Canine-canine Behavior Problems

Vol.18, No. 9

September 1996

Continuing Education Article

FOCAL POINT ★ A program of behavior modification requires from the owner a considerable investment of motivation, time, and understanding.

KEY FACTS ■ Behavior modification can involve avoidance and changing the animal’s response to stimuli. ■ Acquisition of new behaviors is facilitated by a continuous reinforcement schedule. ■ Obedience classes teach commands via a combination of positive reinforcement, negative reinforcement, and punishment techniques. ■ Agility training teaches owners how to apply learning principles effectively by training dogs to navigate obstacles.

Canine Behavior Problems: Behavior Modification, Obedience, and Agility Training* Angell Memorial Animal Hospital Boston, Massachusetts

Pet Behaviour Innovations Toronto, Ontario

Amy Marder, VMD

Pamela J. Reid, PhD

A

treatment program to correct canine behavior problems may include components derived from three main areas: environmental manipulation, physiologic intervention, and behavior modification.1 In general, environmental manipulation is easily accomplished, even by inexperienced owners. Physiologic intervention, by means of surgical procedures (e.g., castration) or drug therapy, also is easy to implement. Environmental manipulation and physiologic intervention do not require much time, motivation, or understanding from the owner. In many cases, however, behavior modification is necessary to realize long-term change in a dog’s behavior. Unlike environmental manipulation and physiologic intervention, a behavior modification program does require considerable time, motivation, and understanding on the part of the owner.

BEHAVIOR MODIFICATION Behavior modification involves the implementation of learning principles to change a behavior. Most behavior modification programs rely heavily on positive reinforcement and on techniques that involve operant and classical conditioning. Avoidance and actually changing the animal’s response to stimuli may be involved. In treating a dominant-aggressive dog, for example, an owner might be advised to first avoid provoking aggressive behavior. Avoidance prevents the owner from being injured and averts aggressive encounters that might *This article originally appeared in Voith VL, Borchelt PL: Readings in Companion Animal Behavior. Trenton, NJ, Veterinary Learning Systems, 1996, pp 56–61.

Small Animal

The Compendium September 1996

DO-IT-YOURSELF SIT–STAYa by Amy Marder, VMD This program is designed to teach dogs to sit and stay and to make learning easy for owners and pets. No punishment is used at the beginning—only verbal praise and food. First, a few basics are necessary. The best way to teach a new behavior is by continuous reinforcement. Every repetition should initially be reinforced with food. Because you cannot always have food available, the food is paired with verbal praise. This pairing of food with your voice makes the latter a more powerful reinforcer. Eventually, intermittent reinforcement is introduced and the food treat is not offered every time. This approach makes the sit–stay behavior more permanent. You should select a food treat that your dog loves, preferably something that is low in calories and can be broken into little pieces. Beef jerky treats, boned chicken, and cheese are reasonable choices. Some dogs work for rice cakes. The reward must be given immediately. Timing is crucial. You should reinforce only desired behavior. If reinforcement is delayed, the dog may bark; at this point, the food rewards the bark rather than the stay. You must proceed in the exact order of the program. The steps are designed to make learning very easy. Dogs learn by being successful. If you proceed too rapidly, your dog will make mistakes and both of you will become frustrated. If the dog makes a mistake (e.g., breaks the stay), you should sternly say “no” or “uh-uh,” should not give the food reward, and should reposition the pet in the sit mode and repeat the task. Sometimes your dog may become bored or refuse to work regardless of what you do. On such days, do not push yourself or your pet. Instead, end the training session on a positive note with a few easy, short stays (e.g., count 5) and try again tomorrow. You should end each session by giving your dog a release word. Most people just say “okay,” but each owner can coin his or her own phrase. A food

