SPANNING AND ITS RELATION TO TEAM
PERFORMANCE MODERATED BY TEAM
SIZE
Master thesis, Msc Human Resource Management University of Groningen, Faculty of Economics and Business
June 21th, 2013 MARLOES FRAIJ Student number: 2211556 Weegschaalstraat 32 9742 NZ Groningen tel.: + 31 (0)6-44022462 e-mail: m.fraij@student.rug.nl Supervisor/ university prof. dr. G.S. van der Vegt
2
ANTECEDENTS OF TEAM BOUNDARY SPANNING
AND ITS RELATION TO TEAM PERFORMANCE
MODERATED BY TEAM SIZE
Abstract
Team boundary spanning consists of the team’s actions to establish linkages and manage interactions with parties outside their own team (Ancona, 1990). Previous research showed that team boundary spanning is a key predictor for team performance, but researchers gave little attention to exploring contingencies of this relationship (Marrone, 2010). This study contributes to existing research by exploring antecedents of team boundary spanning and examining contingency variables that affect the relationship between team boundary spanning and team performance. With the use of two questionnaires I collected multiple data from employees and managers working in 79 teams in an Additional Healthcare Institute in the Netherlands. These data were used to perform a linear regression analysis. The results of this analysis confirm the positive relationship between functional team diversity and team boundary spanning, and the positive relationship between team boundary spanning and team performance. Marginally significant evidence was found for the moderating effect of team size on the relationship between team boundary spanning and team performance. The results of this research give managers more insight in how the composition of their team influences team boundary spanning and how this is related to team performance.
3
INTRODUCTION
Today’s work environment is characterized by task complexity, knowledge work, changing economic conditions, global competition, and flatter work structures (Mohrman, Cohen, & Mohrman, 1995). Therefore organizations rely increasingly on teams to make decisions. The effectiveness of the collective information-sharing process of teams has grave implications for the quality of decisions and performance in organizations (Devine, Clayton, Philips, Dunford, & Melner, 1999). Given the demands of an increasingly diverse workplace (Jackson, 1992), members of work teams must also learn how to pool resources and
knowledge effectively.
In modern organizations, teams need to coordinate interdependent work and bridge disconnected parties by actively managing relationships external to the team (Mohrman et al., 1995). This is also named team boundary spanning, referring to the team’s actions to establish linkages and manage interactions with parties in the external environment. These actions can support the team, and others linked to the team, in meeting performance goals and task objectives (Ancona, 1990).
While a great deal of empirical work on team boundary spanning has emerged over the past two decades, and some theoretical attention has been given to advancing team boundary spanning research (Ancona & Caldwell, 1992a), significant gaps still exist in our
understanding of the team boundary spanning phenomenon (Marrone, 2010). Marrone (2010) pointed out that still little attention is given to exploring contingencies of the relationship between team boundary spanning and team performance. This is remarkable given the strong assertions in early work that team boundary spanning may be critical for certain teams under certain conditions. Research needs to explore if structural elements, such as team size and team diversity, serve as meaningful influences on the team’s ability to carry out team boundary spanning activities more or less effectively.
4
THEORY AND HYPOTHESES
Team boundary spanning takes place with parties in the external environment and differs in a meaningful way from internal team processes. Internal team processes involve the nature of team members’ interactions to internally develop strategies, coordinate workloads and manage interpersonal conflict. External team processes reflect the nature of the team members’ interactions with parties external to the team (Marks, Mathieu, & Zaccaro, 2001). These external team processes can go further than the team’s host organization, although this research is limited to the external team processes within the team’s host organization
For teams involved in team boundary spanning, the challenge is to create team
boundaries that are porous enough to allow resources and information in, but resistant enough to avoid uncertainty about who is in the team and whether these members are accountable for collective outcomes (Hackman, 2002). If teams acquire information and resources,
successfully manage relationships with external stakeholders and protect team resources from external demands they can succeed in this challenge.
Functional team diversity as antecedent of team boundary spanning
Functional team diversity refers to team member differences in organizational functions. Teams with members working in different functions can expose different sources of task information and know-how (Cummings, 2004). Keller (2001) and Joshi, Pandey and Han (2009) found that functionally diverse teams also have more external communication and more links to external resources than less functionally diverse teams.
Team members working in different functions have often different backgrounds in training, education and work experience (Milliken & Martins, 1996). The difference in team members’ background provides the team with access to diverse social networks they have established in their domains (Ancona & Caldwell, 1992b). The breadth of the external network of functionally diverse teams can increase the number of team boundary spanning contacts compared to less functionally diverse teams. Less functionally diverse teams have more team members working in the same organizational functions increasing the likelihood that they have access to the same external networks. This increases the chance that external contacts of less functionally diverse teams are similar and can reduce the number of team boundary spanning contacts.
5
team boundary spanning members of functionally diverse teams are more likely to encounter unique knowledge that had not been previously shared within the team (Cummings, 2004).
The lack of kindred spirits in functionally diverse teams stimulates team members to seek outsiders with whom to communicate, increasing team boundary spanning (Ancona & Caldwell, 1992b). Unless team members developed a shared identity, teams members with diverse functional backgrounds identify themselves more with organizational members in the same functional area rather than with their functionally diverse team. This makes it difficult for team members to discover critical linkages among different perspectives (Sethi, Smith, & Park, 2000) and to develop a shared purpose (Dougherty, 1987). The communication
problems that might occur due to the lack of common vocabulary caused by different
background and experience can make it even more difficult to link different perspectives and to develop a shared purpose (Zenger & Lawrence, 1989), increasing the likelihood that team members participate in team boundary spanning.
In line with previous research I expect that functional team diversity is positively related to team boundary spanning. The arguments presented above are combined to propose the following:
Hypothesis 1. Functional team diversity is positively related to team boundary spanning.
