Document Type



In order to investigate the effects of instructions describing features of schedules on choice under concurrent variable-interval schedules, 15 human subjects, in three groups, were exposed to five concurrent schedules. In addition to basic instructions describing a task, subjects in different groups received different additional instructions. Subjects in one group received instructions describing features of variable-interval schedules such as time-dependent programming and unpredictable availabilities of points (Instruction-S). Subjects in a second group received instructions describing frequencies of reinforcement obtained from two variable-interval schedules (Instruction-F). Subjects in a third group received both Instruction-S and Instruction-F. In contrast to results reported in previous studies on human choice, most subjects showed systematic variations of choice ratios with reinforcement ratios. In addition, the slopes of the functions corresponding to the generalized matching law for the second and third groups were higher than those for the first group, indicating that Instruction-F had a stronger effect than Instruction-S. These results are consistent with the view that human choice conforms to the generalized matching law when subjects verbally formulate correct rules regarding concurrent variable-interval schedules.