Dog breeds by state-interactive map

[michele5611]

www.huffingtonpost.com/2012/03/27/dog-breeds-by-state-inter_n_1382740.html

[catstina]

What's that whole "Pit Bulls make up only 3% of the population" thing the anti-Pit crowd always quotes, again?

[emilys]

Kim Wolf asked them where they got their statistics from
btoellner.typepad.com/kcdogblog/2012/03/american-pit-bull-terriers-among-the-most-popular-us-dogs.html

"I emailed the reporter to ask how they got the data. It's from 950,000 vet records to which Vetstreet has access. Issues with breed ID aside, 950,000 dogs represents less than 2% of the total canine population of 78+ million living in the USA as of 2011. So while it's nice to see that American Pit Bull Terriers (which they consider to be the same breed as American Staffordshire Terriers) are well represented here, I'm not confident in these numbers at all. Plus, they only looked at people who take their pets to vet offices, which as we know does not include a substantial portion of pet owners. It's not that I don't want to see "strength in numbers" when it comes to pit bull ownership (however we're defining "pit bull"), but if we're going to cry foul on "dog bite statistics" studies that use flawed or incomplete data, then we can't have it both ways and use these figures either.
Posted by: Kim Wolf | March 28, 2012 at 12:58 PM"

[catstina]

I feel like vet records is a pretty good way to do it. And I think it's a pretty good sample size compared to many statistical studies. It may not be the best study possible, but it certainly seems to be one of the best available.

[johnr]

Mar 28, 2012 20:17:58 GMT -5 catstina said:

I feel like vet records is a pretty good way to do it. And I think it's a pretty good sample size compared to many statistical studies. It may not be the best study possible, but it certainly seems to be one of the best available.

Unless the sample size is virtually as big as the population size, bigger is not better if it is skewed. Two skewing factors: A) Going to the vet at all, which sadly is not a high priority amongst a lot of people who not only own Pit Bulls but breed the hell out of the; B) Records intentionally concealing rather than revealing true breed identity - even my own Pit Bull loving vet, after years of knowing me, Pablo and my Pit Bull advocacy, asked me if it was "okay" to put Pablo down as a Pit Bull on his last rabies cert. Note that even a complete census of the population would not be representative if there is massive mis-IDing, intentional or not.

[catstina]

I didn't even think of your B point, but that is so true. Especially in areas with BSL. The first Pit Bull I ever knew, Belize, was listed as a Boxer mix with her vet because she lived in Miami.

[emilys]

sample size is less relevant IF the sample is random.
This sample is not.
Even aside from the "does anyone know how to identify a dog" issue. Yes even the owners

[melonie]

I know when I took Lilly into see the vet, he mentioned that she looked like she had a little pit bull in her. I know her parents were both pit bulls, so that tells me that he doesn't know the breed enough to identify one accurately. I could have said she was a full blood something or other or a mix of anything but pit bull labeled breeds and likely gotten away w/ no questions asked.
When Winston first went in to see the vet, I had to tell them what he was, I was asked his breed. Again, I could have said anything. So maybe vet records are only going to be as accurate as the dog owners own breed ID claims.
Also, Ellie's records are odd in that on her rabies cert. it says pit bull mix. Yet her other records labeled her as a sporting breed.

On the other hand, if I lived in a more populated area where vets have a broader canine experience, maybe I wouldn't have been asked what the breed of the dogs were.

[odnarb]

The only thing I liked about this was the fact that it admits that Pit Bulls are a popular breed. Every other one I've seen uses AKC registration statistics. I assume that it's taken from vet records or, more likely, licensing statistics.

Sent from my DROID BIONIC using ProBoards

[emilys]

Apr 1, 2012 15:37:54 GMT -5 odnarb said:

The only thing I liked about this was the fact that it admits that Pit Bulls are a popular breed. Every other one I've seen uses AKC registration statistics. I assume that it's taken from vet records or, more likely, licensing statistics.

Sent from my DROID BIONIC using ProBoards

" vet records to which Vetstreet has access" according to what Kim found.

So all I would describe this as would be "dogs identified in vet records by breed". Leaves out all the vet records NOT included in Vetstreet's "access". Leaves out all dogs without vet records. The problem with this data is the same as with dogbite statistics: without knowing the size of the entire population, you can't know if this sample is representative. It is certainly not a "random" sample which is what you need for a statistic to be reliable.

[johnr]

Apr 1, 2012 11:07:15 GMT -5 melonie said:

I know when I took Lilly into see the vet, he mentioned that she looked like she had a little pit bull in her. I know her parents were both pit bulls, so that tells me that he doesn't know the breed enough to identify one accurately. I could have said she was a full blood something or other or a mix of anything but pit bull labeled breeds and likely gotten away w/ no questions asked.
When Winston first went in to see the vet, I had to tell them what he was, I was asked his breed. Again, I could have said anything. So maybe vet records are only going to be as accurate as the dog owners own breed ID claims.
Also, Ellie's records are odd in that on her rabies cert. it says pit bull mix. Yet her other records labeled her as a sporting breed.

On the other hand, if I lived in a more populated area where vets have a broader canine experience, maybe I wouldn't have been asked what the breed of the dogs were.

There is a widespread assumption that vets are good at breed IDing. They are often are not. It's not part of their training and many really don't care. Similar to the assumption that a vet can give an expert behavioral eval just because s/he's a vet. Not so.

[catstina]

I was super impressed by my vet when she recognized Saxon as a Stafford the first time she met him. I guess it was because she spent time in the UK.