Constructional Aggression Treatment

[CliftonPark]

Does anyone know a trainer in Melbourne (don't care where) that uses C.A.T Constructional Aggression Treatment method for treating aggressive dogs?

[Cosmolo]

I have used a similar method before on a few different dogs (never heard it called that) but much depends on the dog as to what the method is- i don't know of anyone who uses that strategy for every dog every time.

[pipsqueak]

Does anyone know a trainer in Melbourne (don't care where) that uses C.A.T Constructional Aggression Treatment method for treating aggressive dogs?

can you explain what this is? i haven't heard of it before.

[Nekhbet]

you use what is right for the dog. Frankly there are better ways and I have a problem with not stopping a dog showing explosive behavior. Done incorrectly it can push a dog to try harder.

There are plenty of us here on DOL that deal with aggression, myself , Erny, Cosmolo, Kelpi-i etc

[CliftonPark]

I have used a similar method before on a few different dogs (never heard it called that) but much depends on the dog as to what the method is- i don't know of anyone who uses that strategy for every dog every time.

The C.A.T. procedure I am referring to is the one developed by Jesus Rosales-Ruiz and Kellie Snider. Not requiring trainer that only uses this method, just want to talk with a trainer that is knowledgable about this procedure and has experience applying it. I watched the seminar conducted by Jesus and Kellie and am interested in discussing it further with a trainer that has experience with it.

[Cosmolo]

Well to be honest, i have never heard it called that and only looked up and read what you were referring to. Although i'm pretty sure we've done similar things at times, its highly unlikely to be exactly the same and i don't know any trainers who promote that they use that particular system.

When/ where was the seminar you attended?

ETA Just read a little more- using this technique alone may prove difficult in practical situations where you cannot control the other dogs movements

Edited December 9, 2009 by Cosmolo

[CliftonPark]

Does anyone know a trainer in Melbourne (don't care where) that uses C.A.T Constructional Aggression Treatment method for treating aggressive dogs?

can you explain what this is? i haven't heard of it before.

It's difficult to explain properly and be brief. Here some info I have copied. C.A.T is an operant approach to treating aggression and uses negative reinforcement. It is a shaping procedure. The Constructional approach, which was presented by Ruiz-Rosales and Snider at the seminar, maintains that aggression is a learned behavior and it is persistent because it has worked in the past. Aggression makes other dogs go away, that aggressive behavior becomes very reinforcing. The dog learns that aggressive behavior pays off. So, what would happen if the aggressive behavior didn’t pay off?

Since the function of most aggression is to achieve distance from the icky thing (other dogs, humans, skateboards), if you can manipulate the consequences (the icky thing DOESN’T go away when the dog aggresses) then the dog will start to experiment with which behaviors actually will make the icky thing go away. Initially the behaviors may be very small; a sideways glance, blinking the eyes, or a head turn. The dog learns that THOSE behaviors WILL make the icky thing go away. As the treatment progresses the dog begins to feel more comfortable, he will offer more and more alternative behaviors, building a repertoire that is more social and friendlier. All of this is done under carefully controlled circumstances with the trainer, the client and his or her dog. In the cases of dog to people aggression a series of humans are used as decoys to work with the aggressive dog so that the dog can take these alternative social behaviors and begin to apply them across the board. The same technique is used with dogs who are aggressive to other dogs. Decoy dogs work just beyond the dog’s comfort level and move ever closer as the dog softens and becomes more accepting. The end goal for all types of aggression using these techniques is total interaction not just tolerance.

[CliftonPark]

you use what is right for the dog. Frankly there are better ways and I have a problem with not stopping a dog showing explosive behavior. Done incorrectly it can push a dog to try harder.

There are plenty of us here on DOL that deal with aggression, myself , Erny, Cosmolo, Kelpi-i etc

I agree totally that you use what is right for the dog. I have seen the seminar dvd's and would now like to find someone that has experience with it to learn more about it from people that have experience with treating aggression and using not only C.A.T. but other methods to determine it's strengths and weaknesses. I just have not been able to locate anyone yet that has a thorough knowledge and experience with C.A.T. I'm hoping with all the dolers out there that there will be someone that does have that knowledge and experience.

