Thursday, July 3, 2014

Commenter TS Claims to "Destroy" Kellermann Study

TS in comments:

Please. I destroyed you last time you brought up Kellerman. Don’t you remember how I brought up that the same study says you’re 4.4 times more likely to be murdered if you rent your home, rather than own? Then you go off saying it’s because they tend to have less income, etc. but this was right after you explained how Kellerman’s study isolated each factor from each other. Income was supposed to be removed from the gun ownership result, right? So why wasn’t it removed from renting? It was comical.
I'm always amused how gunloons claim to have debunked or disproved or "destroyed" scientific studies without having read or understood them.  It's a real talent.

In 1993, Dr. Arthur Kellerman (and several other co-authors) published a peer-reviewed article in the New England Journal of Medicine.  Among the findings was keeping a gun in the home carries a murder risk 2.7 times greater than not keeping one.  Naturally, such a conclusion runs counter to the gunloon religious tenet that guns prevent all crimes and bad happenings and make you more manly, etc.

Now did Dr. Kellermann and his co-authors use some kind of voodoo pseudo-statistical gimmickry to come up with his findings?  Hardly.  He used the exact same methodology used by epidemiologists the world over.  In fact, it's the same methodology used to ascertain that cigarette smoking cause cancer and other health ailments.

TS's comments regarding renters and income demonstrate he hasn't read or understood the study.  When researchers "control" for various factors--what they are doing is trying to isolate a certain factor from environmental and behavioral variables. For instance, health problems from smoking has to be controlled for certain variables such as genetics, exposure to occupational hazards, etc.

Dr. Kellermann's study showed that illicit drug use in a home carried the highest murder risk increase (5.7X).  Being a renter (4.4X), domestic violence (4.4X),a gun in the home (2.7X), and a household member with an arrest record (2.5X).  Where TS gets confused is that he "thinks"  these murder risks are a combination; they aren't.  IOW, if one has both illicit drug use and a gun in the house, the murder risk increases dramatically.  The more present variables, the greater increase in murder risk.

52 comments:

  1. Hey Jade, how's it going? Got any idea when those big bad post office police are going to kick down my door and arrest me like you promised they would? It's been several years now and I'm still waiting. You wouldn't lie to me, would you?

    ReplyDelete
  2. The notion of "controlling for" is not exactly what is stated above. They are trying to take other variables into account. It is the nature of the regression method. When you have another variable in the model, you statistically remove that variable from the primary relationship. Here, the primary relationship is guns and homicide. By including renting, this statistically removes the rent-homicide relationship from the gun-homicide relationship. It is not perfect, and statistical control is always unsatisfying in ways. But you can say "I found a relationship, even after accounting for the rent situation". It's a standard epidemiology approach. Same appoach as was used to show the tobacco-lung cancer relationship.

    ReplyDelete
  3. Not only did you not address the point I made, but you made up something about me thinking it's a "combination"- which I never said. What I said is that when I brought up the fact that Kellermann's study showed renting to be a higher risk factor than gun ownership, you said it it was because of reasons other than whether or not they rent. You said its because they're poor. But you said he adjusted for income for the gun risk number. You're claiming he adjusted for income for guns, but not for renting. Why would he do that? Is he really in the pocket of the NRA by trying to make the gun risk look better than the renting risk? So why do you trust this academic study then?

    Secondly, I brought up another huge problem with the study, which you didn't even quote me on when challenging me. Probably because you have no goof answer. Kellerman's study only looked at urban data. It was for the counties encompassing Seattle, Memphis, and Cleveland. What do you think happens to that risk factor when you start including rural household where gun ownership is higher and fewer murders happen?

    ReplyDelete
    Replies
    1. Well, for the rural households, we need to talk about suicide.

      TS, you're the most close-minded and stubbornly biased gun nut I've ever run across.

      Delete
    2. Sadly, TS, you insist on showing us you've not read the study.

      Does renting have a greater murder risk than a gun in the household? Yes. And if you rent *and* own a gun your murder risk is even higher. Rent, own a gun and engage in illicit drug use--you're practically begging to be killed.

      This is what you don't understand. Studies like this have to separate out factors in order to determine risk. IOW, it would be pretty hard to attribute someone's death dues to smoking if he also worked in a caol mine and had a family history of lung cancer.

      Re your urban nonsense. The study looked at areas in Washington state, Tennessee and Ohio. It encompassed areas that were both urban and rural. It included areas that were well-off and areas where the povert rate exceeded 15%. There were areas that were predominantly white and ares where the black population exceed 25%.

      Pleas e come back when someone reads you the study.

      Delete
    3. Small correction: areas where the black population exceeded 44%.

      Delete
    4. Jade, you must be a politician. Only a politician can be so blatant about not addressing a point, and instead talking about something completely different. I thought I was pretty clear in my last post that I AM NOT claiming this study isn’t separating out risk factors. That is the intent of this study. You are right about that, so you can stop repeating that. Of course, I was never disagreeing with that. I guess you just felt you needed to answer something, but couldn’t come up with a rebuttal to the criticism I did pose about the study. That’s what a political would do in a debate when there is a real loser of an issue for them-immediately talk about something else.

