Managing Multi-Cultural Teams: From a cross-cultural to a global perspective

thunderingaardvarkΤεχνίτη Νοημοσύνη και Ρομποτική

18 Νοε 2013 (πριν από 3 χρόνια και 10 μήνες)

112 εμφανίσεις

1


FINAL REPORT

GRANT #128

Managing Multi
-
Cultural T
eams:
From a cross
-
cultural to a global
perspective



Miriam Erez



Submitted to the SHRM




June 2011




The two studies summarized in this report were supported by SHRM

Foundation
. We thank
SHRM
Foundation
for their generous
support
.

Th
e

first study summarized in this report was based on a cross
-
cultural team project,
developed and conducted by Prof. Miriam Erez and coordinated by her doctoral student Alon
Lisak.
Data for the field study were partially collected by help of Dr. Michaela Schipper,
Rotterdam School of Management, Erasmus University, The Netherlands. The two studies

were part of the

doctoral dissertation of
Dr. Alon Lisak
, Faculty of Industrial Engine
ering and
Management, Technion
, Israel
.

2






Abstract

As part of the globalization process, a growing number of employees in Multi
-
National
Organizations (MNOs) face the new reality of working in Multi
-
Cultural Teams (MCTs).
Although a plethora of articles concerning MCTs have been published in the last decad
e, most
of these studies didn't consider the role of leaders and followers in the MCTs as part of their

research models (Lisak & Erez, 2009).

In this research, we suggested a model which emphasized

both global leadership behaviors
and followers' openness

to cultural diversity as antecedents for desirable MCT outcomes.
Based on Self
-
Concept
-
Based Leadership Theories (Lord et al., 1999; Shamir et al., 1993)
and on global work values typologies (Erez & Shokef, 2008)
,

we asserted that global
leadership behavi
ors, which convey a collective sense of global identity, interdependence and
openness to cultural diversity, are related to MCT identity. This relation is positively
moderated by followers' openness to cultural diversity. Additionally, MCT identity leads t
o
MCT

effectiveness.

Our research included two studies
:

Study 1 consisted of 282 MBA students from 42
nationalities
,

working in 73 virtual, short term project MCTs. The results of study 1 supported
the suggested model. Study 2 consisted of 274 employ
ee
s, working in 55 on
-
going MCTs in 9
MNOs. In this study, the research model was expanded by adding team trust as a mediator of
the interaction relation between global leadership behaviors and followers' openness to
cultural diversity on team identity. The
results partially supported this model. Global
leadership behaviors were positively related to team trust and team identity. However, the
strength of this relationship decreased as the level of followers' openness to cultural diversity
increased.

Our rese
arch results highlight the importance of considering global characteristics of leaders
and followers in MCT effectiveness models. The theoretical and practical implications of
these findings
are

discussed in the
report
.









3


I) Introduction

As part of
the globalization process, Multi
-

National Organizations (MNOs) form
international teams to pool global talent, meet organizational goals and implement complex
business strategies (Bartlett & Ghoshal 1989; Joshi, Labianca &Caligiuri, 2002). As an
outcome o
f this continuous process, a growing number of employees in MNOs face the new
reality of working in Multi
-
Cultural Teams (MCTs) (Distefano & Maznewski, 2000), which

can be defined as "a group of people from different cultures working together on activities

that span national borders" (Snell, Snow, Davidson & Hambrick, 1998)
.

Early definitions of MCTs created a dichotomy between two types of MCTs: collected MCT,
whose members work as a team in the same physical location, and virtual MCT
s?
, whose
members rel
y totally on computer
-

mediated interaction to complete their team assignments
(Stanko & Gibson, 2009). Recently, studies have found mixed results regarding the relations
between the level of virtuality and team effectiveness (see review at Martins, Gilson

&
Maynard, 2004) and it was suggested that positive team processes appear in different levels of
virtuality (Fiol & O'Conor ,2005; Wilson, Straus & McEvily, 2006). Therefore, scholars have
shifted away from this dichotomy and started to describe most MCTs

on a continuum of
virtuality (Stanko & Gibson, 2009; Webster

& Wong, 2008
).

MCTs face unique difficulties which characterize the global environment, such as: lack of
shared meaning, communication problems, cultural conflicts and differences in regulation
s
and work procedures (Cascio & Shurygalio, 2003; Gibson & Cohen, 2003; Halevy & Sagiv,
2008; Vodosek; 2007). Thus, the ability of both the MCT leader and the MCT members to be
effective in their roles is one of the main factors contributing to the creatio
n of an effective
MCT (Kayworth & Leidner, 2002; Maznevski & Distefano, 2000; Vallaster, 2005).

MCTs play a major role in the success of MNOs (multinational organizations) (Earley &
Gibson, 2002). Therefore,
a plethora of articles concerning MCTs have bee
n published in the
last decade (
Maloney & Zellmer
-
Bruhn, 2006; Stanko & Gibson, 2009). Surprisingly, most of
these studies didn't consider the role of leaders and followers in the MCTs as part of their
research models (Kearney & Gebert, 2009; Malhorta, Maj
charzak & Rosen, 2007) and there
are hardly any empirical studies to test the effects of leaders and followers on the success of
an MCT (Joshi & Lazarova, 2005; Lisak & Erez, 2009).

We suggest two possible explanations for this lack of research literature
, both of which can
serve as a starting point for developing a model of effective MCTs:





4


Focus on Cross
-
Cultural Leadership studies
rather

than Global Leadership studies for
explaining MCTs phenomena

"There are many leaders who keep their local leadership style, when they lead MCTs. It will
not work for them. When I lead a MCT, I compromise, I'm flexible and I'm sensitive. As a
MCT leader, I must focus on the common subjects and not on the differences…
" (Rachel,
MCT leader).


Most research on leadership and culture represents the Cross
-
Cultural Leadership (CCL)
perspective and not the Global Leadership (GL) perspective (Adler, Miller &Von
-
Glinow,
2001; Boyacigiller & Adler, 1991, 1996; Osland, Taylor
& Mendenhall, 2009). In their recent
review, Osland et al., (2009) found only 14 GL empirical studies which had been published.
Hardly any of the
se

were published in peer
-
review journals and most were qualitative.

Most of the CCL research examined differ
ences and similarities in leadership characteristics
across cultures (e.g. Dorfman et al., 1997; House, Hanges, Javidan, Dorfman & Guta, 2004;
Spritzer, Perttula & Xin, 2005; Wendt, Euwema & Emmerik, 2009). Other studies compared
values of expatriates with

those of their subordinates in the host country (e.g. Chen, Choi &
Chi, 2002; Shaffer, Harrison & Gilley, 1999; Van
-
Vianen, Pater, Kristof
-
Brown & Johnson,
2004). In contrast to the above research, MCTs operate in a global context, where people
from diffe
rent nationalities work together towards the accomplishment of a global mission. In
such a case, comparisons between two national cultures can be meaningless. Additionally,
understanding leader
-
followers relations in local teams may not explain leader
-
foll
owers
relations in a global context, such as MCTs (Gelfand, Erez & Aycan, 2007; Erez, 2010).
Therefore, in contrast to CCL studies, most GL studies reflect geocentric and synergetic
research methods (Adler, 1983; Osland et al, 2008). Geocentric research f
ocuses on efforts
towards understanding the complexity of cultural issues related to the function of MNOs.
Synergistic research attempts to uncover the universal processes which occur in the global
context, where people from heterogeneous national culture
s work in the same organization
(Adler & Bartholomew, 1992; Boyacigiller & Adler, 1991). GL research, therefore, suggests
a different way of thinking, which emphasizes that managing in a global context (such as
MCT) is related to the creation of a new enti
t
y

which emerges through the interactions of
individuals from different national cultures who work together (Adler, 1983)

Furthermore, CCL studies focus on the national layer of culture, while GL studies focus on
the global layer of culture (Erez, 2010).
CCL research explores similarities and differences
between leaders' national cultural values and national clusters (see review at Dickson, Den
-
Hartog and Mitchelson, 2003; Dickson, Den
-
Hartog and Castano, 2009). These values (e.g.
Hofstede, 1980; 2001; Sch
wartz, 1992; 1999; House et al., 2004) reflect differences in
leadership perceptions and behaviors in different nations and focus, therefore, on the national
5


layer of culture. Recently, Erez and Shokef (2008; see also Erez & Gati, 2004; Shokef & Erez,
2006
) proposed that
in the context of MNOs and MCTs, the global work culture emerges as the
most macro
-
level culture, shared by members of MNOs and of other international organizations
and alliances, operating globally beyond national boundaries
. This layer of

culture holds a
unique set of global work values, which is shared by global work employees (Shokef & Erez,
2006).

