Getting Behaviour
Dogs are always behaving. Your job is to convince him to do the behaviours you like by reinforcing the behaviours you like. The problem is, you have to get that behaviour in the first place! Fortunaely, there are several ways to set your dog up to start doing the behaviours you want, although some are much more effective than others.
Luring
Luring would involve having the dog follow a cookie or toy in order to produce the behaviour you want. An example of this would be putting a cookie on his nose to entice him to sit or holding it up to your face to get eye contact. There are definite limitations with luring. The first is, you are putting the reinforcer before the intended behaviour. Let's assume you want to teach your dog to look at your face. Your dog looks away and you place a cookie in front of him and draw his attention back to your face, then click and reward him. Your dog looks away again and you place another cookie in front of him to draw his attention back on you. This common routine doesn't make much sense when you remember that consequences drive behaviour because when your dog looks away from you, you present a Primary Reinforcer (food). You are actually training your dog to look away from you!
The second problem with luring is that the dog is not actively involved in the learning of the behaviour. If you lure the dog into a sit with a cookie, the dog is not thinking about the sit, he is thinking about following the cookie. The sit just happens although there is little learning occuring.
Luring can be very reinforcing for the owner because the dog actually does the desired behaviour very quickly. The problem is, the behaviour is less reliable and the reasons will be discussed a little later on.
Prompting
Prompting is similar to luring although you would use body language to persuade your dog to do the things you want. You might blow in your dog's face to teach him to wave or start running to teach him to come when called. Prompts have the same limitations as lures.
Capturing
Capturing would be waiting for the dog to offer the behaviour himself and Reinforcing it when you get it. If you wanted to teach your dog down, you might sit on the couch with a clicker and cookies and wait for him to lie on his bed and reinforce that. Capturing actively engages the dog in the learning process. When the dog lies down and hears the click, he will engage in other behaviours, trying to make you click again. He might sit, spin, lift a paw, and finally lie down - Click! It may take you a bit longer to get the down this way, but because the dog is actively deciphering what earns him reinforcement, he will easily remember it, unlike a dog who is prompted or lured.
The draw back of capturing is that your dog may not feel like lying down at the moment you want to train. Although capturing promotes better understanding in your dog, it can be very inefficient to have to wait for him to offer desired behaviours. To speed the process up, you can use shaping.
Shaping
Shaping is the process of successive approximations. If you want your dog to lie down and he is not freely offering it, then you can start with a muscle movement that would preceed lying down. To shape a down, you might first click your dog for standing still. Then when you have that, you might look for the next step to a down like a movement of the head toward the ground or even a sit or a bow. You can split the down into very small sums and ask for those one by one, until you get the total behaviour, the down. Another example would be to teach your dog to spin. You would first click for any head movement, then a head movement to the side, then moving one foot to the side, then two feet, then curling his spine, then a half a turn, then finally a full turn. This seems like a lot of work but dogs who are clicker savvy learn to offer a lot of behaviours and become very engaged in the learning process. They will keep moving and trying new things in order to get you to click and once your dog is freely offering behaviours (we call this operant), you can shape even complicated behaviours in just 5 minutes. Unlike luring, because the dog is actively involved in the process, the motivation and reliability for the desire behaviour are built in and the behaviour will be very strong.
Let's make the difference between Luring/Prompting and Capturing/Shaping more clear. Let's assume you are in a strange city and need directions from your hotel to the dog show site. Fortunately, there are dog people staying at your hotel and they say you can follow them to the show site. You climb into your car and follow the bumper in front of you, playing with your radio and singing to your dogs as you drive there. You arrive in just ten minutes and haven't made any wrong turns: the "driving to the show site" behaviour would appear to be strong. Now let's say the next morning you have to return to the show site. You go down to the parking lot and look around and your dog friends are gone. You get in your car and try to find your way back to the show site. It suddenly dawns on you that you really don't know how to get there, even though you did it just yesterday.
Now assume that there were no people going to the show the first day so you bought a map. You got in your car and maybe had to pull over once or twice to get to the show site and it took you fifteen minutes instead of ten. When you pack up the following morning to go back to the show however, you have retained the street names, landmarks and turns because you were actively looking for them the previous day. You get to the show site with almost no hesitation.
Luring and Prompting are the same as following a bumper. You will get the behaviour you want but when you try to repeat it, there is difficulty because you were not actively involved in the process. Capturing and Shaping are like following a map. Map following may be slower than bumper following in the initial steps, but retention is much greater and subsequent efforts are easier and faster.
So what methods do clicker trainers use? To be honest, they use all three. Although shaping and capturing create the best environment for learning, they can be a bit slower in the initial stages. It is OK to use a lure or a prompt to get the ball rolling but if you do so, it is important to begin shaping as soon as possible. An important rule of thumb is this: if you must use prompts or lures, use them only three or four times, then move to shaping. This will get you maximum learning in minimum time!
Now you are well on your way to getting the behaviours you want. The next step is to put the behaviours on cue.
Cues