httphum.sagepub.comHuman Relations httphum.sagep.docx

http://hum.sagepub.com/
Human Relations
http://hum.sagepub.com/content/61/11/1593
The online version of this article can be found at:
DOI: 10.1177/0018726708096639
2008 61: 1593Human Relations
Knippenberg
Michaéla C. Schippers, Deanne N. Den Hartog, Paul L.
Koopman and Daan van
The role of transformational leadership in enhancing team
reflexivity
Published by:
http://www.sagepublications.com
On behalf of:
The Tavistock Institute
can be found at:Human RelationsAdditional services and
information for

http://hum.sagepub.com/cgi/alertsEmail Alerts:
http://hum.sagepub.com/subscriptionsSubscriptions:
http://www.sagepub.com/journalsReprints.navReprints:
http://www.sagepub.com/journalsPermissions.navPermissions:
http://hum.sagepub.com/content/61/11/1593.refs.htmlCitations:
What is This?
- Oct 10, 2008Version of Record >>
at LIBERTY UNIV LIBRARY on August 17,
2012hum.sagepub.comDownloaded from
http://hum.sagepub.com/content/61/11/1593
http://www.sagepublications.com
http://www.tavinstitute.org/index.php
http://hum.sagepub.com/cgi/alerts
http://hum.sagepub.com/subscriptions
http://www.sagepub.com/journalsReprints.nav
http://www.sagepub.com/journalsPermissions.nav
http://hum.sagepub.com/content/61/11/1593.refs.html
http://hum.sagepub.com/content/61/11/1593.full.pdf
http://online.sagepub.com/site/sphelp/vorhelp.xhtml

The role of transformational leadership
in enhancing team reflexivity
Michaéla C. Schippers, Deanne N. Den Hartog,
Paul L. Koopman and Daan van Knippenberg
A B S T R A C T Team reflexivity, or the extent to which teams
reflect upon and
modify their functioning, has been identified as a key factor in
the
effectiveness of work teams. As yet, however, little is known
about
the factors that play a role in enhancing team reflexivity, and it
is thus
important to develop theorizing around the determinants of
reflexivity. From an applied perspective, leadership is a very
relevant
factor. The current study is a first step in the development of
such
a model, and addresses this important gap in our understanding
of
team reflexivity by focusing on the role of leader behavior. We
examined the extent to which transformational leadership
influences

team reflexivity, and in turn, team performance, in a field study
conducted among 32 intact work teams from nine organizations.
Team members rated reflexivity and leadership, while external
managers rated team performance. We hypothesized and tested a
mediational model proposing that transformational leadership is
related to the adoption of a shared vision by the team. This in
turn
relates to team reflexivity, which leads to higher team
performance.
Results support this model.
K E Y W O R D S learning � performance � reflexivity �
transformational
leadership � vision
1 5 9 3
Human Relations
DOI: 10.1177/0018726708096639
Volume 61(11): 1593–1616
Copyright © 2008
The Tavistock Institute ®
SAGE Publications

Los Angeles, London,
New Delhi, Singapore
http://hum.sagepub.com
Teams have become the basic organizing structure for
accomplishing work
in many firms, especially for the increasing numbers of
organizations
operating in dynamic and complex environments (e.g.
Edmondson, 1999).
A growing number of teams in the workplace perform
intellectual and
cognitive tasks (Cooke et al., 2000; Hinsz et al., 1997; Salas et
al., 1992),
with information processing as a central aspect of their work,
making it
important to identify factors that influence effectiveness of
those teams.
Recently, reflexivity (a concept related to team learning) has
been identified
as a key factor in the effectiveness of work teams (e.g.
Schippers, 2003;
Schippers et al., 2003, 2007; West, 2000).
At the same time, scholars have noted that individuals and
teams rarely
reflect spontaneously; rather, teams tend to behave in habitual
ways, even

when presented with evidence that this behavior might be
dysfunctional
(Gersick & Hackman, 1990). However, research and theory
regarding the
determinants and outcomes of reflexivity is still scarce.
Therefore, given the
importance of reflexivity for the effective functioning of teams,
it is crucial
to understand what factors motivate teams to become more
reflexive, and to
develop theory about the determinants of reflexivity. In the
present study, we
focused on a factor that may be of particular importance in this
respect: team
leadership (see Bass, 2000; Hirst et al., 2004). More
specifically, we examined
how leadership may motivate group members to become more
reflexive. We
tested the hypothesis that transformational leadership is
positively related to
team reflexivity and team performance, and that this
relationship is mediated
by a shared vision within the team. We expect that
transformational leader-
ship will enhance a common goal and shared vision in the team.
Having the
shared frame of reference inherent in such a shared team vision
will enhance
teams’ ability to collectively reflect on team objectives and the
strategies used
to reach them and, in turn, this should enhance team
effectiveness.
Team reflexivity, transformational leadership and a shared
vision

The concept of reflexivity is rooted in ancient philosophy and is
seen by Greek
philosophers such as Socrates and Epicures as seeing the world
and oneself in
a dialectical manner. In developmental and educational
psychology, a related
concept is metacognitions, referring to reflecting on the
learning process (e.g.
Brown, 1978; Flavell, 1979), whereas in medicine the same
concept is used
to refer to a reflective approach involving stepping back from
the problem at
hand in order to examine and reflect on the thinking process. In
a work
context, reflexivity has taken on the meaning of both
sensemaking and
learning from actions (Senge, 1990; see Daley, 2001). For
instance, Schön
Human Relations 61(11)1 5 9 4
(1983) coined the term ‘reflection in action’ to describe the
interwoven cycle
of reflection and action of professionals completing complex
tasks. At the
team level, reflexivity is defined as ‘the extent to which group
members overtly
reflect on, and communicate about the group’s objectives,
strategies (decision-
making) and processes (communication), and adapt these to

current or
anticipated circumstances’ (West, 2000: 3). A critical difference
between indi-
vidual reflexivity and team reflexivity is that reflexivity at the
team level
necessarily involves discussion and is thus observable behavior,
and so can
be seen as a relational activity (Barge, 2004). According to
Barge (2004: 92),
‘the latter recognizes that managers are continually co-creating
conversational
texts with others and that the tactics they employ influence the
shape and
form of the emerging text’. In this view, reflexive practice
involves making
sense of situations together in a continually changing
environment. In the
current research, we focus on reflexivity in a relational way,
entailing
communicating views and ideas within a team.
Research has consistently found reflexivity to be positively
related to
subjective as well as objective measures of team performance
(Carter & West,
1998; Hirst et al., 2004; Schippers et al., 2003; Somech, 2006;
Tjosvold
et al., 2004). For example, in a study among 19 BBC production
teams,
Carter and West (1998) found that reflexivity predicted team
effectiveness.
A study among three-person experimental groups showed that
teams in the
reflexivity condition performed better than teams in the control
condition
(Gurtner et al., 2007), and a field study among 59 work teams

found that
team reflexivity mediated the (moderated) relationship between
diversity and
team performance, commitment, and satisfaction (Schippers et
al., 2003).
The converging evidence that reflexivity enhances team
performance
suggests that organizations may improve team performance by
fostering
team reflexivity. This gives rise to the question of how team
reflexivity may
be stimulated, and an obvious route would be through team
leadership. Team
leaders carry the responsibility for the day-to-day functioning
of the team
and should be especially well positioned to influence team
processes like
reflexivity. Although it seems an obvious relationship, and
some authors hint
at the relationship between transformational leadership and
learning
(e.g. Bass, 2000), empirical evidence regarding the relation
between trans-
formational leadership and reflexivity (i.e. team learning) is
still lacking. In
a similar vein, Gersick and Hackman (1990) suggested that a
team leader
might help the team to develop meta-routines, which prompt
members to
initiate re-evaluation of first level routines in a regular and
timely fashion,
and thus become more reflexive. Indeed, others also mention
collective infor-
mation processing and team metacognition (i.e. reflexivity) as
important

mediators between leadership processes and team effectiveness
(e.g. Zaccaro
et al., 2001).
Schippers et al. Transformational leadership and team
reflexivity 1 5 9 5
First evidence for the role of team leadership in engendering
team
reflexivity may be found in studies by Hirst et al. (2004), who
found that
facilitative leader behaviors were positively related to team
reflexivity, which
in turn affected customer ratings of team performance, and
Somech (2006),
who found that both directive and participative leadership
moderated the
relationship between functional heterogeneity and team
reflexivity, and that
team reflexivity, in turn, influenced innovation in a sample of
health care
teams. However, the notion that leaders may engender
rethinking or re-
flexivity by fostering a shared vision is found in theories of
transformational
leadership (e.g. Bass, 1985, 2000; Bass & Riggio, 2006; Berson
et al., 2001).
For example, this literature suggests that leaders present a
vision that raises
followers’ awareness of and dedication to the ideals of the

group and that
may help them see old problems in a new light.
Transformational leadership is a style of leadership that
transforms
followers by stimulating them to go beyond self-interest
through altering
their morale, values and ideals, and motivating them to perform
above expec-
tations (Bass, 1985; Yukl, 1999). Since its introduction,
transformational
leadership has been strongly emphasized in the management
literature (Bass,
1985; Bass & Avolio, 1990; Burns, 1978; House, 1996; Lowe et
al., 1996;
Sashkin, 1988; Yukl, 1998). It is often suggested, but hardly
ever tested, that
transformational leadership is related to a shared vision and
learning among
followers. Communicating a compelling vision is seen as an
important part
of transformational leadership, which is supposed to be related
to a shared
vision among followers (e.g. Berson et al., 2001). Our central
argument is
that transformational leader behavior will enhance the
development of a
shared vision among team members and that this shared vision
in turn affects
reflexivity. The inspirational motivation, charismatic, and
intellectual stimu-
lation aspects of transformational leadership seem especially
important for
team reflexivity. For instance, through intellectual stimulation,
transfor-
mational leaders encourage followers to consider new points of

view and
question old assumptions (Bass, 1985).
Prior research has shown that transformational leadership is a
higher-
order construct comprising several components. However, even
after decades
of research little consensus exists in the literature about the
exact
components comprising transformational leadership. An often
used ques-
tionnaire to measure transformational leadership is the
Multifactor Leader-
ship Questionnaire (MLQ) as developed by Bass and associates
(e.g. Avolio
et al., 1999), although other dimensional measures (e.g.
Podsakoff et al.,
1990) or combined scales (e.g. De Hoogh et al., 2005) have also
been used.
Here, such a shorter, combined scale was used. Several reviews
and meta-
analyses showed that the subscales of transformational
leadership are highly
intercorrelated (often around or even above .80), and that
transformational
leadership can also be seen as one higher-order construct (e.g.
Den Hartog

