My Photo


follow us in feedly

Enter your email address:

Delivered by FeedBurner

Best Of KC Dog Blog

Become a Fan

« Weekly Roundup - Week Ending 3-10-13 | Main | Very good sentences...What I'm reading today... »

March 13, 2013



Joel, please explain how the splitting of data techniques used in classification and regression trees (which look at each independent variable separately) can possible estimate the influence of one variable while controlling for the influence of others. They cannot, by design. By design, they will mask the influence of correlation among independent variables. They are useful, but, again, they are your hammer.


PS Sorry, Brent, I was going to play in this sandbox. But.

I wonder how the pro-"pit bull"-BSL gang feels about "huskies" (also not a breed but a shape)? In Canada, where I live, they are by far the leading shape involved in DBRFs; in all but one case the victims have been children.

Is that due to breeding?

We have had one fatality in Canada, ever, attributed to the "pit bull" type. Are our "pit bulls" wimps, or what?

I think it's safe to say that mixed breed dogs are involved in the most biting/mauling incidents overall.


While I'm at it, nobody is actually verifying the 'breed' when a dog is involved in a DBRF.

If a Lab x Collie bites someone (pretending that mutts are F1s which is actually rare), is it the Lab or the Collie? Or is it maybe just a DOG?

The basic flaw in all of this, on both sides, is that when you combine a handful of breeds, over a dozen lookalike breeds and an unknown number of mongrels (the most popular type of dog in the US), the whole 'breed'-based behaviour thing falls apart.

A recent experiment by Maddie's fund showed that even those who are very familiar with breeds of dogs differ significantly when asked to visually ID a mixed breed dog's ancestry.

In a population of roughly 78 million dogs, 30 or so DBRFs a year, while horrific, are not enough to predict any kind of trend around dogs, dog breeds or dog behaviour.

What we look for are ownership practices to try to discern whether there are common factors that lead to bites, attacks and fatal maulings.

Focusing on the dog's shape is misleading and does a disservice to the public at large, espcially the non dog-savvy among us.


Dubv, sorry I'm not going to put in several years of study into anything. But I'm not going to be conducting any dog attack studies either. I'll read your paper if you post it though.

Instead of trying to show everybody how smart you are, why don't you make the proper analysis a project of yours? Seems like you're only going to accept your own methodology anyway.


"A recent experiment by Maddie's fund showed that even those who are very familiar with breeds of dogs differ significantly when asked to visually ID a mixed breed dog's ancestry."

I've seen it.

It is a very bad study. I'll explain why. I am a trained scientist and so am fit to evaluate these things. Every strong statement I make below can be confirmed independently if you so wish.

DNA fingerprinting in order to determine paternity or presence at a crime scene is HIGHLY accurate. Therefore, people get it in their heads that anything using DNA is like this. The identification of breed components using DNA relies upon a totally different approach to the data. It often uses mixture modeling (a math technique) and is much less accurate. It is blatantly stated by the test manufacturers that the tests ONLY work for mixed breeds dogs. That has something to do the mixture models used, but I have a more cynical view: it is so the results cannot be adequately checked easily. If they worked on purebreeds, then the DNA test could be shown to give crazy results. Please take time to watch videos of people reading off the results from the most popular testing company (MARS)where people laugh when the company claims that their Great Dane is really a chi.

So, POINT 1: the accuracy of DNA breed testing is questionable and has not been confirmed. To confirm this, an independent study would need to be done on the particular technique used by the Maddie project to determine the accuracy of the DNA testing. This would involve, at the least, testing mixed breed dogs of known origins (those cases where a registered bassett hound knocks up a registered lab) and see if they are accurate.

Maddie's project, instead, assumes that the DNA tests are accurate and shows that visual identification does not match the DNA results. In reality, the visual match may be more accurate. The project could've done something like brought mixed breeds of known origin in front of breed identifiers and then tested the same with DNA techniques. They did not.

