noted earlier, supported the contention that differences between teams are 

 primarily a function of group dynamics, team composition, and individual 

 differences in motivation. 



General Comments on Team Differences 



It may be useful to make a few preliminary remarks on probable causes of the 

 large differences across teams in performance (marine science and total work) . 



One characteristic of each mission is that groups began working at a given 

 level (for example, 25% marine science per day) and varied around this mean. 

 Psychologically, each group appears to have established an initial norm for 

 work and to have generally maintained this level of output throughout the 

 mission. This determination of performance by expectancy has been widely 

 noted as a cognitive consistency phenomenon derived from the theory of cogni- 

 tive dissonance (Festinger, 1957); Aronson and Carlsmith, 1962). The implica- 

 tion of this finding is that it will be advantageous for mission planners to 

 make explicit and expected norms for performance. In TEKTITE, no transmission 

 of expectancy from program personnel was planned or undertaken. This, of 

 course, was because a goal of the research was to investigate the types of 

 norms established. 



An important goal of later analyses will be to uncover reasons for the estab- 

 lishment of different norms and to specify personal characteristics of those 

 who deviate from group norms. Group influence in small, isolated groups is 

 extremely powerful, and subtle pressures are exerted to enforce conformity. 

 Individual differences will be discussed later. 



One factor which appears to be strongly related to level of performance is crew 

 composition, particularly in regard to the leadership role. In six missions, 

 the team leader was the scientific leader; in two missions the engineer was the 

 team leader; and in two missions leadership changed. In the latter cases, the 

 engineer served as team leader for 10 days; then, after engineer rotation, the 

 scientific leader served as team leader for the remainder of the mission. 



Although the number of leaders in the different leadership roles is insuffici- 

 ent for statistical tests of significance, several tentative hypotheses about 

 the leadership role can be advanced and may be subjected to test in subsequent 

 research. 



The differences in time spent in work between the top three teams and the 

 bottom three teams are large (mean work, top three = 38. 7%; mean work, lowest 

 three = 27.57o). The three teams showing the highest work output had a scien- 

 tist as team leader while the lowest three had an engineer as team leader for 

 all or part of the mission. 



The leaders of high performing teams were not necessarily top performers (in 

 terms of time spent in work). Indeed, the within-team ranks for the leaders 

 of the top three were 1, 3, and 5 on work. In earlier studies (see Radloff 

 and Helmreich, 1968) it has been noted that leaders of effective groups may not 

 themselves perform particularly well. These differences in crew performance 

 are probably due largely to the mission role of the leader. The objective of 

 TEKTITE was to provide a base for scientific work. When a scientist was team 



VIII-35 