GL researchers attempt to reveal dimensions and components which reflect a unified global
context which leaders operate in (e.g. Bird & Oslan
d, 2004; Lane, Maznevski & Mendenhall,
2004) and therefore refer to the global layer of culture. This new way of thinking can
potentially offer explanations for leadership processes which exist in unique global
environments, such as MCTs.

Based on the abo
ve, for better understanding of effectiveness in MCTs, we propose to adopt a
Global perspective, rather than a Cross
-
Cultural perspective.



Focusing on both leaders and followers rather than adopting a “Leader
-

Centric” perspective

"A global leader must r
ely on his/her followers to gain success. An MCT leader is an
orchestra Maestro. Despite the fact that the players are from different cultures, and their
original music was written in different languages, in the end of the day all players need to
write the

music together and play it with the Maestro"


(Zvi, MCT leader).


With a few exceptions (e.g. Graen & Cashman, 1975; Fidler, 1967), most of the empirical and
theoretical studies in the general field of leadership have focused on the behaviors and traits
of the leaders. The majority of leadership literature
-

while ostensibly focusing on the leaders
-

has neglected the important role of the followers in defining and shaping the latitude of the
leader's actions (Bass, 1990; Dvir & Shamir, 2003; Hollander, 1
992; Howell & Shamir, 2005,
Van
-
Knippenberg, Van
-
Knippenberg, De
-
Cremer & Hogg, 2004 ).

The importance of followers to leadership and team effectiveness has gained a wider
recognition in the last two decades (Hollander, 1992; Howell & Shamir, 2005; Kark &

Van
-
Dijk, 2007; Klein & House, 1995; Meindl 1990, 1995). These researchers and others
criticized the "leader
-
centric" perspective dominating the literature and emphasized the
contribution of followers to leadership and team success. Some of these theories

were
"followers
-
centric" (e.g. Meindl 1990, 1995), but other focused on the relationships between
leaders and followers, pointing at the influence of followers on the success of the leader (e.g.
Graen & Uhl
-
Bien, 1995, Howell & Shamir, 2005, Wilson et al.
, 2010).

As the current GL research is mainly a reflection of the general leadership research (Osland,
2008), it is not surprising that most of the research in the field of global leadership attempts to
6


identify leaders' competencies, which can assist in

achieving their tasks (Jokinen, 2005; Joshi
& Lazarova, 2005; Morrison, 2000; Suutari, 2002). Theses studies may be utilized for
understanding effectiveness in MCTs. However, due to the complex global work context
(Shokef & Erez, 2006), the requirement fo
r global integration which influences MCT
effectiveness (Bartlett & Ghoshal, 1992, 1994) and the fact that MCT followers' adaptation to
the global context is crucial for MCT effectiveness, there is a need to include followers'
factors which contribute to M
CT effectiveness in related research models. Hence,
effectiveness of MCTs depends on the global characteristics of both the leader and the
followers and the relations between them.


During the last decades, several MCT studies have examined the manner in w
hich team
members can contribute to team effectiveness. Most of these studies concentrated on
members' contribution to desirable MCT processes and outcomes (see reviews at Gelfand et
al., 2007 and at
Connaughton & Shuffler
, 2007). Other studies connected g
lobal
characteristics of MCT members (e.g. global mindset (Malloney & Zellmer
-
Bruhn, 2006),
cultural intelligence (Earley & Gardner, 2005) and team identity (Shokef & Erez, 2006)) to
MCT effectiveness. However
, these studies neither examined the specific r
ole of followers in
their models, nor did they consider the followers' interaction with their leader for gaining
effective outcomes (Lisak & Erez, 2009). Therefore, in this work, we will consider both the
leader's and the followers' characteristics which c
ontribute to MCT outcomes. Specifically, as
illustrated in Figure 1, we emphasize these factors as two antecedents of team outcomes. The
first factor involves the unique global behaviors of the MCT leader, the second is related to
the MCT followers' openne
ss to cultural diversity, which is a unique global characteristic that
reflects
individuals who are open to cultural diversity and are motivated to actively seek new
cultural experiences (Hartel & Fujimoto, 2000; Shokef & Erez, 2006)
. According to the
mode
l, both global leaders' behaviors and followers' openness to cultural diversity interact
with each other to create team identity, which is a
joint perception of group cohesiveness
(
Earley & Mosakowsky, 2000). Team Identity, in turn, contributes to the crea
tion of MCT
effectiveness.


Figure 1


Research Model (study 1)







Followers' Openness
to Cultural Diversity

Global Leadership
Behaviors


Team Identity


MCT

Effectiveness

7


In the next sections we further elaborate on the influence

of global leadership behaviors and
followers' openness to cultural diversity on the creation of team identity. Then we will
explain how team identity contributes to the creation of MCT effectiveness.


A. Research Literature and Research Hypotheses


1.
Gl
obal Leadership Behaviors


a.
Current Research in the Field of Global Leadership (GL)

Global leadership can be defined as "
A process of influencing the thinking, attitudes and
behaviors of a global community to work together synergistically toward a common

vision
and common goals” (Osland & Bird, 2006, p. 123)
. Due to the growing presence of
globalization in the last two decades (
Govindarajan & Gupta, 2001
), GL research has gained a
wider attention in the international management community

(Osland et al., 2009).

Until recent
years, most GL studies focused on global leaders' competencies. Jokinen (2005) defined GL
competencies in three major categories: a. “a
core of global leadership competencies”, which
are
fundamental

to the development o
f other characteristics and include
self
-
awareness,
engagement in personal transformation and inquisitiveness
, b “
desired mental characteristics
of global leaders”
, which
are consist of optimism, self
-
regulation, social judgment skills,
empathy, motivation

to work in an international environment, cognitive skills and acceptance
of complexity and its contradictions
, and c.
“desired behavioral competencies of global
leaders”
, which

are explicit skills and tangible knowledge that refer to concrete actions and
produce visible results and include social skills, networking skills and knowledge (See also
Suutari, 2002)
.

Another subject which has received attention during the last few years is the
Global Mindset notion, which is the ability to influence others who a
re culturally different
(
Levy, Taylor, Boyacigiller & Beecher, 2007)
. This notion includes an Intellectual
Component (Global Business Savvy, Cognitive Complexity, Composition Outlook), a
Psychological Component (Passion for Diversity, Quest for Adventure,
Self
-
Assurance) and a
Social Component (Intercultural Empathy, Interpersonal Impact and Diplomacy), which
present different aspects of effective global leadership (
Beechler & Javidan, 2007
;

Levy et al.,
2007
).
Despite this theoretical work of the last few

years, Jokinen summarized: "
Although
the researchers focusing on leadership competencies have described the importance and
causalities of different competencies, they have not often been explicit about the process by
which the competencies affect the perf
ormance outcome"
(2005, p.204). The absence of
theoretical and empirical research concerning the contribution of global leaders to their units
is specifically reflected in poor knowledge regarding leadership factors needed to lead MCTs
(Joshi & Lazarova,
2005).

Despite the lack of empirical research, much of the discussion regarding the global leader
focuses on the effectiveness of Transformational Leadership (TL) and Charismatic
8


Leadership; both

empower their subordinates to perform beyond their own expe
ctations and to
achieve the leader's goals as if they were their own, by inspirational behaviors, intellectual stimuli
and individual consideration for followers. (Bass, 1985, 1997; Brake, 1997; Rosen & Digh,
2000). These theories emphasize the importanc
e of
leadership behaviors (rather than leaders'
traits) such as role model behaviors and inspirational behaviors (
Joshi, Lazarova & Lio ,2009
),
as the main leadership tools for transforming followers' perceptions and behaviors towards the
leader's goals
.

S
hamir et al. (1993)
and Lord et al.
(1999) in their Self
-

Concept Based Theories of Leadership
explained why and how leadership behaviors influence followers' perceptions and behaviors
(Kark & Shamir, 2002)
.

The self is a collection of modular processing
structures (self
-
schemas) that are elicited in
different contexts and have specific cognitive, affective and behavioral consequences. The
self schemas may reflect the individual identity, the dyadic interpersonal

identity and the
collective identity (Brewe
r & Gardner, 1996, Lord, et al., 1999; Markus & Wurf, 1987).

Based on the assumption that an individual can activate one specific self
-
identity level at a
time (Baldwin, 1992), Lord et al. (1999) asserted "that
leaders can temporarily influence self
str
uctures through activities that influence the accessibility of various self concepts. For
example, by emphasizing similarities among workers, leaders can increase the activation of
collective identities while inhibiting individual
-
level identities (p.184)"
.

