Changes in Human Performance with Age 159 



requirement may be needed if, instead of dealing with a 

 present problem, we are trying to anticipate one in the future. 

 We might, for instance, need to control education to allow for 

 changes over the years in schooling, or health to allow for 

 changes in the treatment of disease, but the principle is still 

 the same — we are attempting to predict a representative 

 sample of a future population. 



The essential requirement for a theoretical study is a 

 sample in which older and younger people are comparable. 

 This is much more difficult than a representative sample to 

 obtain with human subjects because differential effects of 

 death, disease, education, occupation and so on must be taken 

 into account, and in a rapidly changing world parts of the 

 economic, social, material and educational backgrounds of 

 subjects will inevitably differ with age even among people of 

 the same status in every other respect. The problems of ob- 

 taining comparable samples differ between experimental and 

 industrial researches, and we will therefore deal with the two 

 types separately. 



Experimental Studies 



The obvious procedure is to attempt to control the back- 

 grounds of the subjects so that the different age groups will 

 be equated in all important respects. Unfortunately, just 

 what respects are important appears to differ from one task 

 to another in a manner which, with present knowledge, is not 

 entirely predictable. There is, however, some evidence upon 

 three types of factor: (a) educational and occupational level, 

 (b) occupational "skill" and (c) family relationship. 



(a) Level of education and occupation. Several researches 

 have shown that declines of performance with age are less 

 among people of high educational or occupational level. 

 Effects have been shown for problems involving "ingenuity", 

 translating an artificial language, giving synonyms and anto- 

 nyms, symbol-digit substitution, completing number series, 

 giving meanings of words, giving analogies and solving arith- 

 metic problems (Sward, 1945); recall of a passage of prose and 