Also, MARS and other testing companies state that they have not been able to build a breed profile for pit bulls. This might have political reasons (they do not want their test used for BSL) or it maybe for the reasons they state.

Further, it unclear whether or not the breed identifiers knew that they were in a study designed to test whether or not pit bulls can be visually ID'd. The test should have been double blind. It is unclear whether or not it was.

Lastly, the only reference I see to this study is a poster. It has not past peer review. And the references on the poster, etc. indicate a leaning on the part of the researchers and vets are NOT formally trained in research while receiving their DVM, so it is no wonder they would make mistakes.


I see you've been tripping around the inter-tubes. It sounds as though you might have read some of my stuff LOL

The reason they can't build a profile for "pit bulls" is that it isn't a breed, as they state. American Pit Bull terrier is a breed but those registries did not donate samples to the genome project as the AKC did. But what they are really saying is that they cannot identify mixed breeds. That concurs with the scientific evidence to date.

However, I wasn't talking about the DNA aspect at all, since in my opinion (and I admit I'm only skimming posts here) the study showed that two unvalidated methods disagreed, not that either one was better.

My point, again, was that even experienced dog people, when asked to guess the ancestry of mixed breed dogs, disagreed - a lot - about what that might be. That was the only worthwhile takeaway point. (And it's not the only example out there).

Therefore, the chances that dog bite victims, bystanders and others would even achieve that level of accuracy are slim at best, especially given that eyewitnesses are notoriously poor at reporting detail.

Maybe if you weren't nitpicking, looking for tiny holes, you might actually understand what people here are saying and see the big picture.

Brent, what was the name of that person who used to hang around, from Florida, claimed to be a researcher or something. Was it Doug? This guy reminds me of those heady days.

I'm finis, this is the usual waste of time.


Yeah Selma, I'm going to have a quick hook on Dub from here on out.

Dub, here's the bottom line. Last week, a 7 year old boy was killed by a dog. While a family mourns, I think it's fair to ask the question of 'why did this happen?' and "what can we do to prevent it from happening again?"

In the process, the vast majority of the experts in the animal professional community, including the vast majority of trainers, animal handlers, veterinarians, and animal control officers will point to the facts that the dog was raised in a household where violence existed (the male owner was convicted of murder), that the dogs were malnourished, neglected, and chained in the back yard.

You would have us believe the attack happened because the dogs were pit bulls -- even though there are 5-8 million pit bulls in this country that are not responsible for attacks like this. That not only defies expert opinion, but also common sense and the depth of knowledge and research on canine behavior that exists.

Even you acknowledged that the majority of the expert community is opposed to your stance. I'm all about challenging the status quo, but in order to do so, you must have the experience, expertise, or knowledge to challenge it.

But you don't. You're someone who comments annonymously online. And not only as anonymous commenter can you not have credibility, you have now posted 20 times in the past 52 hours (many I've deleted just to clean things up a bit, because they were not adding to the intelligent dialogue, or because that amount of posting is just excessive) and the only outside knowledge you've brought to the table was an article by a hack editorialist out of Canada.

So instead, instead of challenging the plethora of knowledge that exists within the science and expert community with facts and expertise, you have done what the community has consistently done in comments sections of news sources around the world -- post frequently, excessively and anonymously hoping that by hijacking the comments sections of media outlets and shouting from the rooftops loud enough you can somehow add somehow legitimize a point of view that completely lacks credibility right now.

The victims deserve better than this -- and they deserve intelligent conversation from the expert community, and they deserve it from people who are willing to put their name and personal records on the line and stand behind their point of view.

I'm done letting you use this blog space as a platform for you bully pulpit when you're merely acting as a spamming internet troll


"I'm done letting you use this blog space as a platform for you bully pulpit when you're merely acting as a spamming internet troll"

Convenient for you.

