Two experiments investigated the learning of variability by undergraduate students. In the experiment 1, four groups (n = 5) were submitted to two schedules of reinforcement for pressing the P or the Q keys at a computer. For the Variability group (VAR) the reinforcement was dependent on high variation among sequences; for the Yoked group (ACO) the reinforcement was given independently of the sequence variability. It was manipulated the order and interval (0 and 3 months) for contingencies presentation. In the second experiment, four groups received these contingencies associated with correct or incorrect instructions (rules). The results showed that: 1) the responding was always more variable under the VAR contingency, 2) under ACO, the behavior was partially influenced by order and interval for contingencies presentation, and by the rules. In both experiments, the subjects were not able to report the contingency that was working. These results point the operant control of the behavior variability observed.
behavioral variability; rule-governed behavior; operant contingencies; human behavior
