Individual Differences and Computer Mediated Communication: The Role of Perception
***** OLANIRAN ************* EJC/REC Vol. 3, No. 2, 1993 ***
INDIVIDUAL DIFFERENCES AND COMPUTER MEDIATED COMMUNICATION:
THE ROLE OF PERCEPTION*
Bolanle A. Olaniran
Texas Tech University
Abstract. This study attempted to investigate
how perceived computer mediated communication
(CMC) system attributes and individual
characteristics (e.g., gender) affect users'
perception of five communication outcome variables
(user satisfaction, decision confidence,
immediacy, effectiveness, and ease of use of the
system) across three CMC systems. The results
lend support to the argument that perceived
system attributes influence all five communication
outcome variables. The effect, however, is
moderated by gender for three of these variables
(i.e., immediacy, decision confidence , and
effectiveness). Interaction of system attributes
and gender effects was also found for these three
variables. Discussion, implications, and
research limitations are presented as well.
The degree to which a communication medium influences
the human communication process has been subjected to close
scrutiny in recent years (Rice and Love, 1987). The
computer-mediated communication (CMC) research is at the
middle of the debate. At the center of this debate is the
argument that medium characteristics account for
communication differences found between CMC and traditional
face-to-face (FTF) communication. The theoretical
explanation often used to justify this argument can be
linked to social presence theory. Social presence theory
describes the degree to which other communication
participants are believed to be jointly involved in the
communication process (Short, William, & Christie., 1976).
The proponent of this theory subscribes to the notion that
CMC systems are low in capacity to convey information about
facial expression, posture, and nonverbal cues. The lower
availability of such cues is believed to influence users'
views of the communication medium, contexts, performance,
and message interpretations. Specifically, CMC, with its
few nonverbal cues, is said to be lower in comparison to FTF
communication. Theoretically, social presence theory has
been used to account for task orientation and is less
communication oriented toward social-emotional functions in
CMC (Hiltz, Johnson, & Turoff, 1986; Rice, 1984).
According to Walther (1992b), it is not clear from
social presence theory whether media characteristics are the
causal determinants of communication differences or whether
the users' subjective perceptions of media influence their
behavior. In other words, while there are inherent media
characteristics, it is equally possible that users'
perceptions of these characteristics may be subjective and
socially constructed. Steinfield (1986) alludes to this
when he argues that factors like environmental uncertainty
and need for communication across locations influence the
extent to which CMC messages address socioemotional needs of
users.
Easton, George, Nunamaker, and Pendergast (1990) allude
to the importance of differential software in electronic
meeting systems (EMS). Dennis, George, and Vogel (1988)
define EMS as a meeting process where communication is
relayed electronically (e.g., computer conferencing,
teleconferencing, videoconferencing, etc.,). Easton et al.
(1990) argue that a mismatch between software and project
task could account for differential findings. Therefore,
this study attempts to investigate the role of participants'
perceptions on some communication outcome variables.
However, the perceptions will be measured for variations
across different CMC systems. While comparison among
differential systems is the primary focus of this study, the
role of users' personal characteristics on their perceptions
will also be explored. This is necessitated by the fact
that in some CMC research where these characteristics have
been considered, they are often controlled through random
assignment to treatments (e.g., Easton et al., 1990). Thus
this study focuses primarily on one of those user
characteristics (i.e., gender) to determine if it accounts
for variations in users' perceptions of communication
outcome variables.
The remainder of this paper presents relevant arguments
in the CMC literature followed by discussions of variables
of interest in this study, the research questions, a
description of the study, and the three systems used in the
present study as well as relevant conceptual issues. The
last portion of the paper presents the findings, discussion,
and implications for CMC research. The general consensus
among findings from CMC research lean toward the lack of
socioemotional cues in CMC content for one-shot experimental
studies. On the other hand, the longitudinal studies show
more positive socioemotional tone than those found for
one-shot experiments (Walther, 1992b; see also, Dennis,
Nunamaker, & Vogel., 1990/91). Walther (1992b) attributed
the inconsistency to lack of opportunity or time among
participants to develop a relationship in one-shot
experimental settings where most of the studies involved a
zero base history group. He argued that it is possible that
when participants are given enough time to develop a
relationship, they are more likely to express more social
information. Dennis et al. (1990/91) offered another
explanation for the general differential findings between
the field research and the laboratory experiments. They
posit that the inconsistencies between the two types of
research can be attributed primarily to differential
systems, contexts, tasks, and groups (see also Dennis et
al., 1988). With organizational contexts, they argue that
while the laboratory experiments often followed the rational
model, the field studies seldom followed it. On task, they
argue that task types such as intellectual versus
decision-making (see McGrath, 1984) differ between
laboratory experiments and organizational field studies.
They believe that the difference in task types could
influence the conclusion drawn from those studies.
Furthermore, they suggest that the "goodness" of task
solution, while present and easy to determine in the field
studies, is lacking in the experimental studies. They also
advocate that task clarity is often elucidated in
experiments, while group members in field studies often lack
a common understanding of the problem. On task nature, they
argue that laboratory experiments consist of "low
complexity" tasks that require one single meeting to solve
(one to two hours) whereas field studies are characterized
by "medium to high complexity" tasks that sometimes require
several meetings among the participants. Concerning
electronic meeting systems (EMS), the use and lack of use of
a facilitator in either laboratory or field studies are also
believed to influence findings from those studies
respectively. They indicate that it is a common practice to
see a facilitator taking an active role in the field study,
while laboratory studies are known for controlling the level
of participation by a facilitator (see Dennis, et al.,
1990/91 for a fuller discussion).
Perceived Characteristics of the Medium
Steinfield (1986) proposed two categories into which
CMC literature on medium attributes can be categorized. The
first category assumes the inherent physical characteristic
of a medium, while the second approach focuses on the
perceived characteristics. However, the two can be summed
into one single approach, because the research focusing on
the inherent physical media characteristics (e.g., social
presence) often measured them through the users' subjective
perception of the media. The measurement is collected from
ratings on semantic differential scales defined by such
adjectives such as warm, sociable, personal, and sensitive
(Short et al., 1976). According to CMC literature, some
researchers argue that the limited bandwidth or cues present
in the channels makes CMC a low social presence medium
(Heimstra, 1982; Rice, 1980). On the other hand, others
advocate that the degree of social presence attributed to a
CMC medium might be a factor of individual characteristics
as increased CMC usage is associated with greater perceived
social presence. Therefore, individual characteristics such
as age, sex, educational background, status, and others are
factors that could influence users' perception of CMC medium
attributes. The underlying theoretical perspective for this
reasoning can be traced to the attribution theory. This
theory assumes that the communication process involves
individuals' biological and psychological make up, which are
also influenced by the environment or social context (see,
Jaspars, Hewstone, & Fincham, 191983; Kelley & Michela,
1980). As a result, how one perceives and creates meaning
through encoding and decoding communication processes is
influenced by these factors. Some areas where CMC
characteristics have been explored in the previous research
include usefulness/utility, accessibility, ease of use,
convenience, and the richness of the channel (Steinfield,
1986). These variables are often examined as separate
concepts, while evidence suggests that some of these
variables are related. For instance, "ease of use" was
found to comprise three variables, which are convenience,
accessibility, and ease of use itself (see Ruchinkas, 1982;
Svenning, 1982). Despite the argument about the low social
cues and lack of nonverbal information in CMC media, a
perceived advantageous effect of these features has been
proposed. For instance, previous research indicated that
"nonverbal involvement" and "social cues" drawn from the
physical presence of others (e.g., dress and artifacts)
reveal certain information which attracts peoples' attention
and their increased concern for those features (e.g., Ekman,
Friesen, O'Sullivan, & Scherer, 1980; Mehrabian, 1972).
