CHAPTER 19. 



DATA FROM COLLEGES AND THE STUDENTS' ARMY TRAINING CORPS. 



Data resulting from giving the alpha examination to the Students' Army Training Corps 

 and to the college groups are of interest as bearing upon the approximate intelligence of such 

 groups and, conversely, of the position of these groups on the scale of alpha scores. They 

 show, further, the results obtained from the revision of procedure for the use of the Students' 

 Army Training Corps (see Part I, page 200) . Accordingly it seems desirable to present a number 

 of comparisons based on such data as are available. 



Caution is necessary in the interpretation of the figures in the tables. It is by no means 

 certain that the values given are typical of the groups which they are supposed to represent. 

 The institutions from which the figures were obtained will be seen to be largely the smaller 

 schools and colleges from the West and South. Few data are at hand from the larger universi- 

 ties, and the records suggest that inclusion of figures from a number of large universities of 

 high standing would materially raise the medians and the percentage of the higher ratings. 



Again, it can not be stated with any certainty that the groups to which the examinations 

 were given were truly representative of the institutions. There are from Brown University, 

 for instance, records of only 210 men. This group is such a small proportion of the total 

 registration that the chances are very much against its being representative. At some of 

 the institutions the tests were given only to those who volunteered to take them, a procedure 

 which introduces a factor of selection that we have no means of measuring. Sometimes the 

 records were obtained from only one class. Such a limitation probably gives a selected group ; 

 the figures from the University of Illinois show distinct class d\fferences. 



A third factor making for variability in the results is the fact that tests were given under 

 widely varying conditions and by many examiners, most of whom were untrained. While 

 the chances of variations from this source are not large, yet the possibility of such variations 

 must be held in mind while examining the figures. 



Table 425 shows the distribution of alpha scores in several of the larger groups, and table 

 426 the median scores, quartiles, and percentages of A and B grades for the groups listed in 

 table 425. There are wide differences between these groups. The medians vary from 111 

 for the normal school women to 130 for the college men, and the percentage of A and B grades 

 combined for the same groups varies from 57.4 to 75.2. None of the groups make a record 

 as good as that of the Army officers chosen for the principal sampling for Hollerith analysis, 

 but all groups rate higher than the Army sergeants, who made 50 per cent A and B grades. 

 The men of the Students' Army Training Corps and the Reserve Officers' Training Corps seem 

 to offer good material for the selection of officers for the Army. 



In table 427 an analysis is made of the records of five units of the Students' Army 

 Training Corps which included more than a hundred men. Here again wide differences are 

 apparent. 



Two of the groups of the Reserve Officers' Training Corps were made up of more than a 

 hundred men each, and these two groups are compared in table 428. There is a difference of 

 10 points in the median scores and a difference of 18.5 per cent in the number of A and B grades 

 obtained. 



The figures for various groups of college men are presented in table 429. Here the 

 differences are even more striking, for the medians range all the way from B0 to 150, while 

 the percentages of A and B grades run from 17.4 to 95.5. Several colleges present here a better 

 record than that made by the white officers in the principal sampling. 



869 



