Sunday, March 4, 2012

Jane Killion Seminar: Attention is a Behavior, and You Can Teach It!

Does your dog blow you off, ignore you, or run away? Do you feel frustrated by his lack of focus? Do you want him to pay more attention to you? Well, you’re in luck, because Jane told us at the seminar that attention is a behavior, and you can teach it. Today I’m going to share Jane’s two-step method for creating an attentive dog.

Step 1: Teach the Concept
You might think you’ve done this. After all, you’re a caring dog owner interested in training, so you’ve probably taught your dog to make eye contact when cued. But this, Jane argued, is missing the point. As she put it, attention is like respect. If you have to ask for it, you’ll never get it. If you’re calling your dog, nagging him, or begging for attention, then he’s learned nothing… except maybe that he can disengage from you to do whatever he’d like because you’ll let him know when it’s time to pay attention.

What Jane wants is a dog who chooses to be attentive. To that end, the concept Jane encouraged us to teach is not cued eye contact, nor is it for the dog to ignore distractions when told. Instead, she said that true attention is achieved when the dog learns that distractions themselves are cues to focus on the handler. In other words, when the dog realizes there is something interesting out there, he should not only ignore it, but also make eye contact with his handler without being told to do so.

Jane starts teaching this by letting the dog know that eye contact is a behavior he can offer in order to receive a click and treat. She showed us how to do that by putting a dog on leash and simply waiting. When the dog would make eye contact, no matter how briefly, she would click and treat. She had each dog/handler team repeat this until it was clear the dog understood that he can make the click happen by offering eye contact.

With this foundation in place, Jane began to raise the criteria using the classic technique of introducing distractions. Jane had the handlers hold a fistful of treats out to the side. Most dogs will stare at the food when this happens, and our working dogs were no exception. Again, Jane just had the handlers wait. She cautioned that they should not call the dogs, make noises, or otherwise prompt them to make eye contact. Simply wait. When the dog looked, he would get a click and treat. Again, she had the teams repeat the exercise until it was clear the dog was offering attention as a behavior to earn the treats.

But remember, the concept is not simple eye contact; Jane’s goal was for the dogs to learn that those distractions are a cue for the dog to pay attention to the handler. Because of this, Jane does not want a dog to look away from the handler, check out the distraction, and then look back. Instead, she wants the dog to understand that as soon as he’s aware of the distraction, he should put his complete focus on his handler. So instead of making the distraction bigger or harder, Jane actually had the working teams raise the criteria by only clicking when the dog refused to look at the distraction.

Here’s how the exercise was set up: the handler would present a distraction (usually by holding a handful of treats out to the side), and the dog would be allowed to look at it, then look at his handler and earn a click/treat. The handler would then remove the distraction (by putting her behind her back, for example), and then present it again, which started the exercise over. After a few reps, the criteria would be raised. Now the dog would only get clicked if he maintained eye contact and ignored the distraction. The first few times, the dog would look at the distraction before looking back at his handler. However, now that the criteria had been raised, he would not get clicked for this. Every single dog soon realized that he shouldn’t look at the distraction at all, and thus would choose to keep looking at his handler, which earned him a click and treat. It was pretty awesome to watch.

Step Two: Work the Envelope
Once the dog learned the concept, Jane began to do what she calls “working the envelope.” This is the phrase Jane uses to describe the process of gradually making the training task more difficult. She would change one thing at a time, with the goal being that the change was large enough that the dog would notice, but not so large that he couldn’t quickly return his attention to his handler.

Tibby has figured out that when Jane moves towards her, she should look at her mom.

Again, she would allow a few reps of look-at-the-distraction-then-back-at-the-handler before requiring the dog to ignore the distraction entirely in order to get clicked. Each team worked with different levels of distraction. For more novice dogs, the handful of treats might simply be moved closer to his face, while more advanced dogs were required to ignore Jane while she stepped towards them or crinkled a treat bag.

Although she couldn’t demonstrate the entire evolution of the process due to time constraints, Jane did share that pretty much all training is just an expansion of the attention concept. What’s more, she stated that pretty much every problem you have with your dog’s performance is because he doesn’t understand the concept of attention. She even argued that teaching solid attention skills is at the core of working with reactivity or aggression. While I think attention skills are helpful, I know from my own experiences with Maisy that it's not a one-size-fits-all solution, and I have my doubts that what I saw would have worked.

But that is for another post. In the meantime, I'd love to hear from others. How have you taught attention skills? What do you think of Jane's method? Do you cue attention, or have you taught your dog to view the distraction itself as the cue?

14 comments:

Raegan said...