Shamir et al. (1993)
suggested that through charismatic behaviors, such as role modeling and frame alignment, the
leaders can affect the followers' self concept by activating the followers' personal
identification with the leader, or by activating the s
ocial, collective identity in the followers'
self concept. Additionally, the theory suggests that saliency of personal identification with
the leader will increase when the leader represents desirable personal components. The
salience of collective identi
fication in the followers' self concept will increase when the leader
defines the boundaries of the collectivity by emphasizing its distinctiveness, prestige and
competition with other groups. By connecting followers’ self
-
concept to the mission and to
the

team, the followers are motivated to contribute to the team and to the organizational
success beyond their self
-
interests. Therefore, Shamir et al (1993) argued that "charismatic
leaders change the salience hierarchy of values and identities within the fo
llower's self
-
concept, thus increasing the probability that these values and identities will be implicated in
action." (p. 584).


The GLOBE study (House et al., 2004) examined the leadership characteristics that
contribute to leaders' success in 62 countr
ies and supports the assumption of the universality
of charismatic leadership as a success factor. According to the study, charismatic leaders
reflected charisma, inspiration, visionary and supportive leadership behaviors across countries.
However, the GLO
BE study (House et al., 2004) examined cross
-
cultural differences and
9


similarities in successful local leadership behaviors but did not test for the effect of
TL/charismatic behaviors on the success of leaders of MCTs or other global contexts. Only a
rec
ent study showed a
positive effect of inspirational leadership on MCT outcomes (Joshi et al.,
2009). Additionally, TL was found to be a moderator of the relation between cultural diversity and
collective team identification in MCTs (Kearney & Gebert, 2009)
. Therefore, TL/charismatic
behaviors have a positive impact on MCT processes and outcomes. However, these studies on TL
and charismatic leadership do not explain how leadership behaviors are translated into the specific
values and behaviors which the lead
er must reflect to achieve desirable outcomes in MCTs.
Related to this subject, one of the MCT leaders in this study said:

"As a leader, I know that I need to care for my people, to coordinate the work and to motivate
them in many directions to achieve o
ur goals. The problem is that it is not always clear how
to implement it in the complex global team that I have… Our company gives general
leadership training but gives almost no training to become an effective MCT leader" (Dan,
MCT leader).


Therefore, t
here is a need to identify the specific MCT leaders' behaviors which fit in with the
global work context and should serve as models for their MCTs followers (Berson, Erez &
Adler, 2004; Erez & Shokef, 2008; Shokef and Erez, 2006,).

Based on the assertions of Shamir et al. (1993) and Lord et al. (1999), and in the frame of the
global context, we propose that in order to achieve identity and effectiveness in MCT, the leader
must emphasize both global values and a collective sense of id
entity through his/her behaviors to
bring about the salience of these factors in the MCT, as further explained in the following sections.


b. Global Leadership Behaviors and Their Contribution to Team Identity

The MCT leader faces many challenges which do not exist in local teams. Cultural
differences and geographical dispersion can lead to cultural misunderstanding and sub
-
group
faultlines, which may impede the creation of a cohesive and functioning MCT (Brett
, Behfar
& Kern, 2006; Earley & Mosakowsky, 2000; Malhorta et al., 2007). Certain leadership roles
may be particularly important for MCT settings. Given the "altered" social context, leaders
must be able to build and maintain a social climate necessary for

ensuring an adequate level
of team unity and cohesiveness (Kayworth & Leidner ,2002). Therefore, the leader plays a
central role
in the team identity needed for MCT effectiveness.



(1)

MCT Leader's Focus on a Collective Global Sense of Identity

Followers' pe
rsonal identification with the leader can create dependency on the leader, rather than
on the organization, and may therefore impede organizational and team goals in the absence of the
particular leader (Howell & Avolio, 1992, Conger & Kanungo, 1998; 1987)
.

10


On the other hand, according to the social identity theory, w
hen individuals develop a social
identification with a group (namely, have the perception of oneness with a group of
individuals (Ashford & Mael, 1989)), they base their self
-
concept and self
-
esteem partly on
their sense of belongingness to the group. Therefore, group success and failure are
experienced as personal successes and failures and a high level of identification leads to a
high level of effectiveness (
Ashforth & Mael, 1989
; Mael & As
hforth, 1992;

Tajfel, 1978;
Tajfel & Turner 1979
). The emphasis put by the leader on the creation of social
-
collective
identification, rather than on interpersonal identification gives priority to the considerations of
the organizational unit over the lead
er's self interest (Howell 1988; Howel & Avolio, 1992).
Indeed, in the long run, collective identification leads to success, empowerment and
interdependence of followers and organizational units (Collins, 2001).
By focusing on
collective aspects, the lead
er activates the followers' collective identification (Lord et al, 1999,
Shamir et al., 1993) and increases the members' concern for their behavioral contributions to the
collective interest (Wit & Kerr ,2002).

Empirical studies supported the theory, demo
nstrating a positive relation between the leader's
emphasis on the unit's collective identity and the followers' level of shared values and
identification with their unit (Shamir, Zakay, Breinin & Popper, 1998, 2000). Additionally, team
-
focused TL was rela
ted to high team identification and team effectiveness (Wu, Tsui and Kinicki,
2010). Therefore, the leader has the ability to influence desirable team processes and outcomes
(Fiol & O'Connor, 2005). A body of research suggests that collective identificatio
n is related to
team identity, as team identity reflects a high level of team identification among members (e.g.
Shapiro, Furst, Spreitzer & Van Glinow, 2002
;
Van der Vegt & Bunderson ,2005
; Wu et al,
2010).
Team Identity can be defined as a "common percep
tion of group cohesiveness" and as
a "common sense of entitativity" (
Earley & Mozakowski, 2000, p.35)
. Team identity becomes
salient when team members recognize that their membership in the team is more self
-
defining
than other self
-
characteristics in rega
rd to the team (Ashford & Mael, 1989; Ellemers, Gilder
& Haslam, 2004). Hence, when a salient team identity exists, individuals are motivated to
engage in behaviors that ensure the welfare of their team (Brickson, 2000).
Therefore, by
emphasizing
the impor
tance of having a collective global sense of identity and cooperation as
a cohesive unit in their behaviors

and actions,

can the MCT leaders enhance the saliency of the
collective level in the followers' self and attenuate the

salience of their national di
versity. This
will lead to facilitation of team identity in the MCT. These collective global identity behaviors will
emphasize subjects, such as the added value that of the MCT compared to a culturally
homogeneous local team (e.g. the ability to produce sy
nergetic knowledge which stems from
valued members from diverse countries), the mutual goal of the MCT and its global implications
and benefits for team members who are part of a global team (e.g. being connected to the global
vision of the MNO; the value
of working with people from diverse countries).

11


Furthermore, MCT members may each perceive their leader as very different from their respective
prototypical leader due to their diverse national culture (Kark & Shamir, 2002). This situation may
impede t
he leader's effectiveness (Lord & Maher, 1991). Therefore,
followers' perceptions of
their leaders as sharing the same collective identity has important consequences for leader

effectiveness (Ellmers et al. ,2004).

(2)
MCT Leader's Focus on Global Work Val
ues

An interesting question is how MCT members who come from different cultures can accept
and assimilate common values in the MCT.

Erez and Gati (2004) suggested a new, higher layer of culture, located beyond the national
level. This layer of culture
-

the global work culture
-

can be defined as "the shared
understanding of visible rules, regulations and behavior and the deeper values

and ethics of a
global work context" (Shokef & Erez, 2006). The researchers assert that on this cultural layer,
as on any other layer of culture, global work values facilitate the adaptation of companies to
global demands and help to maintain existence a
nd prosperity (Erez & Shokef, 2008).

Work values can serve as "general constrains of the generation of work related goals and
behaviors" (Lord & Brown, 2001, p.138). Shokef and Erez (2006) argue that the basis of the
team culture in MCTs relies on the g
lobal work values, which are the scaffolds of the shared
understanding system in MCTs.

They generated a typology of global work values that are
functional in the adaptation of employees to the global work context. This typology includes
strategic and task
-

related values, such as competitive performance orientation, quality,
customer orientation and innovation, and relational values, such as openness to diversity and
interdependence. A study conducted in four subsidiaries of a large global organization gave

support to this typology (Erez & Shokef, 2008; Shokef, 2006).

Markus and Kitayama (1991, 1994) suggested that cultural values can influence the construct
of the self. These scholars focused on national level values, such as individualism vs.
collectivism,

and found a significant impact on the individual's perspective regarding relations
with significant others, which has behavioral implications in everyday life and in the work
context (see also Kitayama, Markus, Matsumoto & Norasakkunkit, 1997).

Leaders c
an change the salience hierarchy of values and identities within the followers' self
concept, to achieve desirable behaviors (Shamir et al.,1993) and global leaders should have the
capacity to install values to inspire others (Kets de Vries & Mead ,1992)
.

