Reward based training is a method of training that encourages good ‘wanted’ behaviour by rewarding a dog when they get something right. A reward can be anything from a tasty treat and verbal praise, to a game, to fuss and cuddles, to a walk, depending on what the dog enjoys the most.
We use positive reinforcement to motivate a dog. ‘Reinforcement’ of a behaviour will increase the likelihood of the behaviour recurring. The term ‘positive’ is used in the mathematical sense and simply means that we are adding something following the behaviour; in the training class, this will be a treat.
Where possible, we’ll be using the lure-reward method of training. Lure-reward training is a highly effective way to teach your dog new behaviours. It is easy and fun for both dog and human, and dogs can learn new behaviours quite quickly with this training. This type of training simply involves using a food reward to guide the dog into the desired position or behaviour.
The luring motion can potentially become the hand signal for the behaviour.
We use classical conditioning to form positive associations with lots of different things; people, other dogs, training, commands.
TYPES OF REWARDS
Generally, in a classroom environment, food is used as a motivator, particularly for puppies, as it’s quick and easy to deliver. A food reward can be given less frequently over time, but should never be completely phased out. Some dogs prefer a toy or game as a reward and some will be happy with lots of praise. Praise should be given in conjunction with any other type of reward.
If using food as a reward, be sure to factor this into the dog’s daily food intake. For puppies the rewards should be very small, no bigger than a pea cut into two or even four pieces. The treat should ideally by dry/dried as greasy food can become messy and difficult to deliver. I would recommend home made treats, such as dried liver, chicken, cheese (although this can be difficult to chop up very small and will stick together when it gets warm, but dogs tend to love it).
When training outside of the classroom, consider other methods of reward; a game of tuggy or fetch for example.
Be sure to give the dog lots of verbal praise in addition to any other reward and eventually the verbal praise will become as valuable as any other reward.
The value of a reward is determined by the dog, not the trainer. High value rewards (something that the dog really likes or enjoys) should be used as motivators when the dog finds what you’re asking him to do difficult, such as when he’s learning a new behaviour, or when there are other things going on that the dog finds distracting. Less desirable rewards, like the dog’s kibble, can be used if the dog already knows how to do what you’re asking of him.
I’d recommend a treat pouch that can either be attached to a belt or can be fastened around your waist so that you are ready and able to reward the dog as soon as the dog has performed the desired behaviour.
We use a ‘event marker’ to tell the dog exactly when they got it right, for example if you were training your dog to sit, the exact time that he got it right would be when his bottom hits the floor, not while his bottom was hovering, and not while he was in the process of getting up.
The event marker is usually a short noise, such as a click, or a word. Clicker training is extremely powerful, but a novice can find it difficult to get the timing right and to juggle a clicker, a treat and a lead. Initially I prefer a word. It needs to be a short word, to be able to pin point the exact time that the behaviour was correctly performed, but not a word or phrase that the dog would normally hear in every day life, such as ‘good boy’ or ‘good girl’. I use the word ‘yes’, but I pronounce it in a way I wouldn’t normally do on a day to day basis – ‘yesss!’. A more experienced owner can use a button clicker (as opposed to a box clicker which is more difficult to handle), available from pet stores.
With either method of marking, the word or the clicker, you’ll first need to associate the marker with a treat. I call this ‘charging up’ the event marker; adding value to it.
Initially the ‘reward’ will be food. When you’re teaching something new it’s a good idea to use food of high value to the dog, such as very small pieces of chicken, sausage or cheese (don’t forget to adjust his food portions accordingly) . When perfecting and practicing you can afford to use food that’s less valuable, such as his kibble.
Practice this when the dog is calm. If the dog is jumping or excited stop and try again later.
WORD MARKER METHOD
- Say ‘yesss!’ to your dog (note that you are not asking your dog to do anything to get the treat)
- Repeat this exercise until the dog looks at you reflexively for his reward every time he hears the word ‘yesss!’
It’s extremely important that the treat follows the ‘yesss!’, and is not either before or at the same time.
This exercise needs to be repeated over time many, many times until your dog looks at you expecting a treat when he hears ‘yesss!’.
CLICKER MARKER METHOD
- Click once with your clicker
- Repeat this exercise until the dog looks at you reflexively for his reward every time he hears the click
It’s extremely important that the treat follows the click, and is not either before or at the same time.
This exercise needs to be repeated over time many, many times until your dog looks at you expecting a treat when he hears the click.
Your reward marker will ultimately have two functions; it will be the predictor of a reward as you will have ‘charged up’ the reward marker, and as the dog is exposed to more and more training, he’ll learn that the reward marker tells him exactly what earned him the reward.
The timing of the marker is extremely important. If you mark the dog when he’s in the process of getting up from a sit, rather than when his bottom hits the floor, you’ve given him the wrong information. Having said that, that was your mistake and not the dog’s; if you mark the dog incorrectly, you must still reward the dog as it’s important that the mark is seen as the predictor of a reward, which is what will motivate the dog. Timing is extremely difficult to get right, but it will come with practice.
Your dog will perform many behaviours you want from him naturally, for example your dog will sit, lie down, toilet without being asked. You can use this to your advantage and ‘capture’ this when it happens by marking (‘yesss!’ followed by a reward) as soon as the dog has done something that you like. Remember, reinforcement makes the behaviour more likely to recur.
HAND SIGNALS V VERBAL COMMANDS
Dogs are visually oriented animals and will understand hand signals very well. They can read your body language more easily than understand your spoken words. In one experiment hand signals had a 99% success rate versus an 82% success rate for verbal commands.
Generally we don’t put a verbal command to a behaviour unless we know that the dog is going to perform the behaviour. There’s no point in saying ‘sit, sit, sit, SIT!’ to your dog when your dog doesn’t know what the word means. Not only does it devalue the word, but you will become frustrated which means that neither you or the dog is enjoying the training.
Typically these are the steps that we’ll use to teach a behaviour, which you’ll learn during the training course:
- Lure the dog using a treat
- Phase out the treat lure
- Reinforce the hand signal
- Add a verbal command
In this case, for most behaviours you’ll have both a hand signal and a verbal command which are interchangeable.