DIGIT MATRICES FOR DATA-ENTRY AND RETRIEVAL - STUDIES OF SCANNING DIRECTION AND COLOR CONTRAST IN A TELEPHONE SWITCHBOARD APPLICATION

Two investigations are reported on difficulties encountered by switchboard operators when recording and reading out telephone numbers using matrical arrays of digits. In a laboratory study, subjects either entered 9-digit numbers into a matrix by marking characters in response to an auditory presentation; or they read out numbers previously marked by the experimenter. Half of each group searched the digital array by vertical scanning (simulating the telephone switchboard operator's task), half searched an alternatively oriented array by horizontal scanning. As predicted from previous studies of visual search for single predetermined items, vertical scanning tended to be slightly faster and more accurate than horizontal. However, neither scanning pattern had the advantage when subjects searched for marked digits. There was conflicting evidence from a brief transfer test in which subjects scanned vertically lifter scanning horizontally, or vice versa. After transfer, subjects were reliably better when scanning horizontally, both during search for predetermined and for marked digits. The results are explicable by reference to inappropriately transferred reading habits and to the influence of specific task characteristics on the size of the functional visual field. The practical implication is that the demands of data retrieval from digit matrices are incompatible with those of data entry. It is recommended that switchboard operators be allowed to write down ‘called’ numbers, for use when dialling, to avoid reading out from matrices. The general implications of these findings are that vertical scanning does have a slight advantage when searching homogeneous non-random sequences for single predetermined items. However, this result does not generalise to search for non-homogeneous targets, because of the different processing requirements of data entry and retrieval with matrical arrays. The second investigation simulated normal switchboard operating procedures to study effects of print and paper colour-contrast on search performance. None of the colour combinations tested produced significant effects on speed or accuracy of recording and retrieving telephone numbers. But the results supported subjective reports from operators that the major source of difficulty is reading out marked digits, and the above recommendation has therefore been adopted as standard operating practice. © 1979 Taylor & Francis Group, LLC.