Oh, no. I think we have another reason to be afraid–very afraid–of homeopathy/ Apparently New Agers have developed a homeopathic bomb:
The world has been placed on a heightened security alert following reports that New Age terrorists have harnessed the power of homeopathy for evil. ‘Homeopathic weapons represent a major threat to world peace,’ said President Barack Obama, ‘they might not cause any actual damage but the placebo effect could be quite devastating.’
The H2O-bomb has been developed by the radical New Age group, The Axis of Aquarius. In a taped message to the world, their leader, Professor Hubert Pennington, said: ‘For too long the New Age movement has been dismissed as a bunch of beardy weirdy cranks and charlatans. But now we have weapons-grade homeopathy and we demand to be taken seriously.’
Homeopathic bombs are comprised of 99.9% water but contain the merest trace element of explosive. The solution is then repeatedly diluted so as to leave only the memory of the explosive in the water molecules. According to the laws of homeopathy, the more that the water is diluted, the more powerful the bomb becomes.
I wonder why no one (well, besides Charlene Werner) has ever thought of this before. A homeopathic bomb could render nuclear weapons obsolete. Heck, it could even be like the doomsday bomb in Beneath the Planet of the Apes or Dr. Strangelove, or: How I Learned to Stop Worrying and Love the Bomb. There is that little problem of succussion to overcome, though.
67 replies on “Fear the homeopathic bomb”
Given that “like cures like,” shouldn’t the “merest trace” be of something that PREVENTS explosions? These terrorists may have just invented the homeopathic fire extinguisher by mistake.
The humor is strong with this one…
ftw – âHomeopathic weapons are the ultimate Smart Bombs,â warned President Obama, âThey are so smart that they only affect the gullible. The only defence is for everyone to remain calm, vigilant and to always wear a magic vibrating crystal.â
Wow, thanks for the heads-up on this. Homeopathic weaponry never occurred to me before.
Must now quake in fear.
Just consider the ocean: By dropping the merest hint of a homeopathic poison into the Atlantic, it is eventually diluted to the point where the entire planet’s water supply becomes lethal! It’s the ultimate of chemical doomsday weapons!
But April 1 was more than 3 weeks ago.
It was long thought that a homeopathic bomb would be prohibitively expensive to produce due to the difficulty of “succussion sensitivity” – as the explosives become more diluted they become not only more powerful, but more sensitive. Many home-brew attempts at homeopathic explosives have ended poorly, although, obviously, antiterrorism forces are deeply concerned that a homeopathic suicide bomber might lock himself in an airplane’s bathroom and dilute and succuss a tiny amount of conventional explosive into a lethal device. If you notice carefully, the water taps on most airplanes only trickle out a little bit of water in an attempt to reduce this threat.
Meanwhile, the Iranian homeopathic bomb research continues – though they claim it’s “for civilian purposes” – rooms-ful of “succussatrons” are employed to dilute enriched uranium, to produce massively powerful quantities of water that has the memory of being fissionable. It is truly a threat to civilization.
What concerns me, however, is the more practical danger that a b anthracis might somehow get shaken (not stirred) into a large enough dilution of water and every mammal on earth would die horribly. This could happen quite by accident – all it takes is a farmer with dirty hands, washing them in a lake or something, and we’re all doomed.
Does no one see the peaceful potential in all this?
I have diluted a grain of rice seven hundred million times. I am now ready to end world hunger.
Dave I think you misunderstand, things do the opposite when diluted. For example caffeine which is a stimulant, when heavily diluted becomes a sleeping pill. So your dilution of rice would only make people more hungry.
The real danger here is if the homeopaths get there hands on an antibiotic as Arancaytar pointed out an antibiotic Heavily diluted would become the ultimate biological weapon. Why the hell is everyone so concerned about Iran getting nukes? what we really need to prevent is third world countries getting their hands on antibiotics and cures for contagious diseases. After all with nothing but a vial of water and some shaking they could kill millions!
