Was that contention part of the study’s conclusions, or was that something someone else said? If the former, it certainly reflects badly on the quality of the study. But if the latter, it doesn’t.
Think this through with me: pretty much everyone I know who doesn’t own guns, also doesn’t want friends or relatives bringing guns into their house. Many - not all, surely, but many - people who do own guns would, I assume, have a greater tolerance for that. There’s how outside guns are more likely to be in the homes of people who own guns than the homes of people who don’t.
And once there, they’re just as much of a risk factor as the guns kept there. Maybe more, because they’re on the persons of their owners, rather than locked in a cabinet somewhere else in the house.
Now if a gun owner sells all of his guns, that’s not going to change the presence of outside guns in his home, if he doesn’t also change his tolerance of friends bringing guns into his home. If those guns are still welcome, they’ll still be a risk factor. Getting rid of his own guns hasn’t affected the risk of shootings by other people’s guns in his home.
A control group that’s too different from the test group on key variables is a big enough flaw to kill any study.
Thanks, I read that part but my mind apparently didn’t process it when I typed it.
Ok - I always took ‘plausible’ to be roughly equivalent to probable. I think the control group engaged in a less risky lifestyle and therefore experienced less overall risk, including risk of homicides.
I agree that one cannot take from this study that a gun in the house makes you safer. You also cannot take from this study that a gun in the house makes you less safe. The reason is because the study is badly flawed and those flaws make the findings worthless.
What contention are you referring to? The study specifically says:
The part you quoted from me where I said ‘Since many…’ is not from the study, that’s my extrapolation, though I think it’s a fair one.
This is a good point. I don’t let anyone bring a gun into my house unless it is unloaded. There is no need for a loaded gun in my house. Now, I did have a “no guns” policy entirely for quite a while, but as I have a number of friends who like guns, and that see each other when they come to visit me, they wanted to be able to show off their guns to their friends at my house. So, I allowed that, but only if they are unloaded.
Now, as an analogy, I used to have a “no dogs” policy in my house. It wasn’t that I minded dogs, it’s just that I didn’t really have a need for them to be in my house. Then I got a dog of my own, and now my policy is “dogs welcome.”
My chances of being attacked in my house by a dog went up when I got a dog, as well as my chances of being attacked by a dog from outside of my house. (not by much, of course)
I think that that is a conclusion, not a given, and it is one that there are many different viewpoints on.
What I would say is that this study has very little relevance in any way, regardless of how it was performed, because it was done 25 years ago, largely with data that was a bit dated when compiled.
Demographics have changed, culture has changed, technology has changed, even the technology of guns has changed. An entire generation has gone by since the study.
That is what I would attack about the study, personally, that it no longer has any relevancy. Discrediting it on based on whether you agree with the methods used seems a bit pointless, IMHO.
So the scenarios that I brought up in post#2 and #5 are just too ludicrous give any thought? It’s impossible that an armed person would come to steal your guns? That your gun might turn an angry confrontation into a lethal one? That a person might bring his gun when confronting you because he knows you have a gun in the house? Basically, the idea that a gun in the home attracts violence is so unbelievable that it damns the study because of that suggested linkage?
Often true. But if, for instance, drug use is highly correlated with homicides in houses with guns, the big difference in drug use between the homicide and control groups would not invalidate the study. Someone heavily involved with drugs would have a higher risk of getting killed and would also be likely to own a gun, and so controlling for this by selecting the control group to have the same percentage of drug use would skew the results.
What has changed so much in the past 25 years?
I’m sure many a professor would love to get a grant to redo the study. However wouldn’t a similar result, that gun ownership increases the risk of homicide, be considered as anti-gun propaganda and be banned by the bullshit law?
I’m not sure what thought you think wasn’t given to post #2 and #5. I responded to post #2 and post #5 wasn’t directed at me. Post #2 posits a hypothetical that is along the same lines as the ones that Voyager raised. I think it’s possible, but the data presented in this study doesn’t support it. It doesn’t reject it either.
Post #5 is along the same lines, that the presence of a firearm can act as an escalation turning what wouldn’t previously been a homicide into a homicide. The majority of the incidents were non-forced entry (84%) involving someone known to the victim (77%) as a result of an altercation or romantic triangle (51%). The idea you’re presenting is that the presence of a firearm in the home had some kind of escalatory impact on those circumstances. I think that’s possible. **If **that were true, would the primary recommendation be to not have a firearm in the house, or to not associate with or allow people prone to violence in your house? Something else, both maybe?
Put another way - is the person who associates only with the finest upstanding people but also possesses firearms in more danger than the person who doesn’t possess firearms, but who associates with people prone to violence at more risk? I think it’s misinterpreting the data if one were to say that having a firearm in the house make you less safe, or that it causes some kind of increased homicide risk, etc. And that’s how the study is often used.
