Information Technology Reference
In-Depth Information
perceived and objective decision qualities was
-0.0604. This simply indicates no correlation
between them. The significance of this finding is
that perceived decision quality cannot be used as
a surrogate measure for objective decision qual-
ity in all situations. It must be used with caution.
Any study that measures only perceived decision
quality must make clear why objective decision
quality cannot be measured. It also should state
how perceived decision quality can be used as a
surrogate measure for objective decision quality.
Otherwise, the findings of studies will be mis-
leading. Failure to distinguish clearly between
perceived and objective decision quality may have
contributed to inconsistent findings on decision
quality in previous GSS research.
With regard to satisfaction with the decision
process, there are significant differences in the
communication mode and the presence of a group
leader. Parallel communication mode groups re-
ported a higher level of satisfaction with decision
process than sequential communication groups,
while groups with a leader reported a higher level
of satisfaction with decision process than groups
without a leader. Further analysis with Fisher's
LSD indicates that sequential communication
groups with a leader report a higher level of satis-
faction with the decision process than sequential
communication groups without a leader.
One important finding in satisfaction measure
is that parallel communication groups generally
indicate a higher level of satisfaction than sequen-
tial communication groups. The results of satis-
faction with the decision process are particularly
interesting. Because of novelty effect (Watson,
DeSanctis, & Poole, 1988), parallel communica-
tion groups were expected to show a lower level
of satisfaction with the decision process. These
groups, however, reported higher satisfaction
than sequential communication groups with
more commonly used face-to-face interaction. It
appears that as the use of multiple topics and mul-
tiple-participant CMCS such as e-mail or instant
messengers becomes pervasive, sequential com-
munication no longer may be the preferred mode
of communication, thereby dissipating the novelty
effect. This opens up a new avenue for the use of
GSS in different settings where it may be used as
a knowledge management tool or organizational
memory and not be limited to task-oriented groups
or decision-making groups.
The equality of participation was measured
objectively, by counting the number of groups
in which an informal leader emerged, and sub-
jectively with a questionnaire. The emergence
of an informal leader was used as a surrogate
measure for equal participation because it is
linked to participation rate (Mullen, Sales, &
Drisekll, 1987), and the rule of 50% or more
comments than the group's average was used to
determine the emergence of an informal leader
(Hiltz, Johnson, & Turoff, 1982). The commu-
nication mode made no significant difference in
either case. However, the number of groups where
informal leaders emerged was significantly lower
in groups with a leader than in groups without a
leader. Ho and Raman (1991) and Lim et al. (1994)
argue that the emergence of informal leaders is
low in groups with a formal leader who tends to
dominate the group process. This may impede a
group member from accumulating idiosyncrasy
credits that eventually can confer leadership status
(Hollander, 1974). For the subjective measure of
equal participation, neither the communication
mode nor the presence of a group leader made
any difference. All groups felt that participation
was equal. This finding is in line with previous
findings that CMCS increases the level of equal
participation by reducing or eliminating the social
influence of communication (Steiner, 1972).
The simple presence of a leader, however, did
not make much difference. What was more im-
portant was whether leaders properly performed
their leadership roles. To investigate the group
leaders' performances, the contents of the group
leaders' comments were analyzed to determine
whether they were task-related or leadership-re-
lated, as summarized in Table 5. In doing so, all
Search WWH ::
Custom Search