"Here’s how the exercise was set up: the handler would present a distraction (usually by holding a handful of treats out to the side), and the dog would be allowed to look at it, then look at his handler and earn a click/treat. The handler would then remove the distraction (by putting her behind her back, for example), and then present it again, which started the exercise over. After a few reps, the criteria would be raised. Now the dog would only get clicked if he maintained eye contact and ignored the distraction. The first few times, the dog would look at the distraction before looking back at his handler. However, now that the criteria had been raised, he would not get clicked for this."

I assume that if the dog broke eye contact, the fist with the treats went behind the back again?

I need to finish reading the Training Levels on Focus. I know there's arguments both for and against naming attention. I agree that attention should be automatic. Asking for it implies the dog has discretion to not be attentive. There are context clues that he should be paying attention.

I also agree that attention is at the core of dealing with reactivity, but it's not the entire answer. I imagine in many dogs it makes things worse. See Control Unleashed, where the dogs *need* to check out stuff to feel safe.

Crystal Thompson said...

You assume correctly. If the dog fails, just reset the exercise.

I like 99% of what Jane does. I really like her attention technique; I love that distractions are the cues, especially. That goes along with Denise Fenzi's "work is a privilege" and even the Control Unleashed info on environmental cues (for example, uncued Look at Thats).

K-Koira said...

Pretty much the only thing I took away from puppy class with Koira was her full and total attention. Even when she is not looking at me, she is paying close attention to what I am doing. If I am standing and talking to someone while out on a walk, Koira may lay down to rest a bit. If I so much as move a foot, her head comes up and she checks with me, are we doing something?

Same thing goes for when we go to the park or on hikes off leash. She may run around like crazy and feel free to explore, but if I turn and walk the other way, she notices within seconds and comes to check in. She also checks in randomly.

Pallo is not nearly as good about this. He will check in while on off leash hikes, but not as often, and it takes him longer to realize that I have changed direction. If he is on leash, he is paying constant attention to everything except me. We're working on changing this, but it is a slow process.

Kristen said...

Thanks for sharing!

After some things I heard Kathy Sdao say last year, we pretty much stopped teaching attention on cue in classes. For some teams we still will do it... but not typically.

With my own dogs, I do want distractions to be an attention cue but I also want a way to verbally cue the behavior. We need it for some moments in our various sports and activities were I need my dog to go from focused elsewhere to focusing on me.

Ninso said...

Shaping eye contact(or, capturing, more accurately) is the first thing I do with every new dog, and they catch on SO quick! It makes me cringe every time I hear someone talk about "luring" eye contact by bringing a treat up to your eyes (not that that is wrong, just so much more cumbersome). I want it to be a default behavior and I don't cue it. The idea of treating the distraction as a cue for giving eye contact is new for me though. I will have to think about that one.

Although I don't have an attention cue, once I have shaped eye contact I do teach my dogs that their name means "give me your attention." I feel like having a separate attention cue is redundant. Why would I be calling my dog's name unless I wanted them to pay attention?

Andreja said...

I am confused.
Imagine for a second that you were 100% successful at teaching your dog to look at you every time the she sees/hears/smells *anything* interesting. What would your walks look like?
Dog smells something, therefore she looks up. She sees a stick on the ground, therefore she looks up. She sees another dog... you get the picture. My dog would have to walk looking up at my face the whole time, because sniffing for interesting things would get replaced with attention.

I am asking really seriously because I did teach Ruby to offer attention during walks and ended up with so much attention heeling that I was worried for his neck! He got really strong muscles at the withers, too :)

I agree with you that working on attention can't be a cure for reactivity, at least not in all cases. With all that offered attention it became harder to know when my dog feels uncomfortable around other dogs because he has been trained to look at me instead. It only took a few meetings with loose dogs who came to sniff him anyway for him to get conflicted about it - on one hand he should be looking at me, but on the other his experience with loose dogs tells him he better be prepared for them.
Sure, good focus allows us to work even when there are other dogs around. But walks are not "work" and I have no interest in keeping Ruby at the level of arousal that he needs to work.

So in my opinion attention must be cued to be useful. It can be cued by context if you wish, or by your stance, tone of voice... I use a special cue for start of work and I expect to have Ruby's full attention until I cue end of work. During walks I would prefer to have passive attention - not looking at me, but knowing where I am & responding to motion and verbals.

Shannon said...

Great post! One of the things I like about the old school Koehler method is exactly that - it teaches the dog to see a distraction as a cue to pay attention to the handler. It's awesome to see a positive method to achieve this. You've inspired me to go buy Killions' book, and I will definitely be incorporating this into my training with my dog.


