This is phase 2 coursework training, so the terminology on the checklist doesn’t exactly align with the true definition of avoidance learning. It is prep for it in phase 3. Technically the trainers are doing a form of latent leaning and “correction” without true punishment, where they are teaching where the punishment will happen while purposely not punishing at a level that the dog would avoid in the face of most competing motivators. However, the dogs are learning when it happens and that if “corrected” they are made correct and get good things.

Phase 3 the dogs are introduced to true punishment, but because they understand when it happens and how to avoid it, it becomes a smoother experience and ultimately requires less true punishment and at lower magnitude than if the dogs were exposed to more punishment before they truly understand the contingency.
To my knowledge, Skinner doesn’t cover much on latent learning since it is covered more by the cognitive scientists. Basically there are many associations the dogs can learn before they are motivated to act upon it in specific situations.