Dog trainers can, and frequently do, debate which techniques for teaching a dog to do a specific behaviour is the most effective. There is some flexibility in how we teach our dogs to do something new. But what about those cases where our dog is doing something we don’t want them to do? What training methods or techniques are available for dealing with those cases and how do we choose?
I’ve been talking a lot lately in this column about how we go about training our dogs. Does it really matter which approach you use to get the behaviours you want from your dog? My personal experience says it does and I’m not alone in that opinion. Many dog and animal behaviour experts have made the same case – some training methods produce more well adjusted animals than others.
Putting Out A Fire
When something catches fire, we just react. We don’t debate or consider. We just move to put out the fire with what ever comes to hand. For many dog owners, that’s how they deal with problem behaviours as well. When their dog growls at a child or barks unexpectedly at a guest, they just react to stop the behaviour. “Bad Dog!”, is yelled out to interrupt whatever is going on. It can be more or less effective depending on the situation.
The point is, we resort to the most convenient action to stop the behaviour in the moment. We might grab the dog’s collar, give them a smack on the rump or the nose, pull up on the leash, etc. But this isn’t exactly like putting out a fire. In the case of our dog, something caused them to exhibit that behaviour and the correction-of-the-moment might not be effective. Like throwing water on a grease fire in the kitchen, it might make the situation worse.
Proactive Protocols
Just like dealing with fires, there are different ways to approach dealing with bad behaviours in our dogs. It might be worth considering a few different techniques instead of relying on our one-size-fits-all reflexive action to Fido’s unwanted behaviour. There are some systematic principles all of us learn about fires – smothering with blankets, baking soda for grease fires, soaking surrounding areas to prevent the fire spreading, etc. – there are also some protocols we can borrow from another well-known field. Applied Behaviour Analysis offers us some useful approaches for dealing with problem behaviour.
Applied Behavioural Analysis is being used today in programs for the mentally challenged, autism disorders, and even in mainstream primary and secondary schools. The goal of such programs is to provide effective methods of dealing with problem behaviours to allow learning to take place. It has been remarkably successful. So it should not be surprising that positive training techniques like Mark & Reward or Clicker Training for our dogs are based on these same Behaviour Analysis principles. And they can be equally successful at dealing with problem behaviours as they are in teaching new behaviours to our dogs.
Friedman’s Hierarchy
Dr. Susan Friedman is a psychology professor at Utah State University who has pioneered the application of Applied Behavior Analysis (ABA) to captive and companion animals. In an article for the Association of Pet Dog Trainers (APDT) Journal in 2010, Dr. Friedman discusses the intrusiveness of various training and behaviour management methods. She cites research done in Behaviour Analysis that shows that the least intrusive procedures are the most effective for promoting learning.
The principles on which she bases her article refer to research done in human schools and a “hierarchy of intrusiveness” presented by Paul Alberto and Anne Troutman in their book Applied Behaviour Analysis for Teachers. Dr. Friedman expands this hierarchy into an effective model for dealing with problem behaviour in our dogs. Read from top to bottom, this diagram represents a series of approaches to problem behaviours from first resort to last resort.
Stepping Through The Hierarchy
Unwanted or inappropriate behaviours will happen in our lives with our dogs. Some of those will occur because the dog does not know any better or has not been taught alternative behaviours. Other times inappropriate behaviours can show up almost mysteriously. Friedman’s hierarchy provides a simple, easy-to-use protocol for dealing with unwanted behaviours from growling to pulling on leash to barking.
Medical, Nutritional, Physical – Many inappropriate behaviours our dogs exhibit can be the result of real physical problems. A dog in pain or physical distress may be much more likely to show aggressive or defensive behaviour. They are ill and wish to be left alone or at least not be touched in certain areas. Before taking any other action to correct behaviour, first make sure there isn’t any medical or physical cause that may be triggering it. A trip to your veterinarian or health care provider should quickly determine if this is the cause of the dog’s behaviour.
Antecedent Arrangements – This is just fancy language for the “situation” the dog is in when it exhibits an unwanted behaviour. My dog could be frightened of men in hats or small children or even loud noises. Under those circumstances, she might be more likely to display aggressive or other inappropriate behaviour out of fear. So the next step for me under this hierarchy is to look at what I can change in my dog’s environment to change her behaviour. Can I put more distance between us and the disturbing object? If being too near activity causes my dog to be over excited, perhaps I can move back to a safer distance. Similarly, if my dog is annoying me by stealing food off of our counters, well, I guess I should stop leaving food on the counters!
