If there's one thing that the antivaccine fringe wants above all else, it's legitimacy. They crave it almost above all else. They want to be taken seriously from a scientific standpoint. Unfortunately, what they fail to realize is that to be taken seriously from a scientific standpoint you really need to demonstrate that you actually understand science. At the very least you can't be spouting pseudoscience, but that's what antivaccinationists do constantly. Virtually every argument they make trying to demonstrate that vaccines are the root of all evil (or at least cause autism, autoimmune diseases, and various neurodevelopmental problems) is chock full of bad science, pseudoscience, and anti-science.
With this as a background, I noted with some amusement a couple of posts that appeared on antivaccine crank blogs. The first comes to us courtesy of our old friend, one of the earliest promoters of the myth that mercury in vaccines in the form of the preservative thimerosal causes autism. I'm referring, of course, to Dan Olmsted, who was blowing his own horn (or at least that of the antivaccine crank blog of which he is the editor) over what he considered to be a great honor:
Age of Autism is honored to announce we have been chosen as part of Columbia Journalism School's inaugural Single Subject News Network. An initiative of Columbia's Tow Center for Digital Journalism, the network "connects news websites that focus on one subject on an in-depth level, filling the gaps in mainstream media and innovating models for journalism."On the weekend of November 9, the Tow Center "will host a series of panels amongst students, industry, and an elite class of 20 single subject news publishers selected by the program." I'll be on a panel discussing how to create community around a niche topic.
When I learned of this, I was rather surprised. There is a different between a Single Subject News organization and a single subject crank blog. Age of Autism (AoA) is without a doubt the latter. It is not a legitimate news organization. It does no investigations of its own. Rather, it produces a continuous spew of antivaccine propaganda chock full of pseudoscience, paranoid conspiracy mongering, and misinformation. A far better analogy to AoA would be a single-subject Prison Planet, except that in many ways AoA is less restrained and more full of crazy than even Alex Jones. Indeed, I wonder if there's an organization for single-subject conspiracy sites.
In any case, after seeing that announcement, I wondered if the Columbia Journalism School had lost its collective mind. After all, AoA isn't subtle. You don't have to read very much of AoA to figure out its true nature. True, every so often AoA tries to appear to be a real autism advocacy blog and will publish posts about, for example, the restraint of autistic children or autistic children wandering off and being injured or the like. However, it's never very long at all before the topic flips back to the raison d'être of AoA: Opposition to vaccines and the promotion of the scientifically discredited idea that vaccines cause autism. It's also rarely very long before the paranoid conspiracy mongering rears its ugly head. Then, if that fails, there's always the AoA "media editor" Anne Dachel sending her hordes of flying monkeys off to dive bomb with their poo mainstream news sites and blogs that have the temerity to publish articles supporting the scientific consensus that vaccines don't cause autism.
All of this makes it particularly delicious to see the addendum that Olmsted added to his post the next day:
Well, that was quick. Shortly after this article was posted, Columbia Journalism School withdrew their invitation, saying that after further review, "Age of Autism does take a clear position on the link between vaccines and the incidence of Autism, also engaging in advocacy on that position. Therefore we must disqualify the site from our study." The same person said, when inviting us in August: "I'm also a huge fan of The Age of Autism, how you've built and sustained an enriching and focused platform. It's a huge pleasure to invite you to join a community at Columbia University's Tow Center for Digital Journalism. ... What you're doing is part of a wave in the journalism world that the Tow Center wants to bring together and highlight as a trend." We warned them to expect to hear from critics but were told, "Thank you Dan for the head's up but we are happy to have you!" Que sera, sera.
At least the Columbia Journalist School realized its mistake and corrected it. This has the added bonus of driving the AoA commenters into new heights of ranting and paranoia. For example, one accused the school of uninviting AoA because Columbia University Press is the publishing company that published Paul Offit's book, as if that had anything to do with it. It's also rather amusing given how much the antivaccine brain trust at AoA has attacked the Columbia Journalism School for hewing (mostly) to the scientific consensus regarding vaccines and autism.
This little fit of pique by Olmsted, however, was just a little wafer to cleanse the palate, a tiny little thin one. Next up over the weekend was everyone's favorite "six degrees of separation" conspiracy theorist, our old "friend" Jake Crosby. Hilariously, he is now accusing Mark Blaxill of "interfering" in the Autism Omnibus. As you may recall, the Autism Omnibus proceeding was a large action brought to the Vaccine Court by the antivaccine contingent that fervently believes that vaccines cause autism. Because there were around 5,000 complainants, the Vaccine Court basically told the lawyers to present their very best cases for vaccine causation of autism. So that's what happened. A handful of complainants had their cases heard by the Special Masters of the Vaccine Court, the idea being that, if these very "best" cases could convince the court then the rest could be heard. At the time, it was a rather terrifying prospect, because if the court found for any of these complainants, it might have opened the door to thousands of such claims and potentially bankrupted the National Vaccine Injury Compensation Program. Thankfully, despite the best efforts of the complainants to introduce pseudoscience, ultimately the court found against the test cases.
Some of the pseudoscience used to support the complainants claims that vaccines caused autism and that they should be compensated on that basis came from some old familiar faces. I'm referring to that father-son tag team of execrable science done in the basement of the father's Silver Spring home, Mark and David Geier. We've met them before many times, but they are most infamous for chemically castrating autistic children using Lupron because they proposed that testosterone interferes with chelation of mercury in the brain. I've known that the Geiers were bad news when it comes to medicine and science for a long time. It's because of his utterly pseudoscientific quackery that Mark Geier ultimately lost his license to practice medicine in every state in which he had had a license.
Apparently, if Jake Crosby is to be believed, SafeMinds and the antivaccine movement knew it too, way back in 2003. I note that that's before I started blogging and before I took a serious interest in the antivaccine movement. Crosby somehow got a hold of old e-mails between Mark Blaxill and Mike Williams, who was the lead attorney for the plaintiffs in the Autism Omnibus Proceedings of the Vaccine Court, and, if what Jake says is true (and, believe me, I take anything I see written by Jake with a huge grain of salt), Blaxill showed better scientific judgment than I thought him capable of:
As to the Geiers, I may be a bit of a minority voice here, but I worry very much that they can do our cause more harm than good. They are not very good scientists, write bad papers (both writing badly and reporting in sloppy fashion) and attract too much attention to themselves as individuals. In this last regard, they don’t show nearly as well as Andy Wakefield but they’re trying to play the same role. Frankly, if I were on the other side and were asked to critique their work, I could rip it to shreds. I’m surprised they haven’t been hit harder. So I think you are wise to diversify.
In a later e-mail from February 24, 2004, Blaxill is quoted:
In the interest of full disclosure, I thought you might be interested to see my critique of the Geiers' latest work on VSD. I have not been a big fan of the Geiers. I worry they do not represent our side well. They often do sloppy work.
OK, I said "better" scientific judgment than I thought him capable of, not good scientific judgment. After all, Blaxill was obviously very impressed by Andrew Wakefield and presumably still is. Anyone who thinks Wakefield is a good scientist is a very poor judge of quality in science. He was also speaking out of both sides of his mouth, so to speak. For example, not too long ago, Blaxill was arguing that mercury in vaccines is strongly linked to vaccines while trying to explain away a study that failed to find a correlation. He also has argued frequently for the claim that thimerosal is associated with autism.
One can't help but note that the two e-mails above are dated before the emergence of the quackery that ultimately cost Mark Geier his medical license, namely the "Lupron protocol," or, as I like to call it, Why not just castrate them? Yet, even before the worst of the quackery, there was Blaxill, badmouthing them. True, they richly deserved it even then given the absolutely atrocious quality of their "science," but this was at a time nearly ten years ago when Blaxill and the Geiers were on the same team. They were arguing the same "hypothesis," namely that the mercury-containing preservative in vaccines, thimerosal, was the One True Cause of the Autism "Epidemic." As Matt Carey points out, on his own blog, AoA, Blaxill was promoting the Geiers' quackery as Mercury, Testosterone, and Autism — A Really Big Idea!
Indeed, it's hard not to conclude that Blaxill is a total hypocrite. He thought that the Geiers' work was crap, but he never said anything in public because it might undermine the push to try to link vaccines with autism. Then, even though it's likely he realized just how horrible the Geiers' "Lupron protocol" was, he said nothing and let the Geiers continue to promote the protocol. Such are the ethics of the antivaccine movement.
- Log in to post comments
"Well, that was quick."
Said, I think, with the same sad resignation of the fugitive to the arresting officer, "What took you so long?"
The enemy of my enemy is my friend, especially if we both hate vaccines!
What would be crazy enough to push these people to disavow some of those in their corner? I mean, we have AIDs denialists, castration promoters, a guy who actually made sh*t up, shadow-agencies left and right, and they'll all go hand in hand so long as everyone can agree that "VACCINES = BAD" Pretty crazy. Is there anything so crazy even the crowd at AoA will go, "whoa! Slow down...that is TOO crazy for here." I honestly don't think so, even though I doubt they're all also AIDs denialists, etc.
Well, that was quick. So, clearly they know. I don't know if they were "honored" as Dan stated, so much as "surprised."
Those guys at AoA have no idea what the difference is between journalism & advocacy - you can't be a "journalist" if you are the advocate of solely one position on an issue.....that's just common sense (which seems to be in extremely short supply).
I'm not trying to be super skeptical, but do we have any proof that AoA was invited to this inaugural SSN other than Olmstead's claim on AoA?
OT, but prompted by the mention of Prison Planet, I went over to rationalwiki to look into what they said about that site; their entry begins with this absolute gem of a sentence:
I know it's a bit OT, but does anyone know what's going on with Science-based medicine site? It's been down at leaste since this morning (which would be middle of the night in the US), isitdown claims it's not only my problem.
re "the quest for legitimacy"
Anti-vaxxers, like other woo-meisters, have a problem whenever they seek out legitimacy' i.e. recognition by authorities or sources of expertise because they have previously spent a lot of ink, airtime or electrons regalling their thralls with tales about how corrupt, mis-informed or downright evil the experts truly are. Thus consensus science, governmental agencies or the media are all in the same boat for altie critics- and it's a leaky one.
HOWEVER these pronouncements do not stop them from trying to have it BOTH ways: if an enabler has a legitmate degree or is employed at a well-known university or research facility- it is trumpetted loudly- "She is a professor at HARVARD" or " His doctorate is from CAMBRIDGE". Degrees are listed whether they originate from an accredited university or a website. Fledglings attempt to gain legitimacy by listing their degrees - even if they're in marketting or if the proselytiser is currently twisting himself into knots trying to attain one ( see Louise Kuo Habakus and Jake Crosby, respectively). Alties constantly refer to DR Wakefield, DR Hooker, DR Young and DR Null.
And PAUL Offit, STEVEN Barrett, BEN Goldacre or DAVID G.
So Dan had to first quell his rage at a legitimate school of journalism ( remember how they shrieked about the CJR's article recently?) because he wanted the good press THEN only after the rescinded invitation, was he able to vent freely.
Across woo-topia, I often observe this split: casting aspersion upon experts then declaring an exemption for an expert ( or a poseur) who agrees with you. It's all to woo an audience.
Here's an entirely new one I only ran across in scanning ( @ PRN, yesterday's 'Talkback' show) that resembles the 'secret documents' fad ( see RFK)-
the woo-meister, who was a 'research fellow' at the "Institute" and a 'board member' of the 'Hospital across the street' was approached by a cancer researcher from A FAMOUS PLACE who sought him out because of his criticism of the "cancer industry' and revealed that often he and his associates observe that a new med is not working but they allow the patient to continue anyway to the detriment of said patient's health, often leading to death. Years later, he saw the same guy presenting at an Important Conference. He knew the REAL STORY however.
Thus Mike pretends to be a scientist whilst advocating against actual ones.: their own true OUTof this paradox is to declare themselves 'ahead of the curve' or 'before their time', readying themselves for the anticipated paradigm shift.... Reality must be a strange place for these creatures.
Alia: I got a "404 Not Found" message, when I tried to get on the SBM blog.
Jake says he has no intention of stopping...oh goody. This strife is a rich study in what posturing, dishonesty and too-many-chiefs will fetch you. And Jake's commentariat seems to be comprised of the very best of the tin-foil beanie brigade; comments from "White Rose" are almost poe-like in their paranoia and hyperbole.
@Science Mom - yes, White Rose does seem to be off-her-rocker....though I would love to see her list of "individuals for prosecution."
I still feel sorry for Jake. It's got to be rough for kids raised in woo to see that their parents are... well... wrong. Completely, utterly, totally wrong.
Honestly, I still do wonder if he'll figure it out quite suddenly at a later age; our age of real adulthood has been getting higher and higher lately, so it isn't as if he doesn't still have time.
#4 Goooooood point. Very good point.
@ Khani:
I am always a tad bit suspisious when an adolescent - or in this case a *puer aeternus*- in the process of forging an adult identity and socialising amongst generational cohorts- DOESN'T rebel or distance himself from his parents.
I wonder about the late-starting-adulthood trend as well. It's not just about economic viability I'd guess.
@Science Mom:
Given the courage and intelligence of the real "White Roses," I hope their ghosts come back and kick this woman for attempting to hijack their name.
You know, I kind of miss Jake coming to this blog to debate people.
Shay @ 13, no kidding! Another good example of the delusions of grandeur these nutters have. How they can liken themselves to real persecuted heroes is beyond me and frankly dangerous if that's how they view themselves.
I always get a 'sour grapes' vibe from woos regarding legitimacy. They want it so badly, and when they can't overcome the hurdles (or, more often, refuse to try), they whine about how meritocratic systems are so unfair to their merit-free beliefs.
One of the things the loons do that quite interests me is they give themselves titles: so Jake Crosby is "editor of autism investigated". Like, it's his blog. He's the editor of his blog. I find that so weird.
Orac: "Indeed, it’s hard not to conclude that Blaxill is a total hypocrite. He thought that the Geiers’ work was crap, but he never said anything in public because it might undermine the push to try to link vaccines with autism. Then, even though it’s likely he realized just how horrible the Geiers’ “Lupron protocol” was, he said nothing and let the Geiers continue to promote the protocol. Such are the ethics of the antivaccine movement."
Well, if your movement is akin to a religion (which is how at least some at AoA see it), you must not speak ill of co-religionists, or Satan will triumph.
Except things have been complicated by Jesus (excuse me, Jake) storming in to throw the moneychangers out of the antivax temple.
" they don’t show nearly as well as Andy Wakefield "
If I recall, Wakefield didn't even show at all for the autistic kids in vaccine court.
Jeff1971
Of course he wouldn't. At the risk of being actually cross examined and being cross examined on the record he'd probably open himself up for another lawsuit.
Love watching them eat each other like zombies (sorry, Walking Dead on my brain too much).
Isn't "White Rose" new to the blogosphere? I'm wondering if WR is posting on Jake's site to make Jake appear more popular. (S)he really is quite deranged.
@Denice Walter:
I'd agree. It seems to be about modern life. Around 100 years ago, you could leave school at 16 and get a job that would enable you to earn a living, albeit not a fantastic one. Try doing that today. It's not really possible. Too much has changed.
@AnObservingParty
Well, AoA did eventually get around to banning that hemp troll, though that may simply have been a personality thing, rather than any actual objection to their idea.
@Todd - also insano-thingy troll was banned from there as well...but you're right that it was probably more a personality thing than a disagreement on any particular position.
What I find interesting, as well, is that despite Blaxill's alleged misgivings regarding the Geiers, he didn't think they were sloppy enough to avoid citing them in 2003 to support his claims nor to support them in 2004 by criticizing a study that failed to replicate their results.
Whoops. That second link is borked. Let's try again.
Hey Orac,
I am playing by the rules. This blog is about vaccines. Again, I am late for the party but I am very much curious about your response to the COI interest charge levelled against you by Jake. So what exactly is your connection to the vaccine industry? I promise to keep an open mind and hear you out.
@Julian
Julian or any other VCADODer, what is the precise defense to the claim that in the Danish MMR study autism cases were redefined to mask the positive correlation with the MMR vaccine?
That's three, Greg. I even decided not to count further comments from you on the cancer thread, but you nevertheless hit three.
You are now on record as admitting that you lied when you claimed to know some study design that could be ethically done and would completely settle the issue of whether vaccines cause autism, even for the most die-hard antivaxxers, even if it didn't go the way they expected it.
And your next question is already waiting for you.
@Antaeus
Oh yes, study design... Honestly, at this point I would not even bother with a human population vax/unvax study if pro-vaxxers insist on whining so much about ethics. We simply can get our answer by studying to see if an unvax population has a 1 in 50 autism rate. I do recall reading on AoA a proposed study that would examine home-schooled populations autism rate. As you are probably aware, home-schooled kids are less vaccinated. I think this is a good start.
Also, it's simply incredible that pro-vaxxers don't seem so eager to pursue a vax/unvax study in the animal population. What is the fuss? Interestingly, the only such study that I know of found a link.
@Antaeus
Now that I answered your question about study design I will present this challenge to you: Given your propensity to go on a length critiquing things, how do you respond to this allegation that the Danish MMR study did initially find a link between autism and MMR, but it was conveniently wiped away by how autism cases were counted.
http://www.ageofautism.com/2013/09/round-2-cdcs-poul-thorsen-lying-in-p…
@Orac
Still no response to Jake's COI interest charge?
Try to keep up, Greg. You already answered the question about study design; namely, you admitted you were lying when you said you had one which would settle the issue once and for all, one that anti-vaxxers would have to accept even if the results weren't what they expected.
You may have additional "thoughts", as you call them, on study design, but I urge you to deposit them in the proper receptacle, a latrine for instance, as they're of no interest to anyone here.
I responded three years ago when he made the charge, pointing out how ridiculous it was and how it revealed nothing more than Jake's ignorance. I see no need to repeat myself to you.
Greg—
I call to your attention Hitchen's Razor:
You give no evidence for your assertion, so no defense is needed, or even appropriate. Your unevidenced assertion is dismissed.
As well, you are dismissed.
@Bill
"You give no evidence for your assertion, so no defense is needed, or even appropriate. Your unevidenced assertion is dismissed.
As well, you are dismissed."
Actually Bill, in the link I provided they gave the actual data sheet showing the link between MMR and autism. I would say that pretty much constitute evidence.
@Orac
Very well then, so you don't want to repeat yourself in regards to the COI allegation. Could you then point me to your initial response?
@Antaeus
"You already answered the question about study design; namely, you admitted you were lying when you said you had one which would settle the issue once and for all, one that anti-vaxxers would have to accept even if the results weren’t what they expected."
Please point precisely to where I said such a thing about lying about an unvax/vax study. Antaeus, every time I converse with you my opinion of you drops further.
And you'd be wrong. I see nothing that falls outside the margins of error, and nothing to support their conjecture that autism cases were redefined to cook the books.
Orac, if you're bored with writing about cancer, you should take a look at the AoA post Greg linked to.
Greg—
You show us, quite plainly, the typical antivaxer's disdain for ethics. If there were ethics and honesty involved, they wouldn't be so anti-health for everyone.
You are still dismissed.
@Julian - Sullivan / Matt Carey has a very good explanation of the two studies (and the differences) here:
http://leftbrainrightbrain.co.uk/2013/08/25/autism-denmark-and-again-no…
And for our "guest" here is another article that flies in the face of the "anti-vaccine philosophy" -
http://leftbrainrightbrain.co.uk/2013/08/10/the-amish-may-not-be-a-grea…
If anyone other than yourself expresses a need to see the posts involved, I might dig up the links for that person. Just like the Lurker Challenge, where I publicly announced that if anyone thought there might be some glimmer of merit in one of your ramblings, all they had to do was call it to our attention and I, at the least, would consider myself honor-bound to answer. No one taking up the challenge == not even your fellow anti-vaxxers think you have any sort of point.
You, Greg, were the one who chose to claim that you knew that a study which could settle the issue finally, once and for all, was so easily done that there could be no reason for mainstream science not to do it, other than fear that it would show vaccines causing autism. By making that claim, you were asserting that you knew of such a study design and that there were no legitimate reasons why the study design would be opposed, such as it being profoundly unethical or having absolutely unfixable problems with confounders (as your proposed homeschool study does) or that it wouldn't actually do anything to settle the issue because antivaxxers could find a way to discount the results if they didn't like them.
You were challenged to either pony up that study design, within a comment limit to keep you from simply avoiding the question indefinitely, or admit that you were lying when you claimed to have such a study design. Within the challenge period you didn't post a single thing about study design. An honest person who'd just gotten in over their head might have said "you know what, I'm sorry, I was mistaken, I thought I had a study design with no flaws but now I see the flaws, and it makes sense that pro-vaxxers wouldn't want money wasted on a flawed study that wouldn't prove anything" and we'd have accepted that. You chose to post inane insults and ramble instead of backing up the claims you yourself chose to make and what reasonable conclusion would any reasonable person draw from that except that you were in fact lying, that you COULDN'T back up your claims?
@Lawrence, yeah, I read that post. matt is very good at that sort of analysis.
@Antaeus
I guess I shouldn't be too surprised that you would now deliberately misconstrue my words in a feeble attempt to win an argument. Numerous times I have proposed that a vax/unvax study would be the definitive study to settle the autism debate. At no time did I suggest known of any such precise study design to address pro-vaxxers 'ethical' concerns. I will leave that up to you 'intellectual' elites who really want to resolve the autism debate (insert sarcasm) to devise. I have also made mention to other studies that could settle the debate such as a study of an unvax population to see if they have a 1 in 50 autism rate or an animal vax/unvax study. It is quite telling that Orac's minions have remained so quiet on this matter.
@Julian
I second your request for Orac to address the claim that they cooked the books in the Danish MMR study to erase the link. Over the last several days I have been asking Orac's disciples to weigh in on the matter, and, surprisingly, all I have gotten are 'tip-toeing' denials from even the usual
most outspoken pro-vaxxers.
'suggest knowing....'
Lawrence provided a link to a post on Left Brain Right Brain that addresses the claim. Go and read it.
I may regret this but:
It's evidence that both you and the idiots at AoA have no concept of what statistical significance is, nor any clue why the Danish study looked at life years instead of simply counting numbers of vaccinated and unvaccinated children with autism.
I'll be kind and give you a clue about the latter - the older a child the more likely they will have been given MMR (which is mostly given at around 18 months, with a number of children catching up later) and also, entirely independently, the more likely they will be diagnosed with autism, because most autism diagnoses occur at an age greater than 2 years. This means that the two will be correlated simply because both are more likely with increasing age.
The AoA article completely ignores this, which is more moronic than I would have believed possible were I not familiar with their previous pathetic attempts at science. I actually had to look at it several times just to convince myself that they really had been as stupid as it appeared at first glance. They had been. You might as well look at the correlation between autism and number of days spent at kindergarten or the number of well baby visits and draw some idiotic conclusion about causation from the inevitable correlation you would find.
Already done last month:
http://scienceblogs.com/insolence/2013/09/24/does-a-new-danish-study-di…
Previous mentions of how the antivaccine movement launches fallacious attacks on the Danish studies:
http://scienceblogs.com/insolence/2010/03/12/thorsen-wakefield-fine-art…
http://scienceblogs.com/insolence/2011/04/14/poul-thorsen-vaccines-frau…
It's not provaxxers 'ethical concerns' (love the scare quotes) that prevent the kind of study that antivaxxers demand, it's The Declaration of Helsinki, a set of internationally agreed rules about how medical research can be conducted ethically.
In other words, you were knowingly lying when you claimed 'the only reason anyone would oppose doing this study is that they're afraid of the results'; you knew even if you didn't understand that it simply wasn't that easy, but you decided a lie would serve your purposes better than the truth. Thanks for exposing to all the world just what kind of people anti-vaxxers are.
(Orac ignore this post if it is a duplicate.)
@Kreb
I don't have the figure at hand but do you know the age distribution for the kids in the Danish study? How old were the youngest at the time of the study?
Greg,
See for yourself here.
So the youngest children were under 6 months old, the oldest were 7 years old.
The AoA article added all these together and produced a graph showing the rate of autism, other ASD and combined broken down by vaccinated and unvaccinated, with no correction for age. Do you see why this is extremely foolish?
Reading that last comment back, it's a bit of a non-sequitur. To clarify, I was combining the information I quoted with the information in Table 2 in the paper, which the AoA article derived its data from.
@Kreb
But, wait Kreb! You are suggesting that including the younger ages in the study would have skewed the results because some of these kids would have been too young to be diagnosed with autism. Yet, this would actually help the no-link claim! And, if
the actual data that was done on this methology showed a link does it not make the link claim even more compelling. Again, such a study should have been bias towards the no- link side. It really does seem that the 'autism free years' intervention was indeed a way out of the initial 'horrific' result.
Greg,
No it wouldn't, it would do exactly the opposite! You still don't get this do you? The kids who were too young to be diagnosed with autism would also be too young to have had the MMR, leading to a correlation between having had MMR and having a diagnosis of autism.
Let me explain things to you even more simply. Take 1,000 Danish children aged 12 months; lets's say none of them have had the MMR and none have (yet) been diagnosed with autism. Take another 1,000 Danish children aged 7 years; 900 of them (90%) have had the MMR and 9 of these (1%) have been diagnosed with autism (all figures approximate and for the sake of argument), while 1 of the unvaccinated 7 year olds (also 1%) is autistic.
Add these two groups and of the 1,100 unvaccinated children, only about 1 has been diagnosed with autism, that is less than 0.1%. Of the 900 vaccinated children, there are 9 autistic children, about 1%.
Hopefully you can see why it would be foolish to conclude from this that MMR increases the risk of autism. Now do you see why the author of that AoA article was foolish not taking age into account in his figures?
@Greg: are you deliberately being obtuse? Please go read the links given to LB/RB and Orac's reviews of the Danish study, then come back when you might sound more intelligent about it.
And yes, we are really concerned about the ethics of any study. So until you can work with the ethics of a study, you can't speak to the the results.
@Kreb
Kreb, actually it is helpful to break things down even further. When considering how the figures may have been unfairly skewed in favour of either the link or no-link side, there are three vaccination scenarios to reflect upon. First, there were the kids that were too young to be vaccinated and diagnosed with autism. Second, there were the other children that reached vaccination age in which many would have been vaccinated, but it was still too early to detect signs of autism. Then there was the third and last case of the kids that were old enough to be vaccinated, and it was also possible to detect autism at their age. This third case would not unfairly skew things, and it was indeed the optimum scenario for determining the true autism-MMR relationship.
Kreb, you are right, the first scenario would likely produce a result that bias things in the direction of there being an unfair over-representation of unvaccinated non-autistic kids, when if we were to contemplate the autism being genetic argument (please understand that I am only humoring you here), some of these kids would have eventually develop autism.
Yet Kreb, what is to be made of the second scenario? Of the vaccinated kids who were still too young to detect their autism, we would have nearly zero autistic cases. Yes, we would also have some unvaccinated kids (now by parents' choice) who may also later develop autism, but since most parents vaccinated their kids their numbers would have likely been relatively low. The net effect, Kreb, is that we would have had a whole horde of vaccinated non-autistic kids that unfairly skewed things towards the no-link side.
Considering both these scenarios, Kreb, the million dollar question is which is biasing things more? Based on the age distribution of the study -- 6 months to 7 years-- I am feeling that the sample size of kids that were younger than 15 months (the time of Danish MMR uptake) is less than the sample size of 15mths to 3yrs (time of MMR uptake and when autism was likely to go undetected).
In conclusion, I still firmly believe that a straight comparison of MMR and autism would be skewed towards a no-link scenario, and the fact that the data points to a link is indeed compelling.
@MI Dawn
Regarding the actual argument of whether autism cases in the Danish study were counted in a crafty way to conceal the autism link, I didn't find much in the links that Orac provided that directly touched upon this. Maybe I skimmed too lightly through his ramblings but I didn't find anything. Generally, when reading Orac's inane ramblings I find it best to heed this caution: 'If you keep a too open mind, someone will fill it with BS."
Greg,
Perhaps you missed this in Krebiozen's comment at 49.
with no correction for age
MMR is given at 18 months. Diagnoses of autism increase with age. If this is not corrected for (and AoA didn't), the results will be skewed.
Understand now? Or do I need to break out the crayons and draw you a picture?
how do you respond to this allegation that the Danish MMR study did initially find a link between autism and MMR, but it was conveniently wiped away by how autism cases were counted.
Ah, the "Let's-you-and-him-fight" approach to trolling. I am shocked, shocked! that Greg takes these allegations so seriously that he can't even be bothered repeating them, or even summarising them.
It could just be nostalgia but I'm sure trolls used to put more effort into their provocations.
To put it another way... anyone who sees any value in pointing out the holes in some numpty's claims at AoA will already have clicked over to AoA to do it there.
When one can argue with the organ-grinder, why argue with the monkey?
@HBD - leave it up to the resident troll to fail basic math.....
Does Greg realize that what he's pointing out is only how his buddies at AoA f*cked up, not anything wrong with the original study?
@AF,
No.
Taken from #54.....
"Then there was the third and last case of the kids that were old enough to be vaccinated, and it was also possible to detect autism at their age. This third case would not unfairly skew things, and it was indeed the optimum scenario for determining the true autism-MMR relationship."
'
Has anyone thought of running the numbers to see how these 5-7yrs old would have fared on the autism-MMR question? "Autism free years"? ---What a joke!
Ok Greg, out come the crayons.
False, it would skew things.
Here's an example from a commenter on an earlier post. Suppose I compared shoe size and proportion of children diagnosed with autism. And suppose I then pointed out that the larger the shoe size, the more likely the child was to be diagnosed with autism. And suppose I then claimed that this was proof that shoe size had something to do with autism. All proper scientists would roll their eyes at me.
Not every child diagnosed with autism is diagnosed at 18 months. If you fail to correct for that (and AoA didn't) you will get misleading daat.
My final paragraph is ambiguous. I meant:
Not every child diagnosed with autism is diagnosed at 18 months. If you don't correct for that (and AoA didn’t) you will get misleading data.
Greg,
Good grief. If you still can't understand the clear and simple explanation I gave you, you are beyond my help.
Do you have any idea how moronic that sounds to people who understand this?
HDB,
I'm asking myself that very question.
@Julian
"Not every child diagnosed with autism is diagnosed at 18 months"
Yes Julian, but at 5 - 7 yrs age cohort it is reasonable to assume that nearly all the children would have been diagnosed by then. In fact (even speaking as a layperson), I imagine it's possible to run the numbers and calculate the likely percentage of kids that would not have been vaccinated at those age cohort, and which could be used to assess the confidence interval of the results.
As for the analogy you gave, Julian, yes, we can read those figures as saying that larger shoe sizes are linked to autism. We cannot conclude that shoe size causes it, but we can say they are linked. Similar, we could crunch the nunbers to see if the 5-7yrs cohort vaccinated kids have autism at a disproportionate rate. Again we are not arguing causation but just a link.
Except that that wasn't done, Greg. AoA compared the 5-7yrs cohort to the unvaccinated cohort, and the unvaccinated cohort included those too young to be vaccinated.
Do you now see the problem?
Come on Julian - do you expect AoA to present something in a way that doesn't automatically support their position?
@Julian
I am not concerned with what AoA did. The simply question remains, why did the researchers not limit the study to the older cohorts (5-7yrs) that were at vaccination age and it was possible to detect autism in them. Why resort to the dubious 'autism free years' solution?
Are you listening fence-sitters? 9 out of 10 times if it sounds unnecessarily complicated, chances are you are being taken.
