no. 3.] PSYCHOLOGICAL EXAMINING IN THE UNITED STATES ARMY. 



559 



Group I. These increments were intended to approximate 2,500 each. The counting was 

 done roughly, and in some cases considerably more than 2,500 cards were pulled. An analysis 

 of the group is shown in table 3. 



Table 3. — Group III: While draft, additional, by camps. 



Numbers of cases in Group III of principal sample. Additional selections, approximately 2,500 from each of the main 

 campa used for white draft, intended to increase the size of the total sample. States are for the most part not 

 known. 



It seemed best to keep dataupon negro recruits entirely separate from data upon white recruits 

 throughout this report. This procedure is partly a matter of convenience, since the separation 

 of the men in the Army most usually is reflected by a separation of the records in the files, but 

 the striking differences in intelligence ratings that occur between negro and white groups also 

 indicates that a combination of the two types would simply serve to obscure the fundamental 

 and vital differences. 



Group IV: Negro draft, pro-rated, by States. — This group represents the negro draft at large. 

 It is comparable to Group I for whites and was selected in the same manner. It seemed desirable 

 to have a total group of approximately 20,000 ; hence the pro-rating was made on the basis of 

 1 negro recruit to every 250 negro males, according to the Census of 1910. Table 4 gives the 

 analysis. It will be noted that there are shortages in the cases of Alabama and Mississippi. 

 There were not sufficient cards marked with the State for these two groups. Group IV appears 

 overdrawn in a number of cases. These surplus cases are due to an error by which extra cards 

 reserved for emergencies were by accident included with the main group. The error was not 

 discovered until after the Hollerith sortings had been made. 



Table 4. — Group IV: Negro draft, pro-rated, by States. 



Numbers of cases in Group IV of the principal sample. Figures show for every State numbers taking each examination, 

 the actual total number selected for the State, and the desired total on a pro-rata basis of 1 per 250 negro male 

 population in 1910. 



