I pretty much agree.
It didn’t? Cite?
Kellermann et al., 1993.
You should know it, you brought it up. Just read it.
OK, quote what they said about rifles and shotguns. I have the study. I’ll tell you if you are honest.
If you have it, just read it, bitch.
I read it cocksucker, highlighted it too, and I don’t see anywhere where it says what you just said. So cite, or admit you can’t. 
Not my point. Insurance companies also have a lot of experience with statistical analysis. Perhaps on a par with the best of the CDC, perhaps not, but reliably competent. After all, its what they do.
They assess the risk, crunch the numbers, and come up with how much money they are most likely to pay out. What they add to that in profit doesn’t enter into this. Just that a reliably objective investigation is likely to be conducted, because money. If they do the research, come up with their actuarial numbers, and put their money where their math is…pretty convincing, when objectivity is what you’re after.
You do want that, right, Kable? Objective, unbiased research? Not everybody does, you know. There are people who would rather nail gun their scrotum to a two-by-four that see it happen.
I love objective, unbiased research but I wouldn’t trust and insurance company to do it. Didn’t you see Sicko?
Piffle, 10 y/o’s can be expected to bounce off the walls, get into shit and distract you; you shouldn’t bring a weapon into that setting without an overriding reason, if you do, you are not following safe firearms practices.
This is beautiful. You’re lying, but there’s no good reason why you should. The article is freely available and easily found in a simple Google search.
Actually, I do know why. It’s because you’re just regurgitating something you were told by some gun stroker but you don’t really understand it and don’t have the capacity to problem solve. The limits of your cognitive abilities appears to be pasting YouTube links.
You wouldn’t even be able to figure out where the initial indication of a lower risk for homicide associated with a shotgun or rifle comes from in the first place in that article.
Table 4 in the article is where the authors present their final multivariate model. Table 4 is where you will see that long guns are not statistically significant after accounting for other variables.
Do you need a lesson in the third variable problem, confounding or moderation? I have a couple examples; perhaps I could even help you to understand.
Not that I disagree with this case, or your general premise, but don’t some bullets have the capacity to ricochet? Meaning that with the perfect combination of really stupid circumstances you COULD have a mechanically defective gun that got jostled and fired at a wall, then the bullet ricocheted and hit somebody. So the gun did not have the trigger pulled (mechanically defective) and was not pointed at anyone (ricochet).
I mean, we’re talking .0000001% chances here, but technically.
While one should exercise common sense 10 y/o’s shoot guns all the time without incident. Firearm related hunter safety courses start at age 9 in my state.
http://www.azgfd.gov/i_e/edits/hunter_education.shtml
Age is not a factor in any gun safety rules I’m aware of:
Regarding the example in question the father obviously did not follow the safety rules cited in the above link. Following them, would have prevented the tragedy.
Lying how? Cite where it says the findings of decreased homicides in homes containing shotguns and rifles (of which AR15s are a subtype) did not “hold in the final model.”
I just want to see you cite that the study found what you said it found.
OK, so you did read it and you agree that homes with rifles and shotguns had less risk of homicide. Isn’t that cool.
Table 4 does not mention long guns at all. No stats presented to say if the actual findings of less homicides “held up” or not, rather looks like they were omitted. Or do you see something there I don’t.
That’s OK, just tell me why you think homes with rifles and shotguns had less homicides. Also tell me why you think Kellerman ommitted the effects of alcohol in table 4. Though he found handguns were associated with higher risk, how great was this effect, say in comparison to having someone in the home drinking just one beer, or renting their house? Thanks for your time.
I am happy to see you are actually reading the article now. I hope in doing so you will learn something.
Are you just being obtuse? Firearms safety courses are an overriding reason, cleaning, not so much.
Unpredictable behavior (not age) is a factor in
I learned how to clean a gun when I was a kid.
Of course, and keep your finger off the trigger, and you might want to unload that thing before you clean it. In fact I can’t think of any guns that clean very well while loaded. The father obviously broke not one but a number of safety rules. That’s why the NRA should back gun safety in public schools.
Way to kick the can down to the next generation, Kable. In the mean time, what do we do with all the idiots who can’t handle their gear safely?
Why specifically the NRA, the gun manufacturer’s lobby, and not local law enforcement? If I remember correctly, our local PD did come into the school to talk about guns. As in, if you see one, don’t touch it, and tell an adult immediately. I see no reason why “how to unload and clean a gun” is necessary in schools or would do anything to decrease the likelihood of gun accidents, since that curriculum actively encourages children to pick up guns, which increases the chance it will discharge.
Lying that you read it and highlighted anything it in related to your claim. Lying that Kellermann reported that long guns were related to lower rates of homicide.
I’ve done this repeatedly. Table 4 shows all the variables that remained in the model after accounting for the effects of the other significant predictors. That’s the cite. I’ve said it before and I’ll say it again. The headings on the table and the text makes this clear. Table 3 says “Univariate analysis of hypothesized risk or protection factors…” This is a list of possible factors related to homicide outcomes, and it is where you see long guns show up.
“Univariate” means one variable. It is the independent relationship between a variable and the outcome of interest.
Table 4 says “Variables included in the final conditional logistic regression model…” It includes “home rented,” “lived alone,” “fight in the home,” “house member arrested,” “used illicit drugs,” and “guns kept in the home.” That’s the cite, dummy.
For further explication, the text of the Method states
You’ve allegedly got the cite right there in your hands. You’re just too stupid to read and comprehend it.
Wow, are you dumb. First of all see the above. Secondly, and more fundamentally, LONG GUNS WERE NOT EVEN SIGNIFICANT IN THEIR INDEPENDENT ASSOCIATION WITH HOMICIDE.
Someone told you that, but they didn’t understand what they were reading either. Do you know what an odds ratio is? It’s a measure of association, and an odds ratio of 1.0 means that there is no relationship between two variables. The odds of an outcome are exactly the same for each unit of the predictor.
An odds ratio below 1.0 means that the relationship is inverted, so that increasing units of the predictor are associated with a reduction in the likelihood of the outcome occurring. So, an ignorant person would look at Table 3 and think, “Aha, the odds ratio for shotguns is 0.7 and for rifles is 0.8. This means the odds are lower for homicide when either one is present.”
The problem is that this interpretation requires that you do not look at the numbers immediately to the right of the odds ratio. You see where it actually says “0.7 (0.5 – 1.1)” in the row for shotguns, and “0.8 (0.5 – 1.3)” in the row for rifles? The numbers in parentheses are the 95% confidence interval. This will tell you whether the statistic is significant or not. What it means is that you cannot, at a .05 level, be sure that the actual odds ratio does not include 1.0, or even 1.1 (or 1.3 for rifles). This means that even in the univariate relationship, rifles and shotguns are not protective. That’s even before accounting for confounding by other variables, which is what Table 4 shows you.
Yes. I see the methods section, which explains why variables were omitted. They were omitted because they were not statistically significant. It’s pretty simple, really.
They didn’t. I’ll repeat that for the thinking impaired. THEY WERE NOT SIGNIFICANTLY ASSOCIATED WITH A LOWER LIKELIHOOD OF HOMICIDE.
Please see the Methods section of the paper, and in particular the part I quoted for you. Also, see the Discussion section, because Kellermann specifically discusses the fact that that variable did not survive in the multivariate model on page 1089. The fact that you have to ask makes me say again that you do not actually have the paper and have not read it.
The odds of having one or more guns in the home were associated with a 2.7 times greater likelihood for being the victim of a homicide, after taking into account any variance explained by the other five significant predictors in the model. To draw a conclusion about the relative strength of one predictor in the model compared to another predictor in the model (e.g. your question about having a gun versus renting), you would have to do a specific test comparing the parameter values. I do not see such a test having been done (and unless there’s a specific empirical question, such tests usually are not). They did report testing for interactions, and finding none. That means that the odds associated with having a gun in the home were not different for different levels of the other variables in the model.
Since alcohol was not a significant predictor after accounting for the other predictors that were retained (see Table 4), one can infer that it is not as strong a predictor as having a gun in the home. If that was a specific question of interest, a better method would be to directly compare them, as discussed above.
It would seem to me that not only have I read it, I have understood what it means. I don’t think either is true for you. I think you come off looking smarmy and stupid.
Okay, I realize that it was really stupid of me to waste all the time and energy to explain all of that to Kable. Even if he understands what I wrote, he’ll still be claiming in a week or less that Kellermann reported that having long guns in the home was protective.
I’m going to do something I never do here. Just to make me feel less stupid for investing the effort, could someone else let me know that they found that post helpful? Come on, even if you have to lie to me!
I’m just glad you finally found something to do other than defend Dio the thread derailer.