Examined the influence of group size, task difficulty, and member sex on the relationship between actual productivity and potential productivity proposed by I. D. Steiner (1972), and tested the predictive accuracy of hierarchical and equalitarian latency models proposed by F. Restle and J. H. Davis (1962). 349 undergraduates worked 3 intellectual problems of varying difficulty, either as individuals or in same-sex groups of 2, 3, 6, and 10 members. Actual performance was assessed using indexes of the proportion of solvers and time to solution. Potential performance scores were generated from theoretical models for the same indexes. Group size and problem difficulty were important determinants of group performance, whereas member sex generally was not. Size and difficulty effects varied depending on choice of performance measure. For the proportion index, groups often worked up to potential, but those same groups usually fell below potential on the latency index; they were effective but inefficient. Group performance on both indexes generally fell below potential as size and difficulty increased. The concept of "functional size" is advanced to explain the observed pattern of group performance: As group size increases the number of nonparticipators also increases, resulting in a functional group size smaller than actual size.
Stored only when both papers are in the corpus — 126 citations exist upstream.