      Here’s another chance. The renting risk result is part of a long list of discredits for this study. Whether or not you pay checks to a bank or a landlord should have no bearing on risk for being murdered. I have no doubt that more of the victims in his data set were renters, and that is going to show up in the crude odds-ratio (as it did). But if he correctly identified and recorded the real risk factors, the adjusted odds-ratio should come out near zero. Instead the risk would show up in factors like socioeconomics, neighborhood, and age- all of which would correlate with renting.

      Do you really believe that the 4.4 risk factor is isolated for renting? Say a professional working family wants to upgrade their house so they put it on the market, but it sells before they can find their dream home. So they work out the common deal to rent their home from the new owner until they close on their new house. According to Kellermann, this family is at a significant increased risk of being murdered during this time period. Note the ONLY thing that changed for this family is their rent/own status. Income, neighborhood, whether or not there are guns in the home all remain the same. Do you believe this family is at greater risk for being murdered? Of course you don’t. When I brought this up before, you said the risk factor was for other reasons- being in a worse neighborhood, lower income, etc. In other words, you said Kellerman did a crappy job of isolating factors.

      Now back to my other point about the study being only for urban areas:

      Jade: “The study looked at areas in Washington state, Tennessee and Ohio. It encompassed areas that were both urban and rural”

      What do you mean by “areas”? It looked at three counties- Shelby, King, and Cuyahoga. These are the counties that contain Memphis, Seattle, and Cleveland. There is nothing rural about these counties. At best the outskirts would be considered suburban. Even if there are a few cows here and there, you can’t possibly be arguing that this makes an even distribution of rural/urban households. You’d probably have to have four or five rural counties just to make up for population differences of a big city county. And, had you read the study, you would know that Kellerman himself says the data is for three “metropolitan” counties.

      Delete
    5. You're contradicting yourself, TS. In one statement (" Whether or not you pay checks to a bank or a landlord should have no bearing on risk for being murdered.") and in the next you admit the data showed exactly that.

      Of course, it's easy to show why renters as opposed to buyers are at greater risk. And it does sometimes have to do with where the checks go. Most rental properties involve transient groups of people living in close quarters. Some of these rental properties are not in the best areas--consider public housing. Additionally, which is harder? renting a place or obtaining a mortgage? Try to be honest.

      "There is nothing rural about these counties."

      http://www.shelbycountytn.gov/DocumentCenter/Home/View/664
      http://your.kingcounty.gov/dnrp/library/archive-documents/dnrp/pdf/rural-directory.pdf
      http://www3.uakron.edu/src/Demographics/Cuyahoga-Demographics-IPS.htm

      Delete
    6. Oh, Jade… At least you are trying.

      Let’s start from the bottom. You posted a link to the Cuyahoga county census that shows the population being 99.8% urban. Ninety-nine point eight! Did you not read this before posting it? Or do you somehow think that 0.2% of the population being rural contradicts my point? Heck, even I didn’t think it would be that high.

      On the renting thing, I did not make a contradiction. I said it would show up in the crude odds ratio. For all your condescending remarks about me not “understanding” the study, you are revealing that you don’t understand the difference between crude and adjusted odds ratios. If Kellermann did such a masterful job of isolating risk factors, we would see the adjusted odds ratio for renting be somewhere close to 1.0 (no additional risk). Why? Because the real risk that showed up in the crude odds ratio for renters was things like being in bad neighborhoods, or being transients, lower income households, etc., all those things that you just said are the reasons why someone who rents is more likely to be murdered. It is not because they don’t own their house. It is a classic example of correlation not being causation. When you state other reasons why renters are more likely to be murdered, you are saying that Kellerman did not correctly isolate renting as a risk factor. Do you finally get the point I am making?

      Delete
    7. Let's remember, TS, you claimed there was *nothing* rural about these areas. Yet, King and Shelby counties have extensive rural areas.

      You still haven't read the study and are hung up that renting poses a greater risk of murder. Again, if there existed no difference between owning and renting--the study would have shown that. But it did not.

      Just as if owing a gun and not owning gun presented the same risk. But again, it did not.

      BTW, Kellermann et al aren't the only studies to reach similar conclusions.

      Delete
    8. Extensive rural areas? After you shot yourself in the foot by self-posting the Cuyahoga stat showing it at 99.8% urban, you decide to take another couple of blind shots hoping Seattle and Memphis turn out to have more cows than people. Really, Jade? Did you bother to look it up first?

      http://www.city-data.com/county/King_County-WA.html

      http://www.city-data.com/county/Shelby_County-TN.html

      They both come in at 97% urban. As one internet commentator to another, it’s time to let this one go, Jade.

