As trainers, we talk to dog guardians all the time. They come to us with questions and concerns about their dog and our job is to help them work out the best approach to their canine’s needs. Many have been through training classes and have worked with other trainers before consulting us because they’ve hit a wall. But for the basic dog guardian, the information they receive from various trainers can be very confusing. Why? Because we don’t all speak the same language or apply the same methods. We put ourselves into different categories like traditional trainers, balance trainers, force-free trainers, reinforcement based trainers, etc. Meant to describe different training philosophies, for most people these terms don’t describe much at all. Everyone claims to be an expert with X years of experience, such and such accreditation and blah, blah, blah. But, no matter how we define ourselves, we should all strive to apply basic principles of Efficacy, Efficiency and Ethics in the best interest of the dogs.
When we’re sick we go to the doctor and rely on drugs to treat our medical condition. We trust that the physician is applying the latest science in his field. We also want the drugs he prescribes to make the symptoms disappear as quickly as possible and with little to no side effects. The best drugs are effective but they’re also safe. We might be willing to put up with unpleasant side effects when dealing with a life threatening disease, but we certainly wouldn’t have the same tolerance for side effects when treating a simple headache. When our clients have a behavior concern with their dog, they consult a specialist, a dog trainer or behaviorist and go home with a training or behavior protocol. Just like when we’re sick, we want to make sure the treatment works, but we also want to avoid creating problematic developments from the treatment itself.
How can we make sure we’re making the right choices in our training and behavior protocols that we prescribe to our clients? Three guiding principles should prevail:
Measuring efficacy of a training protocol is evaluating whether or not we have successfully accomplished our goal. Did we successfully teach the dog to sit? Did the behavior protocol stop the dog from barking at visitors? We can teach a dog to sit by applying pressure to the hindquarters or we can lure the dog into position with a treat. Do we accomplish our goal of teaching our dog to sit either way? If we teach a dog to walk on leash without pulling using leash pops or treats, which one of these methods is most likely to produce the desired result? Some trainers are strongly opposed to the use of food that they might consider as bribery. Their belief is that dogs learn better and are more focused when punishment is applied. For others, in contrast, no punishment should ever be used. How do we know which methods to chose? In my years of training I have been from one approach to the other and can speak to the efficacy of the different methods (see “My journey towards becoming a dog trainer and behaviorist“). But more than my experience only, whenever possible, I like to rely on science to determine what works best.
A study of dog training styles on 33 military handlers in Belgium showed that the use of punishment or rewards had a significant impact on the efficacy of the training methods (Haverbeke & al., 2008). The dogs were trained to perform 8 exercises: Heel, Sit, Down, Stand, Positions at a distance, Recall, Down out of view of handler, and Jump. They were also evaluated during 5 protection work activities: Handler’s defense, Attack, Attack with gunshots, Attack with threatening behavior and stand off. The handlers could use any rewards or punishment that they felt effective, from stroking, praising, toys and food, to speaking harshly, leash jerks using a choke or prong collar, hitting or even hanging the dog. The handlers were divided into two groups: a high performance group with good accuracy, control and attention from the dog and a low performance group. Based on performance only, contrary to some military trainer’s beliefs, the dogs receiving the most punishment did not learn as well as the others. They were significantly more likely to be in the low performance group. They were also more distracted and paid less attention to their handler.
Measuring efficacy based on whether or not we have trained a particular behavior is not enough however. There are other effects to consider as well. We should also ask ourselves: will this method help the owner in the long run? In other words, did this method teach the dog behaviors that will sustain over time or is the dog likely to revert back to previous habits? Did learning this protocol provide effective tools to the owner to deal with similar problems? Will the process enhance the relationship between owner and dog? Are the effects of the treatment (positive or negative) also likely to be generalized to other situations?
How much time did it take to get the desired behavior? How many repetitions did it require to get to fluency? How resistant to extinction is the behavior? How likely is the owner to implement the protocol?