et al., 1997; for reviews, see Avolio et al., 1999; Lowe et al.,
1996). Although
often used as a unidimensional scale, sometimes subscales have
been used in
prior research. For example, in recent MLQ research on the
factor structure
of this instrument a five-factor model emerged (Antonakis et
al., 2003).
In much of the research on transformational leadership to date,
three
subcomponents are discerned: a combined scale of
charisma/inspirational, as
well as intellectual stimulation and individualized
consideration. Charisma/
inspirational leadership entails appealing to a collective identity
and express-
ing an energizing vision. Intellectual stimulation is expressed
by encouraging
followers to see things in a new light, and to question the status
quo. Indi-
vidualized consideration entails understanding follower’s needs,
and helping
them to grow to their full potential. Since this latter construct is
more on the
dyadic level, and can be different for followers from the same
team with the
same leader, we decided to focus on the transformational
leadership aspects
that we expected to be positively related to team level
reflexivity. Furthermore,
individualized consideration is not always included in
operationalizations of
transformational leadership. For example, in their meta-
analysis, Avolio et al.
(1999) report a higher-order factor of transformational

leadership, consisting
of charisma, inspirational and intellectual stimulating
leadership, while indi-
vidualized consideration and contingent reward (an aspect of
transactional
leadership) comprise a second higher-order factor.
Transformational leaders articulate a vision that describes a
better
future and is congruent with the values of followers. The
leader’s personal
example serves as a model of the kind of behavior required to
attain the
vision. Visioning is not only seen as crucial to arouse followers
in the leader-
ship literature, the importance of having a shared vision as a
motivating
force is also found in the team literature (e.g. West, 2000).
Where the team
literature focuses on the sharedness of the team vision, which is
held to be
important for the achievement of a long-term orientation and
longer-term
goals of the team (see Guzzo & Dickson, 1996), the leadership
literature
addresses leader’s capacity to develop and communicate such a
vision, which
is attractive and motivating for followers, and which they
collaboratively will
try to attain (e.g. Bass, 1985; see Conger & Kanungo, 1987;
Tourish &
Pinnington, 2002). However, few studies have tested whether
such a com-
municated vision indeed becomes a shared vision in the team.
We argue that having a shared, overarching goal or vision of the

future
ensures a shared frame of reference for team members, which
makes it easier
for teams to reflect effectively on their functioning. If teams
have a clear team
goal (i.e. a shared vision), they will be better able to reflect,
because they will
have more of an idea if they are on track in reaching the goal
(see Locke &
Latham, 1990). The goal will aid the reflective team in deciding
if they are
reflexivity 1 5 9 7
on the right track or need to adapt. A transformational leader
will aid this
process by regularly promoting the goal to the team (i.e.
enhancing a shared
vision) and thus stimulate reflexivity in an indirect way. Thus,
we test
whether transformational leadership (i.e. charisma/inspiration
and intellec-
tual stimulation) is positively related to reflexivity and
performance through
its relationship with a shared vision. In other words, we test
whether leaders
who engender shared norms, aspirations and ideals and show
team members
how this new frame of mind helps to look at problems from new

angles, will
stimulate the formation of a shared vision within teams and,
subsequently,
increase reflexivity within teams.
Hypothesis 1. Transformational leadership is positively related
to team
reflexivity.
Hypothesis 2. A shared vision mediates the relationship between
trans-
formational leadership and team reflexivity.
Besides the proposed relationship with team process, many
researchers argue
that a link between transformational leadership and team
performance should
exist (Yukl, 1998), and several studies have tested this link. For
instance, Lim
and Ployhart (2004) examined the impact of transformational
leadership on
team performance in combat teams and found a positive
relationship. Another
study found that transformational leadership positively affected
group
potency, and in turn group effectiveness (Sosik et al., 1997).
Furthermore, a
study among 47 intact teams found that transformational
leadership was
related to group effectiveness, through the effect on group
cohesion,
empowerment and collective efficacy (Jung & Sosik, 2002). It is
important to
note that, although we do expect a relationship between
transformational
leadership and team performance, other variables that are not

measured in
the current study, such as motivation, group cohesion and
collective efficacy,
are also likely to influence team performance (e.g. Jung &
Sosik, 2002; Sosik
et al., 1997; West, 2000). We thus expect reflexivity (and a
shared vision) to
partially mediate the relationship between transformational
leadership and
team performance. This line of thinking also assumes that
reflexivity mediates
between a shared vision on the one hand, and team performance
on the other
hand. The research model is depicted in Figure 1. Thus, we
expect:
Hypothesis 3. A shared vision and reflexivity both partially and
sequentially mediate the relationship between transformational
leader-
ship and team performance.
Method
Participants and procedure
Thirty-two teams from nine different organizations participated
in this study.
The teams included management teams, service teams,

production teams,
teams in government service, and facilitating teams. The teams
came from
companies in the IT, insurance and banking sector, government
and chemical
industry. Following Hackman (1987), we considered teams as
composed of
individuals who both see themselves and are seen by others as
an inter-
dependent social entity. Furthermore, teams are embedded in a
larger
organization, and the team’s performance affects others, for
instance
suppliers or customers. Only teams that met these criteria were
considered
for participation. In most cases team members were assigned to
the teams
when they were first formed; teams did not select members
themselves. We
purposely sought teams with different, but relatively
knowledge-intensive,
tasks to include in the study. Teams with very routine jobs were
not
considered for inclusion in the study, as reflexivity is likely to
be less relevant
for such teams. The team tasks of the participating teams
differed widely,
from administrative or production work (production teams) to
leading a
company (management teams).
Teams were recruited by phone. For all teams, questionnaire
packages
were mailed to the team leaders who had agreed to participate in
the study.
These team leaders then handed the questionnaires to their team

members,
and ensured that these questionnaires were completed in private.
A cover
letter described the purpose of the study and guaranteed the
respondents
confidentiality. Instructions for completion of the questionnaire
were given
reflexivity 1 5 9 9
H3
H1/H2
Transformational
leadership
Shared
vision
Team
reflexivity
Team
performance
Figure 1 Hypothesized direct and indirect relationships in this
study
Note: Hypotheses in bold are the hypotheses including the
mediator(s)

on the first page. All teams had an appointed team leader, which
enabled the
researchers to ensure that all team members were referring to
the same team
leader when filling out the questionnaire. All individual team
members sent
the questionnaires directly to the researchers. Feedback sessions
with the
teams were held to explain the results.
The response rate was 95 percent. Two questionnaires were
incom-
plete and thus excluded from further analyses. The remaining
respondents
(N = 238) were from 32 teams ranging in size from four to 14
members with
an average of 7.56 persons per team and at least two
respondents per team.
In most teams, all team members returned the questionnaire. Of
these
respondents, 68 percent were male. The mean age of
respondents was
38 years (SD = 9.28).
Measures
Transformational leadership
Transformational leadership was measured using six items
based on Den
Hartog et al. (1997). Because we had access to the teams on the
condition
that the survey would be as short as possible, we were unable to

measure
transformational leadership with a lengthy questionnaire. The
items in the
scale were formulated to measure a combination of intellectual
stimulation,
inspirational motivation and charisma, which we argue are the
key elements
of transformational leadership in this context (Waldman et al.,
2006). Other
studies have used similar short measures to tap such forms of
leadership
(De Hoogh et al., 2005; Den Hartog et al., 2007; Waldman et
al., 2006).
The items were: ‘The team leader serves as a role model for
me’, ‘The team
leader makes me aware of strongly held values, ideals, and
aspirations which
are shared in common’, ‘I have complete confidence in
him/her’, ‘In my mind,
he/she is a symbol of success and accomplishment’, ‘Shows us
how to look
at problems from new angles’, ‘Stimulates me to back up my
opinions with
good reasoning’ (1 = strongly disagree, 5 = strongly agree), α =
.85. The first
pair of items concerns the subscale inspirational motivation, the
second pair
charisma, and the last pair intellectual stimulation.
Shared vision
Shared vision was measured with five items, developed in the
context of this
research and in line with previous literature (e.g. Burningham &
West, 1995;
Senge, 1990; Tindale & Kameda, 2000). The items were ‘This

team has a
vision’, ‘Team members are acquainted with the vision’, ‘Team
members
agree on the team’s vision’, ‘The vision provides team members
with clear
directions with respect to the work that has to be done’, and
‘This team has
a clear vision of what it wants to achieve’ (1 = strongly
disagree, 5 = strongly
agree), α = .92.
Reflexivity
Reflexivity was measured by six items from the reflexivity
measure of
Schippers et al. (2007) that are in part based on the scale
developed by Swift
and West (1998). Examples of items are: ‘We regularly discuss
whether the
team is working effectively’, ‘The methods used by the team to
get the job
done are often discussed’, and ‘We regularly reflect on the way
in which we
communicate’ (1 = strongly disagree, 5 = strongly agree), α =
.86.
Performance

In order to avoid potential common source bias, external
managers or super-
visors (who were not team members) were asked to rate the
performance of
the 32 teams on a scale from one to ten (1 = very bad to 10 =
very good).
We asked team members and team leaders to identify such a
manager who
had detailed knowledge about their team performance. In all
teams, team
members and the leader agreed on a manager that could best
rate their team
performance. The researchers checked this with the proposed
managers,
before asking them to rate the teams’ overall performance. This
relatively
simple measure was used because some managers had to rate up
to six teams.
Results
Confirmatory factor analysis of the transformational leadership
scale and measurement model
In order to test for convergent and discriminant validity of the
trans-
formational leadership scales as well as the measurement model,
we
conducted confirmatory factor analyses, using LISREL.
Although confirma-
tory factor analyses sometimes offer support for the
hypothesized dimensions
of transformational leadership, a major problem is that
transformational
leadership dimensions do not have differential relations with

outcome
variables (Bycio et al., 1995). Over the last decades, the
transformational
leadership scale has also often been used as a unidimensional
construct
(Judge & Bono, 2000; Kark et al., 2003; Lim & Ployhart, 2004).
While the
reflexivity 1 6 0 1
more parsimonious measurement of our hypotheses would thus
reflect the
unidimensional construct of transformational leadership, an
obvious issue is
whether the model with one unidimensional factor reflects the
data satisfac-
torily. Therefore, we conducted a series of confirmatory factor
analyses on
the transformational leadership items, to see if a unidimensional
scale would
fit the data better than a two- or three-factor model. We
estimated the models
using maximum likelihood techniques within LISREL VIII.
Subsequently, we
tested our entire measurement model with all constructs in one
analysis.
Based on prior research, the following subscales of
transformational