I guess the problem is, that you seem to keep hanging your hat on this as a “damning” feature of the study. It’s not. If you want to refute someone using this study as a cite, you could say “correlation /= causation” and that would be fine. But that would not be refuting the study which merely showed correlation. And it’s not wrong to say “based on this strong correlation, we recommend not doing x”. That’s how warnings against smoking started. That’s a principle in medical studies, if you remember the “circumcision=less AIDS” African study. It would have been immoral for them to continue the study when the results were coming in so lopsided.
Your arguments about the differences between the control group and the study group makes way more sense if you want to refute the study.
What Bullshit Law? Look, the CDC is banned from doing gun studies which have the purpose of advocating gun control.
*…
The amendment, which was first tucked into an appropriations bill signed into law by President Bill Clinton, stipulates that “none of the funds made available for injury prevention and control at the Centers for Disease Control and Prevention may be used to advocate or promote gun control.” A similar provision was included in the Appropriations Act of 2012.
Named for Republican Rep. Jay Dickey of Arkansas, a self-proclaimed “point man for the NRA” on The Hill – the Dickey amendment does not explicitly ban CDC research on gun violence. *
And I’m quite aware that it can go the other way. The correlation can be a bit of a red herring, like many studies showing marijuana is a gateway drug because most heroin addicts tried it first but ignoring that they tried cigarettes and alcohol first oftentimes and probably had a bunch of other shitty life situations in common as well. That’s why “didn’t control for other variables” is a reasonable objection to me but not “but you didn’t PROVE ‘X’ caused that”.
I don’t think any case control study of this nature will show causation because it’s not intended to or designed to do that. So while it’s true that correlation != causation, I don’t think that particular criticism is very strong for this particular study, or any case control study.
The issue with the conclusion is that it’s misleading. The recommendation misleads readers and users to imply causation. I’m not trying to refute the study in saying the data is wrong per se. It’s that the weaknesses presented in both the method and analysis are sufficient that the conclusions are not supportable. Here from the discussion section (my bold):
The study doesn’t show that possession for personal protection is counterproductive. In fact, the study says this:
It’s hard for me to parse this precisely, but the way I’m reading it is that 15 (3.6%) of the homicides included in the total results were as a result of self defense, though it’s unclear how many if any were from firearms. Firearms certainly aren’t a talisman against bad outcomes, but I don’t see how the conclusion that personal protection is counter productive can be drawn if the study wasn’t evaluating for personal protection in the control subjects. That’s not what was asked in the controls, and that’s not the basis for choosing the case examples.
The discussion section also says that the greatest threat to the lives of household members appears to come from within. But that’s not supportable either. Based on the data the largest plurality of groups of people were friends or acquaintances (31%). And when looking at the method of homicide, just more than half were with non-firearms (50.2%).
Those are the types of conclusions that are drawn from the data that are objectionable. If the study simply said, firearms present a risk and should be evaluated against other risks and other benefits, then sure I’d agree with that. But when it says, the greatest threat to the lives of household members are firearms [in the home] that’s misleading.
They didn’t say “personal protection is counter productive”, they said a gun in the house is counter productive to personal protection. Or am I missing something?
I’m not seeing the distinction, but I quoted the section of the study that I was referring to. When I said ‘The study doesn’t show that possession [of a firearm in the residence] for personal protection is counterproductive.’ it should be read with the bracketed parts.
Look, ignoring the control group problems, if you are 25%(or whatever) more likely to die when there’s a gun in the house then a gun in the house is counter productive to staying alive. You seem to be saying “but a gun in the house IS personal protection!”. That’s what they were studying! Shotgun traps all over my house may be intended as home protection but that doesn’t mean I’m safer with them.
I’m saying that the study wasn’t intended to measure the impacts or effects of firearms as personal protection. It’s not how they selected the case examples, and it’s not what they asked the control examples. Concluding on personal protection while only evaluating the cost side of the equation (even providing for the fact that they did identify incidents of personal protection) isn’t supportable. Because it’s not supportable, the study shouldn’t draw those conclusions.
I agree. I grew up with guns in the house. My father had a semi-automatic on his side of the bed, my mom had a revolver on hers. I was shown where they were and how to properly handle them when I was 7. I got my first firearm when I was 9.
I’ve had firearms in my home pretty much all my life, with some exceptions. My home is safer as a result.
No self-respecting scientist does a study with the purpose of advocating anything. Any study rigged to try to prove anything is flawed and would be rejected by any decent reviewer no matter what their position on the subject is.
So there was no need for the law, in terms of fairness. Why have it? Obviously to intimidate the CDC from funding studies that might produce results calling the wonderfulness of gun ownership into question. Like the one discussed here.
NRA lackeys are so, so concerned about the integrity of science.
So I stand by calling it a bullshit law.