When you use your training diary, you really have to give a lot of thought to what you mean by your cue. If you don’t know, how is your dog supposed to figure it out?
A question was asked about about doing unreinforced repetitions of a behaviour such hand touches. I wasn’t sure whether the handler meant that one touch is a behaviour, and three touches in a row, with only the last one being rewarded means unreinforced repetitions, sometimes called “going for three-fers”) OR whether she meant that that one cue (“touch”) means do it three times.
The difference is, the first is getting three behaviours for one reward, but each behaviour is cued. The second means one cue asks for several behaviours in a row (and therefore is given one reward, obviously when the behaviour sequence is finished).
I don’t mean to be obscure, but I had this conversation with my friend Richard not long ago, in relation to his mini Poodle Fido speaking. Ask yourself, “what does ‘speak’ mean?” Bark once? Bark twice? Bark three times? Continue barking until asked to stop? There’s no correct answer. It depends on your purpose. What are the current obedience trial requirements? A couple of strong barks? So, whatever it means, one cue means “do it”. For film work, sometimes “bark and go on barking until asked to stop” is more useful behaviour to put on cue. Some people use a single “fingers flash” hand signal to cue that behaviour. Others hold their hand open and the dog should bark until the signal stops (close and lower hand).
Getting several hand touches or barks for one cue is like training for duration. Theoretically, if you do it by shaping, you just gradually increase the duration (as you would with the time of the stay) and reward at the end. But I tend to make use of an intermediate (a quiet “good”) which is called a Keep Going Signal (KGS). Shaping purists would say it’s not necessary to use a KGS, but I don’t think it does any harm. It may help the dog to get over the confusion that can arise when you up the ante. Actually, I do it more for the sake of the handler I am instructing – “lie down – good – good – good – reward, release.” This helps the handler to measure out the length of time. Notice that I make use of a release cue. This will also help the dog to understand that not all rewards signify the end of the behaviour. Basically, the way I like to teach an exercise (especially a control exercise) is that the cue means “do it and go on doing it until released”. So I start this from a puppy sit.
1. go into the sit position
2. go on sitting for longer, and
3. don’t get up until released. This way, the handler doesn’t need to repeat the “sit”, and learns not to say “stay” when all they mean is “go on sitting” without being a jack-in-the-box.
How often have you heard someone sat “my dog will sit but she won’t stay”? All they mean is that the dog hasn’t really learnt to sit, but rather to touch down and take off again. which is not very useful.