Cdc Data On Dog Attacks

[sandgrubber]

I keep coming across anti-BSL discussions saying this or that data are biased and the CDC data show that breed does not predict propensity to attack. I was just on the CDC site and the only dog attack data I could find was a paper by Sachs et al that is rather damning to the pitti and rotti. http://www.cdc.gov/ncipc/duip/dog50.pdf

Where is the data that are supposed to be unbiased?

[bryan_mannix]

I keep coming across anti-BSL discussions saying this or that data are biased and the CDC data show that breed does not predict propensity to attack. I was just on the CDC site and the only dog attack data I could find was a paper by Sachs et al that is rather damning to the pitti and rotti. http://www.cdc.gov/ncipc/duip/dog50.pdf

Where is the data that are supposed to be unbiased?

Yep and here is the data through to 1998 from the same author.

[sandgrubber]

A well written paper . . . thoughtful conclusions. The data shown tend to make me angry with the anti-BSL people who jump on others saying that 'any breed can bite' and APBTs are no more dangerous than a Labrador, etc.

There does seem to be a problem with bull terrier types and Rottis. I . . . and a few other breeds. I agree with Sachs that BSL is not the way to address it. But denial is a bad idea. And the worse the denial, the more likely it is that society will resort to BSL.

I keep coming across anti-BSL discussions saying this or that data are biased and the CDC data show that breed does not predict propensity to attack. I was just on the CDC site and the only dog attack data I could find was a paper by Sachs et al that is rather damning to the pitti and rotti. http://www.cdc.gov/ncipc/duip/dog50.pdf

Where is the data that are supposed to be unbiased?

Yep and here is the data through to 1998 from the same author.

Edited March 24, 2010 by sandgrubber

[poodlefan]

Sandgrubber:

There does seem to be a problem with bull terrier types and Rottis.

Ask yourself if its just the breed or the types of people who buy them and how they are treated..

Karen Delise's book Fatal Dog Attacks concluded that the reasons a dog are acquired, such as image enhancement or 'protection' contribute to the liklihood of it being involved in such attacks. I cannot recommend that book too highly.

Consider also that nearly all research on dog aggression concludes that breed identification is questionable in the data.

Any breed CAN bite. However propensity to bite (bite threshold) and the amount of damage a bite can cause vary significantly between breeds. I've only ever been bitten by one dog - a Pekingese. I doubt anyone would compare a bite from that dog with one from a large powerful breed and conclude that the danger was the same.

Another excellent paper on the subject of dog aggression to people and which includes some discussion of breed, can be found here

Labrador Retrievers and Golden Retrievers rate a mention. Bear in mind that breed popularity plays its part in how frequently such dogs rate in statistics.

Edited March 24, 2010 by poodlefan

[sandgrubber]

Sandgrubber:

There does seem to be a problem with bull terrier types and Rottis.

Ask yourself if its just the breed or the types of people who buy them and how they are treated..

Karen Delise's book Fatal Dog Attacks concluded that the reasons a dog are acquired, such as image enhancement or 'protection' contribute to the liklihood of it being involved in such attacks. I cannot recommend that book too highly.

Consider also that nearly all research on dog aggression concludes that breed identification is questionable in the data.

Any breed CAN bite. However propensity to bite (bite threshold) and the amount of damage a bite can cause vary significantly between breeds. I've only ever been bitten by one dog - a Pekingese. I doubt anyone would compare a bite from that dog with one from a large powerful breed and conclude that the danger was the same.

Another excellent paper on the subject of dog aggression to people and which includes some discussion of breed, can be found here

Labrador Retrievers and Golden Retrievers rate a mention. Bear in mind that breed popularity plays its part in how frequently such dogs rate in statistics.

Breed popularity and the extent to which the dog is likely to be left unsupervised with kids -- some of whom will torture a dog, and many of whom will encourage rough play. My experience in placing Lab pups and getting feedback from puppy-buyers show there is a problem with people thinking that Labbies and Goldies are benign nannas that can be trusted with kids. I've had two who reported child-biting. Both allowed and encouraged rough play and left kids unsupervised in working out their relationship with the puppy --> dog. Run-scream-puppy chase seems fun to the kids, but it easily teaches a dog to bite, not out of meanness, but as an element of being inadvertently trained to treat children as they would a sibling or other young dog. I have rewritten my 'puppy care' sheets to warn against this, but lots of people ignore or forget instructions, so I don't think I've solved the potential problem.

