524 MEMOIRS NATIONAL ACADEMY OF SCIENCES. [vou xv, 



Table 168. — Comparison of officers by ranks — Ratio of the differences between means to the probable errors of the differences. 



Reference to table 164 shows that these ranks have no men falling below 150 (C), whereas 

 the three ranks below them contain men of C — grade and even scattering D cases. (It will be 

 noted that data for the highest two groups are less reliable than those for other ranks owing to 

 the relatively small number of cases. This is especially true for the data summarized by camps. 

 Differences of this sort can not therefore be stressed.) 



Table 168 indicates the degree of significance to be attached to differences between means 

 of the various ranks. In conjunction with table 167, this indicates that the second lieutenants 

 are distinguished with a high degree of certainty only from the first lieutenants, appearing as 

 superior to the latter. The slight inferiority of the second lieutenants to the captains, majors, 

 and officers of higher rank does not appear certainly valid. The first lieutenants, as a whole, 

 show mental rating inferior to that of all other ranks, and the difference is sufficient to be con- 

 sidered valid. (Elimination of medical first lieutenants from this group raises the average by 

 a significant amount above the average for the total group of first lieutenants, but not by a 

 sufficient amount to affect the reliability of the differences between these and other groups 

 except in two instances.) The ranks of captain, major, and higher ranks fail to be significantly 

 distinguished either from one another or from second lieutenants. The small numbers of cases 

 hi the two highest groups are apparently mainly responsible for failure of these ranks to show 

 clear distinctions. From the data at. hand, therefore, the only fact which stands out as con- 

 vincingly evident is that of the relative inferiority of the group of first lieutenants. 



FURTHER DATA ON OFFICERS' TRAINING CAMPS, THIRD SERIES. 



The relationship of candidate officers to officers on the one hand, and enlisted men on the 

 other, has been discussed in an earlier section (pp. 511 ff.). The main further comparisons of 

 significance concerning themselves with (1) differences between various stations, and (2) differ- 

 ences between different types of training groups. 



Differences between stations. — Tables 169 and 170 present data on 22 stations of the third 

 series made up of Infantry and Artillery units, and also on the Engineer reserve officers' training 

 camp at Lee. For convenience the stations have been arranged in order of excellence as indi- 

 cated by median scores in examination a. Table 169 gives values of the medians, first quartiles, 

 and third quartiles. Table 170 shows the percentages in each camp making given grades. The 

 22 camps have been arbitrarily separated into 1 1 better and 1 1 poorer camps to facilitate 

 comparison, although it is evident that steps from one camp to another are very gradual. Dif- 

 ferences between the extremes are, however, marked. 



From Table 169 it appears that the medians range from 299 to 245 — a difference of obvious 

 validity. The average median score of the better 1 1 camps is 283 as compared with 259 for the 

 poorer 11. Figure 25 presents graphically the medians of the various stations. Comparing 

 this figure with figure 18 we note that the medians of these various camps fill in almost the whole 

 range from that of the officers on the one hand (296) to that of the Taylor sergeants on the other 

 (232), the highest of the training-camp group showing a higher median than the general-officer 

 group, but the lowest of the training camps failing to drop to quite the level of the sergeant group. 



