Understanding Conflict in Virtual Teams: An Experimental Investigation using Content Analysis

Abstract

Virtual teams are temporally and geographically dispersed groups, which may have members from varied cultures and backgrounds. Such diversity may cause intra-group conflicts in virtual teams. We analyzed the contents of the transcripts of GSS-based virtual teams and identified the conflict episodes and the approaches followed to resolve intra-group conflicts. The conflict episodes that occurred in the early phases of decision-making were separated from those taking place in the choice phase. The results revealed that conflicts in the choice phase of decision-making were detrimental to global virtual teams. We also found that groups following an integrative conflict resolution style had better performance than those following other conflict resolution approaches, such as a distributive style. While the results contribute towards the understanding of conflict in groups, the area warrants further research.

Full text

Page 1

Understanding Conflict in Virtual Teams: An Experimental Investigation using Content
Analysis
Souren Paul
Southern Illinois University at Carbondale, paul@cba.siu.edu
Priya Seetharaman
Indian Institute of Management, Calcutta, priyas@iimcal.ac.in
Imad Samarah
Southern Illinois University at Carbondale, imad@siu.edu
Peter Mykytyn, Jr.
Southern Illinois University at Carbondale, mykytyn@cba.siu.edu



Abstract
Virtual teams are temporally and geographically
dispersed groups, which may have members from varied
cultures and backgrounds. Such diversity may cause
intra-group conflicts in virtual teams. We analyzed the
contents of the transcripts of GSS-based virtual teams and
identified the conflict episodes and the approaches
followed to resolve intra-group conflicts. The conflict
episodes that occurred in the early phases of decision-
making were separated from those taking place in the
choice phase. The results revealed that conflicts in the
choice phase of decision-making were detrimental to
global virtual teams. We also found that groups
following an integrative conflict resolution style had
better performance than those following other conflict
resolution approaches, such as a distributive style. While
the results contribute towards the understanding of
conflict in groups, the area warrants further research.

1. Introduction

Organizations use teams and groups for decision
making to address a variety of issues. Traditionally, teams
operated in face-to-face (F2F) settings, but with the
proliferation of advanced computer technology, virtual
teams have become commonplace. Jarvenpaa and Leidner
[12] define a virtual team as a temporary, culturally
diverse, geographically dispersed, and electronically
communicating work group. In today’s world, it is not
uncommon for virtual team members to be of different
cultural and functional backgrounds, which may be a
source of intra-group conflict in virtual teams. Moreover,
the geographic dispersion, the temporal separation and the
low richness of the media used may also lead to intra-
group conflict. While current literature has examined the
various antecedents of virtual team conflict, there has not
been ample research on conflict in virtual teams and its
impact on team performance.
The willingness and ability of group members to deal
with conflict is likely to have a bearing on the
performance of the group. Only recently have researchers
begun to address the role that conflict management styles
play in tasks performed by virtual teams[22][25].
The major issues examined in prior research on virtual
teams deal with dynamics of virtual group work and the
characteristics of the formed groups and how these two
factors influence performance measures. Typically,
researchers collected their data through surveys and/or
controlled experiments, and data analyses were mostly
conducted using standard
techniques. In this paper, we utilize “content analysis” to
investigate whether discussion scripts generated by global
virtual teams reveal some insight into the nature of virtual
teams’ effort and performance.
Content analysis is a “research technique for the
objective, semantic, and quantitative description of the
manifest content of communication” [2, pg. 74] . Holsti
[10] defines content analysis as being an inference
technique that is based on an objective and semantic
identification of specified attributes of messages. Thus,
one objective of content analysis is to transform the
qualitative content of textual communications scripts into
quantitative and reliable measures of phenomena.
Although content analysis has been employed in
research for a long time [34], e.g., in the behavioral and
social sciences, its application in information systems (IS)
research is relatively new. The extensive use of
groupware-supported decision-making in the extant group
support systems (GSS) and virtual teams research entails
that a considerable body of communication and message
content must have been generated in the process.
However, very few studies on virtual teams have followed
a content analysis methodology. Content analysis of
quantitative statistical
0-7695-2268-8/05/$20.00 (C) 2005 IEEE
Proceedings of the 38th Hawaii International Conference on System Sciences - 2005
1

Page 2

group communications can help explain performance
differences, and the findings can help improve group
interaction [1].
In this research, we analyze the content of group
discussions to identify conflicts. We also examine the
influence of conflict on team performance as measured by
the level of agreement on the final decision and by team
members’ perceptions of decision quality and decision
making process.
In the next section we review literature on virtual
teams, team conflict, conflict management, and team
performance. In section 3, we present the research
hypotheses. We describe the research methodology in
section 4. The results of hypotheses testing are presented
in section 5. Finally, we conclude the paper with
discussion of the results, limitations of the study, and
future directions.