Please read the Sachs et al article. Their conclusions carefully and strongly qualify the 'anti-breed' reflex that may come out of the data. They speak out against BSL. But the statistical evidence, even if corrected for various biases, is VERY strong. I have done a lot of work with statistical evidence, and rarely find anything as damning as the data in Sachs et al. is to pittis and rottis. Even if corrected for likely biases, this data would be scary. I find it irritating when anti-BSL campaigners come out and say that the statistics are 'crap'. I have yet to see any set of dog fatality statistics that doesn't show a few breeds to be especially problematic, even where the people compiling the data are neutral, and interpret the data away from the obvious breed-specific conclusions. Denial of a problem usually allows the problem to grow.

Edited March 25, 2010 by sandgrubber

[Sticky]

Have a look here:

http://www.cdc.gov/HomeandRecreationalSafe...-factsheet.html

S

[sandgrubber]

Sandgrubber:

Ask yourself if its just the breed or the types of people who buy them and how they are treated..

I never said or came close to saying it is just the breed. Nor would I. I am not dumb and I have observed a lot of people and their dogs. I know environment affects behaviour. Please don't insult me.

Edited March 25, 2010 by sandgrubber

[poodlefan]

Sandgrubber:

Ask yourself if its just the breed or the types of people who buy them and how they are treated..

I never said or came close to saying it is just the breed. Nor would I. I am not dumb and I have observed a lot of people and their dogs. I know environment affects behaviour. Please don't insult me.

That was never my intention.

[justin19801]

If people bothered to read the study they would find it was based on media reports. Given that this dog has been called an APBT, APBTx, SBTx, APBTxSBT which media report would they have used to include this attack in the study? Furthermore they state that any shorthaired muscular dog and and longhaired black and tan can be labelled an APBT or GSD. They at worst only say there may be a potential problem with the breeds. Apart from using a decade old study which CDC has now walked away from nowhere in any of their recommendations do they list breed as a determinant of aggression. In fact they have discounted it completely with Dr Gilchrist the head of that section actually saying "unless there's a study I don'y know about there is no breed problem with aggression" (paraphrased for obvious reasons. All this study proves is that a little knowledge in the hands of the ignorant can be dangerous. E-mail CDC and they will tell you the same. They take a while to answer because of the number of queries they get but they will get back to you. Unlike the R$PCA they are not donation driven and try to stop dog attacks, not make money from them.,

[sandgrubber]

I pay US taxes, and I'd call the CDC driven by involuntary donations . . . but that's a different issue. I don't know Gilchrist from a bar of soap and have not had good experiences in emailing a Center with 1000+ employees and expecting a straight answer. I'll give it a try. But won't hold my breath.

I don't see indication that the CDC has backed away from Sachs . . . who is the first to admit the biases of his study and draws highly tempered conclusions.

I once worked with futurist predictions based on computer models. Data are always dicey when it comes to important issues.

I would love to see data that the anti-BSL people consider 'good'. But until clean data are available, I take the best data I can find and regard it as potentially/probably skewed against certain breeds by media bias. Still doesn't look pretty for the bull breeds and rotti. Unfortunately, this agrees with my extensive personal experiences with dogs (I run a boarding kennel) and I'm inclined to think there is something to the data, even if it is biased.

It p's me off to see thread entries like Decided To Check Nsw Dog Attacks Proving BSL is based on rubbish! . . . and see NO proof whatsoever that the data are rubbish. Biases, sure. Most data are biased. But the leading data are collectd by professionals with attempts to compensate for bias. It p's me off to hear people calling other people's observations crap and accuse them of being ill informed and biased. Bottom line seems to be there are no good unbiased data and we are all biased by our own personal beliefs and experiences. The anti-BSL community, in my observations, are looking strongly like they are in denial. Mind you, not the Rotti community. The pedigree/show Rotti people I know recognise there is a problem and manage their breedings, dog contaimen, dog training, and puppy placements with respect of the problem. But a lot of bull breed people seem to think the way to solve the PR problem is to attack the messenger. Not smart.