The relationship between team boundary spanning and team performance
Several studies (Marrone, Tesluk, & Carson, 2007; Ancona & Caldwell, 1992a) have shown that team boundary spanning is a key predictor for team performance. Through team boundary spanning teams establish linkages and manage interactions with other teams that can support their team, and others linked to their team, in meeting performance goals and task objectives (Ancona, 1990). The interactions between different teams play an important role in gaining access to critical information that is necessary to reach high levels of performance and to successfully coordinate activities.
Ancona and Caldwell (1988) have shown that successful teams try to manage the flow of information and other resources across their boundaries. Successful teams gather
6
assess the quality of the information obtained. In this way teams can avoid the distribution of inaccurate information and ensure the accurateness of information used to perform their tasks.
The more team members are able to span team boundaries and to communicate with people outside their team, the more effective they can be (Ancona & Caldwell, 1992a). Teams that participate in team boundary spanning can bridge diverse and disconnected parties and act as critical conduits for information transfer, knowledge creation and innovation (Argote, McEvily, & Reagans, 2003). Team boundary spanning also helps the team to get informed about externally available support and resources that can be used to increase their
performance. Teams are also less likely to act inconsistent with organizational goals, because through team boundary spanning teams can better understand the goal of the organization what can avoid that teams focus too much on their own goals (Edmondson, 2003).
In line with previous research I expect that team boundary spanning is positively related to team performance. The arguments presented above are combined to propose the following:
Hypothesis 2. Team boundary spanning is positively related to team performance.
Team size as moderator for the relationship between team boundary spanning and team performance
The relationship between team boundary spanning and team performance may vary from team to team. Hare (1952) found that larger teams have more access to information and resources, but coordination problems and process losses (Gooding & Wagner, 1985) might hinder the use of this advantage. In this research I will investigate the moderating effects of team size, defined as the number of people working in a team (Carson, Tesluk, & Marrone, 2007).
Researchers working in recent years have generally agreed on the informational and resource benefits that crossing boundaries brings teams (Ancona & Caldwell, 1992a; Argote et al., 2003; Hansen, 1999). The resources and information obtained from team boundary spanning can help the team to increase their performance (Edmondson, 2003). Smaller teams can better integrate the resources and information obtained from high team boundary
7
bounded” Alderfer (1976 ) means that the team has external knowledge and resources, but not enough cohesion to motivate members to bring different perspectives together.
For larger teams, it is more difficult to integrate the resources and information obtained from high team boundary spanning. Larger teams can suffer from coordination and process losses (Gooding & Wagner, 1985) that will increase with the addition of each team member (Steiner, 1972). If team size increases, there is less opportunity for each individual to speak, resulting in an incomplete exchange of resources and information within the team (Thomas & Fink, 1963). Another explanation is that individuals in larger teams expend less effort to integrate resources and information because their contribution is less critical for team performance (Valacich, Wheeler, Mennecke, & Wachter, 1995) and therefore they take less responsibility for the successful integration of resources and information (Wicker & Mehler, 1971).
Low team boundary spanning provides both smaller and larger teams with limited resources and information. Smaller teams still have the integration advantage, although there are less resources and information to integrate. Larger teams still have the integration
disadvantage, which is less critical because there are less resources and information to integrate.
In line with previous research I expect that a smaller team size strengthens the relationship between team boundary spanning and team performance. The arguments presented above are combined to propose the following:
Hypothesis 3. The positive relationship between team boundary spanning and team performance becomes stronger when team size decreases.
8 FIGURE 1 Conceptual model
METHODS Procedure
For this research the already existing dataset of the Teammonitor 2012 was used. The data were collected using an online questionnaire that consisted of one questionnaire for all employees (including team managers) and one extra questionnaire for only team managers. The items in the questionnaires were formulated in Dutch and all the items had a 7-point Likert rating scale, unless otherwise mentioned. The employee questionnaire contained items about team membership, conflict, team goals, feedback and team boundary spanning. The team manager questionnaire contained items about team performance, innovation and creativity.
The researcher informed all teams during team meetings. Participation was voluntary, but every team member was encouraged to participate by the researcher and by the team manager. The online employee questionnaire was distributed among all team members. The online team manager questionnaire was taken one week after the employee questionnaire was taken. Data were collected in April and May 2012.
The introduction of the questionnaire mentioned that the research was anonymous, data were treated confidentially, and that the questionnaire took about twenty minutes to complete. The employees were blind to objectives and they were asked to answer all items.
9
To increase the number of responses a reminder was sent and team members received extra time to fill in the questionnaire.
This research only used the team boundary spanning data collected with one item in the employee questionnaire and the team performance data collected with six items in the team manager questionnaire. To obtain information about team diversity and team size information was collected from the human resource information system (HRIS).
Respondents
All participants worked in an Addiction Healthcare Institute in the Northern part of the Netherlands. The organization employed around 1000 employees, working in 85 teams
varying from 3 to 31 team members. Almost all organizational teams participated, including staff teams, such as communication and HR, and teams providing addiction healthcare.
Participants had different professions, varying from socio therapist, to communication advisor and general practitioner.
From the 1003 questionnaires distributed across 85 teams, 724 fully completed questionnaires and 44 partly completed questionnaires were returned. At the individual level, 72.2% of the questionnaires was fully completed and 4,4% was partly completed, resulting in a total response of 76,6%.
The average age of the respondents was 44 years (SD = 10.81), ranging from 23 to 65 years. About one third of the respondents was male (35%) and nearly all respondents were of Dutch nationality (99%). The average tenure was 6.58 years (SD = 5.67). More than half of the respondents (54%) completed a higher professional education (HBO) and 11% had a university degree.
There were 83 questionnaires distributed among 44 team managers, of which 82 fully completed questionnaires were returned. This resulted in a response rate of 98.8%. Twenty-seven team managers supervised multiple teams, ranging from one to five teams per team manager.
Measuring instruments
Team diversity. Team diversity was determined based on data from the HRIS.
10
is the fraction of the team in the ith category. Fifty percent of the teams scored .60 or higher on Blau’s index for team diversity (M = .49, ranging from .00 to .85).
