no. a.] PSYCHOLOGICAL EXAMINING IN THE UNITED STATES ARMY. 



703 



Table 222.— Variables: Stanford-Binet score X years in the United States. Group I, II, III: White draft, foreign born. 

 For men who took Stanford-Binet examination only, or following beta, or following alpha, or following alpha and beta. 



For purposes of comparison the data of these tables have been combined in terms of the 

 theoretical scale described in chapter 2. For this purpose the scores on the performance scale 

 were transmuted into mental ages by terms of the table in the Examiner's Guide. Scores on 

 beta were taken for men who had taken both alpha and beta. The distributions by class inter- 

 vals against the theoretical combined scale are shown in table 223, together with their medians 

 and quartdes. 



Examination of the median values of table 223 shows an increase from 11.3 to 13.7, about 

 2.5 of the arbitrary units of this scale. This range is equivalent to something over a difference 

 of 2 years mental age. The increase, it will be noted, is regular from one class interval to 

 another. Successive differences are 0.41, 0.83, 0.97, and 0.24. The dispersion is great and 

 the curves overlap considerably; nevertheless with such a large number of cases the differences 

 between the arrays are mathematically significant. The chances that the difference between 

 the group for 16 to 20 years and that for over 20 years is in the direction noted are 94 percent. In 

 all other cases such a measure of significance is 100 per cent when measured with four-place 

 tables. The ratios of the differences to their computed quartUe deviations (assuming sym- 

 metry) range from 9 to 35, with the exception of the single case of the comparison of the two 

 highest groups. 



