Content uploaded by Ana Ortiz de Guinea
Author content
All content in this area was uploaded by Ana Ortiz de Guinea on Oct 27, 2015
Content may be subject to copyright.
A Meta-Analysis of the
Virtual Teams Literature
INDUSTRIAL RELATIONS
CENTRE
SCHOOL OF POLICY STUDIES
Ana Ortiz de Guinea*
Queen’s School of Business, Queen’s University
Jane Webster**
Queen’s School of Business, Queen’s University
Sandy Staples***
Queen’s School of Business, Queen’s University
Contact Information:
* Email: AdeArana@business.queensu.ca
Tel.: (613) 533-8316
Fax: (613) 533-2325
** Email: jwebster@business.queensu.ca
Tel.: (613) 533-3163
Fax: (613) 533-2325
*** Email: sstaples@business.queensu.ca
Tel.: (613) 533-2314
Fax: (613) 533-2325
We would like to thank Derek Chapman, Peter Dacin, and Sandy Hershcovis for
their advice on the meta-analytic analyses. This paper was financially assisted
by the Industrial Relations Centre, Queen’s University, and by Social Sciences
and Humanities Research Council grants to Sandy Staples and Jane Webster.
The opinions expressed in this document do not necessarily reflect those of the
Industrial Relations Centre, Queen’s University.
Paper presented at the Symposium on High Performance Professional Teams
Industrial Relations Centre, School of Policy Studies, Queen’s University
Kingston, Ontario, Canada
October 12, 2005
1
A Meta-Analysis of the Virtual Teams Literature
Virtual teams are becoming more prevalent in organizations (Horvath and Tobin 1999)
due to interorganizational alliances, globalization, outsourcing and alternative work
arrangements such as job sharing and telecommuting (Saunders 2000). These teams are
most often constructed because organizations require skills, local knowledge,
experience, resources and expertise from employees who are geographically-
distributed. Information technology can support team members’ activities by saving
travel costs, enabling expertise to be captured where it is located and speeding up
team processes. For instance, one study estimated the savings from collaborating
electronically rather than face-to-face: over 123 days, the project team saved between
10 and 23 days using information technology tools (May and Carter 2001). Although
virtual teams offer benefits, they also create many challenges to organizations desiring
to create high performing teams. For example, coordinating interdependent activities
over distance and developing mutual understanding and team spirit are difficulties
virtual teams often experience.
Developments in information technology have allowed the use of virtual teams in
organizations to grow rapidly over the last decade. Not surprisingly then, much of the
research on virtual team effectiveness is relatively new and we are still trying to
understand the unique challenges of virtual teams versus traditional (i.e., co-located
teams). There exist several qualitative, narrative reviews of the existing literature that
have furthered our understanding of virtual teams (e.g., Dube and Pare 2004; Furst,
Blackburn and Rosen 1999; Martins, Gilson and Maynard 2004; Pinsonneault and
Caya, 2005; Powell, Piccoli and Ives 2004; Saunders 2000); however, no quantitative
reviews, or meta-analyses, exist yet of this literature. Our study aims to provide this. In
contrast to a narrative review, the purpose of this paper is to synthesize virtual team
research statistically by conducting a quantitative review of this literature. Our meta-
analytic review adds to the existing narrative reviews by quantifying the strength of
the relationships between team inputs, processes, and outcomes associated with virtual
teams (Schmidt and Hunter 2001). Specifically, the first goal of this study is to
summarize the relationships among variables typically examined in virtual team
studies via meta-analytic techniques. These estimates can be used to assist researchers
in building theory, to guide future research, and to guide practice. Our second goal,
again using meta-analytic techniques, is to explore if selected moderator variables can
help explain differences among existing empirical findings. A third goal of this study is
to test a meta-analytic structural equation model of virtual team effectiveness.
The paper is organized as follows. First, we outline a general model of virtual team
functioning. The methodology used to identify over 200 empirical articles on virtual
teams and to meta-analyze appropriate studies is described next, followed by the
empirical results. We conclude by suggesting implications for research and practice.
A Model of Virtual Team Effectiveness
We define virtual teams as groups of individuals who work together in different
locations (i.e. are geographically dispersed), work at interdependent tasks, share
responsibility for outcomes and rely on technology for much of their communication
(Cohen and Gibson 2003). Virtual teams were originally conceptualized as “fully”
virtual, in contrast to face-to-face (“traditional” or co-located) teams (Griffith and
Neale 2001). Researchers have since viewed virtuality as a continuum, arguing that
many teams in organizations today are characterized by dimensions of virtuality (e.g.,
Griffith, Sawyer and Neale 2003).
We developed a model of team effectiveness to guide our review of the empirical
studies of virtual teams, starting from the existing body of knowledge on traditional
teams and groups.
1
(Some researchers argue that research on virtual teams typically
does not build on the past 50 years of traditional team research and is consequently
rediscovering old truths (Cohen and Gibson 2003) which we strove to avoid). In
contrast to virtual teams, traditional teams are co-located and have easy access to both
face-to-face and electronic communication. These teams have been formally studied for
more than half a century, resulting in thousands of studies and a huge body of
literature (Guzzo and Shea 1992). Fortunately, several reviews of the knowledge in this
field already exist (e.g. Bettenhausen 1991; Cohen 1994; Cohen and Bailey 1997; Guzzo
and Shea 1992; Holland, Gaston and Gomes 2000; Marks, Mathieu and Zaccaro 2001;
Yeatts and Hyten 1998) and we refer to these in the following discussion.
As illustrated in Figure 1, an inputs-process-outputs model based on McGrath’s (1984)
perspective is the dominant way in the literature of thinking about group performance
(Guzzo and Shea 1992). Inputs refer to things that group members bring to the group,
as well as the context in which the group operates. Main inputs are task design, group
characteristics, organizational context and supervisory behaviors. Process refers to
“members’ interdependent acts that convert inputs to outcomes through cognitive,
verbal, and behavioral activities directed toward organizing taskwork to achieve
collective goals” (Marks et al. 2001, p. 357). These interdependent acts among group
members are often categorized into either expressive (interpersonal) or instrumental
(work) interactions (Guzzo and Shea 1992). Expressive interactions are affective or
socio-emotional, such as showing antagonism or affection, being dependent or asking
for support. Instrumental interactions are task-related, including seeking information
and making suggestions. Outputs refer to team effectiveness, and include things such as
performance, the satisfaction and attitudes of group members, and their behavioral
outcomes.
2
1
While much of the academic research uses the term group to describe a set of interdependent individuals who
have common goals, collective responsibility for outcomes and a shared identity, the term team can be used
interchangeably, and has been done so by Guzzo and Shea (1992). Consistent with this approach, we use the terms
team and group interchangeably.
