Sensitivity of Responding to Simple vs. Complex Contingencies of Reinforcement

Abstract

Five rats were trained under second-order schedules in which the completion of 10 short fixed intervals (2, 4, or 8 sec) produced food reinforcement, and then to yoked variable-ratio schedules, in which the individual ratios corresponded to the ratios of responses to reinforcement obtained under the second-order schedules. Changes from second-order to yoked variable-ratio schedules resulted in increases both in overall response rate and running rate for all rats. Response rates decreased whenever the second-order schedules were reinstated for all but one rat. Response rates under the second-order schedules were inversely related to the value of the fixed-interval component. Positively accelerated responding was evident within some interrein forcement intervals under the second-order schedules at all three fixed-interval values studied, but this pattern of responding was more pronounced at the higher fixed-interval values. The yoked variable-ratio schedules, by contrast, generated short postreinforcement pauses followed by uniformly high response rates, and no evidence of positively accelerated responding. To the extent that different rates of responding were maintained by the second-order and yoked variable-ratio schedules, responding can be considered sensitive to the temporal contingency imposed by the fixed-interval component of the second-order schedules.