Team boundary spanning. Team boundary spanning was measured using the
employee questionnaire. The participants were asked to answer one item “with whom do you discuss and/or adjust work related subjects on a regular basis, which means one or more times a week? (e.g. by using e-mail, telephone or personal contact)”. This item was adapted from Hansen (2005) , and focused on the binary measures of the boundary spanning relationship. The full network method was used, collecting information about each actor's ties with all other actors (Hanneman & Riddle, 2005).
UCINET 6 (Borgatti, Everett, & Freeman, 2002) was used to calculate team boundary spanning measures. Before calculating team boundary measures I symmetrized the dataset. A symmetrized dataset only includes contacts between employees that are mentioned by both respondents. However, in reality it often happened that person A reported contact with person B, but person B had not reported contact with person A. In network analysis there are two ways to solve this problem and to make the dataset symmetric.
The most often used approach is from Stork and Richards (1992) who suggested reconstructing the missing part of the network using the observed incoming relations of the missing actors. All missing ties Xij were replaced with the observed values of the opposite tie in the dyad: Xijmiss= Xji. The simple strategy of this approach was assuming that if the
respondent had answered, he would have responded the same way as others did about him. The second approach is the fixed choice design from Holland and Leinhard (1973). Holland and Leinhard (1973) found that network ties can be reported by both persons
(reciprocated nominations), by only one partner (non-reciprocated nominations), and yet some ties will not be reported at all (censored links). The fixed choice design allowed the researcher to exclude non-reciprocated nominations, because they may be qualitatively different from reciprocated nominations. However, fixed choice nominations easily leads to a non-random missing data pattern, what is the main reason why the approach of Stork and Richards (1992) is common in network analysis.
In this research I applied both approaches to symmetrize the dataset, what resulted in two different datasets which both were used to test hypotheses. In these datasets the intra-team contacts were removed, to ensure that only intra-team boundary spanning relations were involved.
11
immediate team, divided into two mutually exclusive categories: horizontal and diagonal contacts. Horizontal contacts are links outside the immediate team at the same hierarchical level as the respondent as defined by the organizational chart. Diagonal contacts go up and down one or more hierarchical levels, but outside the formal vertical chain of command. The numbers of horizontal and diagonal contacts were divided over the total amount of contacts to get percentages (Wilson & Malik, 1995). Hypotheses were tested using both team boundary spanning measures percentage of horizontal contacts and percentage of diagonal contacts. Appendix A explains in more detail how UCINET (Borgatti et al., 2002) was used to calculate team boundary spanning measures.
Team performance. Team performance was measured by asking team managers to
compare their teams’ performance to other teams when assessing their team(s) on six items. Items were for example about achieving goals, quality of the work, productivity and team effectiveness. All items are mentioned in Appendix B. The scale used was developed by Van der Vegt and Bunderson (2005) and the items were assessed on a 7-point Likert rating scale (1=far below average, 7= way above average). Measures were obtained from 79 out of 82 teams due to three missing cases related to team performance. Cronbach’s alpha (α) was .92 showing that the six items were internally consistent.
Team size. Team size was measured objectively by counting the numbers of team
members. This information was provided by the HRIS and checked by the team managers.
Control variables. A review of the literature and consideration of my research context
suggested the need to control for tenure diversity (Bunderson & Sutcliffe, 2002). Several studies have suggested that tenure diversity can negatively affect intra team
communication (Williams & O'Reilly, 1998), motivating employees to involve in inter-team communication. Previous research examined the relationship between tenure diversity and performance outcomes, with some scattered evidence for significant relationships (Williams & O'Reilly, 1998). In this research I controlled for two types of tenure diversity, namely organizational tenure and team tenure.
Data analysis
Based on information obtained, I examined team diversity as antecedent of team boundary spanning, the relationship between team boundary spanning and team performance, and the moderating effects of team size. First, I measured Cronbach’s alpha for team
12
variable functional team diversity and the non-standardized outcome variables percentage of horizontal contacts and percentage of diagonal contacts. Third, I tested the right side of my model with the linear regression analysis. I took the standardized independent variables percentage of horizontal contacts and percentage of diagonal contacts, the standardized moderator team size and the non-standardized outcome variable team performance.
All items were averaged and all data were transformed from the individual level to the team level. From the 85 teams that participated, three teams were removed because they were not included in the network item and three teams were removed due to missing performance evaluations. In the end 79 teams were used to test hypotheses.
In this research I used the two different team boundary spanning measures. I also used two different datasets, one calculated with the approach of Stork and Richards (1992) and one calculated with the approach of Holland and Leinhard (1973). First, I tested all three
hypotheses for the two boundary spanning measures. Second, I used two different datasets to test the hypotheses. In the main result section I will present the results calculated based on the approach of Stork and Richards (1992) and in the supplementary analysis I will present the results calculated based on the approach of Holland and Leinhard (1973). The results of both tests will be presented in the same tables. The results of the approach of Stork and Richards (1992) are presented without brackets (e.g. .45) and the results of the approach of Holland and Leinhard (1973) are presented between brackets (e.g. (.33) ).
RESULTS Descriptive statistics
Table 1 presents the mean score and the standard deviation of the two control variables organizational tenure and team tenure, the independent variable team diversity, the moderator team size, the two mediating team boundary spanning measures and the dependent variable team performance. Also the Pearson zero-order correlations between the seven variables are presented in Table 1.
13
diagonal contacts (r = - .51, p < .01). The results show a significant positive relationship between functional team diversity and team size (r = .29, p < .01).
Non-significant relationships are found between the two control variables
Descriptive statistics and Pearson zero-order correlations among the variables
Mean S.D. 1 2 3 4 5 6 7
1. Team tenure 5.68 2.57
2. Organizational tenure 6.59 2.54 .49**
3. Functional team diversity 0.50 0.27 -.01 .08
4. Percentage horizontal contacts 63.13 (51.63) 26.00 (25.94) -.03 (-.02) -.10 (-.07) .23* (.15)
5. Percentage diagonal contacts 11.63 (9.88) 19.87 (18.28) -.07 (-.05) .16 (.15) -.32** (-.33**) -.51** (-.40**) 6. Team size 9.14 5.50 .11 -.04 .29** .04 (.07) .01 (-.01) 7. Team performance 4.71 0.94 .25* .18 -.02 .17 (.13) -.13 (-.15) -.05 Note: N= 79 †p < .10 *p<.05 **p<.01
of functional team diversity on the percentage of horizontal contacts. Table 3 presents the results of the linear regression analysis to check the direct effect of functional team diversity on the percentage of diagonal contacts.