3
Figure 1: General Model of Virtual Team Effectiveness *
Group Characteristics
* Team Structure
(e.g. amount of face-to-face
interaction, degree of
virtualness)
* Type of Team and
Members
(e.g. team size,
individual characteristics,
group efficacy, group
beliefs. IT skills)
* Stage of Team
Development
Nature of Task
Task design (e.g. signifi-
cance, skill variety), task
demands (e.g. conceptual
versus behavioral), task
duration, team autonomy,
interdependence
Organizational Context
e.g., organizational culture,
rewards, IT resources and
training
Supervisory Behaviors
e.g. transactional versus
transformational, degree of
supervision (directive ver-
sus self-managed teams),
expectations, communication
through IT
Effectiveness
Performance Outcomes
e.g., quality, productivity,
learning
Attitudinal Outcomes
e.g., job satisfaction, trust
Behavioral Outcomes
e.g., turnover, absenteeism
Expressive and
Instrumental
Interactions
(e.g., cohesion,
communication,
coordination,
effort, sharing of
expertise, work
procedures
Inputs Outputs
Process
a
c
b
Adapted from traditional team frameworks (e.g. Cohen 1994; Cohen and Bailey 1997; Marks, et al. 2001;
Yeatts and Hyten 1998) and extended (variables in italics) based on research on individuals working
virtually (through telecommuting or distributed work)
The major components and relationships in the model are as follows (space does not
allow us to describe the detailed reasoning for each relationship; readers are directed to
the review papers listed above for more information). Link a describes the influence of
inputs on process. For instance, although different classification schemes have been
developed to describe team tasks, McGrath’s (1984) well-known task typology ranges
from conceptual to behavioral tasks (that is, from generating ideas, to choosing between
alternatives, to negotiating conflicts, to executing work) (Stewart and Barrick 2000). In
terms of supervisory behaviors, teams may range from hierarchically-led to self-led,
with self-leadership being “the extent to which teams have the freedom and authority
to lead themselves independent of external supervision” (Stewart and Barrick 2000,
p. 139).
Group research has also suggested link b, from process to outputs. For instance,
coordination involves working together without duplicating or wasting effort, as well
as working with team spirit and energy (Cohen 1994). Previous research supports a
direct relationship between such process variables (e.g. coordination) and attitudinal
outcomes (i.e. job and team satisfaction, organizational commitment and trust), and an
indirect effect on withdrawal behaviors (through employee satisfaction).
Link c has two aspects. First, it summarizes the direct effects of inputs on outputs (e.g.
the effects of supervisory behaviors on outcomes). In the context of self-managing
teams, Cohen (1994) proposes that encouraging supervisory behaviors, such as setting
expectations, help a team to lead itself. Bandura’s (1977) social learning theory is the
basis for this: it suggests that cognitive evaluations combine with environmental
conditions to determine human behavior. A leader’s role in a self-managed team is to
facilitate the development of self-controls so that team members can lead themselves.
Cohen (1994) suggests that supervisors’ actions that support self-leadership should
have a direct impact on performance, employee trust in management, organizational
commitment and attendance.
Most group research focuses on one-way effects (that is, from inputs to outputs in our
model) but researchers are calling for more attention to be paid to feedback loops. This
is the second aspect of link c, the reciprocal relationship shown from outputs to inputs
(Marks et al. 2001). For instance, a key team characteristic is group efficacy, or the
shared belief of the group that it can be effective. Empirical research supports the direct
effect of group efficacy on both performance and satisfaction; however, group efficacy is
also directly affected by prior performance; hence, the reciprocal relationship (Cohen
1994).
We extended this traditional team model to virtual teams by examining other research
on virtual work, specifically research on individuals working virtually. We drew on
Pinsonneault and Boisvert’s (2001) review of individual virtual work to suggest several
variables of particular importance to virtual teams that should be included in the
4
model. First, being able to effectively use IT is crucial for both team members and their
supervisors since communications rely heavily on electronic media. Second, the amount
of face-to-face contact possible through periodic meetings has important implications
for virtual team functioning (degree of virtualness). Third, virtual teams need both IT
resources and training, and virtual team training. Finally, effective supervisory
communication and modeling of appropriate behaviors through IT are particularly
helpful for virtual team functioning. Given these findings, we have extended the
traditional models of group effectiveness to include variables of particular importance
to virtual teams.
Research Questions
We were limited in the number of relationships outlined in Figure 1 that we could
study. Since studying virtual teams is a relatively new research area, there are
relationships that have not yet been studied in a virtual team context. For this meta-
analysis, we examined the variables and relationships outlined in Figure 1 that have
been most examined in empirical research on virtual teams. That is, our first research
question is:
Research Question 1: What are the relationships between virtual team inputs, processes,
and outputs?
We also examined the generalizability of these findings. Specifically, by introducing
moderating variables into the analysis, meta-analysis techniques can determine if
between-study differences are partially due to the different conditions in the moderator
variables (Schmidt and Hunter 2001). Research design and sample characteristics are
typical moderators examined in meta-analysis work (Franke 2001). Accordingly, two
possible moderators that varied frequently across the primary studies were investigated
in this study: type of team member (student versus employee) and level of analysis
(individual versus group). It is important to see if student team functioning differs from
employee team functioning and whether team functioning differs based on individual
versus group perceptions (Gully et al. 2002). This will help us understand the
generalizability of different types of studies. For example, some argue that that it is not
possible to generalize from research conducted with individuals to make predictions
about team functioning. Relations studied at the individual level of analysis may be
weaker than at the team level because individual perceptions may not be shared (Gully
et al. 2002). Therefore, our second research question is:
Research Question 2: To what extent do team member type and/or level of analysis
moderate the relationships between virtual team inputs, processes, and outputs?
5
Review of Empirical Research on Virtual Teams
For the meta-analysis, we examined over 200 empirical articles on virtual teams
identified by Staples and Webster (2005b). They found these articles by a variety of
methods: searching the ProQuest and PsycINFO electronic databases (using the
keywords of virtual, distributed, dispersed, global or remote combined with team or group),
searching conference proceedings (including the International Conference on Information
Systems, the Americas Conference on Information Systems, the Association for Computing
Machinery, the Hawaii International Conference on System Sciences, the Computer Supported
Cooperative Work conference, the Society for Industrial and Organizational Psychology, and
the Academy of Management), obtaining working papers through personal contacts,
searching two Web sites focused on virtual team research (VoNet and
virtualteamsresearch.org), reviewing five collections of articles on virtual teamwork
(Beyerlein, Johnson and Beyerlein 2001; Gibson and Cohen 2003; Godar and Ferris 2004;
Hinds and Kiesler 2002; Pauleen 2004) and an annotated bibliography (Sessa, Hansen,
Prestridge and Kossler 1999), conducting a general Web search using the meta-search
engine Google, and examining references in reviewed articles to identify new articles
leads. Of these papers, some report on the same dataset and some report on more than
one dataset (only unique datasets were included in the meta-analysis). Studies were
excluded because they only contained data that (a) were qualitative, (b) involved
relationships beyond the ones selected for this meta-analysis, and/or (c) did not contain
all of the data needed.
2
Fifty studies contained data relevant to the variables included in
the meta-analysis (studies included in the meta-analysis are indicated with *’s in the
References).
Coding the Data
We meta-analyzed correlations between variables. To do so, one author coded all of the
papers based on input from the other two authors. However, we needed to make a
series of coding decisions. For example, an examination of the items used in the
primary studies revealed that variable labels were not always consistent with the
measure content. In these instances, at least two authors discussed the content of the
measure and reconciled any discrepancies. Another issue concerned the necessity of
collapsing many narrow measures into larger categories. For example, some studies
reported group satisfaction while others reported outcome satisfaction; we coded these
different aspects of satisfaction into one overall outcome variable called “Satisfaction”.
Further, some studies captured the degree of virtualness (a continuous variable from
less to more virtual) while others compared face-to-face (0) with distributed teams (1);
in the latter case, we transformed the data to point-biserial correlations which were
combined with the correlations calculated for the former case.
6
2
Many of the articles did not report all of the data needed to conduct a meta-analysis. We contacted all authors
who provided partial data in their papers. In some cases, they were able to send us the required information, but in
many cases they were unable to do so.
We also made coding decisions for studies that presented multiple correlations for the
same constructs. Since multiple correlations in a single study create the potential for
overweighting studies that contain non-independent data, this can distort the results of a
meta-analysis (Hunter and Schmidt 1990). To avoid this, we used a conservative
approach of choosing one correlation/study to minimize the problems associated with
non-independent coefficients. When making these decisions, we used the data that most
closely represent teams in organizations. For example, some studies measured variables
at more than one point in time. When data were reported for multiple points in time, we
used the data that most closely represent employee teams (e.g., if correlations were
presented after 1 month and 3 months for the same variables, we chose the correlation
for 3 months). Also, if there was more than one face-to-face condition in the experiment
(e.g., face-to-face with and without computer support), we chose the condition with
computer support, as this most closely resembles teams in organizations. When more
than one correlation was reported for the same construct at the same point of time for
the same study (that is, multiple dimensions of the same construct), we averaged the
correlations into one correlation in order to capture the whole dimensionality of the
construct (Lipsey and Wilson 2001).
Analyses
The coding process resulted in a total of 161 coefficients representing 34 relationships (for
Tables 1 and 2) and a larger set of coefficients in order to conduct the path analysis (in
LISREL we needed to include a complete matrix of correlations, such as the correlations
between outcome variables). Meta-analysis provides an estimate of the “true” population
effect size (here, correlations) for a given relationship via a series of steps. We cumulated
correlation coefficients into average effect sizes using the meta-analytic techniques
outlined by Lipsey and Wilson (2001). Appendix A outlines the analyses performed to
calculate the effect sizes. These analyses were performed for relationships presented in
two or more studies.
Computation of effect sizes. Anumber of studies did not report correlations but we were
able to obtain some of them either through contacting the studies’ authors or by
converting other reported data (such as t-tests, F-tests, etc.) to correlations using the
procedures described by Lipsey and Wilson (2001).
We calculated overall mean effect sizes (ρ) by following steps (1) through (7) in Appendix
A. This involved weighting correlations (1) by their sample sizes (2), correcting both the
correlations (3) and weightings (4) for unreliability, performing Fisher’s Z-transformations
(5), calculating the overall mean correlations (6), and then un-transforming the overall
mean correlations (7). These were tested for significance using a Z-test (11). (In addition,
we also calculated mean effect sizes (r
xy
) without correcting for unreliability).
Moderator detection and estimation. The second research question was addressed by
determining whether the studies that were included in the meta-analysis came from
7
heterogeneous populations (i.e., whether moderator variables were likely present or
not), and if so, whether the moderator variables accounted for a significant amount of
the residual variance. The Q statistic (Hedges and Olkin 1985) was used as a test of
effect size homogeneity, with a statistically significant Q indicating heterogeneity of
effect sizes (see formula 12 in Appendix A). For statistically significant Q’s, we
separated the results by the moderator variables (both students/employees and
individuals/groups) to determine if the correlations differed across these moderators.
Path Analyses. Structural equation modeling (SEM) was conducted using the
correlation matrix obtained from the meta-analyses. This analysis allowed us to test if
the relationships between the various predictor variables and team effectiveness were
mediated by team processes. This approach is consistent with Viswesvaran and Ones’
(1995) approach and is used by Chapman et al. (in press) and Hershcovis, Turner and
Barling (in press). When conducting meta-analytic SEM, choices must be made as to the
appropriate treatment of empty cells and sample size. One of the cells in the inter-
correlation matrix was empty (the relationship between quality and trust). We followed
Viswesvaran and Ones’ (1995) suggestion of using existing results from related areas for
the missing value in the matrix.
3
The harmonic mean of the studies was used to
determine the sample size (Viswesvaran and Ones 1995).
To rule out the possibility that alternative models fit the data as well or better, we also
assessed three variants of the model. In combination, the models were designed to
assess whether the relationship between predictor variables and team effectiveness were
fully mediated by team processes, or whether direct or partially mediated models
provided a better fit to the data (see Baron & Kenny 1986). The Direct model was a fully
independent model estimating direct paths from the inputs to the outputs without any
mediating processes. In the Fully Mediated model the relationships between the inputs
and outputs were fully mediated by team processes, with no direct paths estimated
from inputs to outputs. The Partially Mediated model was the hypothesized model
outlined in Figure 1 in which inputs affected both processes and outputs, and processes
influenced outcomes. The overall fit for each model was assessed using the Goodness of
Fit Index (GFI), the adjusted Goodness of Fit Index (AGFI), the Comparative Fit Index
(CFI), and the Root Mean Square Error of Approximation (RMSEA).
Results
Table 1 presents the meta-analytic results for the relations between virtual team inputs,
processes, and outcomes. As can be seen, many of the relationships diagrammed in
8
3
We considered two options for estimating the relationship between quality and trust; using the same relationship as
for performance and trust that we calculated from our meta-analyses below (
ρ
= .31) or drawing on related research.
Concerning related research, we were able to find several studies, one examining the relationship between quality and
trust in consumer services (r = .67: Sharma and Patterson 1999) and another in e-commerce ( _ = .50: McKnight,
Choudhury and Kacmar 2002). We took the conservative route and chose the lowest value (.31) to conduct the analyses.
9
Team Inputs
1
Degree of Virtualness
=
Disposition to Trust
Age
Virtual Team
Experience
Transformational
Leadership
k
6
3
2
*
*
442
226
136
-.21
.38
.00
.00
-.19
a
-.41
a
n
Sign.
of
Z
r
xy
ρ
k
7
2
3
*
671
136
333
-.38
.03
-.07
-.07
-.39
a
.03
nZ
r
xy
ρ
k
3
2
264
229
-.01
.00
.00
-.02
nZ
r
xy
ρ
k
7
318
.39
.40
nZ
r
xy
ρ
*
Notes
1. Link a in Figure 1
=Lower virtualness = face-to-face; higher virtualness = distributed teams
k = number of studies
r
xy
= mean weighted coefficient
ρ
= coefficient corrected for the unreliability of predictor and criterion
* z-test: correlation is different from 0 (p < .05)
ª significant Q statistic (homogeneity test)
Table 1: Meta-Analysis of Virtual Team Imputs, Processes and Outcomes
Virtual Team Processes
Cohesion Communication Shared Identity Conflict
continued . . .
10
Team Inputs &
Processes
Inputs
1
:
Degree of Virtualness
Processes
2
:
Cohesion
Disposition to Trust
Age
Communication
Shared Identity
Conflict
Virtual Team
Experience
Transformational
Leadership
k
20
2
3
3
2
*
*
*
*
1024
136
296
207
136
.08
.25
.48
.29
.30
.02
.02
.08
a
.52
a
.26
nZ
r
xy
ρ
k
25
2
5
2
3
2
7
2
3
*
1961
136
550
355
215
136
622
240
96
.16
.02
.04
.06
.29 .31
.19
.18
.21
.17
.04
-.25
-.29
.17
a
.06
a
.23
a
.20
a
.02
nZ
r
xy
ρ
k
14
5
2
2
3
2
1404
136
651
187
136
333
.06
.32
.32
.04
.54
.73
.13
.02
nZ
r
xy
ρ
k
4
5
2
3
2
281
761
287
226
285
nZ
r
xy
ρ
*
*
*
*
Notes
1. Link c on Figure 1
2. Link b on Figure 1
Table 1: cont’d.
Virtual Team Outcomes
Quality Productivity Satisfaction Trust
*
*
*
*
*
.06
a
-.07
-.07
a
.58
a
.32
a
.56
a
.15
a
.56
a
.83
a
.14
a
*
*
*
*
*
.45
.29
.51
.14
11
Figure 1 could not be tested because they had not been examined frequently in virtual
team research. Relevant data were available on five input variables, four process
variables, and four output variables. In the Table, we see correlations ranging from no
relationship between age and cohesion (
ρ
= .00) to a strong relationship between
cohesion and satisfaction (
ρ
= .83, p < .05). Specifically, the first part of Table 1 examines
the relationship between inputs and processes (link a). Significant negative associations
were found between degree of virtualness and cohesion and communication. Significant
positive associations were found between disposition to trust and cohesion and between
degree of virtualness and conflict. This indicates that as team structure increases in
virtuality, it has a negative impact on team processes, whereas if team members have
higher disposition to trust (an individual characteristic), team cohesion will be higher.
The next section of Table 1 examines the direct effect of the input variables on the team
output variables (link c in Figure 1). All statistically significant associations were found
to be positive for these relationships. Specifically, degree of virtualness was found to be
positively associated with quality, productivity and satisfaction and not significantly
related to trust. This implies increased virtuality has a positive effect on team
effectiveness. Disposition to trust was positively associated with quality, satisfaction and
trust, indicating that teams with team members who have a high willingness to trust will
be more effective. The amount of experience team members had working in virtual
conditions was also positively associated with trust. Transformational leadership
behavior was positively associated with productivity and satisfaction, indicating that this
type of leadership is of benefit to teams.
The last section of Table 1 reports the relationships between the four process variables
and four outcome variables (link b in Figure 1). Cohesion and communication were
positively associated with all four outcome variables, indicating the importance of
having a cohesive team and high levels of communication. Shared identity was
positively related to productivity (we did not have data available to examine the
relationship with the other outcome variables). Conflict was negatively related to
productivity, indicating that high levels of conflict are detrimental to the team’s
effectiveness.
Moderator Analysis
The possibility of moderator variables affecting the relationships was examined via the Q
statistic (see Table 1). Heterogeneity was indicated in 18 of the 34 relationships. For
correlations with a significant Q, we further analyzed these relationships by splitting the
data on the two moderators of individual/group and student/employee. As shown in
Tables 2 and 3, not all relationships could be examined since in some pairings, for
example, all of the studies were conducted with only individuals (i.e., no group level
analysis) or with only students (i.e., no employee subjects). Analysis of the level of
analysis moderation (Table 2) was possible for ten relationships. Although the strengths
of the paths varied somewhat for the two groups, consistent results were found at both
12
Team Inputs
1
Degree of Virtualness
Disposition to Trust
Age
Virtual Team
Experience
Transformational
Leadership
k
2
I
G
I
G
I
G
I
G
I
G
4
*
166
276
-.35
-.12
-.07
a
-.35
n
Z
r
xy
ρ
k
nZ
r
xy
ρ
k
nZ
r
xy
ρ
k
nZ
r
xy
ρ
Notes
I: individual
G: group
k = number of studies
r
xy
= mean weighted coefficient
ρ
= coefficient corrected for the unreliability of predictor and criterion
* z-test: correlation is different from 0 (p < .05)
ª significant Q statistic (homogeneity test)
τ
Those with significant Q’s in Table 1 but without reported moderator analyses here were all individual or all group studies.
Table 2: Moderator Analyses: Individual level vs. Group level
τ
Virtual Team Processes
Cohesion Communication Shared Identity Conflict
continued . . .
13
Team Inputs &
Processes
Inputs
1
:
Degree of Virtualness
Processes
2
:
Cohesion
Disposition to Trust
Age
Communication
Shared Identity
Conflict
Virtual Team
Experience
Transformational
Leadership
k
6
2
1
14
*
*
*
400
136
71
600
.33
.57
.26
.27
-.02
.34
a
.64
a
nZ
r
xy
ρ
k
9
3
4
1
1
*
*
1149
712
333
213
197
43
.09
.29
.21
.21
.13
.33
.16
.39
.09
a
.30
a
.02
a
.22
a
.25
a
nZ
r
xy
ρ
k
7
7
1
1
1050
354
158
29
.19
-.33
.20
.59
.17
.17
nZ
r
xy
ρ
k
2
2
4
1
229
52
695
nZ
r
xy
ρ
*
*
*
Table 2: cont’d.
Virtual Team Outcomes
Quality Productivity Satisfaction Trust
*
*
*
-.34
a
.03
.03
-.49
a
.56
a
.62
*
*
*
-.49
66
.69
.73
.52
I
G
I
G
I
G
I
G
I
G
I
G
I
G
I
G
I
G
16
14
k
nZ
r
xy
ρ
Team Inputs
1
Degree of Virtualness
Disposition to Trust
Age
Virtual Team
Experience
Transformational
Leadership
k
S
E
S
E
S
E
S
E
S
E
2
1
*
*
*
*
136
302
369
90
.38
-.52
-.26
.39
.41
a
-.52
a
-.27
a
.41
n
Z
r
xy
ρ
k
n
95%
r
xy
ρ
k
n
r
xy
ρ
Notes
τ
Those with significant Q’s in Table 1 but without reported moderator analyses here were all student or all employee studies.
Table 3: Moderator Analyses: Students vs. Employees
τ
Virtual Team Processes
Cohesion Communication Shared Identity Conflict
4
3
continued . . .
95%
15
Team Inputs &
Processes
Inputs
1
:
Degree of Virtualness
Processes
2
:
Cohesion
Disposition to Trust
Age
Communication
Shared Identity
Conflict
Virtual Team
Experience
Transformational
Leadership
k
18
3
*
1000
88
.11
-.20
.11
a
n
95%
r
xy
ρ
k
19
4
3
*
1160
701
253
369
.26
-.01
.33
.12
.26
a
-.01
a
-.20
.36
a
.12
a
nZ
r
xy
ρ
k
12
2
1049
355
136
.38
.41
-.05 -.05
197
.03
.16
.03
a
n
r
xy
ρ
k
2
2
2
1
1
1
46
90
88
.48
.56
.52
-.04
.52
-.05
.60
235
136
197
n
r
xy
ρ
Table 3: cont’d.
Virtual Team Outcomes
Quality Productivity Satisfaction Trust
*
*
.17
a
-.36
-.41
a
.54
a
-.02
*
*
-.01
S
E
S
E
S
E
S
E
S
E
S
E
S
E
S
E
S
E
6
2
1
95%
9
levels of analysis for the following five relationships: degree of virtualness to cohesion
and productivity; disposition to trust and trust; cohesion to quality; and
communication to productivity. These results are consistent with the full sample
analysis (i.e., same significant direction of association).
Differences were found between the two levels of analysis for the remaining five
relationships. The significant positive associations between degree of virtualness and
quality and transformational leadership and satisfaction were only found for at the
individual level of analysis and the positive associations between shared identity and
productivity was only found for at the group level of analysis. Astrong negative
relationship was found at the group level of analysis for the relationship between degree
of virtualness and trust, while the relationship was non-significant for individual-level
studies. The relationship between degree of virtualness and satisfaction was different for
the two groups: significantly positive at the individual level and significantly negative at
the group level.
Table 3 presents the moderator analysis for the student versus employee subjects. We
had data to test ten relationships. Four of the relationships found the results consistent
for both groups and with the full sample analysis: degree of virtualness to
communication, disposition to trust to cohesion, cohesion to trust, and communication to
productivity. The significant positive relationships between degree of virtualness and
quality and productivity and between communication and satisfaction and trust were
only found in the student samples (the employee groups were non-significant). The
positive relationship between degree of virtualness and satisfaction was only found in
the employee studies. A relatively large negative relationship was found between degree
of virtualness and trust for the student studies (this relationship was non-significant for
employees).
Path Analysis
Amatrix of correlations between all variables was obtained from the meta-analyses and
used to test the virtual team effectiveness model. The hypothesized Partially Mediated
model (Figure 1) provides the best fit to the data (GFI = .98, AGFI = .79, CFI = .97, RMSEA
= .04); in comparison, the Direct and Fully Mediated models fit less well. Figure 2 provides
the path coefficients for the partially mediated model.
16
As can be seen on Figure 2, the path coefficients of the relationships between degree of
virtualness and the process and output variables are relatively consistent with the rho
values obtained from the meta-analyses (see Table 1), again indicating negative
relationships with the process variables and positive relationships with the output
variables. However, the pattern of path coefficients regarding disposition to trust are
somewhat different than the rho values. This can be explained by examining the direct,
indirect and total effects. When the relationship between disposition to trust and
satisfaction is examined in the structural model, it becomes apparent that the effect is
basically fully mediated through cohesion. The relationships between disposition to
trust and communication and productivity are also different in the structural model
than in the simple correlations. The path model indicates a negative direct relationship
between disposition to trust and productivity, positive relationships with both process
variables, and positive relationships between the two process variables and
productivity. The total effect of disposition to trust on productivity equals the sum of
the direct effect and the two indirect effects which totals 0.01, consistent with the rho
value in Table 1.
17
Figure 2: Path Model of Virtual Team Effectiveness
Degree of
Virtualness
Cohesion
Quality
Productivity
Satisfaction
Trust
Communication
Disposition
to Trust
Inputs Outputs
Process
Cohesion: -.19*
Commun.: -.39*
Quality: .27*; Product.: .35*; Satisf.: .23*; Trust: -.04
Quality: .01; Product.: -.16*; Satisf.: .02; Trust: .47*
Quality: .49*
Product.: .21*
Satisf.: .87*
Trust: .42*
Quality: .26*
Product.: .35*
Satisf.: -.02
Trust: -.12*
Cohesion: .35*
Commun.: .29*
Discussions and Conclusions
Our first goal for this study was to quantitatively examine the relationships among
virtual team measures that had been empirically examined in two or more studies. We
found large relationships (Cohen 1988) for: degree of virtualness with conflict;
disposition to trust with cohesion and trust; transformational leadership with
satisfaction; and cohesion with quality, satisfaction, and trust. However, we found that
very few inputs had been frequently examined in a virtual team setting. We were only
able to examine three individual characteristics (disposition to trust, age and virtual
team experience), one leadership behavior (transformational leadership), and one team
structure variable (degree of virtualness). This shows that there is a significant need for
research that studies organizational context and nature of task variables, as well as
other group characteristics and supervisory behaviors. Examining the relationships of
these variables to team processes and team effectiveness would be valuable pursuits
for future research.
For the input variables that we could include in our meta-analysis, disposition to trust
was found to be important, in that it was significantly associated with the most of the
process and outcome variables. This implies that having team members that have an
initial high disposition to trust is important for team functioning. This is consistent
with many suggestions in the literature about the importance of trust in a virtual
setting. Due to the geographic dispersion, the monitoring of behavior is very difficult,
if not impossible. Trust that others will deliver on their commitments and share
knowledge as appropriate is essential for a virtual team to function well. Presumably,
the trust that one has in a past team would influence one’s willingness to trust the
members of a future team. Therefore, even if an organization was not particularly
concerned about team trust as an important outcome for an existing team, low team
trust could reduce the effectiveness of future teams that the team members join.
However, results of the path analysis demonstrate that although high disposition to
trust positively influences cohesion and communication (and thereby team
effectiveness), its direct effects are negative. It may be that those with a high
propensity to trust are less likely to monitor their team members’ activities, with the
result that team productivity is lower.
The degree of virtualness was found to be negatively associated with the process
variables, indicating that more virtuality hurts team processes. This is consistent with
work on computer-mediated communication (e.g., Baltes et al. 2002; Bordia 1997)
which indicates that working through electronic means creates process losses. These
are due to lower social presence, fewer social cues in the communication (i.e., the
medium is leaner), more effort/time required to communicate electronically, higher
social loafing, and less inhibited communications contributing to conflict and
misunderstandings. However, our meta-analytic results indicate that the total effect of
virtualness on team effectiveness (quality, productivity and satisfaction) is positive,
18
indicating overall positive benefits of virtual teams, regardless of the losses in the team
processes. The overall positive effect could be due to team diversity benefits such as
access to more resources and perspectives. The results imply these benefits outweigh
the losses in the team processes. We feel this is an important finding since much of the
literature implies that virtuality in teams will cause problems and negatively impact
team effectiveness. Given that virtual teams are a way of life for many organizations
today, knowing that virtuality can contribute to team effectiveness is encouraging.
However, the moderation analyses raise some questions about the generalizability of
these conclusions. For example, the positive relationships between virtualness and
quality and productivity were only found in the student group, whereas the positive
relationship with satisfaction was only found in the employee groups. It may be that
the diversity found in many student teams (such as students from other campuses and
countries) helps to build quality and productivity, but that students still experience in-
group/out-group biases in distributed teams, leading to lower satisfaction. Arelatively
strong negative relationship was also found between virtualness and trust in the
student groups and at the group level of analysis. It may be that employees’ shared
culture, institutional trust, and longer expectations for future interactions with team
members all help to build team trust, with the result that employees are just as trusting
of their distributed as their local colleagues.
For the moderators, we found potentially important differences for approximately half
the relationships that could be examined (i.e., the nature of the relationship varied
between the two sub-samples being examined). Because of the small number of
studies, these findings should be interpreted with caution. However, they suggest
many potential areas for future research.
Limitations and Future Research
There are many opportunities for future research related to conspicuous gaps in
empirical research (when compared to the model of virtual team effectiveness
diagrammed in Figure 1). For example, as we reviewed the papers, we found that few
fully describe the organizational context studied. If this is the most important group of
variables for performance as some suggest (e.g. Cohen 1994), then it is important to
direct more attention to context. For instance, most studies do not examine cross-
cultural issues for teams made up of participants from multiple countries, despite
these issues being especially important for virtual work (Pinsonneault and Boisvert
2001). In addition, meta-analytical work is only as meaningful as the primary studies
from which it is derived. For example, many of the groups were short-term and most
studies were cross-sectional in nature (a relatively weak design since team
development may evolve over time). Another limitation of meta-analyses is the “file
drawer” problem, or unpublished studies that are not included in the meta-analysis
(Lipsey and Wilson 2001); however, we included working papers and studies from
conferences. Another limitation of meta-analyses arises from the lack of consistency
concerning measures and reporting standards. Because of the large variety of measures
19
used, this necessitates collapsing many narrow measures into larger categories,
possibly reducing the variance in some measures. Afurther weakness of many meta-
analyses is the small number of studies examining the same relationships, and this
meta-analysis was no exception. More disturbing for this meta-analysis was the lack
of reported data needed to conduct the analyses (effect sizes, sample sizes, etc.)
4
: we
had to exclude many papers because of a lack of appropriate data. We hope that this
meta-analysis will help ensure that future researchers will be more aware of the
necessity for full reporting of relationships between measures.
Many studies examine the impact of a specific type of IT through experiments with
student subjects. If we are going to be able to have any confidence that the results for
short-term student teams can be generalized to the field, research should examine
realistic bundles of communications technologies. Most research has focused on e-mail
and the World Wide Web as communication tools, but managers also need to know
about groupware and knowledge management tools for virtual work. More generally,
they need to know which tools are most appropriate for which tasks (Bélanger et al.
2002; Woerner, Orlikowski and Yates 2004). We encourage future research in this area
to balance studies of student teams with those of existing virtual teams in
organizations.
Although we found a considerable body of knowledge on virtual teams, the
multifaceted nature of these teams and the inconsistent reporting of empirical results
make understanding what leads to high performance very complex. There are many
valuable opportunities for more research. Our meta-analysis should help researchers
focus on areas with the greatest need for additional understanding. We agree with
Saunders (2000) and Martins et al. (2004) that there is still a lot to learn about virtual
teams. We hope that our review and suggestions will help researchers meet this need
in order that organizations can more fully realize the potential of virtual teams.
20
4
Some academic associations and journals have standard reporting requirements that require authors to include the
data needed for others to conduct meta-analyses. However, virtual team research occurs in multiple fields (information
systems, organizational behavior, computer science, etc.) with differing reporting expectations.
21
Appendix A. Formulas Used in Computing Effect Sizes (Lipsey and Wilson 2001)
Statistic Formula
22
5
Studies included in the meta-analysis are marked with *s.
References
5
* Aiken, M., and Vanjani, M. “A Comparison of Synchronous and Virtual Legislative
Session Teams Faced with an Idea Generation Task,” Information & Management
(33), 1997, pp. 25-31.
* Alavi, M., Wheeler, B. C., and Valacich, J. S. “Using IT to Reengineer Business
Education: An Exploratory Investigation of Collaborative Telelearning,” MIS
Quarterly (19:3), 1995, pp. 293-312.
* Alge, B. J., Ballinger, G. A., Green, S. G., and Till, G. “The Effects of Dependence and
Trust on the Decision to Electronically Monitor Subordinates,” paper presented at
the 2002 National Academy of Management Meetings, Denver, CO, 2002.
* Anderson, W., and Hiltz, S. “Culturally Heterogeneous vs. Culturally Homogeneous
Groups in Distributed Group Support Systems: Effects on Group Process and
Consensus,” in Proceedings of the 34th Hawaii International Conference on System
Sciences, Hawaii, USA, 2001.
* Andres, H. P. “A Comparison of Face-to-Face and Virtual Software Development
Teams,” Team Performance Management: An International Journal (8:1/2), 2002,
pp. 39-48.
* Baker, G. “The Effects of Synchronous Collaborative Technologies on Decision
Making: A Study of Virtual Teams,” Information Resources Management Journal (15:4),
2002, pp. 79-93.
Baltes, B.B., Dickson, M.W., Sherman, M.P., Bauer, C.C., and LaGanke, J.S. Computer-
mediated communication and group decision making: a meta-analysis.
Organizational Behavior and Human Decision Processes, (87:1), 2002, pp. 156-179.
Bandura, A. “Self-Efficacy: Toward a Unifying Theory of Behavioral change,”
Psychological Review (84), 1977, pp. 191-215.
* Bhappu, A. D. and Giambatista, R. C. “Using Communication Technology to Enhance
Knowledge Exchange and Decision-Making in Diverse Groups,” paper
presented at the Annual Meeting of the Academy of Management, 2004.
23
Baron, R. M. and Kenny, D. A. “The Moderator-Mediator Variable Distinction in Social
Psychological Research: Conceptual, Strategic, and Statistical Considerations,”
Journal of Personality and Social Psychology (51), 1986, pp. 1173-1182.
Bélanger, F., Watson-Manheim, M. B., and Jordan, D. H. “Aligning IS Research and
Practice: A Research Agenda for Virtual Work,” Information Resources Management
Journal (15:3), July-September 2002, pp. 48-70.
* Benbunan-Fich, R., Hiltz, S. R., and Turoff, M. “A Comparative Content Analysis of
Face-to-Face vs. Asynchronous Group Decision Making,” Decision Support Systems
(34), 2002, pp. 457-469.
Bettenhausen, K. L. “Five Years of Group Research: What We Have Learned and What
Needs to be Addressed,” Journal of Management (17:2) 1991, pp. 345-381.
Beyerlein, M., Johnson, D., and Beyerlein, S. Virtual Teams, Elsevier Science Ltd.,
Oxford, UK, 2001.
Bordia, P. Face-to-face versus computer-mediated communication: a synthesis of the
experimental literature. The Journal of Business Communication, 34(1), 99-120, 1997.
* Bos, N., Olson, J., Gergle, D., Olson, G. and Wright, Z. “Effects of Four Computer-
Mediated Communications Channels on Trust Development,” in Proceedings of CHI
2002 (4:1), April 20-25, 2002, Minneapolis, MN, pp. 135-140.
* Burgoon, J. K., Bonito, J. A., Ramirez, A., Jr., Dunbar, N. E., Kam, K., and Fischer, J.
“Testing the Interactivity Principle: Effects of Mediation, Propinquity, and Verbal
and Nonverbal Modalities in Interpersonal Interaction,” Journal of Communication,
2002 (September), pp. 657-677.
* Cappel, J. J., and Windsor, J. C. “Ethical Decision Making: A Comparison of
Computer-Supported and Face-to-Face Group,” Journal of Business Ethics (28), 2000,
pp. 95-107.
Chapman, D. et al. (in press). “Applicant Attraction to Organizations and Job Choice: A
Meta-Analytic Review of the Correlates of Recruiting Outcomes,” Journal of Applied
Psychology.
* Chidambaram, L., and Jones, B. “Impact of Communication Medium and Computer
Support on Group Perceptions and Performance: AComparison of Face-to-face and
Dispersed Meetings,” MIS Quarterly (17), 1993, pp. 465-491.
24
Cohen, J. Statistical Power Analysis for the Behavioral Sciences (2nd Ed.) Hillsdale, NJ:
Erlbaum, 1988.
Cohen, S. G. “Designing Effective Self-Managing Work Teams,” in Advances in
Interdisciplinary Studies of Work Teams, Volume 1, Series of Self-Managed Work Teams,
M. M. Beyerlein, D. A. Johnson and S. T. Beyerlein (eds.), JAI Press, Greenwich,
Connecticut, 1994, pp. 67-102.
Cohen, S. G., and Bailey, D. E. “What Makes Teams Work?: Group Effectiveness
Research from the Shop Floor to the Executive Suite,” Journal of Management (23:3),
1997, pp. 239-290.
Cohen, S. G., and Gibson, C. B. “Putting the Team Back in Virtual Teams,” paper
presented at the 18th Annual Conference of the Society for Industrial/Organizational
Psychology, Orlando, FL, 2003.
* Cramton, C. D., and Webber, S. S. “Modeling the Impact of Geographic Dispersion on
Work Teams,” Working Paper, George Mason University, Washington, DC, 1999.
* Cummings, J. N. (in press). “Leading Groups From a Distance: How to Mitigate
Consequences of Geographic Dispersion,” in Leadership at a Distance, S. Weisband
and L. Atwater (eds.).
Dube, L., and Pare, G. “The Multi-Faceted Nature of Virtual Teams,” in Virtual Teams:
Projects, Protocols and Processes, David Pauleen (ed.), Idea Group Publishing,
Hershey, PA, 2004, pp. 1-39.
Franke, G.R. “Applications of Meta-Analysis for Marketing and Public Policy: A
Review,” Journal of Public Policy & Marketing (20:2), 2001, pp. 186-200.
Furst, S., Blackburn, R., and Rosen, B. “Virtual Team Effectiveness: A Proposed
Research Agenda,” Information Systems Journal (9), 1999, pp. 249-269.
* Gallupe, R. B., and McKeen, J.D. “Enhancing Computer-Mediated Communication:
An Experimental Investigation into the Use of a Group Decision Support System
for Face-to-face Versus Remote Meetings,” Information & Management (18) 1990,
pp. 1-13.
* Galvin, J. E., Ahuja, M. K., and Agarwal, R. “Dispositional and Situational Influences
on Trust Behaviour in Virtual Teams,” Working Paper, Indiana University,
Indianapolis, IN, 2000.
25
Gibson, C., and Cohen, S. “Is There Anybody Out There? The Impact of Virtuality on
Individual Psychological States and Team Enabling Conditions,” paper presented at
the Annual Meeting of the Academy of Management, Denver, CO, 2002.
Gibson, C. B. and Cohen, S. G. (Eds.) Virtual Teams that Work, San Fransicso, CA:
Jossey-Bass, 2003.
Godar, S. and Ferris, P. Virtual & Collaborative Teams: Process, Technologies, & Practice,
Hershey, PA: Idea Group Publishing, 2004.
*González, M. G., Burke, M. J., Santuzzi, A M., and Bradley, J. C. “The Impact of Group
Process Variables on the Effectiveness of Distance Collaboration Groups,”
Computers in Human Behavior (19), 2003, pp. 629-648.
Griffith, T. L., and Neale, M. A. Information Processing in Traditional, Hybrid, and
Virtual Teams: From Nascent Knowledge to Transactive Memory,” in Research in
Organizational Behavior, R. I. Sutton and B. M. Staw (eds.), JAI Press, Stamford,
Conn., 2001, pp. 379-421.
Griffith, T. L., Sawyer, J. E., and Neale, M.A. “Virtualness and Knowledge in Teams:
Managing the Love Triangle in Organizations, Individuals, and Information
Technology,” MIS Quarterly (27), 2003, pp. 265-287.
Gully, S., M., Incalcaterra, K. A., Joshi, A., and Beaubien, J. M. “A Meta-analysis of
Team-Efficacy, Potency, and Performance: Interdependence and Level of Analysis as
Moderators of Observed Relationships,” Journal of Applied Psychology (87), 2002,
pp. 819-832.
Guzzo, R. A., and Shea, G. P. “Group Performance and Intergroup Relations in
Organizations,” in Handbook of Industrial and Organizational Psychology (2nd ed.,
Volume 3), M. D. Dunnette and L. M. Hough (eds.), Consulting Psychologists Press,
Palo Alto, CA, 1992, pp. 269-313.
Hedges, L. V. and Olkin, I. Statistical Methods for Meta-Analysis, Orlando, FL: Academic
Press, 1985.
* Hedlund, J., Ilgen, D., and Hollenbeck, J. “Decision Accuracy in Computer-mediated
versus Face-to-face Decision-Making Teams,” Organizational Behavior and Human
Decision Processes (76), 1998, pp. 30-47.
Hershcovis, S., M., Turner, N., and Barling, J. (in press). “Predicting Workplace
Aggression: AMeta-Analysis,” Journal of Applied Psychology.
26
* Hertel, G. Konradt, U., and Orlikowski, B. “Managing Distance by Interdependence:
Goal Setting, Task Interdependence, and Team-Based Rewards in Virtual Teams,”
European Journal of Work and Organizational Psychology (13), 2004, pp.1-28.
Hinds, P., and Kiesler, S. Distributed Work, London, UK: MIT Press, 2002.
* Hinds, P. J., and Mortensen, M. “Understanding Conflict in Geographically
Distributed Teams: An Empirical Investigation,” (under review) 2004.
Holland, S., Gaston, K. and Gomes, J. “Critical Success Factors for Cross-Functional
Teamwork in New Product Development,” International Journal of Management
Reviews (2:3), 2000, pp. 231-259.
Horvath, L., and Tobin, T. J. “Twenty-First Century Teamwork: Defining Competencies
for Virtual Teams,” paper presented at the 1999 Academy of Management Meeting,
Chicago, IL, 1999.
Huang, W. W., Wei, K.-K., Watson, R. T., and Tan, B. C. Y. “Supporting Virtual Team-
Building with a GSS: An Empirical Investigation,” Decision Support Systems (34:4),
2002, pp. 359-367.
Hunter, J. E. and Schmidt, F. L. Methods of Meta-Analysis: Correcting Error and Bias in
Research Findings, New York: The Free Press, 1990.
* Jarvenpaa, S. L., and Leidner, D. E. “Communication and Trust in Global Virtual
Teams,” Organization Science (10:6), 1999, pp. 791-815.
*Jarvenpaa, S. L., Shaw, T. R. and Staples, D. S. “Toward Contextualized Theories of
Trust: The Role of Trust in Global Virtual Teams,” Information Systems Research
(15:3), 2004, pp. 250-267.
* Kanawattanachai, P., and Yoo, Y. “Dynamic Nature of Trust in Virtual Teams,” Journal
of Strategic Information Systems (11), 2002, pp. 187-213.
* Kirkman, B. L., Rosen, B., Tesluk, P. E., and Gibson, C. B. “The Impact of Team
Empowerment on Virtual Team Performance: The Moderating Role of Face-to-Face
Interaction,” Academy of Management Journal (47:2), 2004, pp. 175-192.
Kline, R. B. Principles and Practice of Structural Equation Modeling, New York: Guilford
Press, 1998.
Lipsey, M. W. and Wilson, D. B. Practical Meta-Analysis, Thousand Oaks, CA: Sage
Publications, 2001.
27
* Majchrzak, A., Malhotra, A. and John, R. “Perceived Individual Collaboration Know-
How Development Through Information Technology-Enabled Contextualization:
Evidence from Distributed Teams,” Information Systems Research (16:1), 2005,
pp. 9-27.
Marks, M. A., Mathieu, J. E., and Zaccaro, S. J. “A Temporally Based Framework and
Taxonomy of Team Processes,” Academy of Management Review (26:3), 2001,
pp. 356-276.
Martins, L. L., Gilson, L. L., and Maynard, M. T. “Virtual Teams: What Do We Know
and Where Do We Go From Here?” Journal of Management (30:6), 2004, pp. 805-835.
May, A., and Carter, C. “A Case Study of Virtual Team Working in the European
Automotive Industry,” International Journal of Industrial Ergonomics (27:3), 2001,
pp. 171-186.
May, A., and Carter, C. “A Case Study of Virtual Team Working in the European
Automotive Industry,” International Journal of Industrial Ergonomics (27:3), 2001,
pp. 171-186.
* McDonough, E., Kahn, K., and Barczak, G. “An Investigation of the Use of Global,
Virtual, and Collocated New Product Development Teams,” Journal of Production
Innovation Management (18:2), 2001, pp, 110-120.
McGrath, J. E. Groups: Interaction and Performance, Englewood Cliffs, NJ: Prentice-Hall,
1984.
McKnight, D. H., Choudhury, V., and Kacmar, C. “Developing and Validating Trust
measures for e-Commerce: An Integrative Typology,” Information Systems Research
(13:3), 2002, pp. 334-359.
* McLeod, P., Baron, R., Marti, M., and Yoon, K. “The Eyes Have It: Minority Influence
in Face-to-face and Computer-Mediated Group Discussion,” Journal of Applied
Psychology (82:5), 1997, pp. 706-718.
* Morris, S. A., Marshall, T. E., and Rainer, R. K. Jr. “Impact of User Satisfaction and
Trust on Virtual Team Members,” Information Resources Management Journal (15:2),
2002, pp. 22-30.
* Mortensen, M., and Hinds, P. “Conflict and Shared Identity in Geographically
Distributed Teams” International Journal of Conflict Management (12:3), 2001,
pp. 212-238.
28
* Ocker, R. “The Relationship Between Interaction, Group Development, and Outcome:
A Study of Virtual Communication,” in Proceedings of the 34th Hawaii International
Conference on System Sciences, Hawaii, USA, 2001.
* Ocker, R. J. “The Mediating Effect of Group Development on Satisfaction in a Virtual
and Mixed-Mode Environment,” in Proceedings of the 35th Hawaii International
Conference on System Sciences, 2002.
* Ocker, R., Hiltz, S., Turoff, M., and Fjermstad, J. “The Effects of Distributed Group
Support and Process Structuring on Software Requirements of Development Teams:
Results on Creativity and Quality,” Journal of Management Information Systems (12:3),
1995-1996, pp. 127-153.
* Olson, J., Olson, G., and Meader, D. “Face-to-face Group Work Compared to Remote
Group Work With and Without Video,” in Video-mediated Communication, K. Finn, A.
Sellen, and S. Wilbur, (eds.), Lawrence Erlbaum Associates, Mahwah, USA, 1997,
pp. 157-172.
Pauleen, D. Virtual Teams: Projects, Protocols and Processes, Hershey, PA: Idea Group
Publishing, 2004.
* Pearce, C., Yoo, Y., and Alavi, M. “Leadership, social work, and virtual teams: The
relative influence of vertical versus shared leadership in the nonprofit sector,” in
Improving Leadership in Nonprofit Organizations, R. E. Riggio and S. Smith Orr (eds.)
Jossey Bass, San Francisco, CA, 2004, pp. 180-203.
Pinsonneault, A., and Boisvert, M. “The Impacts of Telecommuting on Organizations
and Individuals: A Review of the Literature,” in Telecommuting and Virtual Offices:
Issues & Opportunities, N. J. Johnson (ed.), Idea Group Publishing, Hershey, PA,
2001, pp. 163-185.
Pinsonneault, A. & Caya, O. “Virtual Teams: What we know, what we don’t know.”
International Journal of e-Collaboration, 1(3), 2005, pp. 1-16.
* Potter, R. E., and Balthazard, P. A. “Understanding Human Interaction and
Performance in the Virtual Team,” Journal of Information Technology Theory and
Application (4:1), 2002, pp. 1-23.
Powell, A., Piccoli, G. and Ives, B. “Virtual Teams: A Review of Current Literature and
Directions for Future Research,” Data Base for Advances in Information Systems (35:1),
2004, pp. 6-36.
* Rock, K. W., Pratt, M. G., and Northcraft, G. B. “The Effects of Faultlines and
Communication Media on Identification in Virtual Teams.” Paper presented at the
2002 Academy of Management Meeting, Denver, CO, 2002.
Saunders, C. S. “Virtual Teams: Piecing Together the Puzzle,” in Framing the Domains of
IT Management: Projecting the Future Through the Past, R. W. Zmud (ed.), Pinnaflex
Education Resources, Inc., Cincinnati, OH, 2000, pp. 29-50.
* Savicki, V., Kelley, M., and Lingenfelter, D. “Gender and Group Composition in Small
Task Groups Using Computer-Mediated Communication,” Computers in Human
Behavior (12:2), 1996, pp. 209-224.
Schmidt, F. L. and Hunter, J. E. “Meta-Analysis,” in Handbook of Industrial, Work and
Organizational Psychology, (Vol. 1), N. Anderson, D. Ones, H. Kepir Sinangil and C.
Viswesvaran (eds.), Sage Publications: London, UK, 2001, pp. 51-70.
* Schmidt, J., Montoya-Weiss, M., and Massey, A. “New Product Development
Decision-making Effectiveness: Comparing Individuals, Face-to-face Teams, and
Virtual Teams,” Decision Sciences (32:4), 2001, pp. 575-600.
Sessa, V. I., Hansen, M. C., Prestridge, S., and Kossler, M. E. Geographically Dispersed
Teams: An Annotated Bibliography, Center for Creative Leadership, Greensboro, NC,
1999.
Sharma, N. and Patterson, P. G. “The Impact of Communication Effectiveness and
Service Quality on Relationship Commitment in Consumer, Professional Services,”
The Journal of Services Marketing (13:2), 1999, pp. 151-170.
* Staples, D. S., and Webster, J. “Best Practices for Virtual Teams: Exploring the
Effectiveness of Team Members via Their Self-Efficacy for Teamwork.” Working
paper, Queen’s University, Canada, 2005a.
Staples, D. S., and Webster, J. “Comparing Virtual Teams to Traditional Teams: An
Identification of New Research Opportunities”, Working paper, Queen’s University,
Canada, 2005b.
Stewart, G. L., and Barrick, M. R. “Team Structure and Performance: Assessing the
Mediating Role of Intrateam Process and the Moderating Role of Task Type,” The
Academy of Management Journal (43:2), April 2000, pp. 135-148.
* Straus, S. G. “Technology, Group Process, and Group Outcomes: Testing the
Connections in Computer-Mediated and Face-to-face Groups,” Human-Computer
Interaction (12), 1997, pp. 227-266.
29
30
* Straus, S. G. and McGrath, J. E. “Does the Medium Matter? The Interaction of Task
Type and Technology on Group Performance and Member Reactions,” Journal of
Applied Psychology (79:1), 1994, pp. 87-97.
* Tan, B. C. Y., Wei, K-K., Sia, C-L., and Raman, K. S. “A Partial Test of he Task-Medium
Fit Proposition in a Group Support System Environment,” ACM Transactions on
Computer-Human Interaction (6:1), 1999, pp. 47-66.
Tett, R. and Meyer, J. P. “Job Satisfaction, Organizational Commitment, Turnover
Intention, and Turnover: Path Analysis Based on Meta-Analysis Findings,”
Personnel Psychology (46), 1993, pp. 259-293.
* Thompson, L., and Coovert, M. “Stepping Up to the Challenge: A Critical
Examination of Face-to-face and Computer-Mediated Team Decision Making,”
Group Dynamics: Theory, Research, and Practice (6:1), 2002, pp. 52-64.
* Thompson, L. F., and Coovert, M. D. “Teamwork Online: The Effects of Computer
Conferencing on Perceived Confusion, Satisfaction, and Postdiscussion Accuracy,”
Group Dynamics: Theory, Research, and Practice (7:2), 2003, pp. 135-151.
* Townsend, A. M., Demarie, S. M., and Hendrickson, A. R. “Desktop Video
Conferencing in Virtual Workgroups: Anticipation, System Evaluation and
Performance, “ Information Systems Journal (11), 2001, pp. 213-225.
Viswesvaran, C. and Ones, D. S. “Theory Testing: Combining Psychometric Meta-
Analysis and Structural Equations Modeling,” Personnel Psychology (48), 1995,
pp. 865-885.
* Warkentin, M., Sayeed, L., and Hightower, R. “Virtual Teams Versus Face-to-Face
Teams: An Exploratory Study of a Web-Based Conference System,” Decision Sciences
(28:4), 1997, pp. 975-996.
* Walther, J. B. and Bunz, U. (in press). “The Rules of Virtual Groups: Trust, Liking, and
Performance in Computer-Mediated Communication,” Journal of Communication.
* Webster, J., and Wong, W. “Group Identity, Trust, and Communication in Naturally-
Occurring Project Teams,” paper presented at the 18th Annual Conference of the
Society for Industrial/Organizational Psychology, Orlando, FL, 2003.
* Weisband, S. “Maintaining Awareness in Distributed Team Collaboration:
Implications for Leadership and Performance,” in Distributed Work, P. Hinds, and
S. Kiesler, (eds.), MIT Press, London, UK, 2002, pp. 311-333.
Woerner, S. L., Orlikowski, W. J. and Yates, J. “The Media Toolbox: Combining Media
in Organizational Communication,” Paper presented at the 2004 Academy of
Management Meetings, New Orleans, LA, 2004.
Yeatts, C., and Hyten, D. E. High-Performing Self-Managed Work Teams: A Comparison of
Theory to Practice, Sage Publications, Thousand Oaks, CA, 1998.
Yoo, Y., and Alavi, M. “Emergent leadership in virtual teams: What do emergent
leaders do?” Information and Organization, (14:1), 2004, pp. 27-58.
* Ziegert, J. C., Klein, K. J., and Xiao, Y. “Virtual Leaders: The Impact of Distance and
Leadership on Teams,” paper presented at the 2004 Society for Industrial
Organizational Psychology, Chicago, IL, 2004.
31