@Julian
Further Julian, even if AoA did a straight comparison based solely on MMR and autism without adjusting for age, at #54 I explained how, overall, this would likely skew things unfairly towards the no-link side. Yet, the data still found a link, and making the finding even more compelling!
@Lawrence
I was meaning to ask you of a favour. A friend of mine is moving to your area. He is looking for a pediatrician that is cool with delaying his kids' vaccines. Can you refer yours?
@Kreb
"Good grief. If you still can’t understand the clear and simple explanation I gave you, you are beyond my help."
Kreb, I think you are starting to not play so fair in your 'sandbox of life' . Perhaps a time-out is in order for you.
What am I not understanding? I think you are the one in the error because if you refer to my post at #54 your example is not accounting for a large group of vaccinated/non-autistic children. Here is what you said:
"Let me explain things to you even more simply. Take 1,000 Danish children aged 12 months; lets’s say none of them have had the MMR and none have (yet) been diagnosed with autism. Take another 1,000 Danish children aged 7 years; 900 of them (90%) have had the MMR and 9 of these (1%) have been diagnosed with autism (all figures approximate and for the sake of argument), while 1 of the unvaccinated 7 year olds (also 1%) is autistic.
Add these two groups and of the 1,100 unvaccinated children, only about 1 has been diagnosed with autism, that is less than 0.1%. Of the 900 vaccinated children, there are 9 autistic children, about 1%. "
But, actually Kreb, your example should have also contemplated, say, 1000 Danish kids at around 2 years old that reached the Danish MMR vaccination age. Let's say 900 (90%) of them are now vaccinated, and 100 (10%) of them are not vaccinated. Again, since it would be too early to detect autism at this age we would have virtually all of them counted as not autistic.
Kreb, adding things, what we now have is 1800 vaccinated kids with 9 having autism or .5%, which is significantly less that your 1% figure. We would also have 1200 not vaccinated kids with 1 autistic or .08 which is a little less than your .1% figure. Simply Kreb, the biases inherent in 'too early to vaccinat - too early to detect autism' and 'at vaccination age - too early to detect autism' groups are virtually erasing each other.
You might still argue here that the 'too early to vaccinate -- too early to detect autism' group (less than 18mths age cohort) is having a bigger skewing effect, but there are two other things to consider. In your example, to simplify things you assign the groups a sample size of 1000. Yet, given the actual data with the age distribution of 6mths to 7 yr, with kids getting vaccinated at 18mths we would expect a much large sample size of kids that were old enough to be vaccinated and also old enough to detect their autism (5 - 7yrs cohort). Also, within the 'at vaccination-too early to detect autism' group, given that some kids at 3,4, or even 5 yrs autism may have still gone undetected we would expect this group to have a relatively larger sample size than the 'too early to vaccinate -- too early to detect autism (less than 18 mths cohort). These two effects would essentially mitigate the skewing effect of the 'too early to vaccinate/too early to detect autism (less than 18mths cohort), with the large sample size of the 'at vaccination age -- too early to detect autism' group likely biasing things towards the no-link side.
At this point, the really interesting question then to ask is why did the data reveal a link if things were bias towards the no-link side. Answering this, I think it's fair to surmise that the older kids who were old enough to be vaccinated and old enough for their autism to be detected (5-7yrs cohort) probably significantly established the link. And, as I mentioned earlier, this was the prime group to study without resorting to the dubious 'autism free years' counting.
@Kreb
Further to #73, I also don't believe the skewing effects from the different age cohorts would have been entirely a speculative thing. I believe it could have been calculated and used in a meaningful way when making sense of the results -- hence making the 'autism free years' solution even more unnecessary.
Ok Orac,
Why did you not approved these last two comments that are relavant to the discussion that Kreb and I have been having on the Danish MMR study, yet you approved another of my comment on another thread? The comment on another thread was also posted after these.
@Orac,
Also Orac, you still have not yet directed me to your early comments in which you 'defended' yourself against the COI allegations that Jake made against you.
@Orac,
Also Orac, you still have not yet directed me to your early comments in which you 'defended' yourself against the COI allegations that Jake made against you.
Greg:
No. You asserted it. And you were wrong. In fact, the opposite of what you asserted is true.
AoA compared two groups: a group at 5-7 years and a group of unvaccinated children from 0-7 years. The children from 0 to 1.5 years are both too young to be vaccinated and extremely unlikely to have been diagnosed with autism. That, on its own, skews the results to looking as if the unvaccinated are less likely to be autistic In addition, not every child is vaccinated on schedule and not every autistic child is diagnosed at 18 months, so the older a child gets the more likely he or she will be both vaccinated and diagnosed with autism.
I can't make it clearer than that.
@Julian
I posted a detailed comment addressing your take and Kerbiozen's take. To this point Orac as not approved it. If Orac does and after reading I challenge anyone, including Orac's disciples, to show me where my reasoning is flawed.
Well, that should be easy.....
Greg:
There's your error. Both occur at 18 months. The ‘at vaccination age – too early to detect autism’ group is much smaller than the supposed ‘too early to vaccinat[e] – too early to detect autism’ group.
You never learnt statistics, did you Greg? I did, and there are ways to control for sample groups of different sizes.
Incorrect. The ‘at vaccination age — too early to detect autism’ is smaller than you think.
I hope that clears things up.
@Julian
Thanks for implicitly conceding that you and Kerb did indeed neglect the age cohort of 'at vaccination age -- too early to detect autism' that may have skewed the results in favour of 'no-link'. Acquiescing to a mistake is one thing, Julian, but I am sure we can both agree that the nicer thing to do is explicitly admit to it. I will also wait for Kerb to grace us with his presence and also concede the mistake. Still, last I heard, he is back to ignoring me. (Hey Orac, can you solicit your minions to see who is fully signed up to ignoring me? It's starting to get real confusing.)
Anyway, Julian, it appears that you have switched tactics to now argue that skew towards an unfair over-representation of the 'not at vaccination age/too early to detect autism' (less than 18mths cohort) would still win the day and minimize the effect of the 'at vaccination age/too early to detect autism' group. Very well then, Julian, I will leave it up to you to produce the actual findings that this was indeed the case and, second, that the finding of a link was entirely due to this effect. Until you do so, Julian, I am sure we can both conclude that there is no conclusive evidence that the finding of a link was strictly the result of unfair skewing, and rather than there being a true link.
Also Julian, I still have not obtained a satisfactory response of why it would have not been better to limit the research to the older cohort of 'at vaccination age/possible to determine autism' that would avoid all unfair skewing. Surely then it would not have been necessary to resort to the dubious 'autism free years' counting. Even more interesting would be examining the data for the results of this strict group.
Greg,
I'm not sure if you are deliberately playing dumb or if you truly don't understand this. I'm going to give you the benefit of the doubt.
I'm happy to help when someone is ignorant and asks for help, but when someone is so very ignorant and won't even admit it, and makes fun of people who do understand a subject, I get tetchy. For example:
You mean this is too complex for you to understand, but instead of asking for assitance or clarification you assume people are trying to fool you? How did you ever learn anything at school?
Good question, I'm struggling with that myself. Did you study math at school? Statistics? Perhaps you don't understand the study design, and what the AoA authors did with the data. This is a cohort study, which includes children of different ages - is that clear to you?
Since the great majority of children will eventually be vaccinated and the great majority of them will not be autistic, this makes no difference. It is the proportion of unvaccinated non-autistic to vaccinated non-autistic that AoA measured, and there will of course be more of the former than the latter, because even those children who will be later diagnosed as autistic will be included in the former group, but not in the latter group.
The rest of your comment is, frankly, meaningless. You don't appear to understand what is going on here at all.
I'm going to try to explain this one last time with another hypothetical scenario. Let's say the rate of autism is the same in all children, regardless of whether they are vaccinated or not, as I believe is the case in reality. There are 1000 children in each group of children aged 0-1, 1-2 etc.. The number of diagnosed autistic children increases by age, approximately as in reality. The proportion of children who have had the MMR also increases with age, again approximately as in reality.
AgeVaccinatedUnvaccinatedDiagnosed autistic (vax/unvax)
0-1010000
1-26004003 (2/1)
2-37003003 (2/1)
3-48002005 (4/1)
4-590010010 (9/1)
5-690010010 (9/1)
6-790010010 (9/1)
There are a total of 41/4,800 (0.85%) vaccinated autistics and only 6/2,200 (0.27%) unvaccinated autistics. Does this mean that vaccinated children are three times more likely to be diagnosed with autism than unvaccinated children? Of course not! These figures were constructed on the basis that vaccination makes no difference to autism.
It would be very foolish to calculate the figures like this yet this is exactly what the AoA author did.
Sorry typo - that first figure should read 35/4,800 (0.73%), but it makes little difference. I also forgot that tabs don't work in these comments, but hopefully it makes sense anyway.
Another redundant clarification, just because I'm pretty sure Greg will still be unable to grasp this. As I have explained, the figures need to be corrected for age, because age is a confounder; the odds of both having been given MMR and gaving an autism/ASD diagnosis increase with age.
The author of the AoA article that Greg cited at #31 did not correct for age, so his graph is meaningless. The authors of the NEJM paper did correct for age and concluded:
I'm having a typo-ridden day - I meant "having an autism/ASD diagnosis".
@Kreb
"There are 1000 children in each group of children aged 0-1, 1-2 etc.. The number of diagnosed autistic children increases by age, approximately as in reality. The proportion of children who have had the MMR also increases with age, again approximately as in reality.
Age Vaccinated Unvaccinated Diagnosed autistic (vax/unvax)
0-1 0 1000 0
1-2 600 400 3 (2/1)
2-3 700 300 3 (2/1)
3-4 800 200 5 (4/1)
4-5 900 100 10 (9/1)
5-6 900 100 10 (9/1)
6-7 900 100 10 (9/1)"
You said, 'the number of diagnosed children also increased with age". Fair enough, but what is your rationale for breaking down the increase in such a linear way between vaccinated and not vaccinated. For instance at the age 4-5, 5-6, and 6-7 cohorts yes indeed it's fair to say that we would expect 10 cases of autism for each of these cohorts, but if vaccines and autism are strictly random why would we not expect that for some of these cohorts there may be, say, six cases of vaccinated autistics and 3 cases of not-vaccinated autistics?
First, I'm going to attempt to make the table come out more legibly. Here goes...
Age..Vaccinated..Unvaccinated..Diagnosed autistic..(vax/unvax)
0-1..0...........1000..........0
1-2..600.........400...........3...................(2/1)
2-3..700.........300...........3...................(2/1)
3-4..800.........200...........5...................(4/1)
4-5..900.........100...........10..................(9/1)
5-6..900.........100...........10..................(9/1)
6-7..900.........100...........10..................(9/1)
Secondly, Greg asks, "what is your rationale for breaking down the increase in such a linear way between vaccinated and not vaccinated. For instance at the age 4-5, 5-6, and 6-7 cohorts yes indeed it's fair to say that we would expect 10 cases of autism for each of these cohorts, but if vaccines and autism are strictly random why would we not expect that for some of these cohorts there may be, say, six cases of vaccinated autistics and 3 cases of not-vaccinated autistics?"
The answer is "the law of large numbers." This law says that the average of a large number of trials will be close to the expected value, and will get closer and closer as the number of trials increases. If you flip a coin 4 times, the most probable ratio of heads to tails is 1:1, but it wouldn't be impossible or even unusual to get 1:3, 3:1, or even 0:4 or 4:0. Flip a coin 40,000 times, however, and if you get a ratio significantly different from 1:1, it means the coin you're flipping is almost certainly weighted or otherwise rigged.
This means that if you want to do a thought experiment with a small number of trials to demonstrate a principle, and you want the results to "scale up" and be applicable, you cannot put flukes in the data, even if such flukes would be likely (even when they would be certain) to happen at the small scale. To return to our coin flipping example, you would not portray the result of flipping a coin 4 times as anything but a 1:1 ratio, not if you wanted to understand something about flipping a coin 4000 times. If for some reason you had to create the model you wanted to scale up using just one coin flip, the best way to do so would be to pretend that the coin came down half heads and half tails - even though it is clearly impossible for that to happen.
The entire point of this thought experiment is to show that an illusion of correlation can be produced between two things which are not correlated if you misinterpret the data, as AoA misinterpreted the data of the Danish study. Greg is asking, essentially, why we can't add illusory signs of correlation - for instance, a year where for no reason the majority of children diagnosed as autistic had already had their MMR - directly to the data, just to make it look more random to his perceptions. I think the answer is pretty obvious by now.
Greg:
"Switched tactics"? You were the one who raised the argument that the ‘at vaccination age/too early to detect autism’ group would skew results back.
As for your demands for proof:
We have a cohort of children aged 0-18 months, both unvaccinated and undiagnosed.
Let's suppose that we have 1000 children in each 6-month block for a total of 14,000. That gives us 3,000 unvaccinated and undiagnosed children out of a total 14,000. Assuming a diagnosis of 1%, 140 children would have an ASD diagnosis. Even allowing for a 25% delay, that's 35 children.
It should be clear to anyone that the 3,000 would skew things far more than the 35.
Why are you asking us? Address your question to AoA. THEY were the ones who represented things like that.
Finally, I'm going to repeat what Krebiozen said.
Please bring back an edit button.
We have a cohort of children aged 0-18 months, both unvaccinated and undiagnosed.
Let’s suppose that we have 1000 children in each 6-month block for a total of 14,000. That gives us 3,000 unvaccinated and undiagnosed children in the 0-18 month blocks. Assuming an A.S.D. diagnosis rate of 1%, there would be 40 children diagnosed with A.S.D. in the 5-7 years cohort. Even if diagnosis was delayed by 25%, that's just 10 children. It should be clear to anyone that the 3,000 would skew things far more than the 10.
@Kerb,
With the age cohorts 1-2, 2-3, and 3-4 -- what was your reason for splitting the autistics between vax and unvax the way you did? Why did you not also assume that with an autism rate of 1 percent for every 100 vaccinated there would be one autistic?
Greg,
I think Antaeus has explained this. Isn't is obvious that if 1% of children are autistic regardless of whether or not they have had MMR, then 1% of vaccinated children will be autistic, and 1% of unvaccinated children will be vaccinated? Why introduce random variation into the picture when we don't have to?
I was trying to make it clear to you that if you assume that 1% of both vaccinated and unvaccinated children develop autism, you can demonstrate that the AoA method of evaluating the data gives very misleading results. Why can't you see this?
The question of why the authors of the study chose to use the statistical methods they did is more complex. Mostly they were trying to get the greatest statistical power possible by using the greatest numbers possible. Using person years seems perfectly valid to me, it's a measure widely used in epidemiology. If vaccines did cause autism, you would expect more cases after vaccination than before, after age was accounted for, and this would show up when using person years.
Only 7 children were diagnosed with autism at over the age of 6 years (Table 1 of the NEJM paper), which greatly reduces the statistical power of the results if you just compared vaccinated and unvaccinated children over the age of 6. Even then you might still have confounding from age, so I don't see how you could come to any conclusions doing this.
In my opinion the Danish authors did the right thing and came to the correct conclusions.
Julian,
I imagine you are experiencing the same frustration in explaining this as I am.
I'm reminded of trying to explain the concept of gravity to a (somewhat intellectually challenged) 6-year-old with the aid of a globe and a toy soldier. Try as I might, I couldn't get him to understand that what we call "down" actually means "towards the center of the Earth", and Antipodeans don't fall off into space.
@Kreb
You are making two assumptios that I am having trouble with. For the age cohorts of 1-2, 2-3, and 3-4 why are you not assumming a vax rate of near 90 percent. Second, even if there are 800 vax to 200 unvax at a 1 percent autism rate why wouldn't it be 8 vax and 2 unvax, instead of 4 vax to 1 unvax.
@'Kreb
Kreb, it seems you are deliberately dancing between 'rate' and 'ratio' in a sly way.
@Greg, why don't you add a K (for thousand) on the end of each of Antaeus's numbers? Then you will have such large numbers that it is unlikely that you will have much deviation from average, and it does not change the argument.
The whole point of the thought experiment is to show that poor processing of data (like the kind demonstrated by AoA) can lead to the illusion of a correlation where no correlation actually exists.
The reason for "splitting the autistics between vax and unvax the way [Krebiozen] did" is because anything else would be violating one of the base premises of the thought experiment, which is that no correlation between vaccination status and development of autism actually exists.
The whole point is that if there is no vaccine-autism connection, the ratio of vaccinated-to-unvaccinated among autistic children in a particular age group will be the same as the ratio of vaccinated-to-unvaccinated among ALL children in a particular age group - but if you do what AoA did, and fail to correct for age, you'll get the illusion of a correlation.
@Greg: "Second, even if there are 800 vax to 200 unvax at a 1 percent autism rate why wouldn’t it be 8 vax and 2 unvax, instead of 4 vax to 1 unvax."
Because he's assuming only half of the children who are in fact autistic are diagnosed autistic in that age range. All of them are assumed diagnosed by age four when the number of children diagnosed autistic goes to 1%.
Ok Guys let’s recap.
AoA said they compared the rate of autism between vaccinated and not-vaccinated cases. And I quote them….
http://www.ageofautism.com/2013/09/round-2-cdcs-poul-thorsen-lying-in-p…
“You look at one group of people who get a vaccine, and then you look at a group who did not get the vaccine. If there is more of a health problem in the group who got the vaccine than those who didn’t, then there very well could be a problem with the vaccine. Simple.”
The issue then is what’s wrong with this method. Kerb and others pointed out that it is vulnerable to the skew of there being an over-representation of the younger age cohorts (0-1, 1-2) of kids that are counted as not-autistic that may become autistic later when they are vaccinated regardless of vaccines causing it. I agree with this but its impact may be over-stated given that kids get vaccinated at such an early age (18mths). Reading Kerb hypothetical figures, however, we get a different story -- a gradual incremental increase in vaccination rates. These are his figures……
--------------------------------------------------------------------------------------------------------
“Age Vaccinated Unvaccinated Diagnosed autistic (vax/unvax)
0-1 0 1000 0
1-2 600 400 3 (2/1)
2-3 700 300 3 (2/1)
3-4 800 200 5 (4/1)
4-5 900 100 10 (9/1)
5-6 900 100 10 (9/1)
6-7 900 100 10 (9/1)
There are a total of 41/4,800 (0.85%) vaccinated autistics and only 6/2,200 (0.27%) unvaccinated autistics. Does this mean that vaccinated children are three times more likely to be diagnosed with autism than unvaccinated children? Of course not! These figures were constructed on the basis that vaccination makes no difference to autism.”
I consider with an 18mth MMR uptake vaccination at 60% for the 1-2 age cohort is plausible. Still, does Kerb want us to believe that immunization for the 2-3yrs cohort is 70%, and 80% for the 3-4yrs. Kerb we both know that those rates would be closer to 90%.
And Kerb, how did you ever come up with a vaccination ratio that mirrors autistic cases. With a 1% autistic rate, 600 vaccinated kids would translate to 6 autism cases, and 400 unvaccinated cases would mean 4 autistic cases.
Making these adjustments, Kerb, and re-working your figures we would have…..
0-10 1000 0
1-2 600 400 10 (6/4)
2-3 900 100 10 (9/1)
3-4 900 100 10 (9/1)
4-5 900 100 10 (9/1)
5-6 900 100 10 (9/1)
6-7 900 100 10 (9/1)
In total, Kerb, we now have 51/5100 (1%) vaccinated autistics 9/1900 (.47) unvaccinated autistics. Essentially the rates are moving closer together. There is still the bias that we expect from the ‘0-1’ age cohort. Yet, I mentioned another factor that may mitigate this effect, and that is the fact that autism cases are likely to go unnoticed in the younger cohorts.
I will now end with two questions.
1.If we run these figures to approximate the very large sample size of the Danish study would the bias still be significant.
2.Being that it is possible to calculate the bias effect is it possible to use this to determine whether it is accounting for the link in the actual data.
Ummm...this is spectacularly dense, even for you. You do know that rates are ratios, right?
First line of the wiki page: "In mathematics, a rate is a ratio between two measurements with different units"
Greg,
I see others have addressed your points more than adequately, but it seems to me you need all the help you can get, so here's my response as well.
I'm not really making any assumptions, I'm setting out a scenario that is based on there being no difference at all in autism rate between vaccinated and unvaccinated and showing that this leads to misleading results if you do what AoA did with the figures.
Because that's what the NEJM figures show - see Table 2 in the NEJM paper, though the numbers are given as person years it is clear that there are still children getting vaccinated at age 6 and 7. MMR in Denmark is recommended at 15 months, but in practice though children mostly get vaccinated at that age, many don't and instead get vaccinated over the next few years, when they visit the doctor for example; not all are vaccinated even at 7 years.
There isn't a 100% detection of autism at that age. I was guessing at 50%. That's kind of the whole point. If you look at Table 1 in the NEJM paper you will see that 57 children were diagnosed with autism at age 0-2, 218 at age 3-5 and 148 at age 6 or over, so perhaps 60% would have been more realistic, but it doesn't make much difference.
Not at all. There are two rates here: firstly the autism detection rate which will increase between age 0 and 7 years as I replicated in my figures, secondly rate of autism in vaccinated and unvaccinated children which I made the same.
At age 3-4 only 50% of the children who will eventually be diagnosed as autistic will have a diagnosis, that's 0.5%, or 5 children out of 1000. Since there are 4 times as many vaccinated as unvaccinated children there will be 4 times as many vaccinated autistics i.e. 4 as unvaccinated autistics i.e. 1.
Is that any clearer?
Greg:
Unless there was a confounder that you failed to take into account, like age. As we have told you repeatedly, the likelihood of being vaccinated and the likelihood of receiving an A.S.D. diagnosis both increase with age.
To illustrate my point, let's alter the wording in the words I blockquoted.
You look at one group of people who have a large shoe size, and then you look at a group who have a smaller shoe size. If there is more of a health problem in the group with the larger shoe size than the group with the smaller shoe size, then there very well could be a problem caused by shoe size.
Do you now see the problem with failing to control for age?
With his usual instinct for the FIFUDOS, Greg completely misses, not just one, but ALL the points.
First, he quotes the AoA numpty:
“You look at one group of people who get a vaccine, and then you look at a group who did not get the vaccine. If there is more of a health problem in the group who got the vaccine than those who didn’t, then there very well could be a problem with the vaccine. Simple.”
Well, guess what? That AoAer was wrong. Here's a corrected version:
"You look at one group of people who get a vaccine, and then you look at a group who, though substantially similar in all other important aspects, did not get the vaccine. If there is more of a health problem in the group who got the vaccine than those who didn’t, then it provides evidence that there might be a problem with the vaccine."
That is why any legitimate way of interpreting the data must do what the AoA numpty did not do: control for age and other factors, comparing 1-year-olds who didn't get the vaccine to 1-year-olds who did get the vaccine, not to 5-to-7-year-olds who got the vaccine.
Second, all Greg's quibbling about whether Krebiozen's sample figures for vaccination and autism rates are accurate enough evades the key point that in this thought experiment they are defined as unrelated. We could just as easily have done the demonstration by inventing data on what age a child first encounters the "got-your-nose" game.
Would Greg be seriously arguing that we need to provide more accurate figures on exactly when that first playful "got your nose!" is heard, in order to show what we already know, that the AoA understanding of the scientific method is fatally flawed and any results produced from it are junk?
@Kerb
Again, let's clarify. Never did I deny that a straightforward comparison of autism rates for vax/unvax in the Danish study is not susceptible to the bias coming from children younger than the vaccination unfairly skewing things towards the 'link' side. The real issue Kerb is whether this effect could not have been accounted for without resorting to 'life years free of autism' counting. Could we not have adjust for the bias and determine whether there is indeed a link.
Kerb, look again at the example at #94 for the hypothetical sample:
--------------------------------------------------------------------------------
0-1 0 1000 0
1-2 600 400 10 (6/4)
2-3 900 100 10 (9/1)
3-4 900 100 10 (9/1)
4-5 900 100 10 (9/1)
5-6 900 100 10 (9/1)
6-7 900 100 10 (9/1)
In total, Kerb, we now have 51/5100 (10%) vaccinated autistics 9/1900 (.47) unvaccinated autistics.
--------------------------------------------------------------------------
Note Kerb that we are dealing with 10% vaccinatex autistics and 5% unvaccinated autistics. (The rate is high because of the 0-1 age cohort in which our sample of 1000 have zero autism cases because they are too young.) But Kerb, we have determined the bias effect, and a 5% discrepancy in autism rate between the vax/unvax would be considered normal -- the no link situation. If however, the number showed that the vaxed autism rate is 10% rate and the unvax rate is 1% then we would have a good reason to conclude that there is indeed a link. The discrepancy is significantly greater than 5%.
And, this bring me to the next important point. Even if AoA calculated the autism rate for vax/unvax cases in the Danish study and found a link, how can it be concluded that this link is solely a result of skewing caused by the younger age cohorts? What if even after accounting for the skewing effect there is, indeed, a true link?
Kerb, to resolve this question I guess I will have to pose the same challenge that I posed to Julian. Please return with the actual figures demonstrating how the skewing effect is causing the finding of a link in AoA's calculation.
'10% vaccinated'....
'This bring(s) me...
@Kirb
"MMR in Denmark is recommended at 15 months, but in practice though children mostly get vaccinated at that age, many don’t and instead get vaccinated over the next few years, when they visit the doctor for example; not all are vaccinated even at 7 years"
Very well then Kerb, many parents do hesitate and delay their kids vaccines, but are we to believe that the actual Danish data had vaccination rate at 70% for the 2-3 age cohort, and 80% for the 3-4? C'mon Kerb!
You just made exactly the same error AoA made, the same error others here have tried to explain to you in the simplest possible terms. Greg, did you understand Julian Frost's comment @102?
This is what the authors did! It's called correcting for age! Just because the result was not what you wanted it to be does not mean that it must be false.
Could you define, in statistical terms, what exactly qualifies as a 'link?'
Greg @ #104:
I refer you to your earlier comments. Firstly, #51.
#54:
#73:
#82:
Deceiver, deceiver, pantaloons aflame.
Shorter Greg:
"We could do the data analysis correctly, but that doesn't give the answer I want. Why don't we do the analysis wrong instead, and then pretend to 'correct for the error'? That flawed process might happen to go my way, in which case I'll pretend it's the only way to look at the data."
Only appropriate response: "Get bent."
Gerg,
I'm sorry, I have tried to explain this to you, but the questions you ask tell me that you have absolutely no understanding of maths or statistics at all. For example at #99 you asked:
I'm still staggered that anyone could miss the whole point of this exercise so completely. At that age only 50% of the autistic children have been diagnosed - look at Table 1 in the NEJM study, which shows that most children are diagnosed between the ages of 3-5 years with some being diagnosed at over 6 years of age.
You then reworked the figures so that all autistic children were diagnosed at the age of one and then, mind-bogglingly, stated:
This is the very factor that you just removed from my figures! I really don't like impugning people's intellects as I believe almost anyone can learn, but really, how can I possibly avoid concluding from this that you are an idiot?
@Adam
" If however, the number showed that the vaxed autism rate is 10% rate and the unvax rate is 1% then we would have a good reason to conclude that there is indeed a link.
You just made exactly the same error AoA made, the same error others here have tried to explain to you in the simplest possible terms. Greg, did you understand Julian Frost’s comment @102?"
Adam, it's not about whether the actual data reveal a high autism rate due to the younger cohorts bumping the number up. It's about whether the descrepancy between the rates fall within the expected skew range. In my opinion it is rather an easy solution. We can calculate this effect. Heck, we can even graph for it.
@Antaeus
Shorter Greg:
“We could do the data analysis correctly, but that doesn’t give the answer I want. Why don’t we do the analysis wrong instead, and then pretend to ‘correct for the error’? That flawed process might happen to go my way, in which case I’ll pretend it’s the only way to look at the data.”
Only appropriate response: “Get bent.”
I haven't really reflected on the 'autism free years' solution ,yet, to see if it's indeed legitimate. They very fact that it sounds so dubious leads me to wary. I am a firm believer that if a simpler solution is at hand, why resort to a complicated one.
And Antaeus, are you seriously suggesting that we couldn't account for the skewing effect to make accurate calculations without resorting to the 'autism free years' solution?
@Kerb
"Gerg,
I’m sorry, I have tried to explain this to you, but the questions you ask tell me that you have absolutely no understanding of maths or statistics at all. For example at #99 you asked:
And Kerb, how did you ever come up with a vaccination ratio that mirrors autistic cases. With a 1% autistic rate, 600 vaccinated kids would translate to 6 autism cases, and 400 unvaccinated cases would mean 4 autistic cases.
I’m still staggered that anyone could miss the whole point of this exercise so completely. At that age only 50% of the autistic children have been diagnosed – look at Table 1 in the NEJM study, which shows that most children are diagnosed between the ages of 3-5 years with some being diagnosed at over 6 years of age.
You then reworked the figures so that all autistic children were diagnosed at the age of one and then, mind-bogglingly, stated:
Yet, I mentioned another factor that may mitigate this effect, and that is the fact that autism cases are likely to go unnoticed in the younger cohorts.
This is the very factor that you just removed from my figures! I really don’t like impugning people’s intellects as I believe almost anyone can learn, but really, how can I possibly avoid concluding from this that you are an idiot?"
Perhaps, Kerb, you should answer the 'idiots'" question of whether the skewing effect could have not been accounted for to make accurate calculations about the rates, and without resorting to the dubious 'life free years' solution.
@Kerb
I see though how you are picking at 'tidbits' to avoid addressing the legitimate question that I posed in my last post.
Kerb re #110
You were the one that provided a hypothetical sample with consideration for the actual Danish study. You wanted to keep things simple and for some 'inexplicable' reason gave autism cases in ratio rather than at the 1% rate.
"without consideration....."
Ok, listen up drug pushers (sometimes if I didn't know better that you guys don't practice what you preach and vaccinate yourselves, or your kids, I'd think that you took one of Offit's 10,000 injections per sitting, and it has now turned you guys into the most evil, deceitful mutants).......
Anyway, being that the younger cohorts of 'too early to vaccinate and too early to show autism will always skew things towards an unfair over-representation of vaccinated autistic kids, we will not expect the same autism rate for the vax and unvax group. The raw data will always yield a measurable skew -- and I repeat 'measurable'-- of a lower unvax rate. Yet, we would expect the actual figures to fall within a certain 'skew range'. If they don't, with the unvax rate being significant lower and outside of the range then we know that there is indeed a problem, and a link.
Kerb,
Further, in my final analysis, the fact that autistic kids in the younger cohorts are expected to be missed should have no bearing on the results given the large Danish sample size. We would expect both vax and unvax autistic kids to have an equal chance of being missed. We also would not expect the different vaxed kids to be missed in the ratio way that you suggested, but instead respective of your proposed one percent rate. I supposed your ratio idea was an under-handed attempt to unfairly bump up the vax cases and reduce the unvax
ones and hence yielding an unrealistic skew to futher discredit the straight calculation approach. Nice try!
Greg:
Why would such a complication be necessary? Simply exclude the group.
Which, as we have pointed out to you would be overwhelmed by the 0-18 cohort. In addition, not every child was vaccinated at 18 months, and that would have a skewing effect back.
Are you deliberately being dense? The larger the overall sample of 0-7 years, the larger the 0-18 month cohort that is both unvaccinated and undiagnosed. Do you not understand the concept of ratios?