Hypothesis 1 stated that functional team diversity is positively related to team
boundary spanning. The results in Table 2 show a significant and positive relationship (B = 6.34, p < .05) between functional team diversity and the percentage of horizontal contacts, supporting the hypothesis. The results in Table 3 show a significant and negative relationship (B = - 6.96, p < .01) between functional team diversity and the percentage of diagonal
contacts, rejecting the hypothesis.
Table 4 presents the results of the linear regression analysis to check the relationship between the percentage of horizontal contacts and team performance and the moderating effect of team size. Table 5 presents the results of the linear regression analysis to check the relationship between the percentage of diagonal contacts and team performance and the moderating effect of team size.
Hypothesis 2 stated that team boundary spanning is positively related to team
performance. The results in Table 4 show a positive and marginally significant relationship (B = .20, p < .10) between the percentage of horizontal contacts and team performance,
supporting the hypothesis with limited evidence. The percentage of diagonal contacts is unrelated to team performance (B = - .01, n.s.).
Hypothesis 3 stated that the positive relationship between team boundary spanning
and team performance becomes stronger when team size decreases. For the percentage of horizontal contacts no evidence was found (B = .05, n.s.), rejecting the hypothesis. For the percentage of diagonal contacts, the relationship was positive and marginally significant (B = .30, p < .10), rejecting the hypothesis.
Figure 2 presents the marginally significant two-way interaction between the percentage of diagonal contacts and team performance moderated by team size. In creating this figure, I followed the procedure as suggested by Aiken and West (1991).
With the simple slope test (Dawson, 2013) the significance of slopes in Figure 2 was tested. For small teams, team performance significantly decreases when the percentage of diagonal contacts increases (B = - 0.31, t = - 2.00, p = 0.05). For large teams, team
16 TABLE 2
Results regression analysis outcome variables percentage of horizontal contacts Percentage of horizontal contacts
Model Variable R R2 Unstandardized Standard Error Standardized t sig.
Coefficient (B) (SE) Coëfficiënt (Beta)
1 Organizational tenure -2.79 (2.09) 3.53 (3.40) -.10 (-.08) -.79 (-.62) .43 (.54) Team tenure .10 (.07) .01 (.01) .49 (.42) 3.54 (3.40) .02 (.02) .14 (.12) .89 (.90) 2 Organizational tenure -3.51 (-2.54) 3.47 (3.39) -.13 (-.10) -1.01 (-.75) .32 (.46) Team tenure .89 (.67) 3.46 (3.38) .03 (.03) .26 (.20) .80 (.84) Functional team diversity .26 (.18) .07 (.03) 6.34 (4.13) 2.93 (2.96) .24 (.16) 2.17 (1.39) .03* (.17) Note: N= 79 †p < .10 *p<.05 **p<.01
17 TABLE 3
Results regression analysis outcome variables percentage of diagonal contacts Percentage of diagonal contacts
Model Variable R R2 Unstandardized Standard Error Standardized t sig.
Coefficient (B) (SE) Coëfficiënt (Beta)
1 Organizational tenure 5.25 (4.09) 2.64 (2.35) .25 (.22) 1.99 (1.74) .05* (.09†) Team tenure .23 (.20) .05 (.04) -3.96 (-2.94) 2.64 (2.35) -.19 (-.16) -1.50 (-1.25) .14 (.22) 2 Organizational tenure 6.03 (4.80) 2.49 (2.22) .29 (.26) 2.42 (2.16) .02* (.03*)
Team tenure -4.41 (-3.34) 2.49 (2.21) -.21 (-.18) -1.77 (-1.51) .08† (.14)
Functional team diversity .42 (.41) .17 (.17) -6.96 (-6.48) 2.11 (1.94) -.35 (-.35) -3.31 (-3.34) .00**(.00**) Note: N= 79 †p < .10 *p<.05 **p<.01
18 TABLE 4
Results regression analysis outcome variable percentage of horizontal contacts Percentage of horizontal contacts
Model Variable R R2 Unstandardized Standard Error Standardized t sig.
Coefficient (B) (SE) Coëfficiënt (Beta)
1 Organizational tenure .07 (.07) .13 (.12) .07 (.07) .56 (.56) .58 (.58) Team tenure .26 (.26) .07 (.07) .21 (.20) .13 (.12) .22 (.22) 1.69 (1.69) .10† (.10†) 2 Organizational tenure .08 (.07) .13 (.12) .08 (.07) .63 (.57) .53 (.57) Team tenure .22 (.21) .13 (.12) .23 (.23) 1.77 (1.77) .08† (.08†) Team size -.08 (-.08) .10 (.11) -.08 (-.09) -.75 (-.78) .45 (.44) % horizontal contacts .33 (.31) .11 (.09) .18 (.14) .11 (.11) .19 (.15) 1.68 (1.32) .10† (.19) 3 Organizational tenure .08 (.07) .13 (.12) .08 (.07) .61 (.57) .55 (.57) Team tenure .22 (.22) .13 (.12) .22 (.23) 1.72 (1.77) .09† (.08†) Team size -.08 (-.08) .10 (.11) -.08 (-.08) -.73 (-.74) .47 (.46) % horizontal contacts .20 (.16) .11 (.12) .20 (.17) 1.70 (1.39) .09† (.17) % horizontal contacts X .33 (.31) .11 (.10) .05 (.06) .13 (.13) .05 (.06) .40 (.46) .69 (.65) team size Note: N= 79 †p < .10 *p<.05 **p<.01
The numbers placed between brackets (...) are the results of the test done in the supplementary analysis.