A much bigger threat is the consumption of the world’s supply of water by the construction of these homoeopathetic things (be it âmedicinesâ, bombs, whateverâ¦). There’s only a finite amount of H2O available, yet more and more of it is being used by homoeopathetics. Each homoeopathetic substance needs lots and lots of H2O to make the active ingredients more powerful, sensitive, and undetectable. This vastly diminishes the world’s supply of fresh water.
I so want to forward all this to my New Age-y friends, but I do still want some friends (they all know not to bother me with the not-scientifically-proven). I’m from one of the towns on the west coast “known” to be a New Age “vortex” of…..oh, I forget what the vortex thing is all about, but that’s why I have so many new age-y friends.
Just wait ’till I take over. I will commandeer all media and announce my fiendish plan. My mad scientist, Dr. Orac, has created homeopathic bombs. When I push the button, I will set them off all over the country. When I press the button, all the woo-believers will disappear. Call it the woo-rapture.
Do they also want to be paid the Sum of One Million Dollars ?
Fun questions to ask Homeopathic “Doctors:”
What do you do if a disease, and its symptoms are idiopathic, with no known triggers?
Sulfur Dioxide cause severe respiratory problems, and, when dissolved in water, forms the compound Sulfuric Acid. Would you have a man who has Asthma risk horrible burns?
Why bother creating two units of dissolution, if they are, merely, separated by one digit?
What do you give a patient who is allergic to water?
(This question should be directed towards guys who claim that ancient folk redemies are better than modern medicines.) Why, if those treatments were so great, did the ancients not develop medications that allowed them to live thousands of years?
Rorschack: Ozymadius, we have come to stop your evil plans!
Ozymandius: I’m not a B-movie villian, I launched my plans 35 minutes ago: I teleported a genetically-engineered telepathic squid into the right position to explode; its mind will accelerate the explosion, wiping out the population of New York, and presenting a threat that the armies will focus on, rather than each other! I know the explosion will work, because it will be diluted in 2X6C of water!
(Later, after the world has been bombed to shit.)
Ozymandius: What the fuck happened?
Dr. Manhattan: All of the explosion was absorbed by the Atlantic Ocean, and no one died, you dumbass!
Actually, a homeopathic balm might do some people some good.
The real danger here is if the homeopaths get there hands on an antibiotic as Arancaytar pointed out an antibiotic Heavily diluted would become the ultimate biological weapon.
There’s a homeopathic level grain of truth in this: inadequate antibiotic treatment does lead to increased bacterial resistance, so low doses of antibiotics could, in principle, be used to develop bioweapons. But not true homeopathic doses.
LOL, this is funny. Although of course it is true that a placebo effect could cause a lot of problems already, given all the fear it would elicit. I’ve sometimes wondered why terrorists still carry out actual attacks rather than just making it appear like they’re planning them for this reason.
They’re already doing this. Remember ‘homeopathic plutonium’?
That, to me, is the Woo to End All Woos. I imagine some homeo nutter in the UK actually stayed awake in class the day the science master mentioned plutonium, then decided to make some money off the concept years later: “Hey, this plutonium’s supposed to be really strong stuff, why don’t I stick a plutonium label on these old Bach bottles and bring in the coin that way?”
Ah-oh…this new study is another to off-set the apple cart…either homeopathic doses (beyond Avogadro’s # work or mice are psychic? Which one is it?