      Jade: “You still haven't read the study and are hung up that renting poses a greater risk of murder”

      You still haven’t read my comment. You still can’t even address the point I am making. My point is not that the renting risk came in higher than the gun owning risk. I’d make the same point if the numbers were reversed. My point (one more time), is that the renting risk after isolating doesn’t correctly show that the risk has NOTHING to do with actual renting. Jade, answer me just this one question. Why did you say this line: “Some of these rental properties are not in the best areas--consider public housing. “? So you are saying Kellermann did not consider public housing? That he did not consider neighborhood crime rates? He didn’t account for these things?

      Jade: “Again, if there existed no difference between owning and renting--the study would have shown that. But it did not.”

      And you can’t possibly consider that it was a crappy study. Brain can’t go there… Kellermann is infallible.

      Delete
    9. I'm exhausted just reading TS's stubborn nonsense and vain attempts to baffle 'em with bullshit. Thank god Jadegold is back.

      Delete
    10. Of course. Jadegold is back and trying to argue how Seattle, Memphis, and Cleveland are cow country, but I'm the one who is "baffling with bullshit".

      Delete
    11. TS: I'm quite familiar with King County having been there many times. I even have a classmate who owns a farm there. I also worked under a retired Navy officer who owns a farm there as well. So, you are free to claim there is no rural area to King Co--but you'd be wrong. Now, you're free to think I'm lying but I did post a link to King Co's agricultural office. Are they lying too?

      And Shelby Co also has rural areas. I posted a link with a county map designated those rural areas. Is Shelby Co in on my big conspiracy?

      "My point (one more time), is that the renting risk after isolating doesn’t correctly show that the risk has NOTHING to do with actual renting. "

      You are wrong. And you continue to show you haven't read the study.

      "Why did you say this line: “Some of these rental properties are not in the best areas--consider public housing. “? So you are saying Kellermann did not consider public housing? That he did not consider neighborhood crime rates? He didn’t account for these things?"

      Again, the study quite plainly says it accounted for living in public housing. And it accounted for the affluence or non-affluence of the area.

      "And you can’t possibly consider that it was a crappy study. "

      Unfortunately for you, you can't back up that assertion. And if it was so crappy why have many other studies validated its findings?

      Delete
    12. Jade, those counties are 97% urban. If Kellermann's data pool were a 50/50 split of urban and rural households, he would never get the results he intended because so many of those households would own guns without anyone being murder. Try reading what I said sometime. Like this:

      " Even if there are a few cows here and there, you can’t possibly be arguing that this makes an even distribution of rural/urban households. You’d probably have to have four or five rural counties just to make up for population differences of a big city county."

      Jade: "Again, the study quite plainly says it accounted for living in public housing. And it accounted for the affluence or non-affluence of the area."

      Then why did it come out 4.4x risk? His matched pairs for socioeconomics should have filtered this out. And may I remind you that YOU said on many occasions (even just above) that the renting risk was because of other factors- socioeconomic factors that should have been adjusted out.

      Jade: "And if it was so crappy why have many other studies validated its findings?"

      You mean other studies funded by the Joyce foundation? Gee, I wonder. Even when the money source isn't the same, that doesn't mean it isn't approached with an agenda. There are pro-gun agenda studies that come to the opposite conclusion, you know. Not every study can be right.

      Delete
    13. And what do you mean I can't back up that assertion? I point out obvious flaws like how he didn't properly isolate renting as a factor, and you agree with me by saying renters have a greater risk because of socioeconomic factors that were supposed to be adjusted out, but you're too biased to realize you agreed with me.

      Delete
  4. Again and again, this demonstrates that gunsucks are simply not particularly intelligent. The entire point of the study, and other studies like it, is to show that gun ownership is a risk factor. The attempts of gunsucks to show that other factors are also risk factors is TOTALLY BESIDE THE POINT.

    The entire idea of risk-based causality is to show what factors are involved. NO ONE, NOT ANY SINGLE EPIDEMIOLOGIST, EVER SAYS "GUN OWNERSHIP IS THE ONLY FACTOR". This is the fallacy of many readers.

    There is NO PROBLEM in concluding that MANY factors are involved in homicide. OF COURSE THIS IS TRUE. We assume that many factors are involved. The point is not to find the SINGLE factor. There is no single factor. The point is to find the DIFFERENT FACTORS WHICH ALL PLAY A PART. Gun ownership plays a part. So does being in rental housing, drug involvement, etc. None of these are DEFINITIVE.

    This is a limitation that many have in thinking about risk. Risk-based causal considerations DO NOT FIND THE SINGLE FACTOR. They find many factors.

    ReplyDelete
    Replies
    1. What is with you guys not reading what my criticism is? I have no problem with the study showing other factors. It should do exactly that. Nor is my criticism that other factors showed higher than gun ownership. My criticism is how renting wasn’t properly isolated. As Jade said himself, there are other factors that are hidden in the renting risk that the Kellermann study did not pull out. We all know this. My second criticism is over the data pool, and I’ll respond to that below.

