560 



BELL SYSTEM TECHNICAL JOURNAL 



MEAN 



HEARING LOSS 



Fig. 9 — Typical distribution curve of hearing loss. 



than 2(T to the average, because, in this range, many causes are operat- 

 ing and none is predominant. In the range from 2a to Z(x, the dotted 

 area of Fig. 9, there is a good chance of isolating causes, although the 

 attempt might not be justified if it involved great discomfort, danger 

 or expense. Beyond 3a, the cross-hatched area of Fig. 9, there is an 

 excellent chance that the cause, or causes, can be isolated. ^^ This 

 does not imply that findable causes of hearing impairment do not 

 exist in Individuals lying closer to the average than these limits, but 

 merely that the hearing test is not useful in selecting them. The 

 identification of such causes and their treatment are, of course, 

 medical problems. 



TABLE 16 

 Hearing Loss at Which Deafness Begins, in the Sense Given in the Text 



This criterion leads to the limits shown in Table 16, beyond which 

 hearing loss is significant in the sense described. Limits are given 



1' It is important to note that the standard deviation to be used in fixing the limits 

 is to be determined after eliminating the effects of impaired hearing. Methods by 

 which this can be done have been developed in connection with statistical methods 

 of manufacturing control. See, for example, W. A. Shewhart, "Economic Control 

 of Quality of Manufactured Product" (Van Nostrand, New York, 1931). The 

 values of a used in fixing limits in this section are 20 per cent smaller than those 

 given in Table 3. 