[poodlefan]

CP:

Since the function of most aggression is to achieve distance from the icky thing (other dogs, humans, skateboards), if you can manipulate the consequences (the icky thing DOESN’T go away when the dog aggresses)

That might be the case for fear aggression but I can't see it working for other forms of aggression. A dominant dog showing signs of aggression is NOT going to de-escalate if the other dogs doesn't leave.

Analysis of the dog would be critical here.. and it would be potentially dangerous to get it wrong.

[CliftonPark]

Well to be honest, i have never heard it called that and only looked up and read what you were referring to. Although i'm pretty sure we've done similar things at times, its highly unlikely to be exactly the same and i don't know any trainers who promote that they use that particular system.

When/ where was the seminar you attended?

ETA Just read a little more- using this technique alone may prove difficult in practical situations where you cannot control the other dogs movements

The procedure is done in very controlled situations intially before moving to the real world. Yes I agree with you and this is part of what I would like to understand from anyone using this method. How did they generalise it to the real world?

[CliftonPark]

CP:

Since the function of most aggression is to achieve distance from the icky thing (other dogs, humans, skateboards), if you can manipulate the consequences (the icky thing DOESN’T go away when the dog aggresses)

That might be the case for fear aggression but I can't see it working for other forms of aggression. A dominant dog showing signs of aggression is NOT going to de-escalate if the other dogs doesn't leave.

Analysis of the dog would be critical here.. and it would be potentially dangerous to get it wrong.

I'm thinking if the aggression is a learned behaviour it can be manipulated with this type of procedure. I think as you that a dominant dog would be a problem.

[poodlefan]

CP:

Since the function of most aggression is to achieve distance from the icky thing (other dogs, humans, skateboards), if you can manipulate the consequences (the icky thing DOESN’T go away when the dog aggresses)

That might be the case for fear aggression but I can't see it working for other forms of aggression. A dominant dog showing signs of aggression is NOT going to de-escalate if the other dogs doesn't leave.

Analysis of the dog would be critical here.. and it would be potentially dangerous to get it wrong.

I'm thinking if the aggression is a learned behaviour it can be manipulated with this type of procedure. I think as you that a dominant dog would be a problem.

Sounds like a very low key version of "flooding". "Trickling" perhaps?

[CliftonPark]

CP:

Since the function of most aggression is to achieve distance from the icky thing (other dogs, humans, skateboards), if you can manipulate the consequences (the icky thing DOESN’T go away when the dog aggresses)

That might be the case for fear aggression but I can't see it working for other forms of aggression. A dominant dog showing signs of aggression is NOT going to de-escalate if the other dogs doesn't leave.

Analysis of the dog would be critical here.. and it would be potentially dangerous to get it wrong.

I'm thinking if the aggression is a learned behaviour it can be manipulated with this type of procedure. I think as you that a dominant dog would be a problem.

Sounds like a very low key version of "flooding". "Trickling" perhaps?

trickling I like that.

They are shaping behaviour though by selectively reinforcing improved alternate behaviours that are offered. Very minor alt behaviours initially (a look away, sniff of the ground whatever) but if the trainer is really good with timing and selecting what to reinforce from what I've seen (and they show the trials of many dogs) the dogs do learn to offer behaviours other than aggression. I think it really needs a very good trainer. I do think it could easily be done incorrectly in the wrong hands.

The other thing that I liked was that the handler of the target dog (usually the owner but sometimes they tether the dog but with the owner still present if they are concerned about redirection or the handler not being able to hold the dog if a big aggressive response does happen) is that the handler does nothing...they do not speak to the dog, do not reniforce the dog just be there and do nothing. It takes all the pressure off the handler who is usually stressed about their dogs aggression.

[Aidan]

I know Kellie Snider reasonably well and have had several discussions with her about this procedure from the beginning.

Functional Analysis is ALWAYS undertaken as this procedure only works for "distance increasing" behaviours. One of the fears I have for this procedure now that it is popular is that very few people are competent with FA so the method is vulnerable to blame shifting.

It will not make the aggression worse if done correctly. If done incorrectly it can make it worse. One requirement for doing this correctly is that you have a lot of time to complete the procedure, however long that takes, hours.

It is definitely not flooding, however I personally believe (and Kellie Snider disagrees) that there is a lot of habituation going on. It is (or should be!) a Differential Reinforcement of Alternative Response procedure, allowing unwanted responses to extinguish (reasonably quickly) while other responses are reinforced via negative reinforcement. "Spontaneous Recovery" is mitigated to some extent, although there is very little data on this and I think some of the claims need backing up.

