In Part 1, we went through the sorts of things dog-owners often confuse with training, so it is time to look at some data on actual dog training.
Except there isn’t much to look at. From a scan of professional dog-trainer web-sites you can see that the drop-out rate in group classes is a large concern, but that’s about it in terms of quantitative analysis.
Except for this one study from Wichita State University. It is called Selecting Shelter Dogs for Service Training. It’s a well done study on the potential of shelter dogs (of which we have many) to be trained as service dogs (of which we have not enough). Before the dogs could be fully evaluated for their potential as service dogs, they had to be obedience trained.
So, in this study, 4 selectors (they select and test service-dogs for a living) chose 75 dogs from the Kansas Humane Society in Wichita. The selectors walked through the kennels and chose dogs based on breed, age, and size. They chose the breeds that are thought to be easiest to train and have the most potential value as service dogs – retrievers, collies, shepherds, rottweilers and mutts thereof.
These dogs then spent 5 weeks on a farm with highly experienced dog trainers, learning basic obedience. Sit, stay, come. That sort of thing.
At the end of the 5 weeks, 73 of the 75 dogs – when tested on the farm where they were trained – met the basic obedience test standard for those commands.
Then the testers brought the dogs to a PetsMart.
At PetsMart, each dog was again given the obedience test, only now the dogs had to deal with some real-world distractions, and an environment that was less familiar.
At PetsMart, only 44 of the 73 met the same obedience test standard. The others failed basic obedience.
Let’s recap: Experienced dog-training professionals selected 75 dogs – not at random, but from breeds thought easiest to train – and those dogs spent 5 weeks on a farm with those experienced trainers, drilling every day in basic obedience. After 5 weeks, nearly all passed the controlled-environment obedience test, but of those only about 60% passed the real-world basic obedience test.
We are left to wonder what the real-world basic obedience test pass-rate would have been had the pool of dogs included hunting dogs, and sled dogs, and terriers.
We are left to wonder what the real-world basic obedience test pass-rate would have been had the trainers been mere dog-owners, not professional trainers.
We are left to wonder what the real-world basic obedience test pass-rate would have been had the dogs been offered not 5 weeks of intensive training, but the less rigorous training a dog would expect to receive from a normal family taking the dog to obedience classes.
Perhaps most alarming is that the real-world basic obedience test pass-rate was so different from the controlled environment (farm) pass-rate. It suggests that many, possibly most, graduates of dog obedience class have learned nothing more than a (literal) parlor trick, that cannot be relied upon in the real world, when it counts.