Hence, we argue that global work values can influence perceptions and behaviors of MCT
followers, regarding factors related to success in the global context. Related to this subject, an
important question arises: what are the global work values which can

be used as milestones
by the MCT leader to influence the followers to establish team identity in the MCT? We
suggest that the relational global values of interdependence and openness to diversity enable
the MCT leader to gain this effect.

12



(a)
Interdepe
ndence

Interdependence can be defined as "the extent to which team members cooperate and work
interactively to complete tasks" (Stewart & Barrick, 2000, p.137)
.Hence, a high level of
interdependence means that team members depend on each other to accompli
sh their task
(Kiggundu, 1981; Van Der Vegt, Emans & Van De Vliert, 1998).

Interdependence increases the amount of interaction among team members. When the level of
interdependence is high, team members typically communicate more often, support and
influe
nce each other (Somech, Desivilya & Lidogoster, 2009). The intensive interaction
among team members has shown to build team identity (Barrick, et al., 2007) and this pattern
was also found in diverse teams (see review at Van Knippenberg & Schippers, 2007).


On the global work culture level, interdependence is an important global work value which
contributes to the success of MNOs, as it supports the unity of these companies
(Friedman,
2005; Govindarajan & Gupta, 2001).

Interdependence

facilitates collaboration, coordination
and communication across subsidiaries and cultures (Berson et al., 2004). Indeed, several
MNOs adopted the value of "one company", emphasizing the need to operate as one
organization (Erez & Shokef, 2008; Shokef, 200
6).

In MCTs, members face difficulties characterized by national cultural differences and by
geographical dispersion (Earley & Gibson, 2002). This unique context may cause faultlines,
which may impede cooperation and cohesiveness, thus reducing team perfo
rmance (Chatman
& Flynn, 2001; Polzer, 2004). Therefore, to avoid such negative outcomes, the MCT leader
must actively emphasize the importance of interdependence among team members through
words and actions. Installing a sense of interdependence in the MC
T by the leader will lead to
a cooperative frame of action. This means that challenges will be perceived by followers as
common goals and will facilitate the creation of team identity (Earley & Gardner, 2005;
Gibson & Grubb, 2005).



(b) Openness to Divers
ity

Diversity
refers to differences between individuals on any attribute that may lead to the
perception that another person is different from the self (Riordan & McFarlane
-

Shore, 1997;
Van Knippenberg, De Dreu & Homan, 2004).

Openness to diversity can be defined as "the degree of receptivity to perceived dissimilarity"
(Hartel, 2004 p.190). As a value, it
refers to a tolerance of difference
and respect to diverse
others (Hartel ,2004,). As such, diversity

can lead to higher per
formance only when members
are able to understand each other, combine and build on each others' ideas

(Maznevski , 1994).
Hence, in MCTs, acceptance of openness to diversity as a value means that team members are
willing to take the necessary actions to r
educe possible negative effects which stem from
cultural misunderstandings (Fujimoto, Hartel, Hartel & Baker ,2000).

13


As a support, a positive contribution of openness to diversity to group decision effectiveness
and interactions was found (Fujimoto, Hart
el & Hartel, 2004). Additionally, managing
diversity in the MCT was perceived by MCT leaders and followers as a key factor in the
MCT success (Joshi & Lazarova ,2005) .

Therefore, we assert that the MCT leader stresses the value of openness to diversity b
y
serving as a role model and by behavioral actions which enhance the saliency of this value in
the MCT. When MCT members accept openness to diversity as a value, they will be more
open to see other team members' perspectives ( Fujimoto et al., 2004) and
this, in turn, can
facilitate cross
-
understanding among team members (Huber & Lewis, 2010) and a shared
meaning, both leading to the creation of team identity.


c.

Global Leadership Behaviors

A sense of collective identity, interdependence and openness
to diversity can all lead to
positive performance in diverse teams (Roberge & Van Dick, 2010). In such teams,
interdependence is positively related to openness to diversity (Bacharach, Bamberger &
Vashdi, 2005; Hobman et al., 2004) and to the sense of coll
ective identity (Hobman & Bordia,
2006). Additionally, interdependence is positively related to collective identity (Brewer &
Gardner, 1996; Campion, Papper & Medsker, 1996; Cremer & Van Vugt, 1998). These
connections can be explained by the social identit
y theory, which suggests that
interdependence is an important but insufficient condition for the creation of team identity,
and there is still a need to accept other team members and develop a collective sense of
identity to facilitate this process (Lembk
e & Wilson, 1998). Therefore, we assert that when
assessing global leadership behaviors, there is a need to assess the collected impact of
interdependence, openness to diversity and collective global identity to explain the creation of
team identity in MCT
s.



2.

MCT Followership as a Moderator


During the last two decades, a few theories have considered the relations between leaders and
followers (
Graen & Uhl
-
Bien, 1995;
Hollander, 1992; Howell & Shamir, 2005, Klein & House,
1995; Shamir et al., 1993;
Wilson et al., 2010). These theories explained how

followers,
through their relations with the leader, can empower their leader, influence the leader's self
-
concept and behaviors and assist in achieving the organizational goals. These theories received
emp
irical support from studies which consider the followers' characteristics as predictors of
leadership behaviors and outcomes (e.g. Dvir & Shamir, 2003; Ehrhart & Klein, 2001; Pastor,
Mayo & Shamir, 2007;
Schyns

& Sanders, 2007).

Despite their considerab
le contribution to understanding the followership phenomenon, many of
these theories and empirical studies emphasized individual differences between followers, which
14


predict different relational patterns with their leader (e.g. Howell & Shamir, 2005). Cons
equently,
most of these studies didn’t consider team level processes or team level outcomes. Specifically,
these studies didn't define how followers' characteristics moderate the relation between leadership
behaviors and team processes and outcomes.

As mo
st cross
-
culture and global leadership studies apply a cultural lens to extend leadership
theories (Dickson et al.,

2003), it is not surprising that there is almost no research on the role of
followers in the leadership effect on the success of the MCT (se
e review at Osland et al., 2009).
However, in MCTs, which are complex global contexts, it is vital to understand the contribution
of followers to their leader's and team activities (Graen, 2006).

Followers in MCT may interpret or
assist the leader’s behavi
ors in different ways, limiting or enhancing the leader's success in
creating desirable team processes. Therefore, one of the purposes of this study is to suggest how
MCT followers can moderate the relation between global leadership behaviors and team iden
tity
in MCTs.

An explanation for these relations can be based on one of the main concepts in organizational
behavior, which is the Person
-
Environment fit (P
-
E fit).
P
-
E fit can be defined as “the
compatibility between an individual and a particular work en
vironment that occurs when their
characteristics are well matched” (Kristof
-
Brown, Zimmerman, & Johnson, 2005, p. 281)
.
According to this concept, people

develop perceptions of fit over time, due to their level of
adaptation to their environment and these
perceptions drive individual behavior and choices
(DeRue

& Morgenson, 2007).

The construct of Person
-

Team fit (P
-
T fit) stemmed from P
-
E fit and refers to
the compatibility
between individual team members and their teams (Kristof, 1996)
. Studies of P
-
T fi
t found
positive relation between team outcomes of satisfaction and commitment and the congruence of
team members' characteristics and team values
(Adkins, Ravlin, & Meglino, 1996; Barsade,
Ward, Turner, & Sonnenfeld, 2000).

Based on the above, we assert
that global characteristics shared by MCT followers may lead
to desirable team outcomes. Openness to cultural diversity is one such global characteristic.
The cultural diversity of MNOs is embedded in their structure, as they consist of multicultural
subs
idiaries (Shokef & Erez, 2006). Therefore, MNOs need to effectively manage a
heterogeneous workforce both within and across organizational boundaries (Erez & Shokef,
2008; Hartel, 2004).

MCTs embedded in MNOs represent the micro
-
cosmos of the culturally d
iverse and often
geographically dispersed global work environment (Appelbaum, Shapiro & Elbaz, 1998; Iles
& Hayers, 1997). MCT members bring to their team different perspectives regarding leader
-
member relations, members' attitudes and work regulations, wh
ich stem from their diverse
national backgrounds (
Cascio & Shurygalio, 2003
). Studies concerned with the
similarity/attraction paradigm suggest that most individuals prefer to work with similar, rather
15


than dissimilar others. Moreover, dissimilarities amo
ng team members may raise adverse
social categorization processes that impair team functioning (See review at Williams &
O’Reilly, 1998). Therefore, accepting cultural diversity becomes important to team success
(Hobman et al., 2004). Additionally, when op
enness to cultural diversity appeared as a
valuable resource for the organization, members felt more valued and respected and reported
to have high quality team work relations (Ely & Thomas ,2001). Hence,
Openness to cultural
diversity may explain positive

MCT outcomes (Cochavi, 2006).
Individuals who are open to
cultural diversity are motivated to actively seek new cultural experiences, curious regarding
other national cultures and non
-
judgmental regarding other cultural behaviors and
expectations (Hartel
& Fujimoto, 2000; Shokef & Erez, 2006). Furthermore, these individuals
have capabilities to function effectively in diverse cultural settings and contexts (Ang, Van
-
Dyne & Koh, 2006), and are willing to accept new cultural values. Therefore, when followers

in MCTs are open to cultural diversity, they will actively seek to interact with other team
members and will regard these cultural interactions as interesting and challenging, rather than
threatening. Such followers will show, from the beginning, more res
pect to and trust in other
team members despite national cultural differences and they will actively and voluntarily
create positive relations with diverse others, thus increasing team identity in their MCT
(Shokef and Erez (2006).

