308 MEMOIRS NATIONAL ACADEMY OF SCIENCES. [vol. xv, 



TestO. Likenesses and differences. — Finding differences, from Binet; finding likenesses, 

 from Stanford-Binet. The five series and the instructions were arranged by Terman, who is 

 responsible for all but five or six of the fifty items. 1 



Test P. Ingenuity. — Devised by Terman in 1905, used by him in an experimental study of 

 the intellectual processes of bright and dull boys. 2 Included later in the Stanford-Binet. Adapt- 

 ed for present use by Wells, who also prepared the instructions and the alternative problems. 



Test Q. Memory for designs. — Modeled after the Binet test. The designs here used were 

 devised by Terman. In experiments at Stanford University they had yielded high correlations 

 with mental age. 



Test E. Logical memory. — After Binet. Instructions and alternative series were prepared 

 by Wells. 



Test S. Comprehension. — After Binet. Instructions and the alternative series were 

 prepared by Wells and Terman. Wells furnished 1 1 of the 25 questions, Terman 8, and Goddard 

 the remainder. 



Test T. Sentence construction (three words) . — After Binet ; the Masselon test. Instructions 

 and alternative series were arranged by Melville. 



Test V. Arithmetical problems. — A combination of the Binet test of making change and 

 the Bonser type of arithmetical reasoning test. 3 Three of the five problems in the e series 

 came originally from Bonser and had been later incorporated in the Stanford-Binet. The 

 other two of the e series were taken from the group test of arithmetical reasoning. Instructions 

 and alternative series were arranged by Wells. 



Test V. Code learning. — The code test had been used in various forms by various psycholo- 

 gists, notably in the Healy * and Stanford-Binet 5 tests. The instructions and the five forms 

 of the present series were prepared by Wells. 



The 22 tests which have been described were not combined in a single scale. It was sug- 

 gested that they be used as follows: (1) For illiterates, tests A to F; (2) for subnormals, tests 

 J to O, supplemented by H and I, or tests P to U, supplemented by H and I; (3) for psychotics, 

 tests D, E, F, G, I, Q, and R; (4) for supernormals, tests A, C, E, H, I, J, and O. 



Apart from the limitations imposed by the criteria which the committee had agreed each 

 test must satisfy, the choice of tests was determined by three things: (1) By the purpose to 

 secure tests which would be of service with one or another of the types of subjects named 

 above — illiterates, subnormals, psychotics, and supernormals; (2) by the experience which the 

 members of the committee had had with the various types of tests proposed; and (3) by the 

 data contained in two studies of the diagnostic value of certain Binet tests — namely, Brigham's 

 recently published monograph 6 and Knollin's unpublished master's thesis. The latter, which 

 was brought before the committee by Terman, gave the correlation of each test above nine 

 years in the Stanford-Binet scale with mental age as measured by the entire scale for approxi 

 mately 450 adult subjects. Knollin's data showed the following tests to have high diagnostic 

 value: vocabulary, three words, rhymes, absurdities, designs, comprehension, dissected sen- 

 tences, digits backward, giving differences and similarities, arithmetical reasoning, and ingenuity. 

 These were included in the Army series. 



Some 30 other tests were proposed, but for one reason or another rejected. Notes made by 

 members of the committee indicate the rejections listed below, with reasons in some cases. 

 The name in parentheses after certain of the tests indicates authorship of the particular form 

 of test considered. 



Dot cancellation (Abelson): not a good measure of intelligence. Geometric estimation 

 (Thorndike and Woodworth) : too largely on the perceptual level. Line bisection : too largely 



1 Terman, L. M., and Chamberlain, M. B. Twenty-three Serial Tests ot Intelligence and Their Intercorrelations. J. Applied Psychol., 

 vol. 2, No. 4, 1918, p. 347. 



* Terman, L. M. Genius and Stupidity. Ped. Sem., vol. 13, 1906, p. 335. 



• Bonser, F. G. The Reasoning Ability of Children of the Fourth, Fifth, and Sixth School Grades. Teachers' College, Columbia Univ. Contribs. 

 to Educ, no. 37, 1910, p. 2. 



« Healy, W. and Fernald, G. M. Tests for Practical Mental Classification. Psychol. Monog. vol. 13, whole no. 54, 1911, pp. 33-34. 



5 Terman, L. M. The Measurement of Intelligence, pp. 330-332. 



8 Brigham, C C Two Studies in Mental Tests. Psychol. Monog., vol. 24, whole no. 102, 1917. 



