You know, for all the protestation from the Civilian Disarmament Industrial Complex that the ban on federal funds is stifling gun-related research there sure are a ton of gun control studies being published. The latest is from the American Journal of Public Health, and might just be the laziest, most dishonest piece of “science” I’ve seen in a long time. The purported take-away: Hawaii’s strict gun control laws are working, and the proof is that they have the lowest number of non-fatal gun injuries among the states studied. There’s just one small problem: they cherrypicked the states they wanted to investigate . . .
From an article about the study:
Hawaii had the fewest non-fatal injuries from firearms in 2010 in a comparison of 18 states, including California, New York and Florida, a new study shows.
The report found that states with stricter gun-control laws — regulating guns and ammo, requiring background checks before sales, reporting lost or stolen firearms and keeping dangerous people from buying weapons — had the lowest injury rates.
Problem #1: they willfully ignored data from most states. Instead of investigating the non-fatal injury rates from the entire US and comparing them one-to-one, they instead hand-picked 18 states (seemingly at random) and compared the rates in those states only. I can’t even begin to describe how monumentally flawed that approach is. In what world is it morally acceptable to ignore the majority of your data in order to get the result you want? That’s a mortal sin in the world of statistical analysis.
Problem #2: they chose a very strange metric. Non-fatal firearms-related accidents will typically directly correlate with the percentage of gun owners in a given population. Smaller percentage of gun owners = smaller number of non-fatal accidents. Since Hawaii has one of the smallest populations of gun owners it makes sense that, as a population, there would be far fewer non-fatal accidents. But even then, that doesn’t really tell you anything. The common fear is that firearms will be used to kill people, and in this study they are specifically looking at NON-fatal firearms accidents.
Problem #3: their statement is factually inaccurate. The reason why non-fatal gun injuries are so low in Hawaii is that Hawaii has so few guns to begin with, not because their gun laws are so spectacularly effective. QED, you would expect there to be fewer gun-related injuries. Much as how you would expect Florida to have more alligator-related injuries than Montana. That doesn’t tell you anything about the effectiveness of the various laws, but instead about the popularity of the item in question.
The only thing that the gun laws impact are the ability of people to own firearms in the first place. If this study had said that the lower rate of gun ownership in Hawaii was the direct cause of the lower gun-related injury rate, I’d be on board. But it would still be looking at the completely wrong metric and ignoring the real questions people want answered.
Did I mention that they only looked at 18 out of 50 states? I did? OK, just wanted to double check. Because that’s insane.
Problem #4: their conclusion is wrong. The place in the United States with the most restrictive gun control laws is the District of Columbia. That small enclave has the smallest population of gun owners (~3.6%) and the toughest laws on the books. If the authors of this study were correct in their statement that restrictive gun control laws prevent injuries and deaths with firearms, then we would expect DC to be one of the safest cities in the United States.
In fact, the murder rate in the District of Columbia is the highest in the nation, well over twice the murder rate of the next most violent state (California).
Let’s recap, shall we?
- They only looked at 18 of the 50 in the US.
- They investigated the wrong metric, one that no one cares about, and used it as a stand-in for “gun violence.”
- Their results statement was factually inaccurate.
- Their overall conclusion was wrong.
Who the hell gave these people money, much less a degree?
All are associated with the University of Washington & VA. I assume WA and US taxpayers funded the research.
Even cherrypicking States is not a sufficiently granular sampling for useful information. Since the biggest problem with anti gun laws is that they don’t actually work, selecting by state gun laws is deceptive at best, no matter who’s picking and which are picked.
Get better resolution. Look at the output end instead of the input end… As is typical of Democrats/Anti-gunners/Liberals, they defy science by ignoring the observation of cause and effect… They look at their desired input and presume the output all worked out regardless of what actually happened. They perform the experiment, and pay no attention to what happened… Then claim success!
Look at the amount of violence occurring, not just with guns, but in general. Where are the violent PEOPLE? Interesting, that if you break it down by voting districts, regardless of what gun laws may or may not be there, heavily democrat voting districts are where almost all violence occurs… Play with the data a bit and see what you can learn. Remove all the heavily democrat voting districts from the data. Suddenly, the USA becomes the most heavily armed country on Earth, with an occurrence rate of ANY form of violence at least an order of magnitude lower than anywhere else on the planet… Almost 5x more guns than people!
Am I saying “democrat voters are just dirty violent trash?” No. I’m saying the democrat message of entitlement is very frustrating to people dealing with reality that doesn’t match up. They just can get any of the stuff they think they deserve. The education system is a joke, too… Dumb people can take only so much pressure before the revert to their animalistic state and take out their frustrations any way they can… Some self-medicate. Some turn to violence. Some do both… Demcrat politicians are deliberately doing this to their own constituents, and the sooner they pull their heads out of their butts and go a different way, as we beg them to join us in, the better… For everyone.
Got some good sources on the Democrat district => gun violence correlation? I’ve got a friend who regularly sends me every article the Daily Kos has about guns and I’d like to hve this on tap to sling back at him next time.
I don’t have links. It was a concept I read about a year ago and decided to crunch the numbers myself.
Depending on the exact percentage that you define as “heavily democrat,” somewhere between 55% and 60% in the last 5 elections is what I used. It’s an amazing drop off. I modified to to make a easy slider for the “percentage democrat excluded.” Looking district by district, the violence curve starts to drop sharply at exclusion of 53% democrat, and has reached almost zero by the time you’ve excluded >= 62% democrat. I’ve never seen statistical data with such a glaring and obvious result. At the point of excluding all voting districts that are >= 65% democrat, you’re left with a country that has almost 10x more guns than people, and no statistically relevant violence of any kind.
It’s also useful to point out that the curve gets steeper… Meaning that the higher the percentage of democrat voters in a district, the more drastic the violence is. Another way to say it is that the drop from say 55% to 56% is less than the drop from 56% to 57%, and so on. The greater the density of democrats, the greater the violence, and it’s not linear. After the breakover point at about 52%, each 1% more of democrats you have, accounts for a wildly exponential increase in violence. You go from an occurrence rate of less than 1 per 100,000 to almost 15 per 100,000 in just a few percentage points.
I encourage someone with “reputation” to do it because publishing my numbers means nothing. I’m nobody and wouldn’t be taken seriously anyway… It’s even more telling when you overlay that graph with gun ownership… It’s glaringly inversely proportional…
These numbers can wiggle a bit depending on your definition of “violence” and a few other minor factors. But no matter how you try to fudge it in favor of either argument, the difference is so huge that there’s no hiding it…
Study was submitted, but rejected at http://www.jspurc.org/rese2.htm and http://www.improbable.com/2014/05/09/spurious-correlations-by-the-bucketful/
However, they did publish “Strong correlation with manufacturing doctorates awarded and per capita consumption of motzarella cheese” as it was more credible.
” They perform the experiment, and pay no attention to what happened… Then claim success!”
What do you mean, “claim”? They ARE successful, drawing down a big paycheck for nonsense science leading to the result they were being paid to discover.
I was referring to declaring success of their implemented experiment, not these absurdly fake studies…
Good science is good observation, and they don’t observe at all. They don’t even bother to look at the actual outcome of what they do, they just assume that “gun control is good” and declare it a win. No concern for the consequences… It must be a good thing because they just totally know it has to be cuz they heard so on the news and guns are bad…
It’s no big secret why they used such an oddball metric of “gun violence. ” The other ones obviously didn’t give the “correct” result. Even so, they had to hand pick their comparison states – pathetic. It’s like they’re not even trying anymore.
Hawaii is an island that’s six hours from LAX. Let’s compare another island with similar gun laws, maybe Puerto Rico?
+1 PR may be broke, but they lead the world in gun homicides: “Puerto Rico tops the world’s table for firearms murders as a percentage of all homicides – 94.8%. It’s followed by Sierra Leone in Africa and Saint Kitts and Nevis in the Caribbean” http://www.theguardian.com/news/datablog/2012/jul/22/gun-homicides-ownership-world-list
firearm deaths / 100,000
US: 2.97, PR: 18.3
“Who the hell gave these people money, much less a degree?”
SOURCE: Alligator Attacks on Humans in the United States, Ricky L. Langley, MD, MPH. Wilderness and Environmental Medicine, 16, 119 124 (2005). Additional figures taken from news reports.
Ah, but you are looking at standard capacity alligators, not evil black (or possibly a very dark brown/green) high capacity crocodiles… that’s what needs to be banned.
Do the alligator’s shoulders go up?
They have alligator control laws in Florida, and not in Montana.
Therefore their Alligator control laws are not “working”, according to the study.
The “study” is just a variation on the “progressive” theme: “If we take away peoples guns, less people will be killed with guns.”
You can as easily say, if we ban hospitals, no one will die in hospitals!
I hate hospitals. People die there.
Oh look, my nose is bleeding.
If it weren’t for my horse…
I worked in medical research for 38 years and somehow, in spite of spending roughly 10 hours a week chasing citations in the library and online, I have never heard of the American Journal of Public Health. That doesn’t mean that they’re nobodies, they might just be newbies, but they’re pretty damn obscure!
Journals come and go. Some are well respected, and some are crap. Some that were once well respected are now crap, and some get fooled into publishing crap because their unpaid referees are buddies with the author. I was once asked to replicate a study in which the materials and methods section of the article contained no information about the analysis or any reference to it!
When reading research articles it’s a good idea to keep your bullshit filter on high alert.
First, I’ll state the original article is pretty bad. You pretty much can’t determine anything from state level data. You can show some correlations, but causality forget about it. The article proves nothing about state level non-fatal firearms accidents.
However, Nicks rebuttal to this article is even more greatly flawed.
1) They did not cherry pick 18 states searching for a conclusion, never would have gotten published. If, Nick had actually bothered to read the article the data set the used only covered those 18 states.
“For this ecological, cross-sectional study, we used data from 18 US states that reported discharge information to the Healthcare Cost and Utilization Project State Inpatient Databases and the State Emergency Department
Databases in 2010.”
4) Comparing DC to any state is incredibly disingenuous and is a horrendous example of cherry picking in order to try and prove a point. DC has a weird state like legal status, but for all intents and purposes it’s a city and should only be compared to other cities.
Wow, you accused someone of cherry picking to prove a point and then engaged in an egregious example of cherry picking to prove your own point, unbelievable.
Thanks. It’s important to call out erroneous assertions and poor scholarship.
Pete, it is possible all you accomplished in your analysis is to push the question of cherry-picking back one step. The database the researchers used is one created through the willing cooperation only of those states who find the rate of reported accidental and non-accidental firearm discharges to be of interest, i.e. gun-control-eager states. There are all- (or nearly all-) states databases on firearms crimes: Correlating state gun laws to gun crimes would be of interest. It would not, however, serve the purposes of those who, for political goals, seek to identify firearms as an inherent public health risk with no redeeming functions. These same people are unwilling to admit and deal with the reality…that violent people are inherent risks to public health. The same people who wish to ban the means of self-defense seem to be, in my experience, the same people who wish to reduce the sentences of violent offenders, including those who use a weapon, any weapon, in the commission of a crime.
The vagueness of the metric used in the study and its irrelevance to the study of [burdensome gun law/firearms ownership rate/firearms injury rate] correlations…cause the paper to reek of Hemenway.
Straight from their website:
“HCUP provides reliable, comprehensive information that can be used to answer questions about health care use, access, outcomes, and costs related to hospital inpatient stays, ambulatory surgery and services, and emergency department visits. HCUP databases, software tools, and reports enable research on a broad range of current health care issues and trends that are useful to policymakers, health services researchers, administrators, and consumers.”
The HCUP database isn’t remotely a gun dataset, the gun injury data is one of hundreds of data points included in the set.
Besides, you really think the states of AZ, SC, UT etc. are pro gun control states? Please actually glance at the paper, rather than taking Nick’s dubious word for the gospel.
You missed one other major flaw–the apples to oranges comparison, comparing D.C.’s homicide rate to Hawaii’s nonfatal accident rate. What is D.C.’s nonfatal accidental shooting rate? Just like Hawaii, I suspect it is rather low as the total gun ownership is low. The more guns, the more accidents, as a statistical probability. I have to assume that accident rates have nothing to do with gun control laws, except to the extent that those laws artificially limit the gun ownership rate.
What a stinking pile of junk “science” this piece is! I backtracked up to see more from this esteemed journal of knowledge so I could let the editors know what this scientist (me) thinks of their scientific methods (not much). Right there on the homepage: http://ajph.aphapublications.org/ is this little tidbit.
AJPH Collection on Gun Violence Prevention
“In the wake of the tragic shooting at the Washington, DC, Naval Yard on Monday, September 16, 2013, the journal has compiled a collection of papers and other resources to help promote awareness and policy change surrounding gun violence, veterans health, and suicide. For those seeking more information on organizations that work with victims of gun violence and other types of violence, please consult Chapter 11 from APHA’s book, Confronting Violence.”
So it seems that these dingleberries are in the business of converting tax dollars into pseudo-scientific anti-gun agitprop. They should be taunted mercilessly for simultaneously wasting our tax dollars and insulting true scientists who work within the scientific method. To the American Journal of Public Health, I’d like to say up yours with a crooked broomhandle.
In the butt.
With a pinecone.
It’s not just agitprop, promoting these lies has consequences. People die so they can get some dirty guvpaper…
I’d also like to point out, that we have a TON of Gun owners in hawaii. which makes the story even dumber.
I kinda got that impression from all the “defend Hawaii” stickers.
Nick, this was actually peer-reviewed, so each one of the authors and reviewers should be science-shamed and exposed for this deception.
Peer reviewed is meaningless. These peers are other academics looking to brown nose their way into a ever dwindling supply of tenure positions. And you can bet your ass that any “peer” who opposes the anti-gun agenda will never receive any academic post other than janitor. I know because I lived it.
Yep. It’s not “peer reviewed” as much as “pal reviewed” because everyone must never question the orthodoxy.
Peer review is useless when everyone in the group wants to drink from the same poison well.
You guys forgot to add each chief of police for the Hawaiian counties don’t like to issue the common people a ccw permit.
Common people? ANY people. you could be Obama and he still wouldnt give you a permit, guarantee.
You should see the shenanigans they pulled when they replaced their old S&W 5906 service pistols with glocks. They destroyed all of them including 200 NIB pistols, to the tune of 500000 dollars rather than sell them to the public…because he didnt want “more guns on the streets”. in the only state with a gun registry. In a state with no CCW. where each pistol purchase requires a separate and unique permit, and takes 2 weeks. on an island where you can only do registration at a single police station that has literally only 6 public parking stalls. in a state that hasnt had double digit gun murders (or regular murders) in decades.
Ironically what crime there is here is conveniently overlooked when anyone applies for a Ccw in paradise. Two “yoots” double tapped in the head, a multiple stabbing, another multiple stabbing… All within a 2 minute walk from my office in Kalihi.
NY has a pistol registry, and NYC makes you register all guns. Across the state, pistols have to be listed on your permit before you can even pick them up at the store (except for 1 county I think). Sometimes it takes months for the authorities to amend your permit and give you permission to pick up the gun you’ve already bought.
With a Mayor that “supports” gun owners and the shooting sports…..yeah, right.
“they chose a very strange metric”
And that oddity is a giant red flag to anyone who knows anything about fudging studies. I remember I was in a research class once- a liberal arts environment- and a woman was explaining her research proposal. She wanted to study whether there was a relationship between some variable (let’s say oil for the sake of argument) on interstate violence. When I asked how she had picked her three case studies she said that she picked them because they had oil and were violent. This was a person going for a graduate degree at a major university who didn’t understand- or care about- the difference between a study and an editorial.
Whenever someone has a strange metric and when there’s a simpler one it means they’re trying to massage the numbers until one comes out they like.
There you go again using common sense and logic to refute an antigun “scientific study”. By the way, once you bring DC into the conversation, the powers there, like her Honor the mayor and her Chif of police will tell you the rise in gun violence is due to the unregulated Wild West of MD (cough, cough) and Virginia, not DC’s screwed up culture and antigun laws. One can only do so much! LOL.
In other news, Mackinac Island shows fewer incidents of motor vehicle accidents than New York City, Washington DC, and Los Angeles.
At first I thought maybe people are better motor vehicle drivers on Mackinac Island or, assuming the island is small and speed limits are lower, people get into fewer accidents at lower speeds. Nope. My favorite search engine reveals that there are NO motor vehicles on Mackinac Island. Hence there are fewer motor vehicle accidents.
I wonder if the motor vehicle accident rate is lower on the Hawaiian Islands than the mainland?
Reach a conclusion then create the data that supports that conclusion. That is how researchers lie. Today the credibility of a research study is less than that of a t4elevision commercial.
If you think Hawaii is a fun example, try looking into the U.S. Virgin Islands and Puerto Rico.
Also islands and both have strict gun laws.
I’ve got to stand in line and re-permit my weapons every three years (needs approval from the police commissioner) and we still top the murder rates worldwide.
Irrelevant. Such data would skew their predetermined outcome.
“they instead hand-picked 18 states (seemingly at random) and compared the rates in those states only”
OK, this is a general tell, looking for something spun. Picking 18 of 10,000 states, maybe, because sampling all 10,000 is a different scale of effort. Picking 18 of 50 – what’s the difference? Not much. Either sample size is too big to do entirely in your head. Both are pretty easy – just write the piece. You don’t need a supercomputer, or even a calc on your smartphone.
It’s not a guarantee, but whenever the sample / example size is about the same as the whole population, you gotta ask: “So, why not just use everybody?”
It’s easier than all the study, it’s HI, a chain of islands. You don’t escape. It’s difficult to get things in comparatively to continental area. Most shipments come out of CA.
Bottom line: HI can’t be used in a study of this type.
“Who the hell gave these people money, much less a degree?” Probably these math gurus:
‘Sixty-six percent of Americans for Responsible Solutions supporters believe our background check laws are the single biggest impediment to making our communities safer from gun violence.
Forty percent believe state ballot initiatives are the best way to change our gun laws, and twenty-eight percent believe working with Congress should be the top priority. Those are just some of the results from the survey of ARS’ supporters we conducted over the last several days. And the good news is, we’ll have the chance to fight and win on all of these fronts between now and next November. ‘
Poor Gabby Giffords. Someone needs to help her with math, instead of lying about how well she is recovering and naming a ship after her…