Tuesday, March 29, 2011

Clicker Expo 2011 (Chicago): Ken Ramirez- Aggression Treatment and Context, Part 2

In part one of my summary on Ken Ramirez's talk on aggression, we discussed aggression in general. Today, I'll tell you what he said about treating aggression, starting with generalities, and moving on to specific treatments.

Ken said that you can organize all of the various techniques by placing each one into one of three categories: broad scientific approaches to learning, scientific principles, and practical procedures.

The first category, broad scientific approaches, refers to the twin concepts of learning theory: classical conditioning, and operant conditioning. I call them “twin” because, as Ken pointed out, while we can choose to focus on just one of these two approaches, “all animals learn both ways all the time.” As the old saying goes, you have Skinner on one shoulder and Pavlov on the other.

The second category is made up of the various scientific principles which have been developed in the experimental lab. Each one generally falls under either classical conditioning or operant conditioning. Under classical conditioning, we have things such as habituation, flooding, and counter-conditioning. Under operant conditioning, we have both methods that provide consequences to behavior (punishment and reinforcement), as well as redirection techniques (the Differential Reinforcement of Alternative behaviors).

Finally, our third category is made up of the practical applications and strategies borne out of one or more of these scientific approaches. Each of these techniques is “a way that a skilled and talented trainer has operationalized the science to deal with aggression.” They typically have components of both classical and operant conditioning.

Okay, let's dive into some of the specific approaches out there. This is not meant to be an all-inclusive list, but it does include some of the more common ways people respond to and deal with aggression.

Positive Punishment
Using pain or fear to “correct” aggression is a method that people often think of instinctively, probably because it's how parents, teachers, coaches, and yes, by dog trainers have responded for centuries. Done correctly, punishment works, however there are risks, challenges, and fallout to using punishment... including aggression. Seeing as how we were at Clicker Expo, Ken did not discuss punishment in depth, other than to say that trainers should understand how and when to use it. Although trainers should not throw punishment out of the toolbox altogether, they should allow that toolbox to remain on the top shelf, collecting dust.

Classical Conditioning
Falling in the category of broad scientific approaches, and having no real specialized names or operational procedures, classical conditioning is often one of the first tools skilled trainers use. Ironically, it is also the one that inexperienced ones often overlook. This is a mistake; classical conditioning, that is, changing the dog's associations to his triggers, is powerful even if it seems simplistic. Seems is the key word here; in practice, classical conditioning requires thoughtful implementation since you need to keep the dog below his threshold. Though simple, it is easy to screw up if you don't understand the science.

Look at That (LAT)
This technique was developed by Leslie McDevitt and described in her book, Control Unleashed. LAT uses a cue to tell the dog to look at a trigger in order to get rewarded. This changes the dog's associations with his trigger, and is thus largely a classical procedure, albeit one with a strong operant component. It is useful prior to a dog having a reaction to a trigger, and is quite versatile as it can be used in many situations. However, it must be trained in advance so that it can be used sub-threshold. It is not a complete strategy in itself and must be used in conjunction with other tools.

Constructional Aggression Treatment (CAT)
CAT was developed by Jesus Rosales-Ruiz and Kellie Snider. It is a negative reinforcement procedure which rewards the dog for appropriate behavior by having the trigger (human or canine) leave. It tends to work incredibly fast because it treats the root source of the problem- the dog's desire for distance. It is also highly controversial because it exposes the dog to his trigger for long periods of time. Still, Ken feels it's a useful technique when exposure to the trigger is unavoidable on a regular basis. It requires a very skilled trainer who can set up the situation correctly and direct the trigger to leave at precisely the right moment. It is also not right for every dog as it will only work when you have a thorough understanding of both the specific trigger and context in which aggression occurs. It won't work for a dog whose triggers “stack.”

Click to Calm
Emma Parson's book Click to Calm lays out an easy-to-follow program that relies primarily on redirection techniques such as the differential reinforcement of incompatible behaviors, other behaviors, or lower intensity behaviors, but also capitalizes on classical conditioning. What all that scientific mumbo-jumbo means is that the trainer shapes the absence of aggression by clicking the dog when the dog's aggressive display lessens even slightly. This is a highly useful approach because, unlike other techniques, it can be used when the dog is over threshold. However, this does mean that you are clicking the dog for acting aggressively, even if it is for a reduction in aggression. It is also time-consuming and can be difficult for the unskilled trainer. Still, done well, Ken believes it can be a permanent fix.

Training an Incompatible Behavior
This encompasses a broad group of behaviors, including such techniques as “watch me” (where the dog looks at the handler instead of the trigger), U-turns (where the dog is cued to turn and go in the opposite direction away from a trigger), and recalls or whiplash turns (where the dog immediately returns to the handler). Although the behavior itself is different in each case, the goal is the same: to teach the dog something to do instead of being aggressive. If it's trained well, the dog will respond automatically, giving the trainer a chance to intervene and prevent aggression. Unfortunately, it doesn't change the underlying cause, and thus won't cure aggression. It should be followed up by other methods. 

Abandonment Training
Popularized by Trish King, abandonment training is a very specialized tool useful only for dogs whose aggression revolves around their owners. In abandonment training, the dog is on both a leash and a long line. The owner walks holding the leash, and a secondary handler holds on to the long line for safety considerations. When the dog behaves inappropriately, the owner drops the leash and leaves. In scientific terms, this is negative punishment- bad behavior makes the owner go away. Although effective, as noted, it's only effective for a small handful of dogs.

Behavior Adjustment Training (BAT)
Created by Grisha Stewart, BAT has taken the positive training world by storm. Ken really glossed over it since he isn't terribly familiar with it (he's tried at least some version of all the other techniques presented, but hasn't tried BAT yet). All he really said is that it's a negative reinforcement technique that he feels is effective.


These are not all of the ways to deal with aggression, but it is representative of the types of approaches out there. Ken said there isn't any one technique that is the answer to every aggression problem or every situation, which is why new methods are being created all the time. If you are trying to decide if a particular technique would be useful for your dog, you should learn everything you can about it first. Understand the science so you can recognize how it works and compare it to other approaches. Then, decide if it's a good fit with your dog's training history, type of trigger (is it predictable? Controllable?), specific circumstances which provoke the response, the level of risk or danger to all involved, your own experience level (and remember, Ken thinks aggression should be treated by professionals, not the average pet owner), and your own personal ethics.

Personally, I've used Look at That a lot, as well as general counter-conditioning. I've done some work with incompatible behaviors, and although that won't cure the problem, it does allow me to interrupt Maisy before things get out of hand. Maisy is not a candidate for CAT (her triggers stack too much), and her veterinary behaviorist did not think BAT would be a good fit for her, either. Likewise, Click to Calm and Abandonment Training really aren't suited to Maisy, and my ethics do not allow for punishment.

Okay, it's your turn: if you have a reactive or aggressive dog, which approaches have you tried? Did they work? If not, why do you think that was?

7 comments:

Patty said...

Great post summarizing the different treatments for aggression. Very interesting read!

I use LAT with Sophie. Previously I was trying to get her to sit and focus on me whenever another dog approached. She was rewarded for not breaking her focus on me after the look/watch me command was issued. Disaster! Watch/look became a trigger because she knew a dog was around and she couldn't see it otherwise she was breaking her watch command. She would get agitated and react when I said the word, even if she hadnt yet spotted the dog We switched to a LAT approach and there has been significant improvement!!

We have worked U turns and whiplash turns which are good to have in case we can not avoid another dog (off leash dogs can be a nightmare)

I have not really tried other methods. I don't feel I am knowledgeable enough to try CAT, it also seem counter intuitive to me. But again I am no expert.

Tegan said...

Sorry for not answering any of your questions. :P But I just wanted to say that I very much appreciate this summary - a lovely concise description of aggression rehab methods.


Could you explain to me the difference between CAT and BAT?

Crystal said...

Patty- If you don't feel comfortable with a method, then I think it's great you haven't tried it! I have seen some footage of CAT procedures, and it's pretty impressive- it works quickly, but I think I'd have a hard time doing it with my dog. The handler more or less ignores the dog. That's hard for me!

Tegan- I can TRY to explain the difference. Please keep in mind that Ken did NOT describe BAT, and I haven't done it personally, so it is quite possible that I might muck up some details.

In CAT, the handler/owner acts simply as a restraint, and holds the dog's leash while the trigger (can be either dog or human) approaches. The dog will become upset, and the trainer will watch for signs that he is calming down- turning his head away, becoming quieter, etc. The moment he does, the trigger leaves. The dog learns that good behavior (non-aggressive behavior) makes the trigger go away, and that when he acts aggressively, the trigger stays. There is more to it than that, but I'm afraid I don't understand the details well enough to share.

I am even less familiar with BAT, unfortunately. BAT also uses distance as a reward, except instead of the trigger leaving, the dog itself gets to leave. This typically makes BAT a bit easier to do in the real world. I know there are some other things that go with BAT, but again, I don't know enough about it to share. There does seem to be a greater emphasis in keeping the dog subthreshold, which implies that it's easier on the dog, but I'm not sure if that perception is accurate. If you're interested in learning more about BAT, I linked to Grisha's site in my post.

elizabeth deitz said...

In classes I usually start with Counter- conditioning first as it is easiest to understand and gives some level of success and accomplishment. My more experienced students pick up on B.A.T. very quickly, some becoming strong advocates almost immediately.
Nice summary! Thanks.

Joanna said...

When I work with reactive dogs I focus on training various incompatible behaviors and look at that, and also integrate a few principles of BAT. Interesting that he says that they're not complete techniques on their own... I feel like the methods I use (and was guided to use by my mentor) do get at the underlying feelings of fear/anxiety because of classical conditioning always taking place.

Crystal said...

Joanna, I agree- it was interesting to hear which techniques Ken felt were not stand-alone. LAT has been my main training tool, and I think the reason it's worked so well is because of classical conditioning.

Still... I've used other methods, as have you. I suspect there are very few people using only one technique per dog since so many of the approaches complement one another quite well.

ddbb said...

LAT, desensitization. ^^
I think let the dog gain some confident during training is quite important.
Daisy