Thus, the lack of nonverbal cues in CMC or reliance on text
only information is believed to result in communication
depersonalization and "deindividuation." However, Siegel,
Dubrovsky, Kiesler, and McGuire (1986) indicated that the
deindividuation consequence of CMC can reduce the feelings
of embarrassment, guilt, fear of retribution, and empathy
for others. This feature of CMC should facilitate bolder or
uninhibited discussions among communication participants.
As a result, the uninhibited behaviors in CMC should lead to
an objective decision where choices are made based on facts
and quality of participation rather than social influence or
dominance by certain group members (see DeSanctis and
Gallupe, 1987; Kiesler, Siegel, & McGuire., 1984; Kraemer &
Pinsonneault, 1990; Siegel et al., 1986). It is speculated
that the deindividuation by CMC should result in a greater
decision confidence and satisfaction with the decision
outcome. Members are apt to believe that they have made the
best decision possible when less attention is paid to the
physical presence of other participants and the status or
spatial arrangements effects that ensue from such presence
(see Dennis, Nunamaker, & Paranka, 1991; Edinger &
Patterson, 1985 Olaniran, 1991b; Rice, 1984; Sproull &
Kiesler, 1986).
With satisfaction, one would hope that freedom to
participate without being subjected to reprimand would
increase member satisfaction with the group decision.
However, while deindividuation is seen as an inherent
feature of CMC media, an argument could be made that the
extent to which users perceive a medium as capable of
performing this function would influence participants'
ratings of a CMC medium on satisfaction and other
communication outcome variables. For instance, Rice (1987)
discusses the relationship between the "ease of use" of a
medium and satisfaction. The ease of use of a medium refers
to the degree to which a communicating medium helps
accomplish a task with minimal difficulty. Smith and
Vanecek (1988) on the other hand, advocate the simultaneity
of messages and participants' interactions' effect on
satisfaction. They indicate that when there is a lack of
opportunity in a CMC meeting for instant feedback or
"immediacy" in message clarification, a reduced shared
perception concerning progress towards group goals exists;
thus satisfaction is reduced (see also Smith & Vanecek,
1990).
Effectiveness is another communication variable that
could be influenced by perceived medium characteristics.
Effectiveness can be defined as the extent to which a medium
is considered adequate for task accomplishment. As such,
effectiveness could be viewed as a subjective perception of
the users. Poole and DeSanctis (1990) allude to this effect
when they talk about the "degree of respect" or attitudes
that a user has toward technology. For instance, they
indicate that because some users value technology, they
believe that it is useful, while other users have the
premonition that technology has no utilitarian value; thus,
they conclude that the use of such technology will distract
them from focusing on other important issues (see also
Sambamurthy, 1988). One would expect that such attitudes
will influence the users' perceptions of a CMC medium and
its effectiveness. Dennis, Nunamaker, and Vogel (1990/91)
indicate that most CMC studies found that participants
reported improved meeting effectiveness (e.g., Bui, &
Sivasankaran, 1990; Easton, Vogel, & Nunamaker, 1989;
Gallupe & McKeen, 1990; George, Easton, Nunamaker, &
Northcraft, in press; Steeb & Johnston, 1981). However,
these findings were based on studies comparing EMS to FTF
communication. Therefore, it is unclear whether this effect
remains constant across CMC systems or differs with respect
to how CMC systems are perceived by users. Furthermore, it
would be profitable to determine the extent to which
individual characteristics (e.g., gender) influence users'
perceptions of CMC systems on effectiveness, as this
variable is often controlled in group research (e.g., Easton
et al., 1990; Steinfield, 1986). At this juncture, it would
be fair to acknowledge that the inherent features of CMC
systems influence how users form their perceptions. There
are features of CMC, such as the degree to which the medium
allows easy accessibility to messages and feedbacks, that
may influence users' conclusions about suitability of the
medium for a given task. For instance, interactive and
non-interactive media are believed to influence perceived
satisfaction with a medium (see Hiltz, 1988; Hiltz et al.,
1986; Smith & Vanecek, 1988). Therefore, one could argue
that users' perceptions influence and are being influenced
by inherent features of a medium and users' characteristics.
The conclusion that can be drawn from the CMC research
is that communication outcome variables such as
satisfaction, immediacy, decision confidence, effectiveness,
and ease of use are influenced by both physical and
perceived characteristics of a CMC medium. However, the
major factor could be how CMC users perceive the medium
attributes. For instance, Chidambaram, Bostrom, and Wynne
(1990/91) argue that group outcomes may not be so much
affected by the technology itself, but by how the technology
(CMC medium) is appropriated by users.
Given the above review, it seems that CMC has some
features that differ from FTF communication. However, these
features by themselves do not seem to decide the
communication outcomes of a medium. Rather, the users'
subjective perception of CMC features influence those
communication outcomes. For instance, factors such as ease
of use, effectiveness, and others would seem to be
subjectively constructed by users. It is possible that
perceived ease of use of a medium may be attributable to
users' knowledge and level of experience. For instance, a
user's perceived ease of use of a medium may change over
time. The underlying assumption is that users who use CMC
on a regular basis are more likely to have a more positive
perception of the medium on ease of use than those using the
medium for the first time. A requisite of impression
formation in CMC relational development relates to this
(Walther, 1992b). This position assumes that over time, a
less positive interpersonal relationship in CMC during the
initial stages of use becomes more positive and more
developed. Therefore, it would seem that any communication
outcome in CMC would be dependent on individuals' perception
of the medium. At the same time, how we perceive or view
objects is determined by who we are; thus, effects of
physical difference (i.e., gender) on users' perception will
be explored as well.
The following questions help direct explorations of how
users' perceptions of communication medium and their
physical differences influence the perceived communication
outcome variables:
(Q1) How do users' perceptions of a medium
influence users' perceptions of communication
outcome variables such as satisfaction, ease of
use, meeting effectiveness, immediacy, and
decision confidence?
(Q2) Does gender account for any significant
effect on users' perceptions of communication
outcome variables?
(Q3) How does interaction of gender and perceived
medium attributes influence users' ratings of
communication outcome variables?
