RATING TRANSMISSION PERFORMANCE 119 



he is not understanding, the talker may talk more loudly or closer to 

 his transmitter, and also make such adjustments in enunciation and in 

 setting forth his ideas as in the case of direct conversation. Also, the 

 loudness of talking may be affected by the room noise at the location 

 of the speaker, which noise incidentally may not reach the listener and 

 so play no part in his reception. Aside from cases where the room 

 noise at the far end is severe enough to be heard over the telephone 

 circuit, the speaker does not have definite knowledge of the room noise 

 at the listener's end and therefore is not in a position to adjust his 

 manner of talking to this condition except in so far as the listener may 

 indicate difficulty in understanding. 



In listening, the result is of course dependent upon the position of 

 the receiver with respect to the ear. The local room noise reaches the 

 ear to which the receiver is held both by the path between the ear cap of 

 the receiver and the ear, and also through the sidetone path of his tele- 

 phone set. Some telephone users have learned that this effect may be 

 reduced by holding the receiver tightly to the ear and by covering the 

 mouthpiece of the transmitter when they are listening. 



It is evident that the success of telephone conversations depends not 

 only upon the performance of the telephone but also upon the perform- 

 ances of its users, the material of their conversations, the way in which 

 they talk into the transmitters and hold the receivers to their ears and 

 the room noise conditions. In addition, it is seen that the performance 

 of the telephone affects the performances of the users in such important 

 respects as the loudness of talking, the manner of presenting the ideas 

 and the amount of effort exerted to understand. Also, the effect of the 

 room noise is a function of the circuit characteristics. Furthermore, 

 the reactions of the circuit performance on those of the users are not 

 constant but may vary from person to person and from conversation to 

 conversation. In view of the random nature of these factors, which 

 are beyond the control of those who design and operate the telephone 

 system, the service performance rating of a telephone circuit should be 

 on a basis which takes adequate account of their ranges and combina- 

 tions in practice. This points to a rating based on a statistical analysis 

 of results obtained under service conditions. 



To determine and specify these factors so that it may be known how 

 to duplicate the range of service conditions in laboratory investigations 

 would be a prodigious task. Moreover, the duplication of these condi- 

 tions under control is bound to introduce a large element of artificiality 

 which would vitiate the results or at least raise serious questions as to 

 their dependability. 



The practical solution is to get sufficient data regarding the results 

 obtained over telephone circuits of different performance characteristics 



