Asymmetrical Information

"but he hasn't exactly come up with Unified Field Theory or anything."

Well, how do you know? What if your dog had a flash of insight, solved the problem, but didn't know how to convert it into English? Probably would be frustrating. Has your dog ever seemed upset? There you go. Probably figured out Unified Field Theory, but can't understand why you can't understand Dog.

Here's an interesting article that might apply:
http://chronicle.com/free/v51/i32/32a01201.htm
and here is some smarties talking about it:
http://chronicle.com/colloquy/2005/04/dogs/

Someone who has taken Statistics 101 might want to explain to Professor Cowen that the difference between 99 and 95 % is not very significant. But hey, we already know the austrians don't let data stand in the way of their ideas.

Well, consider that (even though the article seems to be aimed at dog owners) the word actually used was "pets", not "dogs". Perhaps the % who think their pets don't communicate all had goldfish or something.

I couldn't understand someone who didn't think their dog (or cat, for that matter) didn't communicate with them.

Hmm, maybe if we replaced the word "pet" with "significant other" the study would make more sense . . .

Even if we take the numbers at face value (assume the differences are not due to statistical noise), they are still amazingly close. The real story here is not the difference in response between men and women but our overall amazing attachment to our pets.

Communicating with dogs is easy, but it happens at a subverbal level. Dogs are very emotional creatures and they're not that hard to understand. They also wear their emotions on their sleeves, so to speak. I certainly find my dogs a lot easier to understand than my wife.

What I find fascinating is that while we communicate with our dogs, in many cases we don't realize what we're saying to them and they don't realize what they're saying to us. Much of a dog's communication to humans is related to the dominance/submission behavior of an animal that runs in very heirarchical packs -- but those communications are easily mistaken by humans for affection. Much of a human's communication to dogs involves affection but is easily mistaken by dogs for dominance/submission behavior.

By lucky chance, evolution made dogs and humans perfect for each other. Dogs think humans act like the dominant members of a joint dog/human pack and follow and obey them accordingly. Humans think dogs act like little furry children and love and care for them as a result.

Because of that, I can't know for sure if my dog truly feels love or affection for me. But I hope he does and I sure love him.

"Someone who has taken Statistics 101 might want to explain to Professor Cowen that the difference between 99 and 95 % is not very significant."

I haven't taken 101, but I did manage 321 (Applied Statistics), 317 (Engineering Probability & Statistics), and 510 (Probability & Statistical Models). Without the benefit of 101, I would say that the difference between 1% and 5% (the percentage who don't frequently talk to their pets) IS significant.

I talk to the cat all the time - "outta my way, furball", "what? what? you're the most annoying cat in the univserse, shut up!", and "yes, I know, you're bored, what do you want me to do about it?".

A tale of communication between my son, my wife, and our cat ...

http://www.livejournal.com/users/p_o_u_n_c_e_r/6772.html

In a similar study, 0% of pets talked back to their owners...

95% vs 99%, or 1% vs 5% -- (statistical) significance all depends on the precision of the estimator.

Now, whether that's significant in a larger 'do we care' sense is another matter. Here, splitting it as 1 vs 5 makes it look more significant, 95 vs 99 less so. Spin and such.

Heck, yes, the cats communicate with us. If nothing else, they come in the bedroom or stand at the top of the steps every morning and make a lot of noise until we go feed them.

I agree with Megan. Some people may not think of this as "communication," because that has become a rather loaded concept having connotations of relationships and feelings and feelings about relationships.

If the difference between 95-99% isn't significant, then the difference between 1-5% in the same sample won't be significant either (in either the statistical or commensensical meaning of the term).

The difference bewteen 99% and 95% isn't statistically meaningful. So, whoop-dee-do on that meaningless stat.
On the other hand, when I had goldfish, I don't think I ever talked to them, and I don't think they ever 'communicated' to me. Not even to clean the tank.

"By lucky chance, evolution made dogs and humans perfect for each other. Dogs think humans act like the dominant members of a joint dog/human pack and follow and obey them accordingly. Humans think dogs act like little furry children and love and care for them as a result."

So, what's the deal with cats? My cat owns me, not the other way around.