Because it is a long procedure where you essentially attempt to "finish the job" in very few sessions, running into uncontrolled dogs should not be an issue. In theory. I have my doubts, and I believe that handlers should be trained in using "non-functional" reinforcers to maintain the alternative responses, but Kellie and Jesus disagree with me on this point.

Some personal criticisms:

1. studies were not long enough to make claims regarding long-term efficacy of procedure

2. habituation and other effects have not been proven to not take place, leading to questions about long-term efficacy

3. no measures of respondent behaviours were taken, therefore no indication of biochemical or nervous system state other than outwardly visually observable phenomena

Criticisms aside, it is an excellent procedure if the practitioner is skilled and has a solid practical ability with functional analysis, putting aside their own interpretations of behaviour. It does have more to offer than the basic use of functional reinforcers which are nothing new to the behaviour world.

[CliftonPark]

I know Kellie Snider reasonably well and have had several discussions with her about this procedure from the beginning.

Functional Analysis is ALWAYS undertaken as this procedure only works for "distance increasing" behaviours. One of the fears I have for this procedure now that it is popular is that very few people are competent with FA so the method is vulnerable to blame shifting.

It will not make the aggression worse if done correctly. If done incorrectly it can make it worse. One requirement for doing this correctly is that you have a lot of time to complete the procedure, however long that takes, hours.

It is definitely not flooding, however I personally believe (and Kellie Snider disagrees) that there is a lot of habituation going on. It is (or should be!) a Differential Reinforcement of Alternative Response procedure, allowing unwanted responses to extinguish (reasonably quickly) while other responses are reinforced via negative reinforcement. "Spontaneous Recovery" is mitigated to some extent, although there is very little data on this and I think some of the claims need backing up.

Because it is a long procedure where you essentially attempt to "finish the job" in very few sessions, running into uncontrolled dogs should not be an issue. In theory. I have my doubts, and I believe that handlers should be trained in using "non-functional" reinforcers to maintain the alternative responses, but Kellie and Jesus disagree with me on this point.

Some personal criticisms:

1. studies were not long enough to make claims regarding long-term efficacy of procedure

2. habituation and other effects have not been proven to not take place, leading to questions about long-term efficacy

3. no measures of respondent behaviours were taken, therefore no indication of biochemical or nervous system state other than outwardly visually observable phenomena

Criticisms aside, it is an excellent procedure if the practitioner is skilled and has a solid practical ability with functional analysis, putting aside their own interpretations of behaviour. It does have more to offer than the basic use of functional reinforcers which are nothing new to the behaviour world.

Thankyou Aidan great information.

I felt the same regarding habituation and that the dog could get used to the decoy dog or person and the environments that the training is occuring in espec due to the amount of time exposed to them with this type of training and that a number of appropriate decoy dogs would be required as well as people handling the decoy dogs to really proof the procedure.

Re generalising in new environments that can't be controlled, this is where I have a concern about uncontrolled dogs. I understand you need to complete the procedure in a controlled environment first and because of this the dog in theory should not be reactive/as reactive. But if habituation of the controlled environments/dogs/people has occured I do think there is a real risk that the dog may react in the uncontrolled environment espec if an inappropriate dog is confronted.

I'm thinking you almost need a team of people monitoring the uncontrolled environment to minimise risk until you have proofed the dog in many environments with many dogs and people.

I thought there may need to be some type of ongoing training (by handler/owner) to maintain the strength of the alternate responses espec if the dog is inadvertantly exposed to inappropriate dogs to help prevent regression. Excuse my ignorance but could you please explain 'non functional reinforcer'

You said that this procedure is now popular. Do you know some good trainers in Melbourne that use C.A.T. you would recommend I talk to.

[Cosmolo]

I have found when doing similar work with dog reactive dogs, that some dogs do get used to our dogs quite quickly. We start with our calm well behaved dogs for obvious reasons but are under no illusion about the need to then transition the client dog to working around more realistic dogs- the bouncy silly type. While we can get our dogs to do those things, the learning history prior to that generally means the reactivity is much less if there at all when we cue our dogs to be 'normal'

Great post Aidan

[Aidan]

Re generalising in new environments that can't be controlled, this is where I have a concern about uncontrolled dogs. I understand you need to complete the procedure in a controlled environment first and because of this the dog in theory should not be reactive/as reactive. But if habituation of the controlled environments/dogs/people has occured I do think there is a real risk that the dog may react in the uncontrolled environment espec if an inappropriate dog is confronted.

