The use of punishment and negative reinforcement in dog training
Ratification Date: 05 Nov 2021
Policy
Punishment and negative reinforcement should not be used in attempts to change the behaviour of dogs. Training of dogs is best achieved through positive reinforcement.
Animal behaviour education and assessment should be incorporated into mainstream veterinary practice.
Purpose
This policy will help to inform veterinarians and the public of the importance of avoiding negative reinforcement and/or punishment in dog training, and the benefits of positive reinforcement.
Background
Definition of training
Training is a term that refers to the act of teaching an animal new or different ways to behave in specific situations. Training involves the animal trainer manipulating the consequences the animal experiences as a result of its behaviour. These consequences can be punishing or reinforcing.
Definitions of punishment, negative reinforcement and positive reinforcement
Punishment is a term that refers to the manipulation of consequences such that the punished behaviour is less likely to be repeated in the future1. In the context of training, the trainer or animal handler inflicts fear, pain or discomfort upon an animal as a consequence of performing undesirable behaviour.
Negative reinforcement refers to an animal becoming more likely to perform a behaviour in the future when trying to avoid or remove something unpleasant1. To teach using negative reinforcement, it is necessary to provide a stimulus to the animal that is unpleasant enough that it cannot be ignored. The animal will then perform behaviours to escape the stimulus and the unpleasant stimulus stops. When negative reinforcement learning has been achieved, the animal performs the desired behaviour when the unpleasant experience is predicted so as to prevent the experience from occurring.
Positive reinforcement occurs when an animal receives something it desires for doing a behavior1. The behaviour is more likely to be repeated in the future because the animal is motivated to do the behaviour to receive the desired reward, such as food, praise, play, rub down or a scratch.
Problems associated with punishment and negative reinforcement in training
The five freedoms include negative welfare states that animals should not experience while in human care2. Using methods that purposefully utilise fear, pain or discomfort to bring about behavioural change can contravene both “Freedom from fear and distress” as well as “Freedom to perform natural behaviours”. Welfare can be reduced in animals who are taught with these training methods3,4.
Emotional learning of associations can occur when an animal is exposed to a stimulus that causes pain, discomfort or fear. Animals develop increased stress responses when exposed to training using positive punishment and negative reinforcement5-7, especially in the hands of people who do not have an advanced understanding of animal behaviour. The freeze response and learned helplessness can be confused with a calm and compliant animal who is happy to obey8.
When the emotional response is high enough for an individual, the ability to respond appropriately to any intervention to change behaviour is reduced9. With increased stress, task learning becomes impaired10 so continued use of techniques that increase fear can lead to difficulties in improving the behaviour or development of alternative future behaviour problems6,11-14.
Punishment - risks
For some animals, punishment effectively changes behaviour, however the timing must be correct and the punishment needs to be significant enough to bring about this change. Each animal will have a different level of motivation and sensitivity, which means that the severity of punishment that is required for motivational change in each individual will vary. That level is an unknown for each animal initially, and so the trainer must establish through trial and error. This approach has inherent risk.
Punishment does not take into account the environment in which the behaviour is performed nor teach the animal what behaviour is appropriate. It aims to reduce the behaviour without a full understanding of the underlying cause for the behaviour, which may be due to an undiagnosed health problem that needs veterinary intervention.
Training techniques that aim to stop or reduce behaviour can lead to other undesirable behaviours emerging11 and reduced welfare3,4. Positive punishment can also result in inhibition of behaviours that serve as communication signals5,14. An example is an animal who is punished for using normal threat displays, for example when a dog growls, snarls or barks. These threat displays are a form of early communication to warn people to stay away. If such displays are punished, the animal will cease warning people; however, the underlying fear causing the behaviour remains or may be even increased due to the punishment. This can lead to what appears to be unpredictable, explosive and sometimes higher intensity aggression. Positive punishment techniques are known to be associated with increased aggression from dogs6,12.
Negative reinforcement - risks
Although correctly applied negative reinforcement can effectively change an animal’s behaviour, increased stress can result from the use of negative reinforcement training.15 When negative reinforcement is used incorrectly or inconsistently, it is difficult for an animal to learn which behaviour gains relief from the stimulus causing pain, discomfort or fear.
When this training technique is used, dogs may change other behaviours in an effort to escape the stimulus they dislike. Because an animal does not know which behaviour is correct during the learning process, the animal can try several different behaviours to escape the stimulus, some of which may be undesirable to the handler. If the behaviours performed are significantly undesirable, some handlers then punish the animals for those behaviours, further increasing the animal’s stress. An example is inappropriate use of a choker chain to create a sense of breathlessness to train a dog to walk more slowly and not pull. The dog may in fact pull more or have increased activity due to the need to escape the feeling of tightening around the neck. In response, some handlers will jerk the choker chain to cause pain in an effort to punish this behaviour
Positive reinforcement - benefits
Positive reinforcement is correlated with a reduction in behaviour problems4, , can be used to change difficult behaviours16 and is less stressful for the animal16. It is successful because the animal is motivated to change its behaviour. It also offers the opportunity for mental stimulation and social interaction with the owners and is known to increase the human-animal bond.
