Signal Detection Performance as a Function of Fourier Description of Symbols
Performance on a signal detection task was explainable by differences between the two dimensional Fourier transforms of the background and target stimuli. A signal detection experiment by Marshak and Osarczuk (1984) used target and background stimuli designed to systematically differ in spatial frequency and orientation. They found that the hypothesized Fourier differences increased sensitivity and decreased decision time. The present paper reports the Fourier analysis of those stimuli which verify and quantify the stimulus manipulation. Multiple regressions were computed using differences in frequency and orientation to explain performance. The results were that 83 percent of d-prime and 74 percent of the decision time variance could be explained by the Fourier differences. These findings indicate that Fourier descriptions of symbols may be used to predict their effectiveness in work station environments.