May 27, 2011 at 9:10 am #64610
This is part of a presentation I have put together to introduce clients to basic reinforcement concepts before starting their training. I’d like to know what you think of the ideas and how they are presented. Our Dog Training Centre is just starting up in Hereford we’ve had many years experience in Dog Sport and now we’re turning our attention to pet dog obedience.
This is our training video (showing off what we do) http://www.youtube.com/watch?v=gJS10MJh-ec
Anyway, heres the theory!
How your dog learns
The pros and cons of positive and negative reinforcement
Part 1: Reinforcement
Every action has a reaction
Any reaction to a dog’s behaviour is reinforcement and will help to shape any future behaviour, good or bad.
Example: if a dog begs for food at the table today and you throw it a morsel, it will probably remember to do the same tomorrow.
Negative reinforcement is any reaction that is aimed at breaking down a behaviour.
Example: if you hit a dog when it jumps on the sofa, it might be nervous about doing it again.
There are many campaigns against using any form of negative reinforcement and we agree that using negative reinforcement as the only tool in teaching your dog will result in bad mental and physical condition, stress, tail chasing, nervousness and so on.
Positive reinforcement is any reaction to behaviour that encourages the dog to repeat the behaviour.
Example: stroking a dog when it brings you your slippers.
There are no campaigns against using positive reinforcement. However, there are just as many drawbacks when only using positive training methods as there are using only negative, though positive training has no media controversy associated with it.
Part 2: Using Reinforcement Effectively
Breaking down pre-conceptions and controversy.
Dogs are not like robots. There is no simple system where a certain input will always yield the same response.
Dogs have a psychology including emotions, instincts and desires all of their own.
To positively shape the relationship you want with your dog, you must learn the ways in which a dog thinks, learns and communicates.
This includes using positive AND negative reinforcement, which all dogs have evolved to understand from domesticated to working breeds.
To ignore one or the other would be like trying to form sentences using only half of the alphabet or painting a picture using only one colour, possible but unnecessarily difficult.
Why not train using only positive?
There are many people who have argued on TV, radio and internet forums that positive reinforcement is absolutely the only method that should be used when training a dog.
To rely on these methods requires that any reward mechanism you use is more desirable than any potential distractions to the dog.
Example: a common reward in positive reinforcement is the ball, which exploits a dogs desire to chase prey. If you have a problem where the dog won’t come back when called and an obedience trainer advises you to use the ball to bring the dog back, you are gambling on the notion that the dogs desire for the ball is greater than, say, a bitch in heat walking along the opposite side of a busy road.
The truth is that inevitably there will come a situation where a dog must obey it’s owner above any other desire, including a bitch in heat, for the safety and well being of the dog, the owner and the general public.
Correct correction. How negative reinforcement makes a positive dog.
You can use negative reinforcement in a number of ways, from a stern ‘no’ to more controversial pinch collars and other gadgets. NR (negative reinforcement) does not mean punishment! A dogs behaviour is always geared towards getting what it wants, be it food, security, a mate etc. In the wild it would learn what behaviour is acceptable through the pack leader, who would use it’s physical dominance to assert control.
The pack leader would appear to us to treat members of his family cruelly, but this form of language and social structure has evolved over millions of years to maintain the well being of the pack and is accepted, if not positively welcomed, by all dogs.
This is how NR should be used, a communication tool used to set clear boundaries and rules that the dog accepts and understands. NR is the only way to correct extreme desires (when used correctly and consistently) because the dog will know that there is no alternative behaviour to the boundaries you have set.
This in turn will boost your dogs confidence and happiness as it finally understands and accepts it’s place in your home.
Where does this leave positive reinforcement?
As stated before, for maximum effectiveness positive and negative reinforcement must be used in combination.
When you feel your dog has demonstrated that it is complying or has understood an exercise, it should be rewarded with whatever motivates the dog (food, toys, fuss etc).
Once a desired behaviour is established and the dog is working well, you should be able to work the dog for long periods of time with maximum concentration for less and less reward (depending on the type of dog).
Part 3: Train Consistently
Maintain control on the field and at home.
If you have understood and accepted the concept of using positive and negative reinforcement, it is now time to implement these methods in practice.
However, learning and using these methods on the training field is not going to mean anything to the dog in the home unless you maintain consistency in your behaviour towards the dog and its behaviour towards you.
Every part of the dogs life from entering and exiting a car to advanced heel work must operate under your rules and boundaries at all times, because the dog’s natural instinct is to test and, if appropriate, contest its position in the pack.
End of Presentation
What do you think? Do you :agree: or :nono: ?June 2, 2011 at 11:56 am #78979
A great idea but you do need to get your learning theory definitions correct.
By definition, reinforcement results in an increase in behaviour, whether that reinforcement is positive or negative.
Any decrease in behaviour is created, by definition, by punishment (whether positive or negative) or by extinction.
It is practically impossible to utilise only positive reinforcement. Again, by definition, if a positive reinforcement is not provided then negative punishment has taken place. Similarly, in order to apply a negative reinforcement, positive punishment must have been applied first in order for it to be removed.
Use of ‘no’ is not negative reinforcement. If it’s effective, it’s a conditioned positive punisher.
Oh dear, just got to the pack leader bit! You might need to update your understanding of this concept as it applies to dogs. It doesn’t really have any place in learning theory.June 3, 2011 at 9:14 pm #78980
This is good stuff thanks for the reply!
Its not me who does the training its my dad, and as brilliant as he is at training dogs he’s not so good with words!
I’ve attempted to translate basic training ideas into a format the vast majority of people can understand, but that is an art-form in itself and is proving to be so so hard.
What you’ve said helps though. I’m going to go back to the drawing board now.
Ash.October 26, 2011 at 3:26 pm #78981
soz …i know…but am catching up…
interesting post…as in poster posts re a presention re his dads knowledge??…well i have to ask myself…is there any hope?? 😮
marion…no mention of extinction…but …hey soz…that maybe an artform 😉
anyone seen the off the drawing board final version??November 9, 2011 at 1:11 pm #78982
Great introductory information, any beginner would benefit from this. Love the video and your website as well. 🙂
You must be logged in to reply to this topic.