reward should not be given when the dog is released, as doing so trains the pet to stand up. Daily practice is important. It is best if every person in the household practices with the dog. Each session need only last 10 minutes. You should teach the sit–stay command first in the house and then outside, in a fenced yard. If the yard is not fenced, the dog can be tied to a long lead for outside practice. When rewarding the dog, offer only verbal praise and food. It may be hard for you to resist petting the animal, but petting can be very distracting and should be avoided at first. Teaching the “sit” command involves prompting and fading techniques and positive primary and secondary reinforcement. Place the food in one closed fist and pass your hand over the end of the dog’s nose, over its head, and toward its rear as you say the word “sit.” Usually, when the dog’s head goes up, the rear end goes down. When the dog’s rear hits the ground, praise the dog immediately and give the food reward. Repeat this exercise 20 times, and then omit passing your hand over the dog’s head. Instead, now use the “sit” hand signal (an open hand held horizontally and raised). Repeat this exercise 20 times, giving a food reward each time. After your dog understands the “sit” command, you can start teaching the pet to stay. Teaching the “sit–stay” command involves prompting. Begin with your dog in the sit position. Hide a handful of food tidbits in one hand (do not dangle the food in front of the dog; keep it hidden). With your other hand, give the stay signal (an open hand with the palm toward the dog) while saying “stay.” Praise the dog as soon as it completes the task successfully (e.g., after count one or after taking one step); you can say “good dog,” “good boy,” “good girl,” or whatever words of praise you prefer. Then, with the hand you used for signaling, offer the pet a food reward. Keep your dog in the sitting position.

The Sit–Stay Exercise Owner Task

Dog Completed

Number of Times Performed

Count 1 Count 2 Count 3 Count 5 Count 7

___________ ___________ ___________ ___________ ___________

___________ ___________ ___________ ___________ ___________

The Compendium September 1996

Small Animal

DO-IT-YOURSELF SIT–STAY (continued) Owner Task

Dog Completed

Number of Times Performed

Count 10 Count 15 Count 20

___________ ___________ ___________

___________ ___________ ___________

___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________

___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________ ___________

Now you are ready to do some dancing! One step left and return One step right and return One step back and return Two steps left and return Two steps right and return Two steps back and return Three steps left and return Three steps right and return Three steps back and return Count 10 Two steps right, count 10 Two steps left, count 10 Two steps back, count 10 Five steps right and return Five steps left and return Five steps back and return One step back, about-face, return Two steps back, about-face, return Three steps back, about-face, return Count 20 Ten steps right and return Ten steps left and return Ten steps back and return Count 10 a

Modified version of Do-It-Yourself Sit–Stay Program, developed by Victoria Voith, DVM, PhD. Copyright © 1977 by Victoria Voith.

cause the dog to assert a dominant position over the owner.2 Physical punishment is usually avoided in these cases because it might cause a dog to become more aggressive and potentially dangerous. The behavior modification program for a dominantaggressive dog also might include conditioning it to assume positive reactions and nonaggressive behavior in specific situations that could otherwise elicit aggressive responses. If the dog displays aggressive behavior during grooming, for example, the owner can first teach the dog to sit or stand and stay for food treats. Grooming implements then can be introduced gradually (desensitization). The dog receives treats for tolerating first the sight and then the use of the implements (counter-

conditioning). A soft brush might be used initially; after the dog tolerates being groomed with this brush, firmer brushes can be introduced. The dog should receive a treat during each stroke of grooming and as a reward after the stroke. Although behavior modification may be the most important part of a treatment program, it is often the most difficult for owners to implement. Most programs are time-consuming and require an owner to understand some simple learning principles. Many owners abandon the behavior modification program early; others fail to initiate a program. Wanting a quick fix, owners may change the environment, give a drug, or avoid confronting problem behaviors but actually do little to

DOMINANT POSITION ■ NONAGGRESSIVE BEHAVIOR ■ GROOMING IMPLEMENTS

Small Animal

The Compendium September 1996

change the pet’s responses to stimuli that elicit the problem. Owners thus fail to understand the cause of the behavior problem and why the problem is continuing or becoming worse. A dog that exhibits territorial aggression when humans enter the house can be handled by avoiding the problem or modifying the behavior. An owner may avoid the problem by confining the dog to the back- Figure 1—A dog practicing the sit–stay command. yard every time a person visits. This method does not change the dog’s response at the door; if the dog is subsequently allowed in the house when humans enter, the aggressive behavior is likely to recur. It may even be more severe because territorial aggression often becomes worse with time. Long-term change will only result from a program that modifies the dog’s behavior toward humans who enter the house.

