AN INPUT-PROCESS-OUTPUT ANALYSIS OF INFLUENCE AND PERFORMANCE IN PROBLEM-SOLVING GROUPS

Member characteristics and social processes that determine influence i n problem-solving groups were examined. University students completed a non-eureka intellective task-the Desert Survival Situation-both indi vidually and in small groups. LEVEL-adjusted correlations and EQS stru ctural modeling revealed that the extent of talking was predicted by d ominance and confidence, but not actual expertise. Perceived expertise was most closely predicted by the extent of talking. Perceived expert ise mediated the relation between talking and influence. Groups attemp ted to base decisions on expertise, but were not proficient at recogni zing expertise. Additional analyses, which examined the impact of 8 in fluence tactics, indicated that the tactics of reason and autocracy pr edicted perceived expertise. Group performance was related to level of member expertise and recognition of expertise.