If people bothered to read the study they would find it was based on media reports. Given that this dog has been called an APBT, APBTx, SBTx, APBTxSBT which media report would they have used to include this attack in the study? Furthermore they state that any shorthaired muscular dog and and longhaired black and tan can be labelled an APBT or GSD. They at worst only say there may be a potential problem with the breeds. Apart from using a decade old study which CDC has now walked away from nowhere in any of their recommendations do they list breed as a determinant of aggression. In fact they have discounted it completely with Dr Gilchrist the head of that section actually saying "unless there's a study I don'y know about there is no breed problem with aggression" (paraphrased for obvious reasons. All this study proves is that a little knowledge in the hands of the ignorant can be dangerous. E-mail CDC and they will tell you the same. They take a while to answer because of the number of queries they get but they will get back to you. Unlike the R$PCA they are not donation driven and try to stop dog attacks, not make money from them.,

Edited March 25, 2010 by sandgrubber

[Sticky]

But the statistical evidence, even if corrected for various biases, is VERY strong. I have done a lot of work with statistical evidence, and rarely find anything as damning as the data in Sachs et al. is to pittis and rottis. Even if corrected for likely biases, this data would be scary. I find it irritating when anti-BSL campaigners come out and say that the statistics are 'crap'. I have yet to see any set of dog fatality statistics that doesn't show a few breeds to be especially problematic, even where the people compiling the data are neutral, and interpret the data away from the obvious breed-specific conclusions. Denial of a problem usually allows the problem to grow.

As statistical evidence goes the data in the Sachs paper is next to useless, firstly there is no denominator and secondly there are too many 'unknown' cases. Without a denominator its not possible to come to any meaningful conclusion on the relative risk from different breeds. This becomes more evident when you look at data from other countries where different breeds are popular. In Canada you find large numbers bites from retriever breeds and in Europe you get more from GSDs. Other papers have attempted to correct for this both here and the US. The results are fairly inconsistent but generally risk seems to be more related to the size of the dog, ie bigger dogs bite harder not the breed. There are also too many other more important factors to become preoccupied with breed alone.

S

[sandgrubber]

Denominators are easy to supply . . .

and the conclusions generally get stronger when data are normalised cause the alleged problem breeds are less common than retrievers in Canada or GSD's in Europe.

Yes, the data are ugly. But I doubt they are off by a factor of 10: Especially if you normalise them by the abundance of the breeds in question.

But the statistical evidence, even if corrected for various biases, is VERY strong. I have done a lot of work with statistical evidence, and rarely find anything as damning as the data in Sachs et al. is to pittis and rottis. Even if corrected for likely biases, this data would be scary. I find it irritating when anti-BSL campaigners come out and say that the statistics are 'crap'. I have yet to see any set of dog fatality statistics that doesn't show a few breeds to be especially problematic, even where the people compiling the data are neutral, and interpret the data away from the obvious breed-specific conclusions. Denial of a problem usually allows the problem to grow.

As statistical evidence goes the data in the Sachs paper is next to useless, firstly there is no denominator and secondly there are too many 'unknown' cases. Without a denominator its not possible to come to any meaningful conclusion on the relative risk from different breeds. This becomes more evident when you look at data from other countries where different breeds are popular. In Canada you find large numbers bites from retriever breeds and in Europe you get more from GSDs. Other papers have attempted to correct for this both here and the US. The results are fairly inconsistent but generally risk seems to be more related to the size of the dog, ie bigger dogs bite harder not the breed. There are also too many other more important factors to become preoccupied with breed alone.

S

btw, if you want some Australian data that are normalised by breed prevailence try this http://www.dlg.nsw.gov.au/dlg/dlghome/docu...June%202005.pdf Rottis come off fine but pittis are still indicated.

Edited March 25, 2010 by sandgrubber

[geo]

Mind you, not the Rotti community. The pedigree/show Rotti people I know recognise there is a problem and manage their breedings, dog contaimen, dog training, and puppy placements with respect of the problem. But a lot of bull breed people seem to think the way to solve the PR problem is to attack the messenger. Not smart.