sponse is rewarded. Acquisition is promoted when the interval between the response and the reward is short (approximately 0.5 seconds for most behaviors). By contrast, maintaining a behavior is best achieved via an intermittent reinforcement schedule; the response is reinforced irregularly. Intermittent reinforcement schedules can involve intervals or ratios. In a ratio schedule, reinforcement follows a certain number of responses. In an interval schedule, reinforcement follows the first response that occurs after a specific interval has elapsed since the animal was reinforced. Schedules can be fixed or variable. A fixed-ratio (FR) schedule provides reinforcement after a specified number of responses has occurred (e.g., an FR 5 schedule means reinforcement occurs after every fifth response). A fixed-interval LEARNING PRINCIPLES Figure 2—Teaching a dog to go over a seesaw during agility (FI) schedule provides reinProfessionals and owners training. forcement for the first remust understand some elesponse after a specified inmentary learning principles terval (in an FI 5 schedule, to design or implement an effective behavior modifireinforcement occurs after a 5-second interval has cation program. The following is a brief introduction elapsed since the previous reinforcement). to frequently used behavior modification techniques. A variable-ratio (VR) schedule provides reinforceMore detailed information and definitions of terms asment after a variable number of responses (in a VR 5 sociated with behavior modification are available in the schedule, on average, every fifth response is reinforced; literature on learning theory.3,4 the actual number of responses before reinforcement Positive reinforcement refers to application of a stimumay vary). A variable-interval (VI) schedule provides lus or event (a positive reinforcer) that follows an anireinforcement for the first response after a variable inmal’s response and increases the probability that the terval (in a VI 5 schedule, reinforcement occurs after response will recur. Negative reinforcement refers to rethe first response following a variable interval; on avermoval of an unpleasant stimulus (a negative reinforcer) age, the interval is 5 seconds). after an animal’s response; such reinforcement also inAll types of intermittent reinforcement schedules creases the likelihood that the response will recur. Reinproduce stronger responses than do continuous reinforcers can be used to modify behaviors that already exforcement schedules. Likewise, variable schedules (VR ist in an animal’s repertoire or to establish new and VI) produce higher rates of response than do fixed behaviors. schedules (FR and FI). Behavioral histories usually Acquisition of a new behavior is facilitated by a condemonstrate that dogs with problem behaviors are on tinuous reinforcement schedule in which every reintermittent reinforcement schedules. For example, a TERRITORIAL AGGRESSION ■ POSITIVE REINFORCEMENT ■ INTERVAL SCHEDULE

Small Animal

The Compendium September 1996

dog that barks to be fed at begins to decrease in frethe table is probably operatquency and eventually dising on a mixed variableappears. After a behavior interval and variable-ratio has been eliminated via exschedule. The dog barks a tinction, it may reappear if variable number of times the pet is not periodically and for variable periods beexposed to the stimulus. fore the owner periodically This reappearance is regives in and feeds it. ferred to as spontaneous Intermittent reinforcerecovery. ment schedules can be efShaping is a procedure fectively incorporated into that is commonly used to behavior modification proestablish new behaviors by grams to change problem reinforcing successively behaviors. Using food re- Figure 3—A border collie descending an A-frame. closer approximations to wards, a continuous reinthe desired response. With forcement schedule can be each successful increment, implemented to teach a dog previous levels of the behavnew, appropriate behavior ior are no longer reinforced. (e.g., sitting when humans Prompting and fading enter the home). Once the techniques also are useful behavior has been estabfor teaching new behaviors. lished, the dog can be A prompt is a stimulus that switched to an intermittent manipulates the animal to reinforcement schedule of perform the desired rerewards that maintain the sponse so that it can be regood behavior. inforced. The strength of Extinction occurs when the prompt is gradually rereinforcement for behavior duced in intensity (faded) is withdrawn and, as a conuntil the animal performs sequence, the animal stops Figure 4—A border collie weaving through poles. the behavior without being the behavior. Many owners prompted. For example, a are quick to use punishment dog can be taught not to to eliminate undesirable take objects by using the behavior; however, serious prompt of a sudden movedrawbacks may be associated ment toward the dog’s face with punishment. Extincwith the hand that holds tion can be used as an alterthe object. The command native. “leave it” is given simultaneAlthough extinction has ously. The dog is rewarded some inherent problems, the with a treat when it moves technique is generally well away from the training obaccepted if the owner is ject. The threatening movewarned of what to expect. ment to the face is faded Typically, when reinforceover time, and the dog is ment for a behavior is withtaught to stay away from drawn, the frequency of the Figure 5—Teaching a dog to down–stay on a table during the training object (or behavior increases dramati- agility training. “leave it”) and subsequentcally (an extinction burst). ly to stay away from other The burst may be accompaobjects. nied by the temporary appearance of displacement acSecondary reinforcers (second-order or conditioned tivities and other behaviors. If the reinforcement conreinforcers) are stimuli that have been associated with tinues to be withheld, however, the unwanted behavior primary, or biologic, reinforcers and have assumed reinVARIABLE-RATIO SCHEDULE ■ PROMPTING AND FADING TECHNIQUES