leadership were distinguished: inspirational motivation,
charisma and intel-
lectual stimulation. Since an often-used subscale refers to the
combination of
intellectual stimulation and charisma, (for a review, see Avolio
et al., 1999)
we composed this scale for the two-factor solution. We
compared the fit of
the unidimensional model to the two- (i.e.
charisma/inspirational motivation
and intellectual stimulation) and three-factor structure (i.e.
inspirational
motivation, charisma and intellectual stimulation). In these
models, the
factors were allowed to correlate. For the unidimensional
model, χ2(9, N =
222) = 13.18 (p < .01), AGFI = .95, and RMSEA = .05; for the
two-factor
structure χ2(8, N = 222) = 8.12, AGFI = .98, and RMSEA = .00;
for the
three-factor structure χ2(6, N = 222) = 1.04, AGFI = .98, and
RMSEA = .00.
These results show that the chi-squares and fit indices do not
differ much
between these three models. Although the improvement in fit of
the two-
factor solution over the unidimensional model was significant
(χ2diff = 8.12,
d.f. = 1, p < .05), the absolute difference in fit is small. The
improvement in
fit of the three-factor solution over the two-factor solution was
not signifi-
cant (χ2diff = 1.04, d.f. = 2, NS). A test of discriminant validity
(recommended
by Fornell & Larcker, 1981, and described in full by Netemeyer
et al., 1990),

is to test whether the variance extracted estimates of the two-
and three-
factor solutions exceed the square of the correlation between the
constructs.
If this is the case, evidence of discriminant validity exists
(Fornell & Larcker,
1981). The variance extracted estimates for the two-dimensional
construct
are .56 for inspirational motivation/charisma, and .51 for
intellectual
stimulation. These are both lower than the square of the
correlations between
the constructs (φ = .86, φ2 = .74), indicating that no support for
discriminant
validity of the two-factor solution exists. The variance extracted
estimates
for the three-dimensional construct are .67 for inspirational
motivation, .49
for charisma and .51 for intellectual stimulation. These are
lower than the
square of the correlations between the constructs (φIM/C = .96,
φIM/C2 = .92;
φIM/IS = .84, φIM/IS2 = .71; φC/IS = .86, φC/IS2 = .74). These
results indicate
that there is no support for discriminant validity of the three
scales. The
standardized loadings suggest convergent validity of a single
factor of

transformational leadership (see Table 1). Concluding, we used
trans-
formational leadership as one factor in subsequent analyses.
Using the same procedures as described above, we tested the
measure-
ment model by comparing the fit of the unidimensional model to
the hypoth-
esized three-factor structure (i.e. transformational leadership,
shared vision
and reflexivity as separate constructs). For the unidimensional
model,
χ2(119, N = 225) = 815.14, p < .001, AGFI = .49, RMSEA =
.21; for the
three-factor structure χ2(116, N = 225) = 158.51, p < .001,
AGFI = .89,
RMSEA = .04. The significant improvement in fit of the three-
factor solution
over the unidimensional model, χ2diff = 656.63, d.f. = 3, p <
.001, offers
support for the discriminant validity of the scales.
We then tested whether the variance extracted estimates
exceeded the
squares of the correlations (Fornell & Larcker, 1981). The
variance extracted
estimates are .50 for transformational leadership, .73 for shared
vision and
.41 for reflexivity. All exceed the square of the correlations
between the
constructs (φs are .10, .11, and .29 respectively), which offers
further support
for the discriminant validity between these three constructs (see
Table 1). The

reflexivity 1 6 0 3
Table 1 Measurement properties
Construct and indicators Standardized loading Variance
extracted estimate
1 Transf. leadership .50
λx1 .81
λx2 .78
λx3 .65
λx4 .73
λx5 .60
λx6 .67
2 Shared vision .73
λx1 .85
λx2 .87
λx3 .85
λx4 .88
λx5 .81
3 Reflexivity .41
λx1 .58
λx2 .61
λx3 .53
λx4 .72
λx5 .65
λx6 .72

standardized loadings for the three scales show that the items
load signifi-
cantly on their respective constructs, offering support for the
convergent
validity of the three scales. Support for the nomological validity
will be
presented in the next section, where we discuss the expected
relationships
(Hair et al., 2006).
Data aggregation
Although we handed out questionnaires to the individual team
members, our
measures were clearly aimed at the team level, and therefore the
variables in
this study are expected to operate at the team level of analysis.
Furthermore,
our hypotheses identified the group as the unit of analysis. ICC-
values
reported in Table 2 support this. James (1982) reports a median
ICC(1) of
.12 for the organizational literature. The ICC(1) values for the
variables in
this study are all higher than .12. In the table, we also report the
ICC(2)
values. However, since the ICC(2) value also depends on team
size, with
higher values of ICC(2) as team size increases (Bliese, 2000),
we chose to
depend mainly on the outcomes of ICC(1) in deciding whether
or not to
aggregate the individual-level scores. To further assess within-
team agree-
ment, we calculated the rwg(j) (James et al., 1984, 1993). A

value of .70 or
above is suggested as ‘good’ with respect to within-group
interrater agree-
ment (James et al., 1993). Rwg(j) averaged .81 for
ship, .74 for vision, and .79 for reflexivity, all were well above
.70 and
suggesting that aggregating to the team level is justified.
The team level correlations between all variables are presented
in
Table 2. As expected, significant positive correlations are found
for trans-
formational leadership and team performance, as well as shared
vision and
team reflexivity.
Table 2 Means, standard deviations, F-values, ICC-values,
aggregate level
intercorrelations and Cronbach’s alphas (N = 32 teams)
Variable M SD F(59, 392) ICC(1) ICC(2) 1 2 3 4
1 Transf. leadership 3.34 .44 2.84** .21 .68 .85
2 Shared vision 3.24 .61 3.51** .32 .79 .43** .92
3 Reflexivity 2.92 .39 2.33** .16 .61 .32* .61** .86
4 Performancea 7.03 .97 – – – .32* .33* .44** –
Note: * p ≤ .05; ** p ≤ .01; one-tailed.
a Supervisor ratings of performance.

Hypotheses testing
Hypotheses 1 through 3 predicted direct and mediating
relationships. We
tested these relationships through series of regression analyses.
These
relationships are described below. We ran all analyses with and
without team
size and kind of team as control variables. Doing so did not
change our
results significantly, and thus, to increase power, the results of
the analyses
without control variables are reported. In order to account for
nested effects
of teams within organizations, we performed hierarchical linear
modeling
using STATA in addition to the regression analysis performed
in SPSS.
We hypothesized a main effect of transformational leadership
on
team reflexivity (Hypothesis 1) and sequential mediational
effects: Trans-
formational leadership is expected to result in a shared vision
amongst
followers and a shared vision is expected to be related to
enhanced team
reflexivity (Hypothesis 2), which in turn is expected to lead to
enhanced team
performance (Hypothesis 3).1
To examine the sequential mediating roles of a shared vision
and reflex-

ivity in the relationship between transformational leadership
and perform-
ance, three steps were followed, in line with the suggestions of
Baron and
Kenny (1986). First, we should demonstrate that there is a
relationship
between the antecedent and the consequence. Regression
analyses showed
significant relationships (see Figure 2). As predicted by
Hypothesis 1, a
relationship between transformational leadership and team
reflexivity was
found (β = .32, p < .01), as well as a relationship between
transformational
leadership and team performance (β = .32, p = .05). Second, the
relationship
between the antecedent and the mediator should be significant,
as well as the
relationship between the mediator and the consequence. A
relationship
between transformational leadership and a shared vision was
indeed found
(β = .43, p < .01), as well as a relationship between a shared
vision and
reflexivity (β = .58, p < .01). Furthermore, the mediator shared
vision was
positively related to team reflexivity, and the mediator team
reflexivity
was positively related to team performance (see Figure 2).
Finally, the unique impact of the mediators (shared vision and
reflex-
ivity) should be demonstrated. In line with this, our hierarchical
regression
analyses revealed that the betas of the simple main effects
declined and

became non-significant when shared vision was added to the
equation
(change in beta from .32 to .08), supporting Hypothesis 2.
Moreover, the
beta values also declined and became non-significant when
reflexivity
was added in the last step (change in beta from .32 to .19; see
Figure 2),
corroborating Hypothesis 3. When reflexivity was added to the
equation, the
relation between a shared vision and team performance also
became non-
significant (change in beta from .33 to .09). With respect to
performance, we
reflexivity 1 6 0 5
expected a partial mediational effect, as other variables besides
the ones
measured in the current study are also expected to influence
performance,
and the remaining beta coefficient seems to point in that
direction, although
it is not significant after adding the mediators.2
We then performed Sobel tests in order to assess whether the
decreases
in the betas of the hypothesized mediational models are
significant

(Goodman, 1960). For the relation transformational leadership–
shared
vision–reflexivity, the z-value (one-tailed) was 2.15, p < .05.
For the relation
shared vision–reflexivity–team performance the z-value (one-
tailed) was
1.62, p < .05.
Our results thus suggest that transformational leadership is
related to
a shared vision among team members, which is in turn related to
increased
team reflexivity. This is ultimately related to enhanced
performance as
proposed in Hypothesis 3.
Mediation can also be demonstrated by a procedure put forward
by
Preacher and Hayes (2004, 2007a), involving bootstrapping
(Shrout &
Bolger, 2002). Bootstrapping is a nonparametric method for
assigning
measures of accuracy to statistical estimates (Efron &
Tibishirani, 1998;
Mooney & Duval, 1993), whereby the standard errors are
estimated using
the available data. It is an alternative test to normal-theory tests
of mediation
(e.g. Shrout & Bolger, 2002), and has been used in previous
research to test
for mediation (Brown et al., 2006), and moderated mediation
(Giessner &
Figure 2 Main and mediating relationships of transformational

leadership with
supervisor-rated team performance (N = 32 teams)a
a Numbers above the arrows represent standardized coefficients
(betas). Betas in bold are based
on regression equations including the connecting mediator.
Note: * p < .05, ** p < .01, one-tailed tests.
.32*/.19
.32*/.08
.43** .58**
.33*/.09
Transformational
leadership
Shared
vision
Team
reflexivity
Team
performance
.52**
van Knippenberg, 2008). This procedure is recommended for

