Maybe at some point you’ve asked yourself this question, and it’s that sometimes without realizing it some behaviors begin to decay, or the animals begin to directly refuse to do them. In these cases, it is important to be observers, and to perceive the root of the problem, because sometimes the cause may come from a precursor that we have not perceived.
We will expose some of the most common reasons.
The moment of the bridge
One of the reasons why the criteria of a behavior decreases is because the bridge is always at the same point. We must bridge in various moments of the topography of behavior, (such as just that the animal responds correctly, we can reinforce that immediate response, even as much as we do when the behavior is about to end).
For example, if we ask the dolphins for bows and we always whistle at the fifth jump, what usually happens is that the animals quickly realize at what point you will give the bridge and therefore the first four jumps will have a lower criteria, and then when you reach the fifth, the criteria will increase, as they know that it is when the bridge will come.
If we do not realize this soon, and reinforce it, we will be creating a reinforcement history with this situation only.
{youtube}https://youtu.be/nYAqWdF4d9U{/youtube}
In this video we can clearly see how the animals are increasing the criteria as the point where the bridge will be made approaches.
In the photo above, we break down the complete behavior so you can see it clearly: In jump 1 the height of the jump is 1 meter, and by photo 4 they are jumping 3 meters high.
Search for variability
The variability in our training is fundamental. It is important that our animals do not know what is next or when we are going to give the bridge, following the example above, to prevent the animals from lowering their criteria, when sending the bows we would have to vary the moment of the bridge by doing it sometimes in the first jump, in others to the third, second or fifth.
With this simple method, we will ensure that the animals do not know when we are going to give the bridge, and therefore that the behavior maintains its original criteria throughout.
Scenario
Another important point is from where we send the animals to do the behavior. We can avoid always doing it from the same point, because if we always send the same behavior from a specific place, the animal will know what behavior follows, and we will be predictable. Therefore, they could increase the latency with which the animal goes to this point or even directly refuse before we give the signal.
For this reason it is important that our animals perform the behaviors from any part of their enclosure: this is known as generalizing a behavior.
Our signals
How we give a signal to an animal will also or will not help the criteria of a behavior to decline. Once we condition a cue, it must be clear and easily distinguishable for the animal, and as we mentioned above, reinforce the animal’s response to that signal on occasion during our sessions. If on the contrary, our signals begin to be confusing, with each trainer performing it in a different way and we do not take care of the details when performing it, for example, if the animal has attention or not before we give the signal, not only will we diminish the motivation of our animals, but we will also create confusion and weaken the criteria of behavior.
Reinforcement criteria
There are times when the most "simple" behaviors are those that fall more easily.
One of the most common mistakes is to label behaviors as "easy” or “difficult". When we do this, unconsciously we ourselves are to blame for the behavior decaying, because of our way of reinforcing those so-called "easy" behaviors.
If we reinforce the animals by the level of difficulty, the animal will lower the criteria, since it will be able to predict how it will be reinforced. By having variability in reinforcement without taking into account the difficulty of the behavior, we will maintain the intensity of each behavior, since we will always have a good expectation of reinforcement.
For that reason we have to be variable when reinforcing any behavior (regardless of the level of difficulty) to avoid being predictable.
And remember, if it’s possible... WeZooit!