Method
Participants
Participants consisted of 104 undergraduate student
volunteers from the University of Oklahoma. The students
were enrolled in several sections of communication classes.
The participants were given extra credit for participation
in the study. This was done in order to attract a diverse
group of subjects and to enhance active participation from
the participants. It is believed that students are active
participants in groups when they have a stake in their
performance, for instance when their grade depends on their
performance (see DeSanctis, in press; McGrath, 1984). The
experiment was conducted over a time frame of five weeks.
The study was conducted over a three week period; however, a
week interval was allowed between each CMC system. Prior to
the first week of the experiment, the participants were
assigned to 21 groups. Twenty groups consisted of five
individuals each, and one group consisted of 4 individuals.
As a result of the sample size, the group with four members
could not be avoided. Nevertheless, the group presented no
problem since individual level analysis was used for all
tests. Participant training was conducted separately for
each of the three conferencing systems. This was done prior
to the inception of the experiment for each conferencing
system.
The CMC Systems
The goal of this study is to make comparisons across
CMC systems. Therefore, three CMC systems were used. These
include the VAX system, the Quickmail conferencing system,
and the Ben system.
The VAX System: The VAX conferencing utility
facilitates a synchronous multi-node electronic message
capability that allows users to send and receive messages to
and from other participants. The VAX terminals were located
at the University computing service center. The room
contains 20 terminals that were arranged in a decision-room
type setting. Because the conferencing system uses headers
and banners that convey the identity of users, pseudonyms
were developed and assigned to each of the participants in
order to hide their real names. The VAX conference system
is equipped with a message notification system that informs
users of new messages. The system also allows individuals
to review all previous messages and provide appropriate
replies. The system is configured in such a way that all
messages are accessible to all group members. Thus, members
are exposed to all information upon which their discussion
could be based.
The Quickmail System: Quickmail is a communication
software system that facilitates synchronous conferencing.
It provides conferencing support for up to five individuals.
The software was available on Macintosh computers in one of
the University's computer laboratories. The room contains
20 terminals that are networked together with the aid of
Appletalk connectors in a decision-room type setting.
Quickmail requires a designated file server terminal that
allows participants to relay messages to one another.
Quickmail divides the computer screen into three panels.
The left top panel displays the pseudonyms of all users; the
lower left panel or the entry panel provides participants
the space to type their messages or reply. The right panel
is the transcript panel that allowed participants
simultaneous access to all members' messages by scrolling
back and forth. One thing that separates Quickmail from
most conferencing software is the provision of simultaneous
access to messages by participants without their having to
leave one screen for another. Like the VAX system,
Quickmail allows participants to review previous messages
and compose replies accordingly in a single process. With
the exception of the simultaneous access to a message as it
is being entered by other users, Quickmail offers the same
features as the VAX system.
The Ben System: The system consists of an electronic
mail system similar to the "Bitnet" system. The system is
available on the university's IBM 3081 mainframe computers
located in another room at the University computing
services. The Ben system is generally conducive to
electronic mail messages from one individual to another.
However, the system was reconfigured to facilitate group
interactions. This was done by creating a group
distribution list for each participating group, such that
when a group member sends a message, it is routed to all the
group members. The Ben conferencing system is a "menu"
driven system that displays help information and directions
in order to guide users in using the system. Although the
Ben system facilitates sending and receiving messages like
the other two systems, some of its features influence
interaction among participants. Unlike both VAX and
Quickmail, Ben system has longer paucity, or takes longer
than the other two systems to receive members' messages.
Furthermore, the system requires that each message line be
entered individually for inclusion, and editing work upon
message composure is problematic.
A second opinion was sought on the systems concerning
the ease of use or their user friendliness. Therefore, five
computer system analysts at the University computer services
were asked to try the three systems for the proposed tasks.
Upon completion of the tasks, they were asked to rate each
of the systems on ease of use. The following indicate the
average rankings on a five point scale with the high number
representing positive rating (Ben = 3.2; VAX = 4.2 and
Quickmail = 4.3). This ranking confirms the speculation of
VAX and Quickmail as being similar to one another and as
more interactive than the Ben system.
The experts' ratings of these systems seem to suggest a
possibility that differences should be expected between the
Ben system and the other two systems (VAX and Quickmail)
jointly; thus,there is a need for analysis between Ben in
comparison with Quickmail and Vax systems. However,
analysis that looked at each system independently was
chosen, in that the experts' ratings were based on only one
of the outcome variables of interest in this study.
Therefore, to conclude the equivalence of Vax and Quickmail
systems for all variables based on ease of use of the medium
alone may be premature, especially when measuring
participants' perception. Given the above difference
between Ben and other systems, it is possible that
participants may perceive the systems differently with
respect to other communication outcome variables. For
instance, participants are first time CMC users; therefore,
it is possible that their attitude is either negatively or
positively predisposed toward CMC. The predisposition
toward CMC may be intensified when users encounter problems
or situations that confirm their expectations while using
these systems. As a result, participants' systems ratings
on perceived ease of use of the medium and other variables
should reflect their individual differences and their
predispositions. Since gender is one such difference, an
attempt was made to determine any differences in perceived
systems' ratings between the two genders. Although
participants were randomly assigned to groups, the final
composition of each group was such that each group had at
least one member of both sex groups (see Hiltz et al.,
1986). The group composition further justifies the need to
explore gender effect and interaction of gender and CMC
systems on dependent variables.
Tasks & Procedures
Three equivalent decision-making (involving creativity
and intellectual skills) tasks[1] were used in the study
(see McGrath, 1984 task circumplex model). The first task
involved the "University Parking Crisis" where the
participants were asked to develop a proposal (as a selected
member of the University executive committee) to solve the
current parking problem on campus. The second task was the
"University Radio Station Management Problem." The students
were asked to develop a proposal on how the school radio
station ought to be managed in order to generate more
revenue for the university. The third task involved a
"Computer Literacy Problem." The participants were asked to
develop a proposal that would help students to become
computer literate. The argument for task equivalence was
based on the results of a pilot study that showed no
significant difference among tasks, and no effect was found
for tasks on perceived ease of use of CMC medium and other
criterion variables. These tasks were developed from some
issues of vital importance at the university at the time of
the experiment. These tasks were chosen to develop tasks
that students are familiar with, in order to enhance their
motivation to perform (e.g., Dennis et al., 1990/91).
Furthermore, different tasks were chosen to prevent boredom
and history effect as participants work from one CMC system
to another.
The three tasks were randomly assigned to each of the
three CMC systems in order to keep task and system
constant/consistent. The results showed that parking task
was assigned to the Ben system; the radio management task
was assigned to the Quickmail system; and the computer
literacy task was assigned to the VAX system. To control
against the "order effect," the 21 groups were divided into
thirds (seven groups each) and randomly assigned to start
the experiment on one of the three CMC systems. In other
words, one-third of the groups started the experiment with
Ben system, another one-third started with Quickmail, while
the final one-third started with the VAX system. The groups
were rotated with the systems until all groups had used the
three systems and completed all three tasks.
