Tag: positive reinforcement training

Aversive to Force Free Dog Training: A Conversation with Stephanie Rombough

Jennifer Berg CPDT-KA, CDBC
Woman in red jacket smiling at brown dog; both are sitting on top of cliff with mountains in view.

Stephanie with Neirah, a Corgi/Heeler mix, on a summit hike called Sulphur Skyline in Jasper National Park. (Photo credit Lars Bjornstad)

I’ve interviewed several dog trainers over the years about their decision to move away from aversive methods that use force, pressure, corrections to focus on positive reinforcement methods. Some common experiences of the trainers I interviewed include:

  • their initial underlying motivation to make the change was based on an emotional connection with a dog;
  • knowledgeable instruction and support from skilled positive force-free trainers was important to their success;
  • they faced significant negatively from others, including aversive trainers in their local community and online;
  • they felt unpleasant emotions as they came to terms with the negative experiences of the dogs they trained previously, but this was off-set by how much better they felt training without aversive methods/tools and how much happier their dogs seemed about training.

I’d also like to point out that not one of the trainers I interviewed said that they were unable to train dogs to the same level of proficiency as previously, and some mentioned that once their new dog training skills were developed, the dogs seemed to learn more quickly. This was glaringly apparent in Sue Ailsby’s experience when she tells of teaching her dog to do the broad jump in a matter of minutes using clicker training, while others in her kennel club had been trying for over a year to teach their dogs to do the broad jump using aversive methods. You can read about Sue Ailsby’s crossover journey “Responding to Change: One Trainer’s Journey From Compulsion to Positive Reinforcement” in the Spring 2022 issue of APDT Chronicle of the Dog, pages 44 – 49.

Most recently, I interviewed Stephanie Rombough, a force-free trainer from Edmonton, Canada. You may have seen her dog training videos on social media. Her instructional videos are clear, concise, and easy to implement, and I have several of her videos bookmarked so I can easily share with dog owners looking for help. Stephanie, like myself, had a parent who was a compulsion-based dog trainer and she learned how to train dogs and was a firm believer that dogs needed to be trained this way. Until she met a very special dog named Burrito. You can read the interview in Stephanie’s blog post

Dog Training: Management vs Teaching

Jennifer Berg CPDT-KA, CDBC
Tan coloured dog looking up at hand holding dog leash
Photo by Ruby Schmank on Unsplash

Some people, especially “balanced trainers”, will insist that punishment-based training methods are necessary sometimes, often equating “emergency management” with “teaching”, using the example of yanking on the leash if a dog is about to run into traffic. I strongly argue that “management” is not the same as “teaching.”

Humane and effective management to prevent unwanted behaviours is an important part of a good training plan, but management is not the same as teaching.

If a dog is about do a behaviour that is dangerous, such as run into traffic or bite someone, and the person holding the leash pulls back on it to prevent the dog from doing the dangerous behaviour, I consider the actions of the person as “emergency management” and not “teaching.”

Setting up a dog to fail is not ethical, humane, or effective training

If someone is training in a way that puts a dog into this situation — on purpose or because of incompetence — then this trainer needs to reconsider how they are training. Setting a dog up to fail and then be corrected is less effective than using positive reinforcement training, and it has a high risk of negative fallout. A person who can recognize subtle canine stress signals is better able to remove a dog from the situation before the dog escalates to extreme and dangerous behaviours. 

Dog Training Fallacies

The use of aversive methods and equipment is often claimed to be necessary because of [insert emergency situation here]. The “emergency measures” defence is a common argument that promoters of aversive training methods (including “balanced trainers”) use to defend their actions. I have encountered this false argument many times over the years from people who insist that the use of coercion, force, and aversive tools is needed to effectively train dogs.

Often the person claims that the aversive method or tool is necessary to save the dog’s life (e.g., to prevent dangerous behaviours or behavioural euthanasia). The argument is usually framed in a way to suggest that anyone who is opposed to aversive methods or tools in dog training would rather see the dog face death (a straw man fallacy). But this argumentative defence relies on false claims. One popular false idea is that dogs that have been euthanized due to severe behavioural problems could have avoided this fate if they had only been trained with aversive methods (shock collars, for example); this is unfounded and is in direct opposition to the massive efforts force-free trainers have been doing to successfully rehabilitate dogs that have suffered from the fallout of aversive training methods.

It is also incorrect to assume that aversive methods and tools are more effective in addressing severe behaviours that would lead to a dog facing behavioural euthanasia. The fact is, many highly regarded organizations, including the AVSAB (American Veterinary Society of Animal Behavior) and the BC SPCA, have explicitly stated that that aversive training methods are not necessary and should be avoided, the main reasons being efficacy in training, safety, and the welfare of the dog. 

Another popular example that proponents of aversive methods use is their insistence that the only way to prevent a dog from performing dangerous behaviours such as chasing cats or wildlife is to use aversive methods or tools. This is absolutely false. See Predation Substitute Training for an example of how to humanely train a dog not to chase wildlife.

People all over the world are successfully training dogs to high levels of performance without resorting to shock collars and other assorted aversive tools and methods that rely on discomfort, coercion, and fear. It’s a choice. When people know better, they do better.

Can E-collars Be Used for Positive Reinforcement Training?

Jennifer Berg CPDT-KA, CDBC

Is it possible?

Can an e-collar be used for positive reinforcement training?

Is it possible to train with an e-collar under these conditions?

