The effect of group size and computer support on group idea generation for creativity tasks: An experimental evaluation using a repeated measures design.

摘要

Most decision making, problem solving, and planning processes begin with the generation of ideas. While the use of groups for these processes is quite prevalent, a number of dysfunctional characteristics often negatively impact a group's performance--this is especially true as group size increases. The main research question addressed in this study was whether idea generation techniques with computer support, as provided by a Group Decision Support System (GDSS), could improve group performance for idea generation, while also maintaining this level of performance as group size increased. A laboratory experiment using a 2 x 2 x 2 full factorial design with repeated measures was conducted. Two hundred forty undergraduate student subjects were randomly assigned to 32 groups, with each group participating in two ideas generation sessions, one with GDSS support and one without GDSS support. Two other independent variables manipulated were: (1) Structure to support the group idea generation process was provided to half the groups (Osborn's brainstorming instructions), and (2) Group size was either five or ten persons. The four dependent variables measured were: (1) Quantity of ideas, (2) Quality of the ideas, (3) Group member satisfaction with the group idea generation process, and (4) Perceived usefulness of the idea generation technique. All other variables were controlled.;Results showed that groups generated significantly more total ideas, different ideas, ideas of higher total quality, more good ideas (as rated by expert judges), and were more satisfied with the group idea generation process when provided with GDSS support. GDSS support also allowed groups of ten to perform at the same per person level as groups of five for performance variables (quantity and quality of ideas) as well as maintaining the same level of satisfaction and perceived usefulness. No significant differences were found between groups provided with structure and those without structure, for supporting the idea generation process.