Use of Aversives in Dog Training

My last post included this, but in case the posers of the Internets block that one...here is the paper again, without links to Positively.com

"AVERSIVE TRAINING AS CRUELTY: FORENSIC BEHAVIOR EVALUATION IN DETERMINING EFFECTS OF +P (AND –R) TRAINING IN DOGS."

FORENSIC APPLIED ANIMAL BEHAVIOR

ESSAY FOR DR. R. LOCKWOOD AND DR. P. REID (12/7/2014)

James W. Crosby

INTRODUCTION

Dog training methods and the question of animal welfare have become issues of major concern for pet owners and Applied Animal Behaviorists. The presence of trainers with marketable programs and the questions-and problems-that these trainers can raise has brought companion pet behavior widely into the public eye. The plethora of methods and the dissemination of non-scientific information over social media has increased the need for scientific evaluation of training methods and the effect of these methods on the welfare of our canine companions. Owners and Behaviorists are both concerned with the welfare of their charges, but there must be consideration for the intended effect of training. The focus of both groups ultimately is behavioral change. The owners want healthy, compliant companions. Are some methods better than others? Are particular training methods effective and humane?

Applied Animal Behaviorists bring the ability to apply scientific scrutiny to claims and methods. Scientific methods allow the assessment and quantification of the efficacy of training methods, and a better ability to establish the line between humane and cruel training tools.

THE DIVIDE

A large divide among trainers and owners currently involves the differentiation between positive and aversive training methods. Are dogs to be treated as sentient beings that deserve to be treated as companions and partners to their human caretakers, or are they unformed, irascible clay in need of forceful shaping to obey. Which method-if either-will reliably produce results, and which method is humane? Which method(s) produce well adjusted, physically and emotionally healthy individuals?

To begin to examine this question we need to define some terms. First, what is behavior?

Behavior can easily be defined as simply whatever an organism does. Eat, sleep, feed, move-behavior is simply what the organism is doing at any particular moment. Behavior has no intrinsic moral value. Good and bad are human values overlaid on behavior. Behavior is either productive or non-productive. Productive behavior results in the organism being able to achieve a goal. Non-productive behavior fails to achieve a goal. Productive behaviors are reinforced and, as such, remain. Non-productive behaviors tend not to be reinforced and fade.  Additionally, behavior can be considered as normal or abnormal. Normal behavior allows an organism to meet its needs.

In companion canines, owners typically seek to produce behavior that allows them to coexist at a level that the human is willing to at least tolerate, if not actually solicit. Behavior training is the process of teaching the dog the difference between appropriate or desired behavior and behaviors that are not wanted. This can be for specific skills such as sit on command, fetch an object, or walking calmly on a leash.

Dedalle (2103) explains the different methods of behavior modification and training: “(the method) can vary by their nature and occurrence: they can be the appearance or disappearance of appetent or aversive stimuli. It follows that there are 4 types of instrumental conditioning procedures: 2 types result in an increase in the rate of responses (positive reinforcement, R+: appearance of an appetent stimulus; negative reinforcement, R−: disappearance of an aversive stimulus) and 2 types result in a decrease (positive punishment, P+: appearance of an aversive stimulus; negative punishment, P−: disappearance of an appetent stimulus.”

This give us two methods of training: reinforcement based, or positive methods, and aversive or punishment based methods. Both of these methods can be used to either produce a desired behavior or to reduce problematic behavior.

To adequately examine these methods we need to define the terms. A reinforcer is defined as “An event that, when made contingent on a behavior, increases or maintains the frequency of that behavior.” (Chance, 1998). Similarly, a punisher is “An event that, when made contingent on a behavior, decreases the frequency of that behavior.” (Chance, 1998).

Hetts (1999) provides a clear definition of positive reinforcement. “This outcome can be defined as the occurrence or presentation of something pleasant immediately following a behavior that will make that behavior more likely to occur in the future.” This is most commonly seen in the use of food (a primary reinforcer) to reinforce a behavior. A dog is cues to sit. The dog sits. A treat is presented immediately upon producing the sit. The sit, if coupled often enough with the reinforcer, becomes a more likely behavior upon the presentation of the cue.