Followers' willingness

to contribute to their team has a positive effect on the relation between
leaders' action and team outcomes (Yun, cox & seems, 2006). A combination between high
level of both leaders' and followers' role behaviors may enhance the level of positive team
co
hesiveness and potency (DeRue & Ashford, 2010; Taggar & Seijts, 2003). When followers
possess high openness to cultural diversity, they will be initially open to accept their leader's
emphasis on openness to diversity, interdependence and collective sense
of global identity
and will be motivated to develop a global collective identity with their multicultural team. On
the other hand, followers who have low levels of openness to cultural diversity will find their
leader's global behaviors potentially threat
ening and may avoid building an MCT identity.


Therefore, we
hypothesize
:

H1:

Followers' openness to cultural diversity will moderate the relationship between global
leadership behaviors and team identity. This relation is more strongly positive when
followers' openness to cultural diversity is high, rather than low.


3
. MCT Iden
tity and MCT Effectiveness


Leadership behaviors are especially relevant in enhancing positive team processes and
outcomes, which lead, in turn, to team effectiveness (Yukl, 2006). Hence, understanding the
relations between leadership behaviors and team e
ffectiveness must consider the influence of
16


global leadership on team processes and team outcomes (Dionne, Yammarino, Atwater,
Spangler, 2004). Therefore, the effectiveness of MCT leaders depends on their ability to
enhance team identity.
Team identity
i
s positively related to Organizational Citizenship
Behaviors (OCB) (Chattopadhyay, 1999; Janssen & Huang, 2008; Van der Vegt , Van der
Vliert & Oosterhof, 2003), and the ability to resolve misunderstandings (Hinds and Weisband,
2003). High level of team i
dentity can reduce "free
-
ride" behaviors (Eckel & Groosman, 2005;
Shapiro et al., 2002) and enhance participants' involvement in team activities (Witt & Kerr,
2002). As such, team identity is positively related to team performance and team effectiveness
(S
cott, 1997; Van der Vegt, Van der Vliert & Oosterhof, 2003) and this relationship was
found in MCTs (
Earley & Mozakowski ,2000;
Vallaster , 2005
).


Therefore, we suggest that high
team identity in the MCT can be very valuable for MCT
effectiveness. When t
eam identity is high, team members view team goals and activities as an
important part of their selves and actions (Ashford & Mael, 1989; Haslam, Powell & Turner,
2000) and
will,
therefore,

contribute to accomplish them (Van Der Zee, Atsma & Brodbeck,
2004
).


Therefore, we
hypothesize
:

H2:

Team identity will be positively related to MCT effectiveness.


4
. Global Leadership Behaviors and MCT Effectiveness


Previous studies found an indirect effect of leadership behaviors on team and individual
outcomes (e.g. Jung & Avolio, 2000;
Podsakoff, et al., 1990).
Our rationale is that due to the
global complexity of the MCT environment, global leadership behaviors wi
ll influence the
emergence of team identity in the MCT, and that team identity will enable team effectiveness.

Indeed, recent empirical finding have suggested that the relation between leadership and team
effectiveness in MCTs is intervened by team identit
y (Kearney & Gebert, 2009). As
followers' willingness to contribute to their team has a positive effect on the relation between
leaders' actions and team outcomes (Yun et al., 2006),
it will be easier for the leader to
enhance team identity when the follow
ers are adaptable to the MCT, given their high
openness to cultural diversity.

In such a case, followers will indirectly contribute to the
creation of team effectiveness through their positive interaction with their global leaders'
behaviors.

Therefore, w
e
hypothesize
:

H3: Global leadership behaviors will be related to team effectiveness through team
identity.


This relation will be moderated by followers' openness to cultural diversity, with a
stronger positive effect of the leader when followers' opennes
s to cultural diversity is high,
rather than low.

17



In our research we examined MCT effectiveness models in two studies. Study 1 was
conducted on 73 MCTs of MBA students. Study 2 examined a wider MCT effectiveness
model on 55 MCTs belonging to 9 MNOs. We wi
ll present these two studies and offer
insights in the discussion sections.


II) Study 1

A. Methods

1.
Participants

The sample consisted of 282 MBA and graduate students from 8 Universities in 6 countries
(USA [3], Israel, HK, India, Spain and Finland).
They belonged to 42 nationalities, thus
representing a high cultural variety. 41% of the participants were Europeans, 26% Asians, 15%
Israelis and 14% North
-

Americans
.
The average age was 27.32 years (SD= 5.66), and 64
percent were men
.
All participants
were students of a cross
-
cultural management course in
their respective universities and they participated in this study as part of their
multi
-
cultural
team project.

2.

Procedure

The study was conducted in English (since it serves as the academic and
business
international language). Therefore, a high level of English proficiency was the preliminary
requirement for participation in this study. Average self
-
report of English mastery was 4.55
(SD=0.70, 1
-
5 scale).

Participants were divided into 73 virtua
l MCTs of four (86 percent) or three members each.
The allocation of members to MCTs was random, with a few constraints: (a). each MCT
member was from a different nationality and held a different native tongue (e.g. there were
no Spanish and Colombian memb
ers in the same MCT). (b). in 61 MCTs (84 percent), all
members were from different universities. In the remaining 12 MCTs (composed of 4
members each), there were no more than two students from the same university. These two
students belonged to different

classes and kept no physical contact during the study. A t
-

test
didn't reveal any differences on the core variables between these groups; therefore, we
gathered all teams to one sample.

Figure 2 describes the time line of study 1, which continued 28 day
s (4 weeks) with two main
phases:

Phase 1:

The first 10 days of the project were the "getting to know each other" phase. MCT
members interviewed each other and got involved in discussions that expanded their
acquaintance with each other. At the end of this

phase each MCT was asked to "
nominate the
most suitable team member" as

a team leader.

Phase 2:

At this stage, the team task assignment began and lasted for18 additional days. The
18


goal of each MCT was to develop guidelines for an expatriate who was goin
g to be posted in
a country selected by the team. The MCT leaders led their teams during this assignment,
following instructions from the study coordinator.

All participants filled out three web
-
based questionnaires, at the following points in time:
befor
e the beginning of Phase 1 ("Time 1"), at the beginning of phase 2 ("Time 2") and at the
end of the project ("Time 3").


Figure 2

Time Line of Study 1









3.

Measures

Global Leadership Behavior Scale (GLBS)

In the absence of an existing measure to empirically test global leadership behaviors, we
developed a short scale; which is the Global Leadership Behaviors Scale (GLBS) (see GLBS
development procedure in Appendix 1)
.

In the end of the assignment phase ("time 3"), all 209 MCT followers filled out the GLBS,
referring to their MCT leader. The GLBS included 12 items , with
a 5 point Likert type scale
(1
-

not at all, 5
-

frequently, if not often)

consisting of three subscales: Collective Sense of
Global Identity (CGI
-

e.g. "
Emphasizes the importance of having a global collective sense of
mission"
), Openness to Diversity (OTD
-

e.g. "
Serves as an example of proper behavior
towards employees from dif
ferent nationalities"
) and Interdependence (IND
-

e.g.

"Inspires
team members to think together about solutions to team tasks
") (see full measure in Appendix
2).

According to Floyd and Widaman (1995), an Exploratory Factor Analysis (EFA) with
Common Factor
Analysis as extraction method should be used to understand the relations
among a set of measured variables in terms of underlying latent variables. Common Factor
Analysis attempts to represent only the common variance of each variable. This common
variance

is shared with other observed variables, as a result of the dependence of the
measured variables on the latent variables.

Therefore, an EFA using Common Factor Analysis method with promax rotation supported
this three
-
factor model, explaining the expected

common variance proportion of 106.00
Time 1

Measures:

Openness to
cultural diversity

Time 2

Measures:

Team Identity

Time 3

Measures:

Team Identity.

Global Leadership
Behaviors.

Team Effectiveness

Day

1

Day 10

Day 28:


End of Study

Phase 1: "Getting to


know each other"


Phase 2: Assignment


19


percent. Alpha Crobnach's reliability was 0.93 for CGI items, 0.85 for OTD items and 0.89
for IND items. The Alpha Cronbach's reliability of the 12
-

item
-

scale was 0.91.