19 TABLE 5
Results regression analysis outcome variable percentage of diagonal contacts Percentage of diagonal contacts
Model Variable R R2 Unstandardized Standard Error Standardized t sig.
Coefficient (B) (SE) Coëfficiënt (Beta)
1 Organizational tenure .07 (.07) .13 (.12) .07 (.07) .56 (.56) .58 (.58) Team tenure .26 (.26) .07 (.07) .21 (.20) .13 (.12) .22 (.22) 1.69 (1.69) .10† (.10†) 2 Organizational tenure .09 (.09) .13 (.12) .10 (.10) .73 (.74) .47 (.46) Team tenure .20 (.19) .13 (.12) .20 (.20) 1.54 (1.57) .13 (.12) Team size -.07 (-.07) .11 (.11) -.07 (-.08) -.64 (-.67) .53 (.51) % diagonal contacts .30 (.31) .09 (.09) -.13 (-.14) .12 (.11) -.13 (-.15) -1.16 (-1.34) .25 (.18) 3 Organizational tenure .10 (.10) .13 (.12) .10 (.11) .75 (.84) .45 (.41) Team tenure .20 (.21) .13 (.12) .21 (.22) 1.60 (1.73) .11 (.09†) Team size -.06 (-.06) .10 (.10) -.07 (-.06) -.62 (-.55) .54 (.59) % diagonal contacts -.01 (.00) .14 (.13) -.01 (.00) -.07 (.03) .94 (.98) % diagonal contacts X .35 (.37) .12 (.14) .30 (.34) .18 (.18) .22 (.26) 1.68 (1.94) .10† (.06†) team size Note: N= 79 †p < .10 *p<.05 **p<.01
20 FIGURE 2
Two-way interaction independent variable percentage of diagonal contacts
Supplementary analysis
In the main result section I presented the results calculated based on the approach of Stork and Richards (1992) and in the supplementary analysis I will present the results calculated based on the approach of Holland and Leinhard (1973).
Descriptive statistics
Table 1 presents the descriptive statistics and Person zero-order correlations among the same variables as tested in the main result section. I will discuss the results that differ from the main analysis.
The percentage of diagonal contacts is negatively and significantly related to the percentage of horizontal contacts (r = - .40, p < .01). The correlation between the percentage of diagonal contacts and functional team diversity (r = - .33, p < .01) is also significant and negative.
Non-significant relationships are found between the percentage of horizontal contacts and organizational tenure (r = - .07, n.s.), team tenure (r = - .02, n.s.), and functional team diversity (r = .15, n.s.). A non-significant relationship is also found between the percentage of diagonal contacts and team tenure (r = - .05, n.s.) and organizational tenure (r = .15, n.s.). The
1 1,5 2 2,5 3 3,5 4 4,5 5 5,5 6 Low Percentage of diagonal contacts High Percentage of diagonal contacts T eam p er for m an ce
21
relationship between team size and the percentage of horizontal contacts (r = .07, n.s.) and the relationship between team size and the percentage of diagonal contacts (r = - .01, n.s.) is also non-significant. Finally, the relationship between team performance and both percentage of horizontal contacts (r = .013, n.s.) and percentage of diagonal contacts (r = - .15, n.s.) is non-significant.
Hypothesis testing
Table 2 presents the results of the linear regression analysis to check the direct effect of functional team diversity on the percentage of horizontal contacts. Table 3 presents the results of the linear regression analysis to check the direct effect of functional team diversity on the percentage of diagonal contacts.
Hypothesis 1 stated that functional team diversity is positively related to team
boundary spanning. The results in Table 2 show a non-significant relationship (B = 4.13, n.s.) between functional team diversity and the percentage of horizontal contacts, rejecting the hypothesis. The results in Table 3 show a significant and negative relationship (B = - 6.48, p < .01) between functional team diversity and the percentage of diagonal contacts, rejecting the hypothesis.
Table 4 presents the results of the linear regression analysis to check the relationship between the percentage of horizontal contacts and team performance and the moderating effect of team size. Table 5 presents the results of the linear regression analysis to check the relationship between the percentage of diagonal contacts and team performance and the moderating effect of team size.
Hypothesis 2 stated that team boundary spanning is positively related to team
performance. The results in Table 4 show a non-significant relationship (B = .26, n.s.)
between the percentage of horizontal contacts and team performance, rejecting the hypothesis. The results for the percentage of diagonal contacts are also not significant (B = .00, n.s.), rejecting the hypothesis.
Hypothesis 3 stated that the positive relationship between team boundary spanning
22
Figure 3 presents the marginally significant two-way interaction between the percentage of diagonal contacts and team performance moderated by team size. In creating this figure, I followed the procedure as suggested by Aiken and West (1991).
With the simple slope test (Dawson, 2013) the significance of the slopes in Figure 3 was tested. For small teams, team performance significantly decreases when the percentage of diagonal contacts increases (B = - .34, t = - 2.27, p = .03). For large teams, team performance was unrelated to the percentage of diagonal contacts (B = .34, t = 1.26, p = .21).
FIGURE 3
Two-way interaction percentage of diagonal contacts and team performance
DISCUSSION AND CONCLUSION
The aim of this research was to investigate if functional team diversity is related to team boundary spanning, and if team size weakens the relationship between team boundary spanning and team performance. This research tried to deal with Marrone’s (2010) concern that little attention has been paid to exploring contingencies of the relationship between team boundary spanning and team performance. Another purpose of this research was to confirm previous results that found that functional team diversity is positively related to team boundary spanning. 1 1,5 2 2,5 3 3,5 4 4,5 5 5,5 6 Low Percentage of diagonal contacts High Percentage of diagonal contacts T eam p er form an ce
23 Research findings and theoretical contributions
This research showed that there is a positive relationship between functional team diversity and team boundary spanning. This is in line with previous research (Ancona & Caldwell, 1992b; Edmondson, 1999) and with what was expected. Cummings (2004) found that members working at different locations, who represent different functions, and who work in different business units have access to and can benefit from unique sources of knowledge outside their team. Lutz (1994) found that functionally diverse teams can bring people from upstream and downstream functional areas together so that they can exchange information and coordinate work activities.
