparity of this attribute and other variables. Therefore, I have controlled for the mean status level in the teams. It is shown that team tenure influences status and status distance among team members as team members with higher tenure often possess or have greater opportunities to attain higher levels of influence, prominence and respect compared to individuals with lower tenure (Anderson et al., 2001; Christie & Barling, 2010). In addition, team tenure is likely to increase social power of individuals (Mehra, Kilduff, & Brass, 2001) and is therefore considered a “marker of status”
(Christie & Barling, 2010: 923). Finally, it is imaginable that team size affected the outcomes of the research, as research has shown relationships between team size and the willingness of team members to contribute to a team task (Alnuaimi, Robert, & Maruping, 2010) and between team size and team performance (Sharma & Ghosh, 2007). Due to the possible influence of these variables, it was important to check whether these control variables affected the results of the study.
RESULTS Data analysis
The hypotheses were tested in two steps. First, the two moderation models as posed in
hypotheses 1 and 2 were examined. Hypothesis 1 predicted that psychological safety moderated
the relationship between status distance and intrateam conflict, such that status distance would be
15
more negatively related to intrateam conflict when psychological safety was higher. The second hypothesis theorized that status distance and helping behavior would be more positively related when moderated by psychological safety. All predictor variables were standardized prior to the analyses (Aiken & West, 1991). Hypotheses 1 and 2 were tested through a regression analysis, using the SPSS computer package.
Second, assuming that hypotheses 1 and 2 were supported, the proposed mediator variable of intrateam conflict was integrated into the relation of hypothesis 2 to analyze the overall moderated mediation (hypothesis 3). Hypothesis 3 concerned a moderated mediation between status distance and helping behavior. To test this hypothesis, the MODMED SPSS macro Version 2.0, Model 2 was used (Preacher, Rucker, & Hayes, 2007). This macro provides a method for investigating the significance of moderated mediation at different levels of psychological safety. Coefficients were estimated using standardized variables. Helping behavior was regressed on status distance, intrateam conflict, psychological safety and the interaction between status distance and psychological safety. Subsequently, bias-corrected bootstrapping was applied (2,000 bootstrap samples) as it produces more accurate confidence intervals (MacKinnon, Lockwood, & Williams, 2004). 1
Descriptive Statistics
Table 1 presents the means, standard deviations, and correlations for all study variables. Status distance, psychological safety and intrateam conflict were standardized. Psychological safety was significantly correlated with intrateam conflict (r = -.59), see Table 1. Team tenure (r = .13), psychological safety (r = .50), and intrateam conflict (r = -.55) were all significantly correlated with helping behavior.
Hypothesis testing
In order to test the hypotheses, status distance was first regressed on intrateam conflict in which psychological safety was added as a moderator. Second, status distance was regressed on helping behavior, again moderated by psychological safety. Finally, the analysis on the conditional indirect effect was conducted (Preacher et al., 2007). All variables in the interaction were standardized (Aiken & West, 1996). The moderated mediation analysis tested the conditional indirect effect of status distance on intrateam conflict and helping behavior at certain levels of psychological safety (-1 standard deviation, mean, and +1 standard deviation). In this step,
1