| Autor: |
Green, David M., Luce, R. D. |
| Zdroj: |
Journal of the Acoustical Society of America; 1972, Vol. 51 Issue 1A, p125-125, 1p |
| Abstrakt: |
The study investigated the speed-accuracy trade for detecting a weak 1000-Hz tone presented in continuous noise. Reaction times were measured in a YES-NO design with a fixed response deadline and various payoffs for the four stimulus-response combinations. The deadline was varied from 300 to 2000 msec in various experiments, and the payoff schedule was varied at a fixed deadline to determine different points on an ROC curve. With 300 msec deadline, performance is near chance; with a 2000 msec deadline, performance is practically perfect. With a fixed-response deadline imposed on both signal and noise trials the data favor a counter model, since the mean reaction time for all events is roughly the same, and the ROC curve has a slope less than unity. When the deadline is imposed on only signal trials, the data favor an interarrival time model, since the mean reaction time on noise trials is a linear function of the mean reaction time on signal trials and the ROC curve has a slope greater than unity. [This research was supported by a grant from National Science Foundation.] [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