Each participant was assigned a workstation or terminal
from which to join their group members. Participants'
workstations were separated from one another to prevent
their staring at each other's terminals. Although the
number of workstations permit it, the experiments were run
nonconcurrenty for each group. On experiment days,
participants were trained on the use of the CMC system, and
they were given time to practice and familiarize themselves
with sending and receiving messages. The meeting took place
synchronously where members were working on the problem at
the same time and in the same room. While participants were
able to see one another, they were restricted from oral
communication. Interaction was limited to the computer
terminals. This was enforced by having the research
administrator monitor participants during experiments.
Participants were also instructed that failure to comply
with this procedure would result in grade forfeiture and
expulsion from the experiment. Upon completion of each
training session, participants were given some time to read
the problem task. In the statement of the problem, they
were instructed to approach each task using the following
structure: First, to individually list or brainstorm their
ideas; second, to pool all members' ideas for development of
a possible solution or proposal to the problem. The groups
spent between one and a half to two hours on each task.
After the completion of each problem task, participants
completed a questionnaire[2] concerning the medium they had
just used, with respect to its suitability on five dependent
variables. Therefore, each participant completed three
questionnaires in total. The questionnaire requires that
students provide their response on a five point Likert-type
scale ranging from strongly agree to strongly disagree.
High scores indicate positive score while lower scores
indicate negative score. The questionnaire was derived from
Walther's (1989) satisfaction measurement in CMC. The
questionnaire consisted of five sub-scales: satisfaction,
ease of use, effectiveness, immediacy, and decision
confidence. The reliability for these scales, using a
Chronbach alpha technique were: satisfaction = .85, ease of
use = .91, decision confidence = .86, immediacy = .80, and
effectiveness = .81.
Conceptual Issues
Given the differential operationalization of concepts
and designs between the laboratory experiments and field
studies, Dennis et al. (1990/91) provide the following
recommendations for enhancing uniformity. Some of these
include: establishment of an ongoing group, the use of a
diverse subject pool, increased group size, use of familiar
and preexisting task, use of incentives for motivating
participants, and use of different incentives to different
participants in laboratory research. Therefore, some of the
factors that were directly under the control of the
researchers were controlled for. The use of diverse subject
pool and familiar and preexisting group tasks was heeded
accordingly. The recommendation that different incentives
for different participants be used was manipulated by
informing the participants that the amount of credit
received will be subject to the quality of their
participation. The participants were measured over three
different time intervals. Furthermore, other physical
characteristics such as age, educational background, and
status were not controlled for in this study because these
factors provide little (if any) variations among the
participants. For instance, the participants' ages ranged
between 20 and 22. At the same time, status and educational
background differences are less likely to be a factor among
student participants as they are still pursuing their
degrees, therefore justifying the examination of gender
effect as the only individual characteristic variable
explored in this study.
Analyses & Results
First, to verify the assumption of dependency among the
dependent variables, a correlation analysis was computed,
and the result showed that the dependent variables were
indeed related (see table 1 for correlation matrix). Since
the dependent variables are related, the use of the
multivariate analysis of variance (MANOVA) was justified for
analyzing the proposed research questions. Therefore, the
MANOVA model was a 3 x 2 design, with the independent
variables being perceived CMC systems' attributes (Ben, VAX,
and Quickmail) and gender (males and females). The
dependent variables consisted of the five communication
outcome variables: perceived satisfaction, effectiveness,
ease of use, immediacy, and decision confidence. A
preliminary analysis of variance was computed, however, as
one might expect in a randomized group, no effect was found
for gender. This procedure justifies the use of
individual-level analyses of data. Therefore, all tests
were based on individual-level differences. The following
procedures were used. The multivariate tests were examined,
and if the tests were significant, the univariate results
were also examined.
============================================================
Table 1
Intercorrelation Matrix for Dependent Variables
Variables: Dec. Conf. Satis. Immed. Ease. Effect.
Decision Conf. .
Satisfaction -.186* .
Immediacy .129 .286* .
Ease of Use .07 .327* .475* .
Effectiveness -.027 .526* .556* .527* .
* p < .01
===========================================================
Q1
In order to determine whether perceived CMC medium
attributes influence participants' perception of the
communication outcome variables, the multivariate test for
system's effect was examined. The multivariate test showed
a significant effect for CMC systems (Wilks = .89; F[10,
908] = 5.21, p < .0001, h2 = .11). Four univariate tests
were significant for the following communication outcome
variables: decision confidence (F[5, 458] = 7.85, p < .001,
eta2 = .94); satisfaction (F[5, 458] = 12.04, p < .0001,
eta2 = .99); immediacy (F[5, 458] = 3.07, p < .03, eta2 =
.94); effectiveness (F[5, 458] = 8.06, p < . 001, eta2 =
.97). The fifth univariate effect, "ease of use,"
marginally approached significance as (F[4,458] = 2.56, p <
.07, eta2 = .98). These results seem to indicate that the
perception of medium attributes influence participants'
ratings of the CMC system on communication outcome
variables.
In order to determine the nature of this effect for
different systems, the means were examined. Therefore, the
overall average means and the average means for each of the
three systems were examined. The mean for decision
confidence was 3.88; however, the means across the three
systems were: Ben = 3.28, SD = .60; VAX = 3.9, SD = .98;
Quickmail = 3.8, SD = 1.15). The means were examined for
significant difference using the least square means
(LSMEANS) method, and the result indicated that the means
for VAX and Quickmail system were significantly different
from the Ben system but not significantly different from one
another. The mean for user satisfaction was 3.58, while the
means across systems were: Ben = 3.2, SD = .68; VAX = 3.74,
SD = .92; Quickmail = 3.82, SD = 1.08. A significant
differential pattern similar to decision confidence was
found for satisfaction. For immediacy, the overall mean was
2.91. Across the systems, the means were: Ben = 2.6, SD =
1.14; VAX = 3.21, SD = 1.5; Quickmail = 3.33, SD = 1.37.
Analysis of the means showed that VAX and Quickmail systems
were significantly different from the Ben system while they
do not differ from one another. Again, this pattern was
similar to those of decision confidence and user
satisfaction. For effectiveness, the overall mean was 3.21.
The means across the system were: Ben = 3.6, SD = .70; VAX =
3.11, SD = 1.01; Quickmail = 3.16, SD = .96. Analysis of
the means showed that VAX and Quickmail systems were
significantly different from the Ben system, while the two
were not significantly different from each other. However,
this result differs from previous trends, in that the Ben
system was perceived to be more effective than both Vax and
Quickmail systems. On ease of use, the overall mean was
2.78. Across the three systems, the means were: Ben = 2.73,
SD = .84; VAX = 2.7, SD = .84; Quickmail = 2.89, SD = .96.