INTRODUCTION: This study was designed to investigate the putative anxiolytic-like activity of ultra-low doses of Gelsemium sempervirens (G. sempervirens), produced according to the homeopathic pharmacopeia. METHODS: Five different centesimal (C) dilutions of G. sempervirens (4C, 5C, 7C, 9C and 30C), the drug buspirone (5 mg/kg) and solvent vehicle were delivered intraperitoneally to groups of ICR-CD1 mice over a period of 9 days. The behavioral effects were assessed in the open-field (OF) and light-dark (LD) tests in blind and randomized fashion. RESULTS: Most G. sempervirens dilutions did not affect the total distance traveled in the OF (only the 5C had an almost significant stimulatory effect on this parameter), indicating that the medicine caused no sedation effects or unspecific changes in locomotor activity. In the same test, buspirone induced a slight but statistically significant decrease in locomotion. G. sempervirens showed little stimulatory activity on the time spent and distance traveled in the central zone of the OF, but this effect was not statistically significant. In the LD test, G. sempervirens increased the % time spent in the light compartment, an indicator of anxiolytic-like activity, with a statistically significant effect using the 5C, 9C and 30C dilutions. These effects were comparable to those of buspirone. The number of transitions between the compartments of the LD test markedly increased with G. sempervirens 5C, 9C and 30C dilutions. CONCLUSION: The overall pattern of results provides evidence that G. sempervirens acts on the emotional reactivity of mice, and that its anxiolytic-like effects are apparent, with a non-linear relationship, even at high dilutions.
Wizard – you have posted an abstract – care to read the paper THEN comment?
Or the the study is so flawed as to be meaningless
Or the study was fraudulent.
Or the result was a fluke –
A p value of 0.05 is normally considered statistically significant in this sort of study, if you do 10 studies of a nonexistent phenomenon there is a 1-0.95^2 = 40% chance of getting a false positive. For 20 studies this probability rises to reaches 64%. Due to the file drawer effect we never see all the negative studies. Also note worthy is the fact that no significant effect was seen on the open field test with homeopathy. If the homeopathic remedy actually was effective, positive results should have been observed in both tests.
Since for homeopathy to work most of what we know about physics and chemistry must be wrong, considerable replication and results that are significant with much smaller p values are required.
Yeah, that’s the ticket, it’s fraudulent. Let’s all say that. Get out the electron microscope and find something/anything that might be wrong with it. If nothing else, let’s blame the researchers.
@23 – Reading comprehension fail – that was only one of the possibilities I presented.
That study is open access:
The journal, Psychopharmacology, must be infiltrated.
The results are non-linear, and in my world, lines rule.
The simple fact is that science is a process of devising hypotheses and finding ways to collect data which will rule out as many incorrect hypotheses as possible.
If you think that the only hypotheses which are consistent with the collected data are “homeopathic doses (beyond Avogadro’s #) work” and “mice are psychic”, that is a symptom of, at minimum, a profound paucity of imagination.
But more likely it derives from a deeply held desire to believe that homeopathy works, which blinds you to the existence of hypotheses much simpler and more likely than the improbable pair you wish to present as a false dilemma. This same desire is also almost certainly deeply held by the researchers who conducted the study (who else, in this day and age, would bother testing such a hypothesis??) and this lends credence to some of the alternative hypotheses that you don’t wish to admit exist and must be considered before “homeopathic doses … work”.
Just imagine the money science labs, just for example biochem and biomedical labs, would save (on enzymes alone)if we switched to using homeopathic quantities and dilutions for all those assays and other experiments.
Also consider all the effort this will save us when writing up our experimental results! Why is nobody doing this?
I wonder…do those attempting to do “research” on homeopathic cures use homeopathic quantities of all reagents in their experimental methods?
Define non-linear and linear (in your own words) and explain what these terms have to do with the study in question.
The suggestion to read the paper is a good one. Maybe I just lack the required training, but to me it seems quite useless.
The differences between controls can be as high as ~30% (14 to 19.5 in Table 4 Permanence time in lit area of lightâdark test (% of total)). So, to reach any kind of conclusion, the results should be pretty dam noticeable.
But they’re not. Sometimes, the homeopathic remedy seems to act as a stimulant, sometimes as a sedative, depending on which which group you look at.
For example, in the aforementioned Table.4, for experimental group 2, the mean of the controls is 21 vs. 24.8 for Gels 4 (4C dilution). For group 5 its 23 vs 19.7. What should that tell me?
And when comparing the dilutions, there doesn’t seem to be any kind of pattern. It just looks totally random as to which will be more effective. In Table 2 Permanence time in the center area of open field (% of total time), group 1 and 2 have Gels 7 being most effective, group 3 has Gels 1 tied with Gels 9, group 4 has Gels 5, group 5 Gels 9 and group 6 has Gels 30. Totally the opposite of a dose response, if you ask me. And if they had reported the mean of the controls as, say, Gels 11, no-one would have noticed.
