Details:
Summary:
From an exam nation of typewritten reports, typeset proofs and published material it is apparent that typographical errors are very often distributed in the fashion of a Poisson Distribution. The four common types of such errors are: an omitted letter, an extra letter, an incorrect letter and transposed letters. Average error rates seem to fall it the range of to 10 per 6000 keystrokes. By using the numerical characteristics of a Poisson Distribution and conventional statistical quality control techniques it should be possible to measure, analyze and improve the performance of typewriting, typesetting and proofreading operations. The statistical relationships outlined in this paper may be used to identify significant differences in performance among people and machines. Preliminary data are given on operating speeds and errors per unit time in a variety of working environments.