An Extended Comparison of Telephone Keying and Dialling Performance

Abstract

The results of a 2y extended experiment on keying and dialling are presented. The experiment was designed to investigate the relative error rates on push-button telephones (keyphones) and dial telephones, and the variation in the error rates as users became more experienced.

Seventy office workers, who were due to have 12-button keyphones installed, and a control group of 20 dial telephone users took part. Their performance was assessed on 11 occasions during the 2 y period, ft was found that the time taken to key numbers (all seven digits long) did not vary significantly between tests. The keying error rate exceeded the dialling error rate throughout the experiment. The keying error rate was related to the amount of keying experience, decreasing during the first 20 weeks of the study and then levelling off. An exponential curve was found to be a good fit to the keying error data. The dialling error rate did not vary significantly during the experiment. The distribution of different types of error and the number of self-detected errors did not vary with experience. Subjects aged fifty years or older made more errors than younger subjects.

It is concluded that, even when performance has stabilised after the learning period, keying errors occur more frequently than dialling errors. The impact of this increase in error rate may be reduced by the introduction of new keyphone designs and an increase in the use of repertory diallers.