Besides, the entire methodology seems suspect, as the drug itself, Busiprone, didn’t seem to have a consistent effect either (please correct me if I’m wrong).
So yea, GIGO.
Not having read through the entire paper Wizard, the first things that leaps out at me is:
From Pharmacology (Rang, Dale et.al)
The main disadvantage of buspirone is that it may take several weeks before its anxiolytic effects become noticeable. Many patients may also require a higher dosage to adequately respond to treatment, which may also be increased in slow increments of 5 mg every three days and up to 60 mg daily, which may be the dose required for adequate relief.
So an anxiolytic drug which takes weeks for its effects to kick in…was used as a positive control in 9 day study?
Do you see a problem with that protocol Wizard? Or am I missing something by having not yet read the entire paper?
Disclaimer: I am rather rusty (or perhaps corroded is the right term) on my pharmacology, so I would appreciate anyone pointing to any error of thought above.
Hehehe…your post wasn’t up when I posted and I really didn’t feel like reading through all of the paper.
That’s right. Drugs always have the exact same response no matter what and dose-response is always precisely linear.
Even though none of us seems to be knowledgeable about pharmacology and we have look things up at wikipedia, let’s condemn this blind trial with anything possibly wrong. Anyway, mice are psychic and they like to confuse researchers.
This could actually be a very intriguing idea.
Drop ‘homeopathic bombs’ on areas composed of alt-med idiots and see what their psychosomatic reaction to it is.
About vortices: new age woo about “subtle” or “spiritual” energies spiraling and funneling to a point of concentration on the earth which we are told facillitates arcane practices.Think of homeopathic “action” at a vortex!(BTW,they call them “vortexes” at Sedona, where even the trees get twisted).Like a whirpool, or water going down a drain or toilet.Like Mother Earth having esoterically relevant cramps. Actually,it does funnel new agey tourists to particularly scenic out-of-the-way small towns.Supposedly one exists about 40 miles north of NYC on the Hudson-“Nyack”(two syllables,please)- which has attracted UFO activity(Hah!),yoga teachers,energy healers,woo-meisters,antique dealers,and trendy restaurants.It’s another instance of people ascribing other-worldly significance to purely psychological phenomena (people enjoy viewing dramatic natural scenery and buying things).(I wouldn’t be surprised if it were claimed that vortices exist in Boulder,CO,Woodstock,NY,and Sebastopol,CA)
Can’t wait to see how JB and co spin this.
DB says “Dave I think you misunderstand, things do the opposite when diluted.”
So if things work the opposite in dilution, then if we dilute a homeopath sufficiently and introduce him into the water supply, we won’t need to be skeptics.
What are you talking about? Please explain.
Even though none of us seems to be knowledgeable about pharmacology and we have look things up at wikipedia
I put the wiki info in because it’s easily accessible and reiterates the information on Buspirone given elsewhere. (for example the text Pharmacology referenced)
If you don’t like/trust the wiki article Wizard, there are some references to journal articles at the bottom of the wiki, which the information given there was gleaned from. Go to it 🙂
Also I have studied pharmacology in both second and third year at uni(bichem/molec bio major),though it’s been 2yrs. Currently am seriously considering doing a post grad in pharmacology (or maybe biotechnology).
How about you?
btw my last post was @Wizard #30
“Even though none of us seems to be knowledgeable about pharmacology and we have look things up at wikipedia”
should have been in blockquote
We must put beta-scintillation counters on every street corner! John Benneth gave us the clue we need to catch these homeoterrorists. Homeopathic is radioactive!
I can only wonder what Wizard is trying to suggest with this clumsy sarcasm that we should do. Is he trying to suggest that real scientists would simply take at face value what the researchers think the correct interpretation of their data is? Does he really think that real scientists would determinedly avert their eyes to anything possibly wrong with the trial, as seems to be his approach?
