 The
birds are tested in a small wire cage placed in a sound deadened
acoustic chamber (photo at left). They are trained by operant
conditioning to peck two sensitive micro switches with light-emitting
diodes (LED's) attached. These switches are mounted on the wall just
above the food hopper (schematic to the right). Test sounds are
delivered from an overhead loudspeaker. Tests are controlled by a
computer and results from each trial are stored for later analysis. The
bird's behavior during test sessions is also monitored at all times by a
video camera.
During testing, a background sound is played repeatedly from the overhead loudspeaker. The birds are trained to peck one key (the left or observation) key until they hear a change in this background sound. When the birds detect the change, they must peck the right (report) key within two seconds of hearing the change in order to obtain access to food. If they incorrectly peck the report key (i.e. when there has been no sound change), all lights in the chamber are turned off for a few seconds. To ensure the birds are not guessing in these tests, a small percentage of the trials are "sham" trials. Sham trials are trials in which the background sound does not change. The bird's behavior in these trials is recorded to determine how often it "guesses". Pecking the report key during a sham trial also results in the lights in the chamber being extinguished. A bird's threshold is determined the number of correct responses, the number of misses, and the number of "guesses". Once the birds are trained on these procedures, they become accomplished psychophysical subjects and can be used for many years in different experiments testing many different aspects of auditory perception. |