Wednesday, 24 September 2014

More deceit from Merritt Clifton attacking shelter pets,

More deceit from Merritt Clifton attacking shelter pets, Information Availability - & how Time magazine got it wrong (again)

This is a weird story. A long story. And it's hard to really decide where it starts. Or if it's really ended. There are more than enough villains to go around. And victims. But we're going to try to make some sense of this as three stories merged into one bizarre news story this week.
Let's start with Merritt Clifton
Long-time readers of this blog are familiar with Merritt Clifton. Few people have spent more time in their lives trying to villify pit bulls than Clifton.
For 30 years, Clifton has claimed to track the number of fatal dog attacks, and "Maimings" and "Maulings" from dogs. Clifton has never made his list of these attacks public (he no doubt is aware it would never withstand public scrutiny). The list is compiled 100% from media reports, which already subjects it to media bias wherein news outlets determine the relative news value of similar stories differently depending on the perceived breed of dog involved.  
Because the information relys on media solely on media reports, it is neither a complete list, nor is it a random sample that could be used for statistical purposes. 
Over the years, I, and many others, have found many snafu's in Clifton's information.
 -- He's misrepresented insurance information to try to show that breed ban pre-emption laws lead to more insurance claims (when, in fact,population of dogs and people is the largest influencer, it just so happens that the six most populous states in the US also have pre-emptions on breed bans). 
-- He's used significant numbers manipulation to try to "prove" that Mandatory-Spay/Neuter for pit bulls was a good idea, but the argument suffers from severe sample bias.
-- Clifton's data represens a very small sampling of  all dog attacks, and in fact, includes less than .05% of the the total number of dog bites that require hospitalization. 
-- He actually had one 3 year stretch of data in which he counted more dog attacks by pit bulls, Rottweilers and Presa Canarios than he actually recorded by the total of all breeds combined, demonstating how inaccurate the entire data collection set is.
-- And then, there is thereality that as Clifton has become more technologically savvy, dog bites have gone up. At one point in 2011, Clifton was championing the claim that 22% of all "maimings" and 24% of all fatal dog bites over the previous 29 years had happened in just the past 19 months.  We'll get to this trend-line in a bit, but this information was definitely a far cry from all reliable statistical studies (hospital reports, CDC reports, etc) that had shown a decrease in dog bites from the 70s through the 80s as pets came to run outdoors less and were moved into the homes, and then slightly increased from the mid-90s to present as the population of people and pets began to grow. No reliable study has ever shown the type of increase Clifton has reported over the past 5 years.
Anyone who would take even just a few minutes to scrutinize Clifton's data would find its credibility severely lacking.
Information Availability
The single most likely factor in this dramatic rise in Clifton's numbers is the availability of information. Clifton started his infamous data collection in 1982. At the time there were 3 networks. There was no internet. While some data would not doubt be collected from Lexus/Nexus searches and scrolling through rolls of microfilm, having spent my college career writing papers this way makes it very apparent that even if the media decided to report dog attacks, it would have been virtually impossible to gather information on all of them.
Enter the 1990s, and you find cable news networks, 24/7 news cycles, etc, and no doubt the amount of time dedicated to news increased. Then, by 2000, the Internet was completely ubiquitous. Now, anyone can type in "dog attacks" into a Google News searh and get a complete run-down of the day's news stories from around the globe. The searches include print and broadcast media, as well as a host of online only news services and citizen journalism sources -- many of which didn't exist even 5 years ago. It would make sense that the number ofdiscovered news stories about dog attacks (very different from the number of actual dog attacks) would have grown substantially. We have access to so much more information, right from our own laptops, than we've ever had before even at the best research libraries. 
Clifton completely denies this reality.
When it comes to access to information, one other thing has changed dramatically -- and that is the amount and TYPE of information that is available. 
Prior to 2006, if a news story happened, you pretty much found out about the story exactly what the media reported in the study. That was about it. However, by 2009, more than 84 million US Citizens were on Facebook. By 2013, that number is up to 152 million. And there seems to be no limit to the amount of information that can be gleened from it.
So, with this new data source, there has become an interesting (and disturbing) trend. Now, when a fatal dog attack occurs, the minions at dogsbite.org's facebook page start trolling the facebook profiles of the people in question. They get pictures of the people. Their significant others. Their kids. The house. The Neighbor's House. It's just flat creepy.
Along with that information is often the information about the dog. There is a picture. And a picture of the owner with the dog. Maybe information about the name of the breeder where they got the dog. Or the rescue organization. Or whatever. And then I've seen the minions, like pirahna on their prey, attack the web & facebook pages. I've seen many pages just shut themselves down completely.
And all of this was information that was virtually inaccessable until 2008 or so. And now, it's all available to anyone who is intersted in diving in.
This brings me back to Merritt Clifton
For the past couple of weeks, Merritt Clifton has one again been using his faulty "data", but this time to villify shelters and shelter dogs. 
Last week, Clifton posted a report on the increase in the number of dogs that originate from shelters involved in dog attacks since 1982. Now remember, in 1982, the only information available to Clifton would have been what he found on microfilm, and what the reporter and editor decided to include. 
However, now, Clifton is able to note a significant increase in the number of US Shelter dogs involved in fatal dog attacks since 2009 -- amazingly, when the availability of the information on where people got their pet became more accessible. The numbers are not reflecting a huge increase in shelter dogs being involved in attacks -- it's reflecting the amount of information that is actually AVAILABLE  for him to garner that type of information.
In the final paragraph of that article, Clifton notes that "35 shelter dogs have participated in killing people since 2010". This data point appears to be a very cognitive effort to a) fear monger by making this number seem really high, b) try to cast doubt on shelter directors who adopt out pets and c) to make shelter pets seems unsafe. 
However, again, upon further scrutiny, it turns out that "35" (assuming it's true, which at this point, is a lot to assume) is roughly 20% of the dog bite related fatalities during that time. When you then note that 30% of dog owners get their pets from shelters, you realize that not only are shelter dogs extremely safe (as are all dogs), but their adopters seem to be more responsible by putting the dogs in tragic situations at a lower rate than non-adopters.
So shelter pets remain an extremely viable option for people in spite of Clifton's fear-mongering.
But it gets worse. 
Late last week, Time Magazine released what was the worst piece of "journalism" I've read in years regarding a 3 year old girl, who was attacked by three dogs, and then supposedly asked to leave a KFC because patrons thought she was too "scary" to be there.
Except the entirety of the KFC sequence was false. 
In the piece, the Time writer, Charlotte Alter, quotes almost exclusively from Merritt Clifton and Dogsbite.org about the this mythical increase in dog attacks, and particularly tageting pit bulls. In Merritt Clifton 31 year Summaryit, they run a new chart based on Clifton's data.
In it Clifton completely fails to recognize that his data is biased based on media bias and availability bias. It's also likely further pronounced because the media bias is now such that clicks, and click throughs matter for determining what stories get published.
And similar to how the dogsbite.org minions handle their affairs on facebook, the glob on to any story involving a pit bull and copy/paste comment after comment to drive up traffic and SEO for the site....
However, Clifton takes it one step further. And blames the increase in pit bull bites on the animal welfare community. Says Clifton:
"The Graphic...starkly illustrated the rising number of pit bull attacks since the Michael Vick dogfighting case caused the Best Friends Animal Society and ASPCA, among others to ramp up advocacy."
So now, Clifton has decided to ignore the complete failures of his own data and take on most progressive animal welfare groups in this country by blaming them (not his own data gathering) for the miraculous increase. And also completely fails to recognize that correlation does not equal causation -- and just because the timing of two events correlate, doesn't mean that one caused the other.
Merritt Clifton 31 year Summary - RevisedBut if Clifton wants to try to draw faulty correlation = causation conclusions based on no evidence of true cause and effecthe may want to be careful, because there is another interesting correlation that is more likely to have had a causal impact....as I have taken a littler liberty with Clifton's chart on the left.
I think a more likely argument could be made that the driving force in the increase in discovered dog bites is the presence of Colleen Lynn, the founder of dogsbite.org, which came online the same year as Mike Vick was convicted.
Let me be clear, Lynn, has no knowledge or expertise in dogs, or dog behavior. She is 100% biased against pit bulls in everything she does and her only experience with dogs is having been bitten by one. However, Lynn, as a web developer is web savvy, and good at gathering information and not doubt helped Clifton in his information gathering. 
She has also become an active voice in the 'pit bull debate' that creates the media's favorite need for two sides to every store (even if one side is filled with experts, and the other side only contains Clifton and Lynn) and encouraged them to write more stories.
She has also filled the debate with mounds of nonsense that makes the situation more confusing for pet owners -- making the debate about "breed" instead of about owner responsibility and canine/human behavior --instead of taking an expert-led community approach
And, as an added bonus, Lynn and her minions are great at filling up comments pages on these articles to "reward" the media outlets for the articles, ensuring further media bias in the types of storiesthey choose to cover.
It's actually pretty obvious where the problem is, and it's not what Merritt Clifton wants to blame and it's certainly not the organizations that are working to save the lives of these animals.
Time Magazine
I would be a bit remiss if I didn't say a little something about the article in Time Magazine. In all honesty, I'm completely embarrassed for Time. For starters, the article is based on a completely inaccurate premise of a girl being kicked out of the KFC (which turned out to be a hoax). It is one thing for a local media outlet to pick up a story based only on the supposed victim's facebook post and no verification from anyone at KFC or that was visiting that KFC. It's quite another for Time magazine to fall to such a standard.
Then, to basically run the anti-pit bull article based on the reports of Clifton and Colleen Lynn with little to no communication with the bevy of experts and science on the opposing side. Heck, they could have done better research had just read the article that had appeared in their own magazine over the past 3 years, including: The Softer Side of Pit Bulls, and Can Attack Dogs be Rehabilitated. Apparently Ms. Alter does even read the magazine she works for.
So from an incorrect premise, to citing the least credible of sources, the Time article is an epic fail in news journalism, and it's a shame they've sunk to such a standard in the hope of keeping up with Buzz Feed for internet clicks (but hey, at least Time may get a mention in the next Buzz Feed List, "The worst researched news stories of all time"). 
For more on the epic fail, where virtually every data point is wrong, please read this post from Sway Love that breaks it down nicely. 
And the Huffington Post reports: "It's time to end the prejudice against pit bulls"

No comments:

Post a Comment