no. 2.] PSYCHOLOGICAL EXAMINING IN THE UNITED STATES ARMY. 301 



Test 5. Information. 



Test 6. Synonym-antonym. 



Test 7. Practical judgment. 



Test 8. Number series completion. 



Test 9. Analogies. 



Test 1 0. Number comparison. 



Many different considerations helped to determine the above selection. As regards sus- 

 ceptibility to malingering, it appeared that the tests differed little. All agreed that the vocabu- 

 lary and opposites tests are both exceptionally valid measures of intelligence; yet for good 

 reasons neither was chosen. The committee accepted the principle that the tests should measure 

 many aspects of intellegence. It appeared that the vocabulary test, the synonyni-antonym test 

 and the opposites test were of the same general type. The synonym-antonym test embodies 

 most of the advantages of both the vocabulary and the opposites tests and it has the advantage of 

 requiring less time and being more easily scored. The vocabulary test is expensive in time and 

 requires so much space that it could not be placed on a single page. It was decided, accordingly, 

 to adopt the Otis synonym-antonym test and to eliminate the vocabulary and opposites tests from 

 further consideration. Again, it was generally agreed that the Trabue type of completion test is 

 a better measure of intelligence than some of the other tests finally accepted, as for example, the 

 number-comparison or memory-for-digits tests. However, the difficulties in securing alternative 

 forms of this test and arranging it for response without writing and objective scoring were 

 too great to be overcome in the time available. Moreover, it seemed undesirable to include 

 in tho scale too many tests of the verbal or language type. It was thought that the absence 

 of a language-completion test would be offset in part by the inclusion of the number-series 

 completion test. 



After the selection of 10 tests to constitute the scale for group examining, it remained to 

 define more specifically the principles to be followed in selecting the items of the tests, and in 

 arranging the form of each test and the general form of the scale. Among the guiding principles 

 formulated were the following: 



Each test should be composed of 10 to 40 items, ranging from easy to difficult. The time 

 limit preferably should not exceed 3 minutes. The time allowed for a test should permit not 

 more than 5 per cent of an average group to attempt all the items. The instructions for each 

 test should be recited by the examiner and, at the same time, read from their examination 

 blanks by the subjects. The directions for each test should be followed by two or three samples 

 with the correct responses given. Ten sets of Alternative "forms" should be prepared as a 

 safeguard against coaching. Speed and accuracy should be weighted empirically after the tests 

 have been given to a sufficient number of subjects. It was desired to avoid overpenalizing 

 subjects who work slowly and carefully. The following points were agreed upon for the separate 

 tests : 



1. Oral directions. — -Ten series consisting of 10 items each. Time for test, 6 minutes. 

 Some of the Abelson and Otis type of material and some directions of military import should be 

 included. The material should not be informational. Responses are to be made otherwise 

 than by writing. 



2. Memory for digits. — Two series each of 4, 5, 6, 7, S, and 9 digits. Credit to be determined 

 empirically, and a series to be scored as either right or wrong without partial credits. Examiners 

 should be drilled to secure uniformity of rate. 



3. Disarranged sentences. — List of 20 sentences. Time limit, 3 minutes. Response by 

 underlining the word "true" or "false." Sentences must state generally known facts. 



4. Arithmetical reasoning. — Ten sets of 20 problems each. Time limit, 5 minutes. 

 Response to require only writing of numbers. Problems should not involve information beyond 

 the four fundamentals and simple fractions. 



