In other words, dogs do what works to get what they want, i.e., attention, petting, food, toys, etc. This powerful principle is a key component of reward-based training. If you praise and/or treat your dog for sitting, he is more likely to sit the next time you ask. If he jumps up on you and you look at him, talk to him, pet him, etc. - which the dog perceives as attention, negative or not - he will be more likely to keep jumping on you.
2. Rewarding Behavior.
When training, it is important that the reward be given during or immediately after the desired behavior is displayed. For example, when teaching your dog to sit, the praise and/or treat should be given as soon as his rear touches the ground, not after he's stood up again. Dogs live in the moment, so you have approximately 0.5 seconds to reward or punish a behavior. Reprimanding your dog for something he may have done hours or even minutes ago, e.g., you come home to find your favorite shoes in pieces, is pointless. Your dog won't associate your yelling with what he's done and, if it happens often enough, he may begin to fear your arrival home as you are always angry for no reason he can fathom.
3. Focus on what you want your dog to do.
Most of us are so busy chasing the dog around the house trying to get him to stop doing "bad" things that we forget to tell him when he gets it right. For example, your dog is bored and trying to get your attention, he barks at you. You turn and look at him and say "ssshhh" and your dog lies down for a minute before he does it again. In your dog's mind, you were ignoring him so he barked at you to make you aware of your mistake. Sure, a barking dog is hard not to notice. But what about when he's lying calmly? Most of us never consider rewarding calm behavior, so the dog only gets rewarded with our attention (even yelling is attention) when he barks. Having been rewarded, of course he keeps doing those things! Make a habit of noticing and rewarding your dog for good behavior.
4. Your dog gets to choose.
Just because you think those expensive new treats are a great reward doesn't mean they are. If your dog turns his nose up at them, they're not much of a reward in his mind. A reward can be petting, verbal praise, a throw of the ball, a quick game with a favorite toy, sniffing grass, saying hello to another dog, etc. - the sky's the limit. Consider what your dog finds rewarding and use it to your advantage!
The jackpot is something really special, head and shoulders above the usual reward. Your dog can earn this amazing prize by doing something especially wonderful. While it is always important to use training treats your dog likes, save the Super-Yummy, Best-Treat-In-The-World as the jackpot. For example, a dog knows what 'sit' means, but doesn't sit very quickly. When you give the sit cue, he watches you for a moment, then languidly lowers his butt to the floor. You can almost hear him sigh, "Okay, if I must". But on the forth repetition, he responds immediately: his butt hits floor in record time. Jackpot! You immediately give him one piece after another of the special treat, along with effusive praise (and petting, if he enjoys it). You can also give a megajackpot by tossing a shower of treats. Jackpot makes an impression - it calls the dog's attention to the fact that he's done something wonderful. He is therefore more likely to perform the behavior better than usual the next time. A jackpot doesn't have to be food, either. If your dog lives for a toss of the ball or a game of tug, use that as your jackpot. Know your dog and use what works for him.
6. Find an alternate behavior.
When you want your dog to stop doing something, give him something else to do that is incompatible with the behavior you don't want. For example: if your dog jumps on you, have him sit instead; he can't sit and jump at the same time. If he chews on furniture, give him an appropriate chew toy instead. Try this: On a piece of paper, draw a vertical line down the center. On the left, list all of the things your dog does that you'd like him to stop doing. On the right, next to each behavior, write down something he could do instead. I've listed a few to get you started.
Behaviors I Don't Like
Behaviors I Would Prefer
1. Jumping up on me and visitors
Sit when greeting visitors
2. Begging for food at the table
Lying down (not staring)
3. Digging in the yard
Playing with toys/being calm
Carrying a toy around/being quiet
7. Raise criteria gradually in small increments, building on each success.
Simply put, that means don't expect too much too soon. Instead, build small steps to get from Point A to Point B. For example, when teaching your dog to down-stay, start with a three-second down-stay. When that is successful, add two seconds, and so forth. Any time your dog does not perform an exercise correctly, consider whether you have proceeded too quickly. Go back to the point at which your dog was last successful, then build gradually. Setting your dog up to succeed eliminates the need for corrections.
8. If trained correctly, the behavior is not contingent on food being present.
This is something that many people who are opposed to food-reward training don't understand. If you phase treats out gradually and use lots of real-life rewards - petting, games, etc. - as well, your dog will do as asked even when no treats are present. Use a lot of treats at first to teach and then practice new behaviors. Eventually, rewards should become fewer and farther between - but they should not stop altogether. You wouldn't want to stop getting paid once you got better at your job, so don't forget to reward at times for a job well done!
9. Have Fun!
Keep training session short - 3 to5 minutes a few times daily is fine. Focus on one behavior in each session. Keep an upbeat attitude when training. Don't train when you're cranky! End each training session on a successful note. Did your dog do seven good sits, with the last one being really great? End the session there. Once a new behavior has been learned, incorporate it into your daily routine.