Master Thesis for Human Resource Management
How does team diversity influence team performance?
Team diversity in football teams
By Xiangru Kong
Supervisor: Dr. P.H. van der Meer University of Groningen Faculty of Economics and Business
ABSTRACT
The purpose of this study is to investigate the impact of football team demographic characteristics on football team performance based on the data of European Champion
League. The literature suggests a positive relationship between team average age, team tenure similarity, proportion of players with coaches' nationalities, variance on
players' matches and team performance and a negative relationship between team average tenure, proportion of non-domestic players, proportion of non-European players, proportion of non-white players, proportion of non home languages, number
of nationalities and team performance. A statistical positive significant relationship was found between team performance and team average age, variance on matches, proportion of non-EU players and proportion of players with coaches' nationalities. No significant results are found for other hypothesis. From the result we can see that
diversity in team composition can influence team performance both positively and negatively. The positive influence of team diversity to team performance is due to improve the quality of team players. The negative influence of team diversity to team
performance is due to block team players’ communication in one team.
INTRODUCTION
In recent years, football has become the most popular sport in the world. And with the development of professional football teams, competitions between teams become intensive. The results of football teams influence money they get and numbers of fans they attract. So all teams try to do anything they can to achieve good results. Changing team players of teams is an important method all teams take. But how to choose the right player for a team and how much improvement the player can make for the teams are difficult problems. Now most teams focus on individual player’s skill. This study tries to find answers in another direction which is the team composition. This study tries to find the relationship between demographic characteristics of team composition and team performance. In order to find the answer to that relationship, this study analyzes teams which attend the European Champion League (ECC). One reason for choosing this data source is that this league includes almost all best football teams in the Europe. Another reason for ECC is that the data can be collected completely. The demographic characteristics related to this study include age, tenure, nationality, language and match.
Research question
There are some researches to study relationship between individual players’ performance and team performance. This study tries to regard team players as a group and answer questions related to team success and team demographic composition. Because the demographic part of team composition is the main focus, this paper tries to answer the following question:
What is the relationship between the team demographic composition and team success?
LITERATURE REVIEW
Section 2 is a literature review. It comprises of three sub-sections. The first one deals with the relation between the degree of team demographic composition and team performance considering communication problems in teams while the second one concerns the relation between team composition and team performance considering quality of players. And the third one concerns relation between coaches and teams.
Team performance and team demographic composition on communication aspect
Demographic characteristics such as age, tenure, education, race, and sex have been related to team performance (Waldman and Avolio, 1986; Wagner, Pfeffer, and O'Reilly, 1984). And the demographic composition of groups has been related to firm-level performance (Wagner, Pfeffer, and O'Reilly, 1984).
The distribution of demographics within an organization unit impacts the amount of conflicts regarding to demographic factors, including race, age, sex, job tenure, and unit performance and turnover (O’Reilly, Williams, & Barsade, 1998).
Demographic dissimilarity was found to be positively associated with perceived intragroup emotion, negatively associated with ratings of group productivity (Pelled, 1996) and negatively associated with team outcomes (Jackson et al., 1995; Milliken & Martins, 1996).
Demographic similarity has been related to high levels of commitment (Riordan and Shore, 1997) and great group productivity (Wiersema and Bantel, 1993). Organization members who are more demographically similar to other members in groups in terms of age, tenure, education, sex, and race will interact more frequently during work than members who are less demographically similar (Chatman, Polzer, Barsade and Neale, 1998; ZEnger and Lawrence, 1989). Weirsema and Bantel (1992) have noted that homogeneity on demographic traits has been shown to lead to a shared language enhancing communication frequency and integration among individuals. As a result, homogeneous teams would be more likely to perform at a high level on tasks that require coordinated activities between team members.
result in avoidance and even dislike which can reduce communication among members (Wagner, Pfeffer and O'Reilly, 1984).
Chiocchio (2007) finds that communication is a key factor in team performance by analyzing 34 teams with time-series analysis. Results suggest that compared to low-performing teams, high-performing teams exchanged more messages. So communication can be regarded as an indicator for team performance. The more communication team members have the better the team will perform.
As mentioned before, demographic similarity can lead to frequent communication; and frequent communication will lead to good team performance. So the demographic characteristics causing good communication will lead to good team performance because good communication can lead to good team performance. Then demographic similarity will lead to good team performance because it causes good communication in teams.