      Delete
  5. And it it totally irrelevant that rural counties were omitted. There is no study which can be done which controls for EVERYTHING. If rural counties were included, the criticism would be that rural counties from some state were omitted. If rural counties from all states were included, then the criticism would be that rural counties are all different, and every rural county in the country needs to be included. Idiotic comments, every time. The study was a properly controlled epidemiological study of factors. It is not perfect, no study ever is.

    ReplyDelete
    Replies
    1. It is not irrelevant. Of course, you are the same person who said it’s irrelevant to compare murder rates to gun laws for each state. Rural areas have higher gun ownership and lower murder rates. This is not in dispute. So what do you think happens to the data when it’s somewhere around 50/50 between urban and rural vs. +99% urban? The data is flooded with gun owning households where no one was murdered, and his adjusted odds ratio plummets. Hell, I’m surprised he only got 2.7 AOR sampling from places like Cleveland and Memphis in the late eighties/early nineties.

      Jade likes to bring up cancer/smoking studies as a comparison of methodology. This would be the equivalent of ONLY sampling data from coal mining towns when linking smoking to cancer. Of course they didn’t need to do bogus studies like that because there is a real risk.

      Delete
    2. Except the vast majority of decedents in Kellermann's stidy were not murdered with guns kept in the household.

      http://hsx.sagepub.com/content/5/1/64.short

      So much for angry spouses grabbing the household gun and firing

      Now how is that a causal effect again?

      Delete
    3. I see Anthony Wilkins has elected to proclaim that he, too, hasn't read the study.

      In point of fact, Kellermann's study doesn't document whether the decedents were killed with the gun kept in the household. As to the "vast majority" claimed by Kleck--this bogus claim actually originated with John Lott who claimed " “they fail to
      report that in only 8 of these 444 homicide cases could it be established
      that the gun involved had been kept in the home.”

      This is simply a gross misrepresentation by Lott. Kellermann's study does refer to 8 out of a subset of 14 cases the police report stated that the gun involved had been kept in the home. But that was anedotal data.

      lott also claims the vast majority of decedents were killed by armed intruders. this alos fails the smell test; Kellerman's data show that less than 8% were killed by intruders.


      But even if assume that everyone is being killed by people who bring guns into the home--why is it that households without guns aren't seeing a similar murder risk?

      Delete
  6. Wow. You didn't read Kellermann's study at all.

    "A subsequent study, again by Kellermann, of fatal and non-fatal gunshot woundings, showed that only 14.2% of the shootings involving a gun whose origins were known, involved a gun kept in the home where the shooting occurred. (Kellermann, et. al. 1998. "Injuries and deaths due to firearms in the home." Journal of Trauma 45:263-267) ("The authors reported that among those 438 assaultive gunshot woundings, 49 involved a gun 'kept in the home where the shooting occurred,' 295 involved a gun brought to the scene from elsewhere, and another 94 involved a gun whose origins were not noted by the police [p. 252].") (Kleck, Gary. "Can Owning a Gun Really Triple the Owner's Chances of Being Murdered?" Homicide Studies 5 [2001].)"

    There you have it. The vast majority were not killed with the same gun kept in the household, but a gun brought into the house by a perpetrator.

    "But even if assume that everyone is being killed by people who bring guns into the home--why is it that households without guns aren't seeing a similar murder risk?"

    I didn't say all. I said the vast majority.

    Look at the univariate analysis.

    Notice how handguns were significantly correlated with homicides but not rifles and shorguns? What sort of gums do people own for self-defense? Handguns.

    Quote from Kellermann: "Third, it is possible that reverse causation accounted for some of the association we observed between gun ownership and homicide -- i.e., in a limited number of cases, people may have acquired a gun in response to a specific threat."

    Which makes sense when almost nobody in the study was killed with a gun that was the same one kept in the household.

    ReplyDelete
  7. Wow. You didn't read Kellermann's study at all.

    "A subsequent study, again by Kellermann, of fatal and non-fatal gunshot woundings, showed that only 14.2% of the shootings involving a gun whose origins were known, involved a gun kept in the home where the shooting occurred. (Kellermann, et. al. 1998. "Injuries and deaths due to firearms in the home." Journal of Trauma 45:263-267) ("The authors reported that among those 438 assaultive gunshot woundings, 49 involved a gun 'kept in the home where the shooting occurred,' 295 involved a gun brought to the scene from elsewhere, and another 94 involved a gun whose origins were not noted by the police [p. 252].") (Kleck, Gary. "Can Owning a Gun Really Triple the Owner's Chances of Being Murdered?" Homicide Studies 5 [2001].)"

    There you have it. The vast majority were not killed with the same gun kept in the household, but a gun brought into the house by a perpetrator. That data is straight from Kellermann, not Lott.

    "But even if assume that everyone is being killed by people who bring guns into the home--why is it that households without guns aren't seeing a similar murder risk?"

    I didn't say all. I said the vast majority.

    Look at the univariate analysis.

    Notice how handguns were significantly correlated with homicides but not rifles and shotguns? What sort of guns do people own for self-defense? Handguns.

