RESULTS OF THE WORLD'S FAIR HEARING TESTS 553 



for taking the test. Observation indicated that the great majority 

 of people took the test seriously, and made a conscientious attempt to 

 test their hearing. 



It was noticed that a very small percentage of the people, mostly 

 in the youngest age group, altered their scores by filling in all of the 

 missing numbers before having them photographed, thus giving a false 

 appearance of a perfect test. Some of these were detected from the 

 differences in writing, and were eliminated from the tabulations, but 

 others were probably included. It is believed that the number of 

 false scores included was too small to affect the hearing loss distribu- 

 tions appreciably. 



Some people undoubtedly secured poor scores in the test on account 

 of failure to understand the test, interruptions, or other causes not 

 connected with hearing. A study of this factor was made by observing 

 about 1200 tests, picked at random, and interviewing all those who 

 failed to fill in more than three squares of their test blanks. About 

 1.5 per cent failed the test in this sense, and were subsequently inter- 

 viewed and watched to see if they permitted their test scores to be 

 photographed. The interview revealed that about two-thirds of this 

 group failed the test because they were definitely hard of hearing. 

 Also, as it happened, about two-thirds of them submitted their test 

 scores to be photographed, so that the number of recorded failures 

 tended to be the same as the number that were actually hard of 

 hearing. 



The noise conditions under which the tests were given were quite 

 favorable. The exhibit building was quiet due to the generous use of 

 sound absorbent walls and carpeted floors. Parts of the building con- 

 taining air-conditioning and other machinery were constructed on a 

 separate foundation from the part containing the hearing test booths. 

 The booths were carefully insulated, the attenuation of the walls to 

 air borne sounds being 30 db or more over a wide frequency range. 

 Additional isolation was provided by a glass partition between the 

 booths and the lobby. Noise in the booths from external sources was 

 nearly inaudible, and it is probable that most of the distributing noise 

 was caused by the people participating in the test. 



Sound level meter measurements with flat weighting were made in a 

 booth where noise from external sources was judged to be most ob- 

 jectionable, and while regular tests were in progress. The average 

 and maximum readings are given in Table 14. After making allow- 

 ance for the attenuation of the telephone receiver covering the ear, 

 the masking computed for the average noise level was less than 5 db 

 at 440 cycles and zero at the higher frequencies. The masking of a 