In training canines there are a number of commonly used positive reinforcers:

Food or treats	Sniffing
Physical touch	Dog-dog social contact
Direct human attention	Verbal marker
Toys	Complex activity (problem solving)
Play	Learning tasks
Free activity or exercise

An alternative to positive reinforcement is negative reinforcement. This involves the deprivation of factors that the organism either desires or needs. Negative reinforcement ranges from the fairly benign such as removing attention from a dog that exhibits unwanted behavior or social isolation (a “time out”) to deprivation of essential needs, such as the food deprivation regularly practiced in the training of marine mammals.

Punishment based training, on the other hand, is centered upon the idea that an organism will act to avoid pain or unpleasant circumstances. An aversive stimulus is presented to the organism that results in the reduction of the associated behavior. The animal learns that the specific aversive stimulus is coupled with the particular behavior. The targeted behavior becomes unsuccessful to the organism, and the organism learns to avoid the behavior.

Punishment can also be used to effect non-compliance to a cued behavior. The dog is cued to perform a behavior (that it hopefully understands). The dog fails to comply, and as a result a punisher is applied. The dog learns to comply to avoid the aversive applied.

Common punishers used in dog training include:

Verbal command (NO!)	Pain (prong collar)
Loud noise	Electric shock
Strike	Physical force
Choke	“Alpha Roll”

Punishment has a substantial chance of unintended side effects. Trust and a positive relationship is threatened by such methods. According to Overall (2007), “Such tools ‘work’ by engendering fear, pain, and distrust, and in doing so they cause long-term damage that make dogs more reactive, less trusting, and less able to reach their full potential in their partnership with humans, no matter what form that partnership takes. These are not my opinions: these are the findings from the scientific literature, and this is an essential point.”

Other problem issues arise in conjunction with the application of punishment to achieve a behavior. In dogs, aggression directed towards another dog or directed towards a human is particularly concerning.

“While on the issue of dog training, one of the most practically significant findings found in this research has to do with the effect that the type of training has on a dog's risk of aggression. There have been a number of studies that have reported that training procedures based on punishment can have negative consequences (Coren 2012). In this study the researchers defined such punitive training techniques as including things like physical punishment (hitting the dog), verbal punishment (shouting), electrical or citronella collars, choke chains and jerking on the leash, prong collars, water pistols, electric fences and so forth. Such punitive techniques apparently increase the risk of aggression in dogs. They are associated with a 2.9 times increased risk of aggression to family members, and a 2.2 times increased risk of aggression to unfamiliar people outside of the household.” (Coren 2014)

Side effects of punishment may include physical injury (Becker 2014).  Whereas positive reinforcement also acts to allow an organism to achieve its needs, punishment may interfere with the ability of an organism to meet basic needs, at least on a short term basis.

In dog training there are currently two main forms of punishment used; electronic shock and physical punishment by use of prong collars and choke chain collars. Both of these methods are based on pain compliance and physical threat.

SPECIFIC PUNISHMENT CONCERNS: PHYSICAL PUNISHMENT

The practice of hanging or choking a dog as a normal training method in the US is common, particularly in police and protection training circles has been long accepted. Many police canine training groups advocate the use of hanging as an effective tool if a dog refuses to disengage with a target. Additionally, the method is also taught and recommended to meet other training goals. The Washington State Patrol Detection Canine Handler Course, in their current Instructor Guide teaches the use of “Choke-Off” as a valid and recognized training practice. To quote their guide “These methods are not designed to teach a release command; rather they are designed to encourage the canine to retain the reward object.” (Washington State Patrol Detection Canine Handler Course)

The effects of using choke chain or prong collars may be immediately life threatening. Dr. Karen Becker (2014) illustrates one possible outcome from the practice of “hanging” a dog to correct problematic behavior in training.

