I finally got to go to ClickerExpo and it was an incredible learning experience! You should go if you can, or buy the VODs.
A Moment of Science: Clicker Training 101, with Kathy Sdao
Intermittent schedules of reinforcement reinforce a subset of behavior. Outside of a lab their only use is to make a behavior resistant to extinction; for example increase the frequency of the behavior in your absence (SAR dog alerting). Use continuous R+ to make a behavior resistant to change.
Uh, what? You mean all this time I’ve been worried about putting behaviors on a VSR and I DIDN’T NEED TO??? My mind was blown by this. I went back to our hotel room to walk Ava with Josh and spend the whole time enthusiastically gesticulating while exclaiming how relieved I was by this discovery. Continuous reinforcement is easy. VSR are irritating and use too much of my brain.
Building Behaviors: Shape the Future, Laura VonArendonk Baugh
I tend to lump behaviors and forget that my dog isn’t a mind reader. They have no idea that my goal is for them to, say, put their front feet on that pot and pivot their rear in a circle. They’re just sitting there like, “uh, is this thing over here important?”
One of the key reminders I got from this session is that the key to amazing shaping is being a good observer. The video below shows the puppet made for the Warhorse theatrical performance, and it moves so realistically because the puppeteers spent so much time observing horses moving that they can replicate it exactly. This is the kind of observation required for effective shaping. If I’m shaping Touch I need to look for movement in the neck and shoulders, not the nose.
The other thing is that I never knew what to do if my dog skipped a step. What I learned is that it depends. If he skips a step due to luck and doesn’t make progress from there, go back to what he knows. If he skips a step and keeps progressing, keep going! He didn’t need those steps in between.