You supposed incorrectly.
Nothing needs to be done to discredit the "straight calculation" a.k.a. the "do not attempt to correct for even the most obvious confounding factors such as age" approach. If it wasn't already discredited by the simple fact that it's wrong, it was already discredited by the observation that it would make shoe size, weight in pounds, whether or not someone had played "got your nose" with the baby, and every other factor that tends to increase between birth and the seventh year appear, erroneously, to have a connection with autism.
Gerg,
Person-years are in no way "dubious". Why should researchers have to abandon a well-established and useful statistical tool just because some people are too lazy or too stupid to learn enough math and statistics to understand it?
I haven't seen a "legitimate question" from you yet, just stupid questions that betray just how clueless you are on this subject. I have no problem with someone admitting they don't understand something and being willing to learn, but I despise a person who pretends they understand something when they don't, and attacks and mocks those that do in frustration at their inability to understand.
I didn't give them "in ratio",. I took into account the number of cases diagnosed at each age, as recorded in the actual Danish study in actual numbers in the actual tables. How can anyone be this stupid? The rest of your comments are more blather about results being skewed, when the study corrected for this perfectly adequately. There is no link between MMR and autism, if anything there were fewer cases in the vaccinated group, though the differences were not statisticallly significant.
In your "final analysis" you reveal that you have no understanding at all of the math and statistics involved in this, not a clue. What grade did you study math to? You are mathematicaly illiterate, which is fine as long as you are willing to learn, which you clearly are not.
Nonsense, the sample size will not affect this at all, apart from reducing randomness and making this effect even clearer.. Both vaccination status and autism diagnosis become more likely as the children get older, therefore there will be a correlation between them, regardless of sample size. I don't know how I can put this any more simply. I truly don't understand how you fail to grasp this.
That's true, which is exactly what I did in the table, I showed that about a third of the 1-2 year-olds, a third of the 2-3 year-olds and about a half of the 3-4 year-old autistic children either vaccinated or unvaccinated would have been diagnosed at that age, to approximately fit the actual figures in the Danish study.
You adjusted the figures to remove this factor, making all autistic children diagnosed by the age of one, which is clearly just wrong, and then complained that this same factor wasn't taken into account! You are an idiot, there is no other explanation.
Respective of my proposed one percent rate? I don't know what that means and I very much doubt you do either. It seems everyone else understand what I did perfectly well. How come only you are struggling with this?
No, it was a perfectly honest attempt to explain this to someone who clearly doesn't understand the math. You are not just an idiot, you are a dishonest ajnd disingenous idiot, pretending to understand something you clearly do not, and instead of making an effort to learn, you insult those who try to help you.
Back into the killfile with Gerg.
Oops.Never mind, it's still readable.
Incidentally, I asked Dr. Madsen if his raw data are available on-line anywhere, but he told me he is no longer involved in that area and sad no longer has access to the data.
Even if I did put the data into a form that Gerg could understand, written in crayon in words of one syllable, I have no doubt at all he wouldn't accept it. His mind is welded shut against anything that doesn't fit with his crackpot beliefs.
Whatever is the matter with me? I meant, "sadly no longer has access to the data".
Update: Crosby is scheduled to be interviewed by Curt Linderman today to spread more dirt about his former allies:
"Interview with Jake Crosby of Autism Investigated
Today’s interview will be eye opening for all of my listeners whether you are affected by autism or not. It is a story of corruption that is found in so many organizations. Jake Crosby is an investigative journalist and a respected member of the autism community who has in recent months been opening the eyes of many in the community to some corruption and shenanigans that I feel people need to know about. The vaccine/autism link is true…it’s real…it needs to be heard. This interview is about why the truth isn’t getting out. Join us at 2 PM EST at Natural News Radio.com and check out the links for this show at our LU FB page or at www.lindermanunleashed.com"
@ lilady:
I KNEW IT! ( And -btw- predicted it)
That Jake would eventually wind up somehow, someway @ NaturalNews...
Now if only he appeared @ PRN....
( he was linked to by those idiots already)
@Kerb
'Has no access to the data'? Kerb, should I place that in the same BS file as 'the cheque is in the mail' and 'the dog ate my homework'. Kerb, I am starting to think we were better off when you signed up to ignoring me.
@Orac
Hey Orac, whatever happened to Elburto? I am also missing ORD. ORD, there were some hard feelings between us last time. Come back and let's have a heart-to-heart.
Jesus Christ, I turn off the killfile out of ill-founded curiosity over the string of Gerglings, and the perceived world happily delivers up the usual consequences. Gotta do something about that sidebar widget.
As for the molding pumpkin's reversion to this bit of drooling idiocy when his latest attempt to pull a rabbit out a hat predictably led ever deepening embarrassment, I'm just bummed out that my doctor's office didn't have the tetravalent* influenza vaccine.
* Well, it is "heptavalent," isn't it?
@Kerb,
Hey Kerb, I just thought of something. Maybe Poul Thorsen's dog ate the missing Danish raw data!!
Narad,
I'm sorry, mea culpa. I honestly thought Gerg was teachable or reachable, but I was horribly wrong. I am saddened that not only does Gerg apparently have no capacity for understanding the math involved in this, which is a beautiful thing to my mind, but he clearly has no desire to do so. I can't bear to see someone mangle the truth so horribly, so my killfile is going back on again. Is there a recent script that works with Firefox? I'm browser-hopping at present and none of the scripts I have work with Greasemonkey.
I turned up late for my flu shot, and was jabbed by a somewhat (understandably) irritable nurse in a corridor (ooh matron!) on her way off duty, but it will hopefully work just as well.
Gerg,
My last words to you.
What's your excuse - the dog ate your brain cells? Learn some math, learn some statistics, look at Madsen's paper again, look again at at the AoA article and I guarantee you will feel very embarrassed indeed. You will also understand why I have labeled you an idiot.
But you're far too lazy to do that, aren't you? It's far easier to read some moronic BS that another idiot at AoA has cooked up and uncritically swallow it whole. You disgust me.
It is difficult to get a man to understand something, when his salary entertainment depends upon his not understanding it.
The one linked before is a Greasemonkey script. That was the bit with switching @include to @match for Chromium-based browsers.
Thanks Narad. I'd forgotten to change @include back to @match. I now have a wonderfully Gerg-free thread.
HDB,
@Kerb
You keep going on about me not understanding without being clear about what I am not understanding. So for the benefit of any lurkers here is my position. Please feel to respond to where I am wrong. First, in the Danish study we would expect that a straight calculation of the rates for vax/unvax groups will result in a skew with the unvax rate being less than the vax rate. The skew is the result of the bias of young kids that are not old enough to be vaccinated and also too young to show signs of autism. Second, the skew effect can be quantified -it can be measured. Third, we can use this measurement to determine if there is indeed a link or not.
I'm listening to today's Jake Crosby radio interview. I usually blow past the commercials, but the first one is selling a "squatty potty", to aid in evacuation of your bowels.
(The things I do for science)
http://radio.naturalnews.com/Archive-LindermanUnleashed.asp
For the benefit of lurkers, not only will we explain where you're wrong, we'll note how many times we have already explained it to you.
Except that what you are calling a "straight" calculation would be more accurately called "the completely wrong" calculation.
Yes, that is indeed why what you misleadingly call a "straight" calculation is actually a calculation nearly guaranteed to give wrong answers.
A moment of theory, for benefit of the lurkers: the ideal scientific experiment is one where you can compare subjects which are identical - except for one variable. This is called the independent variable. You then look to see whether there is a statistically significant difference in the dependent variable - the one that you've hypothesized would be affected by the independent variable. If there's a statistically significant correlation between the independent and the dependent variable, you've got some promising scientific evidence. What makes the evidence convincing is that the subjects are, as mentioned, identical in every way except for the independent variable - we've effectively ruled out any other plausible explanation for the dependent variable varying as it does.
Well, that's the ideal. We don't always get the luxury of working under ideal circumstances. In particular, any study on people runs into difficulties stemming from the simple fact that people are not identical. But we can still get useful information, if we're careful about how we analyze our data. We do what's called controlling for confounding factors; it means that we try to match up subjects in a way that minimizes any differences other than the independent variable that might affect the dependent variable, and then we compare subjects with the subjects that are a good match for them.
In this study, the most blatantly obvious possible confounding factor is age, because so much changes with age. Just take a moment to think of all the ways in which infants between 6-12 months differ from children of 6-7 years. Size, weight, likelihood to have a younger sibling, likelihood to have encountered a wild animal, likelihood to have met a grandparent, likelihood to have lost a grandparent - it's a pretty endless list, isn't it?
If we don't control for the confounding factor of age, comparing the children only to the children who are near them in age and not carelessly comparing infants of 6-12 months to children of 6-7 years, then the conclusions we draw from those comparisons are worthless. Remember that what makes the evidence from an ideal scientific study convincing is that we have ruled out all variables other than the independent variable we think is the key. If we are not ruling out those variables, then we have no justification for pretending we've proven anything about the independent variable.
And what Greg is proposing is that we throw out the window the correct analysis of the data, which did control for the variable of age, and instead substitute a completely wrong analysis where we just do things wrong from the get-go, and then try to fudge the figures to make up for it, and pretend that the results of all that flailing mean something.
Why, yes. We could measure the skew effect pretty precisely, by doing the analysis the correct way, controlling for age and other confounding factors. Then we'd compare that correct analysis to the dumb, pointless, time-wasting, completely incorrect calculation which does not control for confounders, and we'd get a pretty precise result for how stupidly wrong it is to bother doing things a way that we know to be incorrect.
Nope. About the only thing we could learn by doing the calculation the wrong way and measuring how wrong it is, is just what a colossal waste of time it was to do so.
from Linderman:
" Jake Crosby is an investigative journalist and a respected member of the autism community who has in recent months been opening the eyes of many in the community to some corruption and shenanigans...."
Fortunately ( for me) the MP3 refused to co-operate and I- alas- missed out on this enlightening, fact-filled interview.
That is not what an investigative journalist is.
Khani:
Amongst those I survey, calling yourself an investigative journalist is not at all uncommon- neither is calling yourself a researcher, scientist, humanitarian, scholar or gentleman.
Similarly, I can call myself the Queen of the Evil Faeries- doesn't mean that I am.
@Antaeus
"the ideal scientific experiment is one where you can compare subjects which are identical – except for one variable. This is called the independent variable. "
Just curious, Antaeus, if this is the case why did the research bother to include kids too young to be vaxed? Would it not have been better to simple consider only kids that had been vaxed? Even with the 2-3 age cohort we would expect a 90% vax rate. If we were to also consider the 3-4, 4-5,5-6, and 6-7 cohorts would the study not yield a fruitful sample size?
Second, are you saying that other than controlling for age using the 'person years' method there is no other way to measure the skew effect?
@Antaeus
Futher to my point of not including kids that are too young to be vaxed in the study, it would seem that this approach is the best one for controlling for age. The 'let-me-fool-them-by-calculating-autism-free-years-for-every-one-of-the-subjects-and-if-we-pull-a-fast-one-they-likely-will-never-be-able-to-tell-because-who-is-going-to-verify-all-that-tabulations-anyway-person years approach just seems like such a round-about way to do the obvious.
#140 One more reason to prefer "reporter."
Greg:
Ask them.
Not to the level of accuracy required. Any time you introduce a further calculation into statistical analysis, confidence decreases.
@Denice Walter, when I saw "a respected member of the autism community" for Jake, I rolled my eyes so hard I saw my brain.
@Julian
It really sounds like we are starting to speak each other language. Thank you for finally admitting that we could calculate the skew effect -- albeit it may be a issue of how accurate the result is.
@Antaeus
Reading your comment it seems to me that you would have no problem with the straight forward calculation if it was used to determine the rate for kids that all had a chance to be vaccinated and excluding the younger cohort that didn't have a chance. I quote you.....
------------------------------------------------------------------------------
"A moment of theory, for benefit of the lurkers: the ideal scientific experiment is one where you can compare subjects which are identical – except for one variable. This is called the independent variable. You then look to see whether there is a statistically significant difference in the dependent variable – the one that you’ve hypothesized would be affected by the independent variable. If there’s a statistically significant correlation between the independent and the dependent variable, you’ve got some promising scientific evidence. What makes the evidence convincing is that the subjects are, as mentioned, identical in every way except for the independent variable – we’ve effectively ruled out any other plausible explanation for the dependent variable varying as it does."
-------------------------------------------------------------------------------
So Antaeus, if we were to exclude the 0-1, and 1-2 age cohorts in the Danish study and do the straight- forward calculation for the rest of the sample, do you believe that this would give us an accurate result of whether there is a link or not?
Julian Frost,
It's worth noting that in Table 2 of the NEJM paper it shows that the study included 200,000 person-years of children vaccinated before the age of 14 months, with 38 cases of autism diagnosed in those children. That's one reason you can't just omit the younger group.
As a matter of fact the study looked at each age cohort separately, and found no link between MMR and autism in any of them. The vaccinated children actually had a lower risk of autism than the unvaccinated in 6 of the 8 age groups though this was not statistically significant in any of them.
@Kerb
There you go again not addressing the points directly. Are you saying ignore the younger cohorts of kids that are too young to be vaccinated would not have yielded meaningful results. I would think meaningful results are obtained from comparing apples to apples. Sounds like they threw in a whole lot of oranges. Also, where is the raw data so that we can confirm that there are no links from cohorts to cohorts.
This whole study got me thinking of something also: I understand the near sample size of 500,000 kids represent nearly 100 percent of the kids born over the 7 year period. Now you are saying that nearly half (2OO,OOO) were born for the O-1, 1-2 age cohorts? Kerb, how could that be? What about the 2-3,3-4,4-5,5-6 and 6-7 cohorts? It suggests a strange, abrupt decline in Danish fertility rates.
Your objection has exactly no relevance to the statement it supposedly responds to. I don't know why the authors chose to include the 6-12 month age group, nor do I know why they left out groups above 6-7 years. But the fact remains that the correct way to analyze the data you get, when you are studying children at different stages of early development, is to compare children with other children of the same age.
Since you didn't even know that, and since previously you've admitted lying and saying "I know there's a study design which would completely settle the issue, which mainstream science could only be objecting to because they're Trying To Hide The Truth" when in fact you had no such study design in mind, I don't see why any rational person would take your word for it when you say that a good study design would have excluded the youngest age groups entirely.
To my knowledge, there are vaccinations that are recommended for the first year, but MMR is not one of them. However, we all know that there are many parents who try to substitute their own judgment, or the judgment of fringe figures, for the recommendations of medical professionals. Suppose that upon examining the data for the 6-12 month children, the scientists had found that there was indeed a small contingent who were already vaccinated for MMR - and that that small contingent contained all the children in that age range in whom signs of autism were already being detected. You would be screaming if the authors then said "Ooops, we shouldn't have studied that group, so let's just pretend like we didn't collect that data." Yet here you are the one saying "Studying that age group didn't produce the result I wanted, so obviously it should have been excluded from the study."
Again, since children are being compared with other children in their age groups (something which you didn't even realize was important, O One Whose Expertise In Study Designs Varies Widely Depending On The Moment) then the chief effect of including that age group is introducing more possibility for the null hypothesis to be falsified.
If we didn't know you to be utterly shameless and dishonest, Greg, we might well scratch our heads wondering why you are objecting to the collection of more data that might potentially turn up the results you wanted. It's as if a Bigfoot "researcher" examined a study where 75,000 square feet of land were examined for any footprints that might indicate a previously unknown species - and criticized the study because some of those square feet were awfully rocky and thus were unlikely to hold footprints.
You mean, is there another right way to do it besides the right way it was done in the first place, which you don't like because doing it the right way doesn't provide anything useful to your prejudices? Nope, there isn't.
No, the best way to control for age is to control for age. You still don't understand what it even means to control for confounding factors, do you? If you were that Bigfoot researcher, I bet you'd say that excluding the rockier stretches of land was "controlling for terrain", wouldn't you?
You have no idea how science is actually done, and furthermore, you don't want to have one; you just want to throw around terms you don't understand in the hopes that it will fool rubes even less literate than you into thinking your complaints come from something other than just your prejudices.
See? Now you're switching to a completely different accusation in mid-sentence, that some sort of deliberate error was introduced in the process of tabulating the data. You don't understand or even try to understand the terms you throw around, you just commit slander against those who are trying to get real answers from real research. What a repulsive specimen you are, even among antivaxers.
There is no logical way to get that conclusion out of anything I said. I have said, over and over again, that what you are calling "the straight forward calculation" is a completely incorrect way to analyze the data. I have no idea why anyone would believe from anything I've writtenthat I'd suddenly be down with applying that completely flawed process to the data just because the range of ages that are not being properly controlled for is somewhat reduced. Furthermore, since you are an admitted liar, I don't even believe that you believe it.
An analogy, for any lurkers: If you want to weigh a quantity of liquid on a scale without a "tare" button, the easiest way is to take an empty container, put it on the scale, and take the reading. Then pour the liquid you're measuring into the container, and read the scale again. Then subtract the first reading from the second. The container has a weight of its own, but by measuring the difference between the weight of the empty container and the container with the liquid, you come out with the weight of just the liquid.
What Greg is proposing is similar to saying "Well, we measured the weight of the liquid the correct way, by cancelling out the container's weight from our two measurements. However, I didn't get the result I wanted from measuring it that way, so why don't we instead completely throw out those results? Instead we'll take the weight of the container and the liquid together, and we'll do all sorts of complicated calculations where we try to figure out how much of the volume of the liquid plus the container belongs to each of the two, and then we try to multiply those figures by reference values for the densities for the material we think the container might be made of and what we think the liquid is. And then we'll get a figure for the weight of the liquid which might be more convenient to my purposes than the correct result that we got through doing things the correct way."
@Antaeus
Reading your phony tirade 'Freddo' Antaeus what I take from it is an implicit concession that not including the younger kids not at vax age would have avoided the confounding age problem, but you don't know why they didn't. Also, if this was done there would have been no problem with the straight calculation approach.
There is no "would it not have been better" way to analyze the data which will generate the outcome you'd clearly prefer, greg.
There's a right way to do it (which Antaeus and Julian have explained to you multiple times now), and there's everything else.
@Antaeus
Reading your comment it seems to me that you would have no problem with the straight forward calculation if it was used to determine the rate for kids that all had a chance to be vaccinated and excluding the younger cohort that didn’t have a chance. I quote you…..
[completely unrelated quote snipped]
So Antaeus, if we were to exclude the 0-1, and 1-2 age cohorts in the Danish study and do the straight- forward calculation for the rest of the sample, do you believe that this would give us an accurate result of whether there is a link or not?
There is no logical way to get that conclusion out of anything I said. I have said, over and over again, that what you are calling “the straight forward calculation” is a completely incorrect way to analyze the data.
----------------------------------------------------------------------------------
It blows my mind what you are saying. So are you actually saying even had we excluded the 0-1, 1-2 age cohorts we would still have to use the 'person years' approach to control for age even though there is noting to control for because age is no longer a confounder, and the straight-calculation approach would give us an accurate answer. Are you saying we should always use the 'person years' approach because it is the correct way of doing things, regardless of whether it is necessary?
I'm saying that if you think that there is no difference between the average child in the 2-3 age bracket and the average child in the 6-7 age bracket, you're a complete fool - no surprise there.
If you agree, as you would if you were honest and competent, that children in the third year of life differ in many ways from children in the fourth year who differ from children in the fifth year who differ from children in the sixth year, then the answer is obvious: Yes, we do control for age, because the notion that "age is no longer a confounder" is an idiotic falsehood.
This is a keeper:
^ Ah, I failed to divide that properly. I was originally going to go for a "What we say to dogs"/"What they hear" scheme, but Krebiozen's half wasn't actually addressed to the Quiz Kid.
Narad,
It was not, and my killfile remains strictly in place, to maintain my emotional equilibrium. However, I appreciate you drawimg my attention to Gerg's profound idea. How does the fact that the study included 200,000 person-years of children vaccinated before the age of 14 months in any way imply "a strange, abrupt decline in Danish fertility rates"? It's beyond me.
Ok Drug Pushers...
Let me play a little by your game. Work with Greg the 'horrid', 'incompetent' troll here. First, what exactly is a 'person-year'? Second, if I were to calculate for a age group the autism rate for the vax and unvax groups using the straight-forward method would this be a legitimate calculation. Third, let's say I exclude the 'too early to vaccinate kids' and calculate the overall autism rate for the vax and unvax groups and observe that they are out of whack and points to a link, yet you guys report that 'person-year' method reports no such link, what would be the process involved for 'regular Joe', some who may even me as 'incompetent' as me, to obtain the vax rates and compare for himself? Finally, what is the likelihood that there will be a discrepancy between the vax rates calculated by the straight-forward method and the 'person-year' method?
Just out of curiousity, why is it always the MMR that gets blamed? The DTAP (or whatever it's being called now) is administered at a younger age..and yet no one ever seems to question it. Even the denizens of AOA only mention the first diptheria vaccine. Even then, they don't fetishize it like they do the measles.
MMR was the scapegoat of Wakefield but for DTAP, its non-problem was discovered much sooner and in the process, the vaccine court was created after many trials of the DTAP so a new improved vaccine was made in the process.
Alain
Alain: I'm not sure it's improved, since the DTAP in current use seems to have a much shorter period of effectiveness. I'd prefer going longer between jabs to a slightly higher risk of side effects.
Greg:
That whooshing sound is the noise of the point going over your head at Mach 2. I never denied that it was possible to roughly calculate the skew effect. My points were: 1) doing it would lower accuracy; and 2) why do it when excluding the group is simpler and leads to greater accuracy.
I don't think we're speaking the same language at all.
"I never denied that it was possible to roughly calculate the skew effect. My points were: 1) doing it would lower accuracy; and 2) why do it when excluding the group is simpler and leads to greater accuracy.
I don’t think we’re speaking the same language at all."
Guys, I am thinking that you should suggest that Orac moderates all of Julian's comment before they get posted. Being that he is autistic, his honest predisposition may result in him giving away the company's classified secrets.
@PGP
I agree with you that AoA and other anti-vaxxers focus too much on MMR and thimerosal. Taking an investigative approach and trying to unravel the relationship between vaccines and autism points to vaccines in a cumulative way causing autism rather than a specific ingredient of vaccines or one vaccine. We see how MMR has been around for a while and if it is such a big culprit, surely we should have had more autism cases in the past. We also see how removing mercury from vaccines did not result in less autism. We see how autism cases abruptly spiked with the abrupt increase in vaccines in the 90s. All these things point to vaccines in their totality causing autism. Yet, Freddo --ahemmm Antaeus!--said something interesting that got me thinking: To the extent that the Danish MMR study amounts to the examination of one less vaccine there should have been a link.
@Antaeus
Continuing to entertain the 'incompetent' troll please point out the flaws in this thinking: Measuring rates across age to age using straight-forward calculation (what you dubbed the 'wrong approach') should still give us accurate findings. Measuring autism rates for the entire sample, irrespective of age, using this method may not result in accurate results. For instance, we know that autism is likely to go missed in the younger age groups. If a disproportionate amount of the sample contained older vaccinated autistics then the results will be skewed. Given the unlikelihood that birth rates will follow this pattern the straight forward calculation may still carry a fair degree of accuracy -- albeit not as accurate as controlling for age. Also, in the event of such a skew we could also obtain the overall rates by averaging the rates from the age-to-age calculation. Julian, however, did mention that ever extra calculation reduces accuracy. Still, we would have a figure to work with.
Greg: Taking an investigative approach and trying to unravel the relationship between vaccines and autism points to vaccines in a cumulative way causing autism rather than a specific ingredient of vaccines or one vaccine.
I'm sorry, but I just collapsed in laughter. Investigative approach? Bubby, you and your friends wouldn't know an 'investigative approach' if it bit you in the butt. Your crowd whines perpetually because they can't face up to the truth and they can't cope with any child that's less than perfect.
Basically any 'investigative approach' is your crowd whining, someone doing a study to shut them up, and then they whine some more because the study isn't perfect.
They won't be satisfied until every vaccine is gone- and then they'll be whining that their homeopathic remedies didn't work. Or they'll be partying over the grave of their dead children.
Greg, do you really want to attend a two-year-old's funeral? Or that of a baby who didn't make it past their first year? 'Cause that's the kind of world we had in the old days- and don't say it was a world without autism, because it wasn't.
Personally, I may not have any kids, but I think I'd rather live in a world where my neighbor's two daughters and my friend's kids have a 95% chance of making it to age 5. Nothing is guaranteed, but I think the modern world is a lot safer than a world without vaccines would be.
No, I don't think so. Instead, I will just repeat the Lurker Challenge: anyone other than Greg who think Greg's ramblings have some glimmer of a point that should be addressed should feel free to ask about that point, and I, for one, will consider myself honor-bound to answer, and I will encourage others to answer as well.
Of course, the Lurker Challenge has been open for months now, and no one has taken it up, which tells me that even Gerg's fellow antivaxxers realize that his blatherings are crap from beginning to end.
Ok VCADODers, it's time to wrap things up. Let me present this summary of things for the benefit of any interested lurker:
I came requesting info pertaining to 'oddities' about the MMR findings that were pointed out by others. It was then reported to me that such peculiarities extend from researchers for some 'inexplicable' reason introducing a confounding group to the study that should have never been included. Don't believe it was counter productive to do so -- I quote Mr Feldar....
" the ideal scientific experiment is one where you can compare subjects which are identical – except for one variable."
Still don't think it was counter productive to include this group -- I quote Julian.......
"I never denied that it was possible to roughly calculate the skew effect. My points were: 1) doing it would lower accuracy; and 2) why do it when excluding the group is simpler and leads to greater accuracy."
Anyway folks, introducing this confounding group to the study required that researchers utilize an arduous, cumbersome calculation method (dubbed the 'correct method') to nullify the confounding influence of the group and obtain accurate results. Had this group never been included, simple, straight-forward calculations would have sufficed.
Fortunately, however, it is still possible to exclude the confounding group and perform such straight-forward calculations on the rest of the data to see if such calculations 'jive' with the results of the 'correct method'. It's possible to do such calculations, but now we are told that the data is no longer accessible. (Yeah!)
@PGP
I supposed this is the third time that you have come with your admonishment that I will paraphrase as, 'so what if vaccines cause autism Greg -- do you what to go back to the bad, terrible days when everyone was dying in droves from infectious diseases?"
PGP, first let me make it perfectly clear that I sincerely believe that vaccines had a minimal role in overcoming those days. I also believe that vaccines are responsible for the exploding, serious chronic diseases (autism, asthma, adhd, allergies, diabetes, cancers, and others) that are ravaging our kids today. Frankly PGP, by strict risk assessment measures, vaccines just are not worth it.
That said PGP, even if vaccines did play a substantial role in eliminating past infectious diseases (and I don't think they did) should a parent not be given the choice to decide whether their child risk those diseases or the aforementioned current chronic ones? Who gives you, health officials, government, or pharmaceutical industries the right to make that choice for the parent?
I sincerely wonder if Greg understands how stupid this statement makes him seem for people who actually comprehend basic statistics.
Adam G: that's not the only thing he doesn't understand.
Okay, Greg, I will say this again: ADHD IS NOT SEVERE, IS NOT CAUSED BY VACCINESAND RUNS IN FAMILIES.
I'm going to give you the for-dummies rundown on all those disorders, since it's clear you don't understand any of them.
Asthma is caused by chronic lung irritation or by keeping a living environment too sterile. Also not caused by vaccines.
Cancer: Basically an overgrowth of cells that becomes malignant. Tendencies to certain kinds of cancer often are genetically linked, and cancer risk is also linked to radiation. Not linked to vaccines ever.
Diabetes: An inability to process sugars. Type 1 is often inherited and presents in childhood, usually in older children. Type 2 is found in adulthood; linked to poor diet and possibly race. Poorly understood by some people, especially you and Paula Deen.
Allergies: Sometimes caused by environment being too sterile.
Now let me explain why the government has an interest in keeping citizens healthy. The Spanish Flu shut down cities for weeks. Quarantines brought economic activity to a standstill, and workers couldn't work due to illnesses or being unable to enter the city they worked in.
England, between World War 1 and the Spanish flu, lost most of the young men of a generation, and was therefore deprived of any economic contribution those men (and the children of women who died in pregnancy) might've made.
Finally, vaccines free up half of the workforce. Women who might be stuck at home with sick kids no longer have to spend weeks or months at home. (Yes, guys might look after sick kids, but it's statistically and historically unlikely.)
Besides that, there's the obvious benefit of kids being healthy and not needing taxpayer money for a seeing eye dog or hearing aids because they had a bad case of measles or mumps. (Not that seeing eye dogs or hearing aids are bad things, mind you.)
Speaking of, you never answered my question- would you be willing to care for a kid who was disabled from measles or mumps? Would you still love that child? I'm just asking, since your crowd seems to deal in conditional love.
@ paging Denice Walter,
Upon reading Greg's #166 post, it dawned on me that this post is a shining example of stupidity and I'd be perfectly happy to point it out but then, there might be some value to have some explanation of why Science was once very respected and recognized as very valuable and nowaday, peoples of all kinds give the big middle finger to Science and feel empowered as a result. I'm not sure that rampant stupidity account for all of the effect and there might be some underlying factors.
Alain
Geg, you're so uninformed or deliberately deceitful when you piss and moan about "forcing vaccinations". In many states, parents can opt out of vaccinating and that includes mine, in Arizona. But we are now seeing the spoiled fruits of those choices, namely record outbreaks of pertussis and measles directly correlating with where parentally-elected vaccination opt outs are highest. The irony is that these outbreaks you stupidly think aren't related to nonvaccinationy are going to come back to bite you and your "cause". This will re-illustrate pwhat any rational person knows: (1) vaccines are safe, (2) vaccines are effective, (3) the diseased that vaccines prevent are dangerous and fatal, and (4) not vaccinating is dangerous.
It if my fervent wish that when this happens, parents who do vaccinate rise up to publicly proclaim and shame you for the liars and fools you are.
So instead of answering what I said, Greg decides to patronise me.
If instead of answering my points, you decide to insult me, you lose.
No, you were told that MADSEN could no longer get the data. And why do you find this surprising? As a software test analyst I routinely get sent to client sites for months (even years) at a time. When I arrive, I'm given access to their offices and IT network. When my contract expires, my accesses are revoked. It's basic security. That you weave a conspiracy out of Madsen no longer having access to the data is laughable.
For benefit of the lurkers, I will explain: Greg is completely full of sh!t and lying through his teeth.
Anyone can look back over the comments to this very post and see how we've explained to Greg over and over that because the 6-12 month olds were only compared to other 6-12 month olds, that they are not a confounding group.
They would only be a confounding group if we did the math the wrong way, which is of course the way that Greg wants us to do it.
What he keeps calling "the straight forward calculation", even if it excluded the two youngest age groups, would still compare children in their third year of life to children in their sixth year of life, and then attribute any differences detected to vaccination status, rather than, y'know, three of the most crucial years of child development.
Yes, I said that. I also explained that since life doesn't always give us ideal conditions, we do what we can to bring our studies closer to that ideal. One of those steps is called "controlling for age".
And notice, lurkers, that Greg is the one who keeps saying that we should skip that step, deliberately introduce a massive error into our data, and then somehow "measure" the effect of the error and compensate for it.
There is absolutely no reason to prefer Greg's route of "deliberately make a massive error, then try to calculate how big an error you made and adjust your data to try and compensate" to the much simpler "avoid the error in the first place".
