Mark A. Stellmack
Robert A. Lutfi
Waisman Ctr., Univ. of Wisconsin---Madison, Madison, WI 53706
Three experiments compared listeners' detection and recognition in broadband noise of five familiar and five unfamiliar sounds with identical long-term spectra. The familiar sounds were ice in a glass, zipper, crashing wood blocks, creaking door, and gas station bell. The unfamiliar sounds were these same five sounds played backwards. Thresholds were obtained using a single-interval, yes--no procedure. In the first two experiments, detection thresholds were measured for each sound by playing the same sound within each block of trials (fixed signal), and by playing one of the ten sounds with equal probability on each signal trial (random signal). In the third experiment, both detection and recognition thresholds were obtained for three of the familiar sounds played forward and backward. On each trial, listeners were asked first to make a yes/no detection judgment and then to identify which signal was played regardless of their first judgment. In all three experiments, there were no significant differences between thresholds obtained for the sounds played forward and thresholds for corresponding sounds played backwards. There was a small (1 dB) but statistically significant difference between thresholds for fixed and random signals. These results suggest that naturally occurring, familiar sounds have no special status among auditory stimuli in simple detection and recognition tasks [Research supported by NIDCD.]