Group composition and performance: four studies of the relationships amongst member ability, group process, decision schemes and effectiveness

Studies of small group performance have long been dominated bythe view that interaction processes amongst team members are thecritical determinants of task achievement. Yet few studies show thattraining in group dynamics leads to performance improvements (Kaplan,1979). The emphasis on group level processes has overshadowed theinfluence of member task ability on group performance. The researchpresented here aims to redress this imbalance.Four integrated studies are undertaken employing differentresearch methodologies. Study 1 explores the group performance/sizerelationship, and the moderating effects of member expertise andsocial decision schemes. Both simulated and interacting groups areinvestigated. Performance is found to be a positive linear function of log size.The slope of this graph is steeper for high abilitymembers and more rational decision strategies.Study 2 explores the types of decision strategies used byinteracting groups. Team performance is separated into assembly andinformation processing components. The former is a function ofaverage member ability. The latter is determined by the degree offit with a rational non-unit weight decision scheme. Study 3 showsfor a group to achieve this strategy, member participation rates mustbe allocated in proportion to task expertise. Study 3 alsodemonstrates differences between perceived and actual influence: theformer is principally determined by participation level, the latterby expertise.Study 4 builds on the findings of the first three - that groupperformance is a function of member task expertise and strategies forits use. It shows that task training for individuals enhances memberability, group process and decision schemes and performance.All studies employ the same task: the NASA moon exercise. Atotal of 400 subjects working individually and/or in 70 interactinggroups are involved in the studies