Tenure and age
Pfeffer (1981, 1983) suggested that some demographic traits of team members such as age and tenure influence communication frequency and thus produces organizational outcomes. Individuals who enter a group at the same time or who are of similar age may become more tightly bound to one another than individuals who are demographically heterogeneous (Pfeffer, 1985).
Related research has suggested that a shared language in a group determines the efficiency of communication (Allen, 1969; Dearborn, 1958; Katz, 1966; Newcomb, 1953; Runkel, 1956; Triandis, 1960; Tushman, 1978). The shared language in a group reflects similarities in how group members interpret, understand, and respond to information. Members who are unfamiliar with the shared language are likely to distort and misinterpret information received from other group members and disturb communication in the group (Barnlund, 1963; Roger, 1971).
The development of a shared language between team members results from their similar backgrounds and experiences. The backgrounds members share are similar backgrounds and experiences members share outside organizations which can influence members’ attitude, interests and beliefs (Rhodes, 1983; Ryder, 1965). The experiences members share includes common interpretations of vocabulary and events which can facilitate work-related communication (Allen, 1969; Lawrence, 1967).
Members with similar age tend to share common non-work-related backgrounds because they tend to be at similar points in their family lives (Lawrence, 1980). The more similar people are in age, the more likely they are to hold similar attitudes, interests, and beliefs, and thus the more likely they are to communicate with one another (Riley; Ryder, 1965). Zengler and Lawrence (1989) also find that age diversity is negatively related to the profitability of organizational units.
been related to frequency of communication (Zenger and Lawrence, 1989) and some measures of performance (O'Reilly and Flatt, 1989). Tenure diversity can be regarded as a strong predictor of poor performance (Mannix &Neale, 2005).
As mentioned before, the similarity of members’ tenure can increase communication. But when a group's average organizational tenure increases, communication frequency within the group may decrease (Lawrence, 1997). During the initial years of organizational tenure, the group requires frequent communication to complete tasks and develop a common bond among members that facilitates communication. However, overtime, the group develops work routines which decrease the need for frequent communication because of common bonds. So the frequency of communication decreases and leads to decreasing performance of groups.
Today intensive competition among football teams leads to frequent player transfer. Players of one football team often change in order to achieve good team performance. So the tenures of team players often change because of high frequency of players transfer between teams. Those changes can change demographic composition of football teams which will influence communication of teams.
According to aforementioned analysis, the following hypotheses are formulated: Hypothesis 1: The degree of similarity on team players’ age is positively related to the team performance.
Hypothesis 2: The average age of team players is positively related to team performance.
Hypothesis 3: The degree of similarity on team players’ tenure is positively related to the team performance.
Hypothesis 4: The average tenure of team players is negatively related to the team performance.
Nationality and race
In a study of workers in the Netherlands, Verkuyten, de Jong, and Masson (1993) found that individuals who were not Dutch tended to be less satisfied with their jobs than others who are Dutch. In a study in Australia, people who are not Australian perceived more discrimination in their workplace than their Australian counterparts (Dochner &Hesketh, 1994).
Blalock (1957) argued that discrimination by the majority will increase as the proportion of the minority increase. Hoffman (1985) found a tendency toward decreased frequency in interpersonal communication among members of the majority as minority composition increased. The reaction of the majority toward minority may be a result of individuals' needs which maintain and enhance the positively valued distinctiveness of their in-groups compared with out-groups in order to achieve a positive social identity (Tajfel & Turner, 1986).
outward expressions if the minority race outwardly expresses inferiority.
In a football team nationality is a good indicator when players categorize themselves. When players categorize themselves, they will form subgroups where they communicate more than those players who are not belonged to subgroups. When there are many nationalities in a team, there are many subgroups which will be formed. And those subgroups are not good for the whole team communication because they will decrease the frequency of communication between players with different nationalities. And because there are restrictions on players who are not European players in European football teams, whether players are European can be used for categorizing. After the categorization, players from different categorization will interact with each other differently. Because players in one football team are not from one country, players may speak different languages. The dissimilarity of languages will influence communications in football teams.
Race is also another indicator for players to categorize themselves. Race difference may also influence communication in teams. As mentioned before with the increased number of non-white players in one team, the reactions of white players to non-white players will increase and communication between non-white player and white players will decrease.
Several hypotheses are formulated:
Hypothesis 5: The proportion of non-domestic players in football teams is negatively related to team performance.
