Positive reinforcement is the addition of a reward following a desired behavior
in order to increase the likelihood of the behavior continuing.
Today there are two schools of thought regarding training methods - one is coercion, using punishment; the other, using positive reinforcement, is reward-based. Let’s first take a look at coercion.
You put a slip (choke) collar on your new dog. If you look at it with an open mind, that collar works like a hangman’s noose, getting gradually tighter the more you or your dog pulls. Coercion trainers will tell you that it is the noise the collar makes as the chain slips through the ring that corrects unwanted behavior. But the truth is the noise is meaningless until it is associated with something unpleasant - the tightening of the noose. Then the noise will serve as a threat. A threat - this is not very good for establishing a relationship.
When the slip collar doesn’t work, coercion trainers may up the ante by using an e-collar. The “e” is for electric. Many e-collars also have a vibration and/or noise - a beep - which serves as a warning of impending shock. As you can imagine, the shock falls somewhere between unpleasant to extremely painful.
The problem is, the dog doesn’t know what caused the discomfort/pain. He has to figure out how to avoid the shock. He can panic. All his senses are focused on avoiding the shock rather than on his handler. Panic is not a good state for learning nor is it very good for establishing a relationship.
Now let’s look at positive, reward-based training. It operates on the paycheck system at first. You know - you go to work, you bring home a paycheck. And that pay reflects the difficulty of the job. The higher the difficulty, the more the pay is. It is not bribery, as some erroneously claim. A bribe is given in advance of an action - ask any politician. A reward is given after the action, for a job well done.
In positive training, we use praise along with rewards for all dogs. At first, the praise doesn’t mean much to the dog. The value will come as the dog is rewarded because he comes to understand that praise and rewards go together. At first he wants to please to earn another treat, but as time goes on, pleasing his owner is motivation enough for him to do as they ask. And the relationship builds.
In order to learn a new behavior, one must practice. Ask any musician! If a dog repeats an action over and over, soon the action becomes automatic. If, every time you come to a stop on a walk you ask your dog for a sit, eventually, your dog will sit automatically when you stop - the coveted “auto-sit” when walking with your dog. When the dog wants your attention inside the house, and EVERY time, you ask for a sit, you'll get an “auto-sit” instead of a forceful nudge. And we get a dog who, after escaping and wandering a block away, eagerly runs to return to face her owner when called. I know. That happened to me. My dog Dora was trained with positive reinforcement, not coercion. And when the moment of truth came, it was pitch dark. I couldn’t see her anywhere. But when I called, “Dora, come!” she came, running from around the corner, a half block away, straight to me. And I thanked God for every moment we had spent practicing her recall and that it was done positively, so she wanted to do it.
The fact is, positive, reward-based training results in not only teaching your dog a set of behaviors, but also in building
your relationship with him. Coercion training does not build relationships. And I don’t know about you, but I don’t want my
dog to fear me.
I know that many strong, confident professional dogs have been effectively trained using coercion. I even recently heard an e-collar trainer who claimed to have a 100% success rate - sure, if you don’t count the ones who wash out. And of course some do wash out, because dogs, like children, have varying types of temperaments. Some are bold, brave and strong. Some are sensitive and fearful. Obviously, most of the very bold, brave and strong dogs can withstand coercion training. They are the Marines of the dog world. But for the sensitive dog, coercion training can leave them confused, full of fear, and not trusting anything or anyone.
I am currently the owner of a rescue - a Treeing Walker Coonhound who came with the name The Dude. As cool and with-it as his name sounds, when I got him he was afraid of almost everything. When the wind blew in the trees, he bolted. Outside, when he heard neighbor’s voices, he bolted. I ripped aluminum foil off the roll - he bolted. The timer beeped, or my phone dinged, or something clicked - he bolted out the dog door. His noise sensitivity was off the charts, and especially with any "beep" sound.
I recently read a FAQ by an e-collar trainer. He claimed that the e-collar doesn’t hurt or frighten dogs. He said that the e-collar makes a beep when you turn it on, and his dogs come running joyfully when they hear the beep because they want to train.
I’ve had Dude a year now. I believe he had been trained using an e-collar. I’ve used counter-conditioning, desensitization, acclimation, chinese herbs and pharmacology to try to mitigate his noise sensitivity. He is a wonderful dog, and he is much better. He takes most indoor noises in stride now, even coming over to investigate if something falls and makes a noise. But the one noise that can still send him into sheer panic is a beep like that of an e-collar. The other day a power supply lost it’s power and started beeping. In a panic, he flattened himself and frantically tried to dig through the carpet. I couldn’t stop the noise fast enough. It was not joyful.
