THE EFFECT OF COMPUTER-BASED SUPPORT AND ELECTED LEADERSHIP ON INFLUENCE ATTEMPTS IN SMALL GROUP DECISION-MAKING
LIM LAI HUAT
LIM LAI HUAT
Citations
Altmetric:
Alternative Title
Abstract
Group decision support systems (GDSS) for supporting problem formulation and solution in group meetings [DeSanctis and Gallupe, 1987], have received rapidly increasing attention from researchers and practitioners in recent years. A cumulative body of experimental research that examines the effects GDSS has on the quality of decision making has emerged over the decade. Lately, there has been a shift in emphasis from examining the effects of GDSS on the decision outcomes alone, to examining the effects on the interaction process within group, recognizing that the former is actually a product of the latter. Using a two-by-two factorial experiment, this thesis examined the effects of the decision technology and leadership on attempted influence behavior in five-member group interaction. The decision technology variable was varied at two levels, GDSS support and paper-and-pencil support structurally equivalent to a GDSS, i.e. manual support. The reason for experimenting with the two levels of support was to segregate the effects of electronic communication provided by GDSS and the imposition of structure, so that any observed effects may be attributed towards GDSS alone. Leadership was also varied at two levels, without-elected-leader and with-elected-leader, where the elected leaders were given four specific responsibilities to perform. The task employed was a preference allocation task that required participants lo do fund allocation based on personal values. Three dependent variables were measured, all of which centered around influence behavior, an important aspect of group decision process. Another reason for looking into influence behavior is the tight coupling between leadership and influence; specifically, in many instances, leadership has been studied as an exercise of influence. Dependent variables measured were amount of influence behavior, influence distribution, and dominance significance. Putnam's [1981] procedural messages coding system was used to analyze the micro-level interaction. The findings of this experiment show that GDSS groups generated a larger amount of influence behavior than the manual groups, and no-elected-leader groups had a larger amount of influence behavior than elected-leader groups. An interaction effect was observed for the influence distribution: GDSS, compared to manual support, helped to induce a more equal distribution of influence behavior in groups without elected leader. As for elected-leader groups, no significant difference was observed for the influence distribution between GDSS treatment and manual treatment. Dominance significance was found to be higher in manual group than in GDSS groups; and was also higher with elected leadership than without elected leadership.
Keywords
Source Title
Publisher
Series/Report No.
Collections
Rights
Date
1990
DOI
Type
Thesis