Apparently it's not an issue in the short-term, but I just have lingering doubts about what happens in the longer term. Dogs do get pretty adept at picking up little signals about what to expect (or what you expect) when you put them into a controlled set-up if it's too obvious (like in CAT). All reports are that they DO continue their new patterns in uncontrolled environments so far, but this is fairly new.

If you don't maintain a change at an operant level, or fail to make a change at a respondent level (and maintain it), what happens when the dog is pushed a little too far and barking/lunging work for him again?

At least when you use a non-functional reinforcer (e.g food) you always have control over that reinforcer, and can continue to maintain those operants with those reinforcers. The point of controlling the functional reinforcers is very valid, but it's the hardest thing to control. I think it helps to be done early on, but you can't just rely on that forever.

I'm thinking you almost need a team of people monitoring the uncontrolled environment to minimise risk until you have proofed the dog in many environments with many dogs and people.

But that IS a controlled environment You're always moving ahead with CAT, as fast as the dog and trainer are capable. You do go into those situations, and it's a relatively quick process doing it this way. The trainer needs to be prepared for that, btw. Once you've started, you don't stop. I haven't seen the video, I'm not sure how in-depth they go?

I thought there may need to be some type of ongoing training (by handler/owner) to maintain the strength of the alternate responses espec if the dog is inadvertantly exposed to inappropriate dogs to help prevent regression.

Consider that when the dog is responding appropriately to other dogs, most other dogs he meets will also respond appropriately and his new behaviour will work for him - at least most of the time. However, let's not pretend that the sort of dogs who end up needing CAT aren't genetically predisposed to learning aggressive behaviours and experiencing anxiety around other dogs. If they've learned it before, they can learn it again, particularly if you haven't addressed the respondents as well (which is currently an UNKNOWN factor in CAT).

While we're on that topic, this is why I don't use corrections with these dogs, unless I have a good reason (like someone will be hurt if I don't suppress that behaviour promptly).

Something to be very careful of, it should not be assumed that just because we conditioned new operants, or suppressed unwanted aggressive operants, that we made broad-reaching changes to respondent behaviours. We may have, or we may not have. Why do we make the assumption?

Excuse my ignorance but could you please explain 'non functional reinforcer'

You said that this procedure is now popular. Do you know some good trainers in Melbourne that use C.A.T. you would recommend I talk to.

CAT isn't popular within Australia, I'm sorry I don't know of anyone doing it in Melbourne.

Kellie and Jesus differentiate between reinforcers that are currently maintaining the operants you are working with, and reinforcers that have been introduced. I can see their point, but it really is context-specific which can make things confusing.

In the case of a dog who would benefit from CAT with other dogs, the other dog is a functional reinforcer. Let me just clarify because this often trips people up, in a negative reinforcement procedure the aversive is the reinforcer, so if an operant is reinforced when it makes another dog go away, the other dog is a functional reinforcer. The function of the behaviour is to make the other dog go away.

If we come in with some food and reinforce a different behaviour, that reinforcer is functional for the different behaviour, but was not functional for the original behaviour. So if we're referring to the original behaviour, the food is not a functional reinforcer for THAT behaviour. Giving the dog food to reinforce an alternative behaviour does not put the unwanted behaviour that we seek to replace under an extinction procedure in itself, for that matter.

I haven't explained that very well, sorry!

[CliftonPark]

If you don't maintain a change at an operant level, or fail to make a change at a respondent level (and maintain it), what happens when the dog is pushed a little too far and barking/lunging work for him again?

The CAT procedures shown always had placid easy going decoy dogs for obvious reasons but as you state above what happens when the dog is pushed a little too far?

The process shown did not include upping the anti with the type of decoy dogs used and I wonder if the process could be extended to using decoy dogs that are more bouncy, playful types, dogs that bark (friendly/excited barking) dogs that will teach the target dog to cope with a bit more.

The dog is going to encounter this in real life and also needs to be able to cope and realise these are also non threatening behaviours.

Do they incorporate this at all as I didn't see it done or get discussed.