“The owners of a 1-year-old German Shepherd dog brought their pet into a veterinary clinic for incoordination (loss of muscle coordination) and circling to the left behavior. The owners were honest in admitting the dog had been "disciplined" a few hours earlier by being suspended off the ground with a choke collar for almost a full minute. When the owner lowered the dog to the ground, the poor animal was panicked and soon lost consciousness.

At the veterinary hospital, a neurologic examination uncovered severe disorientation and left-sided pleurothotonus, a rare disorder in which there is prolonged and repetitive involuntary contraction of muscles resulting in jerking, twisting and abnormal posturing. Reflexes were reduced in all the dog's limbs, and he was blind. He was also suffering from nystagmus (involuntary eye movements) and paralysis in the left side of his face.

The dog's symptoms indicated a multifocal brain injury, and an MRI showed severe brain swelling due to the prolonged lack of blood flow to the head.

The final diagnosis was strangulation. Due to the extent of the injuries, the dog was euthanized.”

Negative effects of strangulation and “choking out” by use of a chain type collar extend beyond the physical injuries. Emotional damage due to the extreme stress is clear.

“Because oxygen deprivation is a survival threat of the highest urgency, however, the homeostasis-restoration process is not limited to a purely physical response, but also utilizes very strong emotions such as panic and terror. This is why humans-and by all evidence animals-that may be trapped underwater and running out of breath are infused with extremely intense fear and panic, which compels immediate and powerful corrective action.” (McMillan, 2005)

Yet these extreme physical methods are not necessary, even in the specific, high stress environment of police work. In contrast, the Metropolitan London (England) Police Force does not permit the use of choke or prong collars. During my visit to and work with the Met Canine Unit during June, 2014, their policy was explained and illustrated. The Met handlers will, in fact, be subject to termination if they are found to be practicing the use of such tools. Despite these limitations the Canine Unit of the Met is consistently ranked as one of the top performing police canine units in the world.

SPECIFIC PUNISHMENT CONCERNS: ELECTRONIC COLLARS

Electronic collars are designed to produce a measured electrical pulse that is delivered to the dog through electrodes placed tightly against the skin of the dog’s neck on the inside surface of a collar. The electrical impulses are controlled by one of two methods; automatically by sensors that are either triggered by proximity to a boundary (usually established by a wire that emits a short range radio signal) or, in the case of bark prevention collars, by the vibration of a third sensor probe that contacts the dog’s throat near the vocal cords, or by a human-controlled remote transmitter. Both methods usually provide a means of adjusting the level of shock given. In the automatic modes the shock usually escalates based on time, proximity (in the case of electrical fences), and sometimes on the number of “offenses”. As an example of the latter, an electronic anti-bark collar manufactured by Innotek Inc. starts with a warning beep when the dog barks. If the dog barks again within eight (8) seconds of the beep the collar automatically administers an increasing level of shock until the dog stops barking. (Innotek manufacturers’ instructional materials).

On remote training collars directly controlled by the owners the level of stimulus is directly set and adjustable by the handler. The stimulus can also be selected for a momentary impulse (less than 0.5 seconds) or a continuous shock. The continuous mode, in most manufacturers’ models, has an override that stops the impulse after about eight (8) seconds of continuous application. (Innotek manufacturer’s instructional material, Dobbs 1993, Polsky 1994).

WELFARE

Meeting, or failing to meet, a dog’s basic needs can be summarized by the Five Freedoms established by Brambell. To revisit, they are summarized as:

1. Freedom from hunger or thirst by ready access to fresh water and a diet to maintain full health and vigour
2. Freedom from discomfort by providing an appropriate environment including shelter and a comfortable resting area
3. Freedom from pain, injury or disease by prevention or rapid diagnosis and treatment
4. Freedom to express (most) normal behaviour by providing sufficient space, proper facilities and company of the animal's own kind
5. Freedom from fear and distress by ensuring conditions and treatment which avoid mental suffering

Applying the five freedoms to the practicality of dog training, we must look at what our goal is and whether the method of reaching it falls within these guidelines. Our goal (the stated purpose of training) is to elicit or change behavior. Our responsibility is to assess whether our training methods fall within the boundaries of the Five Freedoms, and thereby within the parameters of positive animal welfare.