The means' analyses indicated that VAX and Quickmail systems
were significantly different from one another. However,
neither was significantly different from Ben system (see
Table 2 for the means analyses and level of probability
associated with the significance).
===========================================================
Table 2
Mean Comparisons by CMC System & Gender
p Levels
Variables Systems Means Vax Ben QM
Dec Conf. Vax 3.91 _
Ben 4.29 .01 _
QM 3.75 _ .001 _
Satisf Vax 3.44 _
Ben 4.07 .001 _
QM 3.49 _ .001 _
Immediacy Vax 3.21 _
Ben 2.61 .05 _
QM 3.33 _ .001 _
Effectiveness Vax 3.11 _
Ben 3.65 .001 _
QM 3.17 _ .001 _
Ease of use Vax 2.70 _
Ben 2.73 _ _
QM 2.89 .03 _ _
_Variables Gender Means Male Female_
Dec. Conf. Male 3.74 _
Female 4.22 .001 _
Immediacy Male 2.85 _
Female 3.23 .01 _
Effectiveness Male 3.17 _
Female 3.45 .01 _
Dec. Conf. = Decision confidence, Satisf = satisfaction,
Effect = effectiveness.
_ = p > .05
===========================================================
Q2
This question attempts to determine if gender
influenced the communication outcome in a CMC meeting. In
order to answer this question, the multivariate test for
gender effect was examined. The multivariate test was
significant for gender (Wilks = .89; F[5, 454] = 10.99, p <
.0001, h2 = .11). Three univariate effects were found:
decision confidence (F[5, 458] = 22.28, p < .0001, eta2 =
.95), and the means were males = 3.74, SD = .97; females =
4.22, SD = .89 respectively; immediacy (F[4, 458] = 7.13, p
< .01, eta2 = .94) and the means were males = 2.85, SD =
.1.21; females = 3.23, SD =1.3; and effectiveness (F[5, 458]
= 7.5, p < .01, eta2 = .95) and the mean scores were males =
3.16, SD = .96; females = 3.45, SD = 1.05.
Furthermore, the means were compared for statistical
significance using the LSMEANS techniques. The results
showed that the mean scores by females were significantly
different from those of males for decision confidence,
immediacy, and effectiveness. However, there were no
differences found for gender on perceived users'
satisfaction and ease of use of the medium. The pattern
was reflected in the mean scores' distribution across the
three CMC systems as well (see table 3). Therefore, the
results showed some support for the notion that gender
influences participants' perception of the CMC attributes
and communication outcomes.
===========================================================
Table 3
Mean Comparison By Interaction of Systems & Gender
Variables Sys. Means BM BF VM VF QM QF
BM 3.66 .
BF 4.14 .01 .
Dec. Conf. VM 3.88 _ _ .
VF 4.7 .01 .01 .01 .
QM 3.66 _ .01 _ .01 .
QF 3.83 _ .03 _ .01 _ .
BM BF VM VF QM QF
BM 3.39 .
BF 3.48 _ .
Satisf. VM 4.00 .01 .01 .
VF 4.15 .01 .01 _ .
QM 3.44 _ _ .02 .03 .
QF 3.53 _ _ .02 .02 _ .
BM BF VM VF QM QF
BM 2.80 .
BF 2.78 _ .
Immediacy VM 2.60 _ _ .
VF 4.00 .01 .01 .01 .
QM 3.14 _ _ .04 .03 .
QF 2.91 _ _ _ .01 _ .
BM BF VM VF QM QF
BM 3.03 .
BF 3.20 _ .
Effect. VM 3.30 _ _ .
VF 4.00 .01 .01 .01 .
QM 3.17 _ _ _ .01 .
QF 3.16 _ _ _ .01 _ .
BM BF VM VF QM QF
BM 2.98 .
BF 2.81 _ .
Ease of Use VM 2.64 .01 _ .
VF 2.74 _ _ _ .
QM 2.91 _ _ _ _ .
QF 2.55 .02 _ _ _ _ .
BM = Ben male, BF = Ben female, VM = Vax male, VF = Vax
female, QM = Quickmail male, and QF = Quickmail female.
_ = p > .05.
===========================================================
Q3
This question attempts to determine whether there is an
interaction effect between the medium attributes and gender.
To answer this question, the multivariate test was examined.
The results showed significant effect (Wilks = .88; F[10,
908] = 5.57, p < .001, eta2 = .12). Three univariate
effects were found: decision confidence (F[5, 458] =
3.21, p < .05, eta2 = .92), immediacy (F[4, 458] = 9.
55, p < .001, eta2 = .97), and effectiveness (F[4,
458] = 3.31, p < .05, eta2 = .95). Like gender, no
interaction effect was found for either user satisfaction
and ease of use. The mean scores' distribution is presented
in table three. Therefore, these findings support that
there is an interaction effect for gender and medium
attributes on communication outcomes.
Discussion
The results indicate how users' perceptions of CMC
medium influence the participants' perceptions of the
communication outcomes. This effect was found for all the
dependent variables (e.g., perceived users' satisfaction,
immediacy, effectiveness, and the ease of use a CMC medium).
The findings from the second research question however,
helped shed some light on the nature of the influence.
Analyses of the means across the three systems indicate the
following. Given the level of confidence attributable to
the decision reached by participants in their groups, the
results showed that participants attribute higher levels of
confidence to the decision reached while using either the
VAX or the Quickmail system (3.9 and 3.8) than they did with
the Ben system (3.28 on a 5 point rating scale). Also, the
results indicate that the VAX and Quickmail systems were
perceived by the participants to be equivalent. A pattern
similar to decision confidence was found for perceived
immediacy and user satisfaction. That is, participants
attributed relatively equal ratings to VAX and Quickmail
systems, and these ratings were significantly higher for
both VAX and Quickmail systems than the Ben system (see
table 2).
The implication that may be drawn is that with a system
that is perceived to be less likely to facilitate smooth
interaction as the Ben system (by virtue of being low on
immediacy and ease of use), participants may be tempted to
do whatever is necessary to get by (e.g., rushing through
the activity or selecting the first available decision
proposal) rather than putting forth their best effort.
Thus, participants are likely to express low satisfaction
and confidence in their decision and performance with such
systems. This argument extends those made by Smith and
Vanecek (1988) for the simultaneity of messages and
interaction effect on satisfaction. They indicated that
when there is a lack of opportunity in a CMC meeting for
instant feedback or "immediate message clarification" (as
with the Ben system), a reduced shared perception concerning
progress towards group goals exists, thus creating reduced
satisfaction (see also Smith & Vanecek, 1990).
A surprise finding from the analysis of the means and
perhaps for the entire study is the results on the "ease of
use of the medium." Although evidence showed that despite
the fact that users perceived both VAX and Quickmail
favorably and as equivalent on other outcome variables,
there was a statistical significance difference found
between the two systems (see table 2). The means score
distribution showed that Quickmail was statistically
different and perceived relatively higher on ease of use
than VAX system, while neither of the two systems differ
from the Ben system. Apart from the the statistical
difference, the means distribution scores showed that all
the systems were perceived relatively low on ease of use.
