Abstract
Distinction can be drawn between three types of letter sorting machines and these types can also be used to classify other kinds of machine and task. In one type (‘paced ’) the speed of operation is set by the machine : in another (‘ unpaced’) the speed is determined by the operator's own actions, and in a third (‘ lagged ’) the speed is determined by the operator, but a minimum time (a ‘ time lag ’) has to elapse between one operation and the next. The performance of 7 operators on a lagged letter sorting machine was observed over a period of 9 months. Records were taken of mean sorting rate, and of the distribution of sorting times. Practice effects are shown to have been continuing at the end of this period, and it is argued that the distribution of sorting times has been affected by the machine' lag. By assuming a distribution that would be expected if there were no lag, tables are presented which estimate what the sorting rate would be (a) if there were no log at all, (b) for different values of lag and (c) for different values of paced rate.