no. 2.] PSYCHOLOGICAL EXAMINING IN THE UNITED STATES ARMY. 521 



Figure 21 may also be considered in comparison with the similar diagram for enlisted 

 men (fig. 4). It will be noted that the sharp distinction apparent wherever officers in general 

 are compared with enlisted men in general disappears when classification is made by branches, 

 the steps appearing far more gradual. In fact, the branch ranking highest among the enlisted 

 men (the field signal battalion) surpasses the lowest branch of the officers (the Veterinary Corps). 



Table 162 gives means and standard deviations for the officer groups, affording still another 

 basis of comparison of branches of the service. Table 163, supplementing the preceding table, 

 indicates the importance to be attached to the observed differences by expressing these as 

 multiples of their probable errors. It has seemed necessary to compare only pairs adjacent 

 to one another in order of excellence, except where the difference fails to carry conviction regard- 

 ing its validity. Inspection of Table 163 shows that differences between branches exceed four 

 times the probable error in all cases except in the comparison of the Medical Corps with the 

 dental and of the dental with the veterinary. The relative unreliability of these differences is 

 apparently mainly due to the small number of cases in the latter two groups. The Medical 

 Corps is, however, clearly distinguished from the Veterinary Corps on the one hand as it is from 

 the Quartermaster on the other. 



Table 162. — Comparison of officers by branches of service — Means and standard deviations — Examination a. 



Table 163.- 



-Comparison of officers by branches of service — Ratio of di fcrences between means to probable errors of differences. 



(Cf. Table m.) 



Comparison of ranks. — Distributions of officers by ranks are shown in table 164, the data 

 summarized by grades appearing in figure 24. Table 165 shows the medians of the various 

 ranks by totals and by camps. Table 167 gives means and standard deviations of the different 

 ranks. Table 168, supplementing table 167, indicates the probable validity of observed differ- 

 ences between means by expressing these in terms of probable errors of the differences. 



The various forms of data here presented all indicate the same facts regarding differences 

 between ranks. It should be noted, however, that comparison of ranks is more or less influ- 

 enced by differences between branches, since certain branches have much larger proportions 

 in certain ranks than in others. For example, the Medical Corps has no representation among 

 second lieutenants, whereas it contributes considerable numbers to the group of first lieu- 

 tenants. It is to be expected, therefore, that the standing of first lieutenants as compared 

 with that of second lieutenants should be influenced by the status of the Medical Corps. (See 

 preceding section.) To check the influence of this factor we have figured results for first lieu- 

 tenants both with and without the medical units. (The medical members have not, however, 

 been eliminated from the higher ranks.) 

 121435°— 21 34 