What was I thinking. We all are much smarter and more informed about pharmacology than any of those pharmacology professors who serve on the editorial board of that peer review journal.
And I bet that we can make-up even more silly critique of this other study on the same homeopathic drug, Gelsemium:
Let’s bring out the big guns. Orac, where are you now that we need ya? Take out your electron microscope and let’s find little things wrong, and let’s pretend that this study is meaningless.
Bombs away. Balms away. Balms a way.
Mice? Time spent in light?
Homeopaths claim to cure real diseases (like syphilis) in real humans… and you present a study of mice and how much time they like to be in lighted areas?
Whoa, dude! I also like light! I try to spend as much time as I can in lighted areas. I actually like sunny days. Due to my myopia I prefer to look at things when there is more light. Wow! All of this without homeopathy.
Oh, they also over complicated the test by not figuring out what they were testing for. It is a bunch of bullocks.
I can think of two good reasons to study homeopathy using mice as a model:
1. Mouse studies are cheaper than human studies.
2. Mice are probably less likely to have a placebo effect than humans, so if they can show this remedy works in mice, that’s something they can wave at us skeptics to say “nuh uh, homeopathy is real, not a placebo.”
Unfortunately for them, skeptics know more about effective study design and proper data analysis, as well as how to see if the conclusions can be drawn from the data. We will not be convinced by poorly designed and analyzed studies. The high variation within groups, combined with conflicting effects at different doses, suggests that all we’re seeing is noise from a problematic experiment. (I tend to see this when my experiments aren’t working right. When I clean up the protocol, suddenly the error bars get smaller!)
Whatever the other flaws of this study (and there seem to be many), the basic idea that you can test the efficacy of a drug against anxiety by monitoring how much time the mice spend in lighted areas is absolutely standard in psychopharmacology. You just have to look at it from the mouse’s point of view.
To a mouse, being out in the light means being vulnerable to predators. Therefore, an anxious mouse will balance the risk of getting eaten vs. the benefit of possibly finding food or a mate, and decide to spend more time where it feels safe. A mouse that is not particularly anxious will have lowered inhibitions against going out in the open. (I also suspect that some drugs just make them disoriented enough they don’t pay attention to where they are, but I don’t do these studies. In humans, that would be the difference between relaxed yet alert enough to drive, and being high enough to wander into traffic without realizing it.)
However… I’d like to see why the IRB at that institution approved an animal study of a “drug” with 0% probability of efficacy based on our knowledge of chemistry and physics. I support responsible animal testing, but wasting the lives of mice on this dreck is a shame.
Rather than just making inane comments and trying to change the subject to another study, why don’t you first show us how our “silly critique” is…well silly and flawed? This is what I ask of you in my first response. You’ve avoided answering anything put to you…and rather badly I might add.
Soooo? Let’s have your specific and detailed criticisms of those points so far raised. You could show us all up.
ps You’re attempts at witty reply are merely coming across as childishness and willful ignorance. Just my opinion.
Nobody will try a homeopathic bomb. They are all afraid of the Duchy of Grand Fenwick.
Just a thought, but if one is studying whether a substance has an effect, any effect, upon the body, wouldn’t the best idea be to find the levels at which it is fatal? If a substance has any effect, at high enough doses it will prove deadly. Hell, you can die from too much water. So as long as the fatal does of a Homeopathic solution is smaller than that of water, that would prove it has an effect, right?
I wonder why no one is doing it that way…
Yeah, because everybody knows that passing peer review is the last step in the scientific process and proof positive that all the conclusions of the paper are irrefutable. Everyone knows there’s never been a paper that passed peer review and then turned out to be poorly designed, based on wishfully interpreted data, or fraudulent. (Two can play the sarcasm game.)
Thanx Chris for providing good evidence of what I meant by “silly” (and ignorant, while seeming to be smart).
And ToxicologyKat, how or why does homeopathic research often disprove that 0% possibility of biological activity or clinical efficacy? Why oh why?