Hypothesis 6: The proportion of non-European players in a football team is negatively related to team performance.
Hypothesis 7: The number of nationalities in a football team is negatively related to team performance.
Hypothesis 8: The proportion of languages players speak which are not home language in football teams is negatively related to team performance.
Hypothesis 9: The proportion of non-white players in football teams is
negatively related to team performance.
Team performance and team demographic composition attributing to quality of players
Age diversity and average age
As mentioned before, age diversity influences team performance negatively. But age diversity is also good for organizational units to some degree (Lazear, 1998). Young workers bring new skills and new ideas with them into the organization; and senior workers may have a much better handle on the information that is most relevant to this specific organization. It is possible that some mixture of young and old can produce the most productive work environment. Young workers can introduce new techniques and new ideas to older workers. Older workers can impart the knowledge that they have obtained through years of experience about the industry and especially of the organization in which they work.
force while keeping its stable stance. When a team tries to maximize its competition force, there are three things to think of which are team players skills, experience and coordination. All those things depend on quality of players. Normally a player who has played for several years gains much experience and highest skill level and shows best performance of his life. But when a player achieve some age, such as 32, although he gets a lot of experience and skills, the strength and stamina of the player will decrease. The decrease of players’ strength and stamina will decrease the performance of players.
According to aforementioned analysis, the following hypotheses are formulated: Hypothesis 10: The degree of diversity on team players’ age is positively related to the team performance.
Hypothesis 11: The average age of team players and team performance shows invert U relationship.
Power of money
With the development of modern football, the financial bases of football teams play an important part in team performance. Every team player in football team has his own value depend on his quality. The better quality the player has the more money he is worth. Only those teams with much money can buy the best players in the world. But the number of best players in one country is limited. So the competition between teams which try to buy domestic good players is intensive. To achieve the best benefit of teams, teams can buy better players with the same money which can buy good domestic players. But almost all European leagues have limitation on non-domestic players; all teams pay a lot of attention to choose non-domestic players. For that reason, the overall quality of non-domestic players is better than the overall quality of domestic players in teams. To summarize, the more non-domestic players a football team has mean the more good players the team has, and the better result the team may get. According to this analysis, the following hypotheses are formulated:
Hypothesis 12: The proportion of non-domestic players in football teams is positively related to team performance.
Hypothesis 13: The proportion of non-European players in football teams is positively related to team performance.
Leader’s characteristics and team performance
Tsui and O'Reilly (1989) found that dissimilarity in superior-subordinate demographic characteristics is associated with lower effectiveness as perceived by superiors and less personal attraction on the part of superiors for subordinates. Jackson (1991) found that dissimilarity in demographic characteristics lowered individual performance ratings. The greater the dissimilarity in demographic background between a superior and a subordinate, the more negative will be such job outcomes as performance, affect expressed by the superior toward the subordinate, and role ambiguity and conflict as experienced by the subordinate (Tsui &O'Reilly, 1989).
dissimilarity in demographic characteristics lowers individual performance ratings, coaches may tries to choose players who are similar in demographic characteristics. So coaches will choose players with the same nationality because they higher those players rating. And the demographic similarity between superiors and subordinates are positively related to performance (Tsui &O’Reilly, 1998), so the demographic similarity between coaches and players are possible to lead to good team performance. Another reason for coaches to choose players with their nationality is that they are familiar with those players. Normally a coach tries to choose best players based on his knowledge, and he knows more players with his nationality than international players. So there is bigger chance for him to choose players with his nationality over international players. But there is one more thing to think about which is the nationalities of the team. If teams’ nationalities and coaches’ nationalities are the same the willing of coaches to choose players with his nationality will decrease because there are already many domestic players in that team.
According to aforementioned analyses, one hypothesis is formulated:
Hypothesis 14: The proportion of players with coaches’ nationalities is positively related to team performance.
Coaches’ policy and team performance
After considering team composition, I would like to consider a policy which every coach faces, player rotation. Some coaches like to rotate and others don’t like. In my opinion there are several reasons to apply play rotation. The first one is that players can be tired if they play all matches. If a player can rest in several matches, it is easy for him to maintain his best performance which is good for a team. The second reason is that player rotation can minimize the influence of accidents such as yellow card, red cards and injury. The third reason is that coach can manipulate players by using player rotation. When all players have the threat of not playing for the team, they will work hard in the match in order to play as many matches as possible. From the reasons I mentioned, I assume that player rotation will be positively related to team performance. Variance of matches players play can be a good index for player rotation because the high variance of matches means the more players shows in matches.
