Splitting and Lumping and Reinforcers and the Environment, 3-25-16

I’ve been working with Ollie on his trailering skills a lot lately.  His challenges, as I may have mentioned before, are quite different from Atty’s.  Ollie’s have more to do with generalized fear and anxiety.  He’s afraid of air, really.  And he has physical and proprioception challenges.  He not only doesn’t  know where his body parts are – either in space or relative to himself – but he’s also neither balanced nor coordinated and is somewhat weak in some areas, such as his hind end.  His hocks may be compromised and he has had a stifle injury.

And, silly me, I thought we had made quite a bit of progress on trailer loading, although the biggest problem was backing out.  He doesn’t want to get on the trailer in the first place because that means having to back out.  The scares the bejesus out of him so he rushes it and scares himself even more.  Then he recently scared himself again and nearly fell down in the trailer.

I had been using high-value reinforcers (Stud Muffins) to get him on the trailer.  Then I found I could drop back to Alam cubes to get him on.  But I still had to use pieces of Stud Muffins to slow him down when backing out or he just rushes out backwards and scares himself some more.  I found this fascinating.  Using high-value reinforcers to slow him down!

I had played with two and then one pedestal to get on and off.  I tried once to have no pedestal and that was not a good idea.  I went back to using one pedestal, the larger of the two that I have.

In this process, I was talking to Peggy Hogan about using Stud Muffins to get Ollie to slow down backing off the trailer.  Peggy was working with Eva Bertilsson and Emelie Johnson-Vegh on a talk they were all giving together at the upcoming Cincinnati Clicker Expo (March 18-20, 2016).  Peggy mentioned that Eva commented about using high-value reinforcers might get our animals to do things they were otherwise afraid of.  The question is, “Is it okay to do use high-value reinforcers when working with something that the animal perceives as aversive?”

I agree.  Ollie has his fears around the trailer.  Yes, I was using high-value reinforcers (HVR) to get behavior.  I was going down that path of the HVR because I thought I needed to.  I also thought that, even though we had worked with backing off the pedestal before, we had moved beyond that.  Again, silly me.  How wrong I was.

I started an email conversation with Eva about using HVRs in aversive situations and she talked about using HVR as a way to cheat in the training, not splitting the training plan enough.  I thought she meant splitting behavior as that’s the only way I’ve thought of splitting during training.  Eva said:

“I’m just seeing that this is one of the aspects where I end up in my considerations of high-value reinforcers in potentially aversive contexts; the notion that if I can find the best possible fine-slicing, my “need” for the top value reinforcers is reduced.”

… “But this also, indirectly, reduces the risk that the animal chooses “yes I will get into very aversive event for very valued reinforcer” thus entering the situation with a lot of mixed emotions.

“In my own training I try to remind myself: If I see that the training in a potentially aversive context goes much better with high value reinforcers, I try to look for possible ways of introducing the aversive even more gradually so that it is not the value of my reinforcers as such that gets me the behavior I want.

“There are definitely other aspects to the concept as well, not related to the combination with aversive events. For example, there is the level of arousal or emotionality that is evoked by the reinforcer (or the anticipation of it). Plus, what happens under extinction conditions (whether it is withholding a reinforcer in shaping, in chaining, or after a mistake)? Might it be that missing a highly valuable reinforcer  is more aversive than missing a low-value one? I see this frequently in humans, but again it might not be the high/low value reinforcer in itself that causes the problems, but rather the frequency of reinforcer.”

As I told Eva, I had to read this a couple of times before I really got what she was saying.  And even then, I wasn’t totally sure.  For all I knew, there was something lost in translation.  Eva is Swedish and, even though her English is excellent, I still wasn’t quite sure I understood her fully.

When I finally got a chance to talk to her live, at the Expo, I got a better explanation of what she meant.

I thought I had been fine splitting Ollie’s behavior just great!  Ollie can microshape me into the ground!  He can move by millimeters or repeat behaviors endlessly!  Or, he can create new variations just in case I really didn’t JUST want a hoof lift!

The light bulb finally came on in full 150 Watts when Eva said, “Yes, you were splitting the behavior.  But you were lumping the environment.”  Doh!

What Eva meant by saying that I was “lumping the environment” was that I was overfacing Ollie with actually loading onto the trailer and expecting him to “get used” to backing out. What I really needed to do, and did, was to drop the criteria to practicing more on backing off the pedestal. Even using a lower pedestal or even a flat board, both of which I did very recently.

Criteria can be raised or lowered by how we present or create or manipulate the environment. How high is the jump? How wide is the creek? How many weave poles am I starting with?  If I am going shape backing through an ell of poles on the ground, should I start with a 90-degree angle or something shallower?  Should I start with poles on the ground or poles raised up?  Or should I use something more solid?

Keep in mind that Peggy and I do split the environment and I had dropped the criteria to just working on Ollie backing off a pedestal.  Bob Bailey says that we are “masters of manipulating the environment.”  I just hadn’t thought of it the way Eva phrased it.

It’s amazing how clear something becomes when it shows up in the right wording.

Here are a video of working with Ollie and backing off pedestal.  Count how many times he drops the right hind and lets it dangle before he finally puts it on the ground.