And are you really really saying that there should always be a linear effect from a dose…or that each animal should have a precisely predictable effect from that dose? Really?
They didn’t control for multiple comparisons. Checking many different endpoints, seeing that some of them produce a statistically significant effect when considered on their own, and then hanging one’s hat on those is entirely illegitimate.
Without properly controlling for multiple comparisons, there is no justification to conclude that the observed effects actually were statistically significant.
Look in particular at tables 1 and 2. Five different comparisons between control and “treatment.” Only two give a result greater than one standard deviation.
Then let’s compare the 5 CH vs. 30 CH. Three measures, differing by 1.5, 1.4, and 0.08 standard deviations. And that’s considering the fact that they declined to show the other two tests because they didn’t produce a result distinguishable from control!
The statistics simply do not support the conclusions.
I saw a variant of this back in ’07:
How to be a homeopathic bioterrorist:
According to homeopathy, diluting a substance makes it more potent. While traditional homeopathy creates medicine by diluting harmful subtances, we can apply the same principles to weaponize healthy substances. Since orange juice has lots of vitamin C, a homeopathic dilution of orange juice would induce a crippling vitamin C deficiency in anyone who drank it.
1. Buy a carton of orange juice and 30 1-gallon jugs of water.
2. Place one drop of orange juice into one of the jugs of water. Shake.
3. Take one drop of that dilution and place it into the next jug of water. Shake.
4. Take one drop of that dilution and place it into the next jug of water. Shake.
5. Repeat the process until you reach the last jug of water.
6. Take a drop of that final dilution and place it into your municipalityâs water supply.
7. Everyone gets scurvy!
Orac “Respectful Insolence” Blog
“Fear the Homeopathic Bomb”
Scott, I love it…instead of having multiple endpoints, I assume that you want things to be more simple with just one endpoint because you can then criticize for that. The ole damned-if-you-do and damned-if-you-don’t attack.
You’re wonderfully creative in not seeing differences btw the control and treated groups.
The “wizard” paper is fascinating, ignoring the water effects they proclaim to see you have clear trends of the mice being not all that interested in the experiments (the small numbers showing for all “exp #1”), then getting into it for a while, peaking around exp #4-5, with a drop-off afterward in #6. Makes you wonder how they accounted for that, maybe one should rerun the “data” without the outlier #1 and see if it would still make the significance threshold.
I can’t understand so how Psychopharmacology can accept a paper with zero dose-response dependency (to term it nicely) without more supporting data.
Yeah…why is it that so many respected medical and scientific journals have gone over to the other side? Why is it that they publish research that supports homeopathic medicine? Don’t they know that the results from randomized double-blind and placebo controlled trials are only worthy if they disprove homeopathy?
I mean, look at the extremely shoddy meta-analysis published by the Lancet in 2005 (Shang, et al). If that review was positive of homeopathy, the people here would have condemned it up-and-down-and-sideways…but when it is negative, no one here speaks a harsh word of it.
The best news is that you can then cloak yourselves in the garbs of science and pretend that you are right.
Thanks for proving that you have no conception of proper statistics or science. Having multiple endpoints is fine, but ONLY if you correct for the fact that you have done so! If you have 20 endpoints, for instance, then it is expected that one of them will be 95% significant by pure chance. You might start addressing the deficiencies in your education here:
This, you’ve simply made up. The most cursory review of my post demonstrates your claim here to be an outright lie.
Yeah…why is it that so many respected medical and scientific journals have gone over to the other side? Why is it that they publish research that supports homeopathic medicine?
Ah, so “respected” = either homeopathic or pay-to-publish?
Don’t get all succussed about this, now.
Wizard @53: any actual criticisms of the Lancet study, or are you just going to call it names? Any worthwhile comments at all? Oh, wait, I get it! These are HOMEOPATHIC arguments! The weaker they are, the STRONGER they are!
Scott has the laziest/most simplistic understanding of statistics I’ve ever read. Are you actually saying that no matter what you are testing, there will always be a 5% chance of finding significance? Even when you are comparing placeboes, conventional drugs, and homeopathic medicines? Are you really saying that?