testing of
indirect effects, especially with smaller sample sizes, because it
has no assump-
tions regarding underlying sampling distributions (Shrout &
Bolger, 2002).
The formal test for mediation involves computing confidence
intervals around
the product term (a*b), and if zero falls outside of this 95
percent confidence
interval, the indirect effect is significant and mediation has
occurred. Follow-
ing recommendations, we resampled 1000 times, and used the
percentile
method to create 95 percent intervals (Preacher & Hayes,
2007b). This
approach provided consistent results with the mediation
analyses described
above. Specifically, zero fell outside the confidence interval
around the indirect
effects, ranging from .01 to .94. These results provide
convergent evidence
that, in line with our hypotheses, shared vision mediates
between trans-
formational leadership and reflexivity, and that shared vision
and reflexivity
mediate between transformational leadership and performance.
In order to account for the nested effects of teams within
organizations,
we estimated multilevel regression models using the linear
mixed effects
program xtmixed in STATA 9.1 (StataCorp., 2005), in addition
to the
regression analysis performed in SPSS. Multilevel regression
analysis stat-
istically models both within-groups as well as between-groups

relations (Bryk
& Raudenbush, 1992; Snijders & Bosker, 1999). In our case,
within-group
effects refer to the team-level effects, whereas between-group
relations refer
to the organizational-level effects. Unlike ordinary least squares
regression
(OLS), multilevel analysis considers statistical dependencies of
observations
within groups as well as differences across groups, and hence
provides less
biased estimates for standard errors of regression coefficients.
We estimated
a two-level model with teams nested within organizations, using
maximum
likelihood (ML) estimation. We reported fixed effects (gammas)
analogous
to regression coefficients. To evaluate whether each study
variable signifi-
cantly added to the explanation of team performance, we
calculated likeli-
hood ratio tests. Likelihood ratio tests determine whether model
fit (i.e.
log-likelihood values) of a model with more parameters is
significantly better
than one with fewer parameters. We also computed the
proportional re-
duction of prediction error when predictors were added to the
model, which
is analogous to effect sizes or R2 in multiple regression analysis
(Snijders &
Bosker, 1999).
Results of these analyses showed virtually the same results as
reported
for the multiple regression analyses. Specifically, a relationship

between
transformational leadership and team reflexivity was found (γ =
.36, p < .01),
as well as the relationship between transformational leadership
and team
performance (γ = .69, p < .05). Also a relationship between a
shared vision
and reflexivity (γ = .38, p < .01) was found. Furthermore,
shared vision was
reflexivity 1 6 0 7
positively related to team reflexivity, and the mediator team
reflexivity was
positively related to team performance. Finally, the gammas of
the simple
main effects declined and became non-significant when shared
vision
(change in gamma from .46 to .00) and reflexivity (change in
gamma from
1.07 to .87) were subsequently added in the last steps. These
results show
that our predicted model also holds when the nested structure of
the data is
accounted for.
Discussion
Team reflexivity is seen as a key factor in team effectiveness

and enhancing
reflexivity is therefore important to organizations. A relevant
question related
to this is whether and how reflexivity can be fostered by team
leaders (Hirst
et al., 2004; Somech, 2006; see Gersick & Hackman, 1990). The
current
study therefore focused on potential determinants of reflexivity,
and more
specifically on the relation between transformational leadership
and reflexiv-
ity through the establishment of a shared vision. Results
supported our
hypotheses. Positive relationships between team leaders’
transformational
leadership, a shared team vision, team reflexivity and team
performance were
found, as predicted. The predicted mediational model was also
supported. We
found that where team leaders were rated as more
transformational, teams
reported a stronger shared vision, and having this shared vision
was positively
related to team reflexivity. This was in turn positively related to
team perform-
ance, as rated by an external manager. These results highlight
the direct and
the indirect relations between transformational team leadership,
shared
vision, reflexivity and performance in work teams.
The current study showed that one way in which the team
leader’s
behavior plays a role in enhancing reflexivity and performance
is through
engendering a shared vision within the team. The current

research is the first
to show that transformational leadership is important in
stimulating team
reflexivity and subsequent team performance. Moreover, the
results from this
study suggest that this effect might be mainly due to the
transformational
leader’s role in creating a shared team vision. In our study, the
impact of
transformational leadership (operationalized as a combination
of intellectual
stimulation, charisma and inspirational motivation) on
reflexivity was
mediated by a shared vision. In teams with leaders who inspire
confidence
and awe, present new ways of seeing the world, and who
stimulate their team
members to rethink old habits, team members report having a
shared vision.
In teams with a shared vision, team members see their team as
having a
common view on the goals and vision of the team and are also
aware that
team members share this view. In turn, this stronger shared
outlook of team
members is related to increased reflection on, and
communication about,

objectives, strategies and processes within the team. Sharing
this vision or
basic outlook on end-goals seems to make it easier for the team
to reflect on
whether it is still on the right path as well as on alternative
ways forward.
Finally, in line with previous studies, we found that teams
higher on re-
flexivity outperform those lower on reflexivity. As noted
earlier, several
theorists have advanced such propositions, but the available
empirical
body of knowledge on the role of team leaders as well as the
process of reflex-
ivity in teams is small. Hence, an important contribution of the
present
research is that it provides empirical support for a compelling
argument that
is often advanced but hardly tested.
The current study has several strengths and limitations. An
important
strength of this research lies in the fact that it was done
amongst several
different teams from different kinds of organizations, which
means that the
findings can probably be generalized to several work settings.
However, some
limitations can be outlined as well. A first limitation lies in the
cross-sectional
nature of this study. This design does not allow for testing of
directionality
of the results. Although the mediational tests are consistent with
a causal
chain between transformational leadership, a shared vision,
reflexivity and

team performance, according to Shrout and Bolger (2002: 439):
‘statistical
mediation analyses based on non-experimental data provide
suggestive
rather than definitive evidence regarding causal processes’. In
other words,
reverse causality (e.g. performance increasing reflexivity)
cannot be ruled out
based on these data and the causal ordering should be tested. In
order to test
for directionality, longitudinal and experimental research will
be necessary.
Second, the performance of teams could not be measured
through more
‘objective’ measures, for instance, team output or customer
satisfaction. This
was due to the fact that the teams in our sample had very
different kinds of
tasks and roles that could not easily be compared in terms of
team output
or customer satisfaction (e.g. not all teams had customers or
produced
tangible output). To minimize bias, we did ensure that the team
was rated
by an external (higher level) manager, who had detailed
knowledge of the
teams’ performance, rather than by the team members
themselves or even
the internal team leader, whose behavior was rated by the
relationships with
more comprehensive measures of performance and, of course,
with more
teams as another limitation of our study is that the sample size
at the team
level is limited. Note, however, that the sample size in the

current study is
similar to many other team studies and based on a sizeable
underlying set of
individual ratings and responses.
A final limitation may lie in the questionnaire used to measure
trans-
formational leadership. Some researchers argue that this
construct should be
used as a formative construct (the measures come together to
form the
reflexivity 1 6 0 9
construct, e.g. socio-economic status) as opposed to a reflective
construct (the
measures are seen as reflections of the underlying latent
construct; Jarvis
et al., 2003; MacKenzie et al., 2005; Podsakoff et al., 2003).
However, since
in our research we used items from an existing scale that was
originally
developed as a reflective construct, it may be hard to use that
same scale as
a formative construct, even if this theoretically makes sense.
Research by
Diamantopoulos and Siguaw (2006) showed that the perspective
taken in
developing a measure has a profound influence on the content of

that ques-
tionnaire. In their research, they developed a questionnaire
measuring export
coordination, starting with an item pool of 30 items, adopting a
reflective
versus a formative perspective. Results showed that only two
items of the
original 30-item pool were common to both measures. A key
difference
between the two perspectives is that scale development
procedures (reflec-
tive) tend to retain highly intercorrelated items, whereas index
construction
procedures (formative) tend to eliminate highly intercorrelated
items. The
current version of the transformational leadership scale has very
high inter-
correlations, which makes it hard to consider the scale as
formative. Future
research could focus more on transformational leadership as a
formative
measure.
Overall the results of this study suggest that transformational
leader-
ship can positively influence reflexivity through the formation
of a shared
vision and this in turn may influence team performance. The
finding that
reflexivity is positively related to team performance (in our and
other studies)
is interesting for practicing managers. However, according to
West (1996,
2000), teams in organizations are generally not very reflexive.
Organizational
objectives and the organizational culture are considered as

givens and often
not subject to discussion (Allen, 1996). Teams tend to behave in
habitual
ways, even when faced with evidence that this behavior might
be dysfunc-
tional in reaching team or organizational goals (Gersick &
Hackman, 1990).
An emphasis on action exists in most companies, which might
explain why
in most companies teams do not take the time to reflect and
learn from past
activities. Yet, our results suggest that enhancing team
reflexivity may
provide an important tool for improving team performance. Our
research
suggests that one way to do so is to build a shared vision in the
team, and
that this shared vision can be built through transformational
team leader
behavior. However, other ways to more directly stimulate
reflexivity in teams
may also be relevant. For example, teams could be trained to be
more re-
flexive. Research is needed to assess how, besides through
transformational
leadership, reflexivity of teams can be enhanced and how
reflexivity can
become more customary and built into teams’ daily, rather than
exceptional,
functioning.

Acknowledgements
The authors wish to thank Christian Troester and three
anonymous reviewers for
their helpful comments on an earlier version of this article.
Notes
1 One could argue that intellectual stimulation might be more
strongly related to
reflexivity, whereas fostering a shared vision would be more
strongly related to
charisma. Although the analyses favored a one-factor solution,
we decided to look
at the correlations for the subscales and compare these. Results
show that the corre-
lation between charisma and a shared vision (.43, p < .01) is not
significantly higher
than the correlation between intellectual stimulation and a
shared vision (.37,
p < .05), and the relationship between inspirational motivation
and a shared vision
is .33 (p < .05). Using Steiger’s test for comparing elements of
a correlation matrix
(Steiger, 1980), these differences in magnitude between the
correlations proved not
significant (z = .57, NS, and z = –.42, NS respectively) The
same holds for the
relation between intellectual stimulation and reflexivity, which
is .35 (p < .05), versus
.33 (p < .05) for the relation between charisma and reflexivity,
and.19 (NS) for the
relation between inspirational motivation and reflexivity (z =

.19, NS; and z = –1.3,
NS respectively). Thus, none of the magnitudes of the
differences in correlations is
significant.
2 One could argue that transformational leadership has an effect
on team reflexivity,
which in turn has an effect on a shared vision (see van Ginkel &
van Knippenberg,
2008). We therefore tested whether transformational leadership
affected team re-
flexivity, a shared vision and in turn, team performance.
However, this relationship
did not hold; when adding shared vision in the last step, the
effect of reflexivity
stayed significant, while the effect of a shared vision was not
significant anymore.
References
Allen, N.J. Affective reactions to the group and the
organization. In M.A. West (Ed.),
Handbook of work group psychology. Chichester: Wiley, 1996,
pp. 371–96.
Antonakis, J., Avolio, B.J. & Sivasubramaniam, N. Context and
leadership: An exam-
ination of the nine-factor Full-Range Leadership Theory using
the Multifactor Leader-
ship Questionnaire. Leadership Quarterly, 2003, 14, 261–95.
Avolio, B.J., Bass, B.M. & Jung, D.I. Re-examining the
components of transformational
and transactional leadership using the Multifactor Leadership
Questionnaire. Journal
of Occupational and Organizational Psychology, 1999, 72, 441–

