Download PDF 7–10 minutes to read
The aim of this guide is to provide a basic understanding of how positive reinforcement training works, some jargon busting, the stages of training, and common mistakes made when training.
Many trainers use different words or hand signals for each cue, and this can be quite confusing to those new to dog training. There are also different terms used in dog training that may not make any sense at all! Hopefully this section will cover all of these, and any that are relevant to you will, if required, be covered in more detail in your personal training plan.
A reward is something given to a dog as a thank you for doing well during training. It can be a treat, a toy, a play session, a cuddle, or even being let off the lead. What reward you use is dependent on what drives your dog, and the surroundings. For example if your dog is rewarded by being let off the lead, but you are training next to a busy road, the lead off reward would not be appropriate and another should be used in its place.
Rewards should always be a higher value to the dog than the thing he is trying to do or get to. That doesn’t mean the reward has to have a high monetary value – dogs don’t care about money! This is especially important when training new cues, or working on bad habits or behaviours. These rewards should only be used for training, and different rewards can be used to different things.
Reward marker and non-reward marker
A reward, and a non-reward marker are verbal, auditory or physical signals we use to tell a dog they have done well or have missed the mark. A reward marker would be something such as ‘good boy’, ‘well done’ or ‘yes’. These sounds, once learned by the dog, will let the dog know they have done something correctly. As the name suggests, these markers are precursors to a reward of some kind, be it a toy, treat or a cuddle. In advanced training these markers are used to teach strings of cues, for example teaching a dog to play dead when it’s been ‘shot’ is taught in small parts and eventually strung together.
Non-reward markers are the opposite, again these are verbal, auditory or physical signals given to a dog to let them know that they haven’t done quite what we asked. When a dog is asked to present a ‘sit;, for example, and they go to a lie down, a non-reward marker is used, and the dog is given an opportunity to correct his mistake., before continuing. Many people confuse non-reward markers with a ‘telling off’, and while a non-reward marker means the dog isn’t getting a reward, no shouting, physical force or pain should ever be used. A non-reward marker is a learning technique, and should never be a punishment or ‘telling off’. A common non-reward marker is a gentle ‘no’ or ‘ah-ah’.
A threshold is a term used mostly in behaviour modification, that refers to a dogs tolerance of a particular stimulus. As an example, a dog who is scared of other dogs will display different behaviours the closer to, and the further past their threshold. The distance before the threshold is known as being under-threshold. This is the distance where learning can take place, where the dog is aware of the other dog, but is not overly worried about their presence or their proximity. If we were to move the frightened dog closer, it may pull back, growl, whine, begin panting, snarling or lashing out with warning bites. At that stage the dog would be over-threshold. When a dog is in this state, learning is blocked by neurotransmitter chemicals in their brains and no progress can be made with behaviour modification or training.
An extinction burst is a psychological term describing the re-emergence of an old behaviour in times of stress, worry, or over-excitement. When training a dog to not jump up for attention for example, you may see fantastic progress for the first 6 months, then there’s a thunderstorm and the dog is jumping all over the place. Extinction bursts happen because the dog is unable to think as clearly as they can during training sessions, and so revert to a previous behaviour that is much more ingrained. These bursts are not something the dog has control over, and should never be met with punishment.
Dominance and Submission
Dominance and submission are normal parts of a dog’s social life, and are situational dependent on who they are with and what they are doing. A pup will, in the vast majority of cases, show submission to an adult dog, however that doesn’t mean the adult dog is showing dominance. Pups are submissive to adults out of respect and defer to them in social situations.
Submissive behaviours, called appeasement behaviours, are shown in conflict situations. When a dog is shouted at by its owner, it will move away, lower its head and ears, keep its tail low and avoid eye contact. This is the dog telling the person they are no threat, and asking them to calm down. A lot of owners believe their dogs have a ‘guilty’ look, when in fact the dogs are simply offering appeasement signals to avoid a confrontation.
Dominance and submission are very over-used in old school training, where a dog can be either dominant or submissive all the time, but never both. We know now that it doesn’t work like this.
The Stages of Training
When a command is in the learning stage, a dog will find it very hard to respond if it is in a distracting place, new commands are learned best in the home. When in this stage, rewards should be given each and every time the behaviour is presented, or for more advanced cues, whenever the dog makes a move to nearly the right place. Once the command is understood in the home, take it into garden training sessions. Gardens are full of smells and other interesting things, so serve as a good transition between the home, and the super distracting outside world! Again during garden training a command just learned should be rewarded each and every time the behaviour is given. This command can then be used in the outside world, but be aware that there are a myriad of interesting things out there, so he may be slow to respond or won’t respond at all. If he doesn’t respond, try again after a short while and remember that praise and a reward should be given each time they respond correctly to the command.
Once a command is being reliably responded to, the generalisation process begins. This process involves varying reward value and frequency, while most importantly, practicing in different places – in a friend’s house, in a quiet park, on a hill hike. Up until this stage your dog will only know how to respond to the commands in specific locations. While it’s easy for us to move learned behaviours to other environments, it’s not easy for dogs and this stage is vital to their learning process.
At this stage you reduce the frequency of reward, however don’t stop rewarding completely or the command will be ignored. At this stage rewards should be given randomly, in random amounts. For example if you are reinforcing recall reward every second time with a couple of treats, use lower value rewards, but occasionally use high value rewards. This method uses the same part of a dog’s brain that a slot machine uses in ours – we didn’t win last time, we have to win this time! It encourages faster and more consistent responses to the command, because the ‘jackpot’ might be waiting.
Sadly, most people stop at this stage, without proofing a command.
Proofing is a hard-mode of the command. Can they sit and look at you when there are balls being thrown around, or there are other dogs playing, or kids running about nearby? (but not around, I wouldn’t sit about if some kids were screaming around me in circles, don’t expect your dog to either). At this stage the proofing should happen in short bursts so as not to over stimulate the dog. If they manage to do what they are asking, they need a huge jackpot – after all they are breaking their natural instincts to do what you are asking them, a thank you is the least we can offer them! Once the command is presented, reliably, with various distractions, it is then proofed. After this rewards only need to be used sparsely, and lower value rewards can be used. Training to this stage could literally save your dog’s life!
Put quite simply, overshadowing happens when something used in training takes higher priority to the dog than what is being trained. As an example, if a dog is always trained by first being shown a treat, the dog will only respond when it sees a treat, as treats are much more interesting than the words and hand movements being shown to it. If a person trains their dog to sit, but lowers their head while saying sit, the dog will only sit when it sees the lowered head, even if the cue ‘sit’ is not spoken. Dogs are experts at body language, how you hold yourself when training will affect how the dog responds ‘in real life’.
Blocking occurs when a new stimulus becomes irrelevant when an old one is present. In other words, when something they understand is said, or done, before something we are trying to teach. An example would be if you were training your dog German cues and say “down” and “platz” the dog will likely not rely on “platz” because the familiar word down provides enough information plus it has a history of reinforcement so who cares about ‘platz’. If on the other hand, you say ‘platz” first and then “down” second, your dog will learn that platz means down, and since dogs like to anticipate, after some reps they will attend to the word “platz” alone even if they don’t hear the familiar cue ‘down’. To avoid blocking when training, remember to say the cue before the reward-marker
Hopefully this has given you some understanding of how training works, and how to progress with training. Further details will be given in your training plan, and can be used in conjunction with the information here.
If you need further explanations, or for comments, please get in touch!