According to aforementioned analyses, one hypothesis related player rotation is formulated:
METHODOLOGY
SampleAs this study focuses on team performance in the European Champion league, only teams which attend the group stage are collected. When collecting the data of players, players who do not play in the European Champion league will be erased.
Data sources
The results of European Champion league matches are collected from UEFA official website. The personal information of players and coaches are retrieved from www.footballdatabase.eu. Every team official websites are consulted when some data about team players is missing.
Variables
Dependent variables
Team performance: There are many ways to measure team performance like goal differences and championships won.
Because in the Champion league not all teams meet in one season, it is hard to compare team performance directly. So I will adopt a method which is used for European team league ranking by "Eurotopfoot" (http://www.eurotopfoot.com). As Table 1 and Table 2 show, during a season a team which play at home can get 3 points when it won, 1 point when it draw and -1 when it lost the match. When a team play away it can get 4 points for win, 2 points for draw and 0 points for lost. A team can get 30 points for playing in 1/8 final , 40 points for playing in 1/4 final, 50 points for playing in 1/2 finals, 50 points for playing in final and 50 points for winning the champion. So all points teams get in one season stand for their performance in this season.
Table 1 (The points are distributed during the season):
1/8 Final 1/4 Final 1/2 Final Finalist Champion Total Champion
League
30 pt. 40 pt. 50 pt. 50 pt. 50 pt. 220 pt.
Table 2 (The bonus points are awarded to teams during the season):
Home Away Won 3 4 Draw 1 2 Lost -1 0 Independent variables
get.
2. Team players’ tenure: There are also two variables used to study team players’ tenure which are average on tenure and variance on tenure. Variance on team players' tenure is used to measure the degree of similarity on team players’ tenure. The more variance on team players' tenure shows the lower similarity of team players' tenure.
3. Variables about nationalities: There are three variables which study teams’ nationalities. The first one is proportion of non-domestic players. Whether players are domestic players is judged by the nationalities of teams and nationalities of team players. If a player’s nationality is not the same as his team's nationality, he is regarded as a non-domestic player. The other two variables are number of nationalities and proportion of coach nationality.
4. Number of languages in one team: It is measured by the sum of languages players speak in one team. The languages team players speak are assumed according to the nationalities of the players. If a player can speak several languages, the most useful language other players use will be regarded as that player's language.
5. Proportion of non-white players: The player's race is judged by the author by looking at the player's picture.
6. Proportion of non home language: This variable is used to measure the proportion of players who do not speak the language which the team speaks.
7. Variance on match. This variable is used to test the degree of similarity on team players’ match in one season.
Dummy variables
Because the overall team performance is different every year. In order to incorporate the year difference into the regression model, we include dummy variables, year1, year2, year3 and year4.
DESCRIPTIVE RESULTS
Basic Information of Sample TeamsThere are 32 teams which attend the European Champion League every year. So there will be 160 team records during Season 2003-2007. Because of the limitation of the channel to collect data, there are only 152 team records which can be achieved. Among the 152 team records, there are 58 teams which actually attended the European Champion League during season 2003-2007. Among the 58 teams there are 10 teams which attended all 5 seasons ECC, 7 teams attended 4 of 5 seasons ECC, 10 teams attended 3 of 5 seasons ECC, 13 teams attended 2 of 5 seasons ECC and the rest only attended one season ECC during Season 2003-2007.
Team performance
Among the 152 team records, the least point a team get during one season is -3 and the highest point a team get during one season is 265. The mean point a team gets is 48.05.
Independent variables
Table 3 show descriptive results of independent variables. The youngest team in 58 teams is Ajax whose average is 23.14 in season 2004. The oldest team is AC Milan whose average age is 31.35 in season 2005. The average age of all teams is 27.27 and most teams’ average ages are between 25 and 29 which are usually the best year of team players.
The least average tenure is 1.67 and the biggest average tenure is 5.85. The average tenure on players is 3.26. The mean variance on tenure is 7.04 and most teams' variance on tenure is less than 10. Because the mean variance on tenure is much bigger than the average tenure on players, football teams change many players every year.
