An experimental analysis of jackpot reinforcers.

Animal trainers and others often recommend the use of jackpot reinforcers, which are disproportionally large and come as a "surprise" to the animals. Because the actual behavioral effects of these jackpots remain uninvestigated, many basic questions about their use and even definition are unanswered. This series of experiments explored the definitions of jackpots using several different behavioral tests with both rats and pigeons. Because the original description of a jackpot resembled the reinstatement of previously reinforced and extinguished responding with response-independent deliveries of reinforcer, reinstatement effects of a jackpot, defined by its quality, were examined with rats in Experiment 1. In Experiment 2, response-potentiating effects of response-independent and -dependent deliveries of a jackpot, defined by its quantity, were assessed with pigeons when responding had nearly ceased. The response-potentiating effects of the frequency of jackpot-reinforcer delivery were investigated when responding of pigeons was maintained in single (Experiment 3) or concurrent (Experiment 4) schedules of reinforcement. Effects of jackpots on resistance to change were assessed with rats in Experiment 5. The effects of jackpots in each experiment were either absent or unsystematic across the subjects, casting doubt on their utility in animal training. Possible factors contributing to the negative results are discussed.

Pausing as an operant: choice and discriminated responding.

The effects of intermittent schedules of reinforcement for pausing were evaluated in two experiments. In Experiment 1, across a series of conditions, a variable-interval (VI) baseline schedule, in which pigeons' key pecks produced food, alternated with conditions in which food was delivered according to a concurrent VI (for key pecking) tandem variable-time differential-reinforcement-of-other-behavior (DRO) 5-s schedule. Time spent pausing within a session was proportional to the reinforcement rates associated with the tandem schedule. To examine the control of pausing by antecedent events, Experiment 2 arranged a multiple schedule in which pecking and pausing in either component were maintained according to concurrent schedules like those used in the first experiment. The availability of reinforcement for pausing was signaled in one component while signals uncorrelated with reinforcement were presented in the other. Signaled reinforcement for pausing, relative to the presentation of uncorrelated signals, decreased time spent pausing, a finding consistent with existing research on the effects of signaled VI reinforcement for key pecking in pigeons. The results of the two experiments show that pausing functions as an operant in much the same way that discrete responses, like key pecks, do, and that pausing and other operants are similarly affected by both antecedent and consequent events.

Reduction of biting and chewing of horses using differential reinforcement of other behavior.

Biting and chewing by horses on crossties can result in injury to the handler and damage to equipment. Operant-conditioning techniques have been used to train horses and could be used to reduce or eliminate undesirable biting and chewing. Presently, a differential-reinforcement-of-other-behavior (DRO) schedule, in the context of a reversal design, was effective in reducing biting and chewing in two horses. In DRO schedules, a reinforcer is delivered contingent on the absence of a target behavior for a specified interval. Positive-reinforcement procedures offer an alternative to aversive-control techniques typically used in equine training and may provide for better equine welfare and horse-human interaction.