Training a dog can be frustrating when we’ve done our best and they still don’t get it. Amateur trainers, handlers and owners at that point are hard-pressed not to increase the voltage on the e-collar to try to get the dog to obey. The damage done to the dog’s trust can be lifelong. That’s why I chose positive reinforcement as my training method.
© 2017 Judy Rivard
Paws 4 Fun LLC
(Also known as Nose Magnets or Treat Magnets)
Building Resistance to Extinction
In dog training, when your dog stops offering a behavior, we say the behavior is extinguished. For example, your dog jumps on you to greet you because he wants your attention. If you consistently, every time, don’t touch or talk to him and turn your back to him when he does this, he will stop jumping, because he wants your attention. He does not want you to ignore him. So he stops jumping, and the behavior is extinguished. Behavior that doesn’t result in your dog getting what he wants will go away. Behavior that does result in your dog getting what he wants is repeated (so NEVER give your dog food from the table if you don’t want him to beg at the table. Yep, one time is all it takes.)
Extinction is good when it results in eliminating an undesirable behavior, like jumping, but not good when it eliminates a desirable behavior like sitting or coming when called. In training, we want to build this resistance to extinction into the behaviors we want from our dogs.
Extinction looks like this: if a dog is used to getting rewarded with a treat every time he does a behavior (like sit), there will be an abrupt, behavior-crushing contrast when you stop giving a treat, because reinforcement ceased. The behavior goes into extinction; it stops happening because reinforcement has stopped.
In contrast, if the dog is not used to being rewarded with a treat every time he does what you ask (using a variable schedule of reinforcement,) it will take a lot longer for him to notice that reinforcement has stopped (resistance to extinction.) A common analogy is to compare the use of soft drink machines vs. slot machines. You expect a reward (a drink!) every time you put money into the drink machine. If the machine stops giving drinks, you will not continue to put money in more than once or twice before your money-inserting behavior extinguishes and you try another strategy, such as calling some authority, or body-slamming the machine (which is arguably an extinction burst rather than a strategy. That's a topic for another article.)
Resistance to extinction, however, looks like this: Because you do not expect to get a reward every time, you put your money in a slot machine again and again, hoping that the next time will be the winner. People can actually get addicted to slot machines because human nervous systems are subject to the laws of learning, and the laws of learning state that a behavior that is on a variable schedule of reinforcement is much more resistant to extinction. Dogs' nervous systems are also subject to the same laws of learning. This means once a dog can consistently perform a desired behavior, you the trainer have the option of putting it on a variable schedule of reinforcement, if for no other reason than to obtain this resistance to extinction.
Source: Jean Donaldson's "The Culture Clash" (The Academy for Dog Trainers, 2013), 154-155
© 2016 Judy Rivard
Paws 4 Fun LLC
For urgent needs call: (954) 647-1079 We'll be happy to help!
© 2012 Lili Chin
(Or no, you don't have to give your dog a treat for every behavior for the rest of his life)
Coercion is the practice of forcing another party to act in an involuntary manner
by use of intimidation or threats or some other form of pressure or force. It
involves a set of various types of forceful actions that violate the free will of an
individual to induce a desired response. from Wikipedia
A good skill to have when training dogs is the ability to lure your dog with a treat. And believe me, this is a skill. Many try, fail, and then announce, “My dog just won't follow the treat.” That's because this requires using a very desirable treat with precise movement and control, just like fishing. Here are the steps:
You must present the treat in a way that gets your dog’s attention. Grasp the treat between your thumb and forefinger. Using a fast movement, make a show of presenting the treat close to your dog’s nose. Hold it there and wait a second or two until your dog shows interest in it. I think of a treat as having scent hooks coming off of it that hook into your dog’s nose. You must hold it close enough and long enough that the scent can hook her nose. If you pull it away too soon or move it away too quickly, your dog will lose the connection to your treat, like a dropped phone call.
All of the above should happen in three seconds or less. If your dog does not show interest, try using a much more high-value, desirable treat. Make sure you present it quickly (because movement attracts a dog’s eyes) and allow your dog to really notice it and catch the scent.