When we see the independent variable, 'number of nationalities', we can find that all teams have international players. There are big differences on numbers of nationalities in teams. The least number of nationalities is 2 and the biggest number is 14. The mean number of all teams is 8.18 and more than half of teams have 6 to 11 nationalities. So there is high degree of internationality in teams attending ECC. Because of high degree of internationality, the average proportion of non home language is 45.56 and the average proportion of non domestic players is 52.1. Because of the restriction on non European players the average proportion of non European players in teams is not high compared to high degree of internationality, which is 24.44.
TESTING HYPOTHESES
Before running multiple regressions, the correlation matrix of all variable is examined.
Correlation analysis shows that most variables I select are significant related to team performance. Some variables such as team players' average age, average tenure, age similarity, matches similarity, tenure similarity, proportion of non domestic players, proportion of non home language, number of nationalities in team and coach code are positively related to team performance. Only one variable which is proportion of white players in teams is negatively related to team performance. No significant relation is found for the other variables. The details of correlation test can be found in Table 3.
In order to find the relationship between average age and team performance, a curve estimation analysis is made and the result shows a positive relationship between team performance and average. No invert U relationship is found as expected.
In order to control the correlations among the independent variables, multiple regression analysis is followed.
Because there are high correlation among variables related to age and variables related to tenure, variables related to age and variables related to tenure will not show in the same multiple regression analysis. There are two multiple regression analysis. Similar high correlation happens among variables related to race, nationality. So variable 'proportion of white', 'proportion of non EU' and 'proportion of non domestic' will be analyzed separately. To summarize, there will be 6 multiple regression analysis.
The results show that the independent variables I choose explain team performance quite well. All six F-tests are significant. From those regression analysis, five variables ‘Average on Age', 'Variance on Match', 'Proportion of Non-EU', 'Proportion of Coach nationality' and 'Coach code' are significant in all related regressions. Two Variables 'Variance on age', 'Average on Tenure', 'Proportion of Non Domestic', 'Proportion of Non Home Language' and four dummy variables 'year1','year2','year3','year4' show no significant in all related regression. Three variables 'Variance on Tenure', 'Proportion of White' and 'Number of Nationality' are significant in some of their related regressions. More information about the regression is displayed in Table 5.
team performance. No significant relation is found for the other variables.
Table 3
Correlation Matrix of All Variables
Table 4
Regression Results of Performance focusing on age aspect (152 teams)
Table 5
Regression Results of Performance focusing on tenure aspect (152 teams)
CONCLUSION
This study looked at different characteristic of team composition. It tried to gain the relation between team composition and team performance. Combined with hypotheses, the results I get show useful information for analyzing team performance.
The regression analyses shows the positive relations between teams' average ages, variance on matches, proportion of non EU players, proportion of players with coach's nationality and coach code and team performance. Numbers of nationalities are partly positively related to team performance and proportions of white players are partly negatively related to team performance.
Hypothesis 1, 2, 10 and 11 tries to find relation between age and team performance. The result shows positive relation between team performance and average age.
Hypothesis 3 and 4 tries to find relation between tenure and team performance. But no relation is found. One reason may be that I do not think of the influence of coaches at the same time. A team's style depends on its coach's tactics. Even a team does not change many players, if its coach changes it will take a long time for team players to be familiar with new tactics. So how many coaches football teams change and how long those coaches’ tenures are when players are in football teams influence the effect of players’ tenure. So the result might be significant if both team players tenure and coaches’ tenure are considered at the same time.
Hypothesis 5,6,7,8,9,12 and 13 tries to find relation between nationality and race and team performance. The result shows proportion of non-EU players is significantly positively related to team performance. Although other related variables are not significant, from regression analysis, proportion of non-white players, proportion of non domestic players and number of nationalities are positively related to team performance. This study tries to analysis the relation between team composition and team performance based on communication but the result shows that quality of players may a bigger factor to influence team performance. The positive relation between non-EU players and team performance shows that good quality of non-EU players influence team performance more than communication problems they cause. The other variables related to hypothesis 5,6,7,8,9,12 and 13 are not significant related to team performance. On one hand the demographic diversity may destroy communication in teams; on the other hand the demographic diversity may increase the quality of team players in teams. So the demographic diversity shows both positive and negative influence to team performance at the same time. It is impossible to find the influence of one aspect without erasing the influence of the other aspect unless the influence of one aspect is big enough to overcome the influence caused by the other aspect.
which have international coaches who brought as many players with their nationalities as possible can show best performance among all teams attending ECC.
