Back when the clickryder list was active on yahoo, people would often write in with their questions about clicker training horses. Sometimes those questions were about teaching new behaviors or “getting rid of” unwanted behaviors, but there were also questions about what to do when behavior deteriorated. These were questions like “my horse used to do this and now he doesn’t, what do I do?”
The most common answer I would give is that it was time to do some detective work. One of the basic ideas behind clicker training, and any use of operant conditioning, is that behavior needs to be reinforced (in some way) in order to be maintained. If an existing behavior falls apart, it means that something has changed. This could mean that the behavior is no longer reinforcing, that there is an alternative behavior with a stronger reinforcer, or that the behavior is being punished in some way. As the trainer, it’s my job to try and figure out what to do about it.
Please note that I did not say that it was my job is to figure out why. One of the first things I learned when I started clicker training was that figuring out why something is happening is not the most productive approach to solving the problem. The reason is that unless I saw what happened that led to the change in behavior, I am unlikely to be able to figure out why it happened. That means I am going to be guessing, and coming up with a training plan based on a guess is not the best idea.
I should point out that I am a person who likes to figure out the “why” of something, so I struggled with this for a while. But I have found that my time is better spent focusing on training the behavior I do want, than trying to address possible causes. Having said that, it doesn’t mean that I can’t keep an eye out for clues as to what might have been contributing factors and work to address them as well (either at the same time or at a later date.). As a matter of fact, I think it’s really important that I do this, because otherwise I might find that there are underlying factors that pop up and create problems in other behaviors or the problem re-occurs. I’d also like to add that it’s important to always rule out any physical reason for a change in behavior before addressing it as a training problem.
So what does it mean to do some detective work, if it doesn’t mean figuring out why? In this case, the detective work is identifying under what specific circumstances the behavior occurs. In some cases, it’s also helpful to identify when it does not occur. In Dr. Friedman’s LLA course, she had us write up what are called Functional Assessments (FA for short). An FA is a pithy description of the antecedent, behavior and consequence that describes the conditions under which a particular behavior occurs. It also includes a prediction for what will happen if the A, B and C continue as written.
An FA is not useful unless it is very specific, so this is where my detective work comes in. Can I describe exactly what happens? It’s not enough to say “my horse doesn’t do x…” I need to think in terms of what I DO see. What does the horse do instead of the behavior I want? What happens right before the behavior? And what happens right after the behavior?
It just so happens that I have a great example of working through this with a behavior that deteriorated recently. By sharing my detective work, I am hoping it will help you with yours. About 2 weeks ago Aurora, my 11 month old filly, started being difficult to halter. I have intentionally written the “problem” in vague terms, because that’s often how these questions are asked. And honestly, these problems sometimes creep up on us because there’s always some variability in behavior. That means it might take a few days after the problem started before I suddenly think “wait a minute, you used to be really good at doing this.”
Upon some reflection, here’s my FA for Aurora and the halter:
Background: Aurora has her halter off at night and I put it on every day to take her to turnout. She has been trained to put her nose in the halter and then wait while I slide the crown piece up and over her ears. I click as it goes over her ears and she is reinforced with a treat, after which I do up the throatlatch and attach the lead rope.
A: the crown piece is pulled up toward her ears (when I am on her left).
B: her head goes up
C: the noseband is no longer around her nose
P: She will continue to put her head up for haltering
What’s important here is that I identified exactly what she does (head goes up), when it happens (when I move the crown piece up) and what consequence is reinforcing it. Because the FA is specific about when her head goes up, it also tells me that it is something specific about haltering and not just haltering in general. If she put her head up or moved away as soon as I approached with the halter, I would have to consider that the change in behavior could be about haltering or it could be about what happens after the halter is on. It is also interesting that the behavior only happens on the left. Another clue!
Do you need to know how to write an FA? No, the reason I shared it is so that you could see how helpful it is to be specific. The FA indicates that there is probably something aversive about the crown piece and she is avoiding it.
Here are some of the way I could address this:
- I could change how I put on the halter by unbuckling the crown piece and it and taking it around behind her ears. This could be a long term solution or it could be a temporary fix/diagnostic to see if it the issue is the crown piece sliding over her ears. It’s not a bad idea to change how you do something to avoid triggering the problem, while you are working on a long term solution.
- I could teach an alternative behavior that is incompatible with the undesired behavior (putting her head up). A great one to use would be to teach her head down. If she’s putting her head down, then she will have her head nice and low and it will be easier to slide up the crown piece. Or I could put on the halter as usual, but cue head down as I slide the crown piece over her ears.
- I could ask her to participate more in the behavior. Right now she puts her nose in the noseband but then she waits for me to slide the crown piece up. Waiting is a passive behavior and some horses do better if you let them help more. So I could train her to target into the noseband and push the noseband up her nose farther (right now she just puts it in the opening. That would make it easier to slide the crown piece over her ears and we would be doing it together instead of her waiting while I did it.
- I could go back and re-shape haltering, spending more time building duration with just her nose in and getting her used to the rest of the halter being moved around. This would be a combination of shaping and desensitizing.
- I could spend time making her more comfortable having her ears handled and things sliding over them. This is more of a desensitization exercise and assumes that her reaction to the crown piece is because of how it feels going over her ears.
These are all reasonable things to do and they are not exclusive. I could work on getting her more comfortable having her ears handled, as well as teach her head down, or to push her nose into the halter. Even though I have written these out specifically for this haltering example, they are general strategies that you could adapt to many other behaviors.
Which one(s) did I choose?
In this case, I spent a few days putting her halter on without sliding it over her ears (option 1) and observed her when I took her halter off. Taking the halter off also requires allowing something to move over the ears and it gave me time to see if there was something specific about her ears. I didn’t want to build even more reinforcement history for putting her head up so I would take the crown piece over her left ear first. Her halter was still held on by the right ear. So putting her head up didn’t affect whether or not the halter stayed on her head.
Once she was comfortable with having her ears moved around when taking the halter off, I went back and re-shaped haltering. I clicked and reinforced her for putting her nose in the halter and would repeat that a few times before moving the crown piece toward her ears. This was option 3. I was going to do option 4 but it turned out to be unnecessary.
I didn’t spend a lot of time on any of this. I just worked it into her morning and evening routine so that we worked on some element of it in each session. Over the course of this, I was able to find some clues as to what caused the initial deterioration in the behavior. It was an interesting combination of factors, none of them major, but enough together to change her behavior. They were:
- I had started haltering her with gloves on. When I wear gloves she is noticeably a little less relaxed. I am clumsier with gloves on so everything is just more awkward.
- Her head grew and her halter had gotten a little tighter. It still fit her fine when she was wearing it, but she has nice long ears (the better to hear me with) and I think that it was uncomfortable when I slid the crown piece over her head.
- I had started haltering her with the lead rope attached. It turns out that the weight of the lead actually made a difference in how easily the halter slid up her nose and the crown piece went over her head.
- Ok…this is just a guess, but it’s an educated guess because she is sometimes a bit funny about the left ear.. So my guess is that she has been ear twitched. I say this because when I was at her vet inspection before I bought her, she was moving around too much and her owner held her hand up right near her ear as if to grab it. She didn’t actually do it and I didn’t ask, but it sure looked to me like she was getting ready to do it if needed. Sadly, I know it’s not an uncommon practice.
As of this week, she is now fine with haltering. I will be more careful about adjusting the halter as her head grows and I will continue to work on getting her comfortable with her ears being handled. I have also continued to halter her and then put on her lead rope.
Even though this training was in response to the deterioration of an existing behavior, it was a good project for the two of us. I haven’t had her for that long and by working through it, I learned a lot about what matters to her and how I can help her feel more comfortable about things. She also learned a little bit about participating in training and that she had a voice in our work together.
So the next time your horse doesn’t quite do what you expect or want, put on your detective hat and see what you can learn. I can guarantee that you will both learn a lot from it.