Clicker training - opinions from dog people

Popularized by Karen Pryor’s book first released in 1984, Don’t Shoot the Dog. It’s been gaining ground since then. I think its popularity can be pretty regional.

I got a basenji puppy from a breeder in the mid 90’s, and she taped a clicker to the top of the crate. The only dog training school in my town was very old-school (using aversives like collar jerks, banning treats) so I decided to teach myself.

I would just say Mini-Aussie when you’re in a situation where you’re not sure that someone is interested in the long version (or you’re not interested in explaining). Go with Mini-American when you don’t mind answering follow up questions, or if you’re speaking with someone who is likely to be more knowledgeable about dog breeds.

ETA: On the clicker training, use it if it “clicks” for you (couldn’t resist). Some people love it. Some people find it clunky. The real trick to dog training is consistency, timing, and reinforcement.

I don’t know about “clickers”, but my crew learned in about 2 minutes the “Whistle”. I’ve got 2 whistles from marine issue life vests (very loud!) and they know when they hear them, stop, look and listen!

I can get their attention from about an 1/8 of a mile away, and get them to come back to the Jeep, in a hurry!

But they are the mostest, most smartest dogs out there! A good crew, to be sure! Pics here: bob smith | Flickr

What? You cheatin’ on Ted? :wink: There can be only one, you know?

Rather different concepts. Clickers are used to communicate to an animal “You just behaved in a desirable fashion, and therefore you’ll be receiving a reward in the near future.” In contrast, you’re using your whistle as a command. Perfectly fine; do what works. It’s a very different concept from clicker training, though.

Does a click always need to be followed by a food/treat reward?

My previous understanding has been that reinforcers should at some point vary. Sometimes the treat, sometimes “jackpot”, sometimes praise alone, and eventually randomly sometimes no reinforcer in “variable ratio schedule”. Integrating the clicker into that framework would lead me to expect that the click should at some point after initial association is gained also be sometimes followed with varying levels of stronger reinforcement inclusive of praise alone, eventually not always followed with a separate other stronger primary reinforcer, and of course at some point in mastery variably given itself.

Is that understanding correct?
Also - is there in your minds ever a place for negative reinforcers? I remember a trainer arguing against the exclusively positive treat based approach by saying that when some little kid in a playground unpredictably dumps something on your dog you better have a really big treat available in that approach. I was not so convinced at the time but then agin he had rescued many a dog (and trained into great dogs) that was due to put down as they had been given up by other owners for aggressive behaviors.

Now you stepped into it! :smiley:

If you’ve ever glanced at a dog training (or child discipline) thread around here, you’d have no expectation of achieving any consensus on this topic. :wink:

I’m in the camp that favors both positive and negative reinforcement. Has worked for me and I will continue using it in the future up until my last dog. Many other folk disagree.

Cards on table -

My bias is that there is a role for negative reinforcers (even what I think is called “positive punishment” in the operant conditioning framework) as well but that is like with kids: the goal with kids is to reinforce the doing it right many many more times than telling them they are doing something wrong but setting limits with appropriate negative reinforcements (from time-out to loss of privileges to just one of those painful talks) still has to be part of the mix. The problem with kids for many of us parents, especially when the kids are difficult, is that being good about rewarding the correct behavior enough, even identifying what that correct behavior is, and keeping that balance heavily in the positive column is not as easy in real life as it sounds when you just say it.

Now for dogs it is a bit easier to define what the behaviors are. I think the occasional appropriate correction can go a long way.

A large part of my goal with training for this dog is the awareness that this dog will need the constant cognitive work in addition to adequate exercise to be happy. But I eventually also want to be sure that when there is confusion suddenly happening he automatically looks to us for instructions, that we can be sure that he will drop on command no matter how tempting the distraction he wants to chase or herd is, and that he is reliably recallable. Is the 100% positive reinforcer approach (inclusive of the bridging stimulus) able to get there? Does it vary according to breed?

Again I have not had a puppy to train in many years. I am willing to listen to various perspectives and learn.

Of course it isn’t like this is even mostly my decision to make. I only agreed to getting a new dog with my wife’s promise that she would do most of the training with me in a support role only. I am going to puppy class to watch. She is intended to be the prime handler and she is much more in the positive only camp that I am. I’m in charge of making sure he gets enough exercise!