At this point, you must keep the distance between your dog’s nose and the treat consistent, with the treat not more than 0.5" - 1" from his nose. Move the treat away, slowly at first, in the direction you would like to lead him. Think of pulling taffy - you don’t want to pull too quickly or the taffy will snap. If your dog doesn’t immediately follow, wave the treat in a small circle in front of his nose, then again pull it away slowly. As your dog follows the treat, keep it close to his nose, but moving in the direction you would like him to go. Practice this until you can, with a treat, lead your dog in circles, in a straight line, around a curve, etc. Now you are ready to use luring in your training.
© 2015 Judy Rivard
Paws 4 Fun LLC
Dogs already “speak” a language; it is Dog Body Language (DBL). Most dogs are very skilled at reading Dog Body Language and using DBL to communicate with other dogs. It is their first language, the language they learned from their mother and littermates.
We want them to understand English. But you must remember that no matter how skilled they become at understanding English, Dog Body Language is their mother tongue and their most fluent language.
In my class I teach both languages. When your hands are busy, if your dog understands English, you can still communicate with your dog. When you become skilled at DBL and are in an extremely noisy environment, or at a distance from your dog, you can communicate with her through hand signals. When she gets old and gray and a bit hard of hearing, you will still be able to communicate clearly with her.
I teach hand signals first. This goes along with a dog’s mother tongue; reading hand signals comes very naturally to them. But imagine a 3 year old boy who has just learned how to speak English with a Southern accent. How good is he at understanding someone with a thick Spanish accent? It would be very difficult for him, because the words don't sound like the ones he knows.
Just like the child above, our dogs, though good at reading body language, need precise, consistent signals at first in order to understand what we are saying. Your signals must be given correctly, and exactly the same way each time. After thousands of repetitions, you can get sloppy and your dog will still understand you. But in the beginning, you must be precise every time.
I use treats to train. I like to get a paycheck when I work - don't you? My “paychecks” for my dogs vary with the difficulty of the task I am asking of them. For beginning work (sit, down, stay, watch, touch) in my house, I can use Charley Bears (nickel-sized crunchy treats). But if I take it on the road and am asking for the same behaviors outside, I'd up the ante. Asking your dog to ignore very enticing distractions like those found outside requires a bigger paycheck. Hot dog slices and string cheese slices are examples of what I’d use outside. If I’m going to shop at Petsmart or take my dog to dog class, I’d definitely use hot dogs or something equally valuable. If the task is VERY difficult, I might use freeze-dried liver, or pieces of real chicken or beef. And remember, even “sit” can be EXTREMELY difficult if your dog is surrounded by other dogs that she wants to play with.
By the way, this in no way implies that I regularly give my dogs people food, or encourage them to beg. If you NEVER give your dog food off of your plate, he will not know that could be an option. You can reinforce no begging by training a “place” command, or by crating your dog while you eat.
Dog trainers very often find that they can take a dog that is not paying any attention to her owner and get the dog to sit, down, stay, walk on leash, etc. right away. Why is this? Because trainers are much more skilled at communicating with dogs. This means that their signals are precise, their praise is effusive, and their treats are delivered much more quickly after the behavior occurs. When your trainer takes your dog and works with him, pay close attention to all those things - the exact signal, the high-voiced happy praise, and how quickly trainers pay your dog for good work.
“Effusive praise.” The dictionary defines effusive as “unrestrained” and “heartfelt.” Is your praise unrestrained and heartfelt? I've heard it said that if you don't feel silly, you are not doing it right. Your praise has to be enthusiastic. If you can’t do that, then perhaps now is not the right time to train your dog. Don’t train when you are grumpy, stressed, angry or depressed. When your dog gets it right, you need to be enthusiastic. It has to feel like what just happened is the best thing EVER! Your dog will love you for it.
And remember, your dog's nose is his most important connection with the world. The olfactory center in their brain is proportionally 40 times larger than ours. Dogs have up to 300 million olfactory receptors in their noses. We have only 5 million. Their sense of smell is 10,000 to 100,000 better than ours. So it makes sense to use their sense of smell as a training aid. Learn how to lure your dog with a valuable treat. That is most often the first step in teaching your dog a new behavior.
For more information on To read about a scientific study
dogs’ fascinating noses, using hand signals versus verbal,
© 2015 Judy Rivard
Paws 4 Fun LLC