Time and repetitions required to achieve a particular behavior are probably the biggest reason why some trainers resort to aversive methods. When we put an electronic collar on a dog to treat barking, we get immediate results. Most dogs will quickly stop barking once barking triggers an electric shock. When using positive reinforcement based methods, we’ll use management, classical and operant conditioning techniques that will require repetition and gradual exposure to the trigger. So in terms of repetitions only, the first method sometimes gives us quicker results. However, what happens when the collar is removed? A month or a year later, how does the dog respond to the stimuli? Just like strong pain killers, we have to make sure the side effects aren’t worse over time than the initial problem. When a drug has the potential to be toxic, doctors will only prescribe it if the situation is serious and there is no safer alternative.
When we swallow a pill, the pill doesn’t only affect the area in our body that needs treatment. The chemicals will disperse in all of our organs and also modify our biochemistry in areas that were doing just fine. Similarly, training and behavior modification protocols affect the dog’s neural pathways and emotional states. So if we increase the dog’s cortisol levels, what other behaviors might come out that we’ll then have to treat? If we use a bark collar, what associations may the dog likely develop that could potentially result in an increase in generalized fear? When measuring efficiency, we have to consider not only the targeted behavior, but also the added work that might result from potential side effects.
Our moral principles dictate how we treat others. As Susan Friedman so rightly pointed out, just because a method works, at least some of the time, doesn’t mean that we should use it (Friedman, 2008). When we have become desensitized to certain training methods, we often fail to recognize the harmful effects they might have. But objective assessment needs to be done, especially when working with beings that cannot express their thoughts and feelings. Our only recourse is to pay attention to their body language, to watch for signs of stress and avoidance. But even then, as Eileen Anderson pointed out in her great blog “But every dog is different!” we can’t necessarily tell how much pain dogs are experiencing when punished.
So how to best prevent harming those we work with? By subscribing to the principle of Least Invasive, Minimally Aversive intervention possible (LIMA). We’re invasive when we physically force our way into the dog’s space and impeding on her ability to move freely. In medicine, an invasive procedure would involve cutting a large incision to access the target of the surgery when we could have accessed the same area with a much smaller incision or even through a natural orifice. An aversive refers to anything a dog would rather avoid if he had a chance. Choosing aversive methods, when results could have been achieved through the use of positive reinforcement based methods, reveals either flaws in the understanding of animal psychology and training or a blatant desire to control others by all means vs establishing a relationship based on knowledge, respect and empathy.
Recently, the Association of Professional Dog Trainers (APDT) partnered with the International Association of Animal Behavior Consultants (IAABC) to develop “best practices” for dog trainers and behavior consultants. Those practices involve the application of the LIMA. Positive reinforcement should be applied as a first line of teaching as it has been shown to produce the least incidences of aggression, attention seeking, avoidance and fear in the animals. Relying on punishment mostly focuses on stopping certain behaviors but never asks the question: what do we want the animal to do? As a result, the dog doesn’t have the opportunity to learn what to do instead of the undesirable behavior, leading to much confusion and frustration in the process.
Ultimately, it’s not about the labels we give ourselves. It’s about making sure we choose methods that are in the best interest of the dogs and their guardians. Science is always progressing. No one would go to a doctor who still believes in the value of bloodletting. Physicians work hard at keeping track of the latest advances in medicine and regularly discuss and redefine the ethics and standards of their profession. Similarly, dog trainers and behaviorists should also strive to keep learning with an open mind for new training methods as well as new technology in their field. The dog training industry is unregulated, but as individuals, we can constantly aim at doing the best we can to maximize Efficacy, Efficiency, and good Ethics.
Jennifer Cattet Ph.D.
Haverbeke, A., Laporte, B., Depiereux, E., Giffroy J.-M., Diederich, C. (2008) Training methods of military dog handlers and their effects on the team’s performances. Applied Animal Behavior Science, 113 (1-3), pp. 110-122.