This effect perhaps could be attributed to "change effect"
or newness to the system (see Rogers, 1983; see also,
Olaniran, 1991a, 1991b). For instance, this study engaged
first-time CMC users; therefore, there is a tendency for
resistance as the traditional face-to-face interaction may
still represent the norm on which their subjective ratings
of the CMC systems were based. Fulk, Schmitz, and
Steinfield (1990) reasoned along the same line when they
indicated that "cognitions" arise from prior history of
social interaction. This effect might change, as users
engage in repeat use and become experienced with these
systems over a period of time (see Walther, 1991b).
Based on this over-time change in perception argument,
an exploratory trend analysis was conducted across systems
to determine changes in users' perceptions over the course
of the experiment. This analysis was possible since each
participant participated in three sessions over time.
Therefore, a trend analysis of the mean distribution scores
was calculated for the criterion variables. The results
seemed to lend some support to the speculation that users'
perceptions change over time. This effect, however, was
mixed for perceived ease of use of the medium. The average
mean scores were 2.69, 2.9, and 2.73 (on a five point scale)
from time 1 through time three respectively. While the
means tend to follow the expected pattern from time 1 and
time 2 (at p < .03), there was a directional decline at time
3 where the mean scores revert to a level similar to that of
time 1. These results seem to suggest that perceptual
changes over time may be nonlinear.
Another result that was not anticipated yet is
significant is the perceived effectiveness of the medium.
The results showed that Ben system, which was perceived
lower than the other two systems on user satisfaction,
decision confidence, and immediacy, was perceived higher on
effectiveness than the other two systems. In order to
explain this finding, the means were examined for gender
effect across the systems. The mean distribution scores
indicate that women perceived the Ben system to be more
effective than men (4 to 3.3). The mean comparison
indicated that the mean score for females was statistically
different from any of the other systems even when compared
to themselves on the other two systems (see table 2 and 3).
An explanation that could be offered concerns speculation
that women may prefer less interaction. That is, women may
prefer less interaction and opportunity to work individually
and at their own pace (as is likely the case with Ben
system) such that they felt that this system was more
effective than the other systems that facilitate greater
interaction and, consequently, greater interruptions by
other participants. One feature of the Ben system that
makes it less interactive than the other systems is the
capacity that allows its users to ignore other participants'
messages (selective message process). This capacity also
provides each participant the opportunity to structure
his/her own performance within the group, therefore giving
the users autonomy on how to progress or approach the group
task without having to look to other participants for
approval. This inference may have some merit even though
research on interruptions and gender have been mixed. Some
studies found that women were being interrupted more often
by men (e.g., Eakins & Eakins, 1978; Hall, 1984), while
others found no difference in gender interruption behaviors
(Conger & Dindia, 1985; Smythe & Schlueter, 1986). However,
while there may be no difference in actual interruption
behaviors of men and women, there may be differences in the
perceived interruption behaviors by gender. Smith-Lovin and
Brody (1989) found that women in token group compositions
are three times more likely to "yield the floor" with
negative interruptions than men. Therefore, if women
believe that they are being interrupted more often by men,
there are bound to be differences. This view is consistent
with the Smith and Vanecek (1990) claim that "perceived
freedom to participate may be a better indicator of
satisfaction for the group experience than a measure of the
actual amount of participation per participant" (p. 75).
Nevertheless, one would hope that features of CMC
believed to facilitate uninhibited and equal participation
among users will eliminate this effect. But, this may not
necessarily hold true at the initial stage, especially given
the participants' newness to the CMC technology where
face-to-face interaction may still represent the standard by
which CMC interactions are measured. However, this
perception should dissipate with time and continuous usage
as users become familiar with the system and understand the
system's attributes and usage more comprehensively. An
exploration of means for over-time change lends support for
this speculation. Exploration of the mean scores over time
by gender was more revealing. The mean scores for men on
perceived effectiveness ranged from 3.05 at time 1 to 3.3 at
time 3 with no significant difference between time 1 and 2.
The scores for women ranged from 3.11 at time 1 to 4.01 at
time 3; there was no significant difference between time 1
and time 2 as well. In order to verify that the gender
effect was non-spurious, one would assume that similar
gender effects would be reflected among other dependent
variables, especially since the dependent variables are
related. Thus, other dependent variables were examined, and
the results showed that women also perceived the Ben system
higher on decision confidence and immediacy than men when
comparisons were made across the other two systems (see
table 2 and 3).
If women perceived themselves to be frequently
interrupted by men, it can be argued that women may view
their degree and quality of participation as limited in a
FTF interactive mixed group. Therefore, it would seem that
some CMC features and designs may enhance or facilitate
greater perceived women participation in mixed groups,
especially if the system minimizes interaction or enhances
more individualized participation in the group. Thus, CMC
structures that facilitate anonymity, such as synchronous
but different physical locations or asynchronous meetings,
might appeal to women in mixed groups. On the other hand,
if minimal interaction among participants is a cherished
attribute in CMC system by women, the effect doesn't seem to
be carried over to their perceived satisfaction and ease of
use of CMC medium.
Two explanations could be offered for the result on
perceived satisfaction. First, it is quite possible that
changes or different appropriations by users occur over
different time periods especially for a variable like
satisfaction. Again, over-time mean change was explored.
The results showed that the mean for perceived satisfaction
ranged from 3.4 at time 1 to 4.1 at time 3. There were no
significant changes between time 1 and time 2, while there
was a significant change in time 1 and time 3 and time 2 and
time 3 (P < .01). The changes in mean distribution were in
the right direction, thus supporting the notion of
perceptual change over time. This finding seems to be
consistent with Chidambaram et al. (1990/91) on
appropriation of group outcome variables by participants.
Second, men and women do not differ in their perceived
satisfaction on CMC medium. The merit of this argument lies
in the fact that when gender was taken into account, men and
women do not differ in their mean scores on perceived
satisfaction. Furthermore, the results offer insight on
short-term versus long-term usage of CMC system. It has
been shown that satisfaction increases with CMC over a
period of time (e.g., Hiltz & Turoff, 1985; Walther, 1991a,
1991b,). That is, as participants engage in continuous and
frequent usage of a CMC system, they may find the system
more interpersonally satisfying as they concentrate less on
task oriented activities or find other usage for the system.
Thus, one would expect that, in time, participants would
make different appropriations with communication outcomes as
they gain more understanding of the CMC medium. In other
words, media perceptions are not fixed; rather, they vary
across contexts and individuals. This argument also
corresponds to the one raised earlier concerning the change
effect or users' newness to CMC system.
Furthermore, a closer look at the means distribution
scores showed a tendency that women, in general, were more
positive in their ratings of the CMC systems on all but one
communication outcome variable (ease of use) than their male
colleagues. This tendency was, however, more pronounced for
3 variables which are: decision confidence, immediacy, and
effectiveness. Therefore, both perceived user satisfaction
and ease of use of medium to some degree depend on how a CMC
system is perceived (see Hiltz, 1988; Hiltz & Turoff, 1985).
