Clicker Method Dog Training: Your Opinion-- Does it Work?

I signed Polaris, our 14-week-old puppy up for Puppy Kindergarten last night. The school uses the clicker method, or operant conditioning.

Basically, good behavior is rewarded with a click and a treat and bad behavior is somewhat ignored. As an example, if a dog is jumping, you turn your back on it and ignore the dog until all feet are on the ground. The moment that happens, you click and treat. After a while, you stop rewarding the feet on the floor and only reward when the dog sits. The idea is that the dog will keep trying different things to get the treat.

My older dog was trained with the traditional punish/reward method. If you wanted her to sit, you said the command, and if she didn’t, you popped her choke collar gently until she did, and then rewarded.

The new trainer says that no punishments will be used. You say the command and then wait until the dog obeys. If he doesn’t obey, the only punishment is not getting the treat.

My husband is somewhat skeptical. When I explained the method after coming home last night, he grinned and said, Cartman style, that it sounded “like a bunch of tree-hugging hippie crap” to him. He prefers the traditional punish/reward style, saying that it seems to him that it would work quicker than simply rewarding selected behaviors.

Have any of you trained solely with the clicker method? What were the results? Did you find it easier or more difficult than other methods?
(On a side note to those of you who have given me advice before-- Polaris’ behavior problems have lessened considerably. She no longer chews on me, and she’s accepting her subordinate role much better.)

I was pretty much in your husband’s camp until I read Don’t Shoot the Dog by Karen Pryor. Please urge him to do so. Clicker training is the real deal. Another great book.

You can find them used for under ten dollars. It will be money well spent.

I haven’t clicker trained a dog but I have clicker trained a horse. it is a very interesting technique. I think the one thing it does most effectively is “train the trainer” especially on the incredibly important need to be cosistent. In other words, it teaches the handler to really observe the animal and to react immediately when a desired behavior is happening. I have seen this translate into a real improvement in trainer/handler effectiveness in all areas of handling, not just while training using the clicker method.

It’s what I train SAR, detection and service dogs with.

The method works - it produces animals that think for themselves rather than animals who only “obey”. Get your paws on some books by Karen Pryor (like Don’t Shoot the Dog) and Culture Clash (can’t remember the author now).

Not “Hippie crap” but a method that works very well, though it is frustrating at first. Shaping behaviors takes a while to get used to. We find it produces dogs that are way more confident and “problem-solving dynamos” than those trained using the J&P (jerk and pop) methods.

Good luck with Polaris!

(Ahem. Post #2)

I’ve never been completely sold on the idea, though I’ll admit to not reading the books mentioned so far.

I’ve had some conversations with my breader and read a lot of work from others who don’t believe in this method and my problem with it is that without corrections I don’t see where the dog’s motivation to obey comes from.

Take the stay command for example. You tell a clicker trained dog to stay and he does because he wants your reward [the treat in your hand, or praise]. But what if a squirel runs by? What if he’s tempted with something that he perceives to have greater value than your reward? Why should he stay? With a traditionally trained dog he’s going to stay becuase it’s going to hurt if he doesn’t.

I could see how the role the dog is in could make a difference. Maybe clicker training produces better dogs for SAR and detection work. But for something like a regular patrol dog where you need him to do what you say and do it now, it’s going to be very difficult to clicker train that.

My feeling is that for a house you don’t need a good problem solving dog, you need a dog that’s going to stop jumping up on visitors when you tell him to.

I also think Hello Again hit the nail on the head. Being a good, smart, observant trainer is going to be key no matter what method you use.

Well, I just finished a session with Polaris using this method, and I’m pretty impressed by the results.

We were playing fetch. Over the last few weeks, playing fetch has been less fun than it should, because I had to pry the toy from her mouth every time.

Today, I used the clicker and a pile of small treats. She retrieved the ball, and I clicked. Maybe it was the noise which made her do it, but she dropped the ball, so I clicked again, and gave her a treat and praise. When she brought it back the second time, she looked at me hesitantly and dropped the ball on the floor. BINGO! I clicked, treated and praised.

We played for the next fifteen minutes this way, with her dropping the ball every time in anticipation of a praise/treat.

I’ll have to see if she repeats this behavior tonight when I play fetch with her again to see if it really sank in. We then ran through “sit” and “down” a few times, which she did very well with, and laying a treat on the floor in front of her which she cannot have until I say. She’s still a little sticky on that one, but she seemed to understand that the click meant she had done what she was supposed to do.

Maybe this “tree-hugging hippie crap” might work well for her! :smiley: Luckily, she’s a smart little thing, so that might make it easier.

Servo - the method for clicker training is, initially, one where we try to shape existing behaviors. For example, we will wait for a dog to offer a specific behavior, click, reward the behavior… without a word being said or the dog manipulated. For example, with a puppy, it’s easy to shape a sit. You’ll have an excited puppy who will sit for you, hear a click, get a piece of kibble… and his brain goes… Wait a minute, what did I do, and how can I do it again? After, we associate a word/hand signal to the command. This approach, in my training field, creates “thinking dogs”.

I am just working on the sit/stay with one of my trainees right now. He had a solid sit, so here’s how a first training session could go:

  1. Make pup sit, click, treat, good dog. (the click, by the way, basically is a faster way to say “gooddog”. )
  2. Wait for pup to sit again (hey, it just worked, right)… make eye contact… wait a few seconds, click, treat.
  3. lather, rince, repeat, but make the interval longer. Then, teach the word “Stay”.

When the dog doesn’t do what you were looking for (offers another behavior), I usually use “Wrong, try again!” or “sucks to be you!”

To give you an idea of how powerful it can be for training: my toller is a high-speed agility boy. He used to clip some jumps short in jump series (triples especially.) Using a clicker, my trainer would make us do the jump like we would in competition, and if he tucked his legs up high, he’d get a click at the highest point of the jump. Then, reward. (BTW - dogs learn to wait for their reward… and, even, later, we move away from food completely!) He eventually figured out that what was making us “click” was the leg tucking, and started doing it consistantly. Solved my agility problem in a week.

My current pup just learned to shake paws and lie down with the obedience “long distance down” (hand up) command, in one session. It’s now a reliable behavior. How did we do it? I sat on a chair and stared at him. He tried everything he knew. Got frustrated, whapped a paw on my knee. Click, treat. Then he just sat. Thought. Whapped my knee. Click, treat. Didn’t have to think about it… whapped my knee. Click, treat. Then I started putting my hand where my knee was, if he missed, no click, no treat. when he hit square on, click, treat. 15 minutes later, I have a Shake-a-Paw that is pretty damned reliable. Not bad! :wink:

So. All in all… it works. It works with dumb dogs, it works with smart dogs… and is a LOT of fun if you have a dog who loves problem solving. I have taught many service pups to do the laundry using behavior shaping… no way in hell I could have easily done that with the old J&P stuff!