Mazur J E
Psychology Department, Southern Connecticut State University, New Haven 06515.
J Exp Anal Behav. 1991 Jan;55(1):63-77. doi: 10.1901/jeab.1991.55-63.
Two experiments measured pigeons' choices between probabilistic reinforcers and certain but delayed reinforcers. In Experiment 1, a peck on a red key led to a 5-s delay and then a possible reinforcer (with a probability of .2). A peck on a green key led to a certain reinforcer after an adjusting delay. This delay was adjusted over trials so as to estimate an indifference point, or a duration at which the two alternatives were chosen about equally often. In all conditions, red houselights were present during the 5-s delay on reinforced trials with the probabilistic alternative, but the houselight colors on nonreinforced trials differed across conditions. Subjects showed a stronger preference for the probabilistic alternative when the houselights were a different color (white or blue) during the delay on nonreinforced trials than when they were red on both reinforced and nonreinforced trials. These results supported the hypothesis that the value or effectiveness of a probabilistic reinforcer is inversely related to the cumulative time per reinforcer spent in the presence of stimuli associated with the probabilistic alternative. Experiment 2 tested some quantitative versions of this hypothesis by varying the delay for the probabilistic alternative (either 0 s or 2 s) and the probability of reinforcement (from .1 to 1.0). The results were best described by an equation that took into account both the cumulative durations of stimuli associated with the probabilistic reinforcer and the variability in these durations from one reinforcer to the next.
两项实验测量了鸽子在概率性强化物和确定性但有延迟的强化物之间的选择。在实验1中，啄红色按键会导致5秒的延迟，然后有可能获得强化物（概率为0.2）。啄绿色按键会在经过调整的延迟后获得确定性强化物。这个延迟会在试验过程中进行调整，以估计一个无差异点，即两种选择被选择的频率大致相等的持续时间。在所有条件下，在概率性选择的强化试验的5秒延迟期间会亮起红色房灯，但非强化试验的房灯颜色在不同条件下有所不同。与强化和非强化试验房灯均为红色相比，当非强化试验延迟期间房灯为不同颜色（白色或蓝色）时，受试者对概率性选择表现出更强的偏好。这些结果支持了这样一种假设，即概率性强化物的价值或有效性与每个强化物在与概率性选择相关的刺激出现时所花费的累积时间成反比。实验2通过改变概率性选择的延迟（0秒或2秒）和强化概率（从0.1到1.0）来测试该假设的一些定量版本。结果最好用一个方程来描述，该方程同时考虑了与概率性强化物相关的刺激的累积持续时间以及这些持续时间从一个强化物到下一个强化物的变异性。