    Quote from Kellermann: "Third, it is possible that reverse causation accounted for some of the association we observed between gun ownership and homicide -- i.e., in a limited number of cases, people may have acquired a gun in response to a specific threat."

    Which makes sense when almost nobody in the study was killed with a gun that was the same one kept in the household.

    ReplyDelete
    Replies
    1. Again, you not reading the study makes rebuttal quite easy.

      Again, you elect to quote Kleck by way of John Loot who has claimed that the "vast majority" of decedents were killed by a gun brought into the home by an intruder.

      Table 1 of Kellermann’s paper [p.11] shows that only 14% of the
      homicide victims were killed by intruders. Note that this includes homicide by any means, not just guns.


      Here is what Kellermann actually said: "In 8 of the other 14 homicides, the investigating officer specifically noted that the gun involved had been kept in the home."

      Note the word 'other'. These were 14 homicide cases not included in the study, because the proxy could not answer whether or not the victim kept a gun in the home. As a general check, Kellermann investigated these 14 cases, and found that in 8/14 or 57%, the police noted that not only was there a gun kept in the home, but it was the gun used in the homicide.

      This would normally be interpreted as pretty good confirmation of Kellermann's findings in the cases where the proxies answered, which found that 45% of the homicide victims kept gun in the home, But according to Lott and Kleck because the other 430+ guns weren't identified as being from the home or coming from outside the home--it must mean they all came from outside the home.

      Lott also claims the vast majority of decedents were killed by intruders. Kellermann actually notes fewer than 8% were killed by intruders with a gun. So, if you believe 8% constitutes a vast majority...I'm afraid you live in a different world.

      That aside, you dodged my earlier question. If we assume that all the decedents were killed by a gun brought into the home --why is it those households with a gun had a murder risk three times higher than homes that didn't?




      Delete
    2. Sigh.. for the last time that statistic is no from Lott or Kleck, it's from the 1998 Kellermann.

      Kellermann conducted a study in 1998 which is here: http://www.ncbi.nlm.nih.gov/m/pubmed/9715182/

      Which is almost identical to the 1993 one (same counties covered, slightly different time period)

      In it the authors look at gun homicides in the home and report:

      "among those 438 assaultive gunshot woundings, 49 involved a gun 'kept in the home where the shooting occurred,' 295 involved a gun brought to the scene from elsewhere, and another 94 involved a gun whose origins were not noted by the police"

      So at best only 14% were committed with guns belonging to the household. Now even if you assume that all the 94 guns classified as unknown were the ones kept in the home (highly unlikely because not all spouses and lovers live together) the majority were still killed with non-household guns.

      Don't believe me Jade? Even Kellermann admits it:

      http://www.deepdyve.com/lp/sage/response-to-kleck-PYlvPkLq5x

      There you go. Here Kellermann strongly implies that the majority were NOT killed with household guns. Essentially proving Kleck's point.

      I believe that the effect Kellermann detected just proved that people at a higher risk of homicide acquired handguns for self defense prior to being murdered and/or they tried to use their guns in self defense and failed.

      What is a logical fact is that the majority were not killed with guns kept in the household. Household guns are just as kleck noted, rarely used to commit a murder. The household guns was rarely the cause in most of these deaths.

      It baffles me why people constantly cite this study but are completely ignorant of how the effect functions. By all means continue to be a denialist.

      And lastly it is not three times, it is 2.7 times (2.3 fpr males, 3.6 for women)

      Delete
  8. Sweet Jeebus, AW.

    So you admit you didn't read the study by citing *another, different* study? Wow . Isn't that a bit like doing a book report on a book you didn't bother to read claiming that it's ok because you read another book by the same author? Kind of like claiming Bartleby the Scrivener is just like Moby Dick because, you know, Melville?

    BTW, I strongly suspect you didn't read the 1998 study either--since Kellermann, et al. are measuring something different altogether.

    Re the 1993 study--here's your confusion: " Here Kellermann strongly implies that the majority were NOT killed with household guns. Essentially proving Kleck's point."

    Actually, Kellermann states no such thing. The study did not find that the “family gun is more likely to kill you or someone you know than to kill in self-defense”. It found, and I quote from the abstract “guns kept in the home are associated with
    an increase in the risk of homicide by a family member or intimate acquaintance.”

    BTW, this was another bogus point first raised by John Lott in a WSJ op/ed.

    ReplyDelete
  9. Pretty much all evidence points to homicides involving guns kept in the household being a rarity.

    ""In how many of the homicides was the victim killed with a gun that was kept in the house rather than a gun that was brought to the house by the perpetrator?" The question is a relevant one since, as the letter also notes, the study's authors had stated in part based on their findings that "people should be strongly discouraged from keeping guns in their homes [p. 1090]." In other words, advising people against keeping a gun in the home doesn't make sense unless it causes an increase in homicide risk."

    Kellermann never gave an answer to that. Furthermore the dataset he released in 1997 didn't even contain data on the origin of the guns used in household homicides. Oddly enough such information was available for his 1998 study..