We used Confirmatory Factor Analysis (CFA) met
hod to compare the three
-
factor model with
other possible models, using Mplus Version 6 (Muthe´n& Muthe´n, 2010).

In the present study, a preference was given to fit indices that were less sensitive to sample
size, such as the Root Mean Square Error Appr
oximation (RMSEA), the Comparative Fit
Index (CFI), and the
Tucker
-
Lewis Index
(TLI; also known as NNFI
-

Non
-
Normed fit Index)
(Marsh, Balla, & McDonald, 1988; Marsh & Hau, 1996). For RMSEA, it is suggested that a
value of .08 represents reasonable errors
of approximation (Brown & Cudeck, 1993). For CFI
and TLI, values greater than .90 are usually considered satisfactory (Hoyle, 1995).

The three
-
factor model had significantly better fit than any other possible combination.
Therefore, the three subscales c
onstruct was well established.

Table 1
:

Study 1
-

CFA


Comparison between GLBS Factor Structures
.


Model

χ2

Df

CFI

TLI

RMSEA

Δχ2
-

comparison
with 3 factor model

3 factor model

131.09***

51

.95

.93

.08

-

2 factor model
(CGI/OTD+IND)

327.38***

53

.84

.77

.14

Δχ2

(2)= 196.29***

2 factor model
(IND /CGI+OTD)

477.64***

53

.76

.64

.17

Δχ2

(2)= 346.55***

2 factor model
(OTD/ CGI+IND)

421.04***

53

.79

.69

.16

Δχ2

(2)= 289.95***

1 factor model

659.31***

54

.59

.50

.25

Δχ2

(3)= 528.22***

N=209, *** p<
0.001


OTD
-

Openness to diversity , IND
-

Interdependence, , CGI
-

Collective sense of global identity.


Second order model of GLBS

GLBS is identified by three subscales. However, as suggested in our theory, global leadership
behaviors can be seen as one, unified phenomenon when different global behaviors of the
leader operate in the same direction to achieve effectiveness in the MCT.

A common statistical procedure which can estimate the connection of the three factors (CGI,
OTD, IND) to one global leadership phenomenon is the calculation of a second
-
order CFA
model (House
, et al., 2004; Rindskopf &


Rose, 1988). However, a second order model which
relies on three factors as first order, leads to a just
-
identified model (a model which has one
unique set of parameter estimates that perfectly fit the data (Brown, 2006) and therefore,
measures of fit of

the second order model, give no meaningful information, as they present
20


the same fit as the first order model (Chan & Drasgow, 2001). Hence, we used an alternative
approach based on Chen et al., (2010).

First, we tested our two hypotheses, with each of t
he subscales separately as a dependent
variable (instead of using the full GLBS as a dependent variable). Each subscale exhibited
relationships with outcomes similar to the full GLBS (See Table 3 and Table 4 in the results
section).

Secondly, we aggregat
ed the items of each subscale to one score, then calculated the overall
GLBS Crobnach's Alpha of the three aggregated scores, which demonstrated sufficient
reliability of


= .75. Third, these subscales were highly correlated

(
r

= .71,
p

< .001
,
between
CGI and IND, (
r

= .69,
p

< .001
, between OTD and IND and (
r

= .51
p

< .001
,
between CGI and OTD), suggesting high common variance between the subscales.

In addition to Chen et al.'s (2010) suggestions, a CFA of second
-
order model was conducted,
using the

three subscales as a first
-

order factor. Using the loading of each of the subscales on
the second order factor, construct reliability (an assessment of the variance in the indicators
explained by the common underlying latent construct (Gerbing & Anderson

, 1988)) was
calculated. The construct reliability of CGI, OTD and CGI, explained by the common latent
variable, was 0.83. This result reflects a high level of variance, explained by the common
underlying latent construct. Finally, variance extracted es
timates, which assess the amount of
variance captured by a construct's measure in relation to variance due to random measurement
error (Fornell & Larcker ,1981) was calculated and found to be 0.63. It is acceptable that a
level of .50 or higher supports t
he consistency among items on a scale (Fomell & Larcker
1981).

Thus, given the above results and our conceptualization of global leadership behaviors as a
multidimensional construct which contains three subscales (Collective Sense of Global
Identity, Openn
ess to Diversity and Interdependence), we measured the three subscales as one
12
-

item scale of GLBS (
91
.
0


).


2.
Openness to Cultural Diversity



this scale consists of six items,
using a 7 point Likert type
scale (from 1
-
"very inaccura
te" to 7
-
"very accurate

) (i.e., "I often spend time with people
from cultural groups other than my own"), based on Hobman Bordia and Gallois, (2003), and
Henry (1986). Cronbach's Alpha was 0.77.

All participants filled the scale before the
beginning of the study (time 1). However, only followers' responses were used in the analysis.


3.
Team Identity
-

was measured by a three
-
item scale of
Earley and Mozakowski, (2000)
("The feeling we were all s
haring a common set of beliefs and values was strong in our team";
"Our team members had a strong sense of belonging

to their team
" and "
Our team acted as a
21


single, cohesive team"). Responses were on a 5
-
point Likert scale, ranging from 1 ("Strongly
Disagr
ee") to 5 ("Strongly Agree"). Cronbach's Alpha was 0.85. Followers filled this scale
twice: in the middle of the study (Time 2
-

T2) and at the end of the study (Time 3
-
T3).


Aggregation of GLBS, Openness to Cultural Diversity and Team Identity to the T
eam Level

Following Bliese's (2000) recommendation, a within
-
group coefficient of agreement of Rwg(j)
was used
(James, 1982; James, Demaree, & Wolf, 1984)
. Additionally, i
ntraclass correlations
of ICC(l), and ICC(2), which are statistics commonly used to j
ustify aggregation of data to
higher levels of analysis, were used (e.g., Bartko, 1976; Shrout & Fleiss, 1979). ICC (1)
compares the variance between units of analysis (e.g. MCTs) to the variance within units of
analysis using the individual ratings of ea
ch respondent. ICC(2) assesses the relative status of
between
-
and within
-
variability, using the average ratings of respondents within each unit
(Bartko, 1976).

As a preliminary step, ANOVA ("F
-
test") was used to contrast within
-
group
variance with betwee
n
-
group variance (Bliese, 2000).

For GLBS

the results were: mean Rwg(j)= 0.86, ICC(1)= .29 (F= 2.19, p<.001), ICC
(2)= .54. For
Team Identity

the results were: mean Rwg(j)= 0.82, ICC (1)= .27 (F= 2.08,
p<.001), ICC(2)=.52.
All these were comparable to t
he median or recommended ICC values
reported in the literature (James, 1982; Schneider, White and Paul, 1998). We thus concluded
aggregation was justified for these variables.

For
Openness to Cultural Diversity,

high within
-
group agreement was received (m
ean
Rwg(j)= 0.90). Additionally, ICC (1) = 0.12 (F=1.38, P<0.055), ICC (2)= .27, suggesting
lower ICC(2) level than expected. However, similar levels of ICC (2) were reported in recent
studies (e.g. Bacharach & Bamberger, 2007; Liao & Rupp, 2005;Wu, et a
l., 2010) given
sufficient between
-
group differences (significance or approach to significance F
-
test, see
Ilies ,Wagner & Morgeson, 2007), high within
-
group consensus ( high Rwg(j)), and small
sample size (Bliese, 2000; Liao & Chuang, 2007; Kirkman et al
., 2009 ). Since the Openness
to Diversity measure was defined by all these conditions, aggregation was concluded.



Independence
-

Last, following Podsakoff et al. (2003)
's

suggestions regarding same
-
source
bias avoidance, we compared different CFA models

to assure that these scales were
independent of each other. The five
-
factor model (the three subscales of GLBS, team
-
identity
scale and openness
-
to
-

cultural
-
diversity scale), was the only model which yielded acceptable
fit measures (χ2[179]= 364.45,
p
<
.001; RMSEA=.07; TLI= .91; CFI= .93) and had significant
large chi
-
square difference in comparison to other factor models, thus suggesting low
probability for same
-
source bias.

Team Effectiveness

-

was measured by MCT leaders' report at the end of the st
udy (Time 3).
This scale consisted of seven items based on
Tjosvold, Poon and Yu
's (2005) team
effectiveness scale. The scale consists of two components of team effectiveness, which are:
productivity (e.g. "
Team members met or exceeded their productivity r
equirements"
and
22


commitment (
e.g. "Team members felt highly committed to the goals of their work".)
Items
were on a 5 point Likert scale, ranging from 1 ("Strongly Disagree") to 5 ("Strongly Agree").
Exploratory Factor Analysis using Component Principal An
alysis extraction method revealed
that all items were loaded on one factor, explaining 70.82% of the total variance. The
Cronbach's Alpha was 0.85.

Control Variables


Team Identity (time 2)
: team identity was measured at the beginning of the team assignme
nt
phase (Time 2), before the leader's nomination, and served as a control variable. Cronbach's
Alpha was 0.76.

Participants were randomly assigned to teams; hence, we were not expected any effects of age
diversity or gender composition in teams. Still, we

controlled for age (using Blau’s [1977]
index) and gender (proportion of women in each team). Neither of these variables had main
effects (
.
.
,
15
.
0
s
n




for age and
.
.
,
09
.
0
s
n



for gender) on team effectiveness and on
team identity (
.
.
,
09
.
s
n




for age and
.
.
,
04
.
s
n




for gender). Additionally, their
inclusion didn't change the results concerning our hypotheses. Therefore, we chose not to
incorporate these variables in the analyses reported below (see Homan et al., 2008).


B. Results

1.
Descriptive Statistics

Table

2 summarizes means, standard deviations and correlations at the team level. Team
identity (Time 3) correlated with global leadership behaviors (r=.61, p<.001) and with
followers' openness to cultural diversity (r= 0.27, p<0.05). Additionally, global lead
ership
behaviors significantly correlated with team effectiveness (r= .33, P<0.01). Finally, there was
a positive and significant correlation between team identity (time 3) and team effectiveness
(r=.38, p<0.01).









Table 2
:

Study 1
-

Means, Standard
Deviations and Correlations


Variables

M

SD

1

2

3

4

1. Team Identity (Time 2)

3.48

0.44

-




2. Global Leadership

3.38

0.57

.34**

-



23



Behaviors

3. Followers' Openness to Cultural
Diversity

5.19

0.53

.21†

.12

-


4. Team Identity (Time 3)

3.79

0.58

.55***

.61***

.27*

-

5. Team Effectiveness

3.93

0.72

.19

.33**

-
.05

.38**

N= 73 MCTs . † p< 0.1 *p< .05 **p<.01 ***p<.001


2.
Testing the Research Hypotheses

To examine the research hypotheses, a path analysis model of simultaneous linear
regressions
was conducted.

Our first hypothesis predicted that followers' openness to cultural diversity will moderate the
relationship between global leadership behaviors and team identity, and that this relation will
be more strongly positive when follow
ers' openness to cultural diversity is high, rather than
low. Therefore, we
conducted three steps of linear regression
with predictor variables for
team identity. We entered team identity (Time 2) as a control variable in the first step. For
main effects,
we entered global leadership behaviors and followers' openness to cultural
diversity as the second step and the interaction between them as the third step. Both global
leadership behaviors and followers' openness to cultural diversity were centered (the me
an
subtracted from each variable, leaving deviation scores) to reduce multicollinearity between
these variables and their interaction (Aiken & West, 1991; Preacher & Rucker, 2003).

As expected, a positive and significant interaction effect was found betw
een global leadership
behaviors and followers' openness to cultural diversity on team identity (T3) (



0.17,
p<0.05, Step 3). This model explained 56% of the variance in team identity T3 (F(4,68)=
21.22, P<.001, See Table 3, Step 3), a
nd significantly contributed to the explained variance of
team identity, compared to main effects only model (
Δ

R²= .03, P<.05).

To examine the nature of this interaction, we conducted a simple slope analysis (Aiken &
West, 1991), which presented the regr
essions slopes of team identity on global leadership
behaviors in three levels of followers' openness to diversity (a. High
-

one standard deviation
above the mean, b. at the mean level, and c. Low
-

one standard deviation below the mean).
The results, i
llustrated in Figure 3,

revealed that the higher the level of followers' openness to
cultural diversity, the higher the positive relation between global leadership behaviors and
team identity.

When followers' openness to cultural diversity was high (


0.75, t(68) = 4.72, p<.001), or
at the mean level (


0.46, t(68) = 5.23, p<.001), the global leadership behaviors' scale was
significantly and positively related to team identity. However, when the level of followers'
o
penness to cultural diversity was low, the relations between global leadership behaviors'
24


scale and team identity were not significant (


0.17, t(68) = 0.96, ns). These results
supported Hypothesis 1.


Table 3
:

Study 1
-

Regression
Analysis

of Team Identity on Model Variables

(H1)



DV: Team Identity (time 3)

Variables

Step 1

Step 2

Step 3

Control (step 1)




Team Identity (time 2)

.55***

.36***

.35***

Step 2: Main Effects




Global leadership behaviors (GLBQ)


.48***

.45***

Followers openness to cultural diversity
(OTCD)


.14†

.16*

Step 3: Interactions




GLBQ X OTCD



.17*



.30

.53

.56

Δ



.30***

.23***

.03*

F

30.97***

25.78***

21.22***

N= 73 MCTs. Standardized regression coefficients are reported.

† P<
.10, *P<.05, ** P<.01, *** P<.001


The second hypothesis was that team identity will positively contribute to team effectiveness.
To examine this hypothesis, a simultaneous linear regression was conducted. We regressed
team effectiveness on team identity (
T3), in the presence of all model variables (see Table 4).
The results yielded that this model explained 18% of the variance in team effectiveness
(F(5,67)= 2.99, P<.05). Team identity (Time 3) was the only variable that significantly
predicted team effect
iveness (


0.34, p<.05). These results supported hypothesis 2.

In line with hypotheses 1 and 2, Hypothesis 3 was examined by the conditional indirect effect
of global leadership behaviors on team effectiveness, through team identity, at different levels
of followers' openness to cultural diversity.







25


Figure 3
:
S
tudy 1
-

Followers' Openness to Cultural Diversity as a Moderator of the Relation


between Global Leadership Behaviors and Team Identity



Table 4
:
Study 1
-

Results of Simultaneous Regression of Team Effectiveness on


Model Variables (Hypothesis 2)

Variables



F

df

2
R

DV
:
Team Effectiveness


Team Identity (Time 2)

Global leadership behaviors
(GLB)

Followers' openness to cultural diversity (OTCD)

GLB X OTCD

Team Identity (Time 3)


-
.01

.15

-
.15

.01

.34*


2.99*

5, 67

.18

N=73 MCTs. Standardized regression coefficients are reported.

* p<0.05, ** p<0.01, *** p< 0.001

A conditional indirect
effect can be described as "the magnitude of an indirect effect at a
particular value of a moderator" (Preacher, Rucker & Hayes, 2007, p.186). In this study, the
conditional indirect effect was described by the product (Path a X Path b) of the path from
gl
obal leadership behaviors to team identity (Path a) and the path from team identity to team
effectiveness (path b), at various levels of followers' openness to cultural diversity (Bauer,
Preacher & Gil, 2006; Hayes, 2009) .

To test the significance of this conditional indirect effect, we calculated 95 percent confidence
intervals derived from bias
-
corrected bootstrap estimates (MacKinnon, Lockwood, &
Williams, 2004). Bootstrap is a nonparametric approach to effect
-
size estima
tion and
3.2
3.4
3.6
3.8
4
4.2
4.4
4.6
-1.5
0
1.5
Global Leadership Behaviors (S.D.)
Team Identiy
(Scale 1-5)
-1
0
1
Followers' Openness to

Cultural Diversity (S.D)

26


hypothesis testing that makes no assumptions about the shape of the distributions of the
variables. This approach can circumvent power problems introduced by asymmetries and
other forms of non
-
normality in the sampling distribution of an indirect
effect. It can also be
effectively utilized with smaller sample sizes than methods which assume normality

(as in
this current study)(

Schneider, et al., 2005;
Shrout & Bolger, 2002).

Our results, demonstrated in Table 5,

yielded that the indirect effect
of global leadership
behaviors on team effectiveness through team identity was amplified as followers' openness
to cultural diversity level increased. More specifically, under low level of followers' openness
to cultural diversity (one standard deviation b
elow the mean level), the indirect effect was
non
-
significant (
07
.

B
, Upper limit 95% (UL) = 0.33, Lower Limit 95% (LL)=
-
0.04).
However, under a moderate level of followers' openness to cultural diversity (at the mean),
the indirect effec
t was positive and significant (
20
.

B
, UL = 0.44, LL = 0. 04), and was
amplified under the higher level of followers' openness to cultural diversity (one standard
deviation above the mean level) (
32
.

B
, UL = 0.75, LL = 0.
06).

Hence, these results
supported hypothesis 3.


