Almost every day I talk to someone who is having a training problem with a dog. In every case, there is an underlying misunderstanding that explains it away. It happens to me, too. I'll give you a great example.
Antecedents: I took my puppy to work everyday (yes, my boss is amazing). At lunch time I would tuck him into his crate under my desk. One day, I reached for him and he skittered away. I laughed. Cute, right? So I cornered him and tucked him in and went to lunch. The next day, when I reached for him and then pursued him around the cubicle for a few minutes, my office mate (also a dog person) howled with laughter. She told me she had watched me train my puppy to avoid getting picked up for days. What??
Think about it. Let's assume the puppy has a good life, likes me and likes coming to work with me. Why would he run from me? Because the only time I pick him up is to isolate him ---something a happy puppy will work to avoid by running away. Training 101: if one activity is followed by an undesirable happening several times, a dog will recognize and avoid that activity in the future. Oh boy.
Consequences: The second mistake trainers make is deciding on a reward without asking the dog if the reward is worth working for. It doesn't have to be food, either. If you want your dog to do A, and he doesn't do it, look at what happens immediately following A. No matter what you think, I can assure you, whatever it is, your dog doesn't like it enough to work for it.
This is prevalent when people assume that dogs like to be patted on the head. Handlers who use patting as a reward often end up with poorly trained dogs because most dogs don't like it. I'm not talking about the wonderful ear massages or belly rubs your dog already works for, I'm talking about the trainer who thumps his dog on the head and says, "Good Boy" just a little loudly and is shocked that his dog doesn't work hard for more of that.
Timing: Dogs link events to one another the closer they are in time. You can easily teach your dog to get a treat for barking at a jogger by waiting one second longer than necessary to reward him for noticing the jogger. If your dog is doing more of what you need or a little less, try marking the behavior just a little earlier to see if it changes in the right direction. Went the wrong way? Then mark a little later. Don't use a marker? Add one. All it takes is having the reward ready and pairing it with a word you and your dog come to know means, "that's it!" As we are all aware, many trainers use a clicker or a whistle but a word is just as good.
Punishment: Let's assume you've looked at antecedents and consequences and that's not causing your training issues. The number two training mistake we all make is not using all the tools in the learning box. While it has been scientifically proven that the best way to teach a new behavior is without aversives, this does not mean that we should not use all four quadrants of the learning matrix to train dogs. In a recent blog, Tyler Muto, points out that some force free trainers will go so far as to say that euthanasia is better than using positive punishment to solve dog problems. Um, wow.
Let's clarify. Punishment fails to teach when it is misused. Even force free trainers use punishment. They just don't know it. Failing to deliver the treat based on criteria such as a straight sit, is punishing a crooked sit. Fact. Of course, hurting or intimidating a dog is a bad idea for your relationship. If you are a bully, your dog will be, too and science backs that up. But if a dangerous behavior leads to something a dog would work to avoid, after he has failed to respond to a cue for an alternate behavior he knows well, lives are saved. And this can always be done without hurting or scaring the dog.All four quadrants are tools for proofing any behavior. Refusing to use them is silly.
Generalize: Another training mistake is failing to generalize a behavior to many situations. Ian Dunbar has an excellent test for this. He puts a $50 on the table and tells trainers it's theirs if their dog will sit on cue. Ha! I'll take that bet. Then he says, "I want you to face the wall, stand on one foot and whisper sit. If your dog does it, you get the $". Oh. Well my dog needs me to face forward, lean back and hold my hand up like I have a biscuit. It's also best if we are indoors and no other dogs are playing nearby. In other words, your dog does not understand "sit" exactly. He's on the way but you need to fade the extraneous cues and work in new places and despite distractions. Generalize.
Training time: The last mistake we make is expecting training to take a long time. We plan an session here or a lesson there without seeing results. We don't realize that a dog learns a new task in a few sessions that are a few minutes long throughout our normal lives together. Your dog may not sit on cue with distractions but he knows the sound of your keys from two floors away. You've trained the keys sound without even dragging out any equipment or putting off any household chores for training time. How did that happen? Asking your dog to sit before meals or opening the door for a romp in the yard would increase your practice opportunities to 10 per day, without changing anything. If he doesn't respond to the cue, you punish him by taking away the thing he wanted. Yes, this means you can't ask him to sit for the door opening when you are in a hurry to let him out because if he fails, you can't open the door. So maybe it will only increase your opportunities to 7 per day. That's still enough to train your dog well.
Interesting research shows that dogs learn as much in three sessions per week as they do in 7 so stop delaying training time for life. It's already happening all around you. Use it and avoid these common training mistakes whenever you can.