By "excluding the group" Julian means excluding it from being jumbled in together with all other age groups, which is what Greg keeps saying he wants us to do. That is in fact how the study was done, by comparing the children in each age group only with children from the same age group, excluding each age group from the others.
The effect of including the 6-12 month age group in the study at all, contrary to what you would think from reading Greg's lies, was to give more opportunity for the phenomenon he wants to see to show up. And he was in favor of including that age group, until he found out it didn't produce the results he wanted.
It's not a confounding group. Why was it not a confounding group? Because we didn't do things the incorrect way that Greg is still suggesting we should have done.
Except that Greg has absolutely no understanding of what the correct calculation method is, and if asked to explain what the steps of it are, and where the "arduous, cumbersome" part of the process is, he'd be reduced to wild bluffing and deliberate, desperate lies, because he has no idea.
Once again, the confounding influence of the group is not there, precisely because we didn't calculate using Greg's more complicated and less accurate method.
And as anyone can verify for themselves by reading the comments of this post, when Greg came in, he was claiming that what "would have sufficed" would be the "simple, straight-forward" (and wrong) calculations" with that group included.
Yes, that's right; Greg didn't understand why you can't compare children of 6 months to children of 6 years and claim "well, the only way these children aren't alike is vaccination, so obviously vaccination's to blame for any differences in health outcomes!"
Then, when he thought that including the 6-12 month old age group came out with a result that matched his prejudices, he was all for including that group. Now, however, even though he knows that that age group did include vaccination events and did include autism events and therefore was a legitimate group to study, now he claims that including it was "a fast one" on the part of the researchers. What chutzpah!
Lurkers, I repeat to you what I've been saying for months now: if you have any perception that anything Greg is saying might be sound, please voice your concerns, and I will personally do my best to answer them. But if you read Greg's comments on this post, seeing him among other feats of dishonesty:
- change his position completely on whether the 6-12 month age group should have been included, and
- repeatedly refer to a "confounding effect" from that age group which would only exist if we used Greg's completely wrong methods,
then I can't blame you if you decide it's just not worth the effort to try and extract any putative value from Greg's tripe.
A couple of encouraging comments from a lurker and from Denice on another thread have led me to disengage the killfile, temporarily. I see that Greg has asked some answerable questions which I'm going to answer. I'm going to break this into a couple of comments.
Taking a deep breath and attempting to remain calm; once more unto the breach, dear lurkers, once more:
It is one year of one person's life. If you studied 100 people for one year, that is 100 person-years. If one of them contracted Incoherent Gurgling Syndrome, then the incidence of IGS in that group was 1 per 100 person years, or 0.01 per person year. You could also look at one person for a 100 years, which would also be 100 person years, or 4 people for 25 years. This is just a way of making different numbers of people studied for different lengths of time comparable.
In practise the things studied in epidemiology are usually rarer, so the unit most often used for expressing incidence is cases per 100,000 person-years as, for example, in this study looking at the incidence of lung cancer in those who have never smoked:
Incidentally, in smokers the incidence of lung cancer is more like 100 per 100,000 person-years or more. You see how this is a useful measure of incidence, and how causality affects the incidence measured in person-years?
You mean without correcting for age, which is a known confounder? No, it would not be a legitimate calculation. The age confounding problem isn't just at early ages; it is just much more marked then, as there will be more unvaccinated and as yet undiagnosed autistic children at that age. Age is a confounder at all ages.
If you look at the NEJM study it included 126,749 person-years (6% of the total in the study) of children vaccinated at age 20 months and older, with 16,572 person-years of children vaccinated over the age of 3, and so the problem is smaller in older children, but still significant: the older the child, the more likely it will have had MMR, and we know that some autistic children don't get a diagnosis until they are much older, so the older the child the more likely an autism diagnosis.
Even if you just looked at 6-7 year-olds, the age confounding problem would still be there, as the 6 year-olds are still a little less likely to have been vaccinated with MMR, and a little less likely to have been diagnosed as autistic than the 7 year-olds.
There are legitimate ways for correcting for confounders, Madsen et al used them properly. Where's the problem?
The other problem you would have with this is that your numbers would be too small to have statistical significance. There were only 53 children who were unvaccinated and autistic in total, and though the published figures dont show this, most of them will be in the younger age groups, since by age 6 in Denmark over that period about 90% of children had had their first MMR shot (according to the European CDC). In the NEJM study 82% of children were vaccinated, but this figure is lower because it includes a number of children who would have been vaccinated after the period of the study. This is why Madsen et al chose to use the data they did.
Antivaxxers claim that MMR is a major cause of autism, that children regress into autism within hours or days of being given MMR. If this were true we would expect to see it clearly reflected in this study. There would be children who were autistic who would not have been had they not had the MMR, who would not contribute 'vaccinated non-autistic' person years to the study..
Here are two hypothetical scenarios that are intended to demonstrate my point. I'm not saying that this is what happened in reality, this is a thought experiment, for educational purposes.
In the first scenario MMR causes autism. We take two hypothetical children, both vulnerable to having autism caused by MMR. Child A is vaccinated at age 15 months, immediately develops autism and is no longer followed in the study. His presence in the study contributes 1.25 'unvaccinated non-autistic' person-years. Child B is not vaccinated until he is 2 years old; he immediately develops autism and is removed from follow-up. His contribution to the study is 2 'unvaccinated non-autistic' person-years. Together they add 3.25 'unvaccinated non-autistic' person-years to the study. Neither adds any 'vaccinated non-autistic' person-years to the study.
In the second scenario (what I believe is the real world), MMR does not cause autism. We take the same hypothetical children who are not vulnerable to MMR, because no one is. Child A is vaccinated at 15 months but never develops autism. He is followed up for a further 2 years until the end of the study. He contributes 1.25 'unvaccinated non-autistic' person-years and 2 'vaccinated non-autistic' person-years to the study. Child B is vaccinated at 2 years, and never develops autism. He is followed up for a further year. He contributes 2 'unvaccinated non-autistic' person-years and 1 'vaccinated non-autistic' person-year to the study.
In summary:
In scenario 1 in which MMR causes autism, we have 3.25 'unvaccinated non-autistic' person-years and zero 'vaccinated non-autistic' person-years.
In scenario 2 in which MMR does not cause autism, we have 3.25 'unvaccinated non-autistic' person-years but we also have 3 'vaccinated non-autistic' person-years.
Imagine there are many other children in the study, most of who will be identical in each scenario, so we can ignore them, and a number of children whose contributions will be similar to Child A and B in my thought experiment.
There is a very obvious difference in the number of 'vaccinated no autism' person-years between the two scenarios. If vaccines caused autism we would expect to see that difference, in terms of a higher than predicted ratio between 'unvaccinated no autism' person-years and 'vaccinated no autism' person-years. We don't see this, therefore there is no reason to think there is any link between MMR and autism.
@ Alain:
(answering your page)
I think that that might be more of a question for historians than for me. However, I can dig up an analogy/precedent that might have some value:
our fabled, extremely posh historian of science/psychology taught that after WW1, many people ( esp in Europe) no longer put their faith in science after having seen the destruction modernity had wrought in the form of mechanised warfare- planes, tanks,bombs, machine guns and chemical weapons- this period was called the *Krisis des Wissenschaft* (sp?) as doubts increased, theorists like Freud even postulated a death instinct. A few tried to figure out how psychology could address human problems in the face of these horrors. It didn't really work.
More recently - since the 1960s- western cultures have endured a series of social upheavals, re-evaluations and counter-revolutions- most of which incorporate the underlying theme that authories aren't to be trusted. An additional idea focuses upon the value of individual expression above all else - then there is post-modernism- but I don't want to get into that pail of fish.
Suffice it to say that the achievements of science/ technology are counter-weighted by another trend towards Nature / Simplicity which is often nothing more than nostalgia for a largely fictional paradise lost. This is frequently encountered in pop culture and material written by non-scientists, altho' it might use data from science( re pollution, AGW, loss of habitat, genetics etc)
Which gets us to my area:
ideas like these are used by entrepreneurs to sell their goods and expound upon what has become trendy- thus automatically increasing the liklihood of consumers 'buying' ( in both senses of the term) what they have to offer.
Alt media I survey- almost in split brain fashion- castigates the efforts of SBM and simultaneously self-congratulates its own 'science' ( usually, pseudo-science). The "Back to Nature" trends which have blossomed since the 1960s enable them to address a ready audience as well as the anti-authority, anti-expert, self- expressionists- so we witness woo-meisters trying to be a kinder,gentler*spiritual* sort of scientist- not the Big, Mean Scientist which is largely their invention with some help from pop cultural memes and bad movies. AND if you scan alt media, you'll see how often their ideas do reference movies instead of research in social science- whch does -btw- exist and address many of the ideas we're talking about- tons of this stuff over the past 30 years .
This anti-science/pseudo-science woo-topia is a place where everyone can be a neurophysiologist or quantum physicist-cognitive psychologist ( paging Teresa Conrick and Mike Adams, respectively) without much ado or hard work-just by saying so or become an MPH without once leaving the safety of your own *idee fixee* homeschool, complete with enablers and admirers aplenty.
OBVIOUSLY doing so makes them feel better about themselves.
I know I said a couple of comments, but I lied. Last part.:
Excluding this group will not eliminate the problem, as I explained above. What you are describing is analyzing data and ignoring a known confounder which we know will lead to an incorrect result.
Dr. Madsen told me that the Danish Epidemiology Science Center hold the data, so you could write to them and politely ask for it. Then you could take a course in statistics and epidemiology so you would understand what to do with it, which might take you a year. Maybe longer as I'm pretty sure you would require some basic math study first. There are no short-cuts for understanding this stuff; it's complicated.
You wouldn't ask for the "regular Joe" way of setting up a particle accelerator, given in simple terms because you didn't understand the physics involved, would you? Perhaps you would.
I understand a bit about this kind of statistics because I had to study it to qualify as a biomedical scientist, because I also spent a year studying statistics as part of my social anthropology degree, and I have had cause to use my knowledge a lot over the years, analyzing large amounts of patients data in a number of areas. Even then I struggle with some of the more complex epidemiological stuff, and some of the more advanced statistical analyses used. This isn't simple, and sometimes it is counterintuitive e.g as in this case, what looks at first glance as if it should be simple isn't at all.
It is absolutely certain. We know there is a correlation between age and MMR vaccination status, and we know that there is a correlation between age and autism diagnosis. We know why these correlations occur i.e. because not all parents get their children vaccinated on schedule, and because autism becomes more obvious as a child gets older. These are two separate reasons, and they will inevitably lead to a correlation between vaccination status and autism. There is no doubt at all about this. If you do not take this into account in your calculations i.e. you use what you call "the straight-forward method", you will get the wrong answer, again without any doubt at all.
To add to what Kreb said:
when you are required to study statistics in order to qualify for a degree/s you deal with basic concepts which are then expanded over several courses and related methods/ problems courses- when I listen to some hack talk about how "badly" research is done whilst not even getting the lingo straight I can't help but laugh.
In social sciences, we know that you can very rarely study anything in isolation- thus it is MULTIvariate, partial correlations, controlling for this or that, factors, paths et al.
I was very well-prepared because my secondary school/s - and undergrad work- gave me a good background ( despite the liberal arts) BUT most students despised and feared statistics and I was often consulted on 'how to survive' this course or set of courses. .Grown men and women -often in tears over the arcane, mystical black arts of stat. I was much in demand and assisted the prof.
Let me just say: all of the objections the woo-bent dream up have already been addresssed by researchers YEARS ago- woo-meisters just never really studied the field- so how would they know?
@Denise - Greg has already admitted that he's made up his mind & no amount of evidence showing that his views are incorrect will sway him (he believes that eye-witness testimony trumps real Science - when we know that isn't the case).....it is no surprising that he knows nothing about statistics or how to evaluate a study beyond what his handlers at AoA tell him....
The temptation was just too strong so I am back for just this comment: The way I see it this issue of whether the 'correct method' was the best way to analyze the results, or simply a scheme to obfuscate the results is best resolved by directly analyzing the data. Let the reader decide for his/herself. Guys, again, I humbly request that you make the data available to everyone. (Gone again!)
@ Lawrence:
Agreed. I am reminded of the proverbial dining room table.
At any rate, I do wonder however about how those who are forced against their will to study statistics ( various PhDs who support anti-vax as well as he-who-shall-remain-nameless, MPH) do manage to keep their cluelessness intact despite the rationality of their studies.
Gerg,
What, you run away without addressing my answers to your questions? How disappointing.
By directly analyzing it how, precisely? Using established statistical tools, like Madsen et al. did? Or by using "the straight-forward method", which is guaranteed to give the wrong answer?
How would the reader do that, precisely? Are there any readers following this that are undecided about this? Would any undecided readers care to make their presence known at this point? Or anyone else convinced that "the straight-forward method" would yield anything resembling the truth?
What data? The raw data that you clearly wouldn't have a clue what to do with? I told you who has it, and even suggested how to go about getting it.
If you are so concerned about the matter, and still don't believe Madsenet al.'s results, despite our explanations, prove us wrong. Get hold of the data, figure out a way of controlling for the possible confounding factors, such as age, calendar period, sex, birth weight, gestational age, mother's education, and socioeconomic status, and see what results you get.
You could post your results here in a comment if AoA refuse to publish them, as I have little doubt they would, since if you do the math properly, like Madsen et al. did, you will also find that there is no link between MMR and autism.
In a fit of nostalgia I was fondly remembering some of Gerg's early visits to this blog. Like the times he claimed that "tens of thousands" or parents had reported "their child dramatically regressed into autism following vaccination".
What does the study under discussion show in this regard? Well, they didn't record the number of cases that were diagnosed within hours or days of MMR vaccination, but they did record the number diagnosed within 6 months of MMR vaccination.
Of the 738 cases of autism and other ASDs recorded in the study, only 11 were diagnosed within 6 months of being vaccinated with MMR: 3 autism cases and 8 other ASD cases. Compare that to the 53 cases of autism and 77 cases of other ASDs diagnosed in children who had not been vaccinated with MMR at all.
It makes me wonder why the likes of Gerg continue with their gurgling about MMR, it does.
@Kreb - I always wondered, if the anti-vax folks have "tens of thousands" of people who claim the same thing, why don't they fund a study of them?
Better yet, how do they explain, when real studies are done to examine the issue - whether here or elsewhere, by every type of research group (government, private, educational, etc.) these individuals are nowhere to be seen....
the best way to analyze the results, or simply a scheme to obfuscate the results is best resolved by directly analyzing the data.
Is there some reason why the lazy little numpty doesn't contact Madsen to access the data and perform this analysis himself? Rather than running around like a rat through a sewer-pipe and popping up at blogs to demand that other people do it for him? Oh yes, I forgot, it's Greg.
Ok Kerb,
You tempted me back in. You guys should know by now that I am utterly 'shameless' so let me parade my 'ignorance' again if it will help atleast other 'idiots' to understand. Explain to me Kerb what is so statistically advanced about this study. Explain to me step-by-step where my thinking is going wrong. You take the data and you separate it in six age brackets -- 2,3,4,5,6,7 (why they didn't consider older kids I will never know). You don't include any kids younger than 2, and hence everyone had a chance to get vaxed. You then obtain the autism rates for the vax/unvax (using straight-forward calculations) for each age group. Finally, you obtain the autism rates for the entire sample. Maybe the overall autism rate for the vax group may be hire than the unvax one due to the sheer chance that autism was detected at a disproportionate rate in the younger age bracket. Still, you have the age-to-age rates to make sense of the overall rates.
Greg,
I have tried my best to do so all along, but you still don't appear to get it.
Perhaps that's because the antivaxxers were claiming at that time that, "tens of thousands" of parents had reported that "their child dramatically regressed into autism following vaccination". Why would you look at people decades after they had allegedly had such a dramatic reaction to vaccination? Surely it's best to see if you can pick up a difference in the months and years following vaccination, isn't it?
There are also possible confounders that have to be taken into account if you look at older children who have not been vaccinated. For example, parents who don't get their children vaccinated may be less likely to take their child to the doctor for other reasons, so they may get fewer antibiotics, or have less exposure to contagious diseases in the doctor's waiting room, or they may be home-schooled and have less contact with other children, or more likely to live in a rural area and have less exposure to environmental chemicals, or may be less likely to be diagnosed with an ASD because they are not seen by teachers and doctors who would recognize the symptoms. There are all sorts of problems with this approach.
Everyone may have had a chance to get vaccinated at this age, but not everyone did. Some were vaccinated at an earlier age. A significant minority of children were vaccinated at age greater than 3 years. There's part of the confounding problem right there.
Without correcting for the confounding effect of age that I and others have tried to explain to you? You will get a correlation between vaccination and autism, even if there is no causative link at all. It will be a smaller correlation than you would see for the whole cohort, but it will still be there, invalidating your results.
As Antaeus pointed out, if you looked at children's shoe sizes and autism between the ages of 1 and 7, you would find a strong correlation between shoe size and diagnosis of autism, because both correlate with age. You wouldn't conclude that larger shoe sizes cause autism from this, unless you had some prejudice about shoe size and had developed a delusional conviction that there was some connection, of course.
It's not sheer chance, it is an absolute certainty that there will be younger children who have not yet been vaccinated but who will later be diagnosed with autism, and who will be missed in this study. The older children who are more likely to have been vaccinated will be more likely to have had their autistic symptoms noticed and to have been diagnosed. It is absolutely certain that there will be a correlation between age and vaccination status and between age and autism diagnosis, and therefore a correlation between vaccination status and autism diagnosis. This has nothing to do with chance.
The age-to-age rates will all suffer from the confounding effects of age. If you look at 3-4 year-olds, the youngest in this group are still less likely to have had the MMR than the oldest, and are also less likely to have been diagnosed with autism than the oldest. You can't escape this confounding problem by using even smaller groups, because the smaller the groups the less statistical power you will have.
If you look at 3-4 years olds' shoe size and autism, you would still find a correlation, because a number of children will be diagnosed during this period, and their feet will be growing during this period.
If you wait and look at the children who have never been vaccinated with MMR you have all the confounding effects, that I mentioned at the beginning of this comment, to contend with. You also have the problem that since about 90% of children are vaccinated, you won't have enough children to achieve statistical power to detect anything but a really strong effect of MMR on autism, even assuming you can somehow deal with those confounding factors, which I don't think you can.
In this study of about 500,000 children you will have (eventually, carrying on the study period longer than they actually did) about 50,000 unvaccinated children and about 70 cases of autism in these children, assuming there is no link between MMR and autism. Those aren't huge numbers to try to get statistically significant results out of. There are ways of figuring out what sort of differences you can detect with different sample sizes, but that's beyond me at this point on a Friday night.
Say vaccinated children are 10% more likely to become autistic than unvaccinated children. You would be looking at perhaps 63 unvaccinated autistic children instead of 70. Could that be the result of chance, or some confounding effects you hadn't thought of? Could 7 unvaccinated autistic children have escaped both vaccination and autism diagnosis because their parents were religious nuts who didn't believe in doctors or for some other reason?
I hope some people reading this see my point, though I somehow doubt you do.
Even in this study, using statistics to wring as much statistical power out of the data as possible, one of the commenters on the study pointed out that you can't exclude a possible effect of MMR on a small subset (10%) of children who have some genetic abnormality so that MMR causes autism in these children. This is mere speculation, as we have no reason to believe such a sensitive subset of children actually exists, but it demonstrates the limits of a study using even half a million children.
Any clearer? (Hope springs eternal...)
@Kerb
Still, you have the age-to-age rates to make sense of the overall rates.
The age-to-age rates will all suffer from the confounding effects of age. If you look at 3-4 year-olds, the youngest in this group are still less likely to have had the MMR than the oldest, and are also less likely to have been diagnosed with autism than the oldest. You can’t escape this confounding problem by using even smaller groups, because the smaller the groups the less statistical power you will have.
-------------------------------------------------------------------------------
Let me follow your thinking here: Let's, hypothetically, say we have a bunch of 3 yrs old -- 10. Six have been vaccinated and four of them have autism -- a rate of 66%. Four of them are unvaccinated and 1 has autism -- a rate of 25%. By what reason based on these rates can you say autism is possible not linked to vaccination in 3 yrs old?
Re #187
Please don't tell me that you are contemplating that all the 3yrs old that are vaccinated could possibly be born in December, and all the unvaccinated ones were born in January, and therefore since the vaccinated kids are older there is a greater chance that autism may be detected in them.
Greg,
I wonder if this Wikipedia article about confounding variables might help you.
In terms of the the NEJM study, age is the extraneous variable, diagnosis of autism is the dependent variable and MMR vaccination status is the independent variable. Failure to account for age as a confounding variable led to AoA finding a spurious relationship between MMR and autism, which is statistics is called omitted-variable bias.
You might also find this Wikipedia article on regression analysis useful.
Be sure to let us know how you get on analyzing the data.
@Kerb
Very well then Kerb, let's continue to take it step-by-step for the benefit of other 'idiots' out there. In our sample size of 10 with 6 vaccinated and 4 autistics, and 4 unvaccinated and 1 autistics more autistic in the vaxed group may have been detected by the sheer chance that they had birth-dates in the later months and were older. And, if the ratio between vaxed and unvaxed is greater -- say 9 to 1- we would expect the effect to be amplified. Given our hypothetical sample size then Kerb how would we control for age using 'person-years'?
Gerg,
I very much hope you are in a league of your own, but OK, I'll take you through this as best I can.
That doesn't make any sense at all. I cannot parse that as intelligible English. This is not a good start.
Where did this "sample size of 10" come from? It doesn't seem to bear any resemblance to any hypothetical examples I or anyone else has given, or to anything in reality. Are you talking about a sample of 10, with 6 vaccinated children diagnosed as autistic and 4 vaccinated children not diagnosed as autistic? If so, where do the "4 unvaccinated and 1 autistics" come from? Are you suggesting a group of 10 vaccinated children with a 60% autism rate? That makes no sense. Or an unvaccinated group of 5 with 25% autism?
Are you referring to the table I constructed above, with hypothetically only 50% of autistic children having been diagnosed at age 1-2, so out of a sample of 1,000 you might have 0.5% diagnosed with autism, evenly distributed between vaccinated and unvaccinated?
I need you to clarify what you are referring to here.
Nope, the rest of the sentence helps not at all. As far as I can tell this is gibberish. What have birth-dates in the later months to do with anything?
Grasping at what I think you might mean, I think you still miss my point - it has nothing to do with chance. The older the child the more likely he will have been vaccinated and the more likely that he has had MMR, irrespective of chance, and without any causation at all.
This is what we expect to see if there is no link between MMR and autism, it is what we see in practice, and it is what AoA leaped upon with glee thinking they had discovered that Madsen had covered up the truth, which would be quite funny if so many people didn't believe them.
Let's assume that what I believe is actually true i.e. that autism starts in the womb, and has nothing at all to do with vaccines. A child is born with autism and at some point his parents notice he is not developing as he should, and he gets a diagnosis of autism. The older he is, the more likely it is that someone will notice his developmental delay, and the more likely he will get a diagnosis of autism.
Let's look at the actual figures from Madsen's study.
There were 57 autistic children diagnosed at age less than 2, 218 diagnosed between 3 and 5, and 41 diagnosed at age greater than 6. Put another way, there were 57 diagnosed by the age of 2, 275 by the age of 5 and 316 by the age of 7. You see the increasing number of diagnosed autistics at each age? I know that there were 537,303 children in this study, so I can calculate the odds of being diagnosed at each age.
By age.Diagnosed......Odds of being diagnosed autistic 1 in x
2..........57..................9,426
5..........275................1,954
7..........316.................1,700
I have even plotted the first two columns as a graph, so you can see how this looks, here. As age increases, the likelihood of an autism diagnosis also increases.
Please note that this has nothing to do with MMR, this is what we would see in any population with a proportion of children with autism. More of them are diagnosed between age 1 and 2 than other ages (that's why the graph isn't a straight line) because this is when a number of milestones are expected, and so more autism is detected.
In a large enough group of children aged between 3 years and 4 years, there will be more 4 year-old children with diagnosed autism than 3-year-old children. An autism diagnosis doesn't got away, so the number of children diagnosed with autism as age increases is cumulative, it increases but never decreases.
What effect will be amplified? Again I have no idea what you are talking about and I seriously suspect you don't either.
At age zero no children are vaccinated, by age 7 about 90% of them are vaccinated. If I had the data I could plot it the same as the data for autism diagnoses and it would probably look very similar in shape, but I don't. The recommended age for the first MMR shot in Denmark is 15 months, so most children get it at around that age, a few get it before and some get it later.
Your hypothetical sample size of 10 isn't going to big enough for anything. I'm afraid I'll have to ignore your example figures because they make no sense to me. They reinforce my beliefs that you don't have the faintest clue about this.
You would have to use mutivariate analysis to calculate the contributions that each variable makes or doesn't make to the incidence of diagnosed cases of autism or ASD. Age and MMR vaccination status are the ones we are really interested in, but you should also exclude any confounding effects of other factors just as Madsen et al. did. Once you have established the confounding effect of age, you can correct the figures to remove this confounding effect, just as Madsen et al. did..
Madsen et al. state that:
There's an explanation of this statistical technique here, and you can look up PROC GENMOD and how it is used in a software suite called SAS, but frankly I don't think you have the education or the intellect to understand it.
@Kerb
I don't understand where you are confuse. We are dealing with a hypothetical scenario and it has no bearing on the Danish study. We have a sample of 10 children that are 3 yrs old. Six were vaccinated, and of that six four are autistic -- a rate of 66 percent. Four os them were not vaccinated, and of that four one is autistic -- a rate of 25 percent. Can we say based on these rates vaccination is linked to autism? If not, why?
Obviously these are not realistic rates for children from the general population - not even the least realistic AVer thinks that rates are up to 1 in 2 children.
But let's go along with the thought experiment. If we studied this sample of ten children and found autism occurring at these rates, would it be very strong evidence that vaccination was linked to autism? Let's give Greg's scenario the full benefit of the doubt and say that we could somehow eliminate other possible confounders, even beside age. Is our evidence strong then?
Unfortunately, no. Why? Because we can't eliminate the possibility of probability flukes, the way we could if our sample size was larger. We can illustrate this very simply. Let's say we take one vaccinated child with autism, Child A, and one unvaccinated child without autism, Child B. Child A's parents got him vaccinated in the current year; Child B's parents are going to get him vaccinated, but they won't go do it until the next year.
Now let's say that Child A and Child B's parents switch roles, in terms of which year they get their child vaccinated. Now there's one more vaccinated child without autism, and one more unvaccinated child with autism. Guess what? Now the rates of autism in the two groups are exactly equal. Exactly half the children in each sample group are autistic! Just because two children in the sample changed the scheduling of their vaccinations, our "evidence" completely disappears.
We talked before about "the law of large numbers". Well, one thing to remember about the law of large numbers ... is that small numbers are not subject to it. If we flipped a coin 10,000 times and it came up heads 60% of the time, it would be logical to draw from that a conclusion about whether the coin is fair or not, and what the true probability of it coming up heads is. But no one of sense would flip the coin ten times, see six heads, and say "This must be a rigged coin!"
Even in a hypothetical, thought-experiment world where rates of autism between 25% and 66% were actually plausible, the notion that those must be the actual rates of autism because we picked just ten children as a sample and saw those rates in our extremely small sample is a mistake an educated adult, even one who didn't go on to a career involving math or science, should know better than to make.
A lot of people don't understand clustering in statistics. Frankly, I didn't learn about it in a science class or in my stats class either, so I can see why people don't understand it.
Using the examples of coin flips is a really good way to explain it. Though I suspect it won't be understood/accepted here, good job, Antaeus, for trying! The rest of us do appreciate your effort, and I for one have learned from it. Thanks!
I see now that Greg was referring to the numbers he posited at #191. Unfortunately his comment still doesn't make any sense to me. As Antaeus pointed out, we don't see autism rates of 66% and 25%. I don't see how I am supposed to explain how such bizarre rates might arise, nor how a rate of autism more than twice as high in vaccinated children as in unvaccinated children might arise.
I took another look at John Gilmore's article at AoA about this study, to see exactly what he did, and it isn't as clear as I initially thought.
One thing I noticed is that Gilmore complains bitterly about Madsen et al. using "life-years", and states of the use of this term:
The funny thing is that the term "life-years" doesn't appear in the paper at all. I suppose he means "person-years", which is frequently used in epidemiology as I have previously pointed out. Could Gilmore have deliberately used the wrong term so that if anyone looked it up they wouldn't find out that Madsen used a perfectly acceptable measure? Or is he just incompetent?
I suspect incompetent because he adds:
No it doesn't. When a child was diagnosed with autism he was no longer followed up. There are no "life-years with autism" in the study. Here's what the study itself says:
So even assuming Gilmore meant person-years, no person-years with autism were recorded. Follow-up ended with the diagnosis.
One might almost think that Gilmore either hadn't read the study, or hadn't understood it at all. Your guess is as good as mine. Either way I am tempted to use an unkind word to describe him, but I will resist the urge.
Anyway, I thought I would do what I thought Gilmore had done. Taking data from Table 2 in the study I added all the autistic children in the study together, that's 316. Ignoring the confounding effect of age, since 82% of the children in the study were vaccinated, you would expect 18% of the autistic children to be vaccinated too, that is you would expect 57 unvaccinated autistic children. Instead we see 53 unvaccinated autistic children.
There are 422 ASD children so you would expect 18%, that is 76, of them to be unvaccinated with ASD. Instead we see 77.
Adding autistic and ASD children together we get a total of 738, with an expected 133 being unvaccinated, but in practice there are 130.
Leaving aside the underwhelming association, which I would expect to disappear when corrected for age, where did Gilmore get his figures for the chart he displays? The Y axis of this chart goes from zero to 40, but I have no idea what the units are because Gilmore doesn't say. I also have no idea how he got the unvaccinated autism rate to be 11 units and the vaccinated autism rate to be 16 units. What are these mysterious units? How did Gilmore conclude that, "the data in the study clearly shows that children who got an MMR vaccine had a 45% higher rate of full-syndrome autism than children who were not vaccinated, and a 31% higher rate of less severe forms of autism".
Even if Gilmore calculated the rate of autism in the vaccinated and unvaccinated children as a rate per 100,000, he still wouldn't get those figures - I calculated autism rates in the vaccinated as 59.7 per 100,000 and in the unvaccinated as 51.2 per 100,000, a difference of 17%, not 45%, which can in any case be accounted for by age as a confounder.
I'll leave it there for now, but it seems to me that not only did the AoA author apparently hallucinate units that aren't used in the study, and grossly misunderstand the study design, he also did something very peculiar with those numbers. Perhaps Greg can explain.
A typo leaped out at me just after hitting submit. Should read:
"you would expect 18% of the autistic children to be unvaccinated too, that is you would expect 57 unvaccinated autistic children".
@Kerb
(This is starting to be a lot of fun -- hee hee hee). Okay, if I am understanding you correctly, a straight-forward calculation of autism rates for any age group, or for a collection of ages for that matter, will likely yield a result of a higher vaxed rate. (Just curious but is it 'likely' or 'always'?) This is the result of the confounding factors of vaccination increasing with age, and autism detection also increasing with age. You said researchers in the Danish study had to control for this and that is where person-years measurements came in.
Ok, let's consider my hypothetical sample -- even if it is highly unrealistic. As I am 'highly uneducated' and 'significantly lacking in cognitive abilities' walk me through how one would utilize 'person-years' counting to control for age confounding with this sample.
