BIOLOGY IN HUMAN AFFAIRS 



posters, and group rivalry, as well as adoption of mechani- 

 cal safety devices; but in spite of these efforts the costs 

 were heavy, amounting to more than 3 per cent of the 

 gross receipts. In most other large cities, it should be noted, 

 this percentage is even higher. No wonder the management 

 was concerned to decrease the frequency of accidents. 



The question was raised by the general manager and the 

 claims attorney as to whether accidents might not be 

 reduced by a more careful selection of men to be trained as 

 motormen. It was remembered that in 1908 Miinsterberg 

 had broached the subject of individual differences among 

 motormen and, after working for some time in his labora- 

 tory to devise suitable tests, had examined six of their 

 experienced men — three with a high accident rate, three 

 with a low. His tests did not differentiate perfectly between 

 the two groups, and he retired to revise the methods. His 

 writings on the subject later stimulated much research. 

 In Paris, Berlin, Prague, Milwaukee, and elsewhere, such 

 tests as he proposed were developed into valuable aids in 

 the selection of motormen. So the management of this 

 particular railway called on the Personnel Research 

 Federation for counsel, and the present investigation was 

 begun. 



Our approach to the problem, however, was made along 

 other lines. Instead of focusing on initial selection we 

 decided to study the experienced men who had had more 

 than their share of accidents, to see whether ways might 

 not be found to help them to become safe operators. It is 

 important to keep in mind that the results achieved as a 

 consequence of adopting this approach have been attained 

 by methods other than better selection, or by discharging 

 or transferring accident-prone men. 



The foundations for our work were laid in a com- 

 prehensive job analysis, accident-location studies, and a 

 series of statistical investigations. One of the first of these 



[150] 



