

CHAPTER 8. 



REVISION OF METHODS OF INDIVIDUAL EXAMINATION. 



Section 1. — Original group of methods for individual examination. 



Except at Camp Devens the tests originally proposed for individual examining had only 

 the hriefest trial. At the other camps they were early abandoned in favor of the Yerkes- 

 Bridges, Stanford-Binet, and Pintner-Paterson scales. This action appears not to have been 

 due primarily to unsatisfactoriness of the tests, considered individually, but rather to the fact 

 that they were not welded into a systematic scale or group of scales. The 22 tests were intended 

 as the raw material for at least three scales, which were to be developed and standardized as a 

 result of camp experience. That this expectation was not fulfilled was due to the necessity of 

 securing results which could be immediately interpreted in the light of generally known stand- 

 ards. Such a standard was mental age, and test scores which could not be readily translated 

 into terms of this concept were unsatisfactory both to the psychological examiners themselves 

 and to the neuropsychiatric officers to whom cases were continually being referred. There is 

 no doubt, however, that a good system of tests could have been wrought out of the material 

 which the committee brought together. Five of the non-language tests, in modified form, were 

 later included in the performance scale. 



The validity of the separate tests of the individual examination series is to a certain extent 

 indicated by their correlations with total score of the individual tests combined. Unselected 

 recruits were tested early in October, 1917, as follows: Camp Taylor, 107; Camp Lee, 131; 

 Camp Devens, 70. These 308 men were given the entire series of individual tests except H (free 

 association). 1 The tests were scored according to the supplementary directions given in Part I, 

 pages 147ff. 



The correlation of each test with total score of all the tests was computed for each of the 

 three groups of data separately. The three correlations were then averaged. The correlations 

 were computed by the method of unlike signs. The correlations were as follows; 



Test. 



A. Cube construction 



B. Clock test 



C. Cubeimitation (Knox) 



D. Maze test (Porteus) ! 



E. Form board (Dearborn) 



G. Orientational information 



I. Digits backward 



J. Vocabulary 



K. Letterline 



L. Disarranged sentences 



Correlation 

 with total 



score of the 

 20 tests 

 (unlike 

 signs). 



0.57 

 .57 

 .58 

 .66 

 .57 

 .64 

 .66 

 .74 

 .33 

 .71 



Test. 



M. Absurdities 



N. Rhymes 



O. Likenesses and differences 



P. Ingenuity (Terman) 



Q,. Designs 



K. Logical memory 



S. Comprehension 



T. Sentence construction. J ' 



U. Arithmetical problems 



V. Code learning 



Correlation 



with total 



score of the 



20 tests 



(unUke 



signs). 



0.74 

 .65 

 .70 

 .62 

 .63 

 .70 

 .62 

 .59 

 .61 

 .65 



It will be seen that all the tests except K (letter line) give a good correlation with the total 

 score of the 20 tests. From point of view of size of correlation with the total, the best five 

 tests are, in order vocabulary, disarranged sentences, absurdities, likenesses and differences, 

 and logical memory. 



At Camp Devens 114 men (mental ages 5 to 18) were given tests A, B, C, D, E, and G and 

 in addition either the Yerkes-Bridges or the Stanford-Binet scale. Table 58 shows the corre- 

 lation between mental age and total score for the five tests. 



i Test F (Stenquist construction), though listed with the individual examination methods, was used mainly as a group method for testing 

 illiterates. It is discussed in chapter 6. 



