The Effect of Color Coding on Performance of an Alphabetic Filing Task

Abstract

Color coding in filing tasks has often been accepted as superior to no color coding. The purpose of this experiment was to explore the effect of filing names, typical of a telephone directory, color coded on the first letter and on the second letter of the last name. A no-color-code condition was used as a control. Rate of production, number of errors, magnitude of errors, and user preference served as criteria. Twenty-two subjects filed 375 appropriately color-striped IBM cards into a box of IBM cards appropriately coded. The rate of performance and subject preferences indicated that the second-letter code was better than the other two conditions. Based on quantity of errors, the first- and second-letter codes were significantly (α < .05) better than the no-color-code condition. Considering the magnitude of errors, the first-letter code was significantly better than the two other codes. The results of this experiment suggest a color-coded filing system based on a simple first-letter scheme would significantly reduce the magnitude and quantity of high inconvenience cost errors over the conventional no-color-code system …