In your experience, 'not the rotti community'. Good for you and those rotti owners, it's slightly different everywhere, in the UK rottis have the bad rep.

I hope all puppy placements are with respect to all dog issues.

I think bull breed people just get fed up with all the sensationalism, Most staffy, pity, pigdog owners i know are well aware of their shortcomings.

Lets see how the stats look when all bites get reported, a playful nip from a lab is just that, but god forbid a playful nip from a staffy and that's it the whole world knows about the killer down the street.

I agree bull breed owners need to take a different approach, more awareness is needed when it comes to owning all large breeds, infact all dogs.

Stats don't really do it for me, everyone has an opinion (even statisticians), anything that could be a lab cross generally becomes a pity cross when it comes to reports, thats why i take all info with a degree of scepticism, but i'm not denying any possible flaws in bull breeds.

[justin19801]

CDC have looked at the stats and found they are meaningless, primarily for the denominator. After 20 years of studying the topic I'm sure they would have found a correlation if it existed. The NSW study is meaningless. I'm sure the woman whose limbs were eaten by a Rotti in the UK would beg to differ as would the parents of the child killed by a Rotti in Vic in 12/08.

Everytime BSL is introduced dog attacks increase. SA, UK, Ohio Italy Netherlands the evidence is well and truly in. BSL doesn't work. Calgary is the most successful north American municipality in regards to dog control. No BSL just making people responsible for the actions of their dogs. It seems that is what really scares the BSL supporters.

[Sticky]

This is probably the best Australian study on this. In this study breed id has internal consistency which most don't ie the breed id came from the same source.

http://www.mja.com.au/public/issues/aug4/t...n/thompson.html

BTW there are no pitbulls in this study.

Thompson repeated the study a few years later, with different results (different methodology)

http://www.uam.net.au/PDFs/PUB_Pro04_Thomp...sionEffects.pdf

S

Edited March 26, 2010 by Sticky

[whippets]

A well written paper . . . thoughtful conclusions. The data shown tend to make me angry with the anti-BSL people who jump on others saying that 'any breed can bite' and APBTs are no more dangerous than a Labrador, etc.

There does seem to be a problem with bull terrier types and Rottis. I . . . and a few other breeds. I agree with Sachs that BSL is not the way to address it. But denial is a bad idea. And the worse the denial, the more likely it is that society will resort to BSL.

I keep coming across anti-BSL discussions saying this or that data are biased and the CDC data show that breed does not predict propensity to attack. I was just on the CDC site and the only dog attack data I could find was a paper by Sachs et al that is rather damning to the pitti and rotti. http://www.cdc.gov/ncipc/duip/dog50.pdf

Where is the data that are supposed to be unbiased?

Yep and here is the data through to 1998 from the same author.

I'm still on the fence with BSL. I'm not sure which way to go but with the current situation with bull breed/crosses attacks the public demand action.

It's amazing that when people see statistics that don't suit their purpose then they'll pick it to pieces (as in what sandgrubber posted) yet if they see one where bull breeds/crosses arn't in the top 10 (havn't seen one of those yet) they'll wave it around for the whole world to see.

The tenacity and aggression in the APBT"S (again I only mention this bull breed because it's the one I am most familiar with) could be reduced substantially through proper breeding practice but from the 1980's imports till now that is not many generations. PLUS the breed didn't have the opportunity with BSL. Excepting of course the underground breeders which again won't do the breed any favors by breeding APBT's illegally. I'm sure the ones that frequent here hate me mentioning it and hate it every time I do.

When I don't agree with what was mentioned in the "news" forum, instead of addressing my point:

"APBT's are not for the "average joe" dog owner.

Not ONE person addressed that. Instead posts get selectively misquoted and insults such as

"bog ignorant"

"uneducated and illiterate"

then t-shirt quotes of "judge the deed, not the breed"

People who post on these forums need to be aware that the journalists read these forums as do the general public. I can't remember if it was the Geelong Adi or the Herald Sun but the Australian dog forums were referred to in their newspaper article.

If the DoL people are the same people that are at the forefront of anti-BSL campaigning then they havn't got a hope in Hades of changing BSL legislation with their attitude.

