Friday, November 22, 2013

More Guns = More Gun Deaths = More Overall Deaths by Murder and Suicide

PolicyMic

JAMA Internal Medicine

Although I understand the usual complaint about comparing gun ownership to gun murders as opposed to comparing gun ownership to overall murders, I also realize that since two-thirds of all murders are done with firearms, the complaint is meaningless. More guns = more gun murders = more murders.

41 comments:

  1. They appear to be using Brady scores here, but the reality is that forty of fifty states have basically the same gun laws. But this map asks us to believe that Pennsylvania and California have the same number of gun laws, so something's amiss here.

    ReplyDelete
    Replies
    1. Yeah, something's amiss in your bullshit observation about 40 of the 50 states.

      Delete
    2. Mike, the Brady's 11th best state has a score of only 25 (Michigan). After that, the next state drops to 17.

      What's your observation of the Brady scores?

      Delete
    3. My observation was that Greg's "40 out of 50" remark was bullshit. Didn't I make that clear?

      Delete
    4. What you didn't make clear was your reasoning for saying that.

      Delete
    5. The 11th ranked state is a lot closer to Arizona than it is to California.

      Delete
    6. That may be, TS, but in Greg's over-simplification, you'd need the 40th ranked state to be similar to the 1st. That's a far cry from reality.

      Delete
    7. Mikeb, are you ever going to learn how to analyze data? The fact that murder rates show no correlation with gun laws demonstrates how gun control doesn't work.

      Your side needs an inverse relation between gun laws and murder. That doesn't exist.

      Delete
    8. Greg, how can that be? If you were right that "murder rates show no correlation to gun laws," then it would mean that gun laws have no proven effect on murder rates in either direction. You're so biased it's unbelievable.

      Delete
    9. Mikeb, TS has shown you repeatedly that it is exactly the case that gun laws have no effect on murder rates. Does he have to walk you through the analysis again?

      Delete
    10. What do you guys call that, something like proof by strenuous assertion? I wasn't baffled by the bullshit, I doubt if I would be this time.

      Delete
  2. "Using an ecological and cross-sectional method,
    we retrospectively analyzed all firearm-related deaths reported
    to the Centers for Disease Control and Prevention
    Web-based Injury Statistics Query and Reporting System
    from 2007 through 2010."

    Here's an interesting bit Mike. They used data from the CDC in this study. Just recently you told me that this data is inaccurate. Wouldn't this then make the results of this study questionable?
    Something else I found very interesting is that in this "scholarly" study, they had this really neat graph that seems to show a significant percentage of the states to have a household firearms ownership rate of 50% or higher.

    http://archinte.jamanetwork.com/data/Journals/INTEMED/926956/iiq130121_732_740.pdf

    ReplyDelete
    Replies
    1. You know, your getting more and more like Greg is really becoming a drag. What I said was the CDC number of accidental gun deaths was recently proven wrong. You even agreed with it, remember? Now you're throwing it back at me as if I said all "the CDC data is inaccurate." I didn't say that.

      Delete
    2. Mike,

      In a recent post here you supported a study that contradicted CDC data regarding gun deaths which claimed that the CDC data was inaccurate, and made the comment that the study must have more accurate data than the CDC.
      And now you post a study that also uses CDC data regarding gun deaths and seem to believe it is accurate?
      Can you clarify so that we know when CDC data is accurate and when it isn't?

      Delete
    3. Are you trying to be as tedious as possible? I've been pretty clear in what I think. You're being disingenuous in pretending that you don't understand what I say. Are you telling us you don't recall the discussion about GUN DEATHS BY ACCIDENT being under-reported because some of them are listed as homicides? That's a yes or no question.

      Delete
    4. I do mike, it is cited below. The article you posted suggests both accidental deaths and homicide data being inaccurate because one type is being misfiled with the other. Then you post this study that uses the data you suggested was inaccurate, yet now it's good data somehow.
      You've given no reason why it's accurate in one place, but not the other, much like your claim that the Democrats only use the filibuster for good and light, and the Republicans use it for evil because Elizabeth Warren says so.


      http://mikeb302000.blogspot.com/2013/09/the-under-reporting-of-accidental.html

      Delete
    5. I recall that. I also recall telling you that just because you believe it doesn't make it so.

      Delete
  3. MikeB: “Although I understand the usual complaint about comparing gun ownership to gun murders as opposed to comparing gun ownership to overall murders, I also realize that since two-thirds of all murders are done with firearms, the complaint is meaningless”

    Then why doesn’t this map show murder rates, Mike? Why does it show gun murders only? Why don’t you go and substitute gun murders for total murders and see what the map looks like?

    ReplyDelete
    Replies
    1. I don't know why the authors used gun deaths. I suppose they haven't been following our discussions on the issue. What I do know is this, since nearly 70% of murders are done with guns, the map would still make the same point if it showed overall murders.

      Delete
    2. They used "gun deaths" because using murders doesn't show anything. I showed you this dozens of times. Of all the gun control advocates out there, don't you think at least one of them would put together a map like this showing murder rates? How about you?

      When you automatically assume "gun deaths" follow murders, you make the assumption that gun laws work. You yourself have said on numerous occasions that a patchwork if state laws doesn't work. That's the excuse you gave for why California has a similar murder rate as Arizona.

      Delete
    3. TS, two-thirds of ALL murders are committed with guns. You cannot escape the implications of that with all you "dozens of times."

      Delete
    4. So? That hold true in gun control as well as high freedom states. Stating 2/3rds of murders are committed with guns doesn't change the fact that your little laws don't work. You said so yourself. Remember that line about how Arizona is just a short drive away from California?

      For this map, they've simplified "strength of gun laws" into four different quartile. If we took those four buckets and calculated the average murder rate of this states, would that mean anything to you?

      Delete
    5. What it means to me is where there is greater gun availability there are more murders. This works in spite of the fact that states like California are only a short drive from Nevada and Arizona. And it proves that if we had strict gun laws in every state, we'd have fewer murders.

      Delete
    6. Mike, show me numbers that prove this. An analysis of all 50 state’s murder rates in relation to gun laws. I’ve done it. I’ve showed you my source of numbers. I’ve showed you the methods. I’ve walked you through the methods step by step. You’ve done none of these things. You’ve never disputed anything I showed you. You’ve just said “nu-uh”- that’s it.

      Delete
    7. Except that there are many states with high rates of gun ownership and low murder rates. There is, however, a strong correlation between murder rates and population density.

      Delete
    8. TS, you've got a shitty memory when it's convenient. I always disputed what you had to say in your lengthy and complicated way. I provided alternate numbers to yours. You don't remember that.

      I just had a quick look at FBI stats and guess what, Almost all the top ten or fifteen states for murder rate are gun-friendly states. Is this your game, trying to make me waste time proving what everyone already knows except you gun nuts who pretend not to?

      1. District of Columbia 30.8
      2. Louisiana 14.2
      3. Maryland 9.8
      4. Alabama 8.9
      5. New Mexico 8.2
      6. South Carolina 8.0
      7. Georgia 7.5
      7. Nevada 7.5
      9. Arizona 7.4
      10. Mississippi 7.1
      11. Michigan 6.7
      11. Arkansas 6.7
      13. Florida 6.6
      14. North Carolina 6.5
      14. Missouri 6.5
      16. Tennessee 6.4
      16. Alaska 6.4
      18. California 6.2

      Read more: http://www.city-data.com/forum/city-vs-city/576114-us-states-ranked-murder-rate-per.html#ixzz2lgkqnrIZ

      Delete
    9. His lying game!
      Stop proving all his lies, lies. We know he lies and it's a waste of your time.

      Delete
    10. For Pete's sake, Mike, how many times do we have to tell you that most states are gun friendly? Looking at a partial list of states and claiming "most are gun-friendly" is meaningless. Look at the bottom ten states in that same list you just linked to. How many of those states would you call "gun-friendly"? I'd say all but Hawaii and Rhode Island. That's 8/10 of the safest states for the good guys.

      See this mike? This is effectively picking apart an argument- which you have never done for my correlation posts. Instead you provided something else- and then I picked that apart. I said exactly what was wrong with your methods. You didn't look at all the states (just 20), you didn't account for what the Brady rankings were (you just threw them into two buckets: good and bad), and quite frankly your conclusion was such a slight difference that it favors my position more than yours! When have you ever critiqued what I did? What did I do wrong that yielded such a different result than the narrative in your head?

      Delete
    11. I'm sure you didn't do what I suggested and substitute "gun deaths" for murders to see what this map would look like- but I did. Note, there are many flaws with this method compared to a correlation calculation, or a regression analysis (use of quartiles is inappropriate), but it proves you wrong once again. Here is the average murder rate of each quartile:

      Quartile 1 (0-2 laws): 4.98
      Quartile 2 (3-4 laws ): 3.99
      Quartile 3 (5-8 laws): 3.48
      Quartile 4 (9-24 laws): 4.42

      Aw, you were doing so good until we got to the most restrictive states where we see a bump of 1/100k murders. Remember how in your numbers you said a 0.25 difference was "proof" (but I have a "shitty memory", right?). So what do you want to conclude? That a tiny amount of gun control works, but if you start adding too many restrictions (like more than 8 laws), it turns around and makes things worse? Is that what you conclude? Or do you just want to go with what I did and say "no correlation".

      By the way, when I speak of the flaws in their methods, it is very evident in their distribution of states in each quartile. By definition, this is supposed to be dividing the data set into four equal parts. But they have 16 states in quartile 1, and only 9 in the second. This could be a deliberate attempt to fit Louisiana into quartile 1, because it would look even worse for you when the distribution looks like saw teeth (down, up, down, up). We don't know this without knowing what laws they count (obviously California has more than 24 total gun laws), but as is, it draws a red flag.

      Delete
    12. So, Mikeb, D.C. leads the nation, and Maryland is third? Gun laws aren't doing much for those two. And look at Tennessee and California--nearly identical, despite radical differences in gun laws.

      Delete
    13. TS, the title of this post and the discussion at hand is not whether gun control laws work but whether more guns = more murders. I presented the top murder states as evidence, which except for the two exceptions that Greg was quick to point out, proves the point. You can attempt to turn the discussion to "quartiles" or to the bottom of the list, but all that is is diversion away from the fact. More guns = more murders.

      Delete
    14. You present the top murder states as evidence, and dismiss the bottom murder states which show the exact same thing (8/10 of them being gun friendly). And you call me bias? I showed you what this map will look like substituting “gun deaths” for murder rates, and somehow you still say “More guns = more murders”. Here’s a control question for you, Mike: what is 2 + 2?

      More “baffling with bullshit” coming up. Time to plug your ears/cover your eyes:

      Mike, what I did was a comprehensive comparison of all 50 states of murder rates in relation to what gun laws they have. A methodology where states are put into groups is quite flawed in comparison. What you did a while back was to first ignore 3/5 of the datasets (by doing only a top 10 vs. bottom 10 comparison), and then assign them discrete values of “gun-friendly” and “strict gun control”. We know that gun control is a sliding scale- it is not something that should be classified into discrete buckets, nor do we have to classify it that way because we have a real number that represents “strength of gun laws” provided by the Brady Campaign. Why on Earth would a reputable statistician take a richer dataset, and pare it down to something weaker? The reason of course, is to manipulate the data- what we call “lying with statistics”, and of course one would not be reputable if they did that.

      That’s what this study did. They took a richer dataset and pared it down into quartiles when there was no need to do so. Yet still it doesn’t reach the conclusion you want to see if we substitute “gun deaths” for murder rates. The problem with that becomes evident when we look at the fringe examples, in this case Louisiana. If Louisiana ends up in the second quartile, it totally shifts the result. What was a 1/100K drop between quartile 1 and 2, suddenly becomes an increase! And since the bottom of state gun law rankings is crowded with lots of states, it is open to small changes in a state’s gun laws having drastic effects on the results. It also becomes very susceptible to manipulation. One could define “strong gun laws” in such a way to get the results they want. Compare that methodology to what I did, which is highly stable. One state changing one law would have the result of moving the correlation coefficient a few thousandths of a percent- as one would expect.

      Let us get back to your “numbers” when you compared top 10 to bottom 10 (you know, that thing you say I forgot). Your exact same methodology applied to gun ownership rates instead of Brady Score, leads to the opposite conclusion. The bottom 10 gun owning states have a higher murder rate than the top 10 (as I showed you in one of those megaposts), and by a larger margin than what you dubbed “proof” that gun control works (albeit still small in my book). Of course, when I have looked at all 50 states, I find no correlation to gun ownership, trumping the top10/bottom10 results. But you have hung your hat on top10/bottom 10/average comparison as being acceptable statistical proof, so how do you explain it? So how can you possibly conclude that More guns = more murders? Gun ownership rate is counting the guns.

      (hint: it’s because Louisiana isn’t in the top 10 gun owning states). Do you at least see the flaws in your methods now?

      Delete
    15. Mikeb, you could equally well say that being south of the fortieth parallel causes murder. The fact is that four states out of five have good gun laws, and there's no correlation between gun laws and rates of murder.

      Delete
    16. If I do a proper statistical analysis, then I'm "baffling him with bullshit". If I use the exact same methods he used as "proof" (by counting up the pro-gun states in the bottom of the murder list), then I'm "diverting".

      Amazing.

      Delete
    17. Mike, would you be so kind as to provide us a list of the states you consider to be "gun-friendly"? Maybe that might clear things up.

      Delete
    18. New post, just for you TS.

      http://mikeb302000.blogspot.it/2013/11/more-proof-that-gun-friendly-states.html

      Delete
  4. You need to make a correction, Mike. The title of your post is wrong. The map isn’t showing “gun murders” it is showing “gun deaths” which of course includes “gun suicides”.

    ReplyDelete
    Replies
    1. You're right. That was a mistake. Did you think it was a purposeful attempt to deceive? Did it get your conspiracy alarm buzzing?

      Delete
    2. Mikeb, you attempt to deceive all the time. We're on to you, which makes your efforts all the more pathetic.

      Delete
    3. Pathetic is the guy who promotes criminality, then doesn't even have the courage to stand by his own statements.

      Delete
  5. When Mike uses numbers that are not based on CDC numbers, you blast him and say he is full of it. Now he uses numbers based on the CDC and you still blast him. Irrational and hypocritical. Nothing but brick wall ideological games and dishonesty.

    ReplyDelete