Tue Aug 8 08:49:18 PDT 2006

Reinforcement Timing Explained

Last night was another successful evening as a group-lesson trainer. AS will always interfere with my ability to make social adjustments on-the-fly with other people, but there are moments of brilliance when I feel really good about my ability to distill complex ideas into something palatable to other people.

Last night had a few such moments. One of the most interesting ones was when I was explaining to Robin why I thought a number of the students were having trouble with mastering the lessons.

"It's all about timing," I said. "Timing is the second-most important skill for reinforcement training of any kind. Here, let me prove it to you. What's two plus two?"

Robin hesitated only a fraction of a second, clearly trying to figure out what I was up to, but promptly replied with "Four."

I didn't respond immediately. I just gazed at her steadily for about five seconds or so, deliberately letting the silence drag on. "Right?" she asked, clearly trying to fill the uncomfortable pause.

I let this pregnant moment drag on another few seconds, until the it was stretched to its breaking point. "You're right," I said. "It's four. So, do you feel rewarded?"

She grinned and laughed, having gotten my point immediately. The demonstration had made a very visceral impact with her, as I think it would with most people.

Clearly, there's a window of opportunity when using praise or other reinforcers for training. If you wait too long, anxiety and uncertainty creep in, and the subject is left twisting in the wind.

Some books I've read put the maximum effective window at about 30 seconds when using a conditioned bridging stimulus such as a clicker, but I'd guess that this window is actually a lot smaller for the average non-clicker trainer, and even smaller yet for those who use aversive training methods.

Explaining the importance of timing, both in marking an event and in delivering the reward, is a hard lesson to teach the average student. Most people can't deal in abstracts; human examples with gut-level, emotional reactions are probably much more accessible. In the long run, teaching the average person on an emotional level, rather than an intellectual one, might be more effective.

Posted by Todd A. Jacobs | Permalink