I would certainly find it easy to believe that gun control advocates knew less, in general, about guns than gun ownership advocates - after all, there’s more likely to be hands-on knowledge in that second group, though of course plenty of anti-gunners are owners and users themselves and I’m sure there are pro-gunners who don’t personally own or use.
But that doesn’t actually mean anything. I could be entirely wrong. And for your part, when you say “It’s all anedotal”, you can’t really follow that with a “but”. It’s all anecdotal. A strong anecdotal impression still doesn’t mean anything. You can’t say, “Hey, it’s not really reasonable to use this as data… but look at all this data, it has to mean something, right?”
It depends on what you have to say. We are in great debates, the reasonableness of your explanation or any rebuttal I might have will be apparent to all. If I reject conclusions that ought not be rejected it will undermine my credibility generally and if you insist that there is certainty where certainty does not exist, it will undermine yours. I didn’t find any of your “arguments” in the pit very convincing but maybe I am being unreasonable and a chorus of support from others will drown out my protests and reveal me for an irrational unreasonable gun nut.
You seem ready to believe that gun rights advocates are generally more knowledgeable than gun control advocates based on their relative exposure to andn interest in guns. I agree. My anecdotal experience confirms that theory, thats all.
My arguments in the Pit are the same as they are here. I just also insult you in the Pit. I have no idea why you think that empirical evidence presented here is any different than empirical evidence presented in the Pit. I do ask that you drop this silly suggestion that I do not rely on empirical evidence in the Pit, or that you leave your personal observations to the Pit where I can respond appropriately.
”Associated” means related in some systematic way. Why do you keep asking this? What does the word “associated” mean to you???
A correlation, in statistical terms, is one of many “measures of association.” There are several types of correlations, as well as odds ratios or relative risk ratios, incidence rate ratios, chi squares… The parameter for a variable coefficient generated in a regression analysis, often referred to as a beta value, is another measure of association.
Do you understand this?
The authors address this on page 551
This is related to the argument that Bone made above; the number is small, making it difficult to conclude that the magnitude of change exceeded what one might expect by chance. Statistical significance is important when attempting to establish a systematic relationship between two variables. However, the data also show that you are wrong to say that firearm homicides did not drop; it’s just not demonstrated here that the gun buyback is why they dropped.
For example, this is from the linked Hemenway (2009) article:
Did you read the Washington Post piece? It sounds as if you did not, because the author clearly states his rationale. He also links you to Hemenway’s explanation directly. Did you read the Hemenway piece? Did you understand it? I find your statement here to be informative, since you’re questioning the authority of the blogger and of another researcher (Hemenway) rather than refuting their reasoning. “Who are they? Just a blogger and someone we are just supposed to believe?” No, you try to understand the ARGUMENT they make, you don’t just evaluate and/or dismiss them based on appeals to authority.
If you had read the Hemenway piece, you should understand it. He observed
Now, you can argue that Hemenway is wrong. You can argue that the point is generally right but wrong in this instance. But when your reaction is “What? Are we just supposed to adopt Hemenway’s opinion about which studies are better?” you are revealing a preference for appealing to authority rather than to the merits of an argument. If you don’t understand it, that’s fine. If you reject it out of hand because reasons, that’s ignorance, at best.
It’s like the climate change science, where one side is arguing the science, and the other side is arguing from appeals to authority or because someone somewhere wrote some criticism of the science, all the evidence is of equal weight.
Aggravated? That seems really, really inappropriate to ascribe to the authors of a paper without any evidence whatsoever. This kind of thing is horribly revealing of your bias and your inability to read this kind of material with any sort of balance. The authors have taken a very careful and considered approach, reporting and discussing data and analyses that are consistent and inconsistent in a very thoughtful way, and you want to insinuate bias – a very ugly kind of bias at that. (And by the way, the authors are Leigh and Neill, a man and a woman. Do you understand multiple authorship?)
Again, this is a woefully inadequate description of what the two authors, a man and a woman, wrote:
This careful explanation, with consideration of multiple possibilities and a description of why the method substitution argument doesn’t make sense in light of the data, is boiled down by you to “[h]e doubt”.
Once more, it’s fine to acknowledge that you don’t understand. It’s horribly inappropriate to mischaracterize what was actually said and then add on your own disgusting insinuations.
Being ready to believe something doesn’t mean believing it. I said I would find it easy to believe, because that’s what my loose judgement tells me. But I don’t *actually *believe it, because I have no reason to. I think that’s a mistake you’re making; you’re taking your anecdotal evidence, putting a general caveat over it of “Well, it’s not really evidence.”, and then accepting it anyway.
Anecdotal evidence means nothing when you’re comparing to an entire nation or larger. It confirms nothing. It refutes nothing. Your personal experience, my personal experience, any single person’s experience means jack. Zilch. Nada.
You’re the one that brought up all the time you spent on this (almost all of the effort was spent in the pit).
I am trying to tease out how certain we are that one was the cause of the other.
I come from a different background from you and where I come from beta loosely translates to variance or risk. I have no idea what chi squared is but what I am trying to figure out how certain we are that one thing causes another thing. I know we can never be absolutely certain of anything but there are degrees of confidence on causation aren’t there?
“Gun homicides have dropped. But they do not claim that homicides generally have dropped significantly and the reason for that is that it is not clear that homicides generally have dropped in Australia.”
I am trying to popint out two things here. One is the substitution effect and the other is the fact that there were so few gun homicides to begin with that any reduction in gun homicide rates gets lost in background noise.
Is there any more convincing evidence that getting rid of firearms reduced overall homicides than there is evidence that it didn’t increase overall homicides? Or is it all consumed by the background noise?
Do you understand why I think the overall homicide rate is more important than the gun homicide rate? Do you understand why I think overall suicide rates are more important that firearm suicide rates? Is there a significance to firearm homicide/suicide rates that is not captured by overall homicide/suicide rates?
The blog is more opinion than analysis and is no more compelling that if you had written it yourself. The fact that he linked to some criticism by David Hemenway is almost as convincing as linking to a criticism by Dianne Feinstein.
Hemenway’s criticism in the link is with the study by the folks from the sporting shooters association, not Lee/Suardi.
I don’t really understand (or agree with) Hemenway’s criticism of the study by the sporting shooter’s association.
It looks like Hemenway is criticizing the authors for how they do their trend analysis (or at least that’s what I would call it given my background). They pick 1979 (a relative high point in suicides and homicides) to begin their trend lines and project those trend lines to the present to see if the gun ban altered the trend. Hemenway criticizes this because if you extend the trend lines forever, the line eventually drops below zero and the gun ban eventually starts bringing people back to life.
Hemenway seems to want to start the trendline at the beginning of time and to show that we have seen a drop in homicides (a drop that started around 1979) after the NFA relative to the trendline that you get if you start in 1915.
I don’t do any trend analysis but I do run into it in my work and going back 100 years to establish a trend seems silly. Going to the most recent high or low to start the development of a trend is something I see much more commonly.
And example that I think might be accessible to a lot of people here is stock price trends. If you’ve ever drawn trend lines for a stock price that is dropping, you don’t start with the prices from the beginning of time, you start with some recent relative high and you draw the line down. That doesn’t mean that you think the company is going to go bankrupt or that the shares will have a negative value but it does help establish when something breaks with the trend.
If he isn’t criticizing how they are conducting their trend analysis then what is he trying to get at?
In any case, I’m not really sure why a drop in gun homicides matter if there is not a significant effect on the rate of overall homicides.
AFAICT, the article’s author is simply expressing a preference for the research on one side of the argument over the research on the other side.
Don’t we just have dueling studies with one side of the argument supporting the study by that says that the NFA was a resounding success and the other side supporting the other studies that say that there was little to no effect? Is this like global warming where all the science is lined up on one side of the spectrum with a token shills on the other side? Are Lee/Suardi shills for the gun industry?
What makes the Author’s preference for one set of researchers over the other set any more credible than my own?
Is THAT what we have here? One side that is overwhelmingly supported by the research and the other side with nothing but a few token shills for industry? Is the Lee/Suaardi paper any less credible than the Leigh paper? Is the Shooting sports paper and less credible than the Leigh paper? Or are you trying to make the gun rights side seem like climate change deniers because they disagree with you?
What?!?!? What kind of ugly bias am I insinuating? Is there some connotation to the word “aggravated” that I am unfamiliar with? Here is there exact quote, please explain to me how I am distorting what they say:
“Because there are so many more non-firearm suicides (and
homicides) than firearm deaths, we cannot reject the possibility that there
was 100% method substitution—i.e. that any reduction in firearm deaths
was accompanied by an increase in deaths by other methods. This is
unfortunate from a statistical perspective but is the inevitable result of
the fortunate fact that Australia already had relatively few firearm deaths
relative to non-firearm deaths.”
WTF are you talking about? here are the exact words, tell me how you think I am being woefully inadequate?
“However our panel specification—in Section 4.1.2—suggests that the time path of non-firearms deaths makes it improbable that 100% method substitution occurred.”
Do I have to quote the study verbatim and in whole to do it justice?
What insinuations? I think you might be reading something that isn’t there.
