So you killed your kid while cleaning your gun.

[Measure_for_Measure]

@Kable

OK, so you’re not so interested in decreasing accidental gun deaths. That would take away from your liberal talking points. I get it.

How does that follow? I said I’d have it in schools where there are lots of gun owning families. You may not get this, but lots of people have very little lifetime exposure to bang-sticks: this tends to vary by geography.

It’s self evident. Lots more people handle guns at some point in their lives then own one.

One of the principles of toxicology is, “The dose makes the poison.” So lots of exposure gives you (proportionately) greater risk than less exposure and correspondingly greater need for training.

Great, so if your child is over at a friends house who owns guns but is not trained in gun safety, do you think it would be better or worse if your son knew something about gun safety? Do you think it would be better that your sons friend were not taught gun safety?

I suppose it depends upon how much class time we’re talking about. Part of the problem is that upthread you noted that these things must be practiced: kudos for pointing that out by the way. So while exposure to the 4 rules might take all of 5 minutes of class time, I can’t see devoting more than that in a school district that’s not immersed in gun culture.

This just shows how biased you are against guns. Tobacco is inherently harmful. Guns are not. Guns stop crimes and save lives every day, tobacco not so much. Guns are a constitutional right, tobacco isn’t.

Silly. I’m not attacking guns, I’m attacking those who treat the NRA as a valid source of information. Those clowns want to arm maniacs, crazies and criminals, all the better to jizz up their membership into buying more product. You can see it in their monthly ad circulars, which they call magazines. The NRA opposes routine background checks during gun shows: they think the best thing to do with deviants is to hand them a weapon and turn them loose at the nearest stadium or shopping mall. And from the point of view of their clients, they are entirely correct.

If the NRA cared about the health of their membership, they would support responsible friend and family members separating guns from those in a vulnerable state. But gun safety and public welfare is the furthest thing from their minds.

[elucidator]

@Measure_for_Measure

…they think the best thing to do with deviants…

Which is totally not true. The best thing to do with deviants is to give us lots of money.

[outlierrn]

@elucidator

Which is totally not true. The best thing to do with deviants is to give us lots of money.

The best thing to do with deviants involves pineapples and midgets.

BTW, Kable, you’re an idiot, I’ve lost interest in taking you seriously.

[Damuri_Ajashi]

@Kable

I’m just happy to see my side is winning the gun de..squandered their temporary emotionally charged advantage.

I am also happy that the constitution is winning but I am a bit bummed that the other side squandered an opportunity to pass meaningful legislation on a retarded AWB.

@elbows

The expenses for extensive gun ownership are manifest, emergency room/medical costs, police, court, clean up costs, families losing breadwinners, falling into poverty and social assistance, to name just a few. The burden of those costs should be on gun owners, not those who refrain from owning guns.

Oh you mean like we do with gambling and alcohol (two things that are not constitutionally protected).

What is it the Right is so found of saying to those working full time jobs and still can’t afford health insurance for their families? Oh yeah, I remember, “if they don’t like it they should get a better job so they can afford it!”

You make a mistake if you think that gun owners are all conservatives. You just end up pushing a lot of swing voters in places like Florida, Virginia, New Hampshire, Pennsylvania, Ohio, Indiana, etc.away from the Democrats

@elucidator

Ah, its is my good amigo, Señor Moment. Where I can’t remember where I read this, about a bunch of amendments included in recent legislation that forbid the government from collecting any such data. Because freedom. No, wait, liberty.

Anybody? Beuller? Beaulah? Booyah?

What?

[elucidator]

Just read it in the news, last several days, maybe a week, about how lobbyists had managed to get several amendments inserted into otherwise unrelated bills to prevent the gathering of data about gun ownership, etc. Couldn’t remember where I saw it (hence: “senior moment”) but hoping somebody else had seen the same.

That’s the trouble with the hive mind, if one of us forgets, we all forget.

[Hentor_the_Barbarian]

One more try:

1. If the 95% confidence interval of an odds ratio includes the value 1.0, the relationship is non-significant. (The p-value would be greater than 0.05, if it were shown.)

1a. The odds ratio for shotguns and rifles each include the value of 1.0. (Raise your hand if you don’t see where that is found in the paper.)

2. As I already quoted for you, in the Kellermann study, non-significant variables were not retained in the final model. (This is the standard practice in such studies.)

Kable, you don’t understand what a 95% confidence interval is. You don’t understand how regression analyses are conducted or reported. You don’t understand the basics of very much at all.

Measure for Measure explained why cases were excluded in Table 4. It doesn’t really matter though, because rifles and shotguns were NOT SIGNIFICANTLY RELATED to homicides and BECAUSE THEY WERE NOT SIGNIFICANT would not have been retained in the analyses beyond Table 3.

This is incontrovertible. Find someone in your real life who is smart to explain it to you, if you know of anyone like that.

[Kable]

@Measure_for_Measure

How does that follow? I said I’d have it in schools where there are lots of gun owning families. You may not get this, but lots of people have very little lifetime exposure to bang-sticks: this tends to vary by geography.

