The Effects of Symbol Frequency in Legibility Testing

Abstract

The legibilities of two fonts used for teletyped weather reports are being studied. The first tests were controlled-exposure-time tests, made with a tachistoscope, in which single symbols were shown to subjects randomly with respect to alphabetical and numerical order. The two fonts were tested at symbol brightnesses of 8, 6 and 5 ft-L against a constant background brightness of 1 ft-L. For each font at each symbol brightness, two conditions were compared. The first condition was that all the symbols occurred with the same frequency and the second condition was that the symbols occurred with relative frequencies similar to their frequencies in actual use in the weather reports. The results showed that the subjects’ performances were better when the symbols occurred with the unequal frequencies than they were for the equal frequency condition. Implications which reflect on the validity of legibility testing are discussed.