To change behavior and to assess change of behavior requires description of the behavior and quantification of the change from baseline. To do this the Applied Animal Behaviorist needs to define the behaviors desired. To evaluate whether a method used is humane also requires quantification. The observation that a behavior occurs is not enough to assess the appropriateness of the training method. The overall status and condition of the dog, and the dog’s interactions with humans and other dogs help establish whether the method is humane or not. Measures have been developed that allow the AAB to assess whether the dog’s behavior is normal or abnormal. These measures include:

POSITIVE MEASURES	NEGATIVE MEASURES
Willing human contact	Avoidance/flight
Acceptance gestures	Fear indicators/body language
Positive body language	Warning vocalization
Gaze at owner	Brevity or absence of warning in threatened situations
	Bite
	Negative affective state (shutdown)

Illustrations of the presence of negative behavioral measures resulting from highly aversive training are documented. The use of electrical shock appears to carry more negative behavioral consequences than harsh physical training. For example, during free walking training, police service German shepherd dogs that previously wore a shock collar showed lower ear posture and more stress-related behaviors (lowering of body posture; high pitched yelps, barks, and squeals; avoidance; redirection aggression; tongue flicking) than dogs who never received collar shocks, although they were also trained with harsh methods (Schilder and Van der Borg, 2004).

Chemical support for the observed negative impact of aversive training methods, particularly electric shock, have also been established using cortisol level studies. Beerda et al. (1998) also showed behavioral and cortisol effects on laboratory dogs that were administrated 6 different unpleasant stimuli; “the findings suggest that stimuli like shocks or sound blasts may have been particularly stressful to the dogs because they were associated with a very low posture and an elevated level of cortisol.”

Predictability of a punisher in a dog accustomed to that means of training affects the biochemical status of the dog, and thereby affects the dog’s stress level. In Schalke (2006), the following observations were made: “Three experimental groups were used. Group A (Aversion) received the electric shock when the dogs touched the prey—a rabbit dummy fixed to a motion device. Group H (Here) received the electric shock when they did not obey a previously trained recall command during hunting. Animals of group R (Random) received the electric shock arbitrarily, i.e. the shock was administered unpredictably and out of context…

Group A did not show a significant rise in salivary cortisol levels, while group R and group H did show a significant rise. When the animals were reintroduced to the research area after 4 weeks, the results remained the same.

This led to the conclusion that animals, which were able to clearly associate the electric stimulus with their action, i.e. touching the prey, and consequently were able to predict and control the stressor, did not show considerable or persistent stress indicators.” Even those animals that were able to predict the electrical shock, all three test groups in this study showed an elevated cortisol level related to the administration of the electrical shock, showing a clear stress/distress reaction.”

These physical findings, coupled with the behavioral signs, give a behaviorist evidence to make an evaluation of the dog’s state of welfare as it is defined by the Five Freedoms. The use of pain, be it from physical impact, choking, or electrical shock, appears to be present. This violates the principal of the second and third freedoms, freedom from discomfort and pain or injury: discomfort is shown by lowered body posture, excessive licking, and other indicators. Pain and injury can be proved by the level of pain indicia such as dilated pupils, avoidance, and certainly by clinical signs such as the strangulation cases listed above. These same indicators show that the fifth freedom, freedom from fear, is equally disregarded. Fearful behavior can be simply assessed by the behaviorist and quantified by such measures as speed and willingness to approach humans voluntarily, number and type of interactions, and the display of warning behaviors.

CONCLUSION

These measures should be sufficient to allow the establishment of use of aversive training methods as less than meeting the needs of a dog under the five freedoms. But is it cruelty?