2. Literature Review
Development
2.1. Group Decision Making

The normative model of group decision-making stems
from Simon’s individual rational decision-making model,
which includes the broad phases of intelligence, design
and choice [33]. The group decision-making process
starts with the identification of the problem and the
possible solutions. In this stage, relevant information
concerning the problem is collected. During the design
phase, the group develops and discusses several possible
alternative solutions for the problem. Finally, in the choice
phase, group members choose a final solution from a set
of alternatives.
In contrast to Simon’s model is a decision-making
model that is the garbage can model [4][6]. In this model,
decision- making is viewed as a social-interaction process
rather than a sequence of decision-making activities.
Decision-making is characterized by an element of
chance.
Mintzberg, Raisinghani, and Theoret [20] synthesize
these two perspectives by introducing interruptions, delays
and deviations into the rational, linear decision-making
model. Decision makers may loop back and forth among
decision-making activities.
While a group decision-making process involves
similar phases of activities, it has additional complexities.
It requires discussions among team members and a
decision process structure to arrive at a final solution.
During the course of discussion, group members may
encounter differences in views, beliefs, and objectives
thus engaging in conflicts. The nature of the emerging
conflict is likely to vary depending on the decision phase
in which the conflict arises, and group members follow
different conflict resolution styles to resolve conflicts.
and Theory
The intra-group conflict and group members’ conflict
resolution style is likely to impact the group performance.

2.2. Global Virtual Teams

Virtual teams are groups of geographically dispersed
co-workers who are brought
telecommunication and information technologies to work
on a particular task [38]. Global virtual teams usually
span across space, time and organizational boundaries
[16]; are culturally diverse and geographically dispersed
[12]; and perform tasks of a global nature [19].
Prior research on virtual teams has addressed many
aspects such as conflict and cultural heterogeneity [23],
communication patterns and challenges [31], coordination
and perceived conflict management styles [22], effects of
communication media [35][36], trust [12], and best
practices [17], to list a few. We believe that cultural and
functional diversity in virtual teams lead to differences in
the members’ thought processes. Consequently, conflict
may be inevitable, which necessitates a close examination
by information systems researchers.

2.3. Team Conflict

Conflict refers to the awareness by various parties of
their differences, discrepancies, incompatible wishes or
irreconcilable desires [18]. Behavioral scientists have
argued variedly over the role of conflict in team
performance in organizations. While some contend that
conflict is dysfunctional and must be avoided, others have
put forth that conflict is natural and inevitable and may aid
group performance. A more recent school of thought
maintains that conflict is necessary for groups to perform
effectively. Thus the effect of conflict on decision-
making has not been clearly established [8]. Conflict may
hinder decision-making and disrupt the exchange of
information, thus reducing decision quality [32]. When
used constructively, conflicts may cause a thorough
evaluation of decision alternatives [7][13] as each member
brings unique perspectives and knowledge to the group
discussion, which may question the assumptions made by
the participants..
Two types of conflict have been discussed consistently
in the behavioral science literature – issue-based (task)
and interpersonal (affective) conflict. Issue- based
conflict refers to task-related discord among group
members. Interpersonal conflict refers to discord among
people who differ from each other in their preferred
outcomes, attitudes, values or behavior. Jehn [13] argues
that while interpersonal conflict can be detrimental to
group performance, moderate amounts of task or issue-
based conflict can be beneficial. Groups using computer
support are likely to have less affective conflict and more
together using
0-7695-2268-8/05/$20.00 (C) 2005 IEEE
Proceedings of the 38th Hawaii International Conference on System Sciences - 2005
2

Page 3

task conflict, as they are more focused in their discussions,
than the groups having no computer support [23]. The
anonymous interaction supported by GSS could be a
major reason for low levels or absence of interpersonal
conflict. Anonymity allows greater freedom to participate
openly, and hence it may encourage more productive
conflict [27]. On the other hand, some authors have also
argued that the lean medium does not allow context and
social cues to be conveyed, and hence virtual teams are
likely to experience greater
communication challenges [22]. This may therefore lead
to greater conflict among team members. In addition,
team diversity creates
communication [14][15][19]. Such communication
difficulties may lead to conflict in virtual teams and affect
the team performance.
Conflict in GSS groups may occur in any of the three
phases of the decision-making process. While during the
intelligence phase differences may occur in the basic
definition of the problem or decision situation, during the
design phase, members may argue how to make the final
selection from a set of alternatives. Finally, in the choice
phase each group member, based on his/her perceptions,
values, beliefs, and preferences, may try to select his/her
own preferred solution. When group members favor
different solutions, the group experiences conflicts. We
use the term “choice conflict” to refer to this type of
conflict. Literature has examined certain styles of conflict
management that arise in multi-party decision situations,
which are described next.