This study demonstrates that while CMC system and gender
showed independent effects on perceived communication
outcome variables (e.g., decision confidence, effectiveness,
and immediacy), there was also an interaction effect of
gender and CMC systems on these variables.
In general, the findings from this study lend support
to the argument that participants' perceptions of CMC medium
attributes influence communication outcomes more than the
inherent physical characteristics of the medium (see
Chidambaram et al., 1990/91). Furthermore, support was
found for Rice's (1987) claim that ease of use of a medium
influences user satisfaction. However, there were no
indications that perceived ease of use of a CMC medium
influences the effectiveness of CMC system. Therefore, it
seems that there are other factors besides ease of use of a
system that determine participants' perceptions of a CMC
system's effectiveness.
Research Limitations and Future Directions
A potential limitation of this study has to do with
the confounding of task and CMC systems. Therefore, the
extent to which tasks contribute to the effect found for
medium attributes is not known. The quality of decision
was not addressed in this study. However, research might
benefit from knowing how these systems affect decision
quality. A case in point is the speculation raised
concerning women's preference for a less interactive medium.
The literature will benefit from future investigation of
this issue.
Furthermore, the findings here raise an issue for
future investigations of how individual differences (e.g.,
gender) influence Easton, et al.'s (1990/91) argument that
an interactive medium is advantageous for tasks where groups
are seeking the best unanimous group decision through
consensus or processes leading to it. If perceived freedom
to participate differs by gender as speculated, then women
may perceive a false notion of consensus, especially in a
mixed group where women are the minority. This is likely to
be the case, as women may decide to go along with the
dominant members' views either in an attempt to complete or
terminate the group process as quickly and painlessly as
possible. While this effect is bound to occur in a FTF
interactive mixed group, similar effects cannot be ruled out
in synchronous decision room type settings where nonverbal
information from artifacts or other cues from the physical
presence of other members are discernible. This factor
could affect the perception of other communication outcome
variables.
In light of gender effects found in this study, an
issue that may be raised concerns the need for differential
structures in CMC systems that will mediate the differences
in male's and female's usages. This is essentially
important in that, while variables such as gender can be
controlled in laboratory studies they are difficult to
control in field studies, thus becoming issues that need to
be addressed. Perhaps it may not be so much a problem of
developing different systems but rather provisions of
informational handling and organizational options from which
individual users (regardless of gender) can adopt to suit
their personal communication styles and preferences (see
Hilt & Turoff, 1985). Borgman (1989) suggests the use of
systems targeted for specific user groups, or "multiple
interfaces" with a single system, that are adaptable to user
characteristics in circumventing some of these problems.
The development of such options may not only facilitate
positive perception, but may also give individuals more
control in the communication process.
Finally, it has been shown that perception of CMC
medium attributes influences communication outcomes among
users. It has also been shown that perceptions differ with
individual variables (e.g., gender). However,
generalization from this study should be done with caution,
given the nature of the participants in this study. The
participants in this study were first time CMC users; thus
these findings may not be generalizable to other contexts
where the participants are experienced users. Therefore,
generalization, at best, should be limited to similar
participants and similar systems.
Notes
[1] Copy of the tasks are available from the author by
request.
[2] The questionnaire is available from the author by
request.
References
Borgman, C. L. (1989). All users of information retrieval
systems are not created equal: An exploration into
individual differences. Information Processing &
Management, 25, 237-251.
Bui, T., & Sivasankaran, T. R. (1990). Relation between
GDSS use and group task complexity. Proceedings of the
Twenty-Third Hawaii International Conference on Systems
Sciences (69-78).
Chidambaram, L., Bostrom, R. P., & Wynne, B. E. (1990/91).
A longitudinal study of the impact of group decision
support systems on group development. Journal of
Management Information Systems, 7(3), 7-25.
Conger, J. L., & Dindia, K. A. (1985). A functional
approach to interruptions. Paper presented at the
annual conference of the International Communication
Association, Honolulu, Hawaii.
Dennis, A, R., Nunamaker, J. F., & Vogel, D. R. (1990/91). A
comparison of laboratory and field research in the
study of electronic meeting systems. Journal of
Management Information Systems, 7(3), 107-135.
Dennis, A. R., Nunamaker, J. F., & Paranka, D. (1991).
Supporting the search for competitive advantage.
Journal of Management Information Systems, 8, 5-36.
Dennis, A., George, J., & Vogel, D. (1988). Information
technology to support electronic meetings. MIS
Quarterly, 12, 591-619.
DeSanctis, G., & Gallupe, R. B. (1987). A foundation for
the study of group decision support systems.
Management Science, 33, 589-609.
DeSanctis, G., (in press). Small group research in
information systems: theory and method. In I. Benbasat
(Ed.), Harvard Business School colloquium on
experimental methods in information systems. Boston,
MA: Harvard Business School Press.
Eakins, B. W., & Eakins, R. G. (1978). Sex differences in
human communication. Boston, MA: Houghton Mifflin.
Easton, A. C., Vogel, D. R., & Nunamaker, J. F., (1989).
Stakeholder identification and assumption surfacing in
small groups: An experimental study. Proceedings of
the Twenty-Second Hawaii International Conference on
Systems Sciences (344-352).
Easton, G. K., George, J. F., Nunamaker, J. F., &
Pendergast, O. M. (1990). Using two different
electronic meeting system tools for the same task: An
experimental comparison. Journal of Management
Information Systems, 7(1), 185-100.
Edinger, J. A., & Patterson, M. L. (1985). Nonverbal
involvement & social control. Psychological Bulletin,
93, 30-56.
Ekman, P., Friesen, W. V., O'Sullivan, M., & Scherer, K.
(1980). Relative importance of face, body, and speech
in judgments of personality and affect. Journal of
Personality and Social Psychology, 38, 270-277.
Fisher, B. A., & Ellis, D. G. (1990). Small group decision
making: Communication and the group process (3rd ed.).
New York: McGraw Hill.
Fulk, J., Schmitz, J., & Steinfield, C. W. (1990). A social
influence model of technology use. In J. Fulk & C.
Steinfield (Eds.), Organizations and communication
technology (pp. 117-140). Newbury Park, CA: Sage.
Gallupe, R. B., & McKeen, J. D. (1990). Beyond
computer-mediated communication: An experimental study
into the use of group decision support systems for
face-to-face versus remote meetings. Information &
Management, 18, 2.
Gallupe, R. B., DeSanctis, G., & Dickson, G. W. (1988).
Computer-based support for group problem finding: An
experimental investigation. MIS Quarterly, 12,
277-296.
George, J. F., Easton, G. K., Nunamaker, J. F., JR., &
Northcraft, G. B. (in press). A study of collaborative
group work with and without computer based support.