The positive relationship between team boundary spanning and team performance is also confirmed. This is in line with previous research (Faraj & Yan, 2009; Edmondson 1999) and with the expectations of this research. Ancona and Caldwell (2007) found that activities focussed on communication with outside employees aiming at coordination of the team’s effort, such as obtaining feedback about the team’s progress and discussing problems, are positively related to team performance. Druskat and Wheeler (2003) found that undertaking boundary spanning activities strongly affects team performance. These activities include building relationships, scouting for information, and persuading others to support the team’s work.
The moderating effect of team size on the relationship between team boundary spanning and team performance remains unclear. Team size is not a strong moderator and conclusions cannot be drawn from the marginally significant results found. One possible explanation is that team size is an umbrella concept for other variables that play an important moderating role in the relationship between team boundary spanning and team performance. Future research should focus on other moderating variables than team size to further explore contingencies of the relationship between team boundary spanning and team
performance. One example of such a variable can be the supportive internal team network structure of the team. This internal team network structure is necessary to effectively disseminate resources and information obtained through boundary spanning to other team members (Tushman, 1977). Another possible variable is the number of team boundary spanning contacts (Marrone, 2010). A final possibility is to look at the moderating effects of the external dependency of the team for resources and information (Marks, De Church, Mathieu, Panzer & Alonso, 2005).
24
interesting differences. In the next paragraph I will zoom in on the differences in results between the percentage of horizontal contacts and the percentage of diagonal contacts.
Research findings and theoretical contributions percentage of diagonal contacts and percentage of horizontal contacts
The results showed that the correlation between the percentage of horizontal contacts and the percentage of diagonal contacts was significant and negative. The relative measure of team boundary spanning can explain this negative relationship. Both the number of horizontal contacts, and the number of diagonal contacts were calculated as a percentage of the total number of contacts. As a consequence, the relative measure of team boundary spanning excludes the possibility to have high scores on both the percentage of horizontal contacts and the percentage of diagonal contacts explaining the negative correlation.
A second interesting difference is that functional team diversity was positively related to the percentage of horizontal contacts, but negatively related to the percentage of diagonal contacts. One explanation can be the number of levels of the organizational hierarchy that is used to calculate the diagonal contacts. In this research I distinguished between four
hierarchical levels: employees, team managers, management team members and the chairman of board of directors. The majority of the organizational members is categorized in the
category employees, increasing the chance that the number of contacts with organizational members at the same hierarchical level is higher than with someone at another hierarchical level. Another explanation is the number of functions that is represented at each hierarchical level. The majority of the 69 different functions are represented in the category employees, while at the diagonal level only three functions (team manager, management team member, and chairman of the board of directors) can be distinguished. This classification increases the chance that the functionally diverse contacts take place at the horizontal level, and decreases the chance that the functionally diverse contacts take place at the diagonal level. Finally, I did not distinguish between specialist and generalist functions what can explain the results found. Bunderson and Sutcliffe (2002) showed that teams with specialist functional diversity have complementary knowledge, while teams with generalist functional diversity have primary the same knowledge. For teams with specialist functional diversity the additional value of
25
performance. For future research it is interesting to take the difference between specialist functional diversity and generalist functional diversity into account to see if results remarkably differ.
A third interesting difference is that the percentage of horizontal contacts is positively related to team performance, while the percentage of diagonal contacts is unrelated to team performance. The relative low percentage of diagonal contacts compared to the high percentage of horizontal contacts can explain why the percentage of diagonal contacts is unrelated to team performance. Another possible explanation is that organizational members at higher hierarchical levels have different views and different experiences and use their power to force others to adapt their views (Wamsley & Zald, 1973). This increases the chance of conflicts, and a battle for power can shift the attention away from the effective use of knowledge to improve team performance (Gruenfeld & Keltner, 2003).
A fourth interesting difference is that team size did not moderate the relationship between the percentage of horizontal contacts and team performance, while marginally significant evidence was found for the moderating role of team size on the relationship between the percentage of diagonal contacts and team performance in the unexpected, positive direction. Possibly, the information that smaller teams obtain from horizontal
boundary spanning does not provide the team with new and novel insights needed to solve ill-defined and uncertain problems in complex environments (Daft & Lengel, 1984). Another explanation is that small teams still lack access to critical information to perform well (Hare, 1952). A clear theoretical background related to percentage of diagonal contacts is almost non-existing. A possible explanation why for larger teams the relationship between the percentage of diagonal contacts and team performance is marginally significant and positive is that larger teams have already more resources and information obtained through their team size (Hare, 1952), rather than through team boundary spanning. The specific information of diagonal contacts can be combined with already available resources and information, increasing the performance of larger teams.
Practical implications
26
spanning with team performance. As organizations would like to increase their team performance, it helps to create conditions under which the team boundary spanning is high. Third, the results show that team size is not an important moderator of the relationship between team boundary spanning and team performance. Overall can be concluded that if the organization would like to (re)design their team structure to increase performance, it is more important to take into account the different functions that are present in a team rather than the size of the team.
Strengths, limitations and suggestions for future research
This research has several strengths. The main strength of this research is its design. Data were collected in a field setting and online questionnaires were distributed among almost 1000 employees working in 85 teams. Respondents worked in their natural work
environment, all had experience in performing their job and they represented different demographic areas. Multiple sources were used to collect data. The data of team size and functional team diversity were objective and obtained from the HRIS. The data of team boundary spanning were the experienced subjective contacts measured in the employee
questionnaire. The team managers rated the performance of their own teams, which is positive because the team manager knows the ins and outs of the team and had the possibility to take into account external factors that influenced team performance.
Another strength of this research is that it is unique in the sense that it points at the direction of team boundary spanning contacts and distinguishes between horizontal and diagonal contacts. Previous research of Wilson and Malik (1995) and Ancona and Caldwell (2007) focussed on the direction of contacts as well, but boundary spanning research involving the direction of contacts is still in its infancy. The methods that were used in this research and the results that were found can serve as guidelines for future team boundary spanning research.
