Using rewards and praise effectively when it comes to dog training
At the core of contemporary dog training is a strategy called Positive Reinforcement. It’s simply the law of cause and effect. Or, when the consequence of a behaviour leads to a satisfying outcome, the dog is more likely to repeat that behaviour.
Positive reinforcement does not mean providing praise and rewards indiscriminately. In fact the more a dog is rewarded and praised, just for the act of breathing, the less you’re going to affect anything he or she does.
Dog’s don’t require compliments to build their confidence. What they do require to build confidence is clear communication from you, using a method of communication they can understand, and a reward only after the correct behaviour is achieved, so they know what they’ve done correctly.
If dogs could speak
Pet Parent: “Good boy, I love you”
Dog: "Yes, thank you, so…what was that for, listening to my name five times and the cue to sit five times? I got so tired I sat down, then I got up to sniff out a nearby bug. So did you praise me for sitting down, getting up, or getting the bug?"
Pet Parent: “Hector…focus, focus Hector. Good Boy Hector. Hector…watch me, watch me, Hector!”
Dog: “Sorry, what were you going on about? I watched for a few seconds, then a bird flew by that I couldn’t resist. You praised me for something but what was it that I did that you liked? It’s impossible to understand with all those other sounds, and frankly, it’s just not interesting enough to try.”
Communication and timing versus indiscriminate praise
A big part of positive reinforcement is a lot less about praise and a lot more about rewards provided at exactly the right moment. Not before and not after you’ve spent 10 seconds trying to fish a treat out of a baggie not yet opened, from your pocket.
It’s simple: Cue-mark-reward-release, in half-second increments. That’s the most effective sequence of communication, for both you and your dog to follow, every time, when it comes to training. Get yourself prepared first, before asking a dog to do something. Have the hand with the rewards ready, held behind your back, and make sure that bag of rewards is hidden also. Don’t have two hands out front when providing a hand cue. It’s too confusing for a dog. The hand providing the cue is the only one that should be out front. Use a pouch versus a baggie. With a pouch, rewards are out of sight but readily accessible to dole out at the moment the dog’s in position. Once the dog is in position, count to five, to yourself, then mark the dog’s success, and reward him. Building in that little pause ensures the dog is not immediately popping up before being released.
1) Ask the dog to sit, just “SIT”, don’t load in a lot of extra words for the dog to try to figure out.
2) When she sits, count to five, then mark that behaviour with a clear, upbeat word, like “YES”. Use the same clear, upbeat tone every time to mark that behaviour, or consider using a clicker to mark the desired behaviour.
3) Immediately follow by providing a food reward to the dog. Only provide the reward if the dog is still in position, i.e. still sitting, if you asked her to sit, not if she gets up. If she gets up, withhold the reward until she sits back down again.
This is true communication with your dog, it’s behavioural communication on your part and it’s what you need to learn so your dog can understand what you want.
4) Have plenty of Cheerio-sized food rewards on hand and easily accessible, within a half-second.
5) Finish with that five-second pause and a very specific release cue, like “GOOD JOB”. Don’t skip the release cue. It’s just as important as anything else. It should also be clear, upbeat, and used every single time your dog has finished a task. A dog needs to know when the work is done. Without a release cue, the dog decides when it’s time to get up and leave. With a well trained release cue, they wait until you “release” them to get up and leave.
Dogs understand best when you use the same one-syllable words, as cues. They understand that a specific sound relates to a specific behaviour. They don’t understand the meaning of the actual words.
Using a dog’s name correctly
“Stay, Hector…Hector, stay…No….I said “STAY” Hector.”
Don’t ask your dog to SIT and STAY, then say their name. You’ve just asked them to sit and stay...and to come to you, at the same time. How can they do both at the same time? When a dog hears their name, they think you’re calling them. The more you use their name, the less value it has for when you do actually want to call them. Reserve the use of their name for recalls only.
Hierarchy of rewards
Now let’s look at the hierarchy to that satisfying outcome I mentioned earlier. As much as we might wish for our dogs to perform for our love and attention, let’s get serious, is the “GOOD BOY” sentiment going to hold the same value as a yummy morsel of chicken or cheese?
In order to increase production at your job, which is more likely to compel you to work a little harder? A pat on the back from the boss, or a big cash bonus? Dog’s aren’t saints. They’re guiltless and self-serving just like any other living creature, i.e. the survival instinct. The reward for additional work must appear as “worth the effort” to them.
Food rewards provide the highest motivation and not just any food. Proteins make the top of the list: plain cooked chicken, cheese, chicken wiener, beef, etc. Food holds a very innate biological value with them. Rewards don’t have to be big. Cheerio-sized bits of food work best and will allow you to work longer with your dog, before filling him up. Don’t expect him to work very hard, for a piece of kibble he gets at every meal.
The higher value the reward, the faster your dog learns an associate behaviour. The higher his satisfaction level, the more motivated he will be to repeat the behaviour.
Toys can be fairly motivating, however, in my experience, toys only work if the dog has been conditioned as a young puppy to respond to toys. This means a lot of whipping them up to a high state of excitement as a puppy, and we are usually trying to teach them to “tone the excitement down” as puppies due to excess nipping, scratching, and jumping.
Praise really makes the bottom of the list, mostly because people provide it so often and so indiscriminately that it loses most of its value. As a vocal species, our first impulse is to attempt to modify their behaviour with our vocal praise. Whether this form of communication holds much value is still in question since dogs don’t seem to derive much pleasure even from vocalizations of other dogs. If you really must include praise in training, use it as the release cue. When the dog performs correctly, the behaviour is marked and rewarded, then released with the praise word and a pat, and allowed to move away on his own.
Petting for bonding and engagement
Petting is actually a stronger reinforcer than praise. A dogs’ heart rate and blood pressure decrease when you pet him (assuming he’s used to being petted to a high degree and enjoys it), endorphins, prolactin, oxytocin, and dopamine increase which has a calming effect and which plays a large part in the bonding process between you and your dog.
Consider using various levels of reward. For instance, a piece of kibble or carrot may work in the house but step up the value outside where there is going to be a lot more distraction. Mix it up, dogs will eventually tire of any reward if there is no variety. Dogs love variety in food just as we do. Use different rewards in different places, taking the level of distraction he or she will encounter.
Fading of rewards
The end goal, of course, is to remove most of the reward aspect to cues. You use rewards strategically, and often, at the beginning when your dog is learning something new. That keeps them motivated to work with you. Once they understand the command and concept, it’s time to strategically fade off using rewards.
• Start asking for two or three cues before providing a reward
• Increase the duration of a cue before providing a reward
• When out walking, use some rewards but over time, work down to only a few throughout the walk.
• Ask for a sit then reward, then ask for a stay and provide just praise, then perhaps ask for a down and sit, for a reward
• Make rewarding unpredictable. Less predictability means more anticipation on the dog’s part; more motivation for him to work with you
When it comes to training, words are tools like anything else so use them with precision and for specific purposes.
THE DOG BLOG is a great resource for tips on training and canine well-being. Got questions or a topic in mind? Let me know via the comments box. I would love to hear from you.