As a trainer, you naturally need to know what you are doing. What do you want to train your dog in and how? Understanding and learning the terminology presented on this course will make you a better trainer. Also communication between you and your dog will improve!
Reinforcement is something you are willing to work for.
Reinforcements are the engine that drives all behavior. Treats, petting, touching, playing, odours, being able to move freely, fellow creatures etc. are all examples of reinforcement. Effective reinforcement is pursued, earned, respected and instant. What is effective, is in turn determined by the animal himself. Training can help to make reinforcement more effective since the effectiveness of reinforcement enhances if it is earned. A small kibble can do the same trick as a whole sausage if the dog is not used to getting more valuable treats. Reinforcing with food is different from feeding. Similarly, you cannot use a ball, if you have a ball pit at home, since something so obvious cannot act as a reward.
As you move from home to more interesting environments to train, you realize the value of reinforcement. Environmental stimuli are competing against reinforcement and the trainer for the animal’s attention. If the piece of bread you were using at home doesn’t work in this situation, don’t blame the animal.
Primary i.e. unconditioned reinforcer is something the animal naturally likes and pursues. Primary reinforcers, such as food, water and social contact satisfy innate, breed-specific needs.
Secondary i.e. conditioned reinforcer is something that regularly precedes a moment of rewarding. Examples of these are praising, hand movement, treat bag rustling, whistling, clicker sound, door opening, leash button clicking etc.
Reinforcement is a process
In clicker training, food is a good primary reinforcer. Nutrition is a basic need for all living creatures and usually something animals are willing to work for. Marker is a secondary reinforcer. Examples of markers are clicker, whistle or flashing lights (used when training deaf animals).
To be exact, reinforcement is a process. It cannot be broken down into smaller processes, but must be viewed in its entirety. The impact of reinforcement begins sometime just before the act of reinforcement and it is impossible to say where it ends exactly. Only the animal himself knows the strength and duration of the reinforcement.
Reinforcement should always be given as soon as possible after the animal shows desired behavior. In fact, it is even better to give it while the animal behaves in a desired manner. That is why it is useful to divide reinforcement into two components; primary and secondary. Using the secondary reinforcer helps to make learning more effective and focused.
You need to preserve dependency between reinforcement and behavior. When the animal continually needs to earn reinforcement with his behavior, he begins to respect the reinforcement.
Reinforcement frequency is the time that passes in between each reinforcement that is received. The higher the reinforcement frequency is, the more learning has happened since the animal is showing the desired behavior more. Also, the higher we can keep the reinforcement frequency, the more the animal will focus on training. The reinforcement frequency will determine whether criteria need to be lowered or raised.
When teaching a dog a new behavior, the criterion that determines which behavior will be reinforced needs to be predetermined. As the reinforcement frequency rises, the criterion needs to be raised in order for the training to move towards a whole performance. Thus, after raising the criterion, the reinforcement frequency will decrease until the learning progresses and the reinforcement frequency grows again.
If a dog receives treats often (because the task is too easy), the reinforcement frequency is high, and on the other hand, if the task is difficult to accomplish and it requires many attempts, the reinforcement frequency is low.
A suitable reinforcement frequency is one where a dog is eager to try but will not become satisfied too quickly. When learning something new, a good reinforcement frequency is 12 reinforcers per minute, i.e. one treat every five seconds. When the task gets more challenging, or it takes longer to accomplish, the reinforcement frequency decreases.
When the desired behavior is achieved, reinforcement frequency should vary, for example
one behavior➤ reward, three behaviors ➤ reward, two behaviors ➤ reward, etc.
These are called reinforcement schedules.
The way you naturally behave while reinforcing affects the dog’s behavior too. If you use a toy as a reinforcer, or reinforcing is otherwise wild or frantic, the dog’s behavior will become similar. When training a fast agility dog who focuses on jumping at the same time, you constantly need to balance between using different reinforcers. Food calms the dog down, play speeds him up – a long lasting reinforcer calms him down, something that is eaten quickly speeds him up. When you remember the ratio between increasing the value of reinforcement and earning reinforcement, it is probable that the dog’s behavior will speed up sooner or later.
Criterion determines which behavior will be reinforced. Before training begins, the trainer needs to determine what behavior the animal will be rewarded for. Effective training is not possible unless the criterion is clearly defined. The criterion is raised when the learning has progressed to a certain level (for example, eight out of ten behaviors are accomplished). Raising the criterion is done by adding challenge to the behavior or moving onto the next stage of the task. Moving into a more challenging training environment, i.e. bringing more distractions to the animal’s environment, is also one way to raise the criterion.
If the reinforcement frequency decreases during training, it is a sign of the criterion being raised too quickly, and in that case, it is better to return to the previous stage in training.
Timing refers to the moment when a dog is given reinforcement. The more accurate the timing is, the faster the dog learns. In order for the training to progress, the significance of timing is extremely important because when we want to reinforce desired behavior, the reinforcement needs to be offered at the exact right moment. Even just a little bit too early or too late may mean that we have reinforced the wrong behavior.
Because it is difficult to offer an accurately timed primary reinforcer, it has become more common to use a conditioned, i.e. secondary reinforcer with which the desired action is impossible to mark earlier.
Latency refers to the time interval between when a trainer gives a cue one and when a dog starts to respond to it, and is defined by the trainer.
A short latency makes the behavior seem like it has been very well trained. As an example of this is if you say “Sit” and the latency is one second. In this case, the dog’s bum touches the ground for maximum of one second after the cue has been given.
We never use the word mistake to describe dogs’ unwanted behavior because dogs don’t make mistakes. In other words, a dog is always right. Dogs also don’t trick you on purpose by offering something else than what was desired. A dog’s attempt to find a suitable solution is called variation.