Small Animal

The Compendium September 1996

forcing properties as a result of this association. The most common secondary reinforcer is verbal praise (“good dog!”) because it is often paired with food treats and petting. If it is impossible to deliver a primary reinforcer immediately after the pet’s response, a secondary reinforcer can be used to reinforce the correct response and signal the forthcoming delivery of a primary reinforcer. Periodic pairing of secondary and primary reinforcers is necessary to maintain the reinforcement properties of the secondary reinforcer. Chaining is a procedure that relies on conditioned reinforcers to maintain a series of responses. Backward chaining involves conditioning the final response in a series first, conditioning the response that immediately precedes it next, and so on. This process is continued for the entire response series. The final response is always reinforced with a highly motivating primary reinforcer. Backward chaining can be used to teach a dog to sit and stay when humans enter the house. First, the dog is rewarded for sitting while being greeted by a visitor; then the approach of the person is introduced, followed by the person walking through the door. At each stage, only the final response of sitting while being petted is reinforced with a powerful primary reinforcer. Forward chaining involves working in the reverse order: the first response is initially established, then the second response, and so on. The last response in the chain is reinforced with the primary reinforcer. A behavior modification program will be easier, more enjoyable, and more likely to be implemented if the owner understands learning principles and if the program design allows frequent rewards with small successes. The owner is also likely to view the dog more favorably after working with it in pleasurable ways and seeing frequent (even though small), positive changes. The owner thus may be less likely to abandon the program and the pet. Dr. Victoria Voith has developed a do-it-yourself sit–stay program (Figure 1) (see the box for a shortened version). This autotutorial teaches an owner how to use positive reinforcement and shaping techniques. The program is easy and fosters immediate success for the owner and the dog. After completing each step, the dog is rewarded with a tidbit of food to maintain interest and motivation. The program is repetitious to enable the owner and dog to proceed with few or no mistakes. As owners progress through the program, they feel immediately successful and are usually eager to go further. After they have finished the program, owners may incorporate desensitization and counterconditioning procedures. The stimuli that provoke problem behaviors (e.g., the doorbell ringing, the door being opened, a person entering the house, and the person approach-

ing the dog) are gradually introduced. In each case, the dog continues to sit or down–stay and receives food treats (Figure 1). While the dog is in a nonaggressive mood, it is exposed to stimuli that previously elicited an aggressive response. The same program can be adapted using stimuli that evoke fear or anxiety. A continuous reinforcement schedule is applied first, followed by intermittent reinforcement to maintain behavior.

OBEDIENCE CLASSES Before clinical animal behavior became a recognized specialty, obedience classes were routinely recommended for dogs with behavior problems. Obedience classes, especially puppy classes, probably help to prevent the development of some behavior problems and help owners to gain some control. Once a serious problem has developed, however, routine obedience classes may do little to change the behavior.5 Most classes use combinations of positive reinforcement, negative reinforcement, and punishment techniques to teach commands. Many instructors continue to rely extensively on punishment and have minimal knowledge of the proper application of positive reinforcement or learning principles. Classes thus may become unenjoyable for owners and pets.6 The most common commands taught in obedience classes are sit, down, stay, heel, and come. These commands are rarely applied, however, in situations in which behavior problems occur. For example, commanding a dog to down–stay for a long period may prevent a dog from being destructive when an owner is present but does not stop destruction when the owner is absent (separation anxiety). Similarly, obedience training is often recommended (to prevent dogs from engaging in behaviors) without addressing the underlying motivation for the behaviors. For example, teaching a dog to stay might prevent a fearful dog from biting visitors as they enter the house (just as a gate would) but does nothing to reduce the dog’s fear. By contrast, a good behavior modification program teaches an owner how to control a dog via commands and uses counterconditioning to change the dog’s response to specific stimuli related to the problem. AGILITY TRAINING Agility training is a sport in which dogs are taught by their owners to navigate obstacles. It teaches an owner how to apply learning principles effectively and may maintain an owner’s interest in the behavior modification program. Typically, positive reinforcement, in the form of food and play, is used in the training process. Because dogs must be taught to negotiate the obstacles

