363The composition of 3-man military teams was varied experimentally by assigning members according to team composition in terms of all possible combinations of levels of ability and motivation. The teams (64 in all) performed real military tasks in a military field setting and unit commanders ranked the effectiveness of their performance at the end of 2 months of military activity. It was found that both ability and motivation had an additive effect on crew performance, thus leading to the conclusion that when teams perform highly interdependent tasks, performance is apparently unlikely to be affected in a nonadditive manner by team composition. The data were analyzed with a new statistical procedure developed to substitute for the Friedman two-way analysis of variance, argued to be inappropriate for testing the significance of main effects and interactions in a blocked fractional factorial design.Numerous task settings, such as the combat tank crew or the surgical operating room, require the coordinated efforts of individuals with differing skills and interests in order to perform effectively. However, relatively little is known concerning the manner in which varying ability or motivational levels among members may affect eventual performance. Terborg, Castore, and De Ninno's (1976) comments of more than a decade ago hold true today: &dquo;In spite of the practical value of knowing the effects of group composition on group outcomes ... there exists a lack of attention to the problem&dquo; (p. 783). Downloaded from 364 Past research suggests that skill and motivation are critical factors determining the quality of individuals' performance (see Pinder, 1984). As such, it seems likely that these factors would be critical to group outcomes as well. Yet important questions arise as to the manner in which varying individual characteristics aggregate to influence group behavior.Some writers have suggested that team members contribute in an additive fashion to group performance: The better a given member's qualities, the better group performance, irrespective of other member's qualifications (Hill, 1982;Shaw, 1976;Steiner, 1972). Yet others have argued that some group compositions may lead to performance deviating above (positive nonadditivity) or below (negative nonadditivity) this simple additive sum (e.g., see Hackman & Morris, 1983;Senn, 1971; Szilagyie & Wallace, 1982).One difficulty in assessing these propositions is the relative dearth of adequate empirical research. A cursory overview of the theoretical issues at hand indicates the critical role of task structure and complexity in composition effects. As one recent review of group productivity points out, the thrust of practical and theoretical interests is directed to natural task settings requiring the coordination of unique or distinctive individual participation and performance (O'Brien, 1984). However, the few studies that have attempted to address composition effects have suffered from such drawbacks as laboratory settings, simplistic tasks involving lit...