Choice between fixed and variable delays to reinforcement in the adjusting-delay procedure and concurrent chains

A generalization of the contextual choice model (CCM; R. C. Grace, 1994a) is presented that describes preference between fixed and variable schedules in concurrent chains and the adjusting-delay procedure. In Experiments 1A and 1B, fixed-delay indifference values for pigeons for a series of variable-delay schedules were obtained, using the adjusting-delay procedure, to test a Weber's law invariance and the effectiveness of several models. These data were then used to generate, in Experiment 2, pairs of fixed and variable schedules that should have been equally preferred in concurrent chains. As predicted, deviation from indifference was small. These results indicate that (a) the determiners of value in the 2 procedures are the same, (b) the delay of reinforcement gradient is best represented as a power function with a delay threshold, and (c) a single model (CCM) can accurately describe data in the 3 most popular operant choice paradigms.