718 



Popular Science Monthly 



Putting Speed in Telephone 

 Directories 



A SERIES of experiments were re- 

 cently conducted by the New York 

 Telephone Company to ascertam the 

 quickness with which a telephone num- 



Testing the speed of telephone directones. Names in 



the arrangement adopted were found in 9.28 seconds as 



against 10.36 seconds in the old arrangement 



book 



ber could be found with the 

 printed in three different ways. 



Dr. J. W. Baird, Director of the psy- 

 chological laboratory at Clarke Uni- 

 versity, Worcester, Massachusetts, was 

 called in to supplement the work of the 

 telephpne men by conducting other 

 tests, using a variety of type arrange- 

 ments. Dr. Baird made nearly four 

 thousand experiments to determine the 

 case and speed with which the average 

 person could find a number on pages of 



the directory set up in various forms. 

 Thirty-two men and women were se- 

 lected as subjects for the tests. Care 

 was taken that these individuals should 

 represent radically different occupations 

 and degrees of experience in the use of 

 the directory 



Pages with names begin- 

 ning with the letters I and 

 M and S were selected when 

 tests showed that they va- 

 ried sufficiently in diflficulty 

 to fulfill the purpose of the 

 experiments. 



Twelve pages were sub- 

 jected to experiments, an I- 

 page, an M-page, and an S- 

 page, being printed in each 

 of four different page ar- 

 rangements and mounted on 

 cardboard. Each page was 

 placed in a separate "book- 

 let." While the individual 

 tested was looking up a num- 

 ber, the experimenters held 

 stop-watches measuring the 

 time elapsing from the open- 

 ing of the booklet until the 

 subject found and pronounc- 

 ed the number. 



To find a telephone nuni- 

 ber in the old telephone di- 

 rectory, the pages of which 

 were set in three-column 

 measure, required an average 

 time of 10.36 seconds. When 

 the subscribers' names were 

 printed in a four-column 

 measure without indentation 

 or leading, the finding time 

 increased to 10.69 seconds. 

 When the lines in the four- 

 column page were set in 

 "staggered" arrangement, 

 i.e., in alternate indentation, 

 the finding- time was reduced to 10.14 

 seconds. When the type on the four- 

 column page was made slightly higher 

 and, moreover, narrower, taking eleven 

 lines instead of twelve lines to the mch, 

 the finding time was cut to 9.28 seconds. 

 It was this arrangement of the page that 

 was chosen, cutting i .08 seconds from the 

 10.36 seconds required by the average 

 subscriber to find a number in the old 

 telephone book. This is a gain of more 



than ten per cent. 