  1. the dog is conditioned to have a positive emotional response to the collar (including the sight of the collar and the wearing of the collar even if it’s not activated)
  2. the collar is never used as an interruptor, a warning, a “correction” or to suppress a behaviour (e.g., giving the dog a “stim” as a consequence to doing an undesirable behaviour or stopping the “stim” as a consequence of doing the desired behaviour)
  3. the collar never causes any level of discomfort to the dog. Ever. (The dog gets to decide if the sensation is uncomfortable — physically and emotionally.)

Let’s unpack that.

Conditioning a dog to have a positive emotional response to the e-collar requires several components. Firstly, the collar must be set to deliver a sound or vibration so the dog does not experience any physical or emotional discomfort. The dog is the one who decides if the sound or sensation causes discomfort or anxiety. It doesn’t matter if the human doesn’t think the experience is uncomfortable or worrisome; what matters is the dog’s perspective.

I am not promoting the use of e-collars. This post is to reveal some insights into the claim that some trainers make that e-collars can be used with positive reinforcement.

Determining if a dog is experiencing physical or emotional distress can be very difficult, and in many cases very experienced trainers often miss the signs, even with their own dogs.*  The human MUST be extremely good at reading the very subtle changes in the dog’s body language (e.g. nose licks, yawning, look aways, tightness around the mouth, dilated pupils, slowing down of movement). Dogs have been known to hide their pain, and dogs can show pain or emotional distress in ways that humans misinterpret as “obedience” or “good manners.” For example, dogs that are stressed can shut down emotionally and even “freeze” (Fight/Freeze/Flight response), and dogs in pain can appear more quiet, calm, or still, and some may appear more alert.

If the dog’s experience of the stimulus of the e-collar (the sound or vibration) is not negative in any way, then pairing it with a reinforcer like food (or a conditioned reinforcer) can condition the dog to have a positive emotional response to it. The sound/vibration would then be used as a marker for a desired behaviour, and with correct timing of the human with the remote, will communicate to the dog that the behaviour “correct.”

BUT, if the e-collar’s sound/vibration is later used in a way or in a context that causes the dog any physical or emotional discomfort, the dog’s positive emotional response will quickly degrade. Even after one experience. Instead, the dog will develop a negative emotional response to the sound/vibration (and possibly to the sight of the collar), as well as any other associations the dog may have connected to the experience, including the location, the trainer, the activity.

Cognitive Dissonance Can Make Us Blind to Logic

I just finished listening to a podcast where the two hosts discuss how e-collars work. They clearly state that the device causes enough discomfort to cause the dog to stop a behaviour (or to avoid doing a behaviour) because that’s how aversive tools work, but they insist that they are not causing their dogs physical pain when they use it. Huh? Their defence is that they (the hosts) don’t register it as painful when they give themselves a “stim,” and they don’t see any evidence in their dog’s behaviour to indicate it’s painful.  (Emotional distress is not acknowledged as a consideration. And I would argue that the fact that a behaviour is being suppressed after the delivery of the “stim” is clear evidence that the sensation caused discomfort; that’s how it works: the sensation has caused the dog to stop the behaviour at that moment. Why? Because the dog wanted the sensation to stop. Why? Because it did not feel good.)

The hosts spend a fair amount of time pointing out that one person’s experience of the “stim” of an e-collar can be very different from another person’s experience of the same level of intensity. They acknowledge that the anticipation that the device will cause pain: “An e-collar has the value you give it. If you go into it thinking that the e-collar is a pain tool, it’s gonna deliver you pain no matter what comes out of it. Even if it’s nothing.” Now read that substituting the word “dog” for the word “you.” They recognize from the human end that the anticipation of pain/discomfort can result in actually feeling it, but they do not acknowledge that this can happen from the dog’s viewpoint, as well.  (Note: the anticipation of pain/discomfort is also called “fear“.)

As studies on dogs advance, especially with advances in neuroimaging and functional MRIs, perhaps we’ll have more science-based information on how dogs experience emotional and physical “pressure” when aversive training methods and tools are used, rather than relying on the opinions of dog trainers who insist the dogs are not feeling any physical or emotional distress. But ethics are unlikely to allow this kind of experiment. Why? Because the aversive tools and methods cause the animal emotional and physical discomfort. But wouldn’t the results be worth it — to prove to aversive trainers and advocates of aversive methods and tools? It’s unlikely to change their minds because of cognitive dissonance.

*I remember seeing a video a “balanced” trainer posted online trying to demonstrate how gentle a prong collar was to a dog. (This trainer also promoted the use of shock collars, but not in this particular video. The video isn’t up anymore.) It was obvious to me the dog was showing signs of stress (head hanging low, appeasement behaviours). This same trainer attended a dog class with me — her dog was in a flat collar for the class — and I noticed her dog never really looked happy in the class. Except when the owner asked me to hold onto her dog’s leash while she left the area to use the washroom. That’s when the dog visibly relaxed. My heart broke for this dog and I wondered if the owner even realized how her aversive methods affected her dog, even when the dog was not wearing an aversive collar. I’m sure she loves her dog and would find it very distressing to know that her aversive methods could be causing her dog distress and ruining their bond and relationship. 

The average dog owner is bound to make a lot of training mistakes, even under the guidance of the best dog trainer in the world; this is a big reason why e-collars should not be sold without a licence, but that is a topic for another day.

Why do some people continue to use an e-collar, even after all the evidence that proves they cause discomfort? This leads to another related topic: how using punishment can be reinforcing to the punisher.