As for the 2005 Lancet piece, is anyone out there really suggesting that it has no serious flaws? For instance, does anyone out there believe that the researchers could not find a “matched” trial for people with arthritis in multiple joints? Does anyone really believe that such a trial has never been conducted?
Like I said earlier, why is it that you all take out the electronic microscope to criticize any positive study on homeopathy but never/ever criticize any study that finds a negative result?
Extraordinary claims require extraordinary evidence. Surely you can’t be completely ignorant of that principle?
Studies which indicate that substances diluted in water until the chances are vanishingly small that even a single molecule of the substance remains act just like water — those are not studies making extraordinary claims. Studies which indicate that substances so radically diluted actually become medicines of amazing potency — those are studies making extraordinary claims, and extraordinary claims require extraordinary evidence. That isn’t “evidence that looks extraordinary upon casual examination”, by the way. No, extraordinary evidence is evidence that stands up under extraordinary examination.
You should be happy that everyone’s “tak[ing] out the electron microscope” to examine these results. That is, if you thought the results in question were actually due to homeopathic effects, and could withstand the sort of scrutiny that would be necessary in order to establish a result so contrary to the rest of known science. I wonder why it is you’re instead saying, in essence, “Look!!! LOOK!!! Just look at how these researchers have proven homeopathy to exist!!! But whatever you do, don’t look too closely!!!!“
Wizard is obviously trained in those alternative homeopathic statistics.
Less significance therefore more significant!
You are seriously suffering from an ailment known as tunnel vision. Unfortunately the light you’re seeing at the end of the tunnel is in fact an oncoming freight train.
Thanks for proving that you have no conception of proper statistics or science. Having multiple endpoints is fine, but ONLY if you correct for the fact that you have done so! If you have 20 endpoints, for instance, then it is expected that one of them will be 95% significant by pure chance.
Scott has the laziest/most simplistic understanding of statistics I’ve ever read. Are you actually saying that no matter what you are testing, there will always be a 5% chance of finding significance? Even when you are comparing placeboes, conventional drugs, and homeopathic medicines? Are you really saying that?
q. e. d.
XD Thank you Wizard.
Odd that you should mention comparing placebos in a study of homeopathy. Is that your meaning, Gwen?
For the same reason that an objective party is unlikely to closely examine a study of any unproven treatment that does not demonstrate efficacy.
Unless there is a good reason to believe that the failed study is flawed, it is likely to be ignored. Many are not even published. This is nothing unique to homeopathy.
Of course, the homeopaths have a lot of money at stake, so they will keep trying to come up with a study that works. If homeopathy does not work, and there is no reason to believe it does, then there is nothing for the homeopaths to do but try to establish themselves in some other scam(s). They don’t want to follow that alternative route.
If some other treatment failed as many experiments as homeopathy, but kept on testing, then suddenly produced positive results, there would be similar close scrutiny.
That is the way science works.
I won’t try to explain wizardry to you. Don’t try to explain science to me.
Science is the intelligent alternative to wizardry.
Wizard has apparently decided that since sarcasm is not the “win-every-argument” strategy he thought it would be, he’ll try straw men as his go-to strategy instead. I would say that Wizard has achieved Goofus* status at this point, but for the benefit of others, I’ll try to explain in other words the point Scott is making about multiple endpoints.
As mentioned before, science is not just about devising hypotheses, it’s also finding ways to collect data that will allow us to rule out incorrect hypotheses. Sure, you might want to test the hypothesis “Compound X has Effect Y on human test subjects,” and you might try to test this hypothesis by putting out a call for test subjects willing to take Compound X experimentally, and charting the frequency and intensity with which you see Effect Y in that group, as compared to a “control group” of subjects who don’t get Compound X. And hey, if you see a lot more Effect Y in the group that got Compound X, isn’t that consistent with your hypothesis that Compound X causes Effect Y?