10 Total subjects, all 3 yrs old
6 vaccinated, with 4 autistic
4 unvaccinated, and 1 autistic
@Greg - you really are dumb as a box of rocks, aren't you?
It has explained to you why your "example" isn't valid.
Meanwhile the patient St Krebiostan of East Anglia continues upon his mission- educating the heathen....
Oh Lawrence, Lawrence, Lawrence... It is agreed that my hypothetical sample is unrealistic. Seriously though, are you suggestin that it is not available to 'person-years' calculation?
Greg, what part of "you can't get meaningful information from a sample size that's too small" is hard for you to understand?
@Antaeus - because I think Greg realizes that getting a larger (legitimate) sample size won't get him the results he wants.....again, where are all of these "tens of thousands or hundreds of thousands" of severely autistic individuals he keeps harping on? Because all available research can't seem to find them.......
You've been walked through everything already. Impressively, you still haven't even figured out what incidence is, as demonstrated by your Torheitenexperiment.
It doesn't really matter; just multiply everything by 100,000. The "question" as posed remains meaningless.
Greg,
Don't get over-excited now, I have no confidence your bladder control is any better than your math.
It is obvious to everyone else reading this that the sample size has to be large enough to get some useful information out of it. In a sample large enough to have statistical power age will be a confounder, unless all the subjects are the same age. The larger the variance in age, the larger the effect of the confounder.
No, age is the confounding factor, because the probability of both having had MMR and the probability of having been diagnosed with autism increase with age.
You have to control for this confounding factor, and the Danish researchers did so through a study design that included the use of person years.
How can we have any sensible discussion about statistics when you don't understand what "sample" and "population" mean? The problem here isn't even that your "sample" is highly unrealistic, I'm sure you could find a group of children with these characteristics in Denmark if you looked hard enough, but it it is certainly highly unrepresentative. In statistics your sample has to be randomly selected, otherwise it won't be representative of the population under study, and none of your conclusions will be applicable to that population.
I'm glad you at least recognize this, as I keep finding that I have overestimated my very basic assumptions about your education and intellectual capacities. I may have to reassess my categorization of you as an idiot.
Where to even start? This still makes little sense even taken at face value. How can you talk about person-years when you don't even specify the length of the study? How can age be a confounder when all the subjects are the same age?
I assume you are trying to say saying there are 10 subjects, 6 of whom are vaccinated i.e. 60%, and 4 unvaccinated, 1 of whom i.e. 25% is autistic. That's 50% autistic overall, more than 50 times higher than the less than 1% we see in the actual population. This is not a representative sample, so we cannot come to any conclusions about the larger population from it, and even if it were representative it is far too small to be useful, end of story.
Denice,
I'll end up pinned to a tree by the arrows of ignorance at this rate...
@Kerb
"How can age be a confounder when all the subjects are the same age?"
(hee hee hee) Very well then, you are saying as long as we are calculating rates age-to-age the straight- forward approach is sufficient, because age is not a confounder? If we are calculating across different age groups (say 3-4) then we will need to control for age, and the 'person-age' method will accomplish this?
I am still very curious though about what this 'person-age' method involves, and I think I will understand better if it is demonstrated. Make up any hypothetical sample that you fancy and demonstrate the process involved with calculating rates, while using 'persons-age'. For the sake of clarity (again for someone that might be slightly above being an idiot) please limit the ages to two -- say 3 and 4 yrs old.
@Kerb
" You then obtain the autism rates for the vax/unvax (using straight-forward calculations) for each age group. "
"Without correcting for the confounding effect of age that I and others have tried to explain to you? You will get a correlation between vaccination and autism, even if there is no causative link at all. It will be a smaller correlation than you would see for the whole cohort, but it will still be there, invalidating your results."
------------------------------------------------------------------------------
Here I stated each age group (meaning only 3 yrs old) yet you are talking about controlling for age, yet in your last post you make it clear that if kids are all the same age there is nothing to control for. Please explain.
(Orac, you are moderating my comments in such a way that does not h
---(cont...)
Orac, you are moderating my comments in such a way that is not helping with my exchanges with Kerb. I think it would be more beneficial for interested lurkers if you don't take half a day to approve my comments.
Guess what, Greg? I don't care. I have a life, and sometimes I don't get to comment moderation right away. Sometimes I'm even away from the blog for—gasp!—several hours (as I was today).
As for how you look with lurkers, you should really be grateful that I sometimes am away from the blog for so long. Given the generally low quality of your comments, you'd look a hell of a lot worse if your comments showed up immediately.
@Denice Walter,
Thanks you very much for the explanation for the Greggies of this planet :)
Alain
Any
As I imagine everyone does, since you just made up the term. Do you have similar trouble with other compound forms, such as kilowatt-hours?
QFT.
I have a dilemma:
some weeks ago, I told the brother that I had a network printer and if he needed to have a print of something, I could setup his computer so he can remotely print from his place. Later on during the evening, I have to go at the store so I inquire (always using sms messages) if I have time to go at the store before he come. He lashes at me that he'll be coming WHEN NEEDED...
I go at the store, minding my own business and upon returning back, I tell him that I don't want to see him here, that if he has some issues regarding his life and job that I won't be the drudge bearing his existential crisis. I also told him to not contact me until he change.
2 or 3 weeks later (i.e. this week-end), I write him again wanting to settle thing out (longish texto). Tonight, we go for (apple juice for me and beer for him) and I learn that:
1-: he thought I was needing a lift to the store (I walk 7km each and every day, do I look like a need a lift?).
2-: he treat us the same way as everyone else and if we're not happy, better complain elsewhere as he couldn't be bothered with our complaint. (btw, 9 times out of 10, we hit his voice mail and that's over a period of 2-2 1/2 month for those 10 calls maximum).
3-: If I have some issues with concentration, I just have to throw away my medications and seek a psychologist. The natural falacy.
4-: basically, he's running on the XM frequency while I run at router 2.4GHz frequency, there's no way to understand each other and he won't ever try (see rule 2).
Question, Should I spend christmas with the rest of the close family and having a really bad time because I'll be imposed his company along with a few others a**holes or else, spend it with my cousin and actually have a much better time? Collateral damage is that I'll have to explain my mother that I won't go to her christmas party.
Alain
Alain,
For the sake of your mental health, I'd say go to your cousin's. Your brother doesn't understand you and (crucially) couldn't be bothered to put in the effort to learn to understand you. Going to him would probably provoke a meltdown.
Thanks Julian,
I agree that going to the brother will lead to a meltdown or at least, some bad feelings.
I'll wait until he pay me the 20$ he owes me and after that, I'll give him the jackboot.
Alain
Tell you what, Greggie, why don't you go back and read the comment where Krebiozen already gave you that explanation? Give you a hint, the comment number ends in "4"; if you can't find it, it means you aren't even trying to understand, or that whoever called you just slightly above an idiot was overestimating you, or of course both. In any of these cases, it's not our obligation to overcome all your inadequacies, especially the ones you proudly flaunt.
@Kerb
"In the first scenario MMR causes autism. We take two hypothetical children, both vulnerable to having autism caused by MMR. Child A is vaccinated at age 15 months, immediately develops autism and is no longer followed in the study. His presence in the study contributes 1.25 ‘unvaccinated non-autistic’ person-years. Child B is not vaccinated until he is 2 years old; he immediately develops autism and is removed from follow-up. His contribution to the study is 2 ‘unvaccinated non-autistic’ person-years. Together they add 3.25 ‘unvaccinated non-autistic’ person-years to the study. Neither adds any ‘vaccinated non-autistic’ person-years to the study.
In the second scenario (what I believe is the real world), MMR does not cause autism. We take the same hypothetical children who are not vulnerable to MMR, because no one is. Child A is vaccinated at 15 months but never develops autism. He is followed up for a further 2 years until the end of the study. He contributes 1.25 ‘unvaccinated non-autistic’ person-years and 2 ‘vaccinated non-autistic’ person-years to the study. Child B is vaccinated at 2 years, and never develops autism. He is followed up for a further year. He contributes 2 ‘unvaccinated non-autistic’ person-years and 1 ‘vaccinated non-autistic’ person-year to the study.
In summary:
In scenario 1 in which MMR causes autism, we have 3.25 ‘unvaccinated non-autistic’ person-years and zero ‘vaccinated non-autistic’ person-years.
In scenario 2 in which MMR does not cause autism, we have 3.25 ‘unvaccinated non-autistic’ person-years but we also have 3 ‘vaccinated non-autistic’ person-years.
Imagine there are many other children in the study, most of who will be identical in each scenario, so we can ignore them, and a number of children whose contributions will be similar to Child A and B in my thought experiment.
There is a very obvious difference in the number of ‘vaccinated no autism’ person-years between the two scenarios. If vaccines caused autism we would expect to see that difference, in terms of a higher than predicted ratio between ‘unvaccinated no autism’ person-years and ‘vaccinated no autism’ person-years. We don’t see this, therefore there is no reason to think there is any link between MMR and autism."
----------------------------------------------------------------------------------
Holy Geez!! The implication of using the 'person-years' method is that we would need to calculate figures for all 500, 000 plus kids in the Danish study. Anyone interested in verifying such numbers would also have to make such calculations!
You explained that this procedure is necessary because it overcomes age-confounding. Yet, you seem to be agreeing that if we are studying one particular age group -- say 3yrs -- age is no longer a confounder. Again, my question is why did researchers not simply do age-to-age comparisons for 2yrs, 3yrs, 4yrs, 5yrs,6yrs, 7yrs -- only? They get the unvax/vax rates for each of these groups and compare the figures?
@Kerb
In the second scenario (what I believe is the real world), MMR does not cause autism. We take the same hypothetical children who are not vulnerable to MMR, because no one is. Child A is vaccinated at 15 months but never develops autism. He is followed up for a further 2 years until the end of the study. He contributes 1.25 ‘unvaccinated non-autistic’ person-years and 2 ‘vaccinated non-autistic’ person-years to the study. Child B is vaccinated at 2 years, and never develops autism. He is followed up for a further year. He contributes 2 ‘unvaccinated non-autistic’ person-years and 1 ‘vaccinated non-autistic’ person-year to the study.
But, is this not based on the thinking that kids will display autism shortly after MMR? What if 3 - 4 yrs after MMR these kids become autistic?
Then that shows that the MMR did not cause the autism.
@Julian - actually, I believe the correct term would be, that individual was always autistic, just wasn't diagnosed until later...
Yes, this is called "counting."
Wow - research validation? Who would have thought....perhaps that's why there is a standard process for this. You, Greg, truly are an idiot, aren't you?
Greg,
I am still astonished at the depths of ignorance you continue to plumb.
Would you care to explain exactly what you mean by that? Are you honestly interested in understanding this subject, in learning something and getting closer to the truth? Or are you just interested in irritating as many skeptics as possible by behaving in as infantile and ignorant fashion as you can?
If you really want to learn, I am willing to continue this, otherwise I'm not.
@Kerb
All joking aside, I didn't spend much time considering the 'person-years' rationale so I over-looked your explanation of it. The obvious question is will the length of follow-up (2 or 7yrs) not affect the results. I am assuming in the study it was actually a two year follow-up (what was it?).
Why the hell can't you go read the study yourself and figure it out?
Just a couple of things worth clarifying, regardless of Greg's dubious sincerity:.
I never said the use of person-years overcomes age confounding. How could it? You could take a person-year between the age of 1 and 2, and another person-year between the age of 6 and 7, and the probability of both MMR vaccination and the probability of a diagnosis of autism is higher in the latter group.
I said that Madsen's technique, which included the use of person-years and age correction, overcomes the confounding effects of age.
The use of person-years is a standard epidemiological technique for calculating incidence when you are comparing different subjects who have been followed in a study for different lengths of time. Following one child for a week and observing that he was not diagnosed with autism is not the same as following a child for 6 years and observing he was not diagnosed with autism, is it?
Age correction is an entirely separate calculation, but I can offer a simplified explanation. It depends on calculating what contribution age makes to the likelihood of being diagnosed autistic and then removing that contribution from your figures. As an example, just to explain the principle, say there was a linear relationship between age and diagnosis of autism i.e. that the graph I linked to above was a straight line. Say the chances of a specific child being diagnosed as autistic were zero at age 1 and 1% at age 7. In this case the chances of being diagnosed as autistic at any age would be 0.167 x Age(yrs) - 0.167. By reversing this formula you could correct all your data to a standard age, leaving only the effects of MMR (or other confounders, which you would deal with the same way).
Since the relationship between age and risk of diagnosis of autism is not linear, it would require a more complicated formula to describe it, and the formula to correct for age would also be more complex.
This also affects the issue of whether we are looking for a time-dependent effect of MMR on autism. If a child has MMR and is diagnosed with autism 6 years later, are we still to believe the MMR is the cause? This study looked at whether vaccination per se affected the risk of autism (it didn't), or if age at vaccination affected the risk of autism (it didn't), but it also looked at whether proximity to vaccination affected risk of autism (it didn't). Table 2 shows that the risk of a child being diagnosed within 6 months of MMR was 61% lower than the risk of an unvaccinated child being diagnosed autistic, though this was not statistically significant.
With more than 200,000 person years of observations of children within 6 months of vaccination (that is more than 400,000 children observed for 0.5 years), as I pointed out before only 3 were diagnosed as autistic during that post-vaccination period. So much for the "regressing into autism before their parents' eyes after MMR" claims from Greg and AoA.
Greg, please read the study itself: here's the link.
@ Alain:
I think that it's time to force *votre frere* into line- but YOU have to set the programme- his scenarios are terrible:
-don't offer him anything in the future
-only assist him and/ or interact with him IF AND ONLY IF he behaves reasonably well towards you ( and I don't just mean just in the immediate time frame- he can act decently for 20 minutes- who can't?)
- he now knows that you won't 'cut him off' if he behaves badly ( define this any way you choose) so he repeats
- tell him that you won't help him because..... ( put into your own words)
- as soon as he starts up, cut the interaction off RIGHT THEN AND THERE ( easier over phone or internet )
- YOU set the limits, not him.
- stick to your guns( hold firm)
Here's a sample, " I will not help you with your computer or let you use my tech as long as ( list examples: swear, argue, fight, scream, take my stuff, harass me et al)".
Don't do nice things for people who hurt , aggravate or inconvenience you.
For *votre mere*:
set a date, perhaps just before the holiiday event and take her out _by yourself_ to do something enjoyable, even if it's just for a walk and going to a cafe, your treat. If you have a little extra money, get her a few live flowers- even though it's winter. A film isn't a bad idea ( including a video at home). Just you and her for an hour or a few.
I wish you good luck but your own actions will determine the course more than luck will.
Here's an observation about woo and agenda-bolstering:
detractors of SBM and SB psychology will tie themselves into knots trying to cast suspicion upon studies they disagree with that are done using PERFECTLY reasonable statistical methods and research designs BECAUSE they never got that far and they address an audience which is in precisely the same boat.
The blind misleading the blind.
A recent example- (courtesy of PRN):
( paraphrase) - 'When researchers tell you that the average person eats X amount of sugar a year, it's not true because I eat almost none and diabetics eat almost none so average people eat much, much more than X"
Anti-vaccine advocates are extremely adept at this art form questioning methods and data. ( see Blaxsted's collected oeuvre)
To see how it works, let's look at a historical topic:
'race'** and ability.
those who wanted to push an (incorrect) agenda showed studies that revealed a strong relationship between academic achievement and 'race'. Whites do better therefore they ARE better, it's genetic or suchlike. Not at all.
HOWEVER they leave out discussion of other contributing factors associated with achievement like POVERTY which is associated with race. If you control for income/ social class, a very different picture emerges. Other issues might involve the mother's level of education- which also correlates with poverty which correlates with 'race'. And so on.
Data in the life sciences often look like this: variables are interwoven. Because life is complex. Someone once said it's hard to find a correlation of 0. If you want to research an illness that is common to aging, you'll find that other factors contribute to outcomes as well. If you study ASDs, you'll find that multiple issues are involved NOT ONE.
This stuff is so much easier to understand if one draws Venn diagrammes or other pictorial devices.
Alt media proselytisers will not tell you about this: I will. It's the way they lie and cater to their audiences' misconceptions, prejudices and *idees fixees*.
** there's only one race AFAIK.
Denice,
Au contraire! There's the New York marathon, the Boston marathon, the Big Pig Gig, Formula One, and NASCAR, just to name a few.
Mr O'Brien:
Please let's not bring up the issue of race and the ability to win races.
Alain,
Needless to say, that's excellent advice from Denice at #231.
Denice, it's also interesting to see how these dubious calculations always support the proponent's position. For example the IQ tests devised by upper class white men always found that upper class white men were more intelligent than anyone else.
I remember seeing (in The Mismeasure of Man IIRC) a pictorial IQ test administered to hopeful immigrants at Ellis Island* that included a "what is missing in this picture" of men playing baseball; as if some Eastern European immigrant just off the boat in the 1920s or whenever would have ever seen a game of baseball!
Denice Walter,
I'd hate to be accused of racy comments in this thread.
@PGP
Now let me explain why the government has an interest in keeping citizens healthy. The Spanish Flu shut down cities for weeks. Quarantines brought economic activity to a standstill, and workers couldn’t work due to illnesses or being unable to enter the city they worked in.
England, between World War 1 and the Spanish flu, lost most of the young men of a generation, and was therefore deprived of any economic contribution those men (and the children of women who died in pregnancy) might’ve made.
Finally, vaccines free up half of the workforce. Women who might be stuck at home with sick kids no longer have to spend weeks or months at home. (Yes, guys might look after sick kids, but it’s statistically and historically unlikely.)
Besides that, there’s the obvious benefit of kids being healthy and not needing taxpayer money for a seeing eye dog or hearing aids because they had a bad case of measles or mumps. (Not that seeing eye dogs or hearing aids are bad things, mind you.)
Speaking of, you never answered my question- would you be willing to care for a kid who was disabled from measles or mumps? Would you still love that child? I’m just asking, since your crowd seems to deal in conditional love.
-------------------------------------------------------------------------------
(Oh -where do I start?)
Back to your meme that I paraphrased as, 'so what if vaccines cause autism, things were a lot worse in the bad, terrible days of vaccine preventable diseases. PGP, I already told you that I sincerely believe that vaccines played a minimal role in overcome those days. Please understand that I make this statement while at the same time I will honestly concede that I was once sitting on the fence on the issue.
PGP, yes, vaccines have save lives. I don't want to stand over any 2yr old girl's grave, whose life was taken by a vaccine preventable disease. But PGP, I also don't want to stand over the grave of an infant whose life was taken by SIDS (yes- vaccines cause that too), or a 4yr old autistic boy who wondered and eventually drowned.
As for your spiel about past epidemics ravaging the male population, PGP think about this for a second: Autism at it's current rate is affecting 1 in 30 boys. What life do you think these boys, who will eventually grow into men, will have? How many of them will be able to stand on their own feet and serve as productive citizens?
PGP, you also explained how vaccine spared women in the past from having to attend to sick children. Really, PGP-- with autism, when many fathers bolt, who do you think is left having to look after these sick children?
PGP, I couldn't agree with you more when you said it's in society's interest not to have sick children. With the multitude of chronic diseases plaguing our kids today, I can't say vaccines are delivering in this regard.
@ Krebiozen:
Some 20th century Euro-centric psychologists surmised that natives of Africa did not ever achieve adult thought ( formal operations) because both children and adults fared poorly on standard Piagetian style tests relevant to their ages/ stages.
More sensitive observers noted that these subjects were unfamiliar or uncomfortable with the testing materials, the format and perhaps even the *idea* that you needed to respond to nosey outsiders' interrogation about how much water a bottle held or which container of several was the largest.When the tests were constructed with materials that the subjects used, their answers improved.
The curious idea here is that adults around the world WOULD be different - it's paternalistic to assume that they might remain children emotionally and cogniitively . While it is true that cultures that revolve around literacy will present more strongly on Jean's tasks ( after all, he was a European and focused on his own experiences), even those societies which are not predominately literate have indications of formal thought in their own culture because they have philosophical ideas, trade, government and symbolic and coded speech.
@Kerb
"With more than 200,000 person years of observations of children within 6 months of vaccination (that is more than 400,000 children observed for 0.5 years), as I pointed out before only 3 were diagnosed as autistic during that post-vaccination period. So much for the “regressing into autism before their parents’ eyes after MMR” claims from Greg and AoA."
Yes Kerb, on numerous occasions I have mentioned how parents' report their children dramatic regression into autism, and I do believe them. That said Kerb, if you really want my personal take on vaccines and autism, I think autism is a result of the cumulative effect of vaccines. It's not an ingredient or a particular vaccine alone that causes autism, but the gradual overload of repeated vaccination that may eventually tip things over the edge. And, the fact that many parents do fault MMR may simply be the result of it being one of the last vaccines to be administered when things become noticeable.
Relating to this position of a gradual overload, I don't think it's fair to test for MMR reaction for only up to 6mths. Actually Kerb, let me be really blunt and tell you what I really think:
This study is BS! It was never about controlling for age. Researchers had the data and ran the numbers for age-to-age and found something they didn't like. I believe you mentioned somewhere that the age-to-age data showed no correlations. I don't know if I believe you -- but there is also the possibility that as the kids got older the trend was a movement towards increased correlation. This trend would definitely be an embarrassment to pro-vaxxers. They would have to explain. Could it possible support the notion that increased vaccines were causing autism? Seriously Kerb, reading the study we hear about 'mother socie-economic status', 'birth-weight', 'gestational age' -- yet no one bothered to provide the age-to-age rates?
Anyway, they probably lumped the babies in so they could have the pretext that they needed to control for age. Limiting follow-up to short periods also helped things greatly because it reduced autism cases and therefore resulting in a sample size that is less likely to find a correlation.
Hey Kerb, don't mean to interrupt the 'tango' that we have been having the last few days. Indeed you are quite good on your feet.
Except that they don't, Greg. The studies have certainly been done trying to figure out what causes SIDS - if vaccines played a causative role, we should see a higher rate of SIDS among vaccinated children than unvaccinated. Instead, we see a slightly lower rate of SIDS in vaccinated children, though the difference is not necessarily statistically significant.
Oh, yes, Greg. I know what you'll say. "Because it doesn't fit in with the overall evidence, and by 'the overall evidence' I mean the conclusion that I jumped to, it must be that the studies which don't find a correlation between vaccines and SIDS are rigged studies paid for by Big Pharma!"
But here's why you think that, Greg: you have no integrity. You don't know, and don't care in the slightest, whether a study's methodology is sound or it's full of holes. If someone's saying what you want to hear, then you'll turn a blind eye to clear evidence of wrongdoing. (And of course, you've already admitted that you'll tell lies yourself if they make your position look better than it actually is.) If people aren't toeing your Party line, though, you'll immediately smear them with baseless slander.
How many times so far, Greg, have you accused Madsen's group of doing something dishonest by including the 6-12 month olds in the study? "A fast one," you called it. Well guess who also included the data from that age group in their calculations, Greg? The scientist wanna-bes at AoA whose work you were praising. Why have you never once questioned their inclusion of that group, Greg? Why is it that you'll happily accept antivaxer "scientists" doing the very same things you'll call "fast ones" if they're done by mainstream vaccine researchers?
What could possibly lead you to this conclusion? You've demonstrated over and over again that you can't think; the opinion-blobs floating in the deliquescent slop that fills your brainpan are of vastly less interest than simply watching you make a fool of yourself pretending to try to grasp rudimentary statistics.
Time for lilady's Media Review of the Dachelbot's Media Review.
The bot happened to find a very interesting article about the research being done by a Nobel laureate (Dr. Suchel) to locate a single gene mutation in autistic male siblings. Then the bot had to spoil it, by commenting with her usual Spam:
http://www.wbur.org/2013/10/24/thomas-sudhof-supple-autism
Leaving aside the fact that vaccines do not cause autism:
A quite good life, if I and my father are anything to go by. Oh, and most of them will be able to lead productive lives.
Ah, the old "autism causes marriages to break down" meme. Except it doesn't, Greg. Researchers looked at the hypothesis and found it was false.
Once again, Greg, where is your evidence that vaccines cause these disorders?
@julian - because his "handlers" at AoA say so, that's why....
@greg - you've already admitting you have no grasp of the staristcal method, so besides stating that the researchers ignored "evidence" and you don't like (or understand their methods) is it that you just don't like that they did not confirm your preconceived notions?
And, god help me, a serious question Greg - what is the actual percentage of those that are diagnosed with autism are classified as "severely autistic?" And then how does that relate back to the 1 in ___ number you keep throwing around....and then, how does that compare to the historical rate of severely disabled individuals in the population?
I know, a hard question which involves math and all, but really it should be difficult.
#shouldn't
Greg,
You clearly still haven't read the study. Or if you have you haven't understood it.
There's not a hint of that in this data, only 3 children out of more than 400,000 diagnosed with autism within 6 months after MMR? How is this possible?
If this scenario was accurate, it would have shown up in this study (and in many other studies). It didn't, therefore the scenario is not accurate.
They didn't. They followed up children until more than 60 months after MMR. Read the study. I was simply pointing out that the "regressing into autism within hours of vaccination" tale is not supported by the data looking at time after vaccination with MMR.
Yet you are unable to give any rational explanation for this opinion.
Read the study, it's all there in black and white. They looked at vaccinated and unvaccinated for the whole cohort, for different ages, for those vaccinated at different ages, and found no correlation at all. No competent epidemiologist would run figures "age-to-age" without checking for and correcting for confounding effects.
A movement towards increased correlation between what? They looked at the children by age, and in the older children there was still no significant difference between vaccinated and unvaccinated. In the children vaccinated in 1991-92 i.e. 6-7 years before the end of the study, the autism rate was exactly the same in the vaccinated and unvaccinated (relative risk = 1.0, 95% confidence limits 0.59-1.70).
I often get the feeling that you don't know what any of the terminology you are using means, but stick it in a sentence that looks to you as if it makes sense. Perhaps you think statistics is all a big bluff, and all those complicated-looking formulae don't really mean anything at all. They do. Trying to bluff your way through this just makes you look like even more of an idiot.
The non-existent trend you just made up? Good grief.
Let me get this straight. This study doesn't support your delusional beliefs so you are making up results that would support your delusions, and suggesting that provaxxers have to explain the results you have invented out of thin air? Are you insane?
What information was included about "‘mother socie-economic status’, ‘birth-weight’ and ‘gestational age’" (why the quotes?) that is not included for age? These factors were checked and found not to be confounders. What "age-to-age rates" do you mean? If you mean results that are uncorrected for age, why would anyone provide figures that we know are bound to be wrong and misleading?
If you read the study you will see that there are results for each age group, comparing them to unvaccinated children of the same age. This is the normal way a paper like this is presented. I see nothing suspicious about it in the slightest.
Absolute idiotic BS. They didn't start follow-up until the children were a year old, and they looked at each age group separately. The need to correct for age is because it is a confounder at all ages as I have repeated over and over.
They didn't limit follow-up to short periods, they followed up all children right through the study. The study included of all children born in Denmark during the period from January 1, 1991, to December 31, 1998, and continued until December 31, 1999.
We haven't been dancing a tango, this has been a cookery lesson, and what you haven't knocked over and spilled, you have burned, or smeared on the walls or over yourself.
@Kreb - it is like trying to explain Rainbows to a Juggalo.....
@Kerb
"They looked at the children by age, and in the older children there was still no significant difference between vaccinated and unvaccinated. In the children vaccinated in 1991-92 i.e. 6-7 years before the end of the study, the autism rate was exactly the same in the vaccinated and unvaccinated (relative risk = 1.0, 95% confidence limits 0.59-1.70)."
No significant correlation using 'person-years' method that would include a short-follow-up in which cases get missed? How about just looking at 6yrs separately, 7 yrs separately and running the straight-forward calculation for each of these groups. This is what I meant by age-to-age comparrison, Kerb, comparing 6 yrs old to 6yrs old -- no confounding problem-- not comparing 6-7 yrs old to 6-7 yrs old --confounding problem.
You said sometimes Kerb you think I am bluffing about statistics and I am trying to make a complex problem sound easy. Kerb, I have the opposite problem with you -- sometimes I think you are trying to sell an easy problem as an extremely difficult one.
Anyway Kerb, I think it's only fair that I share what I found out about the study with AoAers. I will be back with my final summary of things that I will also share with you guys. In the meantime, Kerb, I think it's pointless to keep up with this topic. What I would kindly request of you though is to ask your buddies in high places to provide me with some of the raw data for the study. Specifically I want the the vax and unvax groups figures broken down by individaul ages -- again not 2-3, but 2 yrs and 3 yrs. I want their sample numbers, and their autism cases numbers. Please don't provide infor for kids younger than 2yrs.
Kerb, perhaps you may say there is no point giving me such figures because I would not know what to do with them. Try me! Heck -- I even have a friend who teaches statistic at a major university. Perhaps he could help me with the figures.
Again, sterling efforts by Kreb, Antaeus and Julian.
BUT I ask, "Can this person be educated?" Could Alison MacNeil, Dan Olmsted, Cat Jameson or Jake be educated?
I don't think so. Beliefs like this may form a person's identity, solidifying through repetitve use, giving him or her meaning and purpose, either as a general contarian, alt med proselytiser or Mam-to-be-emulated via blogs and facebook. A few write and sell books delineating all facets of their 'expertise'.( see AoA and TMR for examples)
If a person studies Psychology, even at an undergraduate level, statistical analysis and SB research are part and parcel of the curiculum, so I don't know what is wrong here.
( Whilst my undergrad degree is not in Psychology, I studied social sciences as well then)
Thus we have someone trying to make a splash much as John Stone ( today @ AoA), Teresa Conrick or Gary Null do by attempting to poke holes into SB research using smear, artifice and prevarication instead of data or methodology.
The meme of a child observed lapsing into autism right after a vaccine has been eviscerated countless times- Orac has already done that. The validity of eyewitness testimony has been entirely demolished for many years. So why do these ideas persist?
Because they make a good story and garner sympathy for the parent. Because they make a disappointed parent feel better by casting blame on to corrupt doctors, governments and companies rather than accepting the fact that some children have ASDs or DDs or cancer BY CHANCE or via genetic mechanisms, not through malfeasance of others.
Beliefs become more powerful because of repetition and emotionally charged narration: like any memory or internal idea. If you think about it, talk about it or get emotional over it, you strengthen its hold on you. It is irrelevant whether the idea is true or imagined- based on fact or fantasy.
In addiition, these science deniers, like Mike Adams, get an ego-boost because they are elevating themselves above expert (consenus) opinion- which they can never be a part of because they can not be educated and are incapable of putting their fragile egos aside in order to receive instruction.
So, we're talking about psychological needs rather than educational ones. Fortunately, I'm sure that many lurkers and regulars here- including me- will benefit from the discussion and information presented by these fellows- the target, not so much.
Remember that GInger Taylor, Alison MacNeil and TM Saint have also been educated in social sciences.
Actually, it is exactly like trying to explain Rainbows to a Juggalo.....I appreciate its insistence on continuing to try to sound "scientific" even though it continues to fail miserably (but really fun to watch).
First of all, this is coming from the same person who started off claiming it was completely okay to compare 6 years old to 6 months old, as long as the end result looked like damning evidence against vaccines.
Second of all, Greg seems to think that an age cohort labelled "6-7" means "all the children who are 6, plus all the children who are 7". Yet he's been looking at the same tables we have, where the age cohorts are listed as "0-1", "1-2", "2-3", "3-4", etc.
