In dog training, less may be more

It's true, I'm lazy. That's part of why I have sighthounds, because no one really expects you to do obedience or tracking or anything requiring, you know, serious training with a sighthound.



Which isn't to say my dogs aren't well-behaved. They walk quietly on a loose leash, they don't bark, go to the bathroom in the house, or eat the sofa. And I can easily train a dog to the extent required to trot around a show ring and stand pretty for the judge.



But I happily confess I don't have much interest in more formal training, and so I've been pretty happy sharing my life with a type of dog who doesn't, either.



Now, researchers in Denmark have discovered that at least one other breed of hound, the Beagle, learns best when he learns least often. In an article published in Applied Animal Behavioral Sciences, the authors found that dogs trained once a week learned specific behaviors better than dogs trained five times a week.



In a nutshell, 18 laboratory beagles were divided into two groups. The dogs were all trained by the same person, in the same unfamiliar room. The training was done using so-called "positive" methods, including shaping and clicker training.

The results of the study show that dogs trained once a week learned the shaping exercise in significantly fewer training sessions than dogs trained five times a week. In addition, weekly trained dogs tended to have higher success rates at the different steps of the shaping exercise than the dogs trained five times a week. The dogs trained five times a week completed the shaping exercise in significantly fewer days than the weekly trained dogs. It is concluded that for dogs learning a given skill, weekly training results in better learning performance than training five times a week, when performance is measured in the number of training sessions required to reach a certain training level.

My sighthounds approve of these findings about their scenthound cousins. You can read the whole article here (PDF).

Comments

[Luisa]

Christie writes: the authors found that dogs trained once a week learned specific behaviors better than dogs trained five times a week.



Block that extrapolation! This study involved 18 laboratory dogs, and was carried out in the interest of showing how little time needs to be spent on training animals kept in a laboratory setting. The study says absolutely zip about training companion beagles, stockdogs, S&R dogs, assistance dogs, etc., etc.



How often to train depends on the individual dog, the individual trainer, the task, the terrain, the weather, how the dog was raised and a whole lot else. It's nonsense to write, as the authors of this study have written in their abstract, that "surprisingly few studies have been undertaken to analyze the [training] process in detail, e.g. the question of how often training should be done has not been investigated in dogs." The subject has been given an extraordinary amount of thought, observation and testing.



I just blogged about this [and should have titled the post "No Beagle Left Behind," damn]. There's a lot more to teaching and learning than a set of quantifiable variables.

[Gina Spadafori]

Yes, well ... that's why my retrievers would drive Christie crazy when we visited at her old house the country. ("Don't they ever stop moving?") While her sighthounds strike me as extremely large, extremely sweet, long-legged cats.

[Arlene]

Yeah, I have my doubts that it would be similar for retrievers. My experience is that they LOVE the training (whereas when I took care of a friend's beagle, he had a lot of trouble concentrating on training, being highly distractable and independent).

[Alex V]

I wonder if results would be similar for other groups of dogs such as retrievers.

[The OTHER Pat]

Methinks there's a whole lotta' latent learnin' goin' on!

[H. Houlahan]

Oh joy. As if owners needed a "scientific" justification to never do their obedience class homework, and expect their dogs to magically progress when they are simply dragged into class once a week.



As limited as this result is, you know that's not how it's going to filter down to pet owners.



In addition to a single kind of unrepresentative dog, in a highly unrepresentative setting, the researchers were testing only one kind of training methodology (positive reinforcement operant conditioning) and one extremely simple task (touching a target with a paw).



My guess? The dogs who were "trained" five days a week were bored with the laboratory "trainer" and the uninteresting task. The dogs who were "trained" once a week were more bored with their boring everyday lives than they were with their boring "training sessions."

[Luisa]

But they haven't really "discovered that at least one other breed of hound, the Beagle, learns best when he learns least often." That depends on the Beagle, the trainer, the activity and so on. I'd be dishonoring the memory of my own Beagle if I didn't point that out.



If the researchers meant for their conclusion to apply only to the 18 laboratory raised and housed Beagles that were tested, then they should have made that clear -- instead of the sweeping "for dogs learning a given skill, weekly training results in better learning performance than training five times a week." Gah, I hate generalizations.

[Christie Keith]

Be fair, Luisa; in the sentence IMMEDIATELY before that I specified "one other breed of hound, the beagle." I am not the one who extrapolated out to all dogs... I think I framed this pretty clearly to be about a limited set of circumstances, whatever the authors of the study or anyone else may have intended.

[hornblower]

Clicker training my dog to touch a target with a paw took exactly 1 session of about 5-10 minutes, and she's not what anyone would consider a terribly smart or biddable dog.

Then we take the bhvr 'on the road' & increase distractions & distance - that does take a number of reps, but this is not a difficult skill.



I agree with H. Houlahan that what was being observed was utter boredom with the task. I think teaching the same thing everyday does get boring. I prefer to teach a skill, repeat it a few times throughout the day & then drop it for a few days or a week & do it again. And there are many dogs who will only repeat a certain task a few times before calling it quits. My female does this when searching for dropped objects - she'll do it a few times, she understands the task & can perform a grid search but after the 3rd time I drop something for her to find she looks at me like I'm being particularly difficult and dense & pretty much tells me to 'go get it yourself you dolt'....

[straybaby]

Hornblower, my girl is pretty much the same only her look says "get real, I've already proved I can do it. NEXT!" lol!~ we started amusing ourselves with whisper training and taking it out on the streets. She would get a big kick out of peoples reactions :)