for Carrie, as promised
I had the opportunity to go to clicker expo and see Steve White who is a police dog trainer. He was fabulous. He started his career in dogs in the military and police using fairly scary compulsion methods. He now uses clicker training type methods , although obviously not always actually using a clicker since it's hard to hold a leash, a gun, a flashlight and a clicker
He is still training police dogs but also training police dog trainers to use
non-compulsive methods.
His results have been staggering- they are training dogs with a much higher degree of reliability, a higher percentage of dogs complete the program and dogs are completing the program and out working faster. He tours the country as a consultant to other K9 units. He also consults to service dog programs. His website is www.i2ik9.com incase anyone wanst to check him out.
Steve talked about a variety of things. One of the things he talked about was the problems that arose in training dogs with traditional methods for police work and tracking. He had what he called the 8 rules for using aversives - which basically shows how hard it is to use aversives in a useful way, that actually teaches the dog something and without huge negative consequences.
so...The 8 rules for using aversives (in order for them to work in training)
1) It must be something the dog doesn't like and does not expect.
--are collar pops unexpected ? heck no
2) It must actually suppress the behavior
-using aversives that don't work is at best nagging, at worst abuse
3) It must be applied in the perfect intensity
Too much / too harsh = dog shuts down and can't learn
Too little / not harsh enough = development of a "punishment callous" actually teaches dog to become imune to punishment and "tough it out"
4) Must be IMMEDIATE
or it won't be associated with the behavior you're trying to suppress
5) Must be associated with the behavior
and not you-- other wise you get a dog whose behavior is suppressed in your presence, but continues when you aren't present.
"sneaky dog syndrome" = the dog associates the punishment with YOU more so than the behavior.
Steve commented how dogs often don't understand what they are being punished for but how they are remarkably good at figuring it that YOU are involved or connected somehow- even when using hands off tools like electronic collars.
6) Must happen EVERY TIME the behavior does
-otherwise you are just putting the misbehavior on a variable reinforcement scheduel which just makes the behavior stronger and harder to extinguish
7) There must be an alternative behavior trained or available to the dog
- trying to punish away a behavior that is really strong doesn't work unless you replace it with another behavior
ie punishing a dog for jumping up when he greets isn't usually successful unless you reinforce an alternate greeting behavior like sitting - the dog is social and has a strong desire to greet.
8) The aversive must never outweigh the reinforcement in the dogs mind
-or the dog "checks out" and learning shuts down
Steve also talked about teaching crossover dogs (dogs that had been originally trained with force and aversives) and about teaching 'advanced decision making skills' to dogs. Very interesting stuff.
I had the opportunity to go to clicker expo and see Steve White who is a police dog trainer. He was fabulous. He started his career in dogs in the military and police using fairly scary compulsion methods. He now uses clicker training type methods , although obviously not always actually using a clicker since it's hard to hold a leash, a gun, a flashlight and a clicker
He is still training police dogs but also training police dog trainers to use non-compulsive methods.
His results have been staggering- they are training dogs with a much higher degree of reliability, a higher percentage of dogs complete the program and dogs are completing the program and out working faster. He tours the country as a consultant to other K9 units. He also consults to service dog programs. His website is www.i2ik9.com incase anyone wanst to check him out.
Steve talked about a variety of things. One of the things he talked about was the problems that arose in training dogs with traditional methods for police work and tracking. He had what he called the 8 rules for using aversives - which basically shows how hard it is to use aversives in a useful way, that actually teaches the dog something and without huge negative consequences.
so...The 8 rules for using aversives (in order for them to work in training)
1) It must be something the dog doesn't like and does not expect.
--are collar pops unexpected ? heck no
2) It must actually suppress the behavior
-using aversives that don't work is at best nagging, at worst abuse
3) It must be applied in the perfect intensity
Too much / too harsh = dog shuts down and can't learn
Too little / not harsh enough = development of a "punishment callous" actually teaches dog to become imune to punishment and "tough it out"
4) Must be IMMEDIATE
or it won't be associated with the behavior you're trying to suppress
5) Must be associated with the behavior
and not you-- other wise you get a dog whose behavior is suppressed in your presence, but continues when you aren't present.
"sneaky dog syndrome" = the dog associates the punishment with YOU more so than the behavior.
Steve commented how dogs often don't understand what they are being punished for but how they are remarkably good at figuring it that YOU are involved or connected somehow- even when using hands off tools like electronic collars.
6) Must happen EVERY TIME the behavior does
-otherwise you are just putting the misbehavior on a variable reinforcement scheduel which just makes the behavior stronger and harder to extinguish
7) There must be an alternative behavior trained or available to the dog
- trying to punish away a behavior that is really strong doesn't work unless you replace it with another behavior
ie punishing a dog for jumping up when he greets isn't usually successful unless you reinforce an alternate greeting behavior like sitting - the dog is social and has a strong desire to greet.
8) The aversive must never outweigh the reinforcement in the dogs mind
-or the dog "checks out" and learning shuts down
Steve also talked about teaching crossover dogs (dogs that had been originally trained with force and aversives) and about teaching 'advanced decision making skills' to dogs. Very interesting stuff.
Last edited:
