What is positive reinforcement? Hear from our experts
Our behavioral team at the Dumb Friends League uses positive reinforcement training to support the animals in our shelters. If you’ve ever wondered what positive reinforcement entails, you’re not alone. The League’s Director of Shelter Behavior and Veterinary Services, Jes Cytron, and Behavior Team Leader, Zoë Knox, recently explained all the ins and outs of this compassionate, science-based approach.
What is positive reinforcement training? Which species tend to respond to positive reinforcement?
Jes (J): Positive reinforcement training (R+) refers to a conditioning approach to learning where we reinforce behavior that we want to see, increasing the likelihood that the animal will repeat that behavior. I've heard of or witnessed R+ leading to desired outcomes for dogs, cats, rabbits, ferrets, guinea pigs, bearded dragons, fish, pigs, cows, snakes, and many more!
Zoë (Z): R+ is a method of training that utilizes something inherently valuable for the learner in order to encourage an increase in behavior. The trick is to find reinforcement valuable enough for the learner to stay engaged. For your dog or cat, that might be a small piece of plain chicken. For most adult humans, this could be a monetary reward. When I worked with seals and sea lions, we used whole fish. Any animal, [young or old], is subject to Learning Theory.
Also, I should mention that certain breeds are not more defiant than others. However, they often have different needs, skills and intelligence, and activity levels. For example, I keep my training sessions with shepherds or huskies relatively short, and I provide a high rate of reinforcement so they [stay focused on me].
Why is R+ the League’s preferred training method?
J: Generally, we prefer this method because it is well-researched and backed by science. We also achieve a desired outcome from an animal because of their (assumingly) consensual cooperation, rather than a need to avoid an aversive stimulus/punishment.
Z: The League mainly uses R+ methods, with some negative punishment training (P-). P- involves removing your attention to decrease an unwanted behavior, such as turning away from a dog who’s jumping up to interact with you. R+ is effective at building trust and confidence. We don't need to utilize emotional or physical pain in order to see results. We love and care for the animals we work with, and we want to reduce their stress in any way possible.
How does it work? What are ways to reward desired behavior?
J: R+ works by capturing behavior that you want to see repeated, marking the behavior (you can use a clicker or a verbal marker), and providing a reward. Most commonly, this reward is high value food items like hot dogs, cheese, and training treats, but other reinforcers could be toys and praise.
Z: There are two forms of R+: classical and operant conditioning. Classical conditioning is not dependent on the animal’s behavior. We are simply pairing a trigger (e.g., approaching their kennel) with positive reinforcement (e.g., offering treats whenever we approach). In the shelter, we may first use classical conditioning with highly fearful animals.
With operant conditioning, we use a marker to pinpoint a specific behavior and follow it with positive reinforcement. For example, you use a clicker when the animal does the desired behavior and always follow with a treat. This tells them, "YES – that thing you were just doing when I clicked, do that again and you’ll be rewarded!" You can also "capture" behaviors the animal offers normally.
Is it effective? Why shouldn’t I use a tool like a spray bottle or a prong collar to stop the behavior?
J: R+ is highly effective. I do not recommend using tools that punish as they often contribute to an unintentional association between the aversive stimulus (the water, the pain) and the trigger the animal is already having problematic reactions to (strangers, other animals, etc.). It’s easy to understand why tools that fall under the punishment umbrella are alluring to folks who have animals with behavior issues they need to manage. These often suppress unwanted behaviors quickly, but temporarily.
Z: R+ is very effective but requires additional considerations. We need to utilize management tools to prevent unwanted behaviors as much as possible and train the alternative desirable behavior in conjunction with it. For instance, utilizing baby gates will help prevent a dog’s house soiling by making it [clearer] that the correct place to potty is outside.
Likewise, a spray bottle may interrupt a behavior while it's occurring, but the animal may learn to only do that behavior when they or you are out of sight. Also, punishment methods like a spray bottle don't tell the animal what to do instead. Using punishment methods for naturally occurring behaviors (e.g., pottying indoors) only confuses and frustrates the animal and could harm your relationship as they may [grow to fear you].
How long should I expect to wait before I see results and how do I manage the behavior in the meantime?
J: This entirely depends on what you are working on and what resources (e.g., professional trainers, classes) you are using along your training journey. To manage concerning behavior towards strangers, for example, I would recommend choosing walk times strategically, when less triggers are milling about. You can also advocate for your dog according to how they respond [to strangers] in that moment and where they are in their training.
Z: I wish I could give people a more exact time frame, but there are often so many factors influencing the behavior modification process (e.g., intensity of the trigger or behavior, overall stress level, consistency of training), it can be hard to predict. I can say I often see animals catch on to learning new behaviors quickly with clicker training because that communication method is so precise. But it's just one part of the equation.
What are some examples of how to apply R+ for a specific behavior issue?
J: One example is teaching a dog to target your hand with their nose. You offer your palm to the side and start by marking and rewarding (so, clicking and treating) whenever you see your dog move towards your hand with their nose. Then you add some complexity by moving your palm and marking/rewarding for full engagement with the hand target. This can be used to address high arousal or “mouthy” behavior for a puppy learning to interact with people without using their sharp, little teeth. This also teaches puppies impulse control and to take treats with a soft mouth.
Z: If your dog is fearful of strangers, make sure you aren't punishing their communication. We love a growl! It tells us the dog is uncomfortable or feeling fear, and we don't want to lose that warning signal. Instead, try to make sure each stranger your dog sees from a distance is paired with treats. You can then use clicker training to teach hand targeting, which can help the dog build confidence around strangers and interact with them in an appropriate way.
Where do I start with my own pet?
Z: First, make a plan by creating a list of management strategies your household can [realistically] implement to prevent the behavior as much as possible. Then, look for behaviors you can train that are incompatible with the unwanted behavior. Schedule 15-minute training sessions each day and track your progress. If you are new to clicker training, I suggest starting with hand targeting, eye contact/capturing attention, and recall behaviors. Finally, look for local certified trainers and online resources. There are often group classes or private lessons you can take to hone your skills and receive feedback. Additionally, the League offers helpful behavioral resources at ddfl.org/pet-behavior-resource-center.
Lastly, what’s your favorite thing about your job?
J: Currently, my favorite thing about my job is supporting and developing people who share my passion for animal behavior and science and being developed by them in turn. Knowledge sharing for the purpose of bettering the human-animal bond and, therefore, our communities.
Z: I live for those lightbulb moments where we can see [an animal’s] gears turning, and they start to catch on to the behavior I'm training. I especially get a lot of reinforcement from my work during the summer kitten season. Seeing an animal take a brave step to trusting me when they honestly may not have had any good reason to trust humans previously really warms my heart.