Five graduate students from the University of North Texas (UNT) presented their research last May at the annual meeting of the Association for Behavior Analysis in Atlanta. Over 2,000 behavior analysts from all over the world attend this meeting. The students prepared their work under the auspices of their professor, Jesús Rosales-Ruiz of the UNT department of behavior analysis, who is also a popular member of the ClickerExpo faculty.
Students presented their original clicker-related research in a symposium and in poster sessions. Studies included the effects of the so-called poisoned cue (resulting from the introduction of correction procedures in otherwise positively trained behavior) in dogs; a method of using negative reinforcement to reduce aggression in dogs; a use of the clicker in teaching animals such as cows and horses to tolerate being approached; and a study of variability of clicking and reinforcement.
Dr. Rosales-Ruiz reported that a packed crowd of interested professionals filled the room, sat on the floor, and jammed the doorway. This innovative research has potential applications for many areas of human learning and teaching. The abstracts of Dr. Rosales-Ruiz' students are presented below.
Some Detrimental Effects of Combining Positive and Negative Reinforcement During Training
NICOLE BYRD (University of North Texas)
Jesús Rosales-Ruiz (University of North Texas)
Abstract: Karen Pryor suggests that a cue, or SD, established with a combination of negative and positive reinforcement leads to the breakdown of the behavior both preceding and following the cue due to an increase in avoidance behaviors and the uncertainty that exists in terms of the consequence that will follow. The purpose of this presentation is to compare the effects of combining negative reinforcement and positive reinforcement during training with the effects of using positive reinforcement alone. Two dogs served as subjects. Each dog learned the same behavior under two different stimulus conditions associated with two different training methods.
One method involved the presentation of the cue "ven," and the reinforcement of successive approximations to the target behavior. The other method involved the presentation of the cue "punir," the physical prompting of the target behavior by pulling the leash, and the delivery of a reinforcer. Differences in the behavior between the two conditions are documented, as well as differences in the stimulus functions of "ven" and "punir."
Transition from Negative to Positive Reinforcement during Shaping Cows' Approach to Humans
MELISSA MOREHEAD (University of North Texas)
Jesús Rosales-Ruiz (University of North Texas)
Abstract: Negative reinforcement can be a powerful tool for behavior analysts, yet it is often overlooked as a treatment method. Pryor (1999) outlines a method for approaching a "timid" animal using a combination of negative reinforcement and positive reinforcement. When the animal stands still, the human operates a clicker, and then retreats from the animal. Gradually, the human moves closer to the animal through the clicking and retreating shaping process. Once the human is standing close enough, food may be offered as a positive reinforcer, and the negative reinforcer is canceled out. The purpose of this study was to experimentally demonstrate the click-retreat technique with cows. A multiple-baseline design across subjects was used to test this technique. Results show that the click and retreat technique was effective. Results are discussed in terms of the difference between the click-retreat technique and systematic desensitization.
Aggression in Dogs: A Differential Negative Reinforcement Protocol
KELLIE S. SNIDER (University of North Texas)
Jesús Rosales-Ruiz (University of North Texas)
Abstract: Aggression in dogs is a serious problem in the United States, with over 4,700,000 people bitten annually, and over 800,000 requiring medical treatment (K. Delise, 2002). Common aggressive behaviors seen in dogs include barking, chasing, lunging, snarling, growling, and biting. Many such behaviors are maintained by the removal or distancing of other people or dogs. The training protocol presented shows the effects of differential negative reinforcement on aggressive behavior of dogs. Aggressive dogs were tethered on six-foot leashes. An experimenter walked toward the dog until the dog performed an aggressive behavior. She stood still until the behaviors stopped, then walked away contingent upon a desirable behavior such as looking toward the owner, or turning away from the experimenter.
In subsequent approaches the experimenter stopped walking toward the dog before the point at which the dog had previously performed the behavior, and exited based on performance of a desirable behavior.
Proximity was systematically increased. If at any time the approach provoked aggressive behavior, the experimenter stopped and waited until the behavior ended before exiting. The aggressive behaviors diminished or were eliminated with as little as one hour of treatment with some dogs. Successful generalization occurred in several cases.
