326 MEMOIRS NATIONAL ACADEMY OF SCIENCES. tvou xv, 



Various test blanks and the record and report forms used in the original four cantonments 

 are shown on page 20 Iff. 



From the beginning the methods proved themselves practicable. In general the examina- 

 tions commanded the interest and respect of both men and officers. The feasibility of testing 

 all recruits was demonsrated. Indeed, the expected rate of progress was considerably exceeded, 

 due chiefly to the fact that it was found possible to examine men in much larger groups than had 

 been planned for. 



Certain shortcomings in the methods early became apparent: (1) The test which had been 

 prepared for the segregation of illiterates proved to be unreliable and was abandoned after ade- 

 quate trial. Various substitutes, to be described later, were tried. (2) The repetition of the 

 mam group examination with extended time (group examination b) was found to be adminis- 

 tratively objectionable because of the difficulty of recalling men for repeated examination. 

 Investigation also showed that group examination b added little, if anything, to the accuracy 

 of the measure furnished by group examination a. It was early abandoned in all the camps 

 except Devens, where it was given to about 650 men. (3) Chiefly because of lack of norms for 

 the interpretation of scores, the methods prepared for individual examination proved unsatis- 

 factory. Although used to a greater or less extent in all the camps throughout the initial 

 experiment, they were in large measure superseded by the Yerkes-Bridges point scale, the 

 Stanford-Binet scale, and the Pintner-Paterson performance tests. (4) The Stenquist test, 

 although it proved to be of some value in reducing the number of individual examinations, was 

 unsatisfactory because its results correlated so little with those of group examination a that it 

 could not be regarded as a legitimate substitute for the latter. 



Experimentation was early begun in all the camps looking toward the preparation of a 

 more suitable group test for illiterates. Group examination a, on which the success of the army 

 examining chiefly depended, was found to be satisfactory in all except minor details. It was used 

 without modification throughout the official trial, but in January, 1918, it was revised by the 

 elimination of two of the ten tests and by slight changes in several of the others. Its essential 

 nature, however, was not altered. 