Positive Reinforcement – Rather than allowing my dog to decide what she would like to do in a situation where an unwanted behaviour develops, can I teach her to perform a behaviour of my choosing using positive reinforcement? Ideally this is a proactive step taken when the dog first encounters potentially troublesome situations. For example, if my dog exhibits a preference to nip at the pant legs of family members as a puppy, I can use positive reinforcement to teach her an alternate behaviour such as picking up a toy which would be more acceptable to everyone.
Differential Reinforcement of Alternate Behaviours – This is similar to Positive Reinforcement above except that I may be working with an already established unwanted behaviour. The concept of Differential Reinforcement means that I will reinforce my dog for an alternative behaviour at a higher level than she is reinforced for the unwanted behaviour. At that point, the choice should become obvious for the dog. If my dog barks out the window when she sees the mail carrier, I could train her to come to me for a treat when she sees them. In this case, the mail carrier becomes a signal that coming for a treat will pay off. In a short time, my dog will become more reinforced by coming to me after seeing the mail carrier rather than standing at the window and barking.
Extinction, Negative Punishment, Negative Reinforcement – Each of these approaches to behaviour change has a fairly high degree of complexity to implement correctly. Extinction involves providing no reinforcement whatsoever for the unwanted behaviour. This can be difficult especially if the behaviour is caused by fear and the behaviour itself provides some level of stress relief. That said, if the dog does not have the opportunity to engage in the unwanted behaviour (i.e., they are not exposed to the thing that triggers the behaviour) it is possible that the behaviour will become extinct. Both Negative Reinforcement and Negative Punishment involve the removal of something in order to change behaviour. Negative Reinforcement seeks to increase a preferred behaviour by removing something unwanted by the dog when the desire behaviour is performed. Negative Punishment targets the unwanted and seeks to decrease the target behaviour by removing something the dog wants or values – food, companionship, security, etc.
Positive Punishment – This approach should be familiar. It involves targeting the unwanted behaviour by adding something the dog doesn’t want or like when the unwanted behaviour occurs. This could be a tug on the leash, physically pushing or hitting the dog, yelling at the dog, or making threatening gestures to get them to stop doing the unwanted behaviour. In Friedman’s hierarchy, this type of behaviour modification is the most intrusive and therefore the least ideal for teaching the dog. It may learn not to do the behaviour, but it may be learning to stop the behaviour as a way of avoiding the punishment and nothing more. If we are not able to be there to deliver the punishment, there is nothing keeping the dog from doing the unwanted behaviour when we are not around.
First Response and Last Resort
Like putting out a fire, often a dog owner’s first response to a problem behaviour is to stop that behaviour as quickly as possible. All too frequently the method used is Positive Punishment, Dr. Friedman’s last resort in her intrusiveness hierarchy. It may not be the best choice for solving the behaviour problem long term but in the immediate situation, it will very often produce the desired effect. Startling, surprising, or otherwise interrupting the dog in the act of an unwanted behaviour can give the illusion that we have “corrected” the behaviour. But all we have done is deter the outward behaviour without giving the dog any instruction as to what other behaviour we would prefer. We have done nothing to improve the dog’s situation. And by introducing punishment, we may even add to their stress or arousal.
When we were young, our first response to our clothing catching fire might be to run and hope the wind would put it out. But we are quickly taught that the best procedure in this case is to stop, drop to the ground, and roll to put out the flames. It is a more involved solution that might not come to mind if we are not taught the effectiveness of doing it that way. Similarly, I think most dog owners are prone to reacting without much thought as to what the most effective approach would be for an unwanted behaviour. If dog owners understood Dr. Friedman’s hierarchy, they might be able to quickly employ a more effective and longer term solution to behaviour problems.
What Else Could I Do?
All too often dog owners and dog trainers alike will use the reasoning that something needs to be done and it needs to be done fast to stop unwanted behaviour. Unfortunately that “something” is usually the punishment that Dr. Friedman reserves as a last resort. When a dog owner asks me, “What else can I do?” in defence of using punishment to stop unwanted behaviour, I know from the basics of Applied Behaviour Analysis that there is PLENTY that can be tried before resorting to punishment.
Dr. Friedman’s hierarchy provides an excellent, easy-to-understand framework for solving behaviour problems. Although many of these steps are not as dramatic as the most invasive punishment-based solution, they employ techniques that have been proven to produce longer lasting and more effective behaviour changes in our dogs. So although poking, startling, and kicking at a dog might make for good TV, it doesn’t necessarily make for good solutions to behaviour problems.
Using the wrong training method on problem behaviours can be like throwing gasoline on a fire. Using physical force on an unwell dog is a good way to get injured. So, like the “stop, drop, and roll” we were taught in response to fires, Dr. Friedman’s hierarchy gives us guidance for handling problem behaviours without the risk of creating bigger ones. And isn’t it better to be proactive, rather than reactive, when it comes to our dog’s welfare?
Until next time, have fun with your dogs.