CHAINING ■ SIT–STAY PROGRAM ■ OBEDIENCE CLASSES

The Compendium September 1996

gradually, owners can implement and then understand the learning principles involved in most behavior modification programs. Shaping, chaining, prompting, and fading are commonly used in agility training. For example, teaching the seesaw (one of the most difficult obstacles) requires the implementation of several principles. Because most dogs are afraid to walk on narrow, unsteady surfaces, this behavior must be shaped gradually. The dog is first taught to walk on a narrow plank via positive reinforcement. Next, the dog is introduced to a narrow inclined plank, again with positive reinforcement. After it successfully walks the inclined plank, the dog is taught to walk up the plank until it tips and then to tip the plank unassisted. Finally, by means of the forward chaining technique, the dog is rewarded for walking up the seesaw, tipping it, and walking off the seesaw after it touches the ground (Figure 2). Agility training also involves teaching a dog to run through a tunnel, burrow through a fabric tunnel, walk on a narrow elevated plank, climb and descend a 6foot-high A-frame (Figure 3), jump over and through various obstacles, weave through poles in slalom fashion (Figure 4), and down–stay briefly on a table (Figure 5). After the dog learns to maneuver each obstacle, owner and dog walk or run through a course that comprises numerous obstacles. At this point, backward chaining can be used. The dog is first rewarded for completing one obstacle, then for completing two obstacles, then for completing a series of three, and so on until the dog can successfully complete approximately 20 obstacles between reinforcements. Most dogs and owners enjoy running the course. The dog receives deliberate positive reinforcement from the owner as well as self-reinforcement. Agility training is a useful method of counterconditioning various problems. The method can be helpful in dogs that are aggressive toward other dogs, dogs that are afraid of humans, and dominant-aggressive dogs that resist being handled and manipulated. Agility training also teaches dogs to pay attention to their owners. Many owners cannot make their dogs pay attention during obedience classes or when implementing a be-

Small Animal

havior modification program. During agility training, dogs receive reinforcement only for following owner directions on the course; paying attention thus becomes easier.

CONCLUSION Several owner–clients have attended obedience classes, implemented a behavior modification program, and enrolled in agility training. In one case, the owner of a dominant-aggressive wire-haired fox terrier (which had bitten its owner twice, was aggressive toward dogs, and was afraid of humans) implemented all three. The owner reported that the obedience classes and behavior modification program helped resolve the dominance aggression and that the agility training ameliorated the aggression toward other dogs and the fear aggression toward humans. The owner believed that the agility tasks changed his relationship with his dog. Although the pet continued to growl at the owner occasionally, the owner no longer viewed the pet as a major behavior problem but instead as a companion to enjoy. About the Authors Dr. Marder is affiliated with Angell Memorial Animal Hospital in Boston, Massachusetts, and Dr. Reid is with Pet Behaviour Innovations in Toronto, Ontario.

REFERENCES 1. Voith VL: Clinical animal behavior. California Vet 33(3): 21–25, 1979. 2. Voith VL: Aggressive behavior and dominance. Canine Pract 4:8–15, 1977. 3. Domjan M: The Principles of Learning and Behavior, ed 3. Pacific Grove, CA, Brooks/Cole Publishing Co, 1993. 4. Schwartz B, Robbins SJ: The Psychology of Learning and Behavior, ed 4. New York, WW Norton & Co, 1995. 5. Voith VL, Wright JC, Danneman PJ: Is there a relationship between canine behavior problems and spoiling activities, anthropomorphism, and obedience training? Appl Anim Behav Sci 34:263–272, 1992. 6. Myles S: Trainers and chokers: How dog trainers affect behavior problems in dogs. Vet Clin North Am Small Anim Pract 21:239–246, 1991.