"But you don't. You're someone who comments annonymously online. And not only as anonymous commenter can you not have credibility, you have now posted 20 times in the past 52 hours (many I've deleted just to clean things up a bit, because they were not adding to the intelligent dialogue, or because that amount of posting is just excessive) and the only outside knowledge you've brought to the table was an article by a hack editorialist out of Canada."

Brent, if someone is incorrect in their reasoning, then no amount of experience will change that they are wrong in that instance. One person that is anonymous can show them to be wrong. That is simple reality. You are using circular reasoning. I am attacking the premise that these so-called dog experts are actually framing this topic correctly. I feel an outside perspective can show that they are not. These dog experts are investigating human fatalities, but they are all heavily invested in dogs, enough to make them their vocation. That is telling. I am questioning whether the weight of their expert opinion is deserved here. This is a foreign concept to you?

Further, outside knowledge is not needed when one is commenting on the framing or reasoning of another's argument. I am poking holes in the general thinking of you and others. I can do that by simply reading and seeing what goes wrong.

Go ahead and block me. Let the echo chamber continue.


Dub - -it's not a handful of people you think you're poking holes in. It's essentially everyone with knowledge of canine behavior. I mean, it's trainers (on both sides of the spectrum), it's scientists, its humane organizations, it's animal handlers, it's veterinarians, it's animal control officers -- it's nearly everyone who has working knowledge of animal behavior. And it's important to note that it's not group-think with these groups - -most of these groups exist because they're notorious for not agreeing with each other. And yet. On this subject, they're nearly unanimous.

You're not poking holes in it. You're commenting over and over hoping something sticks until you get to a point where you're arguing scientific methodology instead of canine behavior and the impact human-created conditions can have on that.

And while you may think it convenient I keep you on a very tight leash, it's convenient that a site you are a regular on, who is one of the few sites that share your point of view, is notorious for not allowing any counter-points, creating the worst of the worst in group think and excluding expert thinking. I think allowing 15 comments here has more than showed my willingness to listen to all sides of an argument. Your side does not. Nor does it bring expertise in canine behavior. Or even real names and faces. It's just a series of anonymous internet profiles.


Well, dang. My car mechanic is heavily invested in cars, enough to make them his vocation. Reckon that's telling... next time my car breaks down, I'll sure avoid going to him...


Emily FTW!


Dubv, if you want a civil conversation you should removethe evil, nasty comments on what I am guessing is your own website.

You criticize DNA tests yet use journalists' passing comments in a news article as proof of breed of dog?


Dubv is a pretty prime example of the use of babbling jargon instead of, you know, actual fact-based analysis.


Brent, are you man enough to at least let me comment that I do not own a blog? Someone here accused me of having a blog that contains nasty comments. I predict you'll delete this also, but let people continue to refer to me and make accusations.


DubV -- if you have a point in posting, and it's civil, it can stay. But the rambling, obsessive commenting and overall disrespect isn't welcome.

Nancy Tranzow

Ummm...Dub? Are you SURE you know about genetics? Do you truly believe that genetic phenotype expression is changed via genetic drift? I don't mean to insult but when was the last time you were in a lab by chance.

You leave out a great deal of the story. First, there have been no genes or combination that have been found to promote or infer aggression in canines. Second, regardless of genotypic expression, environment plays a tremendous role in phenotypic expression with the exception of basic appearance.

One does not need to throw around big words, nor cut and paste from scientific literature to point out the most basic of concepts, nor to understand them. Clearly, your lack of knowledge is obvious in your need to pound your pseudo-intellectual chest. I am always amused when people like you come into my lab. It's cute. Sad to be sure, but cute.

Verify your Comment

Previewing your Comment

This is only a preview. Your comment has not yet been posted.

Your comment could not be posted. Error type:
Your comment has been posted. Post another comment

The letters and numbers you entered did not match the image. Please try again.

As a final step before posting your comment, enter the letters and numbers you see in the image below. This prevents automated programs from posting comments.

Having trouble reading this image? View an alternate.


Post a comment

Your Information

(Name and email address are required. Email address will not be displayed with the comment.)