A comparison of the performance of structured and unstructured groups in three levels of time pressure

Few comparisons of structured and unstructured group performance in various stressfiil situations exist (e.g., Sprague, 1997; Worchel & Shackelford, 1991; Urban, Bowers, Monday, & Morgan, 1995). Given that time pressure Is the most common stressor in organizations. Experiment 1 compared the performance of three-member structured and unstructured groups in three time-pressure condhions. Structured and unstructured group performance was also compared to the performance of nominal groups.

A 3 (no, low, and high levels of time pressure) x 5 (structured and unstructured groups and the best, middle, and worst members of nominal groups) ANOVA using the quaUty of group performance as the dependent variable revealed that structured and unstructured group performance did not differ significantly. The quality of structured and unstructiued group performance was significantly lower than the best nominal groupmember, indicating that structured and unstructured groups experienced a process loss when working on the logic problem task. Although performance quaUty differed significantly between all three time-pressure condhions, the predicted curvilinear relationship between time pressure and performance did not occur. Finally, the rate of group performance did not differ significantly across the three time-pressure conditions.

Experiment 2, in a test of the social entrainment hypothesis proposed by Kelly, McGrath, and colleagues (e.g., Kelly & Karau, 1993; Kelly & McGrath, 1985; McGrath & Kelly, 1986), compared the performance of structured, unstructured, and nominal groups in two consecutive high time-pressure conditions. A 5 (structured and unstructured groups and the best, middle, and worst nominal group members) x 2 (costume and skiing versions of the logic problem task) mixed subjects ANOVA revealed that groups and Individuals worked at a lower rate in the second consecutive high timepressure condhion than in the first condhion. This unpredicted finding was attributed to the fact that groups and individuals faced a problem of capability when working on the logic problem task (i.e., the demands of the task exceeded available processing resources; Kelly, Futoran, & McGrath, 1990). Possible explanations for the lack of significant differences between structured and unstructured group performance and the failure to find the predicted curvilinear relationship between time pressure and performance quality are explored.