[justin19801]

This is probably the best Australian study on this. In this study breed id has internal consistency which most don't ie the breed id came from the same source.

http://www.mja.com.au/public/issues/aug4/t...n/thompson.html

BTW there are no pitbulls in this study.

Thompson repeated the study a few years later, with different results (different methodology)

http://www.uam.net.au/PDFs/PUB_Pro04_Thomp...sionEffects.pdf

S

The Thompson data are what I believe to epitomise Junk Science when it comes to breed id. The self description of breeds (if you accept that then you would have to accept the Cockapoo as a purebred when owners describe them as such) the numbers involved (does anyone really belive nearly 7% of owned dogs are EBTs) probably explains why it morphed to BTs (all types) leads to comparing half a dozen different breeds to individual breeds, statistically irrelevant. A good example of poor stats being worse than no stats.

Despite this he still claimed Pit Bulls (whatever they are) were the worst of them all because someone told him so. i.e. he didn't even trust his own stats. I know because I talked to him.

BSL or more correctly Breed Discriminatory legislation are not about dogs but about social control (govts) fundraising (R$PCA) and the end of pet ownership (PETA, Humane Society). For many it is about racism as the white anglo middleclass orgs victimise the coloured lower income groups traditionally attracted to these dogs. They can't openly discriminate against them but they can kill their dogs.

[justin19801]

A well written paper . . . thoughtful conclusions. The data shown tend to make me angry with the anti-BSL people who jump on others saying that 'any breed can bite' and APBTs are no more dangerous than a Labrador, etc.

There does seem to be a problem with bull terrier types and Rottis. I . . . and a few other breeds. I agree with Sachs that BSL is not the way to address it. But denial is a bad idea. And the worse the denial, the more likely it is that society will resort to BSL.

I keep coming across anti-BSL discussions saying this or that data are biased and the CDC data show that breed does not predict propensity to attack. I was just on the CDC site and the only dog attack data I could find was a paper by Sachs et al that is rather damning to the pitti and rotti. http://www.cdc.gov/ncipc/duip/dog50.pdf

Where is the data that are supposed to be unbiased?

Yep and here is the data through to 1998 from the same author.

I'm still on the fence with BSL. I'm not sure which way to go but with the current situation with bull breed/crosses attacks the public demand action.

It's amazing that when people see statistics that don't suit their purpose then they'll pick it to pieces (as in what sandgrubber posted) yet if they see one where bull breeds/crosses arn't in the top 10 (havn't seen one of those yet) they'll wave it around for the whole world to see.

The tenacity and aggression in the APBT"S (again I only mention this bull breed because it's the one I am most familiar with) could be reduced substantially through proper breeding practice but from the 1980's imports till now that is not many generations. PLUS the breed didn't have the opportunity with BSL. Excepting of course the underground breeders which again won't do the breed any favors by breeding APBT's illegally. I'm sure the ones that frequent here hate me mentioning it and hate it every time I do.

When I don't agree with what was mentioned in the "news" forum, instead of addressing my point:

"APBT's are not for the "average joe" dog owner.

Not ONE person addressed that. Instead posts get selectively misquoted and insults such as

"bog ignorant"

"uneducated and illiterate"

then t-shirt quotes of "judge the deed, not the breed"

People who post on these forums need to be aware that the journalists read these forums as do the general public. I can't remember if it was the Geelong Adi or the Herald Sun but the Australian dog forums were referred to in their newspaper article.

If the DoL people are the same people that are at the forefront of anti-BSL campaigning then they havn't got a hope in Hades of changing BSL legislation with their attitude.

APBTs are not naturally aggressive. Even dogs from fighting lines lose their gameness after three generations. BDL solves nothing, everywhere it is introduced it results in increases in dog attacks. Of course compassion is only shown to victims of alleged APBTs. If people truly were interested in solving the dog bite epidemic they would look at jurisdictions such as Calgary, the most successful in North America.

BDL is about ending pet ownership, not reducing dog attacks. This link was exposed by Vicky Hearne the author of Bandit: Dossier of a dangerous dog nearly two decades ago.

[bryan_mannix]

A well written paper . . . thoughtful conclusions. The data shown tend to make me angry with the anti-BSL people who jump on others saying that 'any breed can bite' and APBTs are no more dangerous than a Labrador, etc.

There does seem to be a problem with bull terrier types and Rottis. I . . . and a few other breeds. I agree with Sachs that BSL is not the way to address it. But denial is a bad idea. And the worse the denial, the more likely it is that society will resort to BSL.

I keep coming across anti-BSL discussions saying this or that data are biased and the CDC data show that breed does not predict propensity to attack. I was just on the CDC site and the only dog attack data I could find was a paper by Sachs et al that is rather damning to the pitti and rotti. http://www.cdc.gov/ncipc/duip/dog50.pdf

Where is the data that are supposed to be unbiased?

Yep and here is the data through to 1998 from the same author.

I'm still on the fence with BSL. I'm not sure which way to go but with the current situation with bull breed/crosses attacks the public demand action.

It's amazing that when people see statistics that don't suit their purpose then they'll pick it to pieces (as in what sandgrubber posted) yet if they see one where bull breeds/crosses arn't in the top 10 (havn't seen one of those yet) they'll wave it around for the whole world to see.

The tenacity and aggression in the APBT"S (again I only mention this bull breed because it's the one I am most familiar with) could be reduced substantially through proper breeding practice but from the 1980's imports till now that is not many generations. PLUS the breed didn't have the opportunity with BSL. Excepting of course the underground breeders which again won't do the breed any favors by breeding APBT's illegally. I'm sure the ones that frequent here hate me mentioning it and hate it every time I do.

When I don't agree with what was mentioned in the "news" forum, instead of addressing my point:

"APBT's are not for the "average joe" dog owner.

Not ONE person addressed that. Instead posts get selectively misquoted and insults such as

"bog ignorant"

"uneducated and illiterate"

then t-shirt quotes of "judge the deed, not the breed"

People who post on these forums need to be aware that the journalists read these forums as do the general public. I can't remember if it was the Geelong Adi or the Herald Sun but the Australian dog forums were referred to in their newspaper article.

If the DoL people are the same people that are at the forefront of anti-BSL campaigning then they havn't got a hope in Hades of changing BSL legislation with their attitude.

I concur. The Sachs study is thorough, measured and "well" conducted. If it doesn't suit a view point it will instantly be denigrated and another study supporting the view will be posted as an example of robust science.

I don't think BSL will work in the same way that prohibition didn't solve the problems caused by excess alcohol comsumption.

BUT it is so frustrating every time a particular breed is implicated in an attack to see the same tired responses. First denial that it could have been that breed, then mud slinging on the victim suggesting they have provoked the attack and hence deserved to be mauled, then the T-shirt slogans etc etc etc.

Measured, intelligent, articulate responses did much to remove restrictions on GSD and greyhounds. If only some breed enthusiasts would listen.

When we paint dogs as these passive blank canvases whose make ups are totally dependent on their socialisation and training we undermine one of the great attributes that selective breeding has brought us as well as doing the individual dogs a disservice. They each come with their own personality.

Sadly I often find myself agreeing with the authors of the TBL webpage in their assessment of those who embrace anti-intellectualism....

Anti-intellectualism the bludgeoning device the bogan deploys against the nerds of the adult world. It affords the bogan the opportunity to validate its poorly-informed opinion on complex issues, by stating that a lifetime of studying the subject at hand actually serves as an impediment to any ivory tower elitist’s analysis. The bogan believes its knowledge of the ‘Real World’ (which is limited to Today Tonight, explosive domestic arguments, and last summer’s trip to Dreamworld) trumps the intellectual’s access to the university’s considerable research resources and decades of wide reading within the field.

[justin19801]

The Sachs study is meaningless as it relies on media reports. An anti-intellectual position if there ever was one. It's supporters fail to understand in the US the term pit bull refers to a group of breeds, the most defined one being the AST! Apples and pears, not anti-intellectualism just bad research and poor statistics. I have never heard any APBT supporter blame the victim, those supporting BDL definitely attack APBT owners though. No matter which way one swings the cat in the end CDC does not and never has listed breed as a determinant of aggression, despite studying this topic more than any other group.

A cursory glance at Germanys Nuremburg laws will quickly uncover the template for BDL and finally the anti-terror laws. Racism in all but name.