410 



MEMOIRS NATIONAL ACADEMY OF SCIENCES. 



[Vol. XV. 



The abbreviated scale is given and scored in the same way as the original except that six 

 months credit is given for each test passed. The records of 294 subjects ranging up to 12.4 

 years mental age (on the original scale) were re-scored on the abbreviated scale. The deviations 

 of new scores from the old were as follows: 



Deviations^. 

 Number 



3.2 

 1 



1 Median deviation, 0.3 year; mean deviation, 0.46 year; mean deviation, 0.37 year, omitting cases 11.5 to 12.4. 



The correlation of original and abbreviated scales is shown in Table 68. It will be seen 

 that the mental ages agree very closely as far as mental age 10, but above this point the 

 abbreviated scale yields mental ages which are considerably too low and apparently lacking 

 in reliability. As far as the mental age of 9 years the agreement between original and short 

 scales represents a correlation of approximately 0.95. As few rejections were recommended 

 above the mental age of 9 years, the abbreviated scale could be used with a fair degree of 

 safety in the large majority of cases. In order to measure as high as 12 years the abbreviated 

 scale would have to have several tests above the 12-year level. 



Table 68. — Correlation of original and abbreviated Stanford-Binet scales. (r=0.91; for mental ages below 10.5 by short 



scale, r=0.95.) 



The 60-word test was the only one that would fulfill all the requirements at year 11, but 

 in its original form it was too uneconomical of time. Comparison of the number of words 

 named in 1 minute with the number named in 3 minutes gave for 422 subjects a correlation of 

 0.853. Twenty-five words in 1 minute is, in the case of white recruits, approximately equivalent 

 to 60 words in 3 minutes, in the sense that about the same per cent of cases would pass (or 

 fail) by either standard. 



The abbreviated scale has the advantages that its tests are arranged more nearly in order 

 of difficulty for soldiers than is the case with the tests of the original scale. This makes it 

 possible to reduce considerably the range through which a subject is tested. Only 3 subjects 

 who passed two consecutive tests below 8 years failed on a lower test, and only 7 subjects who 

 failed on 4 successive tests above 8 years passed a higher test. As there were 115 subjects 

 who failed on 4 consecutive tests above 8 years, the proportion of these who succeeded in 

 passing a higher test was 1 in 16. 



Another abbreviation 1 of the Stanford-Binet was worked out at Camp Lee by E. A. Lincoln 

 and K. M. Cowdery. The data used consisted of the test records of 358 men, most of whom 

 were beta failures. The mental ages ranged from 4 to 12 years. Thirteen tests were chosen 

 and given a value of 6 months each, in addition to the basal allowance of 4 years. They were 

 as follows : 



' See School and Society, December 13, 1919, pp. 707-709. 