Although this research was based on unbiased data, there were several limitations in this research as well. The main limitation in this research was that relative measures for team boundary spanning were used. A high score on the percentage of horizontal contacts excluded the possibility of a high percentage of diagonal contacts. For future research I would suggest to make use of an absolute measure to avoid this problem and to have information about the number of contacts as well.
A second limitation is that the organizational hierarchy consisted of only four
27
diagonal contacts took place, which influenced both the relationship between functional team diversity and team boundary spanning and the relationship between team boundary spanning and team performance. The challenge for future research is to divide the organization in more hierarchical levels in such a way that the numbers of levels increases and that each level consists of several functions. This can be done for example on the base of salary scales or by distinguishing between junior, medium and senior level employees.
A third limitation is that this research measured the absence and presence of contacts in the organization at one moment in time, while the pattern of activities over time is a better predictor for team performance. Teams can enter cycles of activities early on that reinforce themselves over time resulting in a different pattern of team boundary spanning (Ancona & Caldwell, 1992a). For future research it might be interesting to take into account the
difference patterns in team boundary spanning over time.
A final limitation of this research is that it did not take into account how external conditions shape group processes and outcomes (McGrath, 1997). Sundstrom, DeMeuse and Futrell (1990) found that an external perspective is critical for understanding team
effectiveness, because most teams can perform their tasks and maintain vitality only through their relationships with the external environment. Although this research tried to compensate for the external effects on team performance through the use of team manager rated
performance, I suggest that future research looks at more specific influences on the functional diversity of teams, the possibility to have contacts with external teams and team performance.
Conclusion
Organizations increasingly rely on teams to develop products and processes. This requires that teams span traditional organizational boundaries to gather information and resources necessary to increase team performance. This research looked at functional team diversity as antecedent of team boundary spanning and the relationship between team boundary spanning and team performance. This research showed that functional team diversity is a relevant antecedent for team boundary spanning and that team boundary spanning was positively related to team performance.
28 REFERENCES
Aiken, L. S., & West, S. G. 1991. Multiple regression: testing and interpreting interactions. Newbury Park: Sage Publications.
Alderfer, C. 1976 . Boundary relations and organizational diagnosis. Humanizing
organizational behavior , 109-133.
Ancona, D. 1990. Outward bound: strategies for team survival in an organization. Academy
of Management, 33(2): 334-365.
Ancona, D., & Caldwell, D. 1988. Beyond task and maintenance: defining external functions in groups. Group and Organization Studies, 12: 468-494.
Ancona, D., & Caldwell, D. 1992a. Bridging the boundary: external activity and performance in organizational teams. Administrative Science Quarterly, 37(4): 634-665.
Ancona, D., & Caldwell, D. 1992b. Demography and design: predictors of new product team performance. Organization Science, 3(3): 321-341.
Ancona, D., & Caldwell, D. 2007. Improving performance of new product teams. Research Technology Management, 50(5): 37-43.
Argote, L., McEvily, B., & Reagans, R. 2003. Managing knowledge in organizations: an integrative framework and review of emerging themes. Management Science, 49(4): 571-582.
Blau, P. 1977. Inequality and heterogeneity: a primitive theory of social structure. New York: Free Press.
Borgatti, S., Everett, M., & Freeman, L. 2002. UCINET for Windows: software for social
network analysis. Harvard, MA: Analytic Technologies.
Bunderson, J., & Sutcliffe, K. 2002. Comparing alternative conceptualizations of functional diversity in management teams. Academy of Management Journal, 45(5): 875-893.
Carson, J., Tesluk, P., & Marrone, J. 2007. Shared leadership in teams: an investigation of antecedent conditions and performance. Academy of Management Journal, 50(5): 1217-1235.
Cummings, J. 2004. Work groups, structural diversity and knowledge sharing in a global organization. Management Science, 50(3): 352-364.
Daft, R., & Lengel, R. 1984. Information richness: a new approach to managerial behavior in organizational design. In B. Staw, & L. Cummings, Research in Organizational Behavior, 191-233. Greenwich: JAI Press.
Dawson, J.F. 2013. Moderation in management research: What, why, when, and how.
29
Devine, D. J., Clayton, L. D., Philips, J. L., Dunford, B. B., & Melner, S. B. 1999. Teams in organizations: prevalence, characteristics, and effectiveness. Small Group Research, 30(6): 678-711.
Dougherty, D. 1987. New products in old organizations: the myth of the better mousetrap in search of the beaten path. Sloan School of Management. 1-194
Druskat, V. U., & Wheeler, J. V. 2003. Managing from the boundary: the effective leadership of self-managing work teams. Academy of Management Journal, 46: 435-457.
Edmondson, A. 1999. A safe harbor: social psychological conditions enabling boundary spanning in work teams. Research on Managing Groups and Teams, 2: 179-199.
Edmondson, A. 2003. Speaking up in the operating room: how team leaders promote learning in interdisciplinary action teams. Journal of Management Studies, 40(6): 1419-1452.
Faraj, S., & Yan, A. 2009. Boundary work in knowledge teams. Journal of Applied
psychology, 94(3): 604-617.
Gooding, R. Z., & Wagner, J. A. 1985. A meta-analytic review of the relationship between size and performance: the productivity and efficiency of organizations and their subunits.
Administrative Science Quarterly, 30(4): 462-481.
Gruenfeld, H.H., & Keltner, D. 2003. Power, approach and inhibition. Psychological review, 110 (2): 265-284.
Hackman, J.R. 2002. Leading teams. Boston: Harvard Business School Press.
Hanneman, R., & Riddle, M. 2005. Introduction to social network methods. Riverside: University of California.
Hansen, M. T. 1999. The search-transfer problem: the role of weak ties in sharing knowledge across organizational subunits. Administrative Science Quarterly, 44: 82-111.
Hansen, M.T. 2005. Knowledge sharing in organizations: multiple networks, multiple phases.
Academy of Management Journal , 48(5): 776-793.