OK, so you think we should perhaps teach gun safety to rural kids but not so much inner city kids?

One of the principles of toxicology is, “The dose makes the poison.” So lots of exposure gives you (proportionately) greater risk than less exposure and correspondingly greater need for training.

One exposure to unsafe gun handling can be one too many.

I suppose it depends upon how much class time we’re talking about. Part of the problem is that upthread you noted that these things must be practiced: kudos for pointing that out by the way. So while exposure to the 4 rules might take all of 5 minutes of class time, I can’t see devoting more than that in a school district that’s not immersed in gun culture.

I figure we should give it as much training as we devote to sex-ed.

Silly. I’m not attacking guns, I’m attacking those who treat the NRA as a valid source of information.

Silly, have you taken or read the curriculum of an NRA gun safety course?

Those clowns want to arm maniacs, crazies and criminals, all the better to jizz up their membership into buying more product. You can see it in their monthly ad circulars, which they call magazines. The NRA opposes routine background checks during gun shows: they think the best thing to do with deviants is to hand them a weapon and turn them loose at the nearest stadium or shopping mall. And from the point of view of their clients, they are entirely correct.

We already have routine background checks at gun shows. You must be buying into all that Brady campaign circulars if you think otherwise.

But gun safety and public welfare is the furthest thing from their minds.

Who offers more, and more comprehensive, guns safety courses than the NRA?

[Kable]

@outlierrn

T
BTW, Kable, you’re an idiot, I’ve lost interest in taking you seriously.

You couldn’t recognize sarcasm and I’m the idiot?

[Fear_Itself]

@Kable

Who offers more, and more comprehensive, guns safety courses than the NRA?

And it obviously worked so well for Newtown shooter Adam Lanza:

Reuters reported that according to documents released on Thursday, police discovered NRA certificates in the name of Lanza and his mother, Nancy Lanza — whom he killed before starting the shooting at Sandy Hook Elementary School. The two reportedly frequented gun ranges together as Adam Lanza was growing up.

[Kable]

@Hentor_the_Barbarian

One more try:

1. If the 95% confidence interval of an odds ratio includes the value 1.0, the relationship is non-significant. (The p-value would be greater than 0.05, if it were shown.)

Peabrain, the 95% confidence interval is just an arbitrary cut off often used for statistical significance, it still does not prove an actual effect, nor does an less than one odds ratio with a greater than .05 p-value prove there is no effect. The fact is long guns in the home were associated with less homicides. Researches often comment on effects that don’t reach significance, unless of course they don’t want to talk about it because it don’t fit the story they want to present.

1a. The odds ratio for shotguns and rifles each include the value of 1.0. (Raise your hand if you don’t see where that is found in the paper.)

The average was less than 1.0, but you knew that already.

2. As I already quoted for you, in the Kellermann study, non-significant variables were not retained in the final model. (This is the standard practice in such studies.)

A lot of variables were not put in the final model, and he don’t say which.

Kable, you don’t understand what a 95% confidence interval is. You don’t understand how regression analyses are conducted or reported. You don’t understand the basics of very much at all.

I think you don’t understand human nature and bias in research very much. I could suggest some books for you.

Measure for Measure explained why cases were excluded in Table 4. It doesn’t really matter though, because rifles and shotguns were NOT SIGNIFICANTLY RELATED to homicides and BECAUSE THEY WERE NOT SIGNIFICANT would not have been retained in the analyses beyond Table 3.

Rifles and shotguns were assocated with lesser risk of homicide even if it did not reach significance. Kellerman certainly did not show significantly that they were significantly assocated with elevated risk, but then he used the handgun data to implement all guns in his conclusion. Then morons like you, use that as justification, not to go after handguns, but rather semi-automatic rifles, apparently because you think that’s the low hanging fruit.

This is incontrovertible.

That’s for sure.

[ladyfoxfyre]

I love that the guy whining about “bias in research” is straight faced assuring us that the NRA gun safety training is the best possible option.

[Hentor_the_Barbarian]

@Kable

Peabrain, the 95% confidence interval is just an arbitrary cut off often used for statistical significance, it still does not prove an actual effect,

It is in fact somewhat arbitrary, but it is the standard. You can either argue against the scientific standard or point to a paper using that standard to support your position, but you cannot really do both simultaneously.

nor does an less than one odds ratio with a greater than .05 p-value prove there is no effect.

Well, by definition within the context of that study it means there is no demonstrable effect, yes.

The fact is long guns in the home were associated with less homicides.

Or more, since there’s a non-negligible chance the value was actually 1.1 or 1.3.

Researches often comment on effects that don’t reach significance, unless of course they don’t want to talk about it because it don’t fit the story they want to present.

Authors of peer reviewed publications DO NOT interpret non-significant results as positive evidence for the presence of an effect, whether or not it “do” fit the story they want to tell. Peer reviewers and editors do not allow this to happen.

The average was less than 1.0, but you knew that already.

What average? The statistic in question is an odds ratio, not an average. Dummy.

A lot of variables were not put in the final model, and he don’t say which.