Andreja has a good point, though. Most dogs will pick up contextual cues, and understand that if you're not actively working with the dog, they don't have to focus on you completely - they can check in. It's like with shaping behaviors - if you're not clicking the dog won't constantly be offering you things (one hopes, LOL.) So I think they can understand to give you attention when you're working vs. when you're not.


My dog will give me very focused attention when she wants something from me. But she knows she can look around at the environment, too. I find it useful to have a "go play / go sniff" cue which means "you don't have to give me your undivided focus right now."


I also like having the dog's name be a cue for attention - if you can ask for it, the dog knows he/she doesn't have to be offering it constantly.

Anonymous said...

I wonder if it might be helpful, in this instance, to narrow down what type of "reactive" behavior you're dealing with. For example, if your dog is "reactive" out of fear, it seems like this type of attention, where the distraction is supposed to be the cue, will set your fearful dog up for higher distress, having to choose between looking at the object of the fear or giving you attention.

If, however, your dog is "reactive" more in the sense of being hyper, hard to focus, etc., then distraction as cue seems much more reasonable.

I like the blanket term "reactive" but sometimes I feel like it's helpful to identify the source of the reactivity, too. In this case in particular, while I love the idea, and may use the method to actually work more on our attention skill, I think taking into account what type of reactivity you are seeing with your dog will make a huge difference.

Nicky

Crystal Thompson said...

ANDREJA- You know, that's a really good question. In addition to context clues (like Shannon mentioned), Jane really emphasized the importance of a release word to let the dog know when he doesn't have to pay attention anymore. The implication is that she must have a way to start a session again to indicate to the dog that he needs to start paying attention again. But once she's let the dog know it's time to work, she won't nag him about it. Attention is part of his job, and she expects that he'll give it.

NICKY- SHHHH Don't steal my next post. :) Absolutely, reactivity is a spectrum, from sensitive to the environment to distractable to clinically anxious. Depending on where on that spectrum a dog falls, I think training attention could either be highly successful or a dismal failure.

Andreja said...

Ok, confusion cleared up. I got so hung up on "there's no cue", but there really is a context cue there in addition to distractions. It makes sense now :)

I like how she teaches attention. I don't know if I somehow missed it in her book or if it's a new method since the book was published, but I don't recall reading about it before. Thanks for posting :)

CattleDog said...

Great idea for teaching attention. I realized at my first Obedience show that I never taught attention beyond "look at me when I ask." We never had any problems with "attention" in Rally (where I could frequently remind him to check in), but we were toast in the Obedience ring! Since then we've been working on un-requested attentiveness, and I like the exercise.

But I have a quick question about using the treat out to the side. Maybe Jane addressed this in her seminar?
I use a lot of hand-signals with my dog (finish, drop on recall, moving stand, etc) and I don't want to desensitize him to me moving my hands out to the side (literally my signal for finish right is to swish my arm out in a circular motion). Any thoughts on this? Of course the first time I swung my hand out with the treat, my dog finished right. After that I told him to stay and he's doing great with the exercise, but I'm worried that he'll responding to my hand signal if I do this exercise too much.
Or maybe I'm just too paranoid (this is the first dog I've ever trained beyond "sit, stay") and my dog will know the difference between working on a focus exercise and following signals.
Right now I figure the treat focus game is temporary and even if it momentarily extinguishes or lessens his response to my hand signals, he'll pick it back up when I stop the game?

Laura, Lance, and Vito said...

Cattle dog,
Jane did not address that at the seminar but I think you're over thinking :) Did your dog really see that you had a treat in your hand and still finish? If so, then you could either choose to reward that since your dog left the treat to do the behavior, or not reward and tell the dog stay like you did. The whole point is the dog ignoring the distraction and looking at you so if your dog decides your proofing hand signals than fine! But if you bring the treat from the dog's nose and then out to the side it's really clear to most dogs that you're not asking the dog to finish, or find the jump over there, or...

Crystal Thompson said...

Cattle Dog- As a perpetual overthinker, I appreciate your question! I have two thoughts:

1. I have no cues that involve a fist with a cookie in it. I bet you don't either.

2. If you're really worried about it, simply stand next to a table/bookshelf/whatever and set the treat there.

CattleDog said...

Thanks for the feedback. I'm probably just over-thinking it. It only took about 3 tries for my dog to figure out that a fist going out is not the same as the "swoosh" motion with a flat hand that I usually use for a finish.
We did that a few times and just skipped to the next step (enlisted my roommate with a bag of treats)!
Great exercise.