Also, has anyone seen that reality tv show about crazy dog owners? It's on Bravo. Those people really just need to have kids. Those poor dogs. I wouldn't be shocked to see one of them put the dog in a bonnet in a baby carrier and feed it with a bottle.

Don't trust goldfish. They lie. My wife's fish tell her I haven't fed them even when I have.

My brother's fish decided they'd tell my brother they hate him so they all committed suicide by flinging themselves out of the tank. No, really.

I have to admit I don't know what the deal is with cats. Apparently evolution has made cats and humans perfect for each other -- humans like to read books and cats like to curl up on top of books while humans are reading them.

Animals experience emotions, and they convey those emotions through both body language and vocalization. Of course, the cat who finds you in the easy chair, meows, rubs your legs and then jumps up for a lap-nap is merely expressing some sort of happiness related to a variety of sensory inputs and a memory of previous sensory inputs -- not expressing a complex critique of your wardrobe choices.

But it is most definitely communication.

"I have to admit I don't know what the deal is with cats. Apparently evolution has made cats and humans perfect for each other -- humans like to read books and cats like to curl up on top of books while humans are reading them."

Cats also like to kill mice and rats, and humans like to avoid the plague.

And here I thought that when I was avoiding something "like the plague" I was just avoiding it the way the plague would have avoided it. ;)

I once had a dog with which I could communicate my desire for a hops-seasoned beverage from my small refrigerator on the patio, and it would smartly respond to said communication, avoiding the odd cola-flavored beverage that found it's way into the cooled receptacle. Best damned dog that ever trod the planet (sniff).

Talk to my five cats? Are you kidding? They speak---and I listen. Who in hell do I think I am? We low life human beings must accept our humble place in the universe.

Cats are selfish creatures and I respect that.

The only cat we ever had who spoke to us spoke Spanish. I kid you not. Snozzlegoster had a cry which sounded distinctly like "agua" and if we responded to it by filling the bowl he always drank.

Sebastian, whose picture I posted recently, is our one indoor-outdoor cat (because he came from a household more supportive of going out than is our own) and has learned how to rattle the screen door quite loudly when he wants to come in.

My cat mainly uses her powers of communication to inform me of my many inadequacies as a servant.

"If the difference between 95-99% isn't significant, then the difference between 1-5% in the same sample won't be significant either (in either the statistical or commensensical meaning of the term)."

That's correct. It is significant. The difference between 95 and 99% is big. Looking at the 1-5% shows how significant. Most people think of 4%/95%, call it 1/20th, and think it's not significant because it's samll. The difference between 48 and 52% is much less important.

If we're going for situations of human-dog communication here, here's a story. One of my cousins lived on a farm, and it was a long walk from the house to the letterbox. So he decided to train the dog to fetch the newspaper every morning. So every morning he would take the dog down to the letterbox, train it to get the newspaper out of the box and carry it back to the house.

Eventually, one morning, dog spontaneously turns up with newspaper in mouth. Big excitement, congratulations, treats for dog. Dog enjoys this all immensely. After a bit, dog disappears. Dog returns 1/2 an hour later with another newspaper.

This communication failure is very similar to what often happens between humans.

Currently the puppy is frantically scratching at the back door, clearly communicating that he wants back in the house. And I'm trying to think of a clearer way of communicating to him that holding it in while outside and then putting a huge pile in the middle of the living room is exactly backwards...

"I couldn't understand someone who didn't think their dog (or cat, for that matter) didn't communicate with them."

Obviously I put one too many negatives in there. Blast.

As far as whether 95 vs 99% is significant, there's no way to know since the story doesn't give the confidence interval or experimental error numbers. It could be significant; it could be non-significant. There's not enough information to know.

"As far as whether 95 vs 99% is significant, there's no way to know since the story doesn't give the confidence interval or experimental error numbers. It could be significant; it could be non-significant. There's not enough information to know."

Since I don't have the benefit of Statistics 101, maybe one of you have taken this class can solve this.

The sample size is 901. Without any further information, assume that half are of each gender. So ~450 men and ~450 women. When you have a poll of 450 with a 95% result, what is the margin of error with 95% confidence interval?

Comments are Closed.