What is positive reinforcement?: identifying the four quadrants of training.

Positive reinforcement. Negative punishment. Negative reinforcement. Positive punishment... huh?

Understanding the four quadrants of training can be daunting, but it is incredibly helpful in making training choices that achieve effective results.


The first thing to note is that 'positive' and 'negative' do not mean 'good' or 'bad,' but 'additive' and 'subtractive.' 'Reinforcement' is something that increases the likelihood of seeing a behavior while 'punishment' is something that reduces the likelihood of seeing a behavior. Let's dive in!


Positive reinforcement (R+)

When we use positive reinforcement, we are adding a consequence to an animal's behavior that increases the likelihood of that behavior happening again. For example, when we ask a dog to sit and he performs the behavior, we give him a treat; the next time we ask him to sit, he is more likely to perform the behavior because there was a good consequence the first time.


R+ methods are regarded as the most effective form of training because the dog creates a positive emotional response, achieving long-lasting results. We'll get into this deeper later in this article.


Negative reinforcement (R-)

When we use negative reinforcement, we are taking something away as a consequence of an animal's behavior, increasing the likelihood of that behavior happening again. For example, a dog is being choked by a collar when we ask him to sit, and when he performs the behavior, we release the pressure from his neck; the next time we ask him to sit, he is more likely to perform the behavior because it resulted in the end of discomfort or pain.


R- methods can increase anxiety and fear, which can lead to reactivity and aggression, and are therefore not recommended. As you'll see in the next section, negative reinforcement often goes hand-in-hand with positive punishment.


Positive punishment (P+)

When we use positive punishment, we are adding a consequence to an animal's behavior that decreases the likelihood of that behavior happening again. For example, when a dog walks ahead of us, we pull on a prong collar until he returns to our side; he's less likely to walk ahead of us again because it resulted in pain or discomfort. When we release the pressure of the collar when the dog returns to our side, this is also negative reinforcement - the dog is more likely to stay by our side because doing so resulted in relief from pain or discomfort.


Using P+ methods relies on causing pain, discomfort, or another unpleasant sensation. This often results in an increase in anxiety, fear, reactivity, and/or aggression. While P+ methods can certainly achieve short term results, it does so by suppressing behavior and can create a negative emotional response. For example, a dog barks at other dogs, and we shock him each time he barks; he is now less likely to bark because it causes him pain, but he now associates the other dog with being shocked. This can increase fear and/or anxiety, leading to increased reactivity or aggression. We've also taught him that he shouldn't communicate his fear by barking; the next time he is afraid, he may jump straight to growling, snapping, or biting instead. In many cases in which a dog 'bites out of nowhere,' it is because he has learned that other warning signals do not produce the desired result - making the scary thing go away.


Negative punishment (P-)

When we use negative punishment, we are taking something away as a consequence of an animal's behavior, decreasing the likelihood of that behavior happening again. For example, a dog is jumping on us, so we remove our attention until he stops; over time, he is less likely to jump because it resulted in the loss of something that he finds valuable.


Care should be taken when using P- methods to also redirect the dog to a behavior that we do want to see so that he can then receive positive reinforcement. Without this second step, the dog may become frustrated, experiencing an 'extinction burst' in which the behavior increases or escalates before it goes away. In the case of the jumping dog, we might ask the dog to sit once he has stopped jumping, then reinforce the behavior by giving our attention (and the next time, we want to catch him before he starts jumping and ask for that sit).


Can you identify the quadrant for each of these examples?

Find the answers at the bottom of this article.

Ex 1: Every time a dog drops his ball at our feet, we play with him.

Ex 2: When a dog jumps up on us, we knee him in the chest.

Ex 3: We pull on a dog's collar until he comes to us.

Ex 4: We stop playing with a dog when he mouths us.


So why use positive reinforcement?

In 1898, Edward Thorndike's Law of Effect stated that "responses that produce a satisfying effect in a particular situation become more likely to occur again in that situation, and responses that produce a discomforting effect become less likely to occur again in that situation."* In the 1930s and 40s, B.F. Skinner took this a step further with his study of operant conditioning, in which an individual "makes an association between a particular behavior and a consequence."** In the years since, dozens of studies involving a variety of animals, including humans, have found these statements to be true.


R+ is widely regarded as the most effective form of training as it produces long-lasting results - and without the use of pain, fear, or force.


By using operant conditioning, delivering a reward as a consequence for the behaviors we want to see, we help the dog create a positive emotional response. This is especially helpful in situations in which a dog is displaying reactivity or aggression due to fear or frustration. As the dog starts feeling safer, more confident, and less afraid or frustrated, he no longer needs to display behaviors such as barking, growling, or snapping.

Hierarchy of Procedures for Humane and Effective Practice

Below is the Certification Council for Professional Dog Trainers' guide to their ideal decision-making process for training.*** You'll see that considering medical factors and the environment (antecedent) comes before any training or behavior modification. While the CCPDT does not necessarily forbid the use of positive punishment, many training companies - including K9 Turbo - do not use these methods in any situation.


1. Health, nutritional, and physical factors: Ensure that any indicators for possible medical, nutritional, or health factors are addressed by a licensed veterinarian. The consultant should also address potential factors in the physical environment.

2. Antecedents: Redesign setting events, change motivations, and add or remove discriminative stimuli (cues) for the problem behavior.

3. Positive Reinforcement: Employ approaches that contingently deliver a consequence to increase the probability that the desired behavior will occur.

4. Differential Reinforcement of Alternative Behavior: Reinforce an acceptable replacement behavior and remove the maintaining reinforcer for the problem behavior.

5. Negative Punishment, Negative Reinforcement, or Extinction (these are not listed in any order of preference):

Negative Punishment– Contingently withdraw a positive reinforcer to reduce the probability that the problem behavior will occur.

Negative Reinforcement– Contingently withdraw an aversive antecedent stimulus to increase the probability that the right behavior will occur.

Extinction – Permanently remove the maintaining reinforcer to suppress the behavior or reduce it to baseline levels.

6. Positive Punishment: Contingently deliver an aversive consequence to reduce the probability that the problem behavior will occur.

At K9 Turbo, we believe in using force-free, science-based methods.

Safety and well-being are our top priorities, always. Positive reinforcement and force-free management is the best way to keep all animals and people safe.


And we're here to help! Contact us with questions anytime - we're happy to help you find a Certified Professional Dog Trainer in your area, too.


Answers:

Ex 1: Positive reinforcement. We are adding play, increasing the likelihood that the dog will drop the ball at our feet.

Ex 2: Positive punishment. We are adding pain, decreasing the likelihood that the dog will jump on us.

Ex 3: Negative reinforcement. We are taking away the discomfort of the collar pull, increasing the likelihood that the dog will come to us.

Ex 4: Negative punishment. We are taking away our attention, decreasing the chance that he will use his mouth in play.


References

*Edward Thorndike: The Law of Effect

** Skinner - Operant Conditioning

***Least Intrusive, Minimally Aversive (LIMA) Effective Behavior Intervention Policy

Proudly serving Southeast Michigan

248.733.5738

info@k9turbotraining.com