Updated: Jul 18, 2020
When training a new behavior - or attempting to determine where a behavior has broken down - it‘a best to look at three simple factors:
Rate of Reinforcement
Before I dive into specifics, first thing to remember: animal training is a mechanical skill. As primates, humans tend to use their bodies a lot in communication and sharing information. While this is great when communicating with other humans, the body language we give off can be confusing to our animal learners. The first thing to do is keep your body calm and still. Practice without the animal. Get used to the simple mechanics of clicking (or marking) before moving your treat hand. Limit talking or vocalizing and make your cues extremely salient to the learner
Now, let’s break the rest down step-by-step:
Timing: Be very precise in your clicking/marking the behavior and delivery of reinforcement. if you are using a clicker, know that it is a precise tool, capturing an exact moment of behavior. Think of it as a surgeon's scalpel - one slight move and there are very different outcomes. Click and reinforce within milliseconds. This may require the trainer to be fast (ever trained a chicken?), which is why it’s so important to practice the mechanical skills before working with the animal.
Criteria: Know exactly what you want to reinforce. Be specific! Don’t wait until the animal is performing the behavior, or the click/mark may be late. Not having specific criteria may lead to inconsistent behaviors and outcomes. Be careful not to “lump” too many criteria (more on lumping in another post). One criterion at a time.
Rate of Reinforcement: Time to bring your A-game! If you are training a new behavior, reinforce with a 1:1 ratio. The process needs to be worthwhile for the learner to want to participate. Keep the rate HIGH. If I am working with an extreme leash puller, my rate will be one step, click, treat. I want an extremely high ROR for being by my side with a loose leash.
How does this apply to an undesired behavior? Example: Your dog is constantly pulling you towards bushes to smell while walking. You would prefer if the dog walked by your side until released (this is just an example...it’s really important for your dog to smell things on walks!) First, do we have a second or two where the dog is on a loose leash at the owners side? Great, start there.
Timing: Mark dog for being at your side
Criteria: Dog at handler’s side, leash is loose
Rate of reinforcement: 1:1 ratio = one step:one reinforcer
Note, if the dog is an extreme puller and there’s never a time when the leash is loose, start with marking the dog just for sitting at your side before you even take a step.
As the dog gets more proficient, the criteria can get more specific - mark for head up, mark for being 2” from the owner’s side. For starters, at the owner‘s side with a loose leash is challenging enough.
When behaviors are breaking down, carefully examine them with these three aspects in mind. Then go back and retrain to fluency.
We’d love to hear your success stories, so please post them in the comments.