Effects of noise on signal detection

Abstract
Twelve subjects were tested twice in visual vigilance tasks which lasted 40 min. Employing a two-category confidence rating scale they detected increments in light level from displays of five lights. The display was flashed on simultaneously for 0·5 s every 3·5 s. The subjects performed the task on different days under two conditions of continuous white noise: ‘quiet’ (70 dB) and ‘noise’ (l00 dB). Half of the subjects had the noise treatment in the order of quiet-noise and half in the reverse order. No effects of noise either upon the overall performance or upon the vigilance decrement were observed. For the risky criterion results showed mainly that during a run under the two conditions the percentage of correct and false responses decreased, d' remained unchanged and β partly increased as a function of time. For the cautious criterion only β increased during a run under the two conditions. The results were interpreted in terms of arousal theory