Changes in Human Performance with Age 167 



(c) The most reliable criterion of difficulty for older people 

 at a job is that substantial numbers leave at a relatively early 

 age. The reliability is clearly greater for men's work than for 

 women's and comparisons between jobs in terms of age at 

 which people leave must be made separately for women and 

 men. People transferred within the factory to other jobs must 

 be included as well as those who leave altogether. The reasons 

 given for leaving or transferring should be regarded with 

 caution. Clearly when people leave because of gross disease 

 or move from the district for family reasons we can hardly 

 regard their decision as being influenced by a growing difficulty 

 with age at maintaining performance at their jobs. Reasons 

 such as this are, however, likely to apply equally to all jobs 

 and therefore not to affect comparisons between them. Other 

 reasons such as disagreements with management, dissatis- 

 faction and even, as Richardson (1953) has pointed out, many 

 cases of sickness do seem to be associated with jobs which we 

 should expect on other grounds to become a strain as people 

 grew older. Indeed, it would seem that in the absence of 

 strong indications to the contrary all moves from jobs com- 

 pared should be counted regardless of reason. A possible 

 exception to this is where people leave for promotion or a 

 better paid job, and a more rigorous criterion would be there- 

 fore to count all moves to lower paid work for any reason 

 other than serious incapacity due to disease or accident and 

 genuine redundancy. Anyone looking at changes with age 

 can, however, scarcely fail to wonder whether many moves to 

 better paid work are away from jobs which would soon have 

 become a strain if the move had not been made. 



Two important research possibilities for the future in this 

 field may be mentioned. The first is the detailed study of per- 

 formance at jobs where substantial numbers leave relatively 

 young to see whether early signs of strain can be detected 

 in changing methods or manners of work. The second is 

 to use the cohort technique of taking a group, or set of groups 

 of different ages and following them over a number of years not 

 only at one place of work but also if they leave to go elsewhere. 