The definition of cruelty in Florida State Statutes is: “A person who unnecessarily overloads, overdrives, torments, deprives of necessary sustenance or shelter, or unnecessarily mutilates, or kills any animal…commits animal cruelty.” (Florida State Statute 828.12 – 2012). Although this is a bit looser than the provisions of the Five Freedoms, we can still make a case that, based on the behavioral observations of an animal that is being trained using painful and frightening aversive techniques is being “tormented”. More on point are those cases wherein dogs are being physically injured. Although the provision for “unnecessarily” may come into play, we can revert to the purpose for the training; a change in behavior while maintaining animal welfare and a proper quality of life. Many other states in the US reflect this wording in their statutes, or are at times more stringent. The quantifiable, observable behaviors available to the Behaviorist may well be able to establish that the reactions to these training methods are abnormal and maladaptive and thereby show a needless level of distress.

Further amplification of this is given by other sources.

“Our results indicate that the immediate effects of training with an e-collar give rise to behavioural signs of distress in pet dogs, particularly when used at high settings. Furthermore, whilst best practice as advocated by collar manufacturers mediates the behavioural and physiological indicators of poor welfare detected in the preliminary study, there are still behavioural differences that are consistent with a more negative experience for dogs trained with e-collars, although there was no evidence of physiological disturbance. E-collar training did not result in a substantially superior response to training in comparison to similarly experienced trainers who do not use e-collars to improve recall and control chasing behaviour. Accordingly, it seems that the routine use of e-collars even in accordance with best practice (as suggested by collar manufacturers) presents a risk to the well-being of pet dogs. The scale of this risk would be expected to be increased when practice falls outside of this ideal.” “These findings suggest that there is no consistent benefit to be gained from e-collar training but greater welfare concerns compared with positive reward based training.” (Cooper 2014)

As stated by Coren (2012) in his examination of data regarding the use of physically forceful techniques in training pet dogs, “What does come out of this study is a confirmation that the use of punishing techniques on dogs has much the same effect that the use of physical punishment has on human children — namely an increase in aggressive behavior in general, and specifically increased aggression toward the individual who is applying the punishment. Given these findings, recommendations advocating the use of punishing and confrontational methods as part of dog training and behavior control seem to be ill advised.”

“In conclusion, confrontational or aversive behavioral interventions applied by dog owners before their pets were presented for a behavior consultation were associated with aggressive responses in many cases. Owners of dogs aggressive to family members are especially at risk for injury—and their pets at risk of relinquishment or euthanasia—when certain aversive methods are used. Ultimately, reward-based training is less stressful or painful for the dog, and, hence, safer for the owner. It is important for primary care veterinarians to advise owners about risks associated with aversive training methods, despite their prevalence in the popular media, and to provide resources for safe and effective management of behavior problems.” (Herron 2009)

“More owners using reward based methods for recall / chasing report a successful outcome of training than those using e-collars.” (Blackwell 2012)

“There are alternatives to aversive devices. I recently watched a Schutzhund dog work just as well on a Scruffy-Guider (Misty Pines Dog Park, Sewickley, PA) as he did on a choke collar, but he breathed better. I have seen military dogs learn almost instantly using head collars (Gentle Leader; Premier Pet Products, Midlothian, VA) because the target of their focus was clear. And I have seen my own dog, Flash, recover from being hung from a choke chain until he passed out, after which time he put the trainer in intensive care. That is how he became my dog…he was my patient first. Some people reading this may have met him, and so know what an amazing dog he is. Flash is the individual who first opened my eyes to learning to think in a different way simply because any forceful interaction with him would have resulted in injury to those exhibiting the force. No exceptions. His lessons have benefited many.” (Overall 2007).

CITATIONS:

Beerda, B., Schilder, M.B.H., van Hooff, J.A.R.A.M., de Vries, H.W., and Mol, J. Behavioural, saliva cortisol, and heart rate responses to different types of stimuli in dogs. Appl. Anim. Behav. Sci. 1997; 58: 365–381

Blackwell, Emily J.; Twells, Caroline; Seawright, Anne; Casey, Rachel The relationship between training methods and the occurrence of behavior problems, as reported by owners, in a population of domestic dogs Journal of Veterinary Behavior: Clinical Applications and Research, Volume 3, Issue 5, September-October 2008, Pages 207-217.