A reasonable person who was trying to understand would surely have no difficulty figuring out from those that the designation "n-(n+1)" must mean "children who have reached the age of n, but not yet the age of n+1". If Greg believes otherwise, why is it he had absolutely no objection to the idea of counting nearly every child in the study in two different cohorts?
He claims now to believe that what the study says about the "6-7" age group refers to every child who is either 6 years old or 7 years old, and sanctimoniously informs us that this is an insufficient level of precision. Yet we're supposed to think he didn't have any objection to the 2-year-olds being counted once in the "1-2" cohort and then again in the "2-3" cohort? Much more probable is that, at that time, he hadn't yet had the idea to pretend he believed that.
I believe Greg and all his FIFUDOS antics are here for only one purpose: to annoy us and waste our time. I frankly wonder if it might be time to consider a ban. It's a drastic measure, yes, but hey, if anyone can name even one good thing that has ever come from Greg's puerile antics, name away, because so far as I can tell, he hasn't even provided the meager value that came from having Thing-troll around.
@Antaeus
You guys should know by now that I don't care about your bans. Nevertheless, perhaps you guys should do it for a good reason (and I am sure I already provided you plenty). Seriously, I am confused when Kerb seems to be talking that age-to-age is a confounder that he is actually refer to, say, comparing 6-7s to 6-7s. I am still having a tough time wrapping my mind around what Kerb seems to be saying that I am starting to think that when he says 6-7 he is talking about 6 and 7. Seriously, where is the confounder in comparing only 2 yrs old?
, well, you've already conceded you have no evidence supporting a claim that vaccines cause autism.
What evidence demonstrates that they cause SIDS? Be specific.
@Greg - you keep using "sciency-sounding" words, but they don't mean what you think they mean.....
Antaeus, the only 'good thing' to come from the entire episode are responses from commenters like you.
Perhaps it can serve as a "life study class" wherein we discuss how scoffers deal with SB research. ( I doubt that Orac will ban anyone who doesn't break 'the Rules'... I'm sure you remember what has caused others to be banned. for.being nastier, more obscene or more obviously SMI )
For example, today ( AoA) John Stone** critiques a study by Taylor et al which has just appeared in BMJ Open. Interestingly enough, the study simplifies the issue of age by only including 8 year olds ( longitudinally 2004-2010, UK): it shows a levelling off of autism following the increases seen inthe 1990s and stands in contrast to US CDC data on 8 year olds that has shown increases over a similar time period.
Amongst Stone's many criticisms are the fact that the lead author had previously criticised AJW's work at Royal Free, that the BMJ group is compromised by its 'parnership" with Merck and GSK and that it is currently involved in a suit brought by you-know-who. Also the data comes from GPs and what the frig do they know? Plus it contradicts recent Scottish data that shows increases etc etc etc. In short, it is nearly as "troubling" as the Danish study.
Clearly, it contradicts what John Stone wants to believe- that autism is a rapidly escatling catastrophe that will soon engulf us all. He seems disturbed by the inconsistency between US and UK data. Oddly, I remember inequalities between US and UK diagnoses for SMI ( schizophrenia vs schizophreniform IIRC) which could possibly reflect the fact that slightly differing criteria are used- not just that more serious MI is less noticed or apparent in one place ( lit joke).
Nothing will convince Mr Stone of what he believes to be true and which forms the centring postulates of his existence. Or gets him noticed.
** who was quite polite to your truly - surprisingly.
If this is true, I sincerely hope you show a copy of this thread to your friend.
I really do love how Greg admits that he doesn't understand statistics or the underlying methods - when they are plainly spelled out in the research paper, yet he still tries to claim the methodology is invalid (for the only reason that the results do not match his preconceived notions).
Please Greg, keep it up - this is a very teachable moment for people interested in real Science & not the kind of garbage you are used to being spoon-fed at AoA.
a 4yr old autistic boy who wondered and eventually drowned
Wondering causes drowning? Oh dear.
Now I'm imagining a paper for the Annals of Improbable Research: "Curiosity as a risk factor in feline mortality".
Greg,
It wasn't a short follow-up period, it was until the end of the study, which, I repeat, included all children born in Denmark during the period from January 1, 1991, to December 31, 1998, and continued until December 31, 1999. The person-year, I wearily remind you, is a standard epidemiological measure.
You mean all those children that are 6 years old on just one date? Or do you mean all those children who reached their 6th birthday at some point during the study? This is a very important point that I don't think you appreciate: I don't think you have considered exactly what data was analyzed in this study.
Essentially we have over 500,000 sets of patient notes, with a date of birth (somewhere between 1991 and 1999 inclusive), a date they were vaccinated, and a date they were diagnosed with autism or ASD. Please take a moment to consider how you would analyze this data, using your "age-to-age" method. Feel free to explain exactly what you would do with this data.
OK, explain to me how simple it is to process the data in those 500,000 patient records using your simple "age-to-age" method, so you have enough statistical power to detect a difference between vaccinated and unvaccinated and while correcting for the confounding effects of age.
I have no confidence at all that you will pass on anything at all accurate, just your garbled and distorted version of this debate.
However, please be sure to ask John Gilmore why he wrote:
Please point out to him and the other AoA people that the term "life-years" does not occur anywhere in Madsen's paper, and that follow-up ended with an autism diagnosis, so an 8-year-old diagnosed at 6 would not have been counted as 2 “life-years” (or person-years) with autism. There were no person-years with autism - what would be the point? How can anyone trust anything Gilmore wrote when he made huge blunders like this?
While you're at it you might also ask him exactly how he came up with the 45% increased risk of autism in vaccinated children from the figures in the study. His article doesn't explain, and I can't figure it out.
Do be careful though Greg, as AoA will ban you without hesitation if you post any comment there that looks at all critical of their idée fixe.
Me too. A dark and oppressive tiredness is creeping over me.
I don't think "raw data" means what you think it means. By the way, I don't have friends in high places, I just know that if you write a polite letter to a study's author (their email address is usually included in the study), asking an intelligent question, or even asking for a copy, in most cases you will get a reply. I told you above who has the data; I'm not going to ask for it for you.
Would you like it printed on embossed paper and framed as well? Even assuming that anyone is going to spend the considerable amount of time involved extracting this data (what you are asking for not "raw data") , when you say you want data for those aged exactly 2 and 3, do you mean this age on a specific date, or for all children on their 2nd and 3rd birthdays? This would include the same children twice, once when they are 2 and again when they are 3 - you know that, right?
This isn't a group of 500,000 children, it's a group of 500,000 children followed for 8 years. I think maybe this is what you haven't grasped, and which I wrongly assumed you understood from the beginning. No wonder the concept of person-years confused you if you thought this.
Why don't you ask your statistician friend how he would approach this problem? Ask him how he would process these 500,000 patient records, with date of birth, date of MMR and date of autism diagnosis (if any). How would he go about finding out if there was any association between MMR and autism? I fully expect any statistician worth his salt to propose a similar study design to the one Madsen used. Pleas be sure to let us know what your friend says.
Correction: mine above - that's not *longitudinally* OBVIOUSLY.
also a few typos.
Non-existent lord almighty! I do believe that this thread should be shown to another type of expert rather than to a statistician.
We are getting nowhere fast, gentlemen., Krebiozen, I think that you should charge for tutoring.
That's why I say Greg doesn't even do the meager good that Thingtroll did. Thingtroll was at least bringing sincere arguments that she actually believed to the table, even if the reason she believed them was SMI. With Greg, there's essentially zero chance that anyone not actively dedicated to getting things wrong will ever suffer the "misunderstandings" he pretends he needs to have cleared up for him.
The bar's been set fairly high, yeah, but there comes a point when someone who's clearly trying to be disruptive should be recognized as such and treated appropriately, even if their means of disruption isn't as obvious as some others.
Denice,
My fees are high, but I charge for results, and I fear I have failed in this instance.
I am pretty sure I have wasted my time with Greg, but it did force me to take an even closer look at Madsen's study, which I have found useful. I'm currently working on a study looking at the effects on patient management of putting interpretive comments on thyroid function test reports on hypothyroid patients on thyroxine over a 20 year period, and some similar issues regarding confounders have arisen. What looks simple at first glance often isn't in the realm of epidemiology and statistics.
I also think it was worth looking a little more closely at Gilmore's article. How anyone can take it seriously after having those serious errors pointed out I do not know.
@Kerb
"This isn’t a group of 500,000 children, it’s a group of 500,000 children followed for 8 years. I think maybe this is what you haven’t grasped, and which I wrongly assumed you understood from the beginning. No wonder the concept of person-years confused you if you thought this."
Holy sh!t -- I screwed up! Kerb, give me a few days to read things over and contemplate this 'person-years' method. I will come back with my take on the study. If I did indeed get things wrong, I will apologize to the researchers. I will even communicate all misunderstandings to AoA.
@Antaeus
Did you see that? I admit to when I make a mistake and when I am wrong. I think that's what integrity is about. Now when I asked you if you believe that vaccines in their totality do not play a causal role with autism and you said 'yes', how did you behave?..... Freddo?
@Orac
"Guess what, Greg? I don’t care. I have a life, and sometimes I don’t get to comment moderation right away. Sometimes I’m even away from the blog for—gasp!—several hours (as I was today)"
My bad: Indeed Jake did say that some days your schedule is completely crammed with Merck and Pfizer appointments. (Sorry -- I know it comes a bit late but I couldn't resist.)
Greg #166 and #238:
And you're wrong. There is a polio outbreak in Syria.
Here are some quotes from the article.
And
Once again, a VPD comes hurtling back after a vaccination programme collapses.
@Julian - damn, I hadn't heard that.....though not unexpected, I guess, given the situation. War sucks (understatement of the year, I know).
I'm sorry, Greg, I don't understand what it is you're trying to imply here. Perhaps if you spelled it out in syllogism form it might be clearer.
Hmm, I might have phrased that better. Perhaps:
would have been better.
Such an admission should delight me, but I have a nagging suspicion you are not sincere.
No, please, don't. Talk to your statistician friend and asked him about the study design first.
@Kerb
Ok Kerb, again I am not making any claims. I am just seeing my thinking about this is correct:
--------------------------------------------------------------------------------
"We designed a retrospective follow-up study of all children born in Denmark during the period from January 1, 1991, to December 31, 1998"
"Follow-up for the diagnosis of autistic disorder or another autistic-spectrum disorder began for all children on the day they reached one year of age and continued until the diagnosis of autism or an associated condition (the fragile X syndrome, Angelman's syndrome, tuberous sclerosis, or congenital rubella), emigration, death, or the end of follow-up, on December 31, 1999, whichever occurred first"
-------------------------------------------------------------------------------
So essentially they started with kids that were born as early as Jan 1, 1991. Now since their birth dates would naturally vary they needed to control for age and that is where 'person-years' counting come in. Essentially what they are doing is taking it child by child. They start the 'autism-watch' stopwatch when each child turns one and only stop it when the kid is detected to be autistic or has some other ASD. If the kid does not develop autism or an ASD they run the stopwatch until the end of study period, Dec 31, 1999. Now some of these kids will be vaccinated with MMR and some won't. They then compare the amount of autism-free 'person-years' between the two groups to see if there is a correlation. Is this an accurate summary, Kerb?
@Julian
"And you’re wrong. There is a polio outbreak in Syria.
Here are some quotes from the article"
Julian you are obviously a smart guy, but I think you need to put things into perspective. You are talking about a few cases of 'suspected' polio in war-torn Syria when currently 1 in 50 kids in the US are permanently disabled from autism (and this does not include the other chronic disabilities that vaccines are causing). Drug-pushers wake up! Dachel is right! The real epidemic is right here at our doors.
@Greg - you moron.....confirmed cases now, plus a large outbreak in Somalia that is spreading to both Ethiopia & Kenya...this, with a disease that does spread rapidly, especially in areas where there isn't clean drinking water.
Ever going to answer my question Greg, are is the math too hard for you?
Meanwhile, in other anti-vaccinationist news:
the Internecine War continues... and continues.. apace
Jake ( @ Autism Investigated) responds to Mark's statements( @ Linderman Unleashed @ Natural News.com @internetradiotrashbucketoftripe.com) and
recounts Mark's many sins** -
he had COIs with pharma when he worked for that Boston firm; he still has COIs; he "trashed"** the Geiers; he said Andy's work wasn't "supported by epi data"; he betrayed Brian Hooker; he prevented the exposure of CDC malfeasance; he didn't publish Jake's article; he alludes to Jake's 'problem' indirectly; he un-friended Jake; $40,000 for a congressman, Jennifer, Ginger, Safe Minds, Canaries etc etc etc.
Fellow and sister sceptics:
Occasionally, I become very tired and rather morose, after all I have quite a few serious responsibilities in RL- (counselling people, managing money, assisting relatives, friends and associates) - and I attempt to follow and reveal the slithery, slimey, meandering paths of charlatans..it sometimes feels like I have the proverbial weight of the world upon my well-toned shoulders...
But every now and again, a woo-meister poisons himself with his own supplements or a conspiracy fanboi who shrieks "COI!" incessantly is shown to be rather compromised himself...
Jake, you make it all worthwhile.
And I imagine that in London an investigative journalist, somewhere in the states an actual epi as well as Orac and many, many Oracolytes worldwide, are laughing. Ha ha.
** how about that for a European film title-"The Many Sins of Mark Blaxill"?
*** that shouldn't be difficult
Greg:
Crowded refugee camps, a 32% un- and under- vaccinated population. Perfect conditions for the outbreak to become huge.
A high proportion of autistics can leave productive and fulfilling lives with the proper supports.
Once again, Greg, what evidence do you have that vaccines are responsible for these "other chronic disabilities"?
@Kerb
Two more questions Kerb: First, it is not quite clear, but is this a true longitudinal study in the kids were tracked from birth and at that time the final data was not know? Or did the researchers come after the fact --post Dec 99 -- and go back in the records to obtain figures. Second, why did the researchers start the autism-free years 'stopwatch' at one year and not at birth as you provisioned for in the example that you gave me earlier?
@Kerb
Kerb I also wanted to add that I thin this discussion at the Danish study is quite usefull and it may truly enlighten others. Pro-vaxxers often trumpets the study as stellar evidence of the no-link cliam, but how many of them -- or others for that matter --truly know anything about the study's design.
'can LEAD productive lives'
Le sigh.
And thus one obtains the answer to this question. All ASD diagnoses equate to "permanent disability"* (and point prevalence is retroactive).
It should be straightforward enough to compare this with actual Census or SSA numbers. As Gerg is prepared to rereduce the Madsen data set, this should be a piece of cake. I, however, have an appointment.
* For purposes of U.S. Social Security, this term doesn't exist.
I'm afraid that number is seriously inflated, Greg. About 1 in 88 children are being diagnosed as exhibiting an autism spectrum disorder, and "autism spectrum disorder" refers to a group of developmental disabilities of widely varying severity.
What chronic disabilities are these, Greg, and what evidence demonstrates that they are caused by vaccines? Be specific.
@Narad - once again, Greg refuses to answer my question posed above - to articulate exactly what percentage of the "spectrum" diagnoses correspond to "severe classic autism" and how that compares to the historical rate of "severely disabled."
Perhaps if we could (using, I don't know, math skills) he might be able to put some kind of credence behind the numbers he throws around (or might see how absolutely full of crap he actually is).
I mean "if he could" & the historical rate / incidence of individuals considered "severely disabled."
Because, in the case of this study, we're not seeing "tens of thousands" of regressive autistics - we aren't even seeing thousands....and we have method, data, etc. Which Greg has yet to provide any of his own (oh, that's right, he doesn't have any - he only has whatever AoA allows him to know).
@
"once again, Greg refuses to answer my question posed above – to articulate exactly what percentage of the “spectrum” diagnoses correspond to “severe classic autism” and how that compares to the historical rate of “severely disabled.”
Don't know if you will be satisfied with this, but a while back I provided CDC own figures showing that almost 40% of all autistics (ASDs) have an intellectual impairment (IQ less than 70). Imagaine the true figure for classic autism. Of the general population, only 5% have an intellectual impairment. In case anyone is confused about what 'intellectual imparment' means-- it was formally described as 'mentally retarded'.
'was formerly known as 'mentally retarded'.....
Greg,
Almost. Overall 82% of the children had MMR, but that's partly because some of them would have had MMR after the study period ended. MMR uptake at that time in Denmark was about 90% I believe. So they aren't looking for entirely unvaccinated children to compare with vaccinated.
You still seem to think that the use of person years is intended to correct for the confounding effects of age, but it isn't. Even using person-years the fact that both MMR vaccination and MMR vaccination will correlate with age means that unless you correct for this you will still find a spurious relationship between MMR and autism due to omitted-variable bias. Ask your statistician friend.
I think you mean 'prospective', not 'longitudinal'. It seems clear enough to me; beginning of the Methods section, "We conducted a retrospective cohort study of all children born in Denmark from January 1991 through December 1998", so it was the latter of your scenarios.
I'm not sure, but I hadn't noticed that when I constructed that example. Very few, if any, children would have had MMR or have been diagnosed before 1 year, so I don't think it would make much if any difference to any of the figures. Some problems due to prematurity, for example, are more common in infants so perhaps they wanted to exclude these.
Greg,
The CD got that figure from this surveillance report on 8-year-old children. They found a prevalence of ASDs of 1 in 88, and that 38% of these had an intellectual disability (IQ less than 70). That means that 1 in 232 8-year-olds may have autism and an intellectual disability (note that the intellectual disability may be due to other causes).
Does that mean that 1 in 232 children will grow up into adults with an intellectual disability? Firstly, ASDs are disorders of developmental delay not stasis, also many children with ASDs will find ways to deal with their disability. This study of nearly 5,000 adults with ID found that only 9% of them also had an ASD.
Most importantly, we know beyond any reasonable doubt that vaccines don't cause autism so even if your dire predictions about the tsunami of autism were true, stopping vaccinations would have no effect on it at all, except to bring back contagious diseases, some of which definitely cause not only autism (congenital rubella for example), but brain damage, deafness, sterility, cancer, death etc. etc..
Let me put that even more simply.
No vaccination = current autism rates + more autism + brain damage + deafness + sterility + cancer + death + other horrible sequelae of VPDs.
@Kreb - a resurgence of Rubella would definitely cause an increase in the number of cases of autism, since rubella has been definitely linked as a cause of autism.
Taylor et al @ BMJ Open (currently) plotted the number of year olds with autism in the UK between 2004-2010 and did NOT find a tsunami or a tidalwave or a really flood.
They found a level plain.
Oddly, it is dismissed at AoA. I wonder why that is, John?
...a really BIG flood...
remind me to sleep sometime, 'kay?
Remember Denise - there is no rest for the wicked.....
Convenient of you to mention this. Let's back up for a second:
Now, "1 in 50 kids in the US are permanently disabled from autism." Putting the most charitable spin possible on this, which is assuredly unwarranted, you have just asserted that 40% of 2% of "kids in the U.S." are "retarded" by virtue of being on the spectrum, or 0.8%.
You have one small problem, namely, the actual numbers (PDF) for that to which you directly refer.
In the "best" case scenario for you, you're off by well over a factor of 2.
@Kerb
This snippet was taken from the study......
"The mean age at diagnosis was four years and three months for autistic disorder and five years and three months for other autistic-spectrum disorders."
Kerb, look at the figures below. It shows the birth cohorts and how old the children would have been at the completion of the study -- Dec 31,1999.
Jan 91 - Dec 91 9yrs
Jan 92 - Dec 92 8yrs
Jan 93 - Dec 93 7yrs
Jan 94 - Dec 94 6yrs
Jan 95 - Dec 95 5yrs
Jan 96 - Dec 96 4yrs
Jan 97 - Dec 97 3yrs
Jan 98 -Dec 98 2yrs
Kerb, think about this for a second! By the completion
the follow-up period (Dec '99), the end-date of the study, two of the cohorts wouldn't have even reached the mean age of diagnosis. Overall, just reflect on the amount of kids that are being missed in this study. Kerb, seriously, am I the only one that is having a problem with this? With all these missed cases, how is it possible to obtain a trend or correlation? Incredible!!
Now Kerb, do you seriously want me to consider that this was not intentional? Yes -- I will admit that the design of the study is good but why such a short follow-up after the last birth cohorts? If they had to stick with the Dec 31, 1999 date then start with birth cohorts in the 80s for Christ's sakes. Actually, don't stick with the Dec 31, 1999 date -- move it to Dec 31, 2003. Heck -- move it to Dec 31, 2012! Incredible!
Now Kerb, don't come with the excuse that they were responding to anti-vaxxers' claim that MMR changed their kids in front of their eyes. This was nothing short of a deliberate effort to exclude cases.
@VCADODers
Let me put that even more simply.
No vaccination = current autism rates + more autism + brain damage + deafness + sterility + cancer + death + other horrible sequelae of VPDs.
No vaccination = drastic reduction in autism rates + next to no autism + significantly less SIDS, ADHD, asthma, diabetes, cancers, learning disability, allergies, and other current chronic childhood disabilities + minor spike in VPDs,with very few serious side-effects from them (think what is happening with our fight against HIV and AIDS).
Greg: Stop pretending to care about kids/people with autism. It's disgusting, especially when you then proceed to talk down to people like Julian and Alain. We all know you don't really care, so tone down the hypocrisy a bit.
Greg: (and this does not include the other chronic disabilities that vaccines are causing)
Buddy, I think you skipped over that handy dandy list I provided. And for the love of all that's holy, learn the difference between a disability, a disorder, an immune disorder, and cancer/leukemia/lymphoma.
Comment of mine in moderation.
I'm sorry, Greg, you haven't really made it clear what you're trying to imply here. Perhaps if you spelled it out in syllogism form...
The Trend in AIDS Death: What preventive public health measures and improved medicine can accomplish without vaccines.
http://www.thebodypro.com/content/art2522.html#patterns
In the 1980s and early 1990s, HIV/AIDS was a major cause of death among adults in the U.S., with the mortality (death) rate climbing every year from 1987 to 1994. By the latter year, the disease had become the leading cause of death among adults 25-44 years of age. A significant decrease in AIDS mortality first became apparent in 1996, the year after the first protease inhibitor (PI) was introduced. In 1997 the number of HIV/AIDS deaths fell by nearly 50%, followed by a further 20% reduction in 1998. By 1999, however, the decline had leveled off. According to the Centers for Disease Control and Prevention (CDC), HIV/AIDS deaths fell from more than 51,000 in 1995 to about 16,000 in 2002 (the latest year for which data are available).
"Perhaps the most evident change in the epidemic over the past 20 years has been the increase in the length of time people with access to good health care can expect to live after testing HIV positive or being diagnosed with AIDS"
"While the long-term toxicities of antiretroviral therapy remain an urgent concern, some experts now cautiously predict that HIV positive people who receive optimal care may ultimately live a near-normal lifespan"
Sorry -- at #298 the first paragraph should have also been in quotes.
Greg,
Just when I thought you were beginning to get the hang of this recipe, you have forgotten yourself, and poured the bowlful of ingredients all over yourself again. Here, let's get you cleaned up.
They studied every single child born in Denmark, that's 537,303 children, for 8 years, how many more children and for how much longer would you be satisfied with? They had to end the study at some point.
Remind me how many children Wakefield studied, and how long they were followed up for. The Geiers, Shaw, any of the other studies you antivaxxers get so excited about? Large numbers and time bring statistical power, and the ability to see small effects, if they are there.
So they only followed up over 400,000 children (6/8 of the birth cohorts) for long enough to diagnose autism and ASD, is that your objection?
Are you suggesting they should have carried the study on forever? Whenever the study ended, the children born towards the end of that period would not have been followed up for long. Surely you see that.
They might have had some problems publishing it in 2002 if they had done that, don't you think? Didn't you see date.the paper was published? It's right up at the very top -November 7, 2002.
Really, what is incredible is the stupidity of the statements you come out with to try to justify your delusional beliefs.
Yeah, 12 vs 400,000 - which sample size do you think is going to bring better results?
Just when you think Greg can't get any more vapid & stupid, he dives to new depths....
@Kerb
You said:
"They studied every single child born in Denmark, that’s 537,303 children, for 8 years, how many more children and for how much longer would you be satisfied with? They had to end the study at some point. "
They said:
“We designed a retrospective follow-up study of all children born in Denmark during the period from January 1, 1991, to December 31, 1998″
“Follow-up for the diagnosis of autistic disorder or another autistic-spectrum disorder began for all children on the day they reached one year of age and continued until the diagnosis of autism or an associated condition (the fragile X syndrome, Angelman’s syndrome, tuberous sclerosis, or congenital rubella), emigration, death, or the end of follow-up, on December 31, 1999, whichever occurred first”
Kerb, where am I to get from this that they studied all the kids for 8 years? If the kids were born in the cohort, say, Jan '94 to Dec '94 and followed to 'the end of follow-up, on December 31, 1999' that would've only been 6 years. According to the births, only 2 cohorts could have been studied for 8 yrs or more: Jan 91 - Dec 91 (9yrs) and Jan 92 - Dec 92 (8yrs).
Greg,
Every bit of my equation has solid evidence behind it. Let's take your version apart.
Leaving aside the mangled form of this, why would stopping vaccination lead to "next to no autism"? We know beyond a shadow of a doubt that vaccination isn't a major cause of autism from several studies - it is obvious from the NEJM study we have been discussing.
Vaccination doesn't cause SIDS.
Several studies have found no link between vaccination and ADHD, for example PMID: 15342824.
No evidence, in fact BCG vaccination seems to reduce the risk of a child developing asthma PMID: 19822573 and flu vaccination reduces the chances of a hospital admission with asthma by 76% PMID: 23859413.
No evidence, several studies find no link for example PMID: 22336803.
No evidence for any vaccine causing cancer. Liver cancer from hepatitis B and cervical and other cancers from HPV are definitely prevented by vaccines.
Multiple studies contradict this. A Dutch review study PMID: 15032089 concluded:
There is no convincing evidence to support a link between any of the conditions you list and vaccines.
There would be a huge increase in VPDs, why would you think there would not be?
I believe you live in the US and I'm guessing you are vaccinated so you're probably too young to have had chicken pox, but we don't vaccinate against it in the UK. Nearly every child gets it, with all the misery it brings. About 25 people die from it each year, mostly adults who missed getting it as a child. I dearly hope our government introduces routine vaccination soon.
A colleague of mine in his 20s got chicken pox. He was a healthy, sporty type, never ill, never off sick, but he was off work for 3 months when chicken pox got him. Why would the US be any different? By the way, people in the US who didn't get the vaccine and didn't get chicken pox are at serious risk from the disease - it is bad enough in children, but it is very nasty indeed in adults.
What about the measles epidemics we have seen in the UK and the rest of Europe? About 8,000 cases were reported across Europe last year, with several deaths and thousands of cases of pneumonia and encephalitis. Why would the US be any different?
Think of how many cases of rotavirus there were before vaccination, and of bacterial meningitis. These would return with a vengeance. Polio is transmitted the same way as other stomach bugs, so without vaccination there would be nothing to prevent a traveler from Africa or Asia starting an outbreak in the US.
What is different today from 1952 when the US saw the worst polio epidemic in American history? There were nearly 58,000 cases, 3,145 died and 21,269 were left with mild to disabling paralysis. In 1977 there were still 254,000 persons living in the United States who had been paralyzed by polio. In the 1950s Americans had flush toilets and refrigerators, knew about hygiene and were all eating healthy food straight from the farms, not all the processed rubbish people eat today, according to the CAM crowd. What, apart from vaccination, would prevent polio outbreaks like this from happening again?
I know people with AIDS don't face the death sentence they used to, but we still can't cure the disease. Ask anyone with AIDS if they would prefer to have had a vaccine, even one with side effects.
Have I been banned from commenting? A second comment of mine has now been eaten.
@Krebiozen - I suppose that if we didn't vaccinate then children who would later have been diagnosed with autism, ADHD, diabetes, cancers, learning disabilities, and the like might have died from the various diseases before being diagnosed. Perhaps that's Greg's thought?
No, probably not.
At least Greg hasn't started saying smallpox and polio are still around, they're just under different names. Also, we don't have a moon, NASA just made it up so they could fake the moon landing.
Greg: "The Trend in AIDS Death: What preventive public health measures and improved medicine can accomplish without vaccines."
Your "bodypro" website's stats are a bit dated. From the CDC:
"HIV Diagnoses (new diagnoses, regardless of when infection occurred): In 2011, an estimated 49,273 people were diagnosed with HIV infection in the United States. In that same year, an estimated 32,052 people were diagnosed with AIDS. Since the epidemic began, an estimated 1,155,792 people in the United States have been diagnosed with AIDS....Deaths: An estimated 15,529 people with an AIDS diagnosis died in 2010, and approximately 636,000 people in the United States with an AIDS diagnosis have died since the epidemic began."
http://www.cdc.gov/hiv/statistics/basics/ataglance.html
Take-home message: the death count from AIDS has not changed all that much from 2002 (the last year cited in your linnk), and there are still many people coming down with the disease each year and having to deal with its complications and the morbidity associated with medications (regardless of how much they've contributed to saving lives). Now think how much of that misery and death could be avoided if we had an effective HIV vaccine.
Are you really trying to justify the antivax belief that it's OK to allow people to get sick because meds are available in at least some cases to treat them?*
*the irony here is that antivaxers routinely accuse medicine of promoting vaccines as a profit center. Think about the lifetime cost of antiretroviral medications and all the other spending that goes into keeping HIV-positive people reasonably healthy. A vaccine would be one hell of a lot cheaper.
I must speak up to say that that is one of the most ridiculous analogies I have ever heard BECAUSE hiv/aids
1 is caused by a virus which has defied the many efforts to creat a working vaccine .. there have been multiple attempts and trials. It would be welcomed and used.
2 hiv is not spread by the modes of transmission for diseases like influenza and measles,i.e.casual contact. You can't get hiv if someone sneezes on you or if you use the same computer afterwards- it takes a lot more.( altho' other diseases spread by closer contact have had vaccines developed-for them hpv, hep).
It may be prevented by methods which can be *taught*
Countering hiv/ aids has been a decades long battle.The eGREGious one makes it sound like SBM found an easy non-vaccine solution and that it could do that for other illnesses.
No vaccine being developed for hiv was not the plan. Control of hiv is neither convenient nor inexpensive.
People on ARVs would love a vaccine.
@Dangerous Beacon
*the irony here is that antivaxers routinely accuse medicine of promoting vaccines as a profit center. Think about the lifetime cost of antiretroviral medications and all the other spending that goes into keeping HIV-positive people reasonably healthy. A vaccine would be one hell of a lot cheaper."
Actually, no it would not! If there were a vaccine for HIV, you drug-pushers would probably suggest that all kids get the vaccine (think Hep B). So really what would be cheaper -- medications for an extremely small percentage of the population that develops HIV, or requiring HIV vaccination for the entire population? And again DB, that would be one more vaccine to increase our 'shopping lists' of autism, SIDS, learning disability, asthma, diabetis, degenerative diseases in the elderly (oh - I did forget those in my last posts!), allergies, cancers, ADHD, deaths, and so and so on.
Greg,
You look idiot by asserting all kinds of diseases as being caused by vaccines. I won't bother asking for citations of all that because I know the list is false and that you are lying. It's not even funny to retort to that because I have to look very far and wide at finding any other reason why you'd behave that way because all I can find is that you look idiot.