    I am citing the 1998 study because it contains a section of assaultive deaths by gun shot, which was the same thing Kellermann was measuring in 1993. The only thing the 1998 study did was also look at suicides and accidents in addition to homicides.

    It boggles my mind how you can continue to believe the majority were killed with guns belonging to the household. The fact is that they weren't. That's just how the cookie crumbles I'm afraid.

    So are we done or as you just going to go off on unrelated tangents about Lott?

    ReplyDelete
  10. "Pretty much all evidence points to homicides involving guns kept in the household being a rarity. "

    Not even close. Again, look at what the 1993 study was about. Quote "guns kept in the home are associated with an increase in the risk of homicide by a family member or intimate acquaintance."

    Can't get much clearer. Additionally, if as you say, guns in the house didn't contribute to homicides--this raises 2 questions:

    1. why is the homicide risk in homes without guns 270% less?
    2. ad why aren't these guns protecting their owners from an increased risk?

    Again, the 1998 study studied something altogether different. Hint: it's in the abstract you linked to.

    There is no tangent re Lott--he was actually the first to bring up these bogus aguments--always in op/eds, not studies. Kleck merely parroted them.

    ReplyDelete
    Replies
    1. >1. why is the homicide risk in homes without guns 270% less?

      People generally own handguns for self defense. One can expect a positive association with homicide because people at a higher risk of homicide, or in such a situation would acquire a handgun for self defense. Again even if said handgun that is kept in the home was not the one used in the homicide.

      >2. ad why aren't these guns protecting their owners from an increased risk?

      Beats me and I don't care because I don't own handguns for self-defense.

      What I do know is that angry spouses, or teenagers, etc grabbing the household gun and murdering somebody else in the home is far from common.

      Honestly if people knew the majority of decedents in studies like Kellermann were not being murdered with household guns, it would be cited far less, and far less critically. Guns kept in the household are rarely used to commit a homicide.

      Have the last word.

      Delete
    2. "People generally own handguns for self defense."

      Ok, then why are households with guns associated with a much higher risk of homicide? Can't have it both ways, AW.

      "Honestly if people knew the majority of decedents in studies like Kellermann were not being murdered with household guns, it would be cited far less, and far less critically. Guns kept in the household are rarely used to commit a homicide."

      Again, no evid ence to support this claim. In fact, the 1992 study shows a majority of homicides were committed with the gun in the household.

      Delete
    3. I really should stop but....

      >Ok, then why are households with guns associated with a much higher risk of homicide? Can't have it both ways, AW.

      You most certainly can. It's called reverse causation.

      Again Kellermann says: "Third, it is possible that reverse causation accounted for some of the association we observed between gun ownership and homicide -- i.e., in a limited number of cases, people may have acquired a gun in response to a specific threat. If the source of that threat subsequently caused the homicide, the link between guns in the home and homicide may be due at least in part to the failure of these weapons to provide adequate protection from the assailants"

      "Again, no evid ence to support this claim. In fact, the 1992 study shows a majority of homicides were committed with the gun in the household."

      Simply incorrect. Kellermann does not explictly mention how many homicides were committed with household guns in his 1992 study, but in his 1998 study which also looked at homicides in the home the majority were not killed with household guns.

      This leads us to two things:

      Kellermann had the data for the origin of guns used in homicides in his 1992 study. He didn't report it because of reasons (because most were not killed with household guns).

      Or:

      Kellermann did not have the data (highly implausible he would have such data for his subsequent study which was pretty much identical)

      Now when Dr. Mark Ferris sent a letter asking how many homicides were committed with household guns, Kellermann did not reply.

      People asked Kellermann to a release a dataset with this information. The dataset he released did not contain that information.

      Oddly enough however Kellermann had no problems reporting the percentage of people killed in homicides by household guns in his 1998 study. My guess is that because the study didn't exclusively look at homicides (also suicides, and accidents).

      It's pretty much proven that the 1993 study were not killed with household guns given Kellermann's reluctance and cloak and dagger antics. Why? Because the 1998 study was nearly identical to the 1993 one as it looked at the same counties and gun owning households.

      Kellermann's response to Kleck sealed the deal for me. He basically states that the gun kept in the household doea npt have to be the murder weapon to raise the households risk of homicide.

      Delete
    4. I should really stop...

      >Ok, then why are households with guns associated with a much higher risk of homicide? Can't have it both ways, AW.

      You most certainly can. It's called reverse causation.

      Quote from the K-man: "Third, it is possible that reverse causation accounted for some of the association we observed between gun ownership and homicide -- i.e., in a limited number of cases, people may have acquired a gun in response to a specific threat. If the source of that threat subsequently caused the homicide, the link between guns in the home and homicide may be due at least in part to the failure of these weapons to provide adequate protection from the assailants."

      >Again, no evid ence to support this claim. In fact, the 1992 study shows a majority of homicides were committed with the gun in the household.

      Plenty of evidence. His reply to Dr. Ferris and Kleco, and most importantly his 1998 study which also looked homicides and gun ownership in a household. Prove that the majority involved household guns, you simply cannot.