Hypothesis 15 is supported by multiple regression analysis. It means that player rotation is a good policy for team performance. But in reality no one team change many players in every match. So I compute average matches players played in one team and try to find the correlation. The correlation is 0.411 and significant at the 0.01 level. If the result is negative, it means that total player rotation is best for team performance. If the result is positive, it means that a team should focus several players to play the whole season. Now the results show that both average on matches and variance on matches are positively related to team performance. It means that a team should rotate less than half players while keep its core players playing as many matches as possible in order to get good results.
DISCUSSION
In this section, the limitations of my study and some interesting topics for further research are discussed.
Limitation
The first limitation of my study is the influence of quality of players. Normally international players are better than domestic players. Even international players may cause communication problems in teams, but their high quality may be good for team performance. If people cannot find a way to separate these two aspects, some research can not be successful.
The second limitation of my study is the different financial base of teams. Even teams I select are best teams of all Europe; there is still big difference among those teams. Different budgets of different teams mean different power to select players. Teams with more money can select better players and it will lead better team performance. Those differences are not caused by communication problems. So people should find a way to balance the financial power if they want to find more relation between team composition and team performance.
The third one is the data I select about languages players speak. I can not know what languages players can speak exactly. Some players can even speak 5 languages and even non-EU players normally can speak English. What I can do is to assume players' languages according to their nationalities. So it is difficult to analyze relation between communication problems caused by languages and team performance.
Potentials for Further Research
As mentioned before, quality of players and financial base of teams are two aspects to think about team performance. Quality of players can be balanced by data of individual performance and salary of players. Financial base of teams can be calculated by budgets every teams report every year. If people think of quality of players, financial base of teams and team compositions at the same, team performance might be explained more accurately.
Acknowledgement
I would like to thank Dr Van der Meer. He offers many helpful and constructive comments and suggestions on this paper. His patience and kindness are greatly
Reference
Ancona, D.G., & Caldwell, D.F. 1992. Demography and design: Predictors of new product team performance. Organization science, 3
Bowers, C.A., & Pharmer, J.A., & Salas, E. 2000. When member homogeneity is needed in work teams. Small Group Research, 31: 305
Chatman, J.A., & Polzer, J.T., & Neal, M.A. 1998. Being Different Yet Feeling Similar; The Influence of Demographic Composition and Organizational Culture on Work Processes and Outcomes. Administrative Science Quarterly, 43: 749-780
Chiocchio, F. 2007. Project team performance: A study of electronic task and coordination communication. Project management journal, 27
Horwitz, I.B., & Horwitz, S.K. 2007. The effects of team diversity on team outcomes: A meta-analytic review of team demography Journal of Management, 33: 987
Lawrence, B.S. 1997. The Black Box of Organizational Demography. Organization
science, 8
Lazear, E.P. 1998. Personnel economics for managers. New York
Michalism, M.D., & Karau, S.J., & Tangpong, C. 2007. Leadership’s Activation of Team Cohesion as a Strategic Asset: An Empirical. Journal of Business Strategies, 24
Milliken, F.J., & Martins, L.L. 1996. Searching for common threads: Understanding the multiple effects of diversity in organizational groups. Academy of Management
Review, 21: 402-433
O’Reilly III, C.A., & Caldwetl, D.F., & Barnett, W.P. 1989. Work group demography, social integration, and turnover. Administrative Science Quarterly, 34: 21-37
Rockett, T., & Valenti, M.A. 2008. The effects of demographic differences on forming intragroup relationships. Small Group Research, 39: 179-202
Sacco, J.M., & Schmitt, N. 2005. A Dynamic Multilevel Model of Demographic Diversity and Misfit Effects. Journal of Applied Psychology, 90: 203-231
Tsui, A.S., & O’reilly III, C.A. 1989. Beyond simple demographic effects: The importance of relational demography in superior-subordinate dyads. Academy of
Management Journal, 32: 402-423.
Demography and Organizational Attachment. Administrative Science Quarterly, 37: 549-579
Watson, W.E., & Kumar, K., & Michaelsen, L.K. 1993. Cultural diversity’s impact on interaction process and performance: comparing homogeneous and diverse task groups. Academy of Management Journal 36: 590-602
Zenger, T.R., & Lawrence, B.S. 1989. Organizational Demography: The Differential Effects of Age and Tenure Distributions on Technical Communication. The Academy