Information Systems Research.
Hall, J. A. (1984). Nonverbal sex differences:
Communication accuracy and expressive style.
Baltimore, MD: Johns Hopkins University Press.
Heimstra, G. (1982). Teleconferencing concern for face, and
organizational culture. In M. Burgoon (Ed.),
Communication Yearbook, 6, (pp. 874-904). Beverly
Hills, CA: Sage.
Hiltz, S. R. (1988) Productivity enhancement from
computer-mediated communication: A systems contingency
approach. Communication of the ACM, 31, 1438-1454.
Hiltz, S. R., Johnson, K., & Turoff, M. (1986). Experiment
in group decision making communication process and
outcome in face to face vs. computerized conference.
Human Communication Research, 13, 225-252.
Hiltz, S. R., & Turoff, M. (1985). Structuring computer
mediated communication systems to avoid information
overload. Communication of the ACM, 28, 680-689.
Jaspars, J., Hewstone, M., & Fincham, F. D. (1983).
Attribution theory and research: The state of the art.
In J. Jaspars., F. D. Fincham., and Hewstone, M.
(Eds.), Attribution theory and research: Conceptual
developmental and social dimensions (pp. 3-36).
London: Academic Press.
Kelley, H. H., & Michela, J. L. (1980). Attribution theory
and research. In M. R. Rosenzweig and L. M. Porter
(Eds.), Annual Review of Psychology, 31, 457-501. Palo
Alto: Annual Review Inc.
Kiesler, S., Siegel, J., & McGuire, T. W. (1984). Social
psychological aspects of computer mediated
communication. American Psychologist, 39, 1123-1134.
Kraemer, K. L., & Pinsonneault, A. (1990). Technology and
groups: Assessment of the empirical research. In J.
Galegher., R. Kraut., & Egido, C. (Eds.), Intellectual
teamwork: Social and technological foundations of
cooperative work (pp.. 373-404). Hillsdale, NJ:
Lawrence Erlbaum Associates, Publishers.
Maier, N. R. F. (1970). Problem solving and creativity in
individuals and groups. Belmont, CA: Brooks-Cole.
McGrath, J. E. (1984). Interaction and performance.
Englewood Cliffs, NJ: Prentice Hall.
Mehrabian, A. (1972). Nonverbal communication. Chicago,
IL: Aldine.
Olaniran, B. A. (1991a). Managing organizational change: A
step by step approach. Eighth Annual Southwest
Business Symposium Proceedings (pp. 307- 316). Edmond,
Oklahoma.
Olaniran, B. A. (1991b). Computer-mediated communication in
small group decisional stages. Unpublished doctoral
dissertation. University of Oklahoma, Norman,
Oklahoma.
Poole, S., & DeSanctis, G. (1990). Understanding the use of
group decision support systems: The theory of adaptive
structuration. In J. Fulk & C. Steinfield (Eds.),
Organizations and communication technology (pp.
173-193) Newbury Park, CA: Sage.
Rice, R. E. (1984). Mediated group communication. In R. E.
Rice & Associates (Eds.), The new media: Communication,
research, and technology (pp. 129-156). Beverly Hills,
CA: Sage.
Rice, R. (1987). Computer-mediated communication and
organizational innovation. Journal of Communication,
37(4), 65-94.
Rice, R. E. (1980). The impact of computer mediated
organizational and interpersonal communication. Annual
Review of Information Science and Technology, 15,
221-249.
Rice, R. E., & Love, G. (1987). Electronic emotion:
Socioemotional content in a computer-mediated network.
Communication Research, 14, 85-108.
Rogers, E. M. (1983). Diffusion of innovations (3rd ed.).
New York: Free Press.
Ruchinkas, J. R. (1982). Communicating in organizations:
The influence of context, job, task, and channel.
Unpublished doctoral dissertation, University of
Southern California.
Sambamurthy, V. (1988). A comparison of two levels of
computer-based support for communication and conflict
management in equivocality reduction during stakeholder
analysis. Doctoral dissertation proposal, Carlson
School of Management, University of Minnesota.
Short, J., Williams, E., & Christie, B. (1976). The social
psychology of telecommunications. London: John Wiley &
Sons.
Siegel, J., Dubrovsky, V., Kiesler, S., & McGuire, T. W.
(1986). Group processes in computer mediated
communication. Organizational Behavior and Human
Processes, 37, 157-187.
Smith, J. Y., & Vanecek, M. T. (1988). Computer
conferencing and task oriented decisions: Implications
for group decision support. Information & Management,
14, 123-132.
Smith, J. Y., & Vanecek, M. T. (1990). Dispersed group
decision making using non simultaneous computer
conferencing: A report of research. Journal of
Management Information System, 7, 71-92.
Smith-Lovin, L., & Brody, C. (1989). Interruptions in group
discussions: The effects of gender and group
composition. American Sociological Review, 54,
424-435.
Smythe, M., & Schlueter, D. W. (1986). Can we talk? A
meta-analytic review of sex differences in the language
of literature. Paper presented at the annual
conference of the Organization for the Study of
Communication, Language and Gender. Washington, D. C.
Sproull, L., & Kiesler, S. (1986). Reducing social context
cues: Electronic mail in organizational communication.
Management Science, 32, 1492-1512.
Steeb, R., & Johnston, S.C. (1981). A computer-based
interactive system for group decision making. IEEE
Transactions on Systems, Man, and Cybernetics, SMC-11,
544-552.
Steinfield, C. W. (1986). Computer-mediated communication
in organization setting: Explaining task related and
socio emotional uses. In M. L. McLaughlin (Ed.),
Communication Yearbook 9, 777-804. Newbury Park, CA:
Sage.
Svenning, L. (1982). Explaining predispositions toward
telecommunication innovations: The case of video
conferencing. Unpublished doctoral dissertation.
University of Southern California.
Walther, J. B. (1989). The measurement of satisfaction in
computer meetings systems. Unpublished manuscript,
University of Arizona, Tucson, Arizona.
Walther, J. B. (1992a). Interpersonal effects in
computer-mediated communication: A relational
perspective. Communication Research, 19, 52-90.
Walther, J. B. (1992b). A longitudinal experiment on
computer-mediated and face to face interaction.
Proceedings of the twenty-fifth Hawaii International
Conference on System Sciences. Los Alamitos, CA: IEEE
Computer Society Press.
============================================================
*Acknowledgements: The author would like to extend his
gratitude to Ronald Rice (Rutgers University) for assistance
provided in revision of this manuscript. Also, the author
extends his gratitude to the following people: Robert
Stewart and David Williams (Texas Tech University) for their
help in preparation of this manuscript. An earlier draft of
this manuscript was presented at the 1993 Western Speech
Communication Association Convention Albuquerque, New
Mexico.
------------------------------------------------------------
Copyright 1993
Communication Institute for Online Scholarship, Inc.
 CIOS Support Staffsupport@cios.org
Branch to CIOS home page
|
