Steady State Generalization following Testing with Different Inter-Probe-Test Intervals

Abstract

Rats received discrimination training, wherein the S+ was a light-on (100 ft-c) and the S— was a light-off (0 ft-c). The schedules of reinforcement during S+ were a fixed-interval 30-sec. schedule with a limited hold of 5-sec. for two rats and a variable interval 30-sec. schedule for 3 Ss. After discrimination performance had stabilized, a probe-testing technique was used to test for generalization to three logarithmically arranged test stimuli (1, 25, and 50 ft-c). Tests were conducted with 5- and 2-day intertest intervals. The results indicated no difference in gradient shape after testing with 5- or 2-day intertest intervals. It was concluded that repeated probe testing for generalization virtually eliminates cumulative extinction effects that may distort gradient shapes.