Smokeycat, as long as the movement to the treat bag happens after the click, then everything is ok. If we train with treats in our hand, there is a risk, and it very often happens, that the dog only does the behaviour if we have the treat in our hand. The "treat in hand" has become an integral part of the cue for the behaviour, which we don't want. If I'm training my dog an new behaviour, often I'll have the treat bag a few paces away, clcik the correct behaviour then immediately go and get the treat. It cleans the cue-behaviour sequence. Kay Laurence has some really nice stuff on this.
Princelord - there is a world of difference between this approach and what Cesar does, in fact they are polar opposites. We might achieve the same result, as far as the dog's behaviour appears, but it's how we go about it which is important to me. I can teach you to willingly lie don on the floor or I can punch you in the temple to get there. Same result, two completely different methods and attitudes, especially if you''re the one who is doing the lying down.
Just to clear up also, the terms positive and negative reinforcement are scientific terms with very specific meanings. Positive reinforcement is adding something after the behaviour ossurs to increase behaviour, negative reinforcement is removing something after behaviour occurs to increase that behaviour. The use of corrections as commonly applied is called positive punishment - adding something after behaviour occurs to reduce that behaviour.
It may feel like semantics in this and the previous post addressing Season, but it's really not. I don't mean to be arrogant at all when I say that, but the science surrounding this isn't just important, it's essential.