He “do” say which. I cut and pasted the text from the methods to make it easier for you to get, but that failed. Variables not significant were not retained. If it’s mentioned in the paper but not in table 4, that’s where he do say it.

I think you don’t understand human nature and bias in research very much. I could suggest some books for you.

You don’t know regression analyses. You don’t understand science. You don’t know the difference between odds ratios and averages. You don’t use proper grammar. You’re an idiot.

Rifles and shotguns were assocated with lesser risk of homicide even if it did not reach significance.

No, they are by definition not. Again, that confidence interval says that you cannot be confident, within a 5% chance, that long guns do not actually INCREASE the odds of a homicide in the home.

Kellerman certainly did not show significantly that they were significantly assocated with elevated risk,

The stats are clearly presented in the paper. In Table 3. Which I have quoted for you here, and have explained now more than once to you. It’s incontrovertible.

but then he used the handgun data to implement all guns in his conclusion.

Because they were significantly related, not only independently, but after accounting for the effects that were due to other significant predictors. All of this highlights the dangers associated with firearms in the home. One or more guns in the home is associated with nearly 3 times the likelihood that you’ll be killed.

[Hentor_the_Barbarian]

To further your education, you might consult an article like the following:

Confidence Interval or P-Value?

In particular: “If the confidence interval does not include the value of zero effect, it can be assumed that there is a statistically significant result.”

Because I assume Kable will be confused by the concept of “zero effect” when the odds ratio is 1.0, let me remind you of this:

The odds ratio is the ratio of the odds for one value relative to another. A value of 1.0 means that the top and the bottom values in the ratio are the same (one value divided by the same value = 1, right)? That’s the “zero effect” for an odds ratio.

When the odds ratio is less than one, that means the top value is lower than the bottom value in the ratio.

The odds ratio tells you what the odds are for an outcome for one unit of change in the predictor. When the predictor is dichotomous, as in “the presence or absence of a gun in the home”, the odds ratio tells you the entire effect of the predictor.

You can interpret the odds ratio as telling you the relative percentage increase or decrease in the odds. But you can only make this interpretation if you can be confident that the odds ratio is not actually 1.0 (zero effect).

[Kable]

Hey Hentor, on average, did people with a rifle or shotgun in their home have more or less homicides? Be honest.

[Hentor_the_Barbarian]

@Kable

Hey Hentor, on average, did people with a rifle or shotgun in their home have more or less homicides? Be honest.

Jesus Christ. You are not capable of understanding this.

When you conduct analyses of samples (instead of magically having all the data from the whole population), you have to figure out if the differences you see just happened by chance or not. The accepted limit at which we have chosen to rule out chance is .05. What that means is that we are accepting that if we had randomly drawn 100 samples instead of one, we would see the difference we’re observing come about in fewer than 5 of those samples. We make an inference that the difference in our sample is not just random, but instead has some systematic meaning.

To determine this in the Kellermann study, since the outcome was dichotomous (the occurrence or non-occurrence of homicide), an odds ratio was presented. That odds ratio cannot be determined to be different from 1.0.

That’s honestly what the paper says.

[Trinopus]

@Kable

Hey Hentor, on average, did people with a rifle or shotgun in their home have more or less homicides? Be honest.

This one, I don’t know, but when you include all firearms, people who live in a house with firearms are at greater risk of violent death and homicide.

It would be interesting to know if this differs significantly by type of weapon. Are long-arms actually safer?

But guns, as a whole, are dangerous to own.

(And, yes, I saw the irony in your comment about women. And I sometimes doesn’t use good grammar. But you have made it abundantly clear to the math and stats experts in the audience that you don’t know anything about the subject.)

[elucidator]

Glad to hear that, as a mathtard, I pretty much have to depend on others. Had to take Algebra One twice to pass. A “D”, and I cheated.

[elucidator]

Now that I think on it, that’s a problem. Like this Kelllerman guy. I see a lot of criticism of his work, which may or may not be based on sound statistical principles. And I also know that every political controversy, be it racial, economic or gun-grabistic, more or less generates a cadre of academic shills. Whores with Ph.D 's.

So when they start talking about declined relativistic stochastic models, or blandly assert that the refined ratio of anal retentivity proves that rifles without the precisely correct bayonet mounts are not, technically speaking, “guns”…I got nothing to go on but my relative trust for the speaker.

Math. Not even once.

[Hentor_the_Barbarian]

Whatever problems there are with the Kellermann study are usually resolved by placing a study in the context of the larger body of literature.

Sadly, for some reason, there isn’t much of a larger body of literature on gun and gun violence.

[Trinopus]

@elucidator

Glad to hear that, as a mathtard, I pretty much have to depend on others. Had to take Algebra One twice to pass. A “D”, and I cheated.

I’m a math geek through and through… When I had to take Six Sigma training for my job, I ended up teaching the damn class…

Serious recommendation: Darrell Huff’s famous book “How To Lie With Statistics.” Damn fine read, an eye-opener, and not really all that technical.

Or, y’know, just hang around these message boards: every single damn fallacy will appear, like clockwork. Dependably. And more often than not (a statistic!) by conservatives.