September 08, 2007

Basic skills of clicker training

What: Criteria
Before you start a training session, decide exactly what you are going to reinforce. Unless you know what you want, you’re unlikely to get it! Remember to teach in tiny increments, rather than expecting the whole finished behaviour all at once. The smaller the increments the more successful your dog will be at working the behaviour out for himself – and that produces the best learning.

When: Timing
The click marks exactly the moment that the dog performs the desired behaviour. It tells the dog “That’s what I want. That’ll get you a reward” You will find that you have to anticipate the dog’s action so as to click on time. It is better to click too soon than too late.

Where: Placement
While the clicker is the primary marker, everything that happens between the click and the treat is being reinforced. If you click a sit then your dog gets up for the treat, getting up without a cue is being reinforced. So where possible, reward the dog in position, or use the delivery of the treat to reinforce the behaviour. Eg click backing up, and toss the treat under the dog’s belly so she has to back up to get it.

How Much: Rate of Reinforcement
Keep the rate of reinforcement high, particularly in the early stages of teaching a new behaviour. Your dog knows how to sit. What you are teaching him is that it is worth his while to sit when you give the cue – so make it worth his while! As you raise the criteria you will be demanding more of your dog for less frequent reinforcement.


Anonymous said...

nice! the info is very relevant. you may find this website on dog travel supplies interesting. thanks!

Richard Laplante said...

Good point. This same method is used for teaching children and special needs kids. Hey wait a minute, that's the same tactic my wife uses on me!