Alain
Go ahead, do the math, Giggles. I'm willing to bet that world-wide HIV vaccination would be cheaper than world-wide administration and monitoring of the HIV positive population. Not to mention that taking ARV meds is no picnic, although they certainly extend health and lives. We have already given citations on vaccine safety that blow your "shopping list" all to hell. If you choose to ignore this, that's your delusion.
Oh yes, vaccination potentially could render HIV extinct. I say potentially, thanks to idiots like you.
Wow - again, he gets even more stupid.....over 30 million people have died since the AIDS epidemic began, guess that's just a blip to Greg (you know, with his other "First World Problems and all").
The overall cost of HIV / AIDS, beyond the medical costs, runs into the tens of billions of dollars in lost economic productivity alone, not to mention the damage to society in general (particular in 3rd World Countries).
I recently finished reading "Spillover" a book about the spreading of new diseases from animals to humans (Ebola and HIV being just two of dozens of recent examples) - it isn't a question of if the next big animal to human epidemic will occur, but merely when (and how successful we will be in containing it).
A lot of people don't realize how serious these diseases are - because even something like SARS killed 45 - 50% of the people who caught it.
Greg - which would do more to prevent the spread of the disease - medications for an extremely small percentage of the population that develops HIV, or HIV vaccination for the entire population?
Note that according to the CDC: More than 1.1 million people in the United States are living with HIV infection, and almost 1 in 5 (18.1%) are unaware of their infection. 207,600 people in the US alone have HIV and are unaware of it, so spread the disease unknowingly. There are 50,000 new infections in the US every year according to CDC estimates.
What's the maximum acceptable number for this?
http://www.cdc.gov/hiv/statistics/basics/ataglance.html
Consider as well that 17.3% of the adult population of South Africa is estimated to be infected with HIV. Does that change the math at all?
"Actually, no it would not! If there were a vaccine for HIV, you drug-pushers would probably suggest that all kids get the vaccine (think Hep B). So really what would be cheaper — medications for an extremely small percentage of the population that develops HIV, or requiring HIV vaccination for the entire population?"
Why do you folk invest so much time with this fool, who thinks that one shot or two shots or three shots would be more profitable to industry than chronic drug treatment for life? This Greg character has no grasp of even the most obvious fundamentals. He's just a jerk.
@MoB - I am being to sense that Greg may be an AIDS-denialist as well.....
Jeff1971,
A lurker said she had learned from my attempts to educate him, and I was tempted.
@Jeff - yes, he is a jerk.
M.O'B.,
Again according to the CDC:
Blockquote fail, but I hope it is clear anyway.
@jeff1971
Because it is a way to educate the lurkers and those on the fence the utter absurdity and falsity of greg's ideas?
I certainly have enjoyed the time that the regulars here have used to show how false greg's ideas are and how dangerous his thinking is.
@Krebiozen,
Trying some different numbers, let's assume that an effective vaccine would prevent 90% of the 50,000 infections/year. If we could vaccinate everyone of a certain age for less than US$17,085,960,000/year, that would be break even. Assume we want to vaccinate people at age 14 and that there are about 4 million 14 year-olds in the US (a totally arbitrary age and a rough approximation). As long as the total cost of the vaccine (when and if) were less than US$4,271 it would be more cost effective to immunize than to treat those who became infected.
Now that's an interesting claim to make. Greg's syllogism would seem to be:
1) There is a magic ingredient or ingredient(s) in current vaccines which is known to cause "autism, SIDS ... and so on and so on".
2) Greg knows so much about the making of vaccines that even though no one has made an HIV-preventing vaccine yet, he is 100% sure that any such vaccine which is made at any point in the future must contain that Magic Ingredient of Evil.
3) Therefore any HIV vaccine which ever gets made will cause "autism, SIDS ... and so on and so on."
Now I would like to ask Greg, the vaccine expert, just what that Magic Ingredient of Evil is, because I certainly don't know any chemical which fits that bill, but he has just claimed to. In fact, I think it might be good to make it an ultimatum question: Greg must give his answer within the next three comments he makes, on this or any other thread, or he will be taken as answering by default that he knows of no such chemical; he was just lying because he thought it would serve his purposes.
Also, the CDC is recommending that all baby boomers get tested for hepatitis C, on the theory that at least some of us shared needles and/or had unsafe sex when we were young and don't remember. Or don't want to admit it, but might say okay if a doctor looks at their chart and says "based on your age and records, you should have a hepatitis test and a colonoscopy. The hepatitis test is the easy one." By that same logic, we're all a higher-risk group for HIV.
M.O'B.,
When different approaches to the same problem give you much the same answer, I find it so much more persuasive. If only we had an effective HIV vaccine. Who would have guessed that a virus with just 9 genes would prove to be such a bastard to deal with?
Just wanted to thank Kreb and the gang for the clear explanations of a proper study is done.
By the way, you all might be amused by this comment (http://leftbrainrightbrain.co.uk/2013/07/15/jenny-mccarthy-angling-for-…) by an antivaxxer on LBRB - she has figured out why the parents of autistic children reject antivaccine ideas. Apparently, we're
"borderline autistic yourself and that’s why you have this nerdy obedience to what the borderline autistic “scientists” say. " Who knew?
@DB
Seriously DB, if tomorrow they came out with a HIV vaccine do you think a parent should comply and have their kids get this vaccine just like with Hep B? Let me ask the question another way: At what point do you think it's unwise to recommend an ever expanding cocktail of drugs for our kids?
@Kerb
I was prepared to apologize to Madsen if indeed his study was sound after I suggested it was BS. Upon considering the short follow-up for the younger cohorts, I will offer no such apology. Did you her that, Kerb? I am not going to apologize. Further, I will inform the 'quacks' at AoA that they should have every reason to be skeptical about the study.
@ Antaeus:
That list sounds like the Canary Party party line.
Do you think that it is a single arcane ingredient or a set of substances that conspire together synergistically/ symbiotically/ synchronistically? Janine Roberts has a little list of the evil ingredients, I understand.
Time for lilady's Media Review of Anne Dachel's Media Review:
http://philadelphia.cbslocal.com/2013/10/31/mothers-dream-to-create-lif…
Greg: "Seriously DB, if tomorrow they came out with a HIV vaccine do you think a parent should comply and have their kids get this vaccine just like with Hep B? Let me ask the question another way: At what point do you think it’s unwise to recommend an ever expanding cocktail of drugs for our kids?"
If there was a highly effective HIV vaccine, then yes, responsible parents would want their kids to receive it. After all, intelligent people don't want themselves or their kids exposed to an ever-expanding cocktail of retroviral proteins produced by HIV infection. That plays hell with the immune system, unlike vaccines which provide a temporary challenge to a minute fraction of the immune system.
You could look it up (but you won't find it on www.toxinsundermybed.com).
Greg,
I don't believe you ever had any intention of doing any such thing. You came to this study convinced there must be something wrong with it a) because it came to conclusions that you have already made your mind up are wrong and b) because you don't understand the math involved.
I'm sure he will be heartbroken that an innumerate antivaxxer has failed to understand his study in such spectacular fashion. You are clutching at straws with this, and in a very obvious and rather pathetic fashion.
You don't seem to understand that if you follow a cohort of children for a fixed period of time, some of them will be younger at the end of the study. You have to have an end date for the study - it can't go on forever. You can't publish a study in 2002 that follows up children until 2013 as you suggested, and even if you did, the children born after 2010 would be too young, so you would have to extend the study another 2 years, but then the children born after 2012 would be too young so you would have to extend the study another 2 years... With your logic you would never end the study.
If this had been a study of only 250,000 children born between January 1, 1991 and December 31, 1994 , and all the children had been followed up for a further 4 years, until they were all 8 years old or older, would that have satisfied you?
This is exactly what Madsen did, with the addition of also following the children born between January 1, 1995 and December 31, 1998. Madsen looked at each age group separately, as well as combined, so I fail to see what you are complaining about. I think you are making something out of nothing just so you don't have to admit you and your AoA buddies were wrong about this study.
That's sad, for you, because I truly don't care if you apologize or not. I did think, briefly, that you were going to make the effort to honestly understand this, but it doesn't surprise me that it turns out your prejudices are too ingrained, and you are unable to open your mind sufficiently to see the truth.
What about your statistician friend? Aren't you going to ask him about the study design? You do have a statistician friend don't you?
If you do so, you will be deliberately and knowingly lying to them. I have to say I am a little disappointed, but not surprised, by your lack of any honesty or integrity in this area.
Will you also inform them that Gilmore's article was very badly wrong in several important ways, that I explained above? That there are no "life-years with autism" in the study, and that the term "life-years" does not appear in the study? Do you possess even the tiny sliver of honesty and integrity that would require? I bet you don't, and I bet that even if you did it wouldn't get through moderation. Why don't you find out for yourself how honest the AoA antivaccine camp is.
@Kreb - you have performed yeoman's work here & explained the details of the study in a very straightforward and easy to understand way....for the lurkers here, I'm sure they appreciate the time that you've taken.
Greg, on the other hand, never intended to understand this study - only trying to find one or two nitpicky things he could complain about to attempt to invalidate the entire design (which he has been unable to do - mostly because he doesn't even understand basic statistics).
For the lurkers among here, a vaccine != a drug (in the classical sense), a drug metabolise in the liver while a vaccine target the immune system. Different goals too.
Alain
@Kerb
"If this had been a study of only 250,000 children born between January 1, 1991 and December 31, 1994 , and all the children had been followed up for a further 4 years, until they were all 8 years old or older, would that have satisfied you? "
Actually Kerb, that would've been a start. If the end-date, of the study was the same as the Danish study (Dec 31, 1999), that would mean kids born as late as Dec 31, 1994 would've been around 5yrs at the end-date -- pass the median age of autism detection.
I supposed you may argue that, overall, there would have been less kids in that study than the Danish sample. Yet Kerb, what stopped the researchers from going back even further in the records and, say, also look at kids born in the 80s?
Finally Kerb, I have few more questions. We know the end-date of the study was Dec 31. 1999, and it was published in 2002. Around what date was the actual study done -- we are talking the actual number-crunching period? Also Kerb, in your estimate (maybe it will only be crude), how long would you think it should take to do the number-crunching and produce the results for studies of this nature?
@DB
Greg: “Seriously DB, if tomorrow they came out with a HIV vaccine do you think a parent should comply and have their kids get this vaccine just like with Hep B? Let me ask the question another way: At what point do you think it’s unwise to recommend an ever expanding cocktail of drugs for our kids?”
"If there was a highly effective HIV vaccine, then yes, responsible parents would want their kids to receive it. After all, intelligent people don’t want themselves or their kids exposed to an ever-expanding cocktail of retroviral proteins produced by HIV infection. That plays hell with the immune system, unlike vaccines which provide a temporary challenge to a minute fraction of the immune system. "
---------------------------------------------------------------------------
You still did not really answer my question. At what point will an ever-expanding cocktail of drugs (aka vaccines) become more than 'a temporary challenge to a minute fraction of the immune system'? If there is such a point, maybe you can also enlighten me how it is determined.
@Greg - whatever study decisions were made, were made most likely based on the availability of records. When did Denmark start their centralized repository of pediatric medical records? If the information wasn't available from the same source from the 1980's, it wouldn't have made sense to go back that far.....utilizing the readily available and comprehensive medical records would guarantee the best possible results (either way) - eliminating yet another area of potential discrepancy (i.e. pulling incomplete data from multiple sources).
Every study has a start and end date - they just don't go on forever (though you would seem to prefer it that way - or at least until you think you'll get the results you want).
And since you're running statistical models against readily available data, the actual number crunching shouldn't take that long at all - the most time-consuming part would be checking and re-checking your results to make sure they are accurate and that the study parameters are valid - and to get published, those results do need to be validated.
Again, I'm sorry that you lack a basic understanding of statistical methods, research, and study design.....because it pretty much means that everything Kreb has tried to explain to you (over and over and over again) has gone completely over your head - and you wouldn't accept the results anyway, because they don't match up with your preconceived ideas.
I will try to answer two of Greg's questions.
And...
Time and money constraints would be my guess. Seriously. The larger your sample, the longer and costlier the information gathering, and number crunching, phases, will be. The sample size was over half a million. At that point, the sensitivity would be phenomenal. The reason it took over 2 years between the end date and publication would be due to the sheer quantity of data.
Greg:
As Alain has pointed out, vaccines aren't drugs. Secondly, given the antigen load, even 10,000 vaccinations wouldn't really be more than ‘a temporary challenge to a minute fraction of the immune system’. Such a point would be determined by antigen load, I believe.
@Julian - Greg obviously has no idea what actually occurs within the human body when confronted with an actual disease - the large amounts of antigens that make up a typical infectious agent and how the body responds (and how massive that response can be)......but, since he seems like even a basic knowledge of Science (above and beyond what AoA tells him) - I don't hold out any hope that he'd even begin to understand the biological processes involved.
@Julian
"Secondly, given the antigen load, even 10,000 vaccinations wouldn’t really be more than ‘a temporary challenge to a minute fraction of the immune system’. Such a point would be determined by antigen load, I believe."
And, adjuvants?
@Kerb
Since Julian and Lawrence responded to my few last queries about the Danish study, I won't need you to respond. Thank you for your time explaining the study to me. I will prepare a report about our discussion and my assessment of the study, which I will share with you guys and AoA.
Kerb finally, conversing with you really got me thinking. It is obvious that you are brilliant with numbers and statistics. How do you live with the dissonance knowing that this talent that at one time you likely proudly used to determine facts and uncover truths is now being used in the most sinister manner of suppressing truths and facts?
And Greg has now admitted - AGAIN - that he was just telling lies to serve his purposes when he claimed an HIV vaccine would cause autism and SIDS and diabetes. It's funny, vaccine advocates are often accused of "fear-mongering" for telling the truth about how deadly VPDs can be, but AVers think they're entitled to lie through their teeth, and completely invent consequences to vaccines that don't even exist yet.
@ Antaeus:
Note @ 341/ last paragraph.
It illustrates that Gergeh invents stories out of whole cloth
because:
1. he is not in a position to evaluate talent or proficiency
2. he can't get a simple 4 letter 'nym/ nickname correct
3 he pretends to know what another person he has never met does, knows, feels
3 he assumes that Krebiozen is a person when it is entirely likely that he is acomputer programmed to instruct the un-educable and nicknamed "Higgins"- or an android like me.
Oooops!that's @ 340.
When we run out of dangerous diseases they offer protection against.
Vaccines aren't develped or added to the routine childhood vaccination schedule just for fun, after all, but to address an unmet medical need to reduce incidence of infection.
RE: break even point for an HIV vaccine, figures based on the the lifetime treatment cost of an HIV infection don't reflect other economic costs associated with HIV infection (lost productivity, for example) so the true break cost should be even greater than $4,271.
Greg: "You still did not really answer my question."
This is antivaxer code for "I refuse to acknowledge and make an effort to understand what you're saying".*
Greg: " At what point will an ever-expanding cocktail of drugs (aka vaccines) become more than ‘a temporary challenge to a minute fraction of the immune system’? If there is such a point, maybe you can also enlighten me how it is determined."
As the following article by the esteemed Dark Lord of Vaccines makes clear, that "point" does not and could not exist.
http://www.chop.edu/service/vaccine-education-center/vaccine-safety/vac…
*As to your refusal to acknowledge immune system load from infectious diseases, here's another article for you to ignore (it compares antigenic stimulus from full-blown infectious disease to that from vaccines. It does not help your argument. Run awaaay!!!).
http://www.sciencebasedmedicine.org/the-infection-schedule-vrs-the-vacc…
Incredible!! So it's ok to lace up a newborn with unlimited amount of aluminum, mercury, and other like 'baby-friendly toxins? As this 'safe continual build-up of vaccines' position been adequately tested? Nevermind -- perhaps that will also amount to an excursion in futility. Anyway, I must be on my way shortly. I will provide my report on the Danish study, and also address 'my friend', Ms Walters. Still guys, I have one more question: Does anyone know Chris' sex? S/he won't tell me. Chris what sex are you? Ok -- ok-- I will ask you a question.
Chris, what is the name of your best friend?
'Bobby'? C'mon Chris!
Chris, how do you refer to your spouse?
'As your SO'?
Ok, ok Chris, what's your favourite store?
'The Gap'?
Aaarrgghhh!!!
#346 '(H)as this 'safe....
Greg: Some people, myself included, prefer neutral 'nyms. Inquiring about gender is bad manners- among other courses, you should sign up for an internet etiquette course.
Wow, that's ultra-creepy. Yikes.
kanji - Greg is paraphrasing a routine from Saturday Night Live from some years ago.
Greg - why is Chris's gender (which was in fact disclosed by Chris some time ago in another thread) significant to you in this thread?
Sorry, khani, autocorrect did me in.
Greg,
Good. I would have thought it obvious that the amount of time and energy involved in processing over half a million children's records is huge, and that its statistical power is also very large. Why the need for a larger or longer study?
Madsen's study included 316 children with a diagnosis of autistic disorder and 422 with a diagnosis of other ASDs, that's 738 children with ASDs in all. Please explain to me why those at AoA still believe that Wakefield's disgraced and retracted study of 12 children with ASDs is more reliable than Madsen's study. Please, think about this.
You have given me no reason to believe you have understood anything about this study. Whatever makes you think you are a suitable person to prepare a report on it?
I'm not particularly good with this stuff, there are many people who comment here who know their way around statistics and epidemiology far, far better than I do. What I have been trying to explain to you is really very basic.
I don't believe this is what you really think. I'm pretty sure you are playing a perverse game for reasons I don't understand. Every time you have appeared to be close to understanding why this study was carried out the way it was, you have come out with something very foolish indeed, almost as you have sort of involuntary reaction to anything that challenges your prejudices (the analogy I used of you emptying out the mixing bowl while i was trying to teach you a recipe isn't far from the mark).
Perhaps this is because you feel uncomfortable whenever you begin to understand that you are very, very wrong about autism and vaccination. I do understand that cognitive dissonance can be very uncomfortable.
I think it is more likely that this is just part of your game of 'bait the skeptic'. Either way, it doesn't help anyone get any closer to the truth about vaccination and autism, which is what I am interested in. However, just in case, I will assume that you are sincere.
I looked long and hard at this issue with an open mind before I got involved in skepticism, years before I started commenting here. An acquaintance I trusted drew my attention to the vaccination debate - she was very much in the antivaccine camp, which I didn't even know existed at that time. So, putting on my white coat and wielding my sword of reason I took a good look at it.
What I found was a great deal of fact-twisting and outright dishonesty from those opposed to vaccination, which at first shocked and then annoyed me. Worse, I found that when I tried to politely point out inaccuracies I was attacked or simply censored. It was my outrage at the lies being told about vaccination that led me to finding this blog and other people who are similarly upset at the lies being spread about vaccination.
There is no doubt in my mind that it is AoA and other antivaccine organizations that are suppressing truths and facts in the most sinister manner. Just look at Gilmore's article, and the deliberate lies he told in that - I can't believe they are mistakes, as all he had to do was read the paper to see that the terminology he complained so bitterly about doesn't even appear in it. He knows that the vast majority if those reading his article will simply and uncritically accept what he wrote as true, reinforcing their preconceptions that scientists are covering up the truth about autism. The truth is that the study was well-designed, and Madsen did everything I would expect an honest researcher who was trying to get to the truth to do.
By supporting AoA, and by continuing to misconstrue this study as dishonest, inadequate or in any way attempting to cover up the truth, it is you who are suppressing truths and facts in the most sinister manner. I think you should have a long hard think about that.
@MOB
You're good -- recognizing the routine!
Could you point me to the thread?
Kreb:
My feelings exactly altho' I came upon anti-vax mendacity through general scepticism, being aghast at alt med including hiv/aids denialism- one of its most dangerous forms... there's a place called South Africa...
As you might expect in this kingdom of un-intentional irony, a frequent meme is that SBM *suppresses* the truth and any research that shows its own inefficacy and the superiority of woo-
over the past few weeks this has been wrenched from its mouldly,unsanctified grave and re-animated by SD Wells of Natural News and re-posted and read live @ PRN.
If you think about it, that's AoA's main theme: S
BM's research is entirely fabricated and anti-vax truth is covered up ( see esp Jake Crosby's tussle with Mark Blaxill about the suppression of data about hg) ...
whilst they breathlessly support and laud fraudster par excellence Andy and shriek about Brian Deer, Paul Offit, Orac et al.
Repeating their lies without question doesn't make anyone look trustworthy.
#350, #351
Oh, I recognize "It's Pat."
However, it is ultra-creepy that Greg has a fixation on Chris's gender, and if it goes on, if I were Chris, I would be printing those posts out and making a file for the police.
This isn't the first time Greg has asked that, either, despite it being totally irrelevant to the conversation.
Khani,
I guess I didn't see Greg's current fixation as ultra-creepy so much as part of his standard "keep asking the same question repeatedly despite the fact that it's a silly question and it's either been answered or I've been told it's not going to be answered" style. I can think of much creepier questions that Greg could have fixated on which I will not share in this forum.
As I've mentioned before, I think Greg likes to think of himself as a charming bully. Repeatedly asking questions when he's been told he won't get the answer is his form of taunting.
Except he isn't charming, and he's not succeeding in bullying us.
Julian Frost - true enough. I didn't say he was a success at it.
I doubt Greg thinks of himself as bully. That wouldn't be flattering to him. I suspect he thinks he's educating us by being charmingly provocative - except that, in my perception, only the second of those two words applies.
@Khani
"However, it is ultra-creepy that Greg has a fixation on Chris’s gender, and if it goes on, if I were Chris, I would be printing those posts out and making a file for the police."
Khani, you seriously need to chill! Can't a person be curious about something? Look I don't personally know any of you, and seriously I wouldn't want to because in my honest opinion you have taken the most despicable, unconscionably position of supporting the deliberate, and sustained harm of children. You are not nice people.
Now, do I see myself as a 'charming bully'? I don't know -- perhaps. What I do know for sure is that you guys are disingenuous and honest debate is futile.
Greg: Look I don’t personally know any of you, and seriously I wouldn’t want to because in my honest opinion you have taken the most despicable, unconscionably position of supporting the deliberate, and sustained harm of children. You are not nice people.
Are you familiar with these things called mirrors? Your whole comment sounds like a description of your crowd..
Greg @361
You do excel at projection.
You are curious about someone's gender. That is not an appropriate thing to ask about. If a person has a gender-neutral pseudonym, asking about his/her gender can be considered rude.
You have NO EVIDENCE to support your assertions that vaccines cause more harm than they prevent.
And my latest hypocrisy meter goes thermonuclear.
Strange, I was going to say the same about you. Because of your fact free views, you want to reintroduce diseases that are currently at substantially lower rates than they would be without vaccines. You want to cause millions of people to get sick every year, and you want to cause multitudes of cases to temporary and permanent paralysis, permanent neurological disorders, heart problems, developmental disorders, and, yes, deaths every single year. You want people to live in fear when an epidemic comes around that could have been prevented by a vaccine. You want children to cough so much that they can scarcely breath. All because you believe what all the current best data shows to be untrue. You should be ashamed - but you aren't because you don't care if people die.
Oh, I forgot one - you want to increase the spread of rubella, which causes birth defects and - wait for it - autism spectrum disorders! So you want to cause the very harm you accuse people here of supporting. Well done, sir.
I think most people who know me would say I am basically a kind person. I freely admit, however, that I can be nasty to those who promote their harmful delusions for their own selfish reasons, in the face of massive evidence otherwise. So, yes, to you I am not a nice person - to my credit..
Not only is "what's your gender?" a rude question--if someone wants you to know that, they can and will say so--but you know exactly as much about Chris's gender as you do about mine or Mephistopheles's or Julian's (I know two Julians, one female and one male) or Gray Falcon's. None of which are actually your business, and if they were your business, the people would *tell you*. Also, the fact that you are quoting an old transphobic comedy skit doesn't justify harassing Chris.
Sure, except that you seem to feel yourself exempt when the object of curiosity is your grasp of something that has been explained to you repeatedly. Here, Gerg, answer a question:
Suppose you have two measurements, say, regarding autism diagnoses. The first case was officially diagnosed at age 4, and the second at age 5½, but there's a lot of home video, and a review of this places the true onset at 3 ± 0.5 and 4 ± 0.7 years. Further assume that both were unvaccinated aside from a single dose of the same vaccine permitted as a result of parental bullying precisely at age 2. Missionary activity, perhaps.
Now, proceed on the basis that this is the causal factor. What is the mean time to manifestation of the symptom? Please express your answer to two significan digits.
This strikes me as much more interesting than your sudden revival of interest in the contents of Chris's pants as a distraction.
@Chemmomo - Perhaps I'm using the wrong word and would hope that a mental health professional could supply a better one. I honestly believe that Greg really, really wants people on this forum to read some of his posts and cry. I'm talking big sobs, lots of tears, kicking and screaming kind of crying. Failing tears, he would like people to get so mad they would (metaphorically) like to punch him in the snoot. He would like to do this by benefit of his superior argument, but he is not above using insults and badgering to accomplish this.
Naturally, if someone informs him that he has provoked them beyond reason he tries to show his "charming" side - quotes from popular culture, explains that he was only pulling their legs, and so on. However, it is at that moment that he gets his validation. He has provoked someone, they reacted emotionally, and he got to say "why are you so thin skinned? I'm just funning you."
P.S. I have never reacted emotionally in one of my few responses to Greg's comments.
@Vicki
I will leave this entire 'what's your gender' discussion with this last comment. Please review the thread, 'Chris beat cancer.' It was Chris that brought up the topic that I was commenting without knowing the gender of Chris, the subject of the article. I ignored her (yes - with the way you guys are getting so testy about the issue I have to conclude that Chis is a she-- way to go, you screwed up --hee hee hee) at first, and then she brought it up again. I then decided to go with it and made a comment later to the extent of, 'for what it's worth Chris, I don't know your sex."
As MOB mentioned, it was obviously not a big secret what Chris' gender is, so I didn't see the harm in inquiring. I supposed I could search the threads but to be honest, I am too lazy to do so. Again, Chis' gender only became an issue after Chris brought up the topic in the first place.
Finally, in the big scheme of things I don't care if you are guys, girls, or in between. What stick for me is that you are all disgusting, lying, unscrupulous pharma shills. That I will hold against you.
'I suppose...
Oh come on Greg, tell us how you really feel.....lol
I'm still wondering what color the sky is in your world, because it sure as hell isn't the real world.....
@MoB - Greg is a troll...no more, no less....I actually feel a little sorry for him that he needs to come here for validation and get shown how little he actual knows about the topics at hand.....
At the end of the day, he's just a sad, sorry little man....whom I laugh at, a lot.
#356 MOB
Probably I'm thin skinned about this because the last time I witnessed someone with a similar obsession, the victim's car was mysteriously vandalized.
I'm sure Orac will have access to the IP number just in case anyway.
Greg @ 369
You’re on record now (my emphasis)
http://scienceblogs.com/insolence/2013/10/14/antivaccine-odds-ends/#com…
No more calling for “honest debate,” or pretending to misunderstand questions.
You show no respect for anyone here, and you are posting only to be a pain in the ass.
What in G-d's name are you babbling about? The only thing that "came up" was that you were too stupid to successfully read the title of the post.
Greg,
Now, now, it's not our fault you screwed up your lessons so badly and made such a terrible mess. Lashing out and being so rude when we have all made an effort to help you isn't very nice, is it?
Perhaps you really do believe your bizarre assertions about autism and vaccines, in which case I can only conclude you have "a belief held with strong conviction despite superior evidence to the contrary".
If you don't really believe your assertions, and I truly don't understand how anyone in their right mind could look at the available evidence and come to those conclusions, why are you playing this perverse game? This is something I find far more interesting than what gender any commenter here happens to be.
Kreb and other VCADODers here is my summary on the Danish MMR study, as I mentioned early, I will share this summary with AoA.
Recently an article that I saw here at AoA caught my attention. The article pertained to the Danish MMR study and claimed that based on the figures of the study there was indeed a link between MMR and autism. It was added that this finding was 'craftily' concealed by a creative way of counting autism cases in the study.
http://www.ageofautism.com/2013/09/round-2-cdcs-poul-thorsen-lying-in-p…
Given that the Danish MMR Study is often presented by pro-vaxxers as such a ‘stellar’ piece of 'studies find no link' claim, I decided to solicit the other side for their take. I approached Orac and his minions at RI.
I conversed with Krebiozen and you can find our entire exchanges under the, ‘Antivaccination movement and its quest for legitimacy thwarted’ thread. Krebiozen explained that the calculations were off due to AoA’s John Gilmore calculating rates in the ‘standard way’. This standard process involves taking the total number of vax/unvax autism cases and dividing them by the number of subjects in each group to come up with the rates.
Kreb, explained that with the Danish study of 500,000 plus kids born through the period of Jan 1, 1991 to Dec 31 1998, calculating figures in this way will result in inaccuracy due to two factors. First, the sample of kids included a selection of kids at certain ages or cohorts (0-1 and 1-2 yrs) that did not reach the Danish MMR uptake age of 15 months. Their presence in the study unfairly biased things in that if we were to assume that autism develops by chance some of these kids would naturally go on to develop autism, and likely when they were vaccinated. Autism after MMR in those cases would have been sheerly coincidental and not causal.
Perhaps the following example will better illustrate the effect:
---------------------------------------------------------------------------------------------------------
In this example we have 1000 kids in four different age cohorts. We are also assuming that autism will naturally occur at 1% rate, and the corresponding autism cases are given.
0-1 0 1000 0
1-2 600 400 10 (6/4)
2-3 900 100 10 (9/1)
3-4 900 100 10 (9/1)
4-5 900 100 10 (9/1)
In the first cohort (0-1), the kids are too young to be vaccinated, so the full thousand are counted as unvaxed. In the next cohort (1-2) some of the kids have reached the vaccination age of 15 mths, so we hypothetically assume 600 are vaxed and 400 are unvaxed. The other cohorts have all the kids’ reaching the vaccination age and it is assumed that they will be vaxed at a 90% rate.
The important thing to notice here is that if we total all the autism cases and divide the number of the total sample of vax/unvax kids, we will never get accurate overall rate. The kids in the younger cohorts that will eventually go on to be vaccinated at the 90% rate are skewing things in favour of a higher vax rate.
--------------------------------------------------------------------------------------------------------
The obvious question then is how do we deal with this problem? Well, the most sensible solution would be to remove the two younger cohorts from the study.
The second factor that Kreb pointed out, which also follows on the same logic as the first, is that a ‘straight-forward’ calculation of overall rates among different ages of kids will always result in wrong figures due to autism and vaccination increasing with age. The younger kids that are vaccinated less, as they age will go on to get vaccinated, and at an older age autism is more likely to be detected in them. Autism then after MMR, again, would simply be coincidental and not causal. I must concede that I had a tough time with this concept – so much so that on several times Kreb questioned my mental capacity and whether I was truly and ‘idiot’. (They can be such nice people!) Perhaps another example will best demonstrate the effect:
---------------------------------------------------------------------------------------------------------
We have two age cohorts. In the ‘2-3’ cohort some parents have not vaccinated their kids, yet, perhaps due to delay and neglect. Autism in this cohort, because the kids are so young, is going undetected at a high rate. We then assume a rate .5% -not 1%.
