384 



MEMOIRS NATIONAL ACADEMY OF SCIENCES. 



[Vol. XV, 



in the main far more lenient than alpha, at least for the English-speaking cases who were 

 referred to alpha in the regular course of camp routine. It was felt, however, that no reestab- 

 lishment of grades for beta should be made without securing data on more unselected groups 

 extending over the whole range of the scale. Such a group was accordingly obtained and 

 furnished the basis for the final changes made in beta and in the assignment of letter grades. 

 Since evidence was accumulating to the effect that alpha was probably scaled too severely, this 

 group was also made use of in adjusting the alpha grades. An account of the main facts of this 

 investigation in so far as it bore on the standardization of beta follows. 



Section 3. — Special investigation on relatively unselected group. 



To obtain the desired data a request was sent to 12 camps that approximately 100 relatively 

 unselected men in each camp be given alpha, beta, and individual examination. Preference was 

 expressed for Stanford-Binet as the form of individual examination to be tried, in order to 

 secure a uniformity of data for comparison since it was recognized that Stanford and Point 

 Scale mental ages were not interchangeable. No attempt was made to include the Perform- 

 ance Scale in the present comparison. 



In response to the request data were received from nine camps, which constituted the main 

 basis of the present study. Table 51 shows the numbers of cases who had had all three exami- 

 nations — alpha, beta, and Stanford-Binet — classified according as they were born in English- 

 speaking or in non-English-speaking countries. 



Table 51. — Number of cases received from various camps, examined by alpha, beta, and Stanford-Binet. 



Though each camp was requested to send data on an approximately unselected group, it 

 was recognized that a genuinely unselected group could not be secured by any such method. 

 In the first place, the group chosen in any given camp can not be thought of as typical of the 

 camp; and in the second place, there was the selection involved in the choice of camps. Never- 

 theless there was no such selection of low-grade cases as that involved in the usual beta groups. 



The major part of the work was done with reference to the English-speaking group of 653 

 cases. The reason for this choice was the fact that this is the only portion of the group for 

 whom the alpha and Stanford-Binet examinations can be considered reasonably satisfactory. 

 It seemed, therefore, that estimates of the validity of examination beta, which had for their 

 points of reference these other two forms of examination, must be based on cases without serious 

 language difficulty. It was assumed that beta should justify its value in application to English- 

 speaking subjects at least. Further adjustments to meet the special problems of non-English- 

 speaking cases might be accomplished later. This latter problem was not faced in the present 

 study, partly because of the small number of such cases and partly because of the lack of stand- 

 ards of reference. 



Hereafter, to avoid circumlocution, the terms "English-speaking'' and "non-English speak- 

 ing" are used to distinguish these two groups from one another on the ground of birthplace. 



Since no external standard was available for determination of the value of beta as a measure 

 of intelligence, it was necessary to evaluate this examination by comparison with the two meas- 

 ures already in use, namely, alpha and individual examinations. It seemed reasonable that beta 