In the absense of any evidence to the contrary, reason and and anecdote seems like enough to form an opinion. I point to every fucking talking head on guns that we have seen from Newtown to today and say that the talking heads on the gun rights side are more knowledgeable about guns than the talking heads on the gun control side. I compare that with my personal experience where the average gun rights advocate is more knowledgeable about guns than the average gun control advocate. Why can’t io form an opinion based on that? Why can’t I make a statement based on that evidence? Do I need to cite a study to make a claim that seems to be supported by all available evidence?
If you want to make a claim about the majority of a group of people, you need to cite that majority. Or, at the very least, some statistically significant sample of them. A person’s personal experience isn’t evidence except for their personal experience. If you want to make a claim about the people you know, fair enough. But you’re doing much more than that.
I mean, on this message board alone there are more regular posters than I would figure one person has as friends. Would you consider it reasonable to extrapolate how people in general feel or what they’re like based on the posters on here?
In the absence of evidence to the contrary, or evidence to the affirmative, the correct response is to form no opinion. Not say “Well, we don’t have anything that actually means anything, so let’s just go by what the 100 people I know think.” Not having evidence doesn’t mean we get to lower the reasonable threshold.
I was also referring to every talking head we have seen on TV since December of 2012. Sure thats not a cross section of the population of gun rights side and gun control side but the gun control side seemed relatively ignorant compared to the gun rights side ignorant. Do you disagree?
I think it is entirely rational to form an opinion based on what logic and experience tells you without the benefit of a double blind study of a statistically significant sample of the population where none is available? Logic tells me what it tells you, people who support gun rights are likely to know more about guns. Experience tells me that the people who end up talking about gun control on TV are ignorant about guns compared to the people who end up talking about gun rights on TV. This is reaffirmed by my personal experience. Why isn’t it rational for me to form an opinion on that basis? Is the media conspiring to make gun control advocates look ignorant while making gun rights advocates appear knowledgeable?
In addition to what was posted already, these are some of the relevant parts of the study with regard to substitution:
From page 518:
From page 533:
Yes it’s improbably that 100% method substitution occurred. Conversely it’s improbably that 0% method substitution occurred. This says nothing about the magnitude of method substitution. The actual answer is likely somewhere in between, which is what Damuri is arguing. Yes, firearm homicide and suicide declined after the ban. It went from a very small number to a very smaller number. Simultaneous to this rates were already declining, and Australia was engaged in strong suicide awareness and prevention efforts. This doesn’t draw any conclusions about overall homicide or suicide rate.
(my bold)
Unfortunately I can’t access the actual study beyond the abstract.
Ultimately, I don’t see either study that is able to isolate the impact of the gun ban from the already trending rates of both firearm and non-firearm homicide and suicide. And really, it doesn’t matter. It really makes absolutely no difference. The US is not Australia in very meaningful ways. I’m not aware of any conclusions that can be drawn from one population to the other that are meaningful.
Are you attempting to make some type of analogous conclusion from the Australian example that would apply to the US, Hentor? It seems you’re pointing out a factual error that is only tangential to the actual argument being made by Damuri - one of substitution not of firearm suicide and homicides specifically.
We’ve discussed demonstrating cause in the past. You have failed to retain that information.
:dubious:What background changes the beta value in a regression equation?? What do you mean, “translates to variance or risk”? This sounds like you’re just tossing around terms that you don’t understand.
In a regression equation, the beta value is a coefficient that indicates the change in the slope of the regression line for each unit of change in that variable. The variance in the model is a factor in this, and in certain models, an exponentiated beta value can index risk (like an odds ratio), but the beta value is a mathematical concept. It is what it is, regardless of your “background.”
I told you, the chi square is another measure of association. You keep asking what association means, so I’m explaining to you the concept and the different measures of association that exist.
No. Not in research. Again, we’ve talked before about cause. Empirical studies don’t do “causation,” especially causation by degree.
Gun homicides demonstrably fell. They did not get lost in any “background noise”. The rate of change in gun homicides was also associated temporally with the introduction of the new gun legislation in Australia, but the magnitude of change in the context of the overall small number of incidents, was not sufficient to be considered statistically significant.
What does “background noise” mean?
And, as I have asked you previously, IN GREAT DEBATES, no less: Why should gun legislation that reduces firearm homicide reduce non-firearm homicide as well? It doesn’t make any sense. So, I see your effort to avoid looking directly at the firearms homicide rate as a distraction. You very desperately want to avoid acknowledging that firearms are used to kill people, and that reducing the number of firearms reduces the number of people who are killed by firearms.
I do understand that gun nuts would like to imagine that all killers are just plain killers and if they did not have a gun they would just use some other means to kill. So you are casting about trying to make evidence appear that says that people would just substitute means. But that kind of thinking is just categorical, and stupid. Some times people kill deliberately and planfully. A lot of times people kill impulsively, and guns make that much more likely to happen.
Only because you don’t understand what he says. David Hemenway is a researcher at Harvard. He is much better suited to conduct analyses and review other people’s research than Dianne Feinstein. The fact that you think otherwise is extraordinarily revealing.
So what? This is irrelevant.
This is also extraordinarily revealing. You are explicitly saying “I don’t understand what he is saying but I disagree with it anyway!” Are you able to see how extremely biased you are when you read back your own words? It’s stunning. Remarkable, really.
No. You didn’t understand his critique at all. He did fault the authors for selecting a particular year without justifying their selection. However, he clearly explains what the common practice is that avoids the problem he is talking about.
See, it’s pretty simple.
Now, you don’t like the conclusion and you don’t understand what is being said, so you just disagree on an appeal to authority and an appeal to false equivalence. Exactly like climate change deniers do. Like this:
No, this is how you think. How other people think is that they evaluate the evidence and explain why they think something. In this case, it matters a great deal if the authors’ model is so far from reality that supernatural events are possible AND that even if those supernatural events happened, they would still conclude there was no effect of the variable in question.
Because they understand the issues. You, as you acknowledge, do not. Wow - it’s just fucking stunning.
Please highlight the portion that reveals their emotional state to be “aggravated”. They are not aggravated that there were too few firearms deaths - that would be horribly callous. They in fact say that it’s a fortunate thing!
If you have integrity, you would quote their full discussion of the issue - the discussion I quoted for everyone to see, above.
No - you in fact did put the word “aggravated.” Do you now deny this as well? Sheesh.
Tee hee! We did this one already. I’ll give you a second to think about the problems with this study for this purpose. You don’t even need to see beyond the abstract!
Here’s a hint: When did the gun buy back end in Australia?
No, I don’t disagree - I don’t hold an opinion. As you should not, based on that.
It is absolutely not rational to say “We don’t have any good evidence - so let’s just use all this bad evidence instead!”. We don’t have any good evidence of there being life on other planets - so should we take the personal experience of those who claim to have been abducted? We don’t have any good evidence as to what cures cancer - let’s go buy those energy jewels that guy down the street promises will help! As tempting as it is to want an answer, it’s not worth getting an inaccurate one just so you can say that you have one. You do yourself a disservice.
As far as TV talking heads go; do you believe that talking heads are selected purely on the basis of knowledge on the subject they are talking about? That they’re chosen because they’re experts on the subject, and for no other reason? You don’t need to jump straight to a “conspiracy” to say that it isn’t necessarily so that talking heads are a fantastic sample of the population at large. Beyond that, I don’t know what TV you watch. Beyond* that,* we could start talking about the hypothetical effects of confirmation bias and the like.
One should not necessarily reduce the other, but it is an important factor to consider when evaluating legislation.
Consider the hypothetical - a law is passed that banned and magically confiscated all firearms in the US. After many many years, and controlling for all variables, everyone agrees and determines that pre ban we had X level of intentional firearm death (homicide and suicide), and X + Y level of intentional death (all methods). After the ban, intentional firearm death is now zero, but intentional death (all methods) is still at X + Y. In other words, perfect 100% substitution and no reduction in intentional deaths.
Would you consider that a successful law? If you do not, then it should make sense why it’s important to look at non-firearm homicide. If you consider the hypothetical successful, then looking at non-firearm homicide wouldn’t make sense.
Here’s what I think - if your goal is to reduce prevalence of firearms, then the hypothetical is successful. If your goal is to reduce intentional death, it’s a complete failure, at the cost of liberty.
It’s a fatal flaw. I mean, if you want to measure what happened after you did something, you kind of actually have to measure what happened AFTER you did something, right? It’s more than a limitation.
Well, it’s really just because that verbiage was plain wrong. Specifically, they did identify a relationship between the gun ban and suicide by firearms in that study. They also described specifically why substitution didn’t make any sense. Sure, they couldn’t completely rule it out, but as they said the data suggest that people would really have had to wait about six years before they began substituting.
I quoted this discussion above. It’s right there.
As to the piece about firearms, they found that the size of the effect of the gun legislation on firearms homicide was about the same as it was on suicides - they just were not able to demonstrate statistical significance given the smallish number of incidents.
As to your other point, I think it’s very relevant to look at why other countries don’t have the problem with guns that we have, and particularly to look at what happens when other countries change what they are doing vis a vis guns. Examining between-country differences has value, but is limited by the broad number of factors that make countries different. Examining within country differences over time seems quite valuable.
It’s a fatal flaw. I mean, if you want to measure what happened after you did something, you kind of actually have to measure what happened AFTER you did something, right? It’s more than a limitation.