2.4. Conflict Management Styles

Five types of conflict resolution styles have been
discussed in the literature [28][37]. These are avoidance,
accommodating, competitive,
compromise. The avoidance style refers to intentional
withdrawal from the conflict
accommodating style signifies a focus on areas of
agreement and attempts to smooth out differences. The
competitive style describes an approach in which
members enforce their own views on others. The
collaborative style pertains to integrating the views of all
involved. Finally, the compromising style is associated
with finding a middle ground solution or a common
solution that addresses everybody’s interest. An abridged
form of conflict resolution style categories that includes
integrative, distributive, and avoidance approaches has
also been discussed in the literature [21][30]. In this
classification, avoidance is the absence of an attempt to
resolve the team conflict; the integrative conflict
management style is an attempt to arrive at solutions and
outcomes that are satisfying to all members; and the
distributive conflict management style emphasizes the
enforcement of one person’s choices over those of the
coordination and
obstacles to effective
collaborative, and
situation. The
other. While resolving intra-group conflicts, GSS groups
may tend to follow more integrative and fewer distributive
or avoidance approaches than face-to-face groups [21].
This study focuses on GSS-based virtual teams, and we
restrict our discussion to only three categories of conflict
resolution, which are integrative, distributive, and
avoidance approaches. As discussed below, we
concentrate on integrative conflict resolution style, which
promises best performance of GSS-based groups
[22][25][27].

2.5. Virtual Teams Performance

Prior studies on the performance of GSS-based groups
and virtual teams have frequently included variables such
as decision time, member satisfaction, participation,
consensus, agreement,
[3][17][22][36][39]. In this preliminary study on conflicts
in GSS-based virtual teams, we primarily explore the
issues concerning the effectiveness rather than the
efficiency of virtual teams. Thus, we shall exclude
decision time from our current research and concentrate
on satisfaction with the decision process, agreement, and
perceived decision quality as the indicators of team
performance. The following section develops the theory
and presents the related hypothesis tested in this paper.

3. Theory Development and Hypotheses

Virtual teams are composed of people from varied
backgrounds and cultures, which may cause diverse
beliefs, opinions, perceptions, outlooks and attitudes to
prevail in the activities of the teams, and team members
may have to encounter conflicts. We expect that the nature
of the conflict depends on the task type and, more
precisely, on the phase of the task that a group is engaged
in. In decision-making tasks (the type that is studied in
this research), in the early phases (e.g., intelligence and
design) group members may differ in their views on
various alternatives and the method to be adopted to
eliminate the less promising ones so that a final choice can
be made. Although the prevalence of conflicting
viewpoints may slow down the decision making process, it
enables group members to understand the preferences of
other members. The nature of intra-group conflict
changes once a group decision process enters into the
choice phase of decision-making. As group members try
to make a final choice from a set of filtered alternatives,
each of which has a minimum level of acceptance to one
or more members of the group, the intra-group conflict
may intensify. The conflict that arises in the “choice”
phase of decision-making is referred to as “group choice
conflict” or simply “choice conflict” in this study.
and decision quality
0-7695-2268-8/05/$20.00 (C) 2005 IEEE
Proceedings of the 38th Hawaii International Conference on System Sciences - 2005
3

Page 4

In the early phase of decision-making when the virtual
team members actively discuss the various possible
alternatives, conflict occurs mainly due to variances in
members’ opinions, beliefs and attitude with regard to the
each alternative. These conflicting views may or may not
be intense depending on the preferences of team members.
We expect that the number of conflicting views that
prevail in the early phase of decision-making will
influence the performance of the virtual teams. The
occurrence of too many conflicting views may lead team
members to perceive the decision making process as not
being smooth and that the final decision does not have
unanimous acceptance. We, however, do not expect the
initial conflicting views to have any significant effect on
the agreement level of the team members with respect to
the final choice. Hence:

H1a: In GSS-based global virtual teams, the higher
the number of conflicting views prevailing in the early
phase of decision-making, the less favorable are the
members’ perceptions of the decision quality.

H1b: In GSS-based global virtual teams, the higher
the number of conflicting views in the early phase of
decision-making, the lower
satisfaction with the decision-making process.

Conflicts that occur in the choice phase of a group
decision-making process are expected to be focused and
intense mainly because the decision makers concentrate
on very few selected alternatives, evaluate the features of
the choice alternatives [11] and try to justify their
individual preferences for the final solution. Intense
choice conflicts are likely to have adverse effects on team
members’ perceptions of the decision outcome and
process and will be reflected in the level of their
agreement on the final decision outcome. Hence:

H2a: In GSS-based global virtual teams the higher the
intensity of choice conflict, the less favorable are the
members’ perceptions of the decision quality.

H2b: In GSS-based global virtual teams, the higher
the intensity of choice conflict, the lower is group
members’ satisfaction with the decision-making process.

H2c: In GSS-based global virtual teams, higher the
intensity of choice conflict, the lower is the level of group
members’ agreement on the final decision.

When virtual team members’ preferences converge on
only one alternative in the early phase of decision-making,
the intra-group conflict is less likely to occur in the choice
phase. Teams with no choice conflict are expected to
have higher levels of agreement on the final choice and
is group members’
more favorable perceptions about the decision process and
outcome than those experiencing choice conflicts. Hence:

H3a: GSS-based global virtual teams experiencing
choice conflict will have lower perceived decision quality
than the teams that do not experience such conflict.