I know you can't proof against everything the dog will encounter but if the goal is to be friendly with other dogs I do think they need to be trained to cope with more than placid dogs. Guess it depends on how far the owner is prepared to go with the training and what they want to achieve.

Consider that when the dog is responding appropriately to other dogs, most other dogs he meets will also respond appropriately and his new behaviour will work for him - at least most of the time. However, let's not pretend that the sort of dogs who end up needing CAT aren't genetically predisposed to learning aggressive behaviours and experiencing anxiety around other dogs. If they've learned it before, they can learn it again, particularly if you haven't addressed the respondents as well (which is currently an UNKNOWN factor in CAT).

Perhaps more than one training method could be employed to help cover different situations in dealing with things in the real world,

like a distration technique 'look at me' if a particularly risky confrontation is anticipated so that they don't just have to rely on the CAT training and have other training tools at their disposal.

Something to be very careful of, it should not be assumed that just because we conditioned new operants, or suppressed unwanted aggressive operants, that we made broad-reaching changes to respondent behaviours. We may have, or we may not have. Why do we make the assumption?

agree

I haven't explained that very well, sorry!

Think you did a great job explaining things actually and I thank you.

Assuming the process is suitable for the dog, do you use the CAT procedure as a preferred method for dog aggression or something else?

[Aidan]

The process shown did not include upping the anti with the type of decoy dogs used and I wonder if the process could be extended to using decoy dogs that are more bouncy, playful types, dogs that bark (friendly/excited barking) dogs that will teach the target dog to cope with a bit more.

That's very interesting, I might ask Kellie about this because I'm not sure how far she goes these days (we haven't discussed it for a while), it wouldn't be fair of me to comment without knowing what she does once she gets past this point.

Perhaps more than one training method could be employed to help cover different situations in dealing with things in the real world,

like a distration technique 'look at me' if a particularly risky confrontation is anticipated so that they don't just have to rely on the CAT training and have other training tools at their disposal.

Kellie makes a fairly strong argument in favour of controlling the environment but has also expressed disappointment at how difficult she has found it to achieve, so I would be interested to hear what she has to say about this now. I believe that if you own the DVD you can join her "5th Quadrant" discussion group.

I caution against the use of distraction techniques except when necessary, and there is a very narrow range of scenarios where they are useful in my work. Either you can control the environment (or consequences) in some small way, or it's completely out of your hands and distractions aren't going to help at all. If you can control the environment even a little, you can reinforce operants. If you can't, then you probably can't distract the dog either. The small margin between those two pages doesn't allow a lot of room for doodling, but if you can avoid a potentially explosive situation by tossing some food around then that is better than losing the plot completely.

People like their margin of safety, particularly clicker trainers who don't feel that they have any options if the dog isn't interested in their food any more. One of the challenges for me is to get people to keep moving forward, to at least approach the knife edge if not walk along it. The trick is to keep the dog in the game.

Distractions work if you use them early enough. The problem with that is that it reinforces that sort of behaviour in the owner (or trainer), not moving forward calmly, competently.

Assuming the process is suitable for the dog, do you use the CAT procedure as a preferred method for dog aggression or something else?

I actually hit upon this idea when I read "How Dogs Learn" by Burch and Bailey. They told the story of the dog who acted up in class, the instructor told them to step outside to settle down, then they found it happening more and more. "Hey - that's what I do!" Just not in class...

Cosmolo and others have said they also use something similar to CAT. I am a strong believer in using some "rough and ready" functional analysis to identify actual reinforcers or punishers, rather than relying on my own interpretations. If nothing else it feels good to know I'm right

In fairness to Kellie and Jesus I would have to say that I do not use CAT, but I do use proximity of stimuli to reinforce/extinguish behavioural responses. I also use a couple of kilos of meat and enough clicking to be attractive to cicadas and crickets. I do have a dog in a class at the moment who I will be using something more along the lines of CAT with though. For the record, I have no trouble finding suitable decoy dogs. They just have to be LESS reactive than the dog we are working with

[Kavik]

This has been a very interesting and useful thread

Thank you Aidan and Cosmolo especially. I hadn't heard of this method before.

My biggest difficulty when dealing with Zoe who is dog aggressive (and to a lesser degree Kaos who is not keen on bouncy dogs in his face) is to be able to progress and improve their behaviour around bouncy, playful, barking dogs once we have mastered calmer quiet dogs.

I wish you guys were closer to me!