Tuesday, December 28, 2010

Why Gunloons Have Difficulty With Studies Or, Math and Reading Are Hard

Remember Linoge's simply ludicrous assertion?
Of course, any student of statistics and history would already know that there is no correlation between firearm ownership and crime rates, and what correlation there might be is negative (in that as firearm ownership increases, crime rates typically decrease). But, then, “history” and “statistics” do not exactly fit into the anti-rights nuts’ misappropriated concept of “common sense”, so what do you expect?
Of course, Linoge didn't offer any reference as to what "student[s] of history and statistics" have read or studied to come to Linoge's frankly idiotic conclusion.  OTOH, there are literally scores of studies by real, honest-to-Jeebus professors and researchers that prove his statement false.  In the past, I have posted and discussed many of these studies.  Invariably, gunloons will scoff at them for a number of reasons, such as:
1. they haven't read them (reading is hard) and thus feel free to pretend the study has overlooked or omitted something they feel is crucial;
2. they believe studies are conducted by universities, Government organizations, and professionals who are all biased against guns;
3. the gunloons don't understand statistics(math is hard).  In fact, recently, a gunloon claimed all statistics are "hogwash."

An excellent case in point is Kellermann's case control study on gun ownership and homicide.  This study concluded a gun in the home carries a murder risk 2.7 times greater than not keeping one.  Gunloons have attacked this study vehemently--with little success.  Since they cannot attack the study on its merits or lack thereof--they often attack it by making claims that show they have never read it.  Frequent claims are:
1. the study only looked at criminals and people with drug and alcohol problems;
2. the study did not look at the beneficial effects of gun ownership (e.g., DGUs);
3. many of the victims were black;
4. Kellermann didn't look at enough or the right kind of risk factors.

Each of these claims could easily be dispelled had the gunloon read the study.


  1. Jade, was this one of your studies that you have? For one it is just one study, and we have discussed kellerman at nauseam here. You said you have many, so give us a different one to talk about. If it is Kellerman that you want to talk about, then I will oblige you one more time:

    Kellerman never examined total firearm ownership with violent crime/murder rate. This is where there is no correlation. What Kellerman did was work backwards to reach a conclusion- i.e. he started with murders and then asked if there was a gun in the house to reach his 2.7 number. If he really wanted to determine the risk of firearm ownership, he would simply compare the rate of gun ownership, with the rate of murders. It seems like the obvious thing to do, so why didn’t he? The answer being that he is smart enough to conduct the study in a way to reach the result he wanted; which is to scare people away from gun ownership.

    I’ll give you some examples of why this is faulty. Let’s say I wanted to do a study on car safety. I start by examining all fatal car crashes, and one of the questions I ask is “was the driver wearing a helmet”? In the end I get a shocking conclusion- you are 2.7 times more likely to die in a car crash if you are wearing a helmet! Of course people who think for themselves would realize that those wearing helmets were race car drivers. Do you see the problem with working backwards now?

    Since Mike hates car analogies, I’ll give you another one with guns. If we look at all Americans (including those in Iraq and Afghanistan) killed by guns and then ask the question “was the victim wearing a Kevlar vest”? You might just get a shocking conclusion- you are 2.7 times more likely to die of a gun shot wound if you wear armor that is supposed to protect you against gun shots.

    One more: Start with burn victims. Ask the question, “was the victim wearing Nomex fire retardant clothing”? Wow, you are 2.7 times more likely to be burned if you wear clothing designed to protect you from burns! So why do people wear these?

    Of course these are fictitious studies (no one outside of gun control would even think of doing such a thing), which is why I used the 2.7 number for each. The point is to illustrate the flaw in working backwards. Of course kellerman was 17 years ago, and at a time of this country's highest murder rate. His study is easily refuted not by my logic, or gun bloggers, or Lott studies- it is refuted by REALITY. Since 1993 our murder rates have dropped precipitously, but if it were true that your are 2.7 times more likely to be murdered if you own a gun, it simply couldn’t be the case. You can account for drugs all you want, but if you are working backwards, then you are never going to get a legitimate conclusion.

  2. TS, I tried to get rid of only what was duplicated. Let me know if I messed it up.

  3. TS: No, it's just one study. I wanted to show how gunloons work.

    The problem with your objection to Kellermann's study is pretty clear: you haven't read the study. Your helmet/race car analogy analogy fails because Kellermann conducted his study using a cae control method which is specifically designed to eliminate instances you describe.

    "If he really wanted to determine the risk of firearm ownership, he would simply compare the rate of gun ownership, with the rate of murders. It seems like the obvious thing to do, so why didn’t he? "

    To eliminate the circumstances you describe. BTW, Kellermann uses the exact same methodology used to show a correlation between cigarette smoking and various illnesses. If you don't use this methodology, you risk what you've described in your analogies--attaributing the risk to something when it is something else altogether such as race car driving or being in combat.

    Had you read the study, you'd know Kellerman did exactly as you wish and came up with a much, much higher correlation.

  4. Mike, you got it all, thanks. It is not clear whether long posts go though or not.

    Jade: “Kellermann uses the exact same methodology used to show a correlation between cigarette smoking and various illnesses.”

    You see, this works both ways. You can look at patients with lung disease and find higher rates of smoking. You can also look at smokers and find higher rates of lung disease. Both ways. Kellerman’s problem is you can find higher rates of gun ownership among murderers (of course since it is an effective tool for that), but you can’t look at gun ownership and find higher rates in murders. If you can- then do it. Show us your study that shows it. You said you have them. Show us how across cities, states, and counties there is a positive correlation with gun ownership and violent crime/murders (that means gun ownership goes up- murders go up). Your problem is kellerman has already been debunked by the passage of time. You are yet to address that. Life does not match his study, but still you cling to it with tenacious zeal.

  5. TS: Again, you show you haven't bothered to read the study. The murderer aspect doesn't work for you as the study controlled for those who had criminal records, those who had drug/alcohol problems and over two dozen other variables.

    "You see, this works both ways. You can look at patients with lung disease and find higher rates of smoking. You can also look at smokers and find higher rates of lung disease. Both ways. "

    Nope. Using a case control methodology, you could isolate smoking and lung disease. After all, people who don't smoke also get lung disease and they can get it from things other than smoking such as working in a coal mine, asbestos industry, etc.

    Think of it this way: suppose you were to design a study to find out if smoking leads to lung disease. You'd have to eliminate things like where the person worked or if they were exposed to other behaviors or environmental factors which could lead to lung disease.

    If you do things your way, you'd just say that of all these people who smoke--x% have lung disease. But that proves nothing because their lung disease might be due to some other factor like genetics, occupation, environment or some behavior.

  6. I like Kellerman's studies, not because I've studied them thoroughly and dissected every aspect, considering all the "control groups" and all that crap, but because they make sense to me. Step back and use common sense, ask yourself objectively if the results make sense.

    Do John Lott's conclusions make sense - no. Do those of Prof. Kellerman - yes.

  7. Mike - you like the studies because they provide results you agree with? I think scientific study has been set back a few generations with that comment.

  8. Jim, the same can be said of you. You like the studies which support your already decided-upon opinion.

    But, if you step back and honestly try to be objective, does Kleck's 2.5 million DGUs make sense? When Kellerman says owning a handgun increased the chances of suicide, does that make sense?

    You know what I think, and I assure you it's not some unthinking leap towards that which supports my beliefs. It's just common sense.

  9. Having read (some of) the study (the conclusions part), I guess I'm left with a feeling of "duh." This isn't rocket science.

    They study an urban population of people who have been killed, compare it to an urban population of people who haven't been killed, and find that guns are more prevalent in the killed population.

    Even if their statistics are solid, does this tell us anything other than that city-dwellers who are likely to be killed are more likely to keep loaded firearms in their homes than those who aren't?

    In other news, people likely to be killed are more likely to live in apartments than houses, and to live in houses where at least one person drinks. I'm not sure what I'm supposed to do with that information on a public-policy level....

  10. Kellerman! LOL! For you to bring up Kellerman shows how much you are reaching! The reality is that the states are loosening up the gun laws to allow more people to carry more places because law abiding people carrying weapons for self-defense works!

    There isn't blood running in the streets, crime actually goes down! If it was different, we'd be hearing in the papers, front page nation wide, how another person with a concealed carry permit was in jail for murder during a road rage incident! No? Instead, the anti-gun loons have to really stretch to try to suggest that our gun laws allow the drug cartels to have access to full auto machine guns,hand grenades and RPG's, LOL.

    Sorry, you anti-gun loons have lost the debate;as much as your elitist soul screams at the idea, your just going to have to accept that an average law abiding citizen can be trusted with, deep breath now,a "GUN".

  11. You're pretty funny, TomR. But I think the debate is far from over. Maybe you're the one who needs to read the papers or the internet for those stories.