382 



MEMOIRS NATIONAL ACADEMY OF SCIENCES. 



[Vol. XV, 



Table 49. — Percentage distributions of men who either made scores below 50 in alpha or were eliminated from alpha 



Camp Custer — 250 English-speaking cases. 



1 The maximum scores possible in the various tests are indicated by heavy lines drawn in the respective columns in the row next above that 

 of the maximum score — e. g., the maximum score in 1 is 10, in 2 is 16, etc. 



The desirability of some change designed to increase the difficulty of the tests at the upper 

 end seemed evident. Since changes in the blanks were not feasible at the time, the only method 

 which suggested itself for securing this increase in difficulty was an alteration of time limits. 

 Accordingly the following tentative changes were suggested on the basis of inspection of the 

 correlation plots for the several tests: 



In the hope that these changes might help to reduce slightly the number of high scores, 

 they were incorporated in instructions issued for an experimental trial of beta with unselected 

 groups. Eventually they were issued in the Examiners' Guide as the standard time limits 

 for beta. 1 



Comparison of groups born in English-speaking countries with those born in non-English- 

 speaking yielded results which tended to support the inferences suggested by the data from 

 beta-only and beta-from-alpha groups, respectively (see p. 380) — namely, that the men handi- 

 capped by language difficulties were not the ones who profited most conspicuously by taking 

 beta. Table 50 shows the percentages, making given scores in beta, of men who had either 

 made E in alpha or had been eliminated from alpha. Table 48 has shown that, while this 

 group includes only 13.1 per cent of English-speaking total, it includes 69.2 per cent of all the 

 non-English speaking cases. Inspection of the percentage distributions shows the non-English- 

 speaking to be distinctly inferior to the English-speaking group in their rating by beta, as 

 does also reference to the means of the two groups : 



Mean ( English-speaking) =101. 6±1. 35. 



Mean ( non-English-speaking) = 77. 8 ±1.54. 



1 No evidence was obtained regarding the exact effect of these changes. Inspection of results with the new time limits indicated that they 

 had not been greatly affected by the change, either in the direction of improvement or loss. The changed time limits were retained in the perma- 

 nent instructions mainly because the later modifications of beta, which were adopted, were based on data for which these limits were used, and 

 because further experimentation on time limits was not feasible. 