2-3 600 400 10 (3/2)
3-4 900 100 10 (9/1)
As the kids age we see how more of them are becoming vaccinated in the ‘3-4’ cohort. Autism detection is also increasing, so overall there are more cases. And calculating the rates we have (12/1500) .008 vaxed rate, and a (3/500) .006 unvaxed rate. We would expect the rates to be the same by chance, but they are not and the vaxed rate is higher!
-------------------------------------------------------------------------------------------------------
In the preceding example, we see how even though the overall rates across age groups will be skewed towards a higher vax rate; the age-to-age rates do not suffer from this bias. This fact prompted me to consider whether the age-to-age rates using the standard, ‘straight-forward’ method of calculation was done. Kreb, however, did not seem at all eager to entertain this issue. And when I pressed him about the raw data of the study and the possibility of still doing such calculations, making sure to eliminate the sample of kids that were to young to be vaccinated, he told me that this data is no longer available on the Internet. I quote him…
----------------------------------------------------------------------------------------------------------
Kreb:
Incidentally, I asked Dr. Madsen if his raw data are available on-line anywhere, but he told me he is no longer involved in that area and sad no longer has access to the data.
Even if I did put the data into a form that Gerg could understand, written in crayon in words of one syllable, I have no doubt at all he wouldn’t accept it. His mind is welded shut against anything that doesn’t fit with his crackpot beliefs.
-------------------------------------------------------------------------------------------------------
So, this all brings us to the ‘million dollar question’, how did they calculate the rates in the study?
Well, the Danish MMR study is actually a retrospective study –(I did not initially think of it as such). Research came after-the-fact around 1999 and went back in the records of the kids born during the period of Jan 1, 1991 to Dec 31, 1998. They followed each child from birth, and using ‘life-years’ measurements, they tracked each child to Dec 31, 1999 to see if they developed autism. A ‘life-year’ is determined as 1 year of a person’s life.
I find it most helpful to think of the methodology in terms of as each child is born and they started the ‘autism-free life-years’ stopwatch. They stop the watch whenever the child developed autism. If the child did not, they ran it until the end follow-up date of Dec 31, 1999. Whatever ‘autism-free life-years’ time they obtained, they divided it between how long the child was vaccinated and not-vaccinated. They then compare the total autism-free life-years between the vax and unvax groups to see if there were a correlation.
In fairness, there are actually some sound benefits to this study design. Studying children’s individual records provides a certain degree of precision and accuracy that is not afforded by simply summing figures for a group. The counter-side of this benefit is that it’s not friendly to verification. If someone suspects that a figure may be incorrect, they may have to re-check all 500, 000 plus kids’ figures. Is it likely that they would have fudged the numbers in this way? Even though I would not put that pass them, I have trouble believing that they did. Fudging the numbers involves blatant dishonesty that runs the risk of being easily exposed should someone snitch down the road. They are usually craftier in their deceits.
The other benefit of the study is that it is quite informative. Not only is it able to determine if there is a correlation between MMR and autism --which it did not find-- but it also offers other findings. It’s able to determine if there was a clustering of cases around the time of MMR uptake, and thereby further implicating MMR. The study did not find this. It’ s also able to tell us if there was an age association between MMR uptake and autism. For instance, are younger kids given MMR more likely to develop autism? The study, again, did not find such a correlation.
In all, I would have been inclined to conclude that it was indeed a good study if, later, I did not stumble on a very serious irregularity. Reading the study, it explained that…
--------------------------------------------------------------------------------------------------------
http://www.nejm.org/doi/full/10.1056/NEJMoa021134#t=articleMethods
‘The mean age at diagnosis was four years and three months for autistic disorder and five years and three months for other autistic-spectrum disorders.’
---------------------------------------------------------------------------------------------------------
Reviewing the chart below, we see the ages the children would have been at the end follow-up date – Dec 31, 1999.
Jan 91 – Dec 91 9yrs
Jan 92 – Dec 92 8yrs
Jan 93 – Dec 93 7yrs
Jan 94 – Dec 94 6yrs
Jan 95 – Dec 95 5yrs
Jan 96 – Dec 96 4yrs
Jan 97 – Dec 97 3yrs
Jan 98 -Dec 98 2yrs
So, clearly two of the age cohorts would not have even reached the median age of 4 for autism diagnosis. And, four cohorts would not have also reached the median age for diagnosis of other autistic-spectrum disorders –pass the age of 5. This represents a lot of cases that were potentially missed. Missed cases in any study are not friendly to determining trends or establishing correlations. This situation is extremely disadvantageous for our side.
I did bring this peculiarity up with Kreb and his reaction was the following:
------------------------------------------------------------------------------------------------------------
Kreb: Just when I thought you were beginning to get the hang of this recipe, you have forgotten yourself, and poured the bowlful of ingredients all over yourself again. Here, let’s get you cleaned up.
Greg: Overall, just reflect on the amount of kids that are being missed in this study. Kerb, seriously, am I the only one that is having a problem with this? With all these missed cases, how is it possible to obtain a trend or correlation? Incredible!!
Kreb: They studied every single child born in Denmark, that’s 537,303 children, for 8 years, how many more children and for how much longer would you be satisfied with? They had to end the study at some point.
---------------------------------------------------------------------------------------------------------
Clearly Kerb was implying that each of the cohorts was followed for eight years and this just was not the case; so I had to correct him:
------------------------------------------------------------------------------------------------------------
@Kerb
You said:
“They studied every single child born in Denmark, that’s 537,303 children, for 8 years, how many more children and for how much longer would you be satisfied with? They had to end the study at some point. ”
They said:
“We designed a retrospective follow-up study of all children born in Denmark during the period from January 1, 1991, to December 31, 1998″
“Follow-up for the diagnosis of autistic disorder or another autistic-spectrum disorder began for all children on the day they reached one year of age and continued until the diagnosis of autism or an associated condition (the fragile X syndrome, Angelman’s syndrome, tuberous sclerosis, or congenital rubella), emigration, death, or the end of follow-up, on December 31, 1999, whichever occurred first”
Kerb, where am I to get from this that they studied all the kids for 8 years? If the kids were born in the cohort, say, Jan ’94 to Dec ’94 and followed to ‘the end of follow-up, on December 31, 1999′ that would’ve only been 6 years. According to the births, only 2 cohorts could have been studied for 8 yrs or more: Jan 91 – Dec 91 (9yrs) and Jan 92 – Dec 92 (8yrs).
-------------------------------------------------------------------------------------------------------
In response to my protest I did get a ‘tacit’ admission from Kreb that I was correct, but he, nevertheless, went on the attack:
-------------------------------------------------------------------------------------------------------
“You don’t seem to understand that if you follow a cohort of children for a fixed period of time, some of them will be younger at the end of the study. You have to have an end date for the study – it can’t go on forever. You can’t publish a study in 2002 that follows up children until 2013 as you suggested, and even if you did, the children born after 2010 would be too young, so you would have to extend the study another 2 years, but then the children born after 2012 would be too young so you would have to extend the study another 2 years… With your logic you would never end the study.”
---------------------------------------------------------------------------------------------------------
Yet, in his spiel, what Kreb is doing is craftily changing the issue. It’s not a matter of whether the kids were studied for equal lengths, but whether all the cohorts were studied for a long enough time that autism could have been detected in them. Even if the researchers wanted to respect their Dec 31, 1999 end-date, nothing stopped them from going into the records even earlier and looking at the kids that were born in, say, the 80s. They could have still had the eight cohorts, but an earlier start would have ensured that the last cohorts had a longer time to be studied before the Dec 31, 1999 end-date.
In the end, I feel that the researchers knew precisely what they were doing by having such a short period between the last cohorts and the study’s end date. They were seeking to eliminate cases and they did succeed in doing so. I imagine the data was available and initially they did obtain the age-to-age figures using the ‘straight-forward’ method. I believe those figures were not favourable to their position. Perhaps the figure may not have necessarily indicated a correlation, but something to the extent of a movement towards a correlation as the children aged. Perhaps, the ‘autism-free life-years’ solution was simply the crafty means by which they concealed this inconvenient reality. What is clear to all though is that such a study that suffers from the serious flaw of missing cases is definitely not a sound one.
@Greg - thank for for succinctly laying out all of the areas in which you don't understand statistics, study methodology, or pretty much anything else about the way that this this study was actually conducted.
It really is a bit sad to see how uneducated you are....and how it is so hard for you to accept any actual scientific evidence that shows the wrongness of your position.
(the grasping at straws in your argument is a sight to behold).
Here is Greg's position in very simple terms:
Greg heard that someone flipped a coin 12 times & it came up heads every single time....therefore, when you flip a coin, there is at least a 99.9% chance that it will come up heads.
This was reinforced by a website that told him that more often than not, a coin flipped will come up heads (at least, that's what people said would happen).
When researchers looked at coin flipping - they found that over time, the odds of a coin landing on heads was equal to the odds of the coin landing on tails (even if there were instances where one head or the other was predominate in small samples).
Greg doesn't believe that this is the case & that researchers should continue flipping coins (to infinity) since he knows that there is a 99.9% chance that coins flipped will always land on heads (because people have told him so & their stories make far more sense to him than any old research study).
Of course, it was also found that the original study of the 12 coins was false - that actually a number of those coins landed on tails, but were changed to heads, to make the results of the study more "accurate" to what the researcher wanted to prove....but of course, the original 12 coin study is still considered by the "Age of Head Coins" to be the definitive study proving their theory.
Recently, a study that was done using 500,000 coin flips (balanced to eliminate as many unknowns as possible showed that the odds, again, were about even on where a coin would land.) Greg believes that either the numbers were fudged or they just didn't flip the coins enough times to get the results he knows to be true.
Sound familiar, anyone?
I must admit, I get a chuckle out of the thought that now it's AOA who will be facing Gerg's walls o' misunderstood text, exacerbated by his inability to figure out blockquote.
I nailed the lying POS Troll within hours of his first post on RI...months ago...when he copy/pasted his comment from AoA.
You just need to ignore The Troll, as we did with Thingy and post your comments around him.
Meanwhile, on Jake's blog, Dan Burns one of the co-founders (along with Polly Tommey and Carmel and Andy Wakefield) of The Autism Trust USA, is defending Mark Blaxill. I don't imagine that Mr. Burns will be anymore successful at reasoning with Jake than Ginger Taylor was....in fact, I'm counting on it.
Great Holy Frigg in heaven! He's put his comment on AoA where Jameson perseverates further today.
I'll be back.
Greg, why are you pretending to have a discussion with people you consider
http://scienceblogs.com/insolence/2013/10/14/antivaccine-odds-ends/#com…
Thank you Krebiozen, Lawrence, Antaeus Feldspar for educating the lurkers.
Extraordinary. As I suspected, Greg's "summary" is less than accurate. He appears to have simply reiterated all the errors that I and others have tried to correct, proving that, as I suspected, he has understood practically nothing of our interchange. He has also cherry-picked bits of our interchange to make it look as if I have unfairly criticized him, leaving out his 'amusing' bits of provocation..
Greg even repeats his very worst error. He replicates the table I constructed to show the effect of younger children being both less likely to have been vaccinated and less likely to have been diagnosed with autism, but "corrected" so that every case of autism has been diagnosed by the age of one. He then criticized me for not including the very effect he had just removed. Even though I explained this at length, he repeats this "corrected" table in his "summary". Unbelievable.
I challenged Greg to show some honesty and integrity by pointing out that on AoA Gilmore criticized Madsen of using a term, "life-year", that does not appear in the study. Instead Greg writes:
I think that explicitly demonstrates that not only does Greg have no honesty and no integrity, but he is actually proud of the fact.
About two years ago, RI's minions were treated to an interaction with Jake:
I tried to point out to him the futility of his position which revolved around his belief that research conducted around the world over many years was NOT TO BE TRUSTED.
We find that this general theme forms the bedrock upon which woo and alt media empires are built. Greg adds another facet to this glittering gem:
statistical analysis in the hands of the devious is indeed a deadly weapon which leads to the slaughter of the innocents and incredible wealth for the perpetrators of devastatingly wicked crimes.
They are craftly, those sly foxes! Esp Kreb.
Yes, they conspired to f@ck with data. A cover-up of vast proportions proceeded and now implicates the entire medical profession, most researchers in the life sciences, univeristies, pharmaceutical companies, governments and the media.
If you go over to Jake's blog, Autism Investigated, you'll discover that his perseveration continues.
I asked, " How likely is it that this convoluted imbroglio of lies, pay-offs, sculduggery and cover-ups - involving thousands of people worldwide over decades- could be maintained?
Wouldn't that be ... er.. somewhat difficult? Isn't it more likely that only one or a few people are lying?"
He wouldn't buy it. And neither does his successor to the title of 'RI's Schooled Boy', Greg..
Indeed, to these guys WE are part of the problem. But I ask: How likely is it that a few business people, vitamin sellers and mothers are able to arrive at the heart of the deception and inform the world that something is rotten in Denmark, the UK, the US, Australia et al?
Detectives, they're not.
Added to which, if you had to manipulate the statistics to come up with the conclusion that the original paper did, why even bother to publish the paper? If there's such a conspiracy, why not just make up the raw data?
Poor Greg, sincerely coming here. just asking questions and enduring our insults. I see he has carefully edited his and my comments to make him look like a paragon of virtue, trying politely to deal with a pack of intolerant bigots. In places he has even altered the order of my replies to his idiotic comments to make it look like he was replying to me, not vice versa. Curiously he has omitted all his "hee hee hee"s and his repeated labeling of me and others as liars and hypocritical supporters of child abuse. I can't imagine why.
I do find it amusing that Greg concludes that this is a good study, or it would have been if it had only included more children, and had covered a longer period.
There's little need to remind anyone that this is a study that included 537,303 children studied over a period of 8 years. There was no statistically significant correlation between MMR and autism, whatever way the study's authors looked at it; age at vaccination, time since vaccination or date of vaccination.
Greg's criticism that it didn't follow all the children up for a full 8 years is special pleading in the extreme, especially since the confounding effects of age were accounted for and corrected. Several hundred thousand children were followed up until they were older than the average age at which ASDs are diagnosed. Why doesn't the link show up in these children?
As I mentioned before, Wakefield's study of 12 children is entirely acceptable to Greg and his friends, only because it supports their prejudices, and Madsen's study of over half a million children is unacceptable because it doesn't support their prejudices.
Greg refers to an "inconvenient reality". It seems clear to me that the "inconvenient reality" is that this large and well-designed study shows very clearly that MMR cannot possibly be a significant cause of ASDs, and it is extremely inconvenient for Greg and his fellow AoA antivaxxers. That's why he has spent so much time unsuccessfully attempting to find fault with it.
I was digging around trying to find out how statistical power Madsen's study has when I came across the following, which is worth quoting. Greg, if you're still reading this, take note:
@Kreb - not just 500K children, but all children born in Denmark during the study period....what the hell else does Greg want?
<AoA Loony> Denmark is the only country in the world with more pigs than people. Obviously the Magical Pig Power protects against the ill effects of the Evil Vaccines. This study can't be extrapolated to less High Hog™ environments. I'll take my Nobel Prize now, please. <AoA Loony>
What do you think motivates a person without sufficient study in an arcane discipline like statistics ( or immunology or neurophysiology etc) to 'take on the world' , i.e. consensus, and declare it faulty, misguided or hopelessly corrupt? Or all 3?
I've pondered this question a great deal:
in the case of alt media's web woo-meisters, misrepresenting one's self as an expert or critic is part of the sales pitch. However, I imagine that there's something else afoot- envy and rage about others' accomplishments which fuels the contemptous invective that roils and seethes from each and every word. Listen to the emotion in their speech. One commenter guessed that Greg wanted to "make us cry". I would tend to agree: to put us in our place, to show us the door, shrieking and yelling about our many sins and despicable crimes.
What did we ever do to him? Maybe we represent the smart alecks who argued and won debates with him when he was 14 or the ones chosen for the elite universities or graduate schools which declined his entreaties. Perhaps like one well known woo-meister (without much of a real academic background) the bitter taste of calling an actual physician "doctor" propelled him to buy himself self a mail order degree and proceed to decorate his name fore and aft with both 'Dr' and 'PhD'. Twice the title at one great price! Or to create a self-aggrandising bio like Austin's da Vinci of woo, Mikey,did. They call themselves 'scientists' because no one else ever would.
So is their hatred because of how we behave ( vaccinating the children of unwary parents, lying with statistics, receiving pharma lucre) or because we are accomplished and knowledgeable and it's OBVIOUS.
I invite my gentle readers, sceptical brothers and sisters, to read the next AoA, TMR or AI screed with this idea in mind:
it has little to do with the children but rather with themselves.
Well VCADODers, the time has come when I must be on my way. I did enjoy this outing though. I think it was rather productive. But, before I leave, perhaps I should field this final question. Chemmomo asks...
"Greg, why are you pretending to have a discussion with people you consider
all disgusting, lying, unscrupulous pharma shills"
That's a good question, Chemmomo!
Guys, I am sure it's obvious that I never for a second waiver in my belief that vaccines cause autism. I told you before, and I will tell you again -- there is just no resolution to the countless parents' reports of how vaccines changed their children. The 'bad remembering' and 'coincidence' arguments are shams! Add in all the circumstantial evidence implicating vaccines, and anyone seriously following the issue would have to be an absolute idiot not to come to conclusion that vaccines cause autism.
Now, when I first ventured on this site, I contemplated how to broach the topic and solicit your takes: I thought, should I discuss whether the vaccines/autism studies are sound....whether vaccine ingredients are sufficiently studied....whether vaccines in combinations are examined....whether there really is an autism explosion or just better detection...whether the autism fear is being fueled by con-artist, snakeoil salespeople...whether Wakefield committed research fraud...whether there is sense to Offit's 'safe 10,000 injections per sitting' statement.....whether blah blah blah.....
I thought about asking all these question guys, and then I thought --- screw it! Let me skip the middleman and ask them directly whether they personally believe that vaccines do not cause autism. And, I did ask the question guys, and obtained only five responses--five! Guys, not only do I believe that vaccines do cause autism, but you also believe it too.
Yet guys, even though I am quite confident in my belief, I am also a curious person. I am curious about how exactly the 'denialism apparatus' is maintained. Now, there are certain things you cannot always learn from your friends. Sometimes you need to approach your foes. That's one reason why I come here.
Now, I guess the second reason relates a little to what MOB said. It's not so much that I am seeking remorse or guilt from you guys, but hoping that you see the wrong in your ways, and 'do the right thing' (corny, but I think it fits). You say you are 'protecting the herd', 'the benefits outweigh the risks', 'autism isn't so bad'? Please! Guys, you need to do the right thing!
@Denice
I did intend to have another heart-to-heart with you but maybe Vicki and Khani are right: Maybe we've had enough bad etiquette for the time being. I am sure deep-down you will be curious and disappointed though, but we will talk next time I make a return. Hey Denice, I also appreciate you being a 'good sport'.
Please find a copy of the book "Kluge" by Gary Marcus and read it from cover to cover. I think you'll find that ‘bad remembering’ and ‘coincidence’ are more distorting than you initially assumed.
Circumstantial evidence is not enough. It's a starting point, not an end point. In addition, dozens of large studies were conducted in different countries. None found a link. One of Orac's interests is pareidola. It's caused by the human mind's tendency to see patterns, even where the patterns don't exist.
They are sound, your and AoA's pathetic attempts to discredit them notwithstanding.
Yes. Go to PubMed and search on the subjects. You'll find loads of papers on both.
You forgot greater awareness. You do know that the inventor of Pokemon and the character of Sheldon Cooper in The Big Bang Theory are both autistic?
Lupron therapy; chelation; bleach enemas. Enough said
Well, what would you call "adjusting" the results to point towards a conclusion?
Offit was talking about antigen load. Even with adjuvants, 10,000 vaccines would not overload the immune system.
Silence gives consent? Nope. Most of the commenters here simply couldn't be bothered to answer you because you showed mendacity.
Not quite.
Greg, I don't think you can count. A lot of people responded to you. Simply more evidence, as if we need any more, that you don't do anything but tell lies.
@ChrisP - bad lies at that (he continues to repeat the whole 10,000 injections vs. antigens lie, etc.)
It is very strange to see someone who has cheerfully admitted that no amount of evidence will ever change his hermetically sealed mind saying he is curious about, "how exactly the ‘denialism apparatus’ is maintained". This must be one of the most obvious examples of projection I have ever seen.
Greg, you're wrong. Thousands of doctors, scientists and millions of ordinary people around the world know that you are wrong. There is a huge amount of reliable evidence that you are wrong. Even your claims of "countless parents’ reports of how vaccines changed their children" are wrong, as several studies clearly show, including Madsen's. I'll remind you, out of over half a million children that included 738 cases of autism and other ASDs recorded in the study, only 11 were diagnosed within 6 months of being vaccinated with MMR.
I'll remind you also of this study done in North East London in 2003, five years after Wakefield's fraudulent paper was published. In only 12 of the 118 cases of regressive autism did the parents blame vaccines. It was also clear that some parents had changed their view on what had caused their child's autism only after Wakefield's study was published:
This is good evidence that some parents have edited their memories of the timing, and even the occurrence, of their child's autistic regression after they became aware of the claim that MMR causes autism.
The reason few of us answered your question, "whether they personally believe that vaccines do not cause autism", isn't because we don't want to admit we believe it, it's because it is an extraordinarily foolish question. To my mind it is as foolish as someone asking if they believe that an alien spacecraft crashed in Roswell in 1947. Most skeptics wouldn't bother answering such a question because the mere fact you ask it identifies you as a loon, not worth bothering with. I'll remind you that in an opinion poll last year (PDF) found that more people believe the Roswell story (21%) than believe that vaccines cause autism (20%).
You are one of a tiny minority of delusional loons, a tinier minority than even the space alien Roswell believers, yet you feel justified in pondering "how exactly the ‘denialism apparatus’ is maintained" in those who don't share your delusions? I'm tempted to laugh, but really I think it is very sad.
And Greg seems to think that this is a point in favor of his credibility, rather than the opposite.
If the accelerator is pushed, and the engine roars, and the scenery going by speeds like a flash - if the brakes are applied, and the car slows to a crawl, and pedestrians are looking curiously at the car as they overtake it jogging - and throughout all this, the speedometer reports that the car is going 55 - the speedometer is broken. If that speedometer could talk, I wonder if it would sound like Greg. "Of course we're going 55! Even though there's no gas in the tank and we're up on a garage lift, we're going 55! How can you not believe me, when I've been telling you so consistently that we're going 55 for all these years??"
And as has been pointed out to Greg before, this is exactly the same evidence that was used to put people to death in the Salem Witch Trials. "That person gave my Abigail a mean look and then Abigail got sick, therefore that mean look was an Evil Eye from a witch that caused the sickness!!"
Of course, Greg is including in that "circumstantial evidence" things that he has admitted are in fact only his lies. One of those pieces of "circumstantial evidence", he once told us, is that the medical establishment refused to do an vax/unvax study which could be done, would be both ethical and scientifically sound and would settle the issue so definitively that even antivaxxers couldn't find a way to discount the result if it didn't go their way. Well, eventually Greg admitted that he was lying: he never had such a study design in mind; he just decided to lie and say that there was one and that if the medical establishment didn't do his 'definitive' study it was circumstantial evidence that they didn't want the truth at all.
Greg says we should do "the right thing". I don't see why any reasonable person would think "the right thing" can be learned from someone who doesn't draw their conclusions from the evidence, but starts with their conclusions and shamelessly fakes their evidence, as Greg has shown us numerous times.
I'm still curious as to where all of these "countless" people are....because we keep seeing the same stories & trolls over and over again (involving the same people).
Most interestingly, given over 499,000 children - with the vast majority of them vaccinated, that there would be under 1000 children diagnosed with autism or ASD....so, I ask Greg again, where are the "countless" parents & children you keep talking about?
And we've told you before: the plural of 'anecdote' isn't 'evidence'.
After weeks of back and forth, you're leaving the debate (assuming you stick the flounce) in exactly the same state youentered it--without any evidence which supports your position.
The following shows why it is pointless, other than to educate lurkers, to discuss anything honestly with Greg. First there is this comment after he was told to stop posting off topic, emphasis added: " Let me also take the time to congratulate Chris for his (her) success in fighting cancer."
Here is the title of that particular article, again with emphasis added: "Chris beat cancer? He did indeed, but it wasn’t quackery that cured him."
When I pointed out that this revealed he had not even read the article he responded to me (because I have a very common name): "Actually Chris, for what it’s worth, I don’t know whether you are a he or she either. So which is it? (hee hee hee)."
Now see how he spins it above in a very creepy way in Comment #346.
It is a troll, and nothing else. I commend all of the patience in trying to educate him, and the lurkers. But remember he is just a troll, and a very incredibly stupid troll.
@ Lawrence:
About those "countless" people:
Well, we can count them or at least estimate them. If you survey various vax-autism websites, you can find their facebook numbers- this might give us a hint because the age cohort of parents is rather likely to use social media.
What do we see? AoA gets 7K+, the Canary Party, nearly 10K, TMR and GR about 16K each, IIRC the only one to get a larger number is the Vaccine Machine's at 37K- many of them nervous preggers/ new mommas seeking advice, not autism parents. Remember that these are international numbers. And that there is likely to be overlap amongst the groups; adamant believers may be more likely to add every family member to the count. Also alt media sites' friends may be sympathetic but I doubt that their larger numbers ( e.g. Mercola, Natiral News) will be as directly involved as these anti-vax places are.
We can also look at the number of people who claim vaccine injury through governmental instruments like the vaccine court as well as how many show up at conferences or protests. A Reuters poll estimated that around 30% of parents had ANY questions at all about vaccines 2 years ago.
Mnookin cites1% of parents as non-vaccinating and another 10% vaccinate selectively.
Perhaps in an attempt to invigorate interest, Blaxill has lobbied for the next congressional meeting ( in December) to be about vaccine court- ways to get money always pricks up people's ears- rather than CDC cover-ups about mercury, much to Jake's dismay.
We see that die hards like Jameson are trying to instruct parents on how to use social media to proselytise while AM Dachel comment bombards any news story vaguely associated with ASDs.The Canaries have branched out to diagnoses other than autism- like asthma, allergies, ADHD. They thus appear quite desperate.
Ha ha ha! I notice that Greg plans talking to me next time:
as my late father often remarked when bothered by people on the phone :
"Write me a long letter- tell me ALL about it and send it off. It will delivered here".
@Denise - I did a bit of "back of the envelope" math (something the people at AoA seem to be completely unable to do).....I am doing this right?
Taking the approximate population of the US today of 317 mil...with approximately 6.9% of the population being under 5 years old:
21,873,000 children
Rate of Autism diagnosis (approximately 1 in 88 or 1.13%)
240,603 children 5 and under diagnosed with autism
Of those, approximately 30% could be classified as "moderate to severe" which would give us:
Approximately 72,181 children aged 5 and under out of a total population of more than 21 million....
So, I asked Greg (and the anti-vax crowd in general) where are the "millions" of which you speak? Because even the rough numbers show that you are easily off by a factor of 10 or more........
As the US does not have a monarchy or aristocracy, I suppose we are all countless as well as earlless and dukeless (except for the Duke of Earl, of course).
And since we have no House of Lords, we are without Peer.
@ MOB
Well, there are Duke Ellington, Count Basi and King Oliver.
M.O'B.,
I'm hoping that one day the UK will have a proper revolution* and we too will become countless and peerless. We will also retain our Dukes of Earl - I didn't know what that meant until I looked it up a minute ago.
[Exit singing: Duke, Duke, Duke...]
* Cromwell doesn't count, neither do the Civil Wars or the Commonwealth; no self-respecting revolution still has a monarchy afterwards.
Renate,
We still have Prince Rogers Nelson, but sadly Ellington, Basie, and Oliver are no longer with us. Would Prince fans be a principality?
And, we still have Dr. Jay Gordon, spreading his little gems about vaccines to the mommies of his special snowflakes:
https://twitter.com/JayGordonMDFAAP/status/390297113730371584
Formerly known as a principality...
Formerly known as formerly known as a principality...
Formally known as a vacillating principality...
A note to my American RI friends - I will be visiting your fair country in a few days and just got my flu shot, so as not to carry a nasty infection over the border with me. I did this even though it means I will be nursing a vaccine injury (sore shoulder) for the next couple of days. The things I do for you people....
@Edith Prickly - and we will give you a hearty pat on the shoulder when you arrive!
Here's a story about one of those fervent vaccinators - a villainess of the deepest dye http://www.realsimple.com/health/preventative-health/polio-vaccine-0010…
"And yet within the last decade this 58-year-old has made 20 trips to some of the most rugged and dangerous places in the world—Mali, Nigeria, Chad. Each time, she leads a team of one to two dozen volunteers with the same ambitious goal: to immunize as many children as possible, thereby ensuring that they will not contract polio, an infectious viral disease that can attack nerves and cause paralysis. It’s the same ailment that has wreaked havoc on Ann Lee’s life. “I get exhausted sometimes, but then I remember: I never want another child to endure what I’ve gone through,” she says."
What an awful person, eh...
@Edith Prickly,
Also vaccinated, I feel like becoming more autistic but I will await your visit :)
Alain
Greg #51:
Greg #379:
So which is it? Including younger ages would actually help the no-link (anti-vax) claim? Or is it extremely disadvantageous for the anti-vax side
LW, you are making a mistake expecting any consistency in any of Greg's arguments. The only consistency Greg has is that vaccines cause autism - and numerous other diseases when he wants them to.
LW,
What amuses me is that Greg complains about the confounding effects of age (though he mostly gets it backwards), but then claims it is some sort of conspiracy when these are corrected for using perfectly legitimate statistical techniques.
@ ChrisP:
He also consistently discerns malfeasance whenever he runs up against people with superior arguments and information whilst accusing them of connections to pharmaceutical companies. I mean, I also like certain designers' clothes but I don't get paid to say so.
@ Krebiozen:
Wait, wait... next thing the anti-vaxxers will write up an expose wherein the deep, dark secrets of statistical analysis are revealed....
"In the 1970s, a group of statisticians gathered in darkness on the 5th of November to hatch a fiendish plot to f#ck with people's heads, distort scientific research and create vast wealth in the markets...
their Great Dark Lord opined devilishly- ' Nobody really understands this stuff anyway, so we can say ANYTHING we like- who's gonna know?'
We have uncovered secret documents that demonstrate that researchers in medicine, global industry and governing supervisory bodies frequently consulted with statisticians..."
Denice,
I like it. On several occasions here I have seen antivaxxers and CAMsters reveal, sometimes explicitly, sometimes inadvertently, that if something is too complex for them to understand immediately, it must be suspect. Anything that is true must be easily explainable to a 12-year-old, it seems, and anything more complex is meaningless gobbledygook deliberately contrived to confuse them.
If a person really believes this, what motivation do they have to spend the time and effort required to make sense of something that seems incomprehensible at first? One of my greatest pleasures in life is making sense out of something I once found baffling.
I've seen this too, and called it out. It's a form of entitlement syndrome; the person who suffers from it believe the universe itself owes them easy answers. They argue as if the mere simplicity of an idea is, in itself, evidence for its correctness - whereas common logic would suggest that, if an idea is simple to understand and would be equally simple to verify if it was true and no one who has tried has been able to provide that verification, that's pretty good evidence that the idea isn't true.
Right. Only I'd say " a 10- year-old" because at 12, there's more likely to be a bit more formal operational thought thus more likey to get the hang of scientific constructs and abstractions.
This. So much this. Our 'Libertarian' friend on the other thread is a perfect example of this, like these gems on climate science: