Ranting sadly

SUMMARY: "Balanced training" vs "Positive training"

Update Jan 22 11:30 a.m: List at end of some studies that show the results on dogs and on their relationship with their owners for reward vs punishment-based training.


For all the stuff that I've done with dogs, I'd never heard the phrase "balanced training," which came up in an online discussion. I did some reading via Google. And--what a bunch of uninformed people! Here's one page--http://www.precision-dog-training.com/dog-training...

"Balanced" uses both positive reward and positive punishment and seem to believe that negative punishment isn't a valid training method.  Here's their preferred positive punishment method: "startle tactics-- A simple, humane leash and collar correction (which is tight for only 1 second) is given."

SIDEBAR: Definition of terms. In operant conditioning, you can have 4 types of responses to a behavior (I swear that I wrote this before but can't find it): Positive reward: You give them something that they want.  Example: Sit and you get a treat. Negative reward: You take away something that they don't want. Example: I'll stop shocking you as soon as you Come to me. Positive punishment: You add something that they don't like. Example: A jerk on the leash. Negative punishment: You take away something that they want. Example: Dog jumps up on you, you turn away and ignore them (taking away the attention that they want)  until they put all 4 feet on the ground, then you use positive reward of your attention. So "positive punishment" means that you are explicitly doing something negative to the dog when they don't do what you want them to. Also called aversives. (I just wanted to be clear about that.)

From their page--one of the completely nonsensical quotes:

The problem with the all-positive training method is that you can never teach your dog to obey you simply because you are the "pack leader". Touch-free training, clicker training... they all involve coaxing, luring with treats and simply put - pleading for your dog to listen. 

Oh what a complete misunderstanding! I don't plead for my dog to listen! That's no way to train a dog! I give them opportunities to do what I want them to do and I reward it. And I don't need to be pack leader--I just need to be the person in charge like a parent is in charge, and I need to communicate clearly to my dog what I want. And I don't have to hit them or shock them or jerk them around to do it.

Another quote, this about why "balanced training" is supposedly better:

The dog is trained to respond to your command immediately, the first time you say it. That's right! No more chanting the commands (stay, stay, stay....), no more looking silly when your pooch ignores you and runs away." [stating that "all-positive" training causes these bad results]

Holy moly, more crap--I have competed in agility with four dogs, earned multiple championships, and you betcha that I needed a reliable stay so that I could lead out ahead of my fast beasties--I never chanted "stay stay stay"--and that they responded to my verbal commands immediately. I'm also stunned by her description of what "positive training" schools looked like (if you read it). Wow. I'm appalled and sad that someone who doesn't fully understand operant conditioning is training other people the same way (and maybe the classes that she observed anonymously weren't experts, either--most of what she described I've never seen in any of the many classes or seminars I've taken).

Negative punishment in my view is far better than positive punishment. I'd rather ignore a dog that's doing bad things to get my attention and let them figure out how to behave and reward that--now I have a thinking dog who has figured out what he needs to do and will abandon the behaviors that don't work. It's that simple.

What you get out of the training is what you put into it, no matter what method you use. I've come to believe that hurting or jerking dogs is more than just cheating; it's the lazy way of training that also can have negative effects on your relationship with your dog. When I worked at it, my prey-driven dog would call off of a running squirrel, and all I used was treats and praise. No shock collars, no jerking. No "proving that I was the boss/pack leader," just proving that I was more interesting than running squirrels.

As someone in the discussion board said:

In general it's way better to set a dog up for success and reward the success, rather than set the dog up for failure and punish failure. The dog decides what is aversive. If adding something unpleasant (positive punishment) is sufficient enough to make a dog stop doing something, then it was aversive. Some dogs might enjoy a spritz with water. Other dogs might completely shut down. Another dog might not care either way.

A stern "no" might not be a big deal to some dogs. But can be a huge deal for others.
That's why we just set dogs up for success so we don't need to use it.

I pick the things that I'll concentrate on with each of my dogs at different times, it's true. Don't expect perfect angels by your own definition when you meet them. But they're smart dogs and they learn quickly when I apply myself.

Sorry, I think hat I'm ranting. If you can train small children without "startle tactics" or any other aversive, then you can train dogs the same way. And chickens, and horses, and dolphins, and pigeons--  there is plenty of research to prove it.

So, there you go. I've ranted out now.

SIDEBAR: Some research (if you're not already tired of reading) showing that dogs trained with positive rewards/reinforcement tend to have lower stress levels and react more positively towards their owners than punishment-based: Training methods and owner–dog interactions: Links with dog behaviour and learning ability Effects of 2 training methods on stress-related behaviors of the dog (Canis familiaris) and on the dogeowner relationship Dog training methods: Their use, effectiveness and interaction with behaviour and welfare