62.
Barge, K.J. Reflexivity and managerial practice.
Communication Monographs, 2004, 71,
70–96.
Baron, R.M. & Kenny, D.A. The moderator-mediator variable
distinction in social psycho-
logical research: Conceptual, strategic, and statistical
considerations. Journal of
Personality and Social Psychology, 1986, 51, 1173–82.
Bass, B.M. Leadership and performance beyond expectations.
New York: Free Press, 1985.
Bass, B.M. The future of leadership in learning organizations.
Journal of Leadership
Studies, 2000, 7, 18–40.
reflexivity 1 6 1 1
Bass, B.M. & Avolio, B.J. The implications of transactional and
ship for individual, team, and organizational development.
Research in Organizational
Change and Development, 1990, 4, 231–72.
Bass, B.M. & Riggio, R.E. Transformational leadership, 2nd
edn. Mahwah, NJ: Lawrence

Erlbaum Associates, 2006.
Berson, Y., Shamir, B., Avolio, B.J. & Popper, M. The
relationship between vision strength,
leadership style, and context. Leadership Quarterly, 2001, 12,
53–73.
Bliese, P.D. Within-group agreement, non-independence and
reliability: Implications for
data aggregation and analysis. In K.J. Klein & W.J. Kozlowski
(Eds), Multilevel theory,
research, and methods in organizations. San Francisco, CA:
Jossey-Bass, 2000,
pp. 349–81.
Brown, A.L. Knowing when, where, and how to remember: A
problem of metacognition.
Phoenix, AZ: Oryx Press, 1978.
Brown, D.J., Cober, R.T., Kane, K., Levy, P.E. & Shalhoop, J.
Proactive personality and
the successful job search: A field investigation with college
graduates. Journal of Applied
Psychology, 2006, 91, 717–26.
Bryk, A.S. & Raudenbush, S.W. Hierarchical linear models:
Applications and data analysis
methods. London: SAGE, 1992.
Burningham, C. & West, M.A. Individual, climate, and group
interaction processes as
predictors of work team innovation. Small Group Research,
1995, 26, 106–17.
Burns, J.M. Leadership. New York: Harper & Row, 1978.
Bycio, P., Hackett, R.D. & Allen, J.S. Further assessments of

Bass’s (1985) conceptual-
ization of transactional and transformational leadership. Journal
of Applied Psy-
chology, 1995, 80, 468–78.
Carter, S.M. & West, M.A. Reflexivity, effectiveness and
mental health in BBC production
teams. Small Group Research, 1998, 29, 583–601.
Conger, J.A. & Kanungo, R.N. Toward a behavioral theory of
charismatic leadership in
organizational settings. Academy of Management Review, 1987,
12, 637–47.
Cooke, N., Salas, E., Cannon-Bowers, J. & Stout, R. Measuring
team knowledge. Human
Factors, 2000, 42, 151–73.
Daley, B.J. Learning and professional practice: A study of four
professions. Adult Education
Quarterly, 2001, 52, 39–54.
De Hoogh, A.H.B., Den Hartog, D.N. & Koopman, P.L. Linking
the Big Five-Factors of
personality to charismatic and transactional leadership:
Perceived dynamic work
environment as a moderator. Journal of Organizational
Behavior, 2005, 26, 839–65.
Den Hartog, D.N., De Hoogh, A.H. & Keegan, A.E. The
interactive effects of belonging-
ness and charisma on helping and compliance. Journal of
Applied Psychology, 2007,
92, 1131–9.

Den Hartog, D.N., Van Muijen, J.J. & Koopman, P.L.
Transactional versus trans-
formational leadership: An analysis of the MLQ. Journal of
Occupational and Organiz-
ational Psychology, 1997, 70, 19–34.
Diamantopoulos, A. & Siguaw, J.A. Formative versus reflective
indicators in organizational
measure development: A comparison and empirical illustration.
British Journal of
Management, 2006, 17, 263–82.
Edmondson, A. Psychological safety and learning behavior in
work teams. Administrative
Science Quarterly, 1999, 44, 350–83.
Efron, B. & Tibishirani, R.J. An introduction to the bootstrap.
Boca Raton, FL: Chapman
& Hall, 1998.
Flavell, J.H. Metacognition and cognitive monitoring. American
Psychologist, 1979, 34,
906–11.
Fornell, C. & Larcker, D.F. Evaluating structural equations
models with observable
variables and measurement error. Journal of Marketing
Research, 1981, 18, 39–50.
Gersick, C.J. & Hackman, J.R. Habitual routines in task-
performing groups. Organiz-
ational Behavior and Human Decision Processes, 1990, 47, 65–
97.

Giessner, S.R. & van Knippenberg, D. ‘License to fail’: Goal
definition, leader group proto-
typicality, and perceptions of leadership effectiveness after
leader failure. Organiz-
ational Behavior and Human Decision Processes, 2008, 105, 14–
35.
Goodman, L.A. On the exact variance of products. Journal of
the American Statistical
Association, 1960, 55, 708–13.
Gurtner, A., Tschan, F., Semmer, N.K. & Nägele, C. Getting
groups to develop good strat-
egies: Effects of reflexivity interventions on team process, team
performance, and mental
models. Organizational Behavior and Human Decision
Processes, 2007, 102, 127–42.
Guzzo, R.A. & Dickson, M.W. Teams in organizations: Recent
research on performance
and effectiveness. Annual Review of Psychology, 1996, 47,
307–38.
Hackman, J.R. The design of workteams. In J.W. Lorsch (Ed.),
Handbook of organizational
behavior. Englewood Cliffs, NJ: Prentice-Hall, 1987, pp. 315–
42.
Hair, J.F., Black, W.C., Babin, B.J., Anderson, R.E. & Tatham,
R.L. Multivariate data

analysis, 6th edn. Upper Saddle River, NJ: Pearson Prentice-
Hall, 2006.
Hinsz, V.B., Tindale, R.S. & Vollrath, D.A. The emerging
conceptualization of groups as
information processors. Psychological Bulletin, 1997, 121, 43–
64.
Hirst, G., Mann, L., Bain, P., Pirola-Merlo, A. & Richter, A.
Learning to lead: The develop-
ment and testing of a model of leadership learning. Leadership
Quarterly, 2004, 15,
311–27.
House, R.J. Path-goal theory of leadership: Lessons, legacy and
a reformulated theory.
Leadership Quarterly, 1996, 7, 323–52.
James, L.R. Aggregation bias in estimates of perceptual
agreement. Journal of Applied
Psychology, 1982, 67, 219–29.
James, L.R., Demaree, R.G. & Wolf, G. Estimating within-
group interrater reliability with
and without response bias. Journal of Applied Psychology,
1984, 69, 85–98.
James, L.R., Demaree, R.G. & Wolf, G. An assessment of
within-group interrater
agreement. Journal of Applied Psychology, 1993, 78, 306–09.
Jarvis, C.B., MacKenzie, S.B. & Podsakoff, P.M. A critical
review of construct indicators
and measurement model misspecification in marketing and
consumer research. Journal
of Consumer Research, 2003, 30, 199–218.

Judge, T.A. & Bono, J.E. Five-factor model of personality and
transformational leadership.
Journal of Applied Psychology, 2000, 85, 751–65.
Jung, D.I. & Sosik, J.J. Transformational leadership in work
groups: The role of empower-
ment, cohesiveness, and collective-efficacy on perceived group
performance. Small
Group Research, 2002, 33, 316–36.
Kark, R., Shamir, B. & Chen, G. The two faces of
transformational leadership: Empower-
ment and dependency. Journal of Applied Psychology, 2003, 88,
246–55.
Lim, B.C. & Ployhart, R.E. Transformational leadership:
Relations to the five-factor model
and team performance in typical and maximum contexts. Journal
of Applied Psychol-
ogy, 2004, 89, 610–21.
Locke, E.A. & Latham, G.P. A theory of goal-setting and task
performance. Englewood
Cliffs, NJ: Prentice Hall, 1990.
Lowe, K.B., Kroeck, K.G. & Sivasubramaniam, N.
Effectiveness correlates of trans-
formational and transactional leadership: A meta-analytic
review of the MLQ litera-
ture. Leadership Quarterly, 1996, 7, 385–425.
MacKenzie, S.B., Podsakoff, P.M. & Jarvis, C.B. The problem
of measurement model
misspecification in behavioral and organizational research and
some recommended

solutions. Journal of Applied Psychology, 2005, 90, 710–30.
Mooney, C.Z. & Duval, R.D. Bootstrapping: A nonparametric
approach to statistical
inference. Thousand Oaks, CA: SAGE, 1993.
Netemeyer, R.G., Johnston, M.W. & Burton, S. Analysis of role
conflict and role
ambiguity in a structural equations framework. Journal of
Applied Psychology, 1990,
75, 148–57.
reflexivity 1 6 1 3
Podsakoff, P.M., Mackenzie, S.B., Moorman, R.H. & Fetter, R.
Transformational leader
behaviors and their effects on followers’ trust in leader,
satisfaction, and organizational
citizenship behaviors. Leadership Quarterly, 1990, 1, 107–42.
Podsakoff, P.M., MacKenzie, S.B., Podsakoff, N.P. & Lee, J.Y.
The mismeasure of
man(agement) and its implications for leadership research.
Leadership Quarterly, 2003,
14, 615–56.
Preacher, K.F. & Hayes, A.F. SPSS and SAS procedures for
estimating indirect effects in
simple mediation models. Behavior Research Methods,

Instruments, & Computers,
2004, 36, 717–31.
Preacher, K.F. & Hayes, A.F. Asymptotic and resampling
strategies for assessing and
comparing indirect effects in simple and multiple mediator
models. Working paper,
2007a.
Preacher, K.F. & Hayes, A.F. SPSS and SAS macros for
bootstrapping specific indirect effects
in multiple mediator models, 2007b, available online at:
[http:/quantpsy.org], accessed
May 2007.
Salas, E., Dickinson, T.L., Converse, S.A. & Tannenbaum, S.I.
Toward an understanding
of team performance and training. In R.W. Swezey & E. Salas
(Eds), Teams: Their
training and performance. Norwood, NJ: Ablex, 1992, pp. 3–29.
Sashkin, M. The visionary leader. In J.A. Conger & R.N.
Kanungo (Eds), Charismatic
leadership: The elusive factor in organizational effectiveness.
San Francisco, CA: Jossey-
Bass, 1988, pp. 122–60.
Schippers, M.C. Reflexivity in teams. Unpublished dissertation,
Vrije Universiteit,
Amsterdam, 2003.
Schippers, M.C., Den Hartog, D.N. & Koopman, P.L.
Reflexivity in teams: A measure and
correlates. Applied Psychology: An International Review, 2007,
56, 189–211.

Schippers, M.C., Den Hartog, D.N., Koopman, P.L. & Wienk,
J.A. Diversity and team
outcomes: The moderating effects of outcome interdependence
and group longevity, and
the mediating effect of reflexivity. Journal of Organizational
Behavior, 2003, 24,
779–802.
Schön, D. The reflective practitioner. New York: Basic Books,
1983.
Senge, P.M. The fifth discipline: The art and practice of the
learning organization. New York:
Doubleday Currency, 1990.
Shrout, P.A. & Bolger, N. Mediation in experimental and non-
experimental studies: New
procedures and recommendations. Psychological Methods, 2002,
7, 422–45.
Snijders, T.A.B. & Bosker, R.J. Multilevel analysis: An
introduction to basic and advanced
multilevel modeling. London: SAGE, 1999.
Somech, A. The effects of leadership style and team process on
performance and innovation
in functionally heterogeneous teams. Journal of Management,
2006, 32, 132–57.
Sosik, J.J., Avolio, B.J. & Kahai, S.S. Effects of leadership
style and anonymity on group
potency and effectiveness in a group decision support system
environment. Journal of
Applied Psychology, 1997, 82, 89–103.
StataCorp. Statistical software (Version 9.1). College Station,

TX: StataCorp, 2005.
Steiger, J.H. Tests for comparing elements of a correlation
matrix. Psychological Bulletin,
1980, 87(2), 245–51.
Swift, T.A. & West, M.A. Reflexivity and group processes:
Research and practice. Sheffield:
The ESRC Centre for Organization and Innovation, 1998.
Tindale, R.S. & Kameda, T. ‘Social sharedness’ as a unifying
theme for information process-
ing in groups. Group Processes and Intergroup Relations, 2000,
3, 123–40.
Tjosvold, D., Tang, M.M.L. & West, M.A. Reflexivity for team
innovation in China: The
contribution of goal interdependence. Group & Organization
Management, 2004, 29,
540–59.
Tourish, D. & Pinnington, A. Transformational leadership,
corporate cultism and the
spirituality paradigm: An unholy trinity in the workplace?
Human Relations, 2002, 55,
147–72.

van Ginkel, W.P. & van Knippenberg, D. Group information
elaboration and group
decision making: The role of shared task representations.
Organizational Behavior and
Human Decision Processes, 2008, 105, 82–97.
Waldman, D.A., Siegel, D.S. & Javidan, M. Components of
CEO transformational leader-
ship and corporate social responsibility. Journal of Management
Studies, 2006, 43,
1703–25.
West, M.A. Reflexivity and work group effectiveness: A
conceptual integration. In M.A.
West (Ed.), Handbook of work group psychology. Chichester:
Wiley, 1996, pp. 555–79.
West, M.A. Reflexivity, revolution and innovation in work
teams. In M.M. Beyerlein, D.A.
Johnson & S.T. Beyerlein (Eds), Product development teams,
Vol. 5. Stamford, CT: JAI
Press, 2000, pp. 1–29.
Yukl, G. Leadership in organizations, 4th edn. Upper Saddle
River, NJ: Prentice-Hall, 1998.
Yukl, G. An evaluation of conceptual weaknesses in
transformational and charismatic
leadership theories. The Leadership Quarterly, 1999, 10, 285–
305.
Zaccaro, S.J., Rittman, A.L. & Marks, M.A. Team leadership.
Leadership Quarterly, 2001,
12, 451–83.
Michaéla Schippers is Assistant Professor of Organizational

Behavior
at the Rotterdam School of Management, Erasmus University
Rotterdam,
The Netherlands. She received her PhD from the Psychology
Depart-
ment at Vrije Universiteit Amsterdam, The Netherlands. Her
current
research interests include team reflexivity, team diversity and
team
leadership, as well as team cognition, learning and decision-
making. Her
work has been published in journals such as Annual Review of
Psychology,
Journal of Organizational Behavior and Applied Psychology: An
International
Review.
[E-mail: [email protected]]
Deanne N. Den Hartog is Professor of Organizational Behavior
at the
Amsterdam Business School, University of Amsterdam, The
Netherlands.
She is program director of several Business Studies programs
and
received her PhD from Vrije Universiteit Amsterdam (VU
University),
The Netherlands. Her current research addresses leadership,
especially
cross-cultural, charismatic and ethical leadership processes, as
well as
leadership in project based organizations. Other research
interests
include team reflexivity, employees’ proactive and innovative
behavior at
work and human resource management. She is on the editorial
boards of

Leadership Quarterly, Journal of Occupational and
Organizational Psychology
and Applied Psychology: An International Review.
reflexivity 1 6 1 5
Paul Koopman is Emeritus Professor of the Psychology of
Management
and Organization at Vrije Universiteit Amsterdam, The
Netherlands. He
is interested and actively involved in cross-cultural research, in
particular
in relation to issues of HRM, leadership and organizational
culture.
Daan van Knippenberg is Professor of Organizational Behavior
at the
Rotterdam School of Management, Erasmus University
Rotterdam, The
Netherlands, and co-founder of the Erasmus Centre for
Leadership Studies.
He received his PhD from Leiden University, The Netherlands.
His
current research interests include leadership, in particular the
roles of
self/identity and emotions, work group diversity, group
decision-making,

social identity processes in organizations and creativity and
innovation.
Currently, he is Associate Editor of Organizational Behavior
and Human
Decision Processes and of Journal of Organizational Behavior.
Team climate, empowering leadership, and
knowledge sharing
Yajiong Xue, John Bradley and Huigang Liang
Abstract
Purpose – The purpose of this research is to investigate the
impact of team climate and empowering
leadership on team members’ knowledge-sharing behavior.
Design/methodology/approach – A research model was
developed based on prior knowledge
management studies. Survey data were collected from 434
college students at a major US university,
who took courses that required team projects. The partial least
squares technique was applied to test

the research model.
Findings – Team climate and empowering leadership
significantly influence individuals’
knowledge-sharing behavior by affecting their attitude toward
knowledge sharing. These two
constructs also have significant direct effects on the knowledge-
sharing behavior.
Research limitations/implications – The student sample and US
setting might limit the generalizability
of the findings. Nonetheless, this study is based on and extends
prior research, which provides a
deepened understanding of knowledge sharing in the team
context.
Practical implications – This research has practical implications
for how to design teams to facilitate
knowledge sharing. It suggests that cohesive, innovative teams
with members trusting one another and
led by empowering leaders will have a higher level of
knowledge sharing.
Originality/value – This research originally examines the effects
of both team climate and empowering
leadership on knowledge sharing. Little prior research has
carried out such an integrated analysis. This

paper will have significant value for organizations trying to
redesign teams to enhance knowledge
management.
Keywords Team working, Empowerment, Leadership,
Knowledge management
Paper type Research paper
1. Introduction
As an organizational process, knowledge sharing plays a
fundamental role in generating
new ideas and creating business opportunities (Grant, 1996).
Effectively communicated
knowledge benefits all of the involved organizational actors by
improving their performance
and eventually improving financial, marketing, and general
outcomes of the organization
(Alavi and Leidner, 1999). Yet, in practice, inadequate sharing
has been found to be a major
impediment to effective knowledge management (Davenport and
Prusak, 1998). Therefore,
understanding employees’ knowledge sharing behavior has
important implications for
organizations.
Although empirical evidence has uncovered some of the

complex dynamics of knowledge
sharing in general (Ipe, 2003), there is a paucity of research that
explains how individuals
share knowledge in organizational settings. In particular, an in-
depth understanding of
knowledge sharing within teams is desirable because team-
based design is widely adopted
by contemporary organizations. About 82 percent of companies
with 100 or more
employees have team-based mechanisms (Gordon, 1992) and
new collaborative job
designs and work practices require teamwork (Capelli and
Rogovsky, 1994). Therefore, it is
DOI 10.1108/13673271111119709 VOL. 15 NO. 2 2011, pp.
299-312, Q Emerald Group Publishing Limited, ISSN 1367-
3270 j JOURNAL OF KNOWLEDGE MANAGEMENT j PAGE
299
Yajiong Xue is an Assistant
Professor, John Bradley is
Professor of Management
Information Systems and
Huigang Liang is an
Assistant Professor, all in

the College of Business,
East Carolina University,
Greenville, North Carolina,
USA.
Received: 16 June 2010
Accepted 18 October 2010
imperative to understand how team-related factors influence
individuals’ knowledge sharing
behavior.
Given that it is people who actually create, share, and use
knowledge, an organization
cannot effectively exploit knowledge unless its employees are
willing and able to share their
own knowledge and assimilate the knowledge of others (Ipe,
2003). Thus, knowledge
sharing within organizations can be seen as a multifaceted,
complex process that involves
intricate human behaviors (Hendriks, 1999). It implies a
voluntary act by individuals who
participate in the exchange of knowledge even though there are
no compulsory pressures

