NEW YORK, NY May 2 (DPI) – Both the Wall Street Journal and New York Times picked up a story this week about a survey of dog owners – published initially in the journal Science – who widely believe breeds are pretty much all alike.
Reader comments in both publications produced skepticism and outright snickers that such a survey and its conclusions should be taken seriously. Moreover, that the media re-reported the survey – with its built-in dog-owner bias making the survey largely meaningless – struck many readers as silly, as it accorded too much importance to what dog owners think about their dogs.
In fact, such a survey struck some readers as another example of a national desire to see things as we want them – rather than as they are. As one WSJ reader opined: “Wokeness has reached pet ‘science’.”
Five of the six most popular of 266 related comments as of today on NYTimes.com:
Go to Wikipedia and search for the list of fatal dog attacks in the US, where it also lists the breed of the dog. Then count how many times the breed “pit bull” appears. One of the more recent fatalities caused by pit bulls was in February in Texas. They killed a four-year old boy. The breed collie is found nowhere on the list. Don’t tell me breeds don’t matter.
I really don’t get it. One reads this and thinks, what about the 284 human fatalities over 13 years caused by pit bulls, at a rate of about 10 to 1, over other breeds? Please, one must account for those statistics, before stating breed doesn’t matter.
Sorry, this study and the headline are misleading almost to the point of being dangerous. It’s the combination of behavior and physical characteristics that makes a breed dangerous. An aggressive beagle might be a nuisance but an aggressive Doberman or Pit Bull can kill an adult. 9 percent genetic prediction is not small, and is directly in contradiction of the headline. Totally irresponsible headline. Second, a dog is still an animal and never completely predictable. There too, an unpredictable parrot might provide comic relief, an unpredictable German Shepherd can maim a child for life.
As a dog owner, I find a few design holes here big enough to drive a truck through. First, self-report data with no prior validation is notoriously unreliable: Does anyone ever have a “bad dog?” Do they report undesirable behaviors using the same criteria and accuracy they report desirable? Second, all the authors are really reporting is that behavioral variance within groups (breeds) is greater than variance between. That’s unsurprising, given that breeds loosely correspond to what humans call races. Appearance has little to do with total genetic variability. No news there. Finally, the article doesn’t report the likelihood of interaction effects; some breeds are chosen and trained with different intentions than others. If I own a mutt, I may look for generalized sociability that predates modern breeds and ignore other potentials. If I seek a guard dog or a bird dog, I won’t care so much about sociability as specific traits that arose later: Assertiveness in limited territories, or ability to retrieve precisely without eating. Over time, even if a breed has great total variability, the “successful” animals will be the ones that fit our ideas about what breeds are supposed to do. A herder will want a collie that herds. He’ll be less likely to find poodle that could be good at it. So color me unimpressed.
“After conducting owner surveys for 18,385 dogs…” It sounds like this study reflects the biases of dog owners with respect to their preferred breeds. I watched for years as dog bite cases made their way through our local court system. The offending breed in the overwhelming majority of the cases were unambiguously pit bull type dogs. The owners of these violent dogs would often insist their pet was friendly and sociable – despite the evidence to the contrary. Asking these owners to complete a survey about their dog’s behavior would yield useless data.
Most popular WSJ comments:
My standard poodle, Annie, raised serious doubts about this study. She questioned the scholarship, then she conned me into taking her on yet another walk. 😉
As a veterinarian I find this study highly dubious. I guess 30 years of individual bias.
What would we do without experts? This is beyond silly Breeds were husbanded into existence on the basis of both physical and behavioral traits So, it’s axiomatic that breed does indeed affect behavior at least somewhat. I’m happy to see that my fellow opiners are smart enough to see that any study that contradicts common sense… is suspect at the least