I’d say since your wife is the primary trainer, you do what she wants. In the interests of both marital harmony and not confusing your pup!

I’ve been married to the same woman for a few years more than three decades. I long ago learned the art of compromise … which means, of course, I do things her way as a matter of course! I may not be a weak man but she is a strong woman.

Seriously though, yes, like with parenting, we will stick to speaking with the same voice. But I would like to have my own informed opinions and to be able to express them to she, who is to be the prime trainer, who will decide what that voice will be. And reality is that while she may be in charge of training we will both be home with the dog when the other is not and both working with it, even if my role is troops.

My understanding is that you want to gradually fade out the treat portion of the event, largely because you don’t want to have keep funneling treats every time your dog does something right. Remember that you’re going to be using the treat/click for a range of behaviors. You want it to be available even if you don’t have a treat handy and you don’t want your dog to look like porker.

The actual point of the treat is to help the dog associate the “click” with “YES!” in the dogs mind. Once the dog makes that connection, you might want to start phasing out the treat so that you can use the clicker by itself.

I wouldn’t vary the treat. I would just figure out what motivates your dog the most and use that. I think switching things up leads to confusion. Dogs don’t really enjoy variety for its own sake.

Just to be clear, when using a food treat, I would only use a crumb about the size of a pencil eraser. I used to use frozen hotdogs, that I’d slice in circles and then quarter the circles. I’d put them in a baggie and the baggie in my fanny pack. Now a days I’d probably use something like these Zuke treats, which are small and can be broken in half. I hate getting meat juice on my hands.

So he shouldn’t be getting a cupcake every time he gets a click. It shouldn’t interrupt the training session.

I believe that it’s important that a dog understands the word “No” as well as yes. I find the best way to teach No is to release your Inner Darth Vader and have him tell the dog that he is A Very Bad Boy. The key is to act shocked at the appalling gaucherie of his manners. A touch of Lady Bracknell would not come amiss. After that, I ignore the dog for five minutes or so, and then let him make up to me.

I think it’s mostly not worth doing if you don’t catch the dog in the act. Like a toddler, there’s too much room for confusion if you try to come in later.

I wouldn’t swat the dog, because dogs wouldn’t necessarily make the connection, but I might lay hands on him, depending on the circumstance, to get his attention. I wouldn’t hesitate to grab him by the scruff or put my hand down his throat, if he swallowed something.

The main thing is that you want to be communicating with your dog about your expectations. If you do use negative reinforcement, make sure that you’re getting your point across about his behavior, not about your anger. Negative reinforcement can result in a lot of mixed messages, if you’re not careful.

On the subject of housebreaking: If a puppy is having accidents in the house, it’s the owner’s fault. You need to do a better job of crating him and taking him out. If his schedule is shorter than you’d like - too bad. Take him out as often as he needs and slow down on your attempts to get him to hold it. Patience up front will pay off in the end.

If you catch him in the act, you can scold him with a No!, but the important point is to grab him, hustle him outside and then click or praise him. Don’t spend time scolding him; grab him and get him outside so you can praise him and make a fuss over his success.
Cute little critter you have there! I’m glad you’ve decided to keep his tail!

Variable reinforcers can be very useful. Sticking to super high value rewards loses you some opportunity for nuance. A very food-motivated dog can be TOO focused on getting that treat, and can work themselves up into a very aroused state, which can make learning more difficult.

In that case, a lower value treat, e.g., pieces of kibble, even slices of carrot, can still give the effect of a positive reinforcement, without the arousal of the “cupcake.” And the cupcake then makes a great jackpot, and/or the signaled end of the training session.

Positive only: This is a deep schism in dog training, and not one I’m keen to wade into.

Honestly, OP, I don’t really want to try to convince you there’s one true and correct way to train a dog. Which of course there isn’t. You’ve said your wife will be taking a class – let her instructor be your guide. And, if you want to learn more, read widely – I mentioned Pryor’s classic Don’t Shoot the Dog, but I also enjoyed her Reaching the Animal Mind, which illustrates her understanding of behavioural science with stories of its implementation in lots of different circumstances and species.