H3b: GSS-based global virtual teams experiencing
choice conflict will have lower satisfaction with the
decision-making process than the teams that do not
experience such conflict.

H3c: GSS-based global virtual teams experiencing
choice conflict will have lower levels of members’
agreement than the teams that do not experience such
conflict.

Of the various conflict resolution styles pursued by
virtual teams, the collaborative (integrative) approach has
been consistently favored in IS research. Prior research
has consistently shown that collaborative (integrative)
conflict management style has a positive influence on the
performance of global virtual teams [21][22][25][27]. In
virtual teams that attempt to follow an integrative
approach, members discuss and attempt to patch
differences, and try to arrive at a win-win solution. The
choice would thus appear fair and acceptable. Hence:

H4a: GSS-based global virtual teams following an
integrative conflict resolution style will experience higher
perceived decision quality than the teams following other
conflict resolution approaches.

Members of the teams pursuing integrative conflict
resolution style are also expected to perceive that the
decision-making process has been fair and lead to a
solution that is acceptable to all members. Hence:

H4b: GSS-based global virtual teams following an
integrative conflict resolution style will have higher levels
of satisfaction with the decision-making process than the
teams following other types of conflict resolution
approaches.

Integrative conflict resolution allows group members to
resolve the conflict without having to compromise their
respective viewpoints and opinions. When such a
resolution style is followed, members are likely to
experience higher levels of agreement on the final
decision. Hence,

H4c: GSS-based global virtual teams following an
integrative conflict resolution style will have higher levels
of member agreement on the final decision than do the
teams that follow other conflict resolution approaches.
0-7695-2268-8/05/$20.00 (C) 2005 IEEE
Proceedings of the 38th Hawaii International Conference on System Sciences - 2005
4

Page 5

4. Research Method

We used the data collected in a laboratory experiment
on global virtual teams to test our hypotheses. Volunteer
subjects enrolled in MBA programs at a major
Midwestern US university participated along with
graduate students from a premier management school in
India. All subjects were experienced with information
technology, including internet/Web skills as measured by
a questionnaire completed by all subjects. The student
subjects were distributed among the following categories:
•
US-homogeneous: Consisting of students who had
lived in the US since birth and the students who
had resided in the US for a long duration (typically
over 10 years).
•
Indian-homogeneous:
residing in India.
•
Heterogeneous: A mix of students who had lived
in the US since birth, the students residing in India,
and the foreign students who had lived in the US
for a shorter duration (typically less than 10 years).
For the purpose of conducting statistical analyses, the
US- and Indian-homogeneous were considered culturally
homogeneous groups, whereas the heterogeneous groups
were labeled as culturally heterogeneous groups. A total
of 83 persons completed all phases of the study as
follows: 4 US homogeneous, 9 Indian homogeneous, and
9 heterogeneous groups. Seventeen groups had four
persons; five groups had three.
Once the availability of the students in each location
was known, accounting for the 10 ½ hour time difference
between the two countries and class schedules of the
participants, students were randomly assigned to either
homogeneous or heterogeneous groups, based on their
availability.
Subjects participated in a detailed training session.
The session consisted of a demonstration of each
command of the GDSS software package to be used in the
experiment and of a dummy task. The training session
was conducted in a computer laboratory that was also
used for the actual experiment. The session was
conducted by two of the researchers who had extensive
experience with the software.

4.1. Task identification and description

Given that the participants were students, it was felt
that the involvement of the students would be stronger if
the task was one they would easily relate to. Accordingly,
the task chosen was the selection of a computer use fee for
students enrolled in an online university. Groups were
provided with a printed and online task description. It
identified reasons why the university’s administration was
considering the implementation of a fee, e.g., need for a
Consisting of students
help desk. The description also identified five fee
options: a flat fee for all courses; one fee for
undergraduate courses and a higher fee for graduate
courses; a graduated fee, based on intensity of computer
use in a course; a “fee for use;” and a fee based on the
country of origin of the student.
As part of the experimental procedures, described
below, students in each group discussed these options
among themselves online and selected one option as the
final choice of the group. Experimental procedures did
not provide for identifying and recommending any option
other than the five stated above.