Hare, A. 1952. A study of interaction and consensus in different sized groups. American
Sociological Review, 17(3): 261-267.
Holland, P., & Leinhard, S. 1973. Structural implications of measurement error in sociometry.
Journal of Mathematical Sociology, 3(1): 85-111.
Jackson, S. 1992. Team composition in organizational settings: issues in managing an increasingly diverse force. In S. Worchel, W. Wood, & J. Simpson (Eds.), Group processes
30
Joshi, A., Pandey, N., & Han, H. 2009. Bracketing team boundary spanning: an examination of task-based, team-level and contextual antecedents. Journal of Organizational Behavior, 30(6): 731-759.
Keller, R. 2001. Cross functional project groups in research and new project development: diversity, communications, job stress and outcomes. Academy of Management Journal, 44(3): 547-555.
Ladika, S. 2008. Bigger isn't always better. PM Network, 22(3): 34-39.
Lutz, R. 1994. Implementing technological change with cross-functional teams. Research
Technology Management, 37(2): 14-18.
Marks, M.A., Mathieu, J.E., & Zaccaro, S. 2001. A temporally based framework and taxonomy of team processes. Academy of Management Review, 26(3): 356-376.
Marks, M. A., De Church, L. A., Mathieu, J. E., Panzer, F. J. & Alonso, A. 2005. Teamwork in multi-team systems. Journal of Applied Psychology, 90: 964-971.
Marrone, J. 2010. Team boundary spanning: a multilevel review of past research and proposals for the future. Journal of Management, 36(4): 911-940.
Marrone, J., Tesluk, P., & Carson, J. 2007. A multilevel investigation of antecedents and consequences of team member boundary-spanning behavior. Academy of Management
Journal, 50(6): 1423-1439.
McGrath, J. 1997. Small group research, that once and future field: an interpretation of the past with an eye to the future. Group Dynamics: Theory, Research and Practice, 1(1): 7-27.
Milliken, F., & Martins, L. 1996. Searching for common threads: understanding the multiple effects of diversity. The Academy of Management Review, 21(2): 402-433.
Mohrman, S., Cohen, S., & Mohrman, A. 1995. Designing team-based organizations: new
forms for knowledge work. San Fransisco: Jossey-Bass.
Sethi, R., Smith, D,C., & Park, C.W. 2001. Cross-functional product development teams, creativity and the innovativeness of new consumer products. Journal of Marketing Research, 38(1): 73-85.
Steiner, I. 1972. Group process and productivity. New York: Academic Press.
Stork, D., & Richards, W. 1992. Non-respondents in communication network studies. Group
Organization Management, 17(2): 193-209.
Sundstrom, E., DeMeuse, K. P., & Futrell, D. 1990. Work teams: applications and effectiveness. American Psychologist, 45(2): 120-133.
31
Tushman, M. L. 1977. Special boundary roles in the innovation process. Administrative
Science Quarterly, 22: 587-605.
Valacich, J., Wheeler, B., Mennecke, B., & Wachter, R. 1995. The effect of numerical and logical group size on computer-mediated idea generation. Organizational Behavior and
Human Decision Processes, 62(3): 318-329.
Van der Vegt, G., & Bunderson, J. 2005. Learning and performance in multidisciplinary teams: the importance of collective team identification. Academy of Management Journal, 48(3): 532-547.
Wamsley, G.L., & Zald, M.N. 1973. The political economy of public organizations. Public
Administration Review, 33(1): 62-73.
Wicker, A., & Mehler, A. 1971. Assimilation of new members in a large and a small church.
Journ al of Applied Psychology, 55(2): 151-156.
Williams, K., & O'Reilly, C. 1998. Demography and diversity in organizations: a review of 40 years of research. In B. Staw, & L. Cummings, Research in organizational behaviour. 70-140. Greenwich: JAI Press.
Wilson, D., & Malik, S. 1995. Looking for a few good sources: exploring the intra-organizational communication linkages of first line managers. Journal of Business
Communication, 32(1):31-48 .
32 APPENDIX A
In this appendix I will explain step by step how I calculated the percentage of
horizontal contacts and the percentage of diagonal contacts with the use of the EI-index. The EI-index is the ratio between the number of external contacts and the number of internal contacts normalized to a value between -1 (only internal contacts) and +1 (only external contacts). With the EI-index I calculated if organizational members had contact with their own supervisor (internal contact) or with other supervisors (external contact), and if they had contact with organizational members at the same hierarchical level (internal contact) or with organizational members at another hierarchical level (external contact). This resulted in the number of internal contacts, external contacts and total contacts, and it resulted in the EI-index score for both supervisor and hierarchy. These data were imported in Microsoft Excel, in the same order as presented in Table 6.
The number of horizontal contacts is identical to the internal score of the calculations for hierarchy. The number of diagonal contacts is the sum of the number of external contacts of the supervisor plus the number of external contacts of the hierarchy, minus the total contacts of the supervisor. The number of vertical contacts is number of external contacts of the hierarchy minus the number of diagonal contacts. Table 6 presents this information in formulas.
33 TABLE 6
Formulas for calculating (the percentage of) horizontal and diagonal contacts
Supervisor Hierarchy Number of contacts direction of ties
Measure Intern Extern Total
EI-index
Intern Extern Total EI-index Vertical contacts Horizontal contacts Diagonal Contacts Total contacts Name A* B C D* E F G* H* I J K L Formula =F-K =E =(B+F)-C =I+J+K
* not used for calculating the percentage of contacts, but to avoid confusion everything is presented in the same order as UCINET delivers the output when calculating the EI-index.
Percentage of contacts
Measure Percentage of horizontal contacts Percentage of diagonal contacts
34 APPENDIX B
Team Performance
Als u de prestatie van dit team vergelijkt met andere teams die soortgelijke taken uitvoeren, hoe goed scoort dit team dan op…
1. Het bereiken van de teamdoelen? 2. Het behalen van deadlines? 3. Werksnelheid?
4. De kwaliteit van het werk? 5. Productiviteit?
