Use positive reinforcement when the dog performs the behavior you want.
Positive reinforcement is neither good nor bad. It's the addition of something when a behavior occurs.
Negative reinforcement is the removal of something when a behavior occurs. Think of it this way: If you were a teenager, would you be more apt to take out the trash if it meant your parents quit nagging you, or if it meant you could play video games?
This can work for or against you. For example, if your dog is barking and you give him a treat to get him to be quiet, the dog will think you like the barking and will continue.
Training needs to be broken down into small parts that the dog can understand.
Let's say you want to teach your dog to sit. Before mealtime, get the dog in a quiet room where he won't be distracted. Show the dog a treat. Let him sniff it a little, and then move the treat up and back: he should sit back so he can reach for it. As soon as he does this give him the treat and pet him, saying something like "Good boy!" This is doing a couple things:
The dog is associating the action of sitting with praise.
The dog is associating weaker reinforcement of petting and verbal praise with the stronger reinforcement of the treat. Since these methods aren't dependent on hunger, it will be easier to reinforce correct actions.
Now that you have the correct action, repeat a few times, lessening the reinforcement.
You may have trouble getting the dog to sit. There are a couple bad things you could do here:
You might try to push on the dog's back to get him into the right position. Instead of learning to sit, the dog will think the word "sit" means he should expect you to push on him.
You could punish the dog. The dog doesn't know what he is supposed to do and will find this confusing. He will associate the word "sit" with doing something bad.
Posted 4468 day ago