4.2. Experimental procedures

The subjects were randomly assigned to groups and
were informed as to when they would participate.
Anonymity among group members was maintained. Each
session consisted of the following activities:
•
Activity 1- commenting
disadvantages, etc. of each option. The software
allowed students to read each option, comment on
options as desired, comment on other group
members’ statements, etc. When finished, each
group member rated the five options from 0 (least
appropriate/worst) to 4 (most appropriate/best).
Subsequently, each group member viewed the
rating results for his/her group.
•
Activity 2 – commenting on the group’s rating in
activity 1. This Activity centered on discussions
about why or why not the highest-ranked option
was good, etc. and resolved conflict if more than
one option ended up having almost a similar rating.
Near the end of Activity 2, each group identified an
option as a choice of the group, and members voted
“yes” or “no” to accept or reject the proposed
choice.
•
Activity 3 – each group completed a short post-test
questionnaire.
Groups could select an option for computer use fee
following multiple paths: selecting the option that had
maximum average rating or minimum standard deviation
or both in the group rating process of activity 1; or select
an option that might not have the highest rating in activity
1 but appeared to be satisfactory to the majority of the
group members during the course of the discussion in
activity 2.
These activities were
Consensus@nyWARE, a web-based GDSS. Each group
was under the control of a facilitator who communicated
with team members using “instant messaging.” The
facilitator monitored the discussions and dealt with any
technical software questions; the facilitator did not
on advantages,
implemented using
0-7695-2268-8/05/$20.00 (C) 2005 IEEE
Proceedings of the 38th Hawaii International Conference on System Sciences - 2005
5

Page 6

interject anything into the discussion regarding the task
and the computer use fee options.

4.3. Variable identification

In this research, the independent variables are the
number of intra-group conflicts in the intelligence phase
of decision-making and the intensity of intra-group
conflict and the corresponding conflict resolution
approach followed by group members in the choice phase
of decision making. The dependent variables are the level
of group agreement on the final decision, perceived
decision quality, and satisfaction with the decision making
process. The control variable is cultural diversity
consisting of two levels: homogeneous and heterogeneous
groups.
The GSS software captured a record of the groups’
communications and discussions. The group discussions
were analyzed to identify the number of conflict episodes
in activities 1 and 2. For each conflict episode in activity
2, the contents were examined to identify the
corresponding resolution approaches followed by the
group.
In order to measure the intensity of choice conflict,
meeting transcripts were coded using the rules based on
the classification system used by Valacich and Schwenk
[40]. A brief description of the coding rules used to code
the group discussion is given in appendix 1. The extent of
supportive and critical remarks and arguments made by
the group members in activity 2 reflects the level of group
choice conflict that prevails among them [26]. We hence
measured group choice conflict as the proportion of the
total comments (in activity 2) on group choice that were
critical in nature, i.e., (CR+CA)/(CR+CA+SR+SG). Two
coders independently identified the categories of the
comments and confirmed the comments belonging to the
supportive and critical categories. The intercoder
correlation for number of supportive and critical
comments is 0.82 (n=22, p< 0.0001), which is acceptable
in view of the exploratory nature of the study in the
context of virtual teams.
Degree of agreement was measured as the percentage
of members voting in favor of the final decision proposed
by a group. The other two dependent variables were
measured using 5-point Likert-type scales: perceived
decision quality and satisfaction with the decision-making
process. The indicator items of these scales are presented
in appendix 2.

5. Results

As some of the measures in this study were based on
self-reported data, it was necessary to assess reliability
and validity of the instruments used to capture the data.
Cronbach Alpha coefficients were calculated for each
instrument. As the measurement scales used had not been
tested and validated before and the research is exploratory
in nature, a cut-off value 0.70 was considered acceptable
[24]. An alpha of 0.833 was found for “satisfaction with
the decision making process” and 0.903 for “perceived
decision quality.” In order to determine construct
validity, we conducted factor analysis employing
VARIMAX orthogonal rotation for each instrument.
Factor analysis for each instrument resulted in a single
factor structure with high factor loadings.

5.1. Hypothesis Testing

The hypotheses were tested using t-test and regression
analyses with a level of significance of 0.05. Any weak
significance level in the range of .05 to .10 was treated as
suggestive of the nature of relationship between the
variables. We also conducted t-tests to ensure that the
independent variables did not vary significantly across the
homogeneous and heterogeneous groups.
Twenty-two groups participated in the experiment.
Seventeen of these groups had conflict on the final choice
whereas the remaining 5 groups did not have any choice
conflict. The results of t-tests demonstrated that groups
having choice conflict have lower perceived decision
quality than their counterparts with no choice conflicts
(table 1).

Table 1. Results of t-Tests for groups with and
without choice conflict
Mean (Std Devn) Dependent
Variable
Choice
Conflict Choice
Conflict
Perceived
decision
quality
Satisfaction
with the
decision
making
process
Level of
agreement
(0.156) (0.097)
No
t-
statistic,
df
Prob. (F)
Hypotheses
Support
3.39
(0.573)
4.17
(0.136)
2.49, 20 0.022
H3a: Yes
3.64
(0.376)
3.86
(0.448)
1.12, 20 0.276
H3b: No
0.59 0.68 1.75, 19 0.097
H3c: Yes
[Weak]

We also found weak support for hypothesis 3c, which
proposes that virtual teams experiencing choice conflict
have a lower level of agreement. Hypothesis 3b remains
unsupported in this study.
Next, we considered the 17 groups that had choice
conflict and focused on their conflict resolution styles.
The results of the t-tests (table 2) demonstrate that groups
following an integrative conflict resolution style had a
0-7695-2268-8/05/$20.00 (C) 2005 IEEE
Proceedings of the 38th Hawaii International Conference on System Sciences - 2005
6

Page 7

higher level of agreement, perceived decision quality, and
satisfaction with the decision-making process than their
counterparts who followed distributive approaches.