(Davenport, 1997). It often involves the communication of tacit
knowledge that cannot be
reported through formal channels and is difficult to be
compelled. Hence, the most
appropriate measure to enhance knowledge sharing seems to be
‘‘soft’’strategies relying on
the climate and leadership role of the specific organizational
unit (Hulsheger et al., 2009;
Srivastava and Bartol, 2006).
In this research, the authors investigate knowledge sharing in
light of two team-related
factors: team climate and empowering leadership. First, team
climate refers to an implicit
frame that shapes individual perceptions, attitudes and
behaviors within the group context
(Seibert et al., 2004). It has long been known as one of the most
important sources of social
influence that affects individual behavior in the team
environment (Hulsheger et al., 2009;
West and Anderson, 1996). Second, prior research has
highlighted the importance of the
leader’s role in organizational processes (Liang et al., 2007),
particularly in knowledge

management (Crawford, 2005; Singh, 2008; Srivastava and
Bartol, 2006). A variety of leader
behaviors have been studied, among which empowering
leadership is found to improve
employee’s job autonomy (Bennis and Townsend, 1997). Such
autonomy is essential for
employees to undertake conscious, voluntary knowledge
sharing.
Specifically, in this paper the authors investigate the impact of
both team climate and
empowering leadership on individuals’ knowledge sharing
behavior. This will help
researchers understand how knowledge sharing within the team
environment is
influenced by team-related factors. It will also help
organizations attend to the team
environment and team leaders when trying to enhance
knowledge sharing within
organizations.
This paper proceeds as follows. The next section presents the
literature review and research
model development. The construct operationalization, data
collection, data analysis, and
model testing results are described in the method section, after

which the authors interpret
the findings and discuss their implications for research and
practice. Finally, limitations and
directions for future research are discussed.
2. Theoretical development
Based on the extant research on knowledge management, the
authors develop a research
model (Figure 1) to explain why team members engage in
knowledge sharing. It is proposed
that team climate and empowering leadership help to shape
individuals’ attitudes, which in
turn lead to the desired knowledge sharing behavior. In
addition, team climate and
empowering leadership both have a direct impact on the
knowledge sharing behavior. The
major constructs and hypotheses are discussed as follows.
‘‘ As a social behavior, an individual’s knowledge sharing is
inevitably susceptible to social influences arising from other
people. ’’
PAGE 300jJOURNAL OF KNOWLEDGE MANAGEMENTj
VOL. 15 NO. 2 2011
2.1 Team climate

As a social behavior, an individual’s knowledge sharing is
inevitably susceptible to social
influences arising from other people. Individuals need to be
proximal to the referent others to
be exposed to social influences. Salancik and Pfeffer (1978)
suggest that the immediate
social environment is an important source of information which
individuals use to construct
reality and formulate perceptions, attitudes, and behaviors.
While multiple social networks
can exert influences on individual behaviors, prior research in
social psychology shows that
stronger social influence takes place in work teams because
individuals are likely to identify
most closely with their work team and thus are more willing to
comply with team norms (Fulk,
1993). Liang et al. (2010) also find that team climate
significantly influences individuals’
perceptions, normative beliefs, and technology usage. In this
research, therefore, the
authors contend that a desirable team climate can create an
environment in which
knowledge sharing is encouraged.

The extant literature shows that team climate is a composite
construct consisting of three
dimensions: affiliation, trust, and innovation (Bock et al.,
2005). Affiliation, equivalent to
cohesion in nature, refers to the perception of a sense of
togetherness among members.
Cohesion, defined as members’ attraction to the team (Hogg,
1992), can be considered as a
psychological force that binds people together (Keyton and
Springston, 1990). The sense of
affiliation or cohesion tends to enhance team members’
willingness to care for or help one
another. As a result, they are more likely to share knowledge
with one another.
Trust in the team environment is defined as a member’s
willingness to accept vulnerability
based on a confident expectation of teammates’ competence,
integrity, and benevolence
(Pavlou et al., 2007). Effective communication occurs in an
environment in which trust and
commitment are prevalent (Te’eni, 2001). Huemer et al. (1998)
argue that team members
with stronger trust are more likely to work together
cooperatively and conscientiously. Zand

(1972) finds that team members share information more freely
when they trust one anothers’
capabilities and competencies. Similarly, Weick and Roberts
(1993) argue that to coordinate
knowledge among team members, they need to trust one
anothers’ capabilities. Hsu et al.
(2007) find that social relationships based on trust have a
significant influence on an
individual’s attitude toward sharing knowledge.
Figure 1 Research model
Team
Climate
Empowering
Leadership
Attitude KS
behavior
Control variables
Gender Age
Note:
Second-order
constructs
H1
H3

H4
H5
H2
Participative
Decision Making
Coaching
Trust
Cohesion
Innovativeness
Showing
Concern
Lead by
Example
Informing
VOL. 15 NO. 2 2011 jJOURNAL OF KNOWLEDGE
MANAGEMENTj PAGE 301
Innovativeness in a team refers to the degree to which change
and creativity are actively
encouraged and rewarded within the team. Innovative teams
emphasize learning, open

information flows, and reasonable risk-taking (Bock et al.,
2005). Members of such teams
approve innovations and provide practical support to peers’
innovative initiatives.
Consequently, individuals in the innovative team environment
are more empowered to
share new and creative ideas with each other than individuals in
a non-innovative
environment (Kim and Lee, 1995).
In this study these three constructs – cohesion, trust, and
innovativeness – were used as
measures of team climate with the expectation that it not only
influences an individual’s
attitude toward knowledge sharing, but also facilitates the
actual knowledge sharing
behavior. Therefore, the following hypotheses are derived:
H1. Team climate has a positive influence on knowledge sharing
attitude.
H2. Team climate has a positive influence on knowledge sharing
behavior.
2.2 Empowering leadership
One of the ways organizations could improve efficiency and
performance is to empower

their employees. Knowledge sharing is a critical aspect of
empowered teams (Argote,
1999). Prior research has shown that knowledge sharing is a
significant determinant of
organizational performance and a team’s leader plays a pivotal
role in making knowledge
sharing possible in the team (Srivastava and Bartol, 2006). In an
empowering organizational
structure, leaders are capable of increasing team members’ self-
efficacy and control over
their work environment. When team members are empowered to
make job-related decisions
on their own, they need to possess adequate information to
ensure that the decisions are
reasonable and justifiable given the decision contexts. As a
result, they are more likely to
share knowledge with one another before and during the
decision process. Therefore,
empowering leadership is the enzyme that stimulates and
nurtures the occurrence of
knowledge sharing.
Arnold et al. (2000) show that empowering leadership has five
dimensions:
1. leading by example, referring to a set of behaviors that show

the leader’s commitment to
his or her own work as well as the work of his/her team
members;
2. coaching, referring to a set of behaviors that educate team
members and help them to
become self-reliant;
3. participative decision making, referring to a leader’s use of
team members’ information
and input in making decisions;
4. showing concern, referring to a collection of behaviors that
demonstrate a general regard
for team members’ well-being; and
5. informing, referring to the leader’s dissemination of company
wide information such as
mission and philosophy as well as other importation
information.
An empowering leader who possesses these attributes will be
seen as a supportive leader
who provides guidance to followers, treats them fairly, and
recognizes the value of their
input. Given that team members expect to receive fair
recognition by an empowering leader
for their contribution of ideas and information, they are likely

to be motivated to share their
unique knowledge with others (Srivastava and Bartol, 2006).
All of the five dimensions of empowering leadership contribute
to knowledge sharing. First,
an empowering leader can set an example for subordinates by
sharing his or her own
knowledge first, which signifies his or her support for team-
wide knowledge sharing.
Second, the coaching behavior of an empowering leader
includes teaching team members
how to effectively communicate with one another and
encouraging them to collaboratively
solve problems, thereby providing opportunities for them to
share their knowledge (Arnold
et al., 2000). Third, when a leader advocates participative
decision making, team members
have more opportunities to voice their opinions and provide
suggestions (Locke et al., 1997).
VOL. 15 NO. 2 2011
Under such leadership, team members are likely to see
themselves as an important part of

the decision process and more motivated to share their
knowledge. Fourth, employees
might have concerns when sharing knowledge with peers
because their social status in the
organization is often related to their unique knowledge. An
empowering leader is able to
identify and alleviate such concerns, thus removing barriers to
knowledge sharing. Finally,
Srivastava and Bartol (2006) suggest that informing motivates a
search for solutions both
inside and outside a team and a greater collaborative attempt to
help one another through
knowledge sharing. Overall, the preceding points suggest that
empowering leadership will
strongly influence individuals’ attitudes toward knowledge
sharing and increase the extent of
their knowledge sharing behavior:
H3. Empowering leadership has a positive influence on
knowledge sharing attitude.
H4. Empowering leadership has a positive influence on
knowledge sharing behavior.
2.3 Attitude
Based on theory of reasoned action, attitude is defined as an
individual’s positive or negative

feelings about performing knowledge sharing (Fishbein and
Ajzen, 1975). The theory of
reasoned action posits that attitude determines behavioral
intention, which in turn
determines behavior. Numerous empirical studies have
confirmed the significant influence
of attitude on intention (e.g., Bock et al., 2005). In this
research, the authors decide to
investigate the direct relationship between attitude and behavior
because they are
interested in explaining individuals’ actual knowledge sharing
behavior rather than
predicting their future behavior. Behavioral intention, as a
predictor of actual behavior, has
limitations (Venkatesh et al., 2008). It has been criticized that
there exists an
intention-behavior gap (Sheeran, 2002). Therefore, the authors
propose that the more
favorable individuals’ attitude toward knowledge sharing, the
more likely they will share
knowledge with others. Hsu et al. (2007) state that the biggest
challenge in knowledge
sharing is the willingness (attitude) of the individual. That is,
negative attitude tends to

decrease the likelihood of knowledge sharing. Overall, it is
suggested that there is a positive
relationship between knowledge sharing attitude and behavior:
H5. Knowledge sharing attitude has a positive influence on
behavior.
2.4 Control variables
Team members’ knowledge sharing behavior is likely to be
influenced by their demographic
characteristics. Gratton et al. (2007) find that at large
companies in Europe and the USA
many failures in collaboration and knowledge sharing result
from subgroups that have
emerged within teams based on age and gender. Miller and
Karakowsky (2005) show that
team members’ gender has a significant impact on their
feedback seeking from others.
Therefore, the authors control for the influence of age and
gender on knowledge sharing
behavior.
3. Method
3.1 Measurement development
Measures for the four constructs were developed based on prior

research. Consistent with
Bock et al.’s (2005) study, team climate is modeled as a second
order formative construct
consisting of three first order reflective constructs: cohesion,
trust, and innovation. Cohesion
was measured using three items adapted from (Xue et al.,
2004/2005). Innovativeness was
adaptedfromBock et al.
(2005),measuredbytwoitems.Trustwasmeasuredusingthreeitems
adapted from Langfred (2004). Following Arnold et al. (2000),
empowering leadership is
modeledasasecondorderreflectiveconstructcomprisingfivefirstor
derreflectiveconstructs:
lead by example, participative decision making, coaching,
informing, and showing concern.
Themeasurementitemsfor
theseconstructswereadaptedfromArnold et al. (2000).Thescale
for attitude includes three items adapted from Bock et al.
(2005). The scale for knowledge
sharing behavior was adapted from Hsu et al. (2007). Except
attitude items that are evaluated
byafive-pointsemanticscale,alloftheitemswereevaluatedbyafive-
pointLikertscalewhere1