Sure it is. Problem is, it’s also consistent with these other hypotheses:
* The experimenters are influenced by knowing who’s in the Compound X group, and are more likely to judge members of that group to have Effect Y than members of the control group;
* The people who know they’ve gotten Compound X either expect to experience Effect Y and are more prone to report that they’re experiencing Effect Y, or because they know they’ve gotten the Compound X, act in different ways that are what bring about the Effect Y;
* The kind of people who would volunteer in the first place to be test subjects for Compound X are different in subtle but significant ways from the general population (more adventurous, for example) and these differences are what lead to a greater incidence/intensity of Effect Y;
* dozens of others.
That’s one of the reasons scientists like “double-blind” testing; it helps rule out hypotheses similar to the above. If the experimenters don’t know who’s in the Compound X group, they can’t judge those who got the compound differently from those who didn’t get it; if none of the subjects themselves have the knowledge of whether they got the compound or a placebo, they will not be behaving differently because of that knowledge; and so on.
Unfortunately, there’s a hypothesis that can never, technically speaking, be ruled out. That is the hypothesis of pure chance. Take gravity, for example. Suppose that whether objects move towards or away from the sources of greatest mass around them is not determined according to one consistent universal principle, but is in fact determined by the flip of a cosmic coin somewhere in the vicinity of Alpha Centauri. And the only reason we think objects always move towards the greater mass is because, so far, the coin has just coincidentally come up heads for all Earth-related flips.
Of course, no one really takes such an idea (can’t even be dignified with ‘hypothesis’) seriously. But the question remains: when do you have to consider the possibility that the results you got, you got only because of chance? The answer is complicated but the ultra-simplified version is “when the probability that the results you see came from chance is less than 5%, then chance as a hypothesis can probably be ruled out.” That’s a somewhat arbitrary threshold; there’s nothing magic about 5% as opposed to 4% or 6% and in fact scientists recognize this, but 5% is pretty much the value that everyone acknowledges as the borderline.
Of course, the issue doesn’t end there, because of a key fact about probability that most people don’t appreciate: the most “improbable” event, with sufficient trials, becomes certain. Winning a gambling game with odds of one-in-a-million on one try is highly improbable. Winning the same game with one million tries is not improbable in the least; on the contrary, the chances are better than three out of five.
And this comes back to scientific studies. Attempts to test scientific hypotheses are not gambling games, of course. But each one nevertheless is a gamble! Why? Because each one runs the risk that an illusory effect, caused by random chance, will look like a real effect, derived from some causal relationship (either the one predicted by the hypothesis or some other.) Gamble enough times, and you will lose.
… except that for dishonest people, getting a false positive, a result that looks like it supports their hypothesis when it actually doesn’t, isn’t a “loss”. They care about appearances, not truth, so they’re delighted by this effect. In fact, what they love to do is to actually do multiple comparisons within one study; that way they can make it look like their record is one-for-one (one positive looking result out of one study) when it’s actually one-for-many (one positive looking result out of a substantial number of factors being compared.)
I have no idea why any sane person would think that the mathematical factors just discussed are changed in the least if the subject matter is ‘placeboes, conventional drugs, and homeopathic medicines.’
* One whose only contribution to a conversation of reasonable adults is as a negative role model.
In part. It’s kind of hard for me to explain without being imprecise (English is my second language) but what I was mainly aiming at is that by definition Î± = 0,05 means that your chance of rejecting the null hypothesis wrongly (i.e. you see an effect even though there isn’t one) is 5 %, so Scott is correct and the answer to Wizards question is yes. You’d expect to have a 1 out of 20 chance to find a statistically significant effect even if there isn’t one, e.g. when comparing placebo with placebo (or a drug with itself). Another point was that instead of telling Scott where he thinks Scott is wrong, Wizard is saying “Duh, your view is so simplistic” (I’m paraphrasing) – not really the way to convince others. If he knew what he was talking about, Wizard should have shown it, right now it looks like his “answer” confirms Scott’s opinion of him.
I still don’t know what a bomb’s sexual preference has to do with its lethality.