Table 5
:
Study 1
-

Conditional Indirect Effect of Global Leadership Behaviors on Team


Effectiveness, Through Team Identity, in Different Levels of Followers'


Openness to Cultural Diversity

Level of followers'
openness to cultural
diversity

Indirect effect
estimate

(B)


Lower Limit CI

(.95)


Upper Limit CI

(.95)


High (+ 1 S.D)

.32

.06

.75

Mean

.20

.04

.44

Low (
-

1 S.D.)

.07

(
-
.04)

.33

N= 73 MCTs


C. Discussion

Study 1 proposed a research model for MCT effectiveness. The results supported the research
hypotheses.

Hypothesis 1 predicted that global leadership behaviors will be positively related to team
identity. Study results
demonstrated the relation of global leaders' behaviors to team identity.
Demonstrating behaviors which emphasize the collective identity (
Brewer and Gardner’s,
1996)
, such as interdependence, openness to diversity and a collective sense of global identity
by the leader, led to a higher team identity level. These results contribute to the existing
literature in two ways; first, they give support to self
-

concept based theories of leadership
(Lord et al., 1999; Shamir et al., 1993) in the global context. Thes
e theories propose that the
27


ability of the leader to influence team processes and outcomes depends on leadership
behaviors which enhance the collective level of followers' identity. Second, this study
contributes to the global leadership field, by being t
he first to present global leadership
behaviors which stem from typologies of global work values (Shokef & Erez, 2006) and are
connected to the global context.

Moreover, as predicted in hypothesis 1, the relation between global leadership behaviors and
team identity was positively moderated by the level of MCT followers' openness to cultural
diversity. The higher the level of followers' openness to cultural diversity, the stronger the
positive relation between global leadership behaviors and team identit
y. When followers'
openness to cultural diversity was low, the relation between global leadership behaviors and
team identity was not significant. This moderated effect of the followers' openness to cultural
diversity demonstrated the importance of taking

into consideration the followers' role in teams
in general (Baker & Gerlowski; 2007) and in MCTs in particular. Followers' adaptation to the
global context by demonstrating high levels of openness to cultural diversity facilitated the
emergence of team id
entity. This finding suggests that in MCTs, the fit of the followers'
global characteristics to the global context facilitates the global leader's effect on team
identity, which further influences team effectiveness. As such, this study supports the Person
-
Team fit model (Kristof, 1996).

Additionally, the results supported hypothesis 2, which predicted that team identity will lead
to MCT effectiveness. The results provide additional support to the crucial role of team
identity in achieving MCT effectivenes
s (Gelfand et al., 2007; Van Der Zee et al., 2004;).
Additionally, these findings support the social identity theory (
Ashford & Mael, 1989
)
regards team identity as necessary for team effectiveness. Team identity reflects the
willingness of team members to

"strongly subscribe to a common set of values or beliefs"
(p.22), and therefore, act toward team goal accomplishment.

Finally, the results which supported hypotheses 3, highlight the importance of understanding
the contribution of intervening variables,
such as team identity, to the relation between
leadership behaviors and team effectiveness (Shamir et al., 1993). Additionally, the
moderated effect of followers' openness to cultural diversity on this relation indicates that the
followers have an indirect

ability to influence team effectiveness. As such, followers'
influence must be taken into consideration in addition to their direct contribution to team
processes.

To summarize, the results of Study 1 highlighted the important role played by both leaders
and followers in MCTs. The results demonstrated that global leadership behaviors indirectly
influenced team effectiveness through team identity and this indirect effect is positively
conditioned by followers' openness to cultural diversity.



28


Limitations
:

Despite these valuable results, there are limitations to this study. Participants in
Study 1 were MBA students who worked in short


term virtual MCTs. Therefore, additional
studies should test for the research model in stable, long term MCTs. Study 2, w
hich was
conducted in ongoing industry MCTs, was designed to overcome this limitation and therefore
expanded both the research model and its external validity.


III) Study 2

A. Introduction, Research Literature and Research Hypotheses

Study 2 aims to
test the research model presented in Study 1 in a real work context (MNOs)
of multicultural team members, who work together on ongoing projects and whose level of
virtuality may vary on a continuum between collected teams and pure virtual teams. (Le
-
Nobel,

2010; Stanko & Gibson, 2009; Webster

& Wong, 2008
). Short
-
term project teams,
such as in Study 1, differ from ongoing teams with regard to relational team processes. In
ongoing teams,
members anticipate future interactions with each other beyond the immi
nent
deadline and outcome
. As such, ongoing team members are led by the "shadow of the future"
(Axelrod, 1984), which is the anticipation of future interactions that have an impact on
present relational team processes.
This is a key element in ongoing team
s that does not exist
in temporary teams. Anticipation of working again with team members is likely to alter the
behaviors in such a way that it may encourage relationship processes that contribute to team
identity (Bouas & Arrow, 1996).
In contrast,
tempo
rary, short
-
term action teams are mainly
concerned with effectively accomplishing the goal of the current task, and therefore may
differ in quantity and quality of relational processes in the teams (Saunders & Ahuja, 2006).

Study 2 aims at overcoming the
limitations of study 1 in three ways: first, the participants in
this study are employees who work for MNOs as members of MCTs. Second, the MCTs in
study 2 represent a wide range of virtual levels. Third, MCT members work on ongoing
projects for relativel
y long periods.

Ongoing team processes emphasize interpersonal relationships more than short
-
term
temporary teams. For example, building team trust seems to be important in ongoing teams in
order to build team identity. Therefore, in Study 2 we included t
eam trust in our research
model. The meaning of team trust differs between ongoing teams and short term teams
(especially, short term virtual MCTs, as in study 1) as will be further discussed (Jarvenpaa,
Shaw & Staples, 2004; Panteli & Duncan, 2004; Saunde
rs & Ahuja, 2006).

Most studies conceptualized and measured trust as an expectation or a belief that one can rely
on another person's actions and words and/or that this person has good intentions toward
himself (Dirks, ,2000).

In their review, Rousseau,
Sitkin, Burt and Camerer (1998) suggested
two necessary conditions for trust building. The first is
risk

which is the perceived probability
of loss, as interpreted by a person who conveys trust (Chiles & McMackin, 1996). The second
29


is
interdependence
, sugg
esting that the interest of an individual or a party can't be achieved
without reliance upon others who share the same interest (Wageman, 1995).

Early trust studies focused on the basic dyadic relations arising from attributes associated with
the trustful
person (Dirks, 1999). However, in the last decades, research has shifted towards
the collective level of trust, which consists of multiple members and is more complex than the
dyadic trust (Mayer, Davis & Schoorman, 1995; Rousseuu et al.,1998). Team trust
is defined
as a "shared psychological state in a team that is characterized by an acceptance of
vulnerability based on expectations of intentions or behaviors with others within the team"
(Gibson & Manuel ,2003, p.59). Perceptions of trust reside on the in
dividual level, but the
meaning of trust as a team level construct comes for a shared quality of these individual level
perceptions (De Jong & Elfring , 2010) .

Jarvenpaa and colleagues (Jarvenpaa et al., 1998; Jarvenpaa & Leidner, 1999), who studied
trust

in short term virtual teams, found that members in such teams developed models of
"swift trust". Swift trust is a form of depersonalized action that allows team members to act as
if trust were present from the start of the project, as it enables members t
o take action and deal
with uncertainty, ambiguity and vulnerability that arise while working with strangers on
complex, interdependent tasks (Meyerson, Weick & Kramer, 1996; Saunders & Ahuja, 2006).
However, this kind of team trust is focused on task comp
letion and not based on the trust
associated with embedded relationships in social networks. In other words, the process of
trust creation in short term teams may be temporary, fragile and may lack the antecedents of
trust building as in ongoing teams ( (J
arvenpaa et al., 1998; Jarvenpaa & Leidner, 1999;
McKnight, Cummings & Chervany , 1998
). Short
-
term teams reported either a negligible or
conditional effect of trust on team outcomes (e.g., Aubert & Kelsey, 2003; Dirks, 1999;
Jarvenpaa, Shaw, & Staples, 2
004), unlike ongoing teams that reported a positive effect of
trust on team outcomes (De Jong & Elfring, 2010; Rispens, Greer, & Jehn, 2007; Spreitzer
Noble, Mishara & Cooke ,1999
).

Therefore, in Study 2 we plan to include trust as a factor in
the research

model.


Team Trust as a Mediator of Team Identity

The research on trust examined its direct (e.g. Friedlander, 1970; Klimoski & Karol, 1976)
and
indirect effects on
other team processes and outcomes (Dirks (1999). For example, team
trust positively
influenced team identity ( Fiol & O'Connor ,2005). Moreover, team trust
enhanced team members' involvement in team activities and contributed to the collective
action (Spreitzer et al., 1999). O'Hara
-

Devereaux and Johansen (1994) argued that trust in
MCTs

prevents the geographical and organizational distance of global team members from
becoming a psychological distance. They viewed trust as "the glue of the global workspace"
(p.243).

30


Trust in MCTs reflects the willingness of team members to take risks desp