WHAT PSYCHOLOGY CONTRIBUTED 385 



an educated non-tradesman could pass, on the basis of his 

 general knowledge or intelligence. 



" All of these test records, which show exactly what each of 

 the eighty or a hundred men did with every element of the test, 

 are turned over to the. statistician who computes for every ele- 

 ment its diagnostic value. He then chooses those elements 

 which are found to differentiate most sharply between novice 

 and apprentice, apprentice and journeyman, or journeyman 

 and expert. He determines the best numerical weighting to be 

 attached to each of these selected elements. And finally, com- 

 bining these scores, he ascertains the critical rating which is 

 found to separate the largest number of known apprentices, 

 from the novices, the journeymen from the apprentices, and 

 so on. This stage of the process we have called * calibrating ' 

 the test. Like the calibration of a thermometer, the critical 

 points of the test score are located, not by theory or by the 

 opinion of the deviser of the test, but by actual trial. Only 

 thus are we confident that the test will really measure trade 

 proficiency with the degree of reliability required by the army. 



" Not infrequently the tentative formulation of the test has 

 proved inadequate, and after all the labor and expense of an 

 elaborate tryout it had to be thrown into the waste-basket and 

 a fresh start made. Only after a test had been devised which 

 was found on thorough trial to measure up to the requirements, 

 was it turned over for use with the soldiers. 



4< While these trade tests were being developed, two astonish- 

 ing discoveries came to light. The first of these is the rarity, 

 the practical nonexistence, of the exclusively motor-minded type 

 of tradesman, the man who can do the job with his hands but 

 cannot tell you about it in words. In beginning the trade test 

 development we had expected to meet numerous difficulties due 

 to the prevalence among manual laborers of this variety of 

 mental constitution. We expected to find that the oral type 

 of tests would prove useful with the more verbally minded 

 men; but we anticipated meeting many tradesmen of high 



