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



It will be seen that the digit-symbol test gives by far the best correlations. It not only 

 correlates highly with examination a but, with the exception of the puzzle test, it correlates 

 very well with all the others. However quite a number of illiterates and foreigners fail to do 

 anything with it. The correlation figures, particularly those showing its high correlation with 

 examination a, show that it is a promising group performance test. 



The test correlating next highest is memory for designs and here again the lowest correla- 

 tion is with puzzles. It correlates highest with the digit-symbol test and next with examination 

 a. It also shows a fair correlation with the Stenquist and maze tests. 



The maze test ranks next. But in this there appears a grouping of scores around the 

 lower values. 



The cube construction test correlates very little with examination a, but moderately 

 well with Stenquist, digit symbol, and designs. It is believed that its relatively low correlations 

 with other tests is due to lack of uniformity in methods of scoring and to the fact that bright 

 subjects because of a slight error sometimes "wreck" their cubes near the end of the time 

 allowance. 



The Ruger puzzle test correlates very little with anything. Men of inferior intelligence, as 

 measured by the other tests, received almost as high scores in the puzzle test as men of superior 

 ability. The test is cumbersome, far from "coach-proof," and tends to create disorder in the 

 group. 



All the above tests were given by T. H. Haines to 30 inmates of the State Colony for 

 Feeble-minded Males at Four Mile, N. J. The Binet mental ages of the subjects ranged from 

 4 to 11 years. From a study of the score distribution and the correlations with mental age 

 Haines drew the following conclusions: (1) The use of the puzzle test is contra-indicated, as it 

 gives a distribution of scores for feeble-minded subjects which closely resembles that for unse- 

 lected soldiers. (2) The digit-symbol test should have further trial with illiterate foreigner. 

 It is of value as a group performance test for English-speaking subjects. Smaller groups and 

 improved procedure may render it satisfactory for use with foreign subjects. (3) The maze 

 test, adapted for group use, should prove of value in the diagnosis of mental deficiency. (4) 

 The test of memory for designs lends itself as well to group as to individual examining. "It 

 is a piece of performance we can not dispense with in an attempt to sift out the mentally 

 incompetent by the group method." (5) The cube construction test "commends itself as a 

 most useful group test by which to gauge the intelligence of the individual." 



As it turned out, only the maze test and the digit-symbol test (modified form) of the Dix 

 series for illiterates found a place in the beta scale finally adopted. Undoubtedly, however, an 

 excellent and workable beta scale could have been based upon the supplementary tests used in 

 the Dix iiiYestigation. The maze test, the digit-symbol test, and the test of memory for 

 designs supply a fairly satisfactory measure of the intelligence of illiterate and foreign subjects. 



Camp Taylor. — J. Crosby Chapman devised what he called the "cube distribution test" 

 for the group examination of English-speaking illiterates. Materials for each subject: Four 

 small boxes (painted red, blue, yellow, or green) and 10 cubes. Size of group to be 

 examined, 50 to 100. 



The following directions are given, each beginning with "attention" and ending with "go": 



Part 1. (a) Put one cube in each of the boxes except the blue box. (b) Take the cube in 

 the red box and put it in the blue box. (c) From the pile in front of you put one cube in the 

 red box and two cubes in the yellow box. (d) From the pile in front of you put one cube in 

 the blue box and the rest in the green box. 



Part 2. (a) Put one cube in each of the boxes except the blue and yellow box. (b) From 

 the pile in front of you put one cube in the blue box, two cubes in the yellow box, and one in the 

 green box. (c) From the pile in front of you put one cube in the blue box, two cubes in the 

 green box, and one cube in the yellow box. 



The test was given to 102 men in the Jeffersonville (Ind.) Reformatory who had been 

 tested by the Stanford-Binet. The per cents passing for the different intelligence quotient 

 groups were as follows: 



