Teaching dogs to behave in response to a command is an integral part of enjoying canine companions. Learning to sit, stay and walk on a leash make it possible for dogs and humans to co-exist peacefully in a variety of situations. In training these behaviors, reward based reinforcement is recommended as the safest and most effective option.
Reward based training is the process by which each desired behavior results in a positive outcome and undesirable behavior does not yield the desired outcome. For example, the dog sits and is given a treat. The dog does not sit and no treat is forthcoming.
Using positive reinforcement or rewards to train behavior creates a pleasant training environment for both trainer and dog. Dogs that are given treats, physically rewarded (petting), and praised in response to offering desired behaviors are typically more willing and anxious to work to learn new things than those who are trained via use of aversive stimuli.
Rewards vs. Punishment
A simple example is the dog trained with positive reinforcement happily wagging his tail and offering to sit or shake while waiting for his owner to start dispensing treats. Alternatively, dogs who are accustomed to receiving punishment for bad behavior often shy away from additional training, such as moving away from a trainer applying a shock collar.
Both methods can work to modify behavior, but rewards offered as positive reinforcement are arguably more effective, more fun for owner and dog, and don’t have the potential negative side effects often seen with punishment.
Reward based training, also referred to as a form of operant conditioning by behaviorists, is focused on altering a dog’s voluntary behavior. The principle is that any behavior that is positively reinforced will occur with increased frequency. The dog that is petted when he sits quietly at the owner’s feet will continue to offer that behavior in an effort to achieve the desired outcome. Dogs that receive an owner’s attention after jumping on the owner will continue to offer that behavior.
Making Reward Based Training Most Effective
- The optimal time for reinforcement is 0.5 seconds after behavior has occurred.
- Giving the treat or patting the dog’s head immediately following the behavior helps the dog associate the reward with the desired behavior.
- Intermittently reinforcing behaviors is more effective that reinforcing them every time.
- Rewarding an animal every time they perform a desired behavior leads to rapid learning of a new task, but behaviors that are rewarded intermittently persist longer. Behaviors that are not rewarded will become extinct.
- High quality rewards, those the dog desires most will be most effective.
- Most dogs would prefer a piece of meat over lettuce.
- Training should be done when motivation to obtain the reward is high.
- A dog that is hungry will be more interested in working for food and a dog desiring companionship is more likely to work to be petted.
- Use small rewards to avoid satiation, which will decrease motivation.
- Small, “pea sized” pieces of food or a short duration of physical contact will keep a dog motivated for a longer period of time.
- Keep commands short.
- Dogs respond to simple commands such as “sit” or “stay” better than long sentences, such as “Please sit by my feet, Spot.”
- Be consistent.
- All those interacting with the dog should use the same words or phrases for the desired behavior.
- Be sure to reward only the desired behavior.
- Make sure to intermittently pet a dog sitting quietly at your feet. Using your hands to remove a dog only after he jumps, rewards him with attention and physical contact.
Most reward based training is done via successive approximation where the desired behavior is trained gradually in a series of increasingly more complex steps.
For example, when training a dog to go into a crate, the dog might first be rewarded for going near and investigating the crate. The dog then associates the crate as being a good place to be where rewards are given. Next, the trainer might reward the dog when he touches the crate, enters the crate, lies down in the crate, etc.
Some trainers also use a conditioned reinforcer to aid in training. A popular example is the clicker. With the clicker, an irrelevant sound is made by the trainer alongside the delivery of the reinforcement, reward or treat.
An example might be that the clicker sounds each time the dog sits or lies down in the crate. Over time, the clicker can replace the treat as the method of reinforcement. Intermittently offering the clicker or food when the dog is in his crate can increase the longevity of the dog completing the desired behavior. In this instance the behavior would be lying down in the crate.
Rewards All Around
Operant conditioning using positive reinforcements or rewards is one of the most effective methods for training dogs and is devoid of serious side effects that might damage a human-animal bond. Reward based training is excellent at training dogs to do new tasks, but not as efficacious at solving pre-existing behavior problems.
Operant conditioning can be used to suppress undesired behaviors such as excessive barking or property destruction, but it ignores emotional motivation which is what drives behavior. Excessive barking may be due to anxiety or fear and property destruction may be related to separation anxiety.
Dogs that are in a good emotional state are the most capable of learning new and desired behaviors. If a dog is behaving in an aggressive or destructive manner that threatens its well-being or the well-being of other people and animals, consultation with an animal behavior expert and qualified dog trainer is strongly recommended.