544 



BELL SYSTEM TECHNICAL JOURNAL 



The average real age in any estimated age group depends on the 

 distribution of ages within the group, the amount of random error 

 involved in making the estimates, and any consistent tendency to 

 estimate high or low. The distribution of estimated ages at the two 

 Fairs is shown in Table 6, together with the age distribution of people 



TABLE 6 



Percentage of Individuals Over 10 Years of Age Falling in Various 

 Age Groups According to Estimates at the Two Fairs, 



AND FROM THE 1930 U. S. CeNSUS 



over 10 years of age in the U. S. population (1930 census). At New 

 York, the youngest group was considerably larger than the corre- 

 sponding group in the population while the oldest group was very 

 much smaller. The same tendency, but to a lesser degree, is seen in 

 the San Francisco distribution. The distribution of ages fluctuated 

 greatly on different days of the week and at different seasons. The 

 table is based on counts on about 35 days at each Fair, well scattered 

 through the season. 



From the results obtained with the control group described above 

 and the age distributions of Table 6, the average real age in each age 

 group was judged to be as follows : 



These values should be reliable to the extent that it is reasonable to 

 assume that the accuracy of the age estimates of the control group 

 was representative of the whole group at each Fair. The oldest group, 

 which is designated 50-59 throughout this paper, included that range 

 at New York, but at San Francisco included all persons over 50. 

 Because of this the average real age in this group is judged to be 54 at 

 New York and 60 at San Francisco. The value in the table is the 

 weighted average of these two figures. 



