Home › Blogs › Aidan's blog

Conditioned Reinforcers - Part 3

Aidan's picture
Sat, 08/11/2007 - 00:00 — Aidan

In the last two blogs in this series we looked at what conditioned reinforcers are and how to use them sensibly.

This time we're going to look at some 'out of left field' ideas for using them.

Something I've been working on recently is a training plan for using conditioned reinforcers in scent-work because I have experienced that scent-work is a very important and meaningful exercise for dogs and I want to make it easier to teach.

One traditional method of teaching tracking is to lay food on the track. Essentially you create lots of very short reinforced tracks. You gradually increase the length of those short tracks by placing food less often on the track.

One problem with this method is that inevitably the dog is learning to search for food scent at first. This is solved when the food is faded from the track, that is when the dog really starts learning to look for the scent of human or crushed vegetation (this is also where the training starts to break down, but fear not, a competent instructor can help you through that).

Steve White of the LAPD decided to refine his approach to teaching tracking by spraying an obvious water track that acted like a 'glue' to hold human scent in concentration on the track. He then used clicker training to shape the dog to follow the track. Right from the word 'go' the dog was learning to follow human scent, even on hard surfaces. In fact, Steve White started on hard surfaces and moved to grass later when he was sure that the dog had learned to follow human scent.

Steve White is a very clever man. If you are experienced with a clicker, know a bit about tracking, and are willing to measure your results I think you can't go past his method. This makes it particularly suited to professional police K9 handlers, or those hoping to earn a VST or FHIII later on.

Recently I have been working with an old idea of mine that I didn't get around to testing until now (and boy, do I wish I had done this before!) My idea is to condition articles on the track as conditioned reinforcers. This removes the food from the track, doesn't require you to have great clicker timing, and teaches articles right from the beginning.

I know, I'm a genius. It's only my humility that has prevented me from announcing this idea before now.

I think if you combine my idea with Steve White's idea you have a winning combination for novices, and right now I have a few people testing this plan for me. More on that later...

So what else can we do with conditioned reinforcers?

I have noticed that sometimes people can't get the hang of clicking their clicker at the right time, or don't know when to click. There is very little in the way of automated clicker training devices out there, but the Manners Minder (formerly Treat'n'Train) from Premier is a great example. This is a clicker and food dispenser that is remotely operated. It even has a sensor that knows if your dog is in front of the machine, and reinforces automatically on a schedule!

I have one of the older Treat'n'Trains which unfortunately was not very reliable, I am hoping that the new manufacturers (Premier didn't make the Treat'n'Train) have rectified these design faults.

I think technology in training is a good thing. I suspect as more and more people embrace positive training ideas we will see more positive training technology readily available. So long as people make some attempt to understand the theory behind the machines, this will be a good thing.