The new wave of pet animal training focuses on positive or reward-based training only. The idea is to train your dog to perform certain desired behaviors rather than to punish unwanted behavior. For example, you would not train a dog to stop barking by punishing the unwanted behavior; instead you would reward the silence that inevitably follows.
Timing of rewards is critical. If a dog stops barking for three seconds and you have to reach into your pocket for a food treat and then walk toward the dog to deliver it, the moment may have already passed. Yet it is difficult to have primary rewards (food, water, toys) handy at all times, so how can this best be managed?
The answer is using a “secondary re-enforcer,” like praise or a neutral cue that signals that the primary re-enforcer is due. In animal training, whistles and clicks have been used as secondary re-enforcers, though you can also use your voice. Clicks made by small plastic clickers (“frogs”) are probably the best and most consistent way of marking the successful accomplishment of a behavior.
Initially the click is meaningless to the dog but it doesn’t take long for it to realize that a click signals something good. At this stage, a click can be used to reward a desired behavior instantly, accurately, and even from a distance. Engaging in clicker training is fun for the pet owner and the pet. It’s constructive, produces rapid results, and the learning is indelible. Clicker training doesn’t mean you have to spend hours a day teaching yourself and your pet what to do – a few minutes a day is all it takes. Treat yourself: Go out and buy a clicker kit and get started, and throw away that old choke chain.
Step 1. Pair a click with a reward – for nothing at first. Click-treat; click-treat; and so on. By the end of this stage you should:
a) Have your dog’s undivided attention
b) Notice that your dog has learned to associate the new sound with a reward.
Step 2. Begin to click and treat only when the dog has engaged in some desired behavior. Initially, chose a behavior that the dog readily performs, like sitting or looking at you (i.e. go with the natural flow). There are two circumstances in which a click (followed by reward) is indicated:
1). Immediately on completion of a finished behavior (e.g. sitting).
2). When your dog takes a step or two in the right direction of a behavior that you would like to encourage (e.g. if he moves toward that newspaper you want him to pick up), subsequently rewarding progressively closer moves toward your eventual behavioral goal for him. This process is called “shaping.”
Add a verbal cue to signal your willingness to reward a finished behavior. Click only when the dog performs the desired behavior when the verbal cue is given. E.g. click coming to you only after you have cued, “come.”
Note: Vary the time between the click and the treat from immediate to a second or two later. The dog learns from this maneuver that a click signals that a treat is coming even though he might not know precisely where or when. He also learns that if he performs a behavior that you approve of he can make you click… and that means food.
Sit If your dog does not sit quickly enough to allow you to click and reward, you can lure him into sitting. First, put a food treat in your closed hand and position your hand over the dog’s head. As he shows interest in the enclosed treat gradually move your hand back over the dog’s head so that, in the process of following it, the dog will naturally assume the sitting posture. Then click and reward.
It doesn’t have to be a great sit at first – just an excuse for a sit will be fine. You can refine the sit later by progressively raising the bar on what will be clicked and rewarded. Remember that the click signals the successful completion of any phase of the behavior and that a “real” reward is due – but not when that real reward (e.g. food) will be given.
Off If your dog jumps up on you during training don’t react. Don’t do anything until he has “four feet on the floor.” Then click and reward. Repeat this as necessary.
Down Similar to training sit but with a different trajectory for the food lure.
a) Show the dog your clenched fist containing the food treat.
b) Slowly lower your fist toward his chest, between the elbows. The dog’s head will follow the treat so that he assumes a “hunkered over” posture.
c) Next, move your fist (still clenched) slowly away from the dog so that he slides forward… and down.
Don’t worry if it doesn’t go fully according to plan at first attempt. Remember to start by rewarding even approximations toward the desired behavior, “shaping” the complete behavior in stages.
Long Sit/Down Once your dog has learned to sit or lie down to make you click, you can start shaping the behavior toward longer durations of these behaviors. To accomplish this, do not click right away but rather delay the click and reward by a few seconds. The length of delay can be increased clicking only once to signal the end of the required behavior. The dog will learn that if he sits or lies down for long enough a click and treat will eventually come.
Fading the Lure Of course, you don’t have to keep food in your clenched hand and wave it around forever in order to get your dog to perform. Once a behavior is occurring with the appropriate hand movements, simply stop using them and wait for the response to occur without direction and of the dog’s own volition.
Come Start by sitting on the floor or crouching down and calling your dog to you. Look enthusiastic and pat or scratch the floor in front of you. “Buddy, come here, good boy.” If Buddy comes, click – reward and move to another location. Repeat this exercise many times. If you have more than one person you can practice this recall between 2 or 3 people. Each one calls the dog in turn and whenever he responds to the call successfully – click and treat.
Dog owners will probably want to take this show on the road. First – try the same exercise in a quiet yard, with a long leash attached. But never jerk the leash or haul the dog to you. The leash is just to keep him within a certain radius of you and limit his area of interest. If all successful “comes” are clicked and well rewarded the behavior will become almost automatic.