Positive reinforcement training can improve an animal’s overall quality of life 6,12,15 improve learning17,18 and reduce the risk of future behaviour problems. This can in turn reduce the incidence of relinquishment and euthanasia.
Veterinarians can play an essential role in promoting and using best evidence-based training methods. This is an important aspect of preventative veterinary medicine.
Education of the public on this issue will reduce the risk of physical and psychological harm to dogs, thus improving animal welfare into the future. This will have positive benefits on the client-animal bond as well as public safety.
Other relevant policies and position statements
- The responsible ownership of dogs and cats and the human–animal bond
- Use of behaviour modifying collars in dogs
- reward-based-training-brochure-web.pdf (ava.com.au)
References
- Skinner, B.F., Science and human behavior. 1953: Simon and Schuster.
- McCulloch, S.P.J.J.o.a. and e. ethics, A critique of FAWC’s five freedoms as a framework for the analysis of animal welfare. 2013. 26(5): p. 959-975.
- Schilder, M.B. and J.A.J.A.A.B.S. van der Borg, Training dogs with help of the shock collar: short and long term behavioural effects. 2004. 85(3-4): p. 319-334.
- Hiby, E., N. Rooney, and J.J.A.W.-P.B.T.W.-. Bradshaw, Dog training methods: their use, effectiveness and interaction with behaviour and welfare. 2004. 13(1): p. 63-70.
- Gray, J.A. and N. McNaughton, The neuropsychology of anxiety an enquiry into the functions of the septo-hippocampal system, in Oxford psychology series no 33. 2000, Oxford University Press,: Oxford. p. 1 online resource (xviii, 424 p.).
- Ziv, G.J.J.o.v.b., The effects of using aversive training methods in dogs—A review. 2017.19: p. 50-60.
- Schalke, E., et al., Clinical signs caused by the use of electric training collars on dogs in everyday life situations. 2007. 105(4): p. 369-380.
- Bradbury, G. Getting to grips with correct rabbit handling. 2016 [cited 2019 3 March]; Available from: https://www.vettimes.co.uk/app/uploads/wp-post-to-pdf-enhanced-cache/1/getting-to-grips-with-correct-rabbit-handling.pdf. Accessed June 2019
- Overall, K., Manual of clinical behavioral medicine for dogs and cats. 2013: Elsevier Health Sciences.
- Lubin, F.D., T.L. Roth, and J.D.J.J.o.N. Sweatt, Epigenetic regulation of BDNF gene transcription in the consolidation of fear memory. 2008. 28(42): p. 10576-10586.
- Mason, G., J.J.F. Rushen, and U.C.P. applications to welfare. Oxfordshire, Stereotypic animal behaviour. 2006: p. 286-324.
- Herron, M.E., F.S. Shofer, and I.R.J.A.A.B.S. Reisner, Survey of the use and outcome of confrontational and non-confrontational training methods in client-owned dogs showing undesired behaviors. 2009. 117(1-2): p. 47-54.
- Overall, K.L.J.J.o.V.B.C.A. and Research, Considerations for shock and ‘training’ collars: concerns from and for the working dog community. 2007. 2(4): p. 103-107.
- AVSAB, A.V.S.o.A.B. AVSAB Position Statement: The Use of Punishment for Behavior Modification in Animals. 2018 [cited 2019 3 March]; Available from: https://avsab.org/wp-content/uploads/2018/03/Punishment_Position_Statement-download_-_10-6-14.pdf. Accessed June 2019
- Deldalle, S., F.J.J.o.V.B.C.A. Gaunet, and Research, Effects of 2 training methods on stress-related behaviors of the dog (Canis familiaris) and on the dog–owner relationship. 2014. 9(2): p. 58-65.
- Cooper, J.J., et al., The welfare consequences and efficacy of training pet dogs with remote electronic training collars in comparison to reward based training. 2014. 9(9): p. e102722.
- Rooney, N.J. and S.J.A.A.B.S. Cowan, Training methods and owner–dog interactions: Links with dog behaviour and learning ability. 2011. 132(3-4): p. 169-177.
- Innes, L. and S.J.A.A.B.S. McBride, Negative versus positive reinforcement: an evaluation of training strategies for rehabilitated horses. 2008. 112(3-4): p. 357-368.
Date of ratification by AVA Board: 5 November 2021