No, the CDC did not just apologize and admit that this year’s flu vaccine doesn’t work, part 2

fluvirus

It looks like this year's going to be a bad flu season.

Hard as it is for me to believe, it was only five weeks ago when I discussed an announcement by the CDC that this year's flu vaccine would likely be less effective because it isn't a good match for the influenza strains in circulation this year. Those familiar with how the flu vaccine is developed every year know that the composition of the vaccine depends on the WHO's choice of the three or four strains that its experts deem most likely to cause significant human suffering and death in the coming flu season. Basically, the WHO has to make an educated guess many months in advance about what strains will be circulating during next year's flu season. When it guesses well, the flu vaccine is pretty effective. When it guesses poorly, not so much. Every year, even with the most careful deliberations going into making the choice of strains of flu virus to include in the vaccine, it's always a bit of a crapshoot.

The reason is antigenic drift. This year, it just so happens that a strain of H3N2 is the predominant strain circulating in the population. Unfortunately, H3N2 has undergone drift, such that the strains in circulation now are different than the H3N2 strain chosen to be part of the flu vaccine and thus a poor match. The current H3N2 strains didn't make an appearance until it was too late to alter the formulation of the flu vaccine, and they had increased greatly by September. Basically, only 48% of the circulating H3N2 strains causing most of the disease are a good match.

Unfortunately, yesterday we learned that this year it's even worse than expected. When I said that the CDC didn't just admit that this year's flu vaccine doesn't work, it was true. However, there's no denying that this year's flu vaccine has been quite disappointing as measured by a the CDC's most recent study of vaccine effectiveness (VE), which the CDC has been carrying out each year since 2004. According to a press release from the CDC, year's VE study revealed:

A report published in the January 16 Morbidity and Mortality Weekly Report (MMWR) estimates that getting a flu vaccine this season reduced a person’s risk of having to go to the doctor because of flu by 23 percent among people of all ