no. 2.] PSYCHOLOGICAL EXAMINING IN THE UNITED STATES ARMY. 393 



Examiners' Guide. It was put into effect in the camps through instructions issued August 8, 



19 1 S. The ' ' suggested form " of het a is, therefore, the " final form ' ' from point of view of the army 



situation. 



Section 5.— Establishment of letter grades for beta (suggested form). 



The fact of the special importance of letter grades for practical army purposes has heen 

 noted, as has also the fact of marked discrepancy in rating, as grades were first assigned for alpha 

 and beta. With the adoption of the modifications in beta described above, it became 

 essential that new letter grades be assigned. Since final revision of alpha grades was being 

 made at this time no attention was paid to old relationships, but adjustments were made directly 

 in the light of the proposed ratings for alpha. Assignment of grades on individual examinations 

 was also made at this time. 



It was assumed that we should aim to secure the .highest degree of consistency possible 

 between alpha and beta from C— on up through A; that discrepancy between alpha and beta 

 for the D and E grades was to be expected; that these grades (D and E) must therefore be 

 established mainly with reference to the individual examinations, and that moreover as a 

 practical consideration it must be remembered that an E grade in beta means actually recall 

 for final measurement by individual examination. Meeting these requirements involved a 

 process of cutting and fitting to secure the greatest consistency. An additional arbitrary 

 limitation was set up by the desirability of making the dividing lines between grades always 

 multiples of 5 — e. g., 30 rather than 29, 85 rather than 83, etc. This was important for its 

 bearing on the clerical work of the camps, since it was considered that greater accuracy of 

 work was secured if more irregular points were not introduced. 



As a result of the cutting and fitting process new grade bases were adopted, issued August 

 8, and incorporated in the Examiners' Guide (see Pt. I, p. 153). (Substitution of the D-grade 

 for E was also made at about this time, but it has no bearing on the relationship discussed 

 here.) Table 57 shows the percentages located in the various letter grades by each of the 

 examinations, in accordance with the new assignment of grades. In addition to the total 

 English-speaking group figures are offered for two types of subgroups included in the total — 

 /viz, the Meade and Dix groups on one hand and the Jackson and Lee groups on the other — as 

 representing camps with large percentages of high-grade and of low-grade cases, respectively. 



Table 57. — Percentages making given grades on alpha, beta, and Stanford, by grade standards, as finally assigned English- 

 speaking cases. 



From inspection of this table it is evident that from the grade of C— on there is a much 

 higher degree of consistency between beta and alpha grades than there was by the old assign- 

 ment of letter grades. The two subgroups are offered as an indication that this approximate 

 consistency is not determined solely by the fact that grades were adjusted to the needs of the 

 total group. Considering the total group, we note the following percentages: (1) Making A 

 or B grades, on alpha, 14.7 per cent; on beta, 14.2 per cent; (2) making grades of C or above, 

 on alpha, 54.2 per cent; on beta, 51 per cent; (3) making grades of C— or above, on alpha, 

 73.5 per cent; on beta, 72.8 per cent. A simdar degree of consistency holds for alpha and beta 

 for the two subgroups. Stanford deviates slightly more from the other two examinations, 

 121435°— 21 26 



