These are my notes from the ORCA Great Minds conference at UNT in March 2009. The Great Minds conference brought half a dozen top trainers together to talk about animal training, clicker training, operant conditioning and the future of training. This is the seventh post of my thoughts and notes and the second post about Kay Laurence. To read my first page of notes about Kay Laurence, please see this post. To read more about the conference in general and to find the rest of the ORCA posts, please see this post.
This is the second post about a lecture I heard by dog trainer Kay Laurence.
During the second part of her talk, Kay Laurence highlighted what she believes influences the effectiveness of a reinforcer. My notes from this part of the talk are not nearly as thorough, so much of this post is my own thoughts, based on things Kay talked about.
One of the important concepts she talked about was location. It’s important to consider both where the animal is when you reinforce and where you are keeping the reinforcer. If you are teaching a dog to sit, and you always reinforce the dog for sitting in front of you, the dog only learns to sit in front of you. Later, you’ll have to generalize behavior to include things like sitting beside you, sitting from a distance, etc. This isn’t a bad thing, so much as it’s just something the trainer needs to be aware of. The position of the animal when we reinforce is what we are reinforcing.
I struggle with this sometimes with my dog, Ginger. Her default behavior, many times, is to sit. On many other behaviors, we’re currently working on increasing duration. So, I’ll ask her to down or stand, and she knows very well how to do either. However, between the click and the delivery of the treat, she defaults back to sitting. It’s horribly frustrating sometimes, but I’m learning to deal with it. For stand, if I ask for stand and she does, and then I take a step or two back, I can usually get her to move a step or two, and prevent her from going back to a sit. But not always, so we’ll keep working on it.
As well, the location of the reinforcer is equally important. Where the reward is, such as in your hand, in your pocket, on the counter and so on, will all affect how the animal focuses on it and the timing and efficiency of your delivery. With my dog, I have to keep the treats either on the counter or in a treat pouch that is behind me. If the treat is in my hand, or if the treat pouch is on the side or front of my body, my dog focuses on the treat, instead of on me. With the horses, I use a treat pouch that usually sits on my side or behind me. However, I have to be careful because unlike the dog, the horses can very easily reach the treat pouch!
One important thing to consider is how long it takes us to get to the reinforcer. If you have to go in the next room, dig through a coat pocket or do something else that takes causes you to physically or mentally leave the animal, you could end up punishing the animal by being inefficient with your delivery.
This goes along with timing of delivery. As long as we are engaging the animal in the process, a bit of delay in delivery causes anticipation. The anticipation then becomes part of the reinforcement. For instance, many horse trainers use peppermints as special rewards. The unwrapping of the peppermint and crinkling of the plastic becomes part of the reinforcement–the animal is actively waiting and anticipating the reward because it can hear the wrapper and start getting excited about the peppermint. This is different than fumbling around in your jacket pocket because you can’t find a treat, which frustrates the animal because it doesn’t know when it’s going to get the reward.
The next concept Kay talked about was the speed and pace of our training and reinforcing. This should always match the lowest denominator. Meaning, a fast, energetic, trainer might need to slow down to accommodate a dog who is a slower learner. Likewise, a eager-beaver, over-enthusiastic dog can learn to slow down if it’s paired with a trainer who prefers not to work at a break-neck speed.
I thought this idea was absolutely fascinating, given what the mantra in the horse world usually is. I’ve heard from many trainers that we must up our energy to match that of the horse, but, I think this is usually from a non-clicker point of view and often includes upping our energy when teaching in order to make demands of the horse and keep the horse in line.
Although, I think what Kay was talking about was less about energy levels and more about speed of learning and the pace of the session. There’s a happy-medium that must be found where both animal and trainer are comfortable with the speed the training is progressing at. If we train at too fast of a pace, we risk creating a frustrated animal who does not completely understand the behaviors. If we train at too slow of a pace, the animal gets frustrated out of boredom. Somehow, we must find a comfortable speed where animal and trainer can both remain engaged.
As trainers, we must also consider just how much of a reinforcer is necessary to maintain the behavior. At the beginning, the animal could require a huge reinforcer to maintain the behavior. Later on, a much smaller reinforcer could be needed to continue to sustain the behavior. Yet, we often forget to decrease the size of the reinforcer. By doing this (giving a high value reinforcer for a behavior that only requires a low value reinforcer), we devalue our reinforcers.
Also, to some extent, our reinforcers work to reinforce both the behavior that occurred and the behavior that occurred just previous to that. Sometimes, a dog jumps up on a person and the person asks for a sit, then reinforces the dog. The bulk of the reinforcer goes to reinforcing the sit. However, to some extent, especially if this chain of behaviors gets repeated, the jumping up is also getting reinforced.
A better alternative would be to turn and walk away for a few seconds and then ask for the sit. Better yet would be to anticipate the jumping behavior and to ask for a sit before the dog begins the jump.
By looking closer at exactly how and what we are enforcing, we can work towards more efficient reinforcement and training.
Recently, I bought a rock climbing chalk bag from the outdoors store. (Like these that REI sells.) Previously my delivery of reinforcers with the horses was often quite awful. I was keeping treats in the front and back pockets of my jeans, and was always fumbling and digging around for a treat. The chalk bag is great because my rewards are right there by my side–all I have to do is reach in and grab one. As well, the bag has a drawstring on it, which makes it easy to open and close. (This is advantageous because for some of the horses, the temptation to stick their nose in the bag is still too great. We’re working on this, but it’s nice to have a bag that closes!)