Table 2. Results of t-Tests for integrative and
distributive conflict resolution style
Mean (Std Devn) Dependent
Variable
Integrativ Other
e styles styles
t-
statistic,
df
Prob. (F)
Hypotheses
Support
Level of
agreement
0.72
(0.164)
0.47
(0.142)
3.86, 14 0.002
H4a: Yes
Perceived
decision
quality
Satisfaction
with
decision
making
process

Finally, we examined if group performance, measured
in terms of levels of agreement, perceived decision
quality, and satisfaction with the decision-making process,
had negative relationships with the number of conflicting
views prevailing in the early part of decision-making
process and with the intensity of choice conflict. We
found that perceived decision quality had negative
relationships with the number of conflicting views and
with the intensity of choice conflict. We also found that
both the level of agreement and satisfaction with the
decision making process had negative relationships with
the intensity of choice conflict. Table 3 presents the
results of the regression analyses.

Table 3. Results of regression analyses
Independen
t Variable with Decision
Outcome
Intercept 4.64****
# of
conflicting
views in the
intelligence
phase of
decision
making
Intensity of
choice
conflict
R-Square 0.44
F 7.47
N 22
Hypotheses
Supported? Yes
[Weak]
* p<0.10; ** p<0.05; *** p<0.01; **** p<0.001

4.02
(0.518)
3.45
(0.505)
2.28, 15 0.038
H4b: Yes
3.87
(0.401)
3.47
(0.267)
2.49, 15 0.025
H4c: Yes
Satisfaction
Satisfaction with
Decision Making
Process
3.99****
-0.03
Level of
Agreem-
ent
0.90****

-0.12*




-1.52***

-0.68*

-0.47**
0.16
1.82
22
0.21
5.15
21
H2c:
Yes

H1a:
H2a:
Yes
H1b:
No
H2b:
Yes
[Weak]
6. Discussion

This study aims at enhancing our understanding of
conflicts in virtual teams by examining the communication
content of several sessions of virtual teams work.
Haggarty [9] argues that content analysis can provide
valuable knowledge especially when used to complement
other techniques. We determined the conflict episodes
and the conflict resolution styles through content analyses
and conducted statistical analyses to demonstrate that
conflicts in the choice phase of decision making are
detrimental to global virtual teams. The segregation of
conflicts occurring in different phases of a decision-
making process is an interesting aspect of this study.
We also found that groups following integrative
conflict resolution styles had better performance than
those following other conflict resolution approaches, such
as a distributive style. The findings of this research are
similar to those of the prior studies on conflicts in GSS
supported groups [21][22]. However, unlike prior studies
we analyzed the contents of group discussion to identify
conflict episodes and conflict resolution styles.
In this study we did not find any support for the
relationship between satisfaction with the decision-making
process and the number of conflict episodes in the early
phase of decision making. We also found that process
satisfaction in teams with and without choice conflict was
not different. One possible explanation is that satisfaction
with the decision-making process is not directly
influenced by intra-group conflicts. Team members are
satisfied with a process that results in a consensual
solution. We carried out a regression analysis and found
that satisfaction with the decision-making process is
positively related to the level of agreement (?=1.30;
p=0.0006; R-square=0.469). Next, we included the
intensity of choice conflict as a regressor variable in the
regression model. We found that the original weak
relationship between the intensity of choice conflict and
satisfaction with the decision-making process (table 3)
disappears when level of agreement is present in the
regression model. The findings indicate that the
relationship between intensity of choice conflict and
process satisfaction is mediated by the level of agreement.
In-depth study of this relationship is a future research
agenda.

6.1. Limitations

Content analysis has limitations in its scope and
ability to fully capture the meaning and essence of a group
discussion. Moreover, content analysis is subject to the
interpretation of the coder. Multiple coders can be used to
overcome the interpretation bias.
0-7695-2268-8/05/$20.00 (C) 2005 IEEE
Proceedings of the 38th Hawaii International Conference on System Sciences - 2005
7

Page 8

Also, GSS enabled, anonymous group discussions
may sometimes generate unproductive discussions and
low levels of ownership to views and opinions. This can
be reduced if partial anonymity is introduced in GSS
based group interactions.