Chance, Paul, First Course in Applied Behavior Analysis, Waveland Press, Groveland Illinois, 1998

Cooper JJ, Cracknell N, Hardiman J, Wright H, Mills D (2014) The Welfare Consequences and Efficacy of Training Pet Dogs with Remote Electronic Training Collars in Comparison to Reward Based Training. PLoS ONE 9(9): e102722. doi:10.1371/journal.pone.010272

Coren, Stanley PhD Dog Aggression Is Predicted by Training Methods and Breed: Using punishment during training predicts aggression toward people. Published on March 18, 2014 in Canine Corner, Psychology Today

Dedalle, Stephanie; Gaunet, Florence Effects of 2 training methods on stress-related behaviors of the dog (Canis familiaris) and on the dog–owner relationship Applied Animal Behaviour Science Volume 105, Issue 4, July 2007, Pages 369–380

Dobbs, Jim and Phyllis and Woodyard, Alice TriTronics Retriever Training, published by TriTronics, Inc. 1993)

Grohmann, K., Dickomeit, M. J., Schmidt, M. J., & Kramer, M. (2013). Severe brain damage after punitive training technique with a choke chain collar in a German shepherd dog. Journal of Veterinary Behavior: Clinical Applications and Research, 8(3), 180-184.

Herron, M. E., Shofer, F. S., & Reisner, I. R. (2009). Survey of the use and outcome of confrontational and non-confrontational training methods in client-owned dogs showing undesired behaviors. Applied Animal Behaviour Science, 117(1), 47-54.

Hetts, Suzanne PhD., Pet Behavior Protocols, American Animal hospital Association Press, 1999

Hiby, E. F., Rooney, N. J., & Bradshaw, J. W. S. (2004). Dog training methods: their use, effectiveness and interaction with behaviour and welfare. ANIMAL WELFARE-POTTERS BAR THEN WHEATHAMPSTEAD-, 13(1), 63-70.

Lindsay, S. R. (Ed.). (2008). Handbook of Applied Dog Behavior and Training, Procedures and Protocols. John Wiley & Sons.

Matthijs B.H., Schildera, B. , Joanne A.M van der Borga Training dogs with help of the shock collar: short and long term behavioural effects Journal of Veterinary Behavior: Clinical Applications and Research Volume 9, Issue 2, March–April 2014, Pages 58–65

Overall, K. L. (2007). Why electric shock is not behavior modification. Journal of Veterinary Behavior: Clinical Applications and Research, 2(1), 1-4.

Overall, Karen L. Considerations for shock and ‘training’ collars: Concerns from and for the working dog community Journal of Veterinary Behavior: Clinical Applications and Research July–August 2007

Polsky, R. H. (1994). Electronic shock collars: are they worth the risks? Journal of the American Animal Hospital Association, 30(5), 463-468.

Polsky, R. (2000). Can aggression in dogs be elicited through the use of electronic pet containment systems? Journal of Applied Animal Welfare Science, 3(4), 345-357.

Rooney, N. J., & Cowan, S. (2011). Training methods and owner–dog interactions: Links with dog behaviour and learning ability. Applied Animal Behaviour Science, 132(3), 169-177.

Schalke, E., Stichnoth, J., Ott, S., & Jones-Baade, R. (2007). Clinical signs caused by the use of electric training collars on dogs in everyday life situations. Applied Animal Behaviour Science, 105(4), 369-380.

Schilder, M. B., & van der Borg, J. A. (2004). Training dogs with help of the shock collar: short and long term behavioural effects. Applied Animal Behaviour Science, 85(3), 319-334.

Washington State Patrol Detection Canine Handler Course, author unknown, provided in discovery materials by the Washington State Patrol in Criscuolo v. The City of Moses Lake Washington, 2011.