334 MEMOIRS NATIONAL ACADEMY OF SCIENCES. 



Table 19. — Total score, examination a, and mental age for feebk-minded (r=0.657). 

 [Table includes following cases: Lapeer, 50; Sonoma, 48; Waverly, 40; Faribault, 31; Vineland, 20; total, 193.] 



[Vol. XV, 



Section 6. — Correlations with age. 



Each test was correlated with age for the 1,162 California pupils used in securing age norms 

 from 8 to 16 years. The results were as follows: 



Table 188 (Chapter 15) gives the age medians for the same data. It is certain that these are 

 higher than would be found for white school children in general, for two reasons: (1) All children 

 of foreign parentage were eliminated; (2) the cities in which the tests were made represented 

 better than the average social and cultural conditions for the country as a whole. 



Section 7. — Correlations with schooling and with teachers' estimates. 



Correlation of total weighted score of examination a with school grade completed is ordinarily 

 not far from 0.70 for unselected literate soldiers. In the correlation array shown on page 318 

 it is 0.67. Such comparisons are of course affected by the degree of accuracy with which school 

 grade is reported. The true correlation would doubtless be somewhat higher than the figures 

 given. 



Table 20 gives the total score by school grade for 717 unselected children in the schools of 

 Palo Alto, Calif. These included practically all the children enrolled in the grades above the 

 third, in which only one class was tested. The correlation is 0.83. Norms for school grades, 

 based on 4932 cases, are given on page 537, chapter 15. 



The correlation between total score and school grade becomes still more significant when 

 confined to unselected pupils of a given age. Such correlations were computed for boys and 

 girls separately for each age from 11 to 15 in the case of the California school children of Ameri- 



