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



401 



satisfaction. A more accurate method of scoring was devised and a demonstration^ form 

 prepared for use with. non-English speaking subjects. 



7. Digit symbol.— A test of this kind was specially recommended by examiners at Camp 

 Dix, where it was found to correlate 0.7S with group examination a in the case of approximately 

 900 unselected men. It was also a well-proved test in pre-war work. The form of the test 

 selected was that arranged by Otis for the beta group examination. 



8. The maze test (Porteus) 1 . — The Porteus mazes, adapted byYerkes (test D, old series) 

 had been extensively used in the first four camps and had given fair satisfaction. They were 

 specially recommended by the Camp Dix examiners. They were, therefore, retained, but 

 with improved scoring and rath instructions modified so as to provide for a demonstration of 

 part (a). 



9. Picture arrangement {"Foxy Grandpa" series).— The picture arrangement test had been 

 used by several investigators, pictures of the "Foxy Grandpa" series having been utilized by 

 Miss Bowler, Whipple, and others. The pictures selected for the performance scale are far 

 from satisfactory, but they were the best immediately available. Scoring offered some diffi- 

 culty and the method adopted is not entirely satisfactory. 



10. Picture completion (Healy) 2 . — An earlier form of the picture completion test, also 

 devised by Healy 3 , had proved satisfactory in the.Pintner-Paterson scale 4 and was recommended 

 by the Camp Lee examiners. The form of the test chosen is the new Healy test, with the 

 author's procedure and scoring. 



Instructions for giving these 10 tests were prepared by Bridges, with suggestions from 

 the camp examiners. The scale was then tried out at Camp Lee on 228 subjects, 200 of whom 

 had scored D or E in group examination a. The remaining 28 were A, B, or C men. It seemed 

 desirable to base the trial of the scale chiefly on data from the kind of subjects for whom it 

 would ultimately be used. 



The method of scoring adopted provided for the assignment of credit for both speed and 

 accuracy. As a rule, however, no score for time was given unless a degree of accuracy ap- 

 proaching perfection had been attained. The result is, of course, that inferior subjects get 

 their scores chiefly on accuracy, while the superior get a greater and greater percentage of 

 their total scores for time.- Furthermore, points were assigned for time and accuracy in such a 

 way as to obtain something approximating a normal distribution of scores (see table 60). This 

 was done, not so much because of any a priori assumption as to the distribution of the functions 

 measured among this group of men, as because it seemed very probable that a large number of 

 zero scores Would indicate failure of the measuring instrument to register small differences in 

 amount of the trait measured, and because a large number o'f perfect scores would most likely 

 indicate an undue limitation of the scale at the upper end. 



The method of weighting adopted was that provided by Otis for the tests of the old indi- 

 vidual examination series ; — a method which made it possible to combine the scores of any number 

 of tests into a total score. The median of each test Was assigned 15 points, the upper quartile 

 20, the lower quartile 10, and other ranks corresponding scores. The maximum score on each 

 test is approximately 30 (see p. 182 ff). 



In order to determine what tests best represented the scale as a whole, correlation of each 

 test (weighted score) with total score was computed. The figures obtained by the method 

 of unlike signs were as follows : 



Designs 0. 73 



Digit symbol 62 



Maze 64 



Picture arrangement 71 



Picture completion 71 



1. Ship 



2. Manikin and feature profile. 



3. Cube imitation 



4. Cube construction 



5. Form board 



i Porteus, S. D. Mental Tests for Feeble-Minded: A new Series 4J. of Psycho-Asthenics, vol. 19, No. 4, 1915, pp. 200-213 . 



' Account of this test not yet published. 



» Healy, W. A Pictorial Completion Test. Psychol. Rev., vol. 21, 1914, pp. 1S9-203. 



' Op. cit., pp. 61-63. Also Pintner, R. and Anderson, M. M. The Picture Completion Test. Educ. Psychol. Monog. 