Photo credits –
Intense – colemama 2008 from Flickr
The Stare – Lon Fong 2010 from Flickr
Feature banner – theoelliot 2009 from Flickr
Deb says
When I saw the title and started reading I thought – a ha, finally some answers. Unfortunately for me and my dog I didn’t find any. I know what not to do but still haven’t found out what to do in the case of serious unwanted behaviour. Maybe I should be reading between the lines?
Eric Brad says
Hi Deb –
I’ve never been a believer in 0ne-size-fits-all solutions and that’s probably why you didn’t find recipes to fix specific problem behaviours. Instead, I like to point people at tools they can use in their own situations to deal with problem behaviours in a way that works for them and their dog. This particular article defines an approach to assessing and coming up with a plan for changing problem behaviours. Let me give you an example.
If someone came to me with a dog that was growling at children, my first question would be is this a new behaviour that has started recently? If so, it could be that the dog is not feeling well or is in pain from an injury that is not obvious. If that were the case, the dog would not want to be touched by unpredictable strangers. So I would want to rule out any health issues first. If it’s not a new behaviour, we have to look elsewhere for a solution.
Next I would look at the life situation of the dog. Will the dog be required to be around children in daily life? If not, is there a way to arrange schedules, situations, etc. in a way to keep the dog separate from children? If we can’t arrange things to be sure the dog will not be faced with situations where the unwanted behaviour happens, then we will have to look at behaviour modification alternatives.
Following the hierarchy in the article, is there a way to use positive reinforcement to encourage a behaviour that we would prefer? Could we mark and reward first looking at the children and then later moving closer to them and finally taking treats directly from them. Can we work with the dog and children in a way that is safe and low stress for both?
If we can’t find a way to reinforce positive interaction with the children, can we use differential reinforcement to create an alternate behaviour when the dog becomes concerned about the presence of children? Perhaps we can reinforce the dog for sitting or lying quietly next to the handler using the stimulus of the children as the cue. In other words, “When they stress you out, instead of growling come over and sit by me.”
Only after trying those four steps above would I look at the last two alternatives. Using negative punishment might mean that when the dog growls at a child, we take him way and put them in a crate or room by themselves for a “time out.” We’re hoping to reduce the growling behaviour by taking away something the dog wants, our company. I’m reluctant to recommend this approach because I would have a concern that if I used it too frequently, the dog might come to see the presence of children as a predictor of being sent to a “time out” and might escalate the behaviour to barking or worse.
The last resort of using positive punishment might be to use a leash correction or verbal correction to stop the dog from growling. While this might at first seem like the most expedient way to get the desired result, we really haven’t addressed the cause of the behaviour. We have just tried to suppress the outward expression of the dog’s discomfort. By reducing the growling but not working on the underlying discomfort, we are just creating a shorter fuse in the dog. They may appear to tolerate a closer presence of children but react just as strongly or more strongly when their tolerance limit is reached.
As I said, this provides an overall guideline for approaching problem behaviours. If you find you can’t work out how to address specific problems from this approach, I can recommend two great books.
How To Right A Dog Gone Wrong – Pamela Dennison
Click To Calm – Emma Parsons
Both are available at Amazon.
Hope that helps and thanks for reading!
Eric
Me says
If a person does the first five principals (minus the negative reinforcement) correctly, then it renders the sixth principal and also negative reinforcement unnecessary, and therefore both cannot be justified.
I love your articles, by the way. 🙂 Very coherent and thought provoking.
Kim says
Thanks for putting this out there so clearly. I still need to be reminded of this regularly because it’s too easy to slip back into old habits.
Eric Brad says
Thanks Kim. We won’t always be perfect in our responses but if we take time to think about this stuff once in a while, we’re going to get better and better.
Dana says
Thanks Eric! Another great article I can share!
I think we have all had those moments of frustration where we have just wanted to STOP a behaviour, and we react without thinking….
You have clearly explained some of the things we need to consider when we are dealing with some form of unwanted behaviour (it’s unfortunate that most of the ‘bad behaviour’ is caused by us in the first place!)
Cheers
Eric Brad says
Thanks Dana! Glad you enjoyed the article. And I agree with you. I’m sure a lot of our dog’s “bad behaviour” comes from the confusing ways we handlers deal with them!
Tanya says
Great article, Eric!
It is nice to have this so clearly laid out It is human nature to “respond”. I tend to the knee jerk reaction to barking with a “knock it off!” which, by the way, doesn’t work at all.
The average pet owner may not consider the medical or physical component since their dogs just do stuff to make them mad.
Thanks for clearly explaining all thee steps and things to consider when dealing with unwanted behaviors.
Eric Brad says
Thanks Tanya! Glad you enjoyed it.
Eric