people often confuse terms like "positive" and "negative" and I just wanted to clear up (or increase) the confusion.
Hopefully we will get some good contributions from the people who've studied this stuff and use it in their work.
The words "positive" and "negative" in training are often misused. I confuse them too.
B.F. Skinner | Operant Conditioning - Simply Psychology
Positive - doesn't mean "good". It means "you're adding something" as a consequence of the what the dog is doing. So you can
add a treat or praiseor
add a smack or scold.
negative - doesn't mean "bad" or "nasty". It means "you're taking something away"
you withhold or stop the rewards - or or
you take the naughty teen's phone away.
you can stop the jack hammer (take away the noise and vibration), or
you can stop pulling on the choke chain (release it). (take away the pressure) - also known as escape training.
So the positive and negative - is "good" or "reinforcing" or "bad" and "punishing" depending what you add or subtract. (trust scientists to use maths based not emotion based words).
Reinforcing - is something the dog (or human) finds encouraging and will work to get more of.
punishing - is something the dog finds discouraging and will work to avoid.
so you pair a behaviour with a positive reward (treat) - and ideally the dog will try to do more of the behaviour.
or you pair a behaviour with a positive punisher (smack) - and ideally the dog will try to do less of the behaviour - if the dog figures out the connection between the behaviour and the smack - they don't always do this successfully.
for example - of how the connection is harder to get right with punishers:
you are driving along with your dog barking in the back of the car - and a cop pulls you over and shoots your dog... would you risk driving at all ever again? Or would you know it was for speeding? and not the barking? or just a bad day for the cop? If the police siren goes off just as you plant your foot on the accelerator thinking faster faster - then maybe you'd make the connection. Timing is everything.
So it is true to say you don't want to be "always positive" - sometimes you want to withhold the reward to signify a behaviour you don't want to encourage. And with this phase of the training - you want to test your dog's understanding of the task - so you wave the lead temptingly in front of the dog - and if she leaves it alone - you treat, and if she grabs it you laugh at her, call her silly and don't give a treat.
So anyone who says they are "Purely positive" usually means they use the reward / treat method of training ie give a reward or withhold a reward parts of the operant conditioning quadrant (combinations of positive and negative with reinforcer and punisher - you get four squares in a box). They are also using that term to please people who think that positive is "always good" when in fact the positive side of the box - includes smacking the dog.
Or they say "purely positive" because they don't understand or are confused by operant conditioning quadrants or their customers are. Which is why I say "reward based training" not "purely positive".
with the rewards (something good) and aversives (something nasty)
what one dog (or person) finds good - isn't the same for another
and some things are more rewarding than others. Eg $10,000 is better than $10 for most people
and freshly cooked sausage is better than kibble for most dogs.
but not all. Some (dogs?) prefer to work for attention, or praise or toys (some toys are better than others)...
So if you want to use reward based (or even aversive based) training - you need to know what your dog likes, what your dog loves... and use these things appropriately. Do not hand over the thing the dog LOVES, when they've strolled back to you stopping to sniff at every bush and maybe roll in something stinky on the way back. Give them some praise, and maybe not a pat.
You can also use something called "response cost" or delay the reward... the dog has done what you want, but slowly? tell them good dog - but delay the reward...