Effects of Unconditioned and Conditioned Reinforcement Ratios on the Behavior of Dogs: 1 to 1 vs. 1 to 2 (Applied Behavior Analysis)
KATHRYN KALAFUT (University of North Texas )
Michelle Lamancusa (University of North Texas )
Jesús Rosales-Ruiz (University of North Texas)
Abstract: In the clicker training community it is a common practice to deliver a click (conditioned Sr+) and a treat (unconditioned Sr+) after every correct behavior. At the same time, some advocate the delivery of several clicks before a treat is delivered. There is much controversy over whether there is a difference in the effects of these two procedures. Recent research, however, has shown that the ratio has an effect on both the topography and frequency of behavior (Dunham, et. al., and Wennmacher, et. al., ABA 2005). Dunham at. al., studied the effects using a free-operant under FR1 as a baseline and Wennmacher, et. al., used two restricted operants under a FR1 schedule. The present research investigates the effects of these different reinforcement ratios using a free operant under a FR2. A multiple baseline across three dogs with reversals was used. After a baseline was established with a FR 2 schedule (clicking and treating after every two correct responses), each of the dogs correct responses was followed by a click, but reinforcement was only given after two correct responses were completed. Results are in progress.
Variability: Generalizations Across Stimuli and Contingencies (Experimental Analysis)
FRANCISCO MANUEL GOMEZ (Ideal Companion Canine Behavior)
Jesús Rosales-Ruiz (University of North Texas)
Abstract: Variability as an operant has been evidenced in the behavior analytic literature over the last thirty years. The present study was designed to evaluate the generalization effects of programming a variability contingency on a dog's interaction with one object and generalization of variability across other objects and contingencies. A six-year-old Border Collie's interactions with four different objects were measured under two different contingencies:
ANY (where any physical contact with the object would be reinforced on an FR1) and VAR (where only the novel responses per trial would be reinforced). All four dog-object interactions were first studied under the ANY contingency. This contingency produced stereotyped responding of behavior with all objects. Then, one of the dog-object interactions was changed to VAR while the other three remained under the ANY contingency. The VAR contingencies yielded a marked decrease in stereotypic behavior and an increase in novel responses. Similar effects were also seen with the other object-dog interactions under ANY, where reinforcement criteria didn't require variable behavior. An analysis of variability of interactions in both VAR and ANY clearly show that novel topographies generate from the adduction of already existing behavior. There were no novel responses that lacked topographical elements of already trained behavior.
The Effects of Two Methods of Conditioning a Clicker on Behavior Variability During Training (TPC; Applied Behavior Analysis)
MICHELLE LAMANCUSA (University of North Texas)
Kathryn Kalafut (University of North Texas)
Jesús Rosales-Ruiz (University of North Texas)
Abstract: The stimulus-stimulus contingencies to establish an event as a conditioned reinforcer are clear in the literature. However, the reinforcer-behavior contingency during the conditioning process is not clear or considered relevant.
In animal training, non-contingent reinforcement (NCR) has traditionally been used in clicker training. Recently it has been suggested that the contingency be response specific from the beginning of clicker conditioning. The purpose of this research is to determine the effects of two different methods of conditioning the clicker on behavior variability during training. The subjects were three dogs. In the first condition, the clicker is conditioned in the traditional way, beginning with non-contingent reinforcement (NCR) and then proceeding with reinforcement contingent on eye contact (the target response). In the second condition, clicking and reinforcement is contingent on eye contact without a previous NCR schedule. After establishing eye contact with reliability, a two-minute period of extinction is in place for each condition. Following extinction, differential reinforcement of any behavior is in effect. It is expected that pairing with NCR will produce increased variability as compared to the contingent pairing. Results will be discussed in terms of side effects resulting from the pairing procedures (i.e., prompt dependency). Results in progress.
Karen Pryor is the founder and CEO of Karen Pryor Clicker Training and Karen Pryor Academy. She is the author of many books, including Don't Shoot the Dog and Reaching the Animal Mind. Learn more about Karen Pryor or read Karen's Letters online.
Post new comment