The trainer I worked with, who I thought was a great trainer, was a big fan of positive only for MOST dogs. However, she used a shock collar to train her own Shiba Inu to avoid snakes, as she said it would be much quicker and safer than waiting until she and her dog encountered a possibly poisonous snake in the wild and teaching it then. And she also recommended a kind of choke collar for someone in the class with a husky pup to teach them proper leash walking. She said it could be done with positive training and a lot of patience but it would be much faster with the collar. I don’t know if that was negative reinforcement or positive punishment. (The dog’s owner was not to yank the leash ever, just let the dog run to the end of it and the collar would do the work, and the dog would make the connection; I don’t know if it worked for that dog because I only saw him in the class and he was pretty well behaved there.)

I used positive conditioning and the clicker to teach my dog the boundaries of the yard and 99 percent of the time it’s like there’s an invisible fence there. The other 1 percent there is a certain cat across the street and he does not give a fuck about the invisible boundary, or whether there are cars coming, so…he’s always on a lead in the front yard. Because 99 percent is not enough for that. In fact for this dog I don’t think there’s going to be 100 percent of anything. He’s a border collie mix, and whatever he was mixed with is stubborn and probably a primitive breed.

Mini Zuke’s are our main treat actually, along with small bits of low fat string cheese. Pencil eraser sized would be big for that treat, maybe half that. My concern about excess calories mainly comes from the peanut butter frozen into the mini-Kongs …

Right now the negative reinforcement discussion is centered around helping teach bite inhibition. He tends to want to jump up and try to bite my wife’s face as part of his natural play when he sees her and bite/chew all of her clothes edges hanging down. For whatever reason I don’t get that much. I am not sure that the praise for “nicer” play approaches and encouraging the bite/chew toy use with praise, and getting up and walking away or ignoring for a minute or so for the bite play attempt, is enough to get the message across and am encouraging my wife to use the deep “no” and then minute of ignore, or even the one second hold of the muzzle shut with the “no” before the ignore, after those bites, to teach that that particular biting play is not acceptable. She is in charge and does not want to be “negative.”

Housebreaking is going great. Only issues are completely our fault. He gets very excited when my wife gets home and we have learned that even if I just took him out right before she came in the door she should take him out with her until he gets past his excited to see her piddle session (it takes maybe three minutes and then calmer). And when my daughter is on duty she is I think not as attentive to watching him for signs closely nor automatically taking him out as frequently so there are sometimes one two pee accidents found when I get home. He’s otherwise pretty perfect already. No scolding for accidents had been used or were needed, if seen starting to squat a mildly loudish “whoa!” to startle and try to stop in progress as whooshed up from underneath and rushed outside to complete which he then does followed by praise or treat plus praise if I’ve been able to grab a treat on the way out the door. Otherwise after play sessions, before and after nap times, every 20 minutes after meals for at least three times, and no longer than two hours apart otherwise except when asleep in his crate. He sleeps all night from the third night home.

Did I say he is smart? :slight_smile:

araminty oh I am very aware that there is no one true way. Hearing the variety out there as well as reading it (I’ve liked Dawn Antoniak-Mitchell’s “Teach Your Herding Breed to be a Great Comanion Dog” the most so far, and have found Pryor’s website as an intro to not appeal so much) is useful. Yes the puppy training class will have lots of sway if just for consistency’s sake. First class tomorrow.

Thank you all for the opinions, the information, and the new puppy owner support!

FWIW the instructor is not big on clickers and advises the word “yes” as the bridging stimulus used in the same way as a clicker is. So it shall be, at least for now, for so my wife hath spoken!

I am though personally convinced by posters here that clickers can offer a slight advantage over the word “yes.” But yes consistency is key.

Big issue for now remains the dog’s play bites of my wife and daughter and on their clothes for attention. (Not sure why he does not do either to me.) The instructors advice was to yelp loudly and then briefly ignore when his teeth touch skin and time out for biting clothing.

Thanks again for the input and information.

I would also suggest having several chew toys around the house. When he is biting, try yelping (works for some dogs, but not for all), and then immediately re-direct by offering him the chew toy. Praise him when he starts chewing on that instead of you/wife/daughter/furniture/carpet/wall.