264 HUMAN BEHAVIOR PROGRAM 



Sealabs. Correlations between before-and-after choices were very high, but there were 

 changes. The correlation between men chosen as leaders on the predive and postdive measures 

 was +0.94, while the same correlation for teammates was +0.79. It is not surprising that these 

 correlations were so high in view of the previously mentioned good intra-team relations. 

 Climbers on the American Mt. Everest team evidenced the same stability in predive and post- 

 dive sociometric choices. The higher correlation for leader than for teammates is probably 

 due to the fact that fewer men were seen as potential leaders. 



Although the correlations between before-and-after choices were high, there were changes 

 in men chosen. First, there is an indication that different characteristics were used in naming 

 leaders and teammates, since the correlations between predive leader and predive teammate 

 choices was +0.42 and that between post leader and post teammate choices was +0.49. While 

 these correlations are quite high, they do not approach the level of predive-postdive correla- 

 tions on either the leader or teammate choices. 



Age was a significant variable in the choice of leader and teammate. Older men also 

 tended to be preferred for leaders and teammates on before-and-after measures. As could be 

 expected, the correlation between age and choice for leader is higher than that between age and 

 choice as a teammate. At first glance it is somewhat surprising to see that choice as leader 

 and years of diving experience are not significantly correlated. This lack of relationship, how- 

 ever, is probably artificially determined by one man who was highly chosen as leader even 

 though he had relatively little diving experience. Of considerably greater interest is the fact 

 that there is a high correlation between years of diving experience and choice as teammate on 

 the predive measure (r = +0.55), but that this correlation drops to +0.27 (not significant) on the 

 postdive measure. This means that while diving experience played a large part in choosing 

 teammates in the predive measure, other factors assumed greater importance as a result of 

 exposure in Sealab. 



A brief look at some other variables provides clues as to what might have made the differ- 

 ence in pre and post choices. There was a tendency not to choose men who complained about 

 conditions in Sealab or who made frequent telephone calls from the capsule. Measures of 

 complaints are taken from answers to the postexperiment questionnaire. The number of tele- 

 phone calls is a straight frequency count of personal calls initiated by divers from Sealab. 

 While extreme caution is necessary in attributing causative relations to such correlations, it 

 does appear that men who were most satisfied and content with their lot in Sealab were the ones 

 who were chosen as desirable future teammates. This interpretation is supported by the fact 

 that the amount of time a man spent in the water was correlated with postdive sociometric 

 choices. Finally, sociometric choices were correlated with a number of predive measures of 

 mood and attitudes. Although these measures are too obscure in meaning to warrant discus- 

 sion at the present stage of analysis, the existence of such relations does indicate that it will 

 be possible to specify the characteristics of divers which are deemed desirable by their team- 

 mates. 



Audiometric Test * —An examination of the pre-exposure data provided a basis for several 

 general observations. Only four of the 28 subjects had normal hearing, conventionally defined 

 as no more than 15-db loss at any of the test frequencies in either ear. Four other subjects 

 had normal hearing for one ear, with below-normal hearing levels for the contralateral ear 

 only at test frequencies above the speech range (3000, 4000, and 6000 cps). Five of the subjects 

 had hearing loss (greater than 15 db) at one or more of the test frequencies in the speech range 

 (below 3000 cps), in addition to high-frequency hearing loss. The remaining subjects (about 

 half) all had high-frequency loss, with some extreme instances (up to 70 db). The average 

 hearing levels for all ears (subjects combined) were within the normal range, except at 4000 cps 

 and 6000 cps, but these averages are inherently deceptive, as they tend to conceal the wide 

 range of individual differences. As the main interest and concern are with individuals rather 

 than group data, statistical analysis of the data is not appropriate. 



'■"This test was conducted and the results analyzed by Dr. George Harbold, Life Sciences Divi- 

 sion, Naval Missile Center, Point Mugu, California. 