represents ‘‘strongly disagree’’ and 5 represents ‘‘strongly
agree.’’ The Appendix, Figure A1
shows the measurement items. In addition, age was measured as
a ratio variable and gender
as a categorical variable with male coded as 1 and female as 2.
3.2 Procedure
An online survey was developed to measurethe theoretical
constructs. A total of 650 students,
undergraduates and graduates, who were taking business courses
at a large university in the
USA were invited to take the survey. These students were
recruited from both management
and management information system (MIS) courses where
teamwork assignments were
major course requirements. These team assignments range from
case studies, requirement
analyses, and essay writing, to project design and development.
Students were grouped into
teams to work on several projects. In each team, a team leader
was chosen to act as a
coordinator between the instructor and team members. Team

leaders were responsible for
reporting their team members’ activities and involvement and
workload allocation within
teams.Peerevaluationwas used toassess eachmember’s
performance onthe team projects.
To simulate the real work environment, team leaders’
evaluations were given a higher weight
when aggregating all of the peer evaluation scores. The survey
was administered at the end of
the semester. Participation in the survey was completely
voluntary and anonymous. Extra
course credits were employed as an incentive for completing the
survey. A total of 434
completed surveys were collected resulting in a response rate of
66.8 percent.
Among the 434 respondents, 219 are male (50.5 percent), and
215 are female students
(49.5 percent). Their average age is 25.81, ranging from 18 to
63 (SD ¼ 7:78). Most of them
have some work experiences ranging from 0 year to 35 years
(mean ¼ 5:54 and SD ¼ 7:33).
3.3 Data analysis
The authors used partial least squares (PLS) to validate the
measurements and test the

hypotheses. PLS employs a component-based approach for
model estimation and is not
highly demanding on sample size and residual distribution
(Chin, 1998). It is best suited for
testing complex structural models as it avoids two problems:
inadmissible solutions and
factor indeterminacy (Fornell and Bookstein, 1982). Both
reflective and formative constructs
can be estimated by PLS (Chin, 1998). Hence, this method was
chosen to accommodate the
formative second-order construct (team climate) since
covariance-based SEM techniques
do not allow formative constructs to be estimated easily.
4. Results
4.1 Measurement validation
The reliability of the measurements was evaluated using
Cronbach’s alpha and the
composite reliability scores. As Table I shows, the reliability
scores of all of the constructs are
considered adequate as they exceed the recommended cutoff of
0.70 (Nunnally, 1978).
The convergent and discriminant validity of the measurements
were confirmed by four tests.

First, as Table I shows, the square root of the average variance
extracted (AVE) of each
construct is much larger than all cross-correlations between the
construct and other
Table I Construct reliability, AVE, and correlations
Construct Cronbach’s alpha Composite reliability AVE 1 2 3 4 5
1. Team climate – – – –
2. Empowering leadership 0.98 0.98 0.71 0.49 0.84
3. Attitude 0.89 0.93 0.82 0.44 0.38 0.91
4. KM behavior 0.83 0.89 0.74 0.36 0.36 0.41 0.86
5. Gender – – – 0.07 0.13 0.10 0.05 –
6. Age – – – 0.10 0.03 0.12 0.10 0.10
Note: The diagonal elements (in italics) are square roots of
AVE
VOL. 15 NO. 2 2011
constructs (Chin, 1998). Second, all AVEs are well above 0.50,
which suggests that the
principal constructs capture much higher construct-related
variance than error variance
(Hair et al., 1998). Third, the correlations among all of the
constructs are well below the 0.90
threshold, suggesting that the constructs are distinct from each
other (Bagozzi et al., 1991).

Fourth, PLS analysis shows that each item’s loading on its
underlying construct is above the
recommended 0.70 level (Chin et al., 2003) and significant at
the 0.01 level (Table II). Jointly,
these tests suggest adequate convergent and discriminant
validity of the measurements.
The authors paid particular attention to the two second-order
constructs – team climate and
empowering leadership. Since empowering leadership is a
reflective second-order
construct, its validity is indicated by the path weights of its five
first-order constructs
(Jarvis et al., 2003). As Figure 2 shows, the path weights of
lead by example, participative
decision making, coaching, informing, and showing concern are
0.92, 0.92, 0.94, 0.93, and
0.91, respectively (p , 0:01), suggesting that they are
significantly determined by the
underlying higher order construct.
Traditional methods assessing construct validity and reliability
are inappropriate for
formative constructs whose causal direction flows from
measures to constructs

(Diamantopoulos and Winklhofer, 2001; Jarvis et al., 2003).
Following the formative
measures assessment guidelines recommended by Petter et al.
(2007), the authors
evaluated team climate’s construct validity and reliability. First,
the PLS analysis shows that
all of the three first-order constructs of team climate have
significant weights (Figure 2),
providing evidence for construct validity (Diamantopoulos and
Winklhofer, 2001).
Specifically, the weights for affiliation, trust, and innovation
are 0.41, 0.46, and 0.25,
respectively (p , 0:01). Second, to assess multicollinearity, the
authors computed latent
Table II Factor loadings
Construct Item Mean SD Loading
Innovation 1 3.85 0.92 0.89*
2 3.57 0.86 0.87*
Cohesion 1 3.86 0.81 0.85*
2 3.76 0.97 0.88*
3 4.17 0.79 0.84*
Trust 1 3.80 1.01 0.93*
2 3.79 1.11 0.93*
3 3.80 1.04 0.93*

Lead by example 1 3.90 0.96 0.88*
2 4.09 0.91 0.94*
3 4.03 0.96 0.91*
4 4.07 0.91 0.94*
Participative DM 1 4.03 0.89 0.93*
2 4.10 0.87 0.95*
3 4.14 0.83 0.95*
Coaching 1 3.89 0.93 0.92*
2 3.97 0.87 0.93*
3 3.93 0.92 0.92*
Informing 1 3.94 0.92 0.92*
2 3.90 0.93 0.95*
3 3.80 0.95 0.91*
4 3.94 0.90 0.90*
Showing concern 1 3.73 0.95 0.86*
2 3.89 0.88 0.92*
3 4.11 0.86 0.84*
4 3.91 0.88 0.90*
Attitude 1 4.06 0.71 0.90*
2 3.97 0.69 0.93*
3 4.01 0.66 0.89*
KS behavior 1 3.75 0.85 0.88*
2 3.42 0.90 0.82*
3 3.73 0.81 0.87*
Note: *p , 0:01

variable scores for each first-order team climate component and
then tested its variance
inflation factor (VIF)[1]. The VIFs for cohesion, trust, and
innovation are 1.62, 1.58, and 1.36,
respectively. It is recommended that the VIF statistic for
formative measures should not
exceed 3.3 (Diamantopoulos and Siguaw, 2006). All of the VIFs
are under 3.3, which suggest
that the formative measure is reliable.
Since all of the constructs are measured by single-source self-
report data, common method
variance (CMV) may bias the construct relationships (Podsakoff
et al., 2003). The authors
conducted the Harmon’s one factor test (Podsakoff and Organ,
1986) to evaluate whether
CMV is a serious concern. All of the measurement items were
entered into a factor analysis
using the Varimax rotation. No single dominant factor emerged
from the analysis. Ten
components were extracted and their explained variance ranged
from 2.4 percent to 36.4
percent, indicating that common method variance is unlikely to
be serious.

4.2 Model testing
The structural model testing results are shown in Figure 2.
Team climate is found to
significantly affect knowledge sharing attitude (b ¼ :34, p ,
0:01), as is empowering
leadership (b ¼ 0:21, p , 0:01). These two factors account for 23
percent of variance in
knowledge sharing attitude, thus supporting H1 and H3. Team
climate is found to
significantly affect knowledge sharing behavior (b ¼ 0:14; p ,
0:05), thus supporting H2.
The link between empowering leadership and knowledge sharing
behavior is significant
(b ¼ 0:18, p , 0:01), providing support to H4. Knowledge
sharing attitude is also found to
have a significant positive influence on behavior (b ¼ 0:28, p ,
0:01), thus supporting H5.
About 24 percent of variance in knowledge sharing behavior can
be explained by the three
determinants. The two control variables, age and gender, do not
have a significant effect on
5. Discussion
This study examines the impact of team related factors on
individuals’ attitude and
knowledge sharing behavior, which makes important theoretical

and practical contributions
to team based research. Our results highlight the importance of
both team climate and
empowering leadership on individuals’ knowledge sharing
attitude and behavior. Previous
Figure 2 Model testing results
Team
Climate
Empowering
Leadership
Attitude
KS
behavior
Control variables
Gender Age
* p < 0.05; ** p < 0.01; ns = non-significant
0.34**
0.21**
0.18**
0.28**
0.14*

0.23 0.24 Participative
Decision Making
Coaching
Trust
Cohesion
Innovation
Showing
Concern
Lead by
Example
Informing
0.41**
0.46**
0.25**
0.92**
0.92**
0.94**
0.93**
0.91**
ns ns

Notes:
Second-order
constructs
VOL. 15 NO. 2 2011
research only studied the impact of one of these two important
team factors independent of
the other. By putting them together, this study integrates two
important perspectives – the
social environment of the team and the value of the team leader.
In addition, our research shows that the impact of these two
factors are complementary –
they can work together to cultivate individuals’ knowledge
sharing attitude and lead to more
knowledge sharing behavior. The authors find that both team
climate and empowering
leadership have two pathways to influence knowledge sharing –
besides the indirect
influence via attitude, they also have a direct impact. This
suggests that their effects are both
internal and external. Internally, they sway individuals’
subjective attitude which in turn

increases knowledge sharing. Externally, social pressures from
team climate or facilitating
conditions from empowering leadership can be created to
directly encourage knowledge
sharing.
These findings extend the existing literature on knowledge
sharing. For example, Bock et al.
(2005) found that team climate affects knowledge sharing by
influencing attitude, but they
focused on intention to share knowledge and did not examine
Srivastava and Bartol (2006) uncovered the direct relationship
between empowering
leadership and knowledge sharing behavior, but they did not
examine the mediating role of
attitude. Thus, our research integrates discrete findings of prior
research and should deepen
our understanding of the dynamics of knowledge sharing within
teams.
Practically, this study draws special attention to team design in
organizations. In order to
promote knowledge sharing, besides considering other relevant
organizational and
individual factors, managers need to cultivate a nurturing team

httphum.sagepub.comHuman Relations httphum.sagep.docx

httphum.sagepub.comHuman Relations httphum.sagep.docx

Recommended

Recommended

More Related Content

Similar to httphum.sagepub.comHuman Relations httphum.sagep.docx

Similar to httphum.sagepub.comHuman Relations httphum.sagep.docx (20)

More from AASTHA76

More from AASTHA76 (20)

Recently uploaded

Recently uploaded (20)

httphum.sagepub.comHuman Relations httphum.sagep.docx