He is a two-door service dog. The latest model.
While I spend the vast majority of my time in bed, I also make frequent trips to the adjoining “master bathroom,” which has a difficult-to-open door. It’s actually not as bad as it used to be, but I can never fully shake off the fear of my first experiences with this door.
When I first used the bathroom in this house was when I was a potential home-buyer. I went into the bathroom, shut the door, and did my business. Then, I tried to open the door, and I couldn’t. It was stuck. It was summer, and the wooden door had expanded and become too tight. I’m not super strong. I yelled for help. Nobody heard me. I banged on the walls. I tried repeatedly to tug the door open with its obnoxiously unhelpful egg-shaped door knob.
I don’t remember how I got out. Either someone noticed I’d been gone a while and came to look for me, or — using that extra boost of adrenaline that comes with a combination of fear and humiliation — I finally managed to free myself. Forever after, I was nervous about getting locked into that bathroom.
I made changes: I changed all the egg-shaped knobs to levers and hung door pulls on them for Gadget to use to open and shut the doors. The levers were also easier for me to open. And most importantly, a locksmith friend of mine adjusted the door so that it fit better in the frame and didn’t stick in the summer.
Even with all this, that bathroom door is still the most difficult-to-open interior door in my house. It takes more torque to release the bolting mechanism than any other door does. And even though Barnum has become quite accomplished with the other doors in the house, I hadn’t yet taught him this one because it presents an additional challenge due to the size and configuration of the bathroom.
So, until I taught Barnum how to open this door, I have mainly been dealing with the problem by almost never shutting the bathroom door. This doesn’t allow me a lot of privacy when my PCAs or other people are around, but I’d rather lose some privacy than get trapped in the bathroom. It’s so undignified! (And the location of this bathroom, combined with the very thick, insulated walls mean that when I do have to yell or pound for someone’s attention in there, it’s very hard to be heard.)
The reason this door was the last bastion of dog-door-opening difficulty is that I couldn’t use the same training technique I used with others. The way to make the job of opening a door easiest on Barnum is to have him approach the handle from the side furthest from the lever’s end, as opposed to pulling straight on. This way, he uses maximum leverage with minimal force to release the bolt. (Physics is your friend.) You can see this technique in action in the video below, where it takes less than three seconds for Barnum to open and exit the door. (From 0:03 to 0:06.)
Transcript of the video is here.
However, the master bathroom has a built-in cabinet right next to the door, so Barnum’s only options are to pull from the front or to pull from the lever-end side, which is even worse.
Here’s the bathroom door and the counter immediately on its left that prevents Barnum from getting good leverage.
So, I messed around with it for a while. I tried partially filling the latch hole on the theory that if the bolt had less distance to travel, it wouldn’t require as much torque to release. For whatever reason, that hasn’t worked.
Meanwhile, I started shaping* this behavior with a very high rate of reinforcement so that Barnum would be VERY EXCITED to open the door. I actually began with his favorite PCA sitting on my bed and only partially shutting the door, asking him to find her (as I previously discussed here and also here). This is Barnum’s Very Most Favorite Skill in the World. He LOVES to find people, get a treat from them, and then run back to find me. This also happens to be the most likely real-world application of this skill — if I’m in the bathroom and need Barnum to go get me help. So, I was tweaking the circumstances for maximum thrill.
Once Barnum was whining with excitement every time he flew at the door and tugged, I switched to just shaping a very enthusiastic approach to taking and pulling the cord. Then I shaped for longer holds and harder tugs. Occasionally, seemingly by complete chance, the door would fly open, but most of the time, Barnum was throwing his terrific enthusiasm (and considerable strength) into the job, without success.
I did notice, eventually, that the times that the door opened “out of the blue” did have something in common — Barnum was approaching from further away. So, I went back to my frenemy, physics, to try to figure out the problem. It seemed clear that Barnum needed to pull DOWN more BEFORE he pulled back. He also needed to approach as close as possible to, and parallel with, the cabinet. And there was something about approaching from farther away that helped. Shaping him to approach from the side was easy — I could manipulate each approach by where I threw the treat from the previous attempt. I realized eventually that the distance of the approach often simply meant a more enthusiastic, energetic pull. But why that was so crucial I still wasn’t sure.
I wanted to make the pulling easier on him. Someone on a training list I’m on once mentioned that a very long pull cord works better for her SD than a short one, so I switched to a long cord. That made things worse, which helped me realize that Barnum needed to choke up HIGHER on the cord to be able to pull down more easily. This wasn’t something I’d figured out with Gadget, who naturally had a tendency to grab high and who was also a bit shorter and more naturally wild/enthusiastic in his grabs. Eventually I realized that the two key ingredients were to shape Barnum to grab higher and to pull down hard at the beginning, versus a slow, steady pull that tended to be back as much as (or more than) down. That’s why the “running start” made a difference; he naturally tended to grab higher and pull down more when he was excited.
So, today I moved the knot higher up the pull cord (or tug strap, as some call them), and I tossed treats as far behind him as possible to get him coming at the door faster/further away and as close to the cabinet as possible. Success! Once he understood that grabbing up higher was the key, he was very excited about it. I jackpotted any time the door opened, not least because the door suddenly swinging open was a bit startling to Barnum the first few times.
Then, each time he opened the door, I had him run to find my PCA. Creating this behavior chain served two functions:
1. He loves this behavior, so it added value as a positive reinforcer for opening the door.
2. Most of the time when I really will need him to open the door, it will be to go find help, so it’s good to forge the links in this behavior chain now.
After a few rounds of this, Barnum was getting mentally fatigued (he was still extremely enthusiastic, but he was starting to get cues mixed up and just throwing behaviors at me), so I ended with BOTH the bathroom door and my bedroom door shut, which — again — most closely simulates what I will need in a real situation. He also has such a strong positive reinforcement history of opening my bedroom door to find a PCA that I thought it would be exciting to him.
Well, he did it! He opened the bathroom door. I said, “Where’s [person]?” And he raced into my room, whined with excitement in his hurry to get my bedroom door open faster than was caninely possible and found her. She praised and treated, asked him where I was, and he ran back to me! I was very proud and pleased.
I wanted to pet him or thump him on the chest in celebration, but he really does not like to be touched while in training mode, so I asked him for a “high nose,” which is the behavior I have settled on for when I want some celebratory physical contact at the end of a training session and he doesn’t want to be touched. I do a “high-five” position with my hand, and he bonks it with his nose (because even though I say, “High nose!” which means nothing to him, a palm facing him is our nonverbal cue for “touch”), and he gets a treat, and everyone feels good. (I have been giving a lot more thought to how and when Barnum wants to be touched and how we can both have our needs met and respected since I read this post by Eileen and Dogs.)
Of course, we will need to practice this and get the entire behavior chain on one cue (“Where’s [person]?” leading to opening both doors, finding and nudging the person, sitting down, waiting for the “Where’s Sharon?” cue and then returning) but I feel very confident that we are close to that now.
High nose!
– Sharon, the muse of Gadget, and Barnum, SD/SDiT
P.S. I know I haven’t been posting much lately. I have a lot of posts that are mostly done, and I hope to get back to blogging and other writing soon, which I will be filling you in on. . . .
* For those of you who are new to my blog or to clicker-training lingo, a few explanations/definitions:
Shaping, sometimes referred to as “free shaping,” is, in my opinion, the most creative, advanced, and fun form of clicker training because there is no prompting by the trainer. Instead, we use a dog’s offered behaviors and reward those that resemble — in tiny ways, at first — the end result we want. The dog has to do more thinking than in any other form of training. It is a step-by-step way for dog and trainer to problem-solve their way to a solution. In my experience, behaviors that are shaped are the strongest behaviors when they’re finished than those achieved by luring or other methods, possibly because they tend to involve such a high rate of reinforcement (sometimes referred to as RR).
Rate of reinforcement (RR) means, quoting Karen Pryor’s Clicker Training Glossary: “The number of reinforcers given for desired responses in a specific period of time. A high rate of reinforcement is critical to training success.” Here is a much longer discussion of RR and its importance in dog training.
After Gadget 