834 



MEMOIRS NATIONAL ACADEMY OF SCIENCES. 



[Vol. XV. 



Inspection of the table also indicates that in every case the distribution of cases for those pass- 

 ing the trade tests is skewed toward the A and B end of the scale. For those not trade tested 

 we have a relatively larger number of cases located toward the D and D — end of the scale. 



The significance of the differences of these intelligence distributions was determined by 

 Mr. Brown with Pearsons method used above. The results are as follows: 



1 Note.— 0.020000 or less is considered significant— viz, odds are 50 to 1 against chance as the sole factor in causing the difference. 



By reference to table 379, we see that only a small number of cases of plumbers, general 

 blacksmiths, and cooks have been trade tested. It is largely for this reason that the differences 

 turn out to be insignificant. This lack of significance means that it is necessary to obtain 

 more cases in order to show any real difference that may exist. The results in the other five 

 occupations indicate that there are real differences in the direction of superiority in intelligence 

 of those qualifying in a trade on the basis of trade tests as compared with those qualifying on 

 the basis of personnel interview. The facts seem to indicate that intelligence is a factor in the 

 successful passing of the trade tests and that this is true even in such a trade as 23 1 (heavy 

 truck driver) where a performance trade test is in use. 



An analysis of the relation between degrees of skill in an occupation (as measured by 

 trade tests) and intelligence ratings is possible on the basis of table 380, which gives a per- 

 centage distribution of intelligence ratings of apprentices, journeymen, and experts as classified 

 by trade tests. Only four occupations had a sufficient number of cases in each group to 

 warrant presentation. 



Table 380. — Intelligence and occupation. Percentage distribution of letter grades of apprentices (A), journeymen (J), and 



experts (E) separately, as classified by trade, tests. 



The significance of the differences of these distributions was also worked out in the same 

 manner as above, and the results were as follows: 



