Tonight I did a little shaping work with her around the back door, trying to get her to stay put while I opened the door. This time, I left the other dogs in the basement and just took Mira to the door. I also grabbed a handful of treats to help her think more clearly. It occurred to me that de-escalating the release as the main reward might be useful. Since she is so food motivated, I thought getting her to work for food, instead of for being released, might change the dynamic.
Apparently I thought correctly! MIra was very keen on the treats and was much more focused on the task at hand. Since I also discovered that she can hold her sit easily on the doormat, but not two or three steps back from it, I started her there then shaped her to move further back. This worked quite well and she dramatically improved in just 2-3 minutes.
The other thing I discovered is that the trigger is not me stepping to the door, but me making an obvious movement to open it. That involved me reaching for the door AND looking at the handle. If I do either on its own, she wouldn’t react. Combine the two, and she’d pop up. Isn’t it amazing how much you can break down these behaviours?
So we practiced having her sit back from the door while I moved around, waved my hand towards the handle or looked at the doorknob. When she could hold for all three very well, I started doing combinations. That’s where she still had a little trouble, but she did much, much better than before. I expect just a few more sessions of this and we’ll have the behaviour I’m looking for. Just in time to head back home, I’m sure!
One other positive outcome of this sessions: by the time I released her, she really didn’t want to go outside anymore. Instead of exploding out the door, she stepped out and turned around, wanting to come back in again. Hurray! What a dramatic difference. I told her to go potty – which she did – and then let her back in. We then did a little more training and that was that.