544 



MEMOIRS NATIONAL ACADEMY OF SCIENCES. 



[Vol. XV, 



(3) The E cases also make a very slight mean gain in the one group where such oases appear. 

 It is important also to note the gain by grades — i. e., the per cent of E men raising their 

 grades to D, of D to C — , etc. Table 201 shows in summary form the grade changes: 



Table 201. — Changes in grade on second examination in comparison xciih grade on first. 



Grade change. 



Enlisted 

 men (190). 



Devens, 



officers 7 



training 



camp 



(331). 



Taylor, 



officers' 



training 



camp 



(224). 



Devens, 

 officers 

 (202). 



Same grade 



One grade higher . 

 Two grades higher. 

 One grade lower . . 



Per cent. 



56.3 



40.5 



2.1 



1.1 



Per cent. 



58.3 



34.1 



6.9 



.6 



Per cent. 



50.7 



41. S 



6.2 



1.3 



Per cent. 

 73.3 

 22.7 



3.9 



Section 5. — Equivalence oj forms, examination a. 



On the basis of the original standardization of examination a, scores on form B were at 

 all times increased by 13. In the following discussion, form B scores have always received this 

 addition. There is no suggestion in any data bearing on differences among the five forms of 

 examination a that form B was thus made relatively too easy. 



Within a few weeks after examining was begun the equality of the forms of examination a 

 was called in question by a report from Camp Lee. "The data indicate," it was stated "not 

 only that the E form is more difficult than the others, but also that the forms increase progres- 

 sively in difficulty from A to E." 



A letter was accordingly sent to the chief psychological examiners at Camps Devens, Dix, 

 Lee, and Taylor, asking that similar data be tabulated and forwarded as promptly as possible. 

 Camp Dix sent the statement that the groups taking each form among the Three hundred 

 and third Engineers gave evidence of "a large variation between forms, form A being the highest, 

 form E the lowest, and B, C, and D in between, the latter three being of approximately 

 the same difficulty. Although these totals in general are in accordance with the results from 

 other camps, large differences in the performances of those taking form A and those taking 

 form E on tests 2 and 10 (tests in which the performance should be quite independent of the 

 contest) would indicate an actual difference in intelligence between the groups." 



Camp Devens sent distributions for two groups, Camp Lee for a second group, and Camp 

 Taylor for one group. There is, in addition, a record of the medians and quartiles of a com- 

 bined group (Nashville, Syracuse, Indiana, and the Mosquito Fleet) from the preliminary 

 summer examining. These are summarized in Tables 202 to 212: 