7. References

1. 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.
2. Berelson, B., Content Analysis in Communication
Research, New York: Free Press, 1952.
3. Chidambaram, L., and Jones, B., “Impact of
communication medium and computer support on
group perceptions and performance: A comparison
of face-to-face and dispersed meetings”, MIS
Quarterly, (17:4), 1993, pp. 465-516.
4. Cohen, M.D., March, J.G., and Olsen, J.P. “A
Garbage Can Model of Organizational Choice,”
Administrative Science Quarterly, (17:1), 1972, pp.
1-25.
5. Coppola, N. W., Hiltz, S. R., and Rotter, N. G.,
“Becoming a Virtual Professor: Pedagogical Roles
and Asynchronous Learning Networks”, Journal of
Management Information Systems, (18:4), Spring
2002, pp. 169-189.
6. Cyert, R.M., and March, J.G. A Behavior Theory of
the Firm. Englewood Cliffs, NJ: Prentice-Hall, 1963.
7. Cosier, R.A., and Schwenk, C.R., “Agreement and
thinking alike: ingredients for poor decisions”,
Academy of Management Executive, (4:1), 1990.
8. Eisenhardt, K.M., and Zbaraki, M.J., “Strategic
Decision Making”, Strategic Management Journal,
(13), 1992.
9. Haggarty, L., “What is Content Analysis”, Medical
Teacher, (18:2), June 1996.
10. Holsti, O. R., “Content Analysis”, in Linzey G. And
Aronson, E. (Eds), The Handbook of Social
Psychology, Vol. 2, Addison-Wesley, Reading, MA,
1968.
11. Houstan, D.A. and Doan, K., “Comparison of paired
choice alternatives and choice conflict”, Applied
Cognitive Psychology, (10), 1996, pp. S125-S135.
12. Jarvenpaa, S.L., and Leidner, D.E., “Communication
and trust in global virtual teams”, Journal of
Computer-Mediated Communication, (3:4), 1998,
available
http://www.ascusc.org/jcmc/vol3/issue4/jarvenpaa.ht
ml.
13. Jehn, K. A., “A multimethod examination of the
benefits and detriments of intragroup conflict”,
Administrative Science Quarterly, (40:2), 1995, pp.
256-282.
at
14. Kankanhalli, A., Tan, B., Wei, K.-K., “Technology,
Culture, and Conflict in Virtual Teams: A Case
Study”, Proceedings of the Eleventh Australisian
Conference on Information Systems, 2000.
15. Kayworth, T.R., and Leidner, D., “Leadership
effectiveness in global virtual teams”, Journal of
Management Information Systems, (18:3), 2001-2,
pp.7-40.
16. Lipnack, J., and Stamps, J., Virtual Teams-Reaching
Across Space, Time and Organizations with
Technology, John Wiley & Sons, New York, 1997.
17. Lurey, J.S., and Raisinghani, M.S., “An empirical
study of best practices in virtual teams”, Information
& Management, (38:8), 2001, pp.523-544.
18. Mannix, E.A., Griffith. T. and Neale, M.A. “The
Phenomenology of Conflict in Distributed Work
Teams”, in Distributed Work, ed. P. Hinds, and S.
Kiesler, The MIT Press, Massachusetts, 2002.
19. Maznevski, M.L., and Chudoba, K.M., “Bridging
space over time: Global virtual team dynamics and
effectiveness”, Organization Science, (11:5), 2000,
pp. 473-492.
20. Mintzberg, H., Raisinghani, D. and Theoret, A., “The
Structure of Unstructured Decision Processes”,
Administrative Science Quarterly, 1976, pp. 246-
275.
21. Miranda, S.M., and Bostrom, R.P., “The impact of
group support systems on group conflict and conflict
management”, Journal of Management Information
Systems, (10:3), 1993-94, pp.63-95.
22. Montoya-Weiss, M.M., Massey, A.P., and Song, M.,
“Getting it together: Temporal coordination and
conflict management in global virtual teams”,
Academy of Management Journal, (44:6), 2001,
pp.1251-1262.
23. Mortensen, M., and Hinds, P.J., “Conflict and shared
identity in geographically distributed teams”, Journal
of Conflict Management, (12:3), 2001, pp. 212-238.
24. Nunnally, J. C., Psychometric Theory, McGraw-Hill
Book Company, New York.
25. Paul, S., Seetharaman, P., Samarah, I., and Mykytyn,
P. P., “Impact of heterogeneity and collaborative
conflict management style on the performance of
synchronous global virtual teams”, Information &
Management, (41:3), 2004, pp. 303-321.
26. Paul, S., Seetharaman, P., Samarah, I., and Mykytyn,
P. P., “Conflict in GSS-based Virtual Teams:
Findings from an Experiment”, Proceedings of the
Americas Conference on Information Systems, 2004,
(forthcoming).
27. Poole, M.S., Holmes, M., and DeSanctis, G.,
“Conflict management in a computer-supported
meeting environment”, Management Science, (37:8),
1991, pp. 926-953.

0-7695-2268-8/05/$20.00 (C) 2005 IEEE
Proceedings of the 38th Hawaii International Conference on System Sciences - 2005
8

Page 9

28. Rahim, M.A., “A measure of styles of handling
interpersonal conflict”, Academy of Management
Journal, 26, 2, (1983), pp. 368-376.
29. Samarah, I., Paul, S., Mykytyn, P., and Seetharaman,
P. “The collaborative conflict management style and
cultural diversity in DGSS supported fuzzy tasks: An
experimental investigation”, In R.H. Sprague Jr. (ed.),
Proceedings of the Thirty-Sixth Annual Hawaii
International Conference on Systems Sciences, Los
Alamitos, CA: IEEE Computer Society Press, 2003,
available
http://csdl.computer.org/comp/proceedings/hicss/200
3/1874/01/187410040a.pdf
30. Sambamurthy, V., and Poole, M.S., “The effects of
variations in capabilities of GDSS designs on
management of cognitive conflict in groups”,
Information Systems Research, (3:3), September
1992, pp. 224-251.
31. Sarker, S., and Sahay, S. “Information systems
development by US-Norwegian virtual teams:
Implications of time and space”, In R.H. Sprague Jr.
(ed.), Proceedings of the Thirty-Fifth Annual Hawaii
International Conference on System Sciences, Los
Alamitos, CA: IEEE Computer Society Press, 2001,
available
http://csdl.computer.org/comp/proceedings/hicss/200
2/1435/01/14350018.pdf
32. Schweiger, D.M., Sandberg, W.R., and Ragan, J.W.,
“Group approaches for improving strategic decision
making: A comparative analysis or dialectical inquiry,
devil's advocacy and consensus”, Academy of
Management Journal, (29:1), 1986.
33. Simon, H. A., The New Science of Management
Decision, 1st Edition, Harper and Row, New York,
1960.
34. Smith, A. D., “Another Look at Content Analysis: An
Essay Review”, Social Work Research and Abstracts,
(18:2), 1982, pp. 5-10.
35. Sosik, J.J., and Jung,
characteristics and performance in collectivistic and
individualistic cultures”, The Journal of Social
Psychology, (142:1), 2002, pp. 5-23.
36. Straus, S.G., “Getting a clue: The effects of
communication media and information distribution on
participation and performance in computer-mediated
and face-to-face groups”, Small Group Research,
(27:1), 1996, pp. 115-142.
37. Thomas, K.W., and Kilmann, R.H., Thomas-Kilmann
Conflict Mode Instrument, Tuxedo, NY: Xicom,
1974.
38. Townsend, A.M., DeMarie, S.M., and Hendrickson,
A.R. “Virtual teams: Technology and the workplace
of the future”, Academy of Management Executive,
(12:3), 1998, pp. 17-29.
at
at
D.I., “Work-group
39. Turoff, M., Hiltz, S.R., Bhagat, A.N., and Rana, A.R.,
“Distributed group support systems”, MIS Quarterly,
(17:4), 1993, pp. 399-417.
40. Valacich, J. S. and Schwenk, C., “ Structuring
conflict in individual, face-to-face, and computer-
mediated group decision making: Carping versus
objective devil's advocacy”, Decision Sciences, (26:
3), 1995, pp. 369-394.
0-7695-2268-8/05/$20.00 (C) 2005 IEEE
Proceedings of the 38th Hawaii International Conference on System Sciences - 2005
9

Page 10

Appendix 1
Coding Rules

Type of Discussion Statement
Supportive remark
Expresses support for the option chosen by the
group, without adding evidence or remark. (“I
am for this option”; “I agree with you”)
Supportive argument
Supports the option chosen by the group, and
gives evidence or argument to justify (“I support
this option because it will eliminate many
problems.”)
Critical remark
Expresses opposition to the option chosen by the
group but does not add evidence or argument (“I
don’t like that”; “I don’t agree with you”)
Critical argument
Expresses opposition to the option chosen by the
group and gives evidence or argument to justify
(“I don’t like this option because it has the
following drawbacks.”)
Query
Requests clarification of another person’s
comment or about the option chosen.
Group comment
Remark about the interpersonal process of the
group (“let’s summarize”, “lets try to agree on
something, anyway”).
Remark about the system
General remark about the computer system or
the software used for the task.
Off the track comments
Remarks that are “off the topic” and do not fit
into any of the above categories.
Uncodable text
Uncodable text
Code
SR
SG
CR
CA
QS
GC
COM
OTT
UC


Appendix 2
Indicator Items for Perceived Decision Quality
1. The decision made by my group is practical
2. The decision made by my group is fair
3. I am confident that the final decision we came up
with is the best decision
4. I feel that the quality of the group's decision
would have positive effects on the performance
of the university
5. Overall, it is my opinion that our final decision is
of high quality

Indicator Items for Satisfaction with the Decision
Making Process
1. I believe my contribution to be significant in our
group arriving at the final decision
2. Our group was able to reach a consensual
solution without any major conflict
3. I feel that the group members converged on the
final decision
4. The decision making process of the group was
complete
5. The progress of the group towards the stated
goals of the task was satisfactory
6. Overall, as a member of our team, I am satisfied
with the process I employed in arriving at the
final solution
7. Overall, I am satisfied with the solution process
0-7695-2268-8/05/$20.00 (C) 2005 IEEE
Proceedings of the 38th Hawaii International Conference on System Sciences - 2005
10