3 ms·

The only way to get around these obstacles would be to feed one group of people doses of lead and compare them to another in a controlled setting, or isol

• lutusp 12y ago
  > The exact same is true of this topic. No, this is false. Try to imagine a conclusive falsification of the claim that's being made in the linked article. Falsification doesn't mean casting doubt on a hypothesis, it means proving it wrong. In fields such as the one under discussion, with p-values typically in the range 0.01 to 0.05, nothing is ever falsified. Some ideas are abandoned through embarrassment, but none of them are ever falsified. If I say all swans are white, someone can falsify my claim by locating a black swan. If i say that a reduction in the crime rate might have a cause-effect relationship with a reduction in the lead content of gasoline, that's a perfect article for social psychology, because it's not possible to argue against it in a scientific sense. > Please refrain from the ad-hominem attacks. My posts aren't being downvoted because of their tone, they're being downvoted because my position is correct and it makes people uncomfortable to hear that so much of modern science ... isn't science. Apropos, here's Richard Feynman making the same points, for the same reason, 40 years ago: http://neurotheory.columbia.edu/~ken/cargo_cult.html http://neurotheory.columbia.edu/~ken/cargo_cult.html > the only one of them (outside of the one that they're obviously promoting, drug prohibition) which held up with a consistent correlation was environmental lead. [emphasis added] I shouldn't have to say this, but correlation, however consistent, doesn't equal causation, and causation -- testable, falsifiable theories -- is the foundation on which science is built. Ask yourself why Creationism isn't taught in public school science classrooms. This is so because Creationism isn't science. That, in turn, was established in a trial -- several, actually -- in which the court's decision was based on the fact that science was shown to require falsifiability, not mere argument. Here's one such court ruling: http://en.wikipedia.org/wiki/McLean_v._Arkansas http://en.wikipedia.org/wiki/McLean_v._Arkansas Quote: "The judgment defined the essential characteristics of science as being: * It is guided by natural law; * It has to be explanatory by reference to natural law; * It is testable against the empirical world; * Its conclusions are tentative, i.e. are not necessarily the final word; and * It is falsifiable." Think about this. If your position were accurate, the Creationists would have the right to present Creationism -- argument without falsifiability -- as science in public schools.
  • lambda 12y ago
    No, this is false. Try to imagine a conclusive falsification of the claim that's being made in the linked article. Falsification doesn't mean casting doubt on a hypothesis, it means proving it wrong. In fields such as the one under discussion, with p-values typically in the range 0.01 to 0.05, nothing is ever falsified. Some ideas are abandoned through embarrassment, but none of them are ever falsified. Are you trying to say that the only work that can ever be falsified is that which is black and white, cut and dried, you will just be able to pull a counter example out an someone will go home in shame because their pet theory was disproven? That's not how the majority of science works. The majority of science is statistical, and noisy, with lots of confounding variables. That doesn't mean that it's not falsifiable; just that it's messy, it doesn't follow some neat perfect model of how science is supposed to work. This topic, whether environmental lead leads to increased crime, absolutely is falsifiable, just as falsifiable as the laws of thermodynamics, which are likewise statistical in nature. Now, the laws of thermodynamics are a lot easier to test, and you can get much bigger sample sizes much more easily; but that doesn't have any bearing on whether a theory is falsifiable, just how easy the process is. Lack of falsifiability means that a theory makes no predictions. It means that there is no way to distinguish a world in which the theory is true, from one in which it is not. If i say that a reduction in the crime rate might have a cause-effect relationship with a reduction in the lead content of gasoline, that's a perfect article for social psychology, because it's not possible to argue against it in a scientific sense. Of course it's possible. You would just need to show sufficient evidences of cases of long term, widespread (population wide for a large enough population) increases in environmental lead where the crime rate did not increase 20 years later, or likewise widespread decreases in environmental lead where it did not decrease 20 years later. If you managed to come up with enough data of that sort, without having selectively chosen the data solely for that purpose (or with some other biased methodology that caused you to prefer data that prefers one result over another), that would be a falsification. Other possible falsifications would be to test blood lead levels on a controlled population, and then observe that population's tendency towards criminal activity. If environmental lead were strongly correlated, but blood lead levels were not, then that might indicate that the causation was not ingested lead, but some other variable that was predicting both, such as industrial activity or development levels. My posts aren't being downvoted because of their tone, they're being downvoted because my position is correct and it makes people uncomfortable to hear that so much of modern science ... isn't science. There is a lot of modern science that is flawed. On the other hand, there is a lot of science from any era that is flawed. Pseudo-science has existed as long, if not longer, than science. No one is afraid to admit this. I am well aware of poor science, junk science, and pseudo science, as are many people in this thread. You have provided a lot of arguments against junk science in general, but have not provided one bit of evidence that this study is junk science. I don't downvote, I prefer to make my points explicitly in writing (unless it's just someone who is a blatant troll). You have been behaving poorly in this thread; trying to claim that I don't know what science is, rather than actually addressing my points. You have been criticizing the research by pointing out completely different fields which produce dubious results for sensation headlines (epidemiology and social psychology are quite different). You seem to be arguing from emotion, as you have some kind of prejudice, possibly well deserved, against certain fields of study; but you're now applying that against other, completely unrelated research, instead of actually engaging with the science to try to learn something about how the world works. I shouldn't have to say this, but correlation, however consistent, doesn't equal causation, and causation -- testable, falsifiable theories -- is the foundation on which science is built. Falsifiability does not imply that you have established causation. A correlation is just as falsifiable as a causal relationship; all you have to do is show sufficient evidence of the same two variables without the given correlation. Likewise, you can have completely unfalsifiable causal statements; like "God made the heavens and the earth." You are drifting further and further from actually engaging the topic at hand; all you are doing is rambling about some of your pet peeves about bad science and pseudo-science, but you haven't established that the study in question has anything to do with them. Think about this. If your position were accurate, the Creationists would have the right to present Creationism -- argument without falsifiability -- as science in public schools. I have no idea how you are getting a lack of falsifiability from this study at all, and thus no idea how you're linking it to creationism. Creationism doesn't make predictions. This study does. Those predictions can be tested. Yes, they are statistical in nature, and subject to confounding factors and measurement error, but that's the nature of science. Feynman's example of the oil drop experiment was likewise subject to similar kinds of confounding factors and measurement errors; luckily, that was one that was a lot easier to collect data on so it could be much more easily refined. Would you say that the relationship between smoking and cancer is bad science? The science there has all of the same issues; there are ethical issues with asking someone to smoke in order to determine causation, so all you can do is population wide observational studies and observe a lagged rate of cancer among a population that is strongly correlated with the smoking rate 20 years earlier, and the higher rate of cancer among those who smoke than those who don't. The combination of lagged effect, strong correlation, and explanatory science based on known mechanisms of acute toxicity can all combine to make a fairly compelling argument.