      Posting again in case my previous comment does not go through.

      Delete
  11. "You most certainly can. It's called reverse causation."

    As Kellermann notes it is possible. However, you badly misinterpret the Kellermann quote. He notes it's possible in a *limited* amount of cases. Even so, he notes the fact a gun in the home may not be the self-defense magic talisman gunloons claim. But after noting this, Kellermann, et al conclude even the possibility of limited reverse causation don't alter the findings.

    Again, the 1998 study you haven't read measured something altogether different, in a different time and using different locales. Pretending you can cherrypick data from one study and apply it to a wholly different study is foolish.

    ReplyDelete
    Replies
    1. I never said that. It's just that when you couple it with the fact that all of the risk was attributable to handguns, and that the majority of the killings were not with the household gun. It all falls together in place.

      The 1998 study also looked at homicides and two of the same cities as the 1993 study.

      The problem is that you are assuming that only intruders were using non-household guns, you assume that all intimate acquaintances and family members live together with the decedent, when that is highly unlikely. Thus they would obviously use non-household guns, or at least a significant amount of them would.

      Delete
    2. Not assuming a thing. Intruder has a specific meaning; you can look it up in the dictionary but the definition is one who enters without permission or forcibly enters.

      Again, reading the survey would help your confusion immensely. Kellermann never claims the intimate relations or family members lived with the decedents. He was merely describing the relationship between the decedent and the killer.

      AW, you've got to read the studies--not just critiques from discredited hacks like Lott, Kleck, Suter. They all mistate what Kellermann studied and then proceed to claim Kellermann didn't have the evidence to support what he didn't say.

      Delete
    3. I only care about the homicides which involved household guns, and it doesn't seem that the majority did. Kellermann's subsequent study which studied pretty much the same thing proves that this is the case, and Kellermann's response to Kleck heavily implies that this is the case.

      I mean even David Hemenway says that most homicides do not involve household guns.

      How much evidence does one need? If the vast majority were committed with household guns, why won't Kellermann just say so?

      Delete
  12. What you care about is immaterial. If you're killed by a gun you own or a gun brought into the home--I'd suggest it wouldn't make a whit of difference to you.

    This is what Kellermann measured--the risk of homicide in homes with guns.

    Hemenway said that? Odd, here's a quote from him: " Case-control studies, ecological time-series and cross-sectional studies indicate that in homes, cities, states and regions in the US, where there are more guns, both men and women are at higher risk for homicide, particularly firearm homicide."

    ReplyDelete
    Replies
    1. Read his editorial in response to Anglemyer's meta analysis that was released in Janurary.

      "The evidence that Anglemyer and colleagues summa- rized about the relationship between guns and homicide is not nearly as strong as that between guns and suicide, per- haps largely because the authors focused exclusively on individual-level studies of victims. Most victims of firearm- related suicide shoot themselves with a gun from their own home. In contrast, victims of firearm-related homicide are, by definition, shot by someone else and most perpetrators do not use a gun from the victim’s home."


      You are right. I don't care. Even if the vast majority of homicides in Kellermann's study were committed with household guns, I'm still going to own guns because according to Kellermann the risk he found was 2.3 for males. The absolute risk (not relative) is still very tiny and something I would not worry about.

      It must suck living in constant fear.

      Delete
    2. Man, you really haven't read either Anglemyer or Hemenway. And how you're trying to tie in Kellermann, who didn't address suicide at all, is silly.


      Who lives in constant fear? The person who needs a gun because he fears, well, everything or the person who doesn't. As for the risk being tiny--not really. If a doctor told you a certain behavior increased your chance of a premature death by 300%--I suspect you'd listen. But, hey, maybe you just need to roll the dice and continue a behavior that actually increases the risk of that which you fear most.

      Delete
    3. Kay.

      "Who lives in constant fear? The person who needs a gun because he fears, well, everything or the person who doesn't."

      I don't own guns for self defense. Nice strawman. It's just a hobby for me, like everything else in life.

      "If a doctor told you a certain behavior increased your chance of a premature death by 300%"

      Not three times. It is 2.7 times and even lesser for males at 2.3 times.

      So it's 130% for males or 170% if you're looking at both genders combined. If it was 300% it would be four times.

      That's the relative risk, not the absolute risk anyways. Go look up what relative and absolute risks are.

      "Even if (and that's a big if), Kellermann's estimate is in the ballpark, a very conservative estimate of the actual homicide risk to each household member being killed per year, where no family member has a criminal record, is in the range of three-eighths of one-thousandth of 1 percent to three-quarters of one-thousandth of 1 percent (.000375 - .00075 percent). Over a forty year period that risk translates to between one-and-one-half hundredths to three one-hundredths of 1 percent of homicide risk for each family member (.015 - .03 percent"

      Pretty damn small.

      We are going around in circles. Ta.

      Delete
    4. Wow, I leave for a few weeks, and Jade still can’t let this thread go.

      Anthony Wilkins: “So it's 130% for males or 170% if you're looking at both genders combined. If it was 300% it would be four times.”

      Jade, Anthony is saying it right here. If you are going to keep quoting Kellermann (which I am sure you will) you should strive to at least get the math parts right. Kellermann claims an adjusted odds risk ratio of 2.7. If you are going to phrases this as a percent increase, you would say you are “170% more likely to be murdered if you keep a gun in the home”, not 270% (or 300% to be less precise).

      I’m just trying to help you out here, but in case you can’t see what we’re talking about, think about it this way: if you have $100, and you increase it by 10%, what are you left with? $110 right? If you increase it by 100%, what are you left with? This is also called “doubling”, or “2x”. In short, if you talk about “percent increase”, don’t just multiply the multiplier number by 100. You have to take off 100% after that to be right. A lot of people get this wrong. I’ve tried explaining percent increase/decrease to Mike, and I’m not sure I got anywhere.

      Then there’s this:

      Jade: “1. why is the homicide risk in homes without guns 270% less?”

      This is egregiously wrong. Mike does this all the time. You can’t just take the percent increase number and then use it on the other side as a percent decrease. If you say something decreases by more than 100%, the end result is a negative number. Kellermann most certainly did not say the dead will rise, or more likely to be reincarnated if you don’t own guns in the home. If you read the study you would know that. To convert Kellermann’s 2.7 risk factor into a percent decrease it would be (1 - 1 / 2.7). You would say you are “63% less likely to be murdered if you don’t own a gun”. Use this number next time.

      …Of course, we’d have to add the disclaimer that you are 63% less likely to be murdered if you don’t own a gun AND you live in an urban environment, since 98% of the population in Kellermann’s data pool lived in an urban environment where more people are murdered and fewer people own guns. And no Jade, having a handful of households that may have been rural doesn’t change the fact that the data is highly skewed. We can estimate that a grand total of five of his 388 households in the study were rural based on the population mix and rates of murder in rural vs. urban environments. Not enough to even need a second hand to count them, but yeah that makes a fair study you say.

      Delete
    5. I was thinking that Anthony was TS. You guys sure sound alike, but when you, TS, say you haven't been around for a few weeks, I believe you.

      I can't say I missed your nit-picking, double-talking bullshit. That windy comment took us completely away from the main point, generally speaking, a gun in the home makes people less safe.

      Delete
    6. A gun in an URBAN home makes people less safe. You kind of missed the main point of what I was saying. And of course if you have to add that urban disclaimer then it's not really the guns that's making you less safe, is it?

      Delete
    7. And you don't have to wonder if I changed my identity, or posting anonymously. I don't do that stuff.

      Delete
    8. TS: A sample size that is limited to cities is one problem. Not controlling for gang membership and drug dealing is another..

      By far the most important one is that Kellermann did not ascertain whether any of the guns involved in the homicides were kept in the household. Kellermann' s 2.7 figure is only valid if you assume that the vast majority were guns kept in the victims household. Kleck says this isn't the case, Kellermann heavily implies that this isn't the case and even Hemenway says that this isn't the case.

      It just boggles my mind why Arthur won't tell us the percentage of guns that belonged to the victims household.

      Delete
    9. Anthony, you and I could probably have a good debate on what his biggest mistakes were.

      Delete
    10. I'm certain you could--given neither of you have read any of his studies.

      I'd pay to see that for the amusement value.

      Delete
    11. TS and AW debating Moby Dick.

      TS: So, the novel is kind of about a whale, right?
      AW: What?? Melville is writing about scribes, for God's sake.
      TS: Which makes sense since Melville is a writer--which is kind of a scribe.
      AW: Exactly! This proves conclusively Melville was wrong about Whales!
      TS: Precisely! And Melville also wrote about ships!
      AW: Which we all know has zip to do with whales!

      Delete
  13. Another quote from Hemeway: "Using survey data on rates of household gun ownership, we examined the association between gun availability and homicide across states, 2001-2003. We found that states with higher levels of household gun ownership had higher rates of firearm homicide and overall homicide. This relationship held for both genders and all age groups, after accounting for rates of aggravated assault, robbery, unemployment, urbanization, alcohol consumption, and resource deprivation (e.g., poverty). There was no association between gun prevalence and non-firearm homicide."


    AW, this is what happens when you choose not to read the doctors you wist to criticize and, instead, parrot bloggers.

    ReplyDelete
  14. In the 1998 study by Kellermann et al he reported that 438 gun assaults and homicides took place in the home.

    He reported that one in five assaults resulted in a death (20%) which would be 87 deaths.

    Amongst those 438 assaults, 49 were committed with a gun kept in the household, 295 brought from elsewhere, and 94 which were not noted by police.

    20% of 49 equals 10 deaths due to household guns.

    20% of 295 is 59 deaths due to external guns.

    20% of 94 is 18 deaths due to unknown guns.

    10 + 59 + 18 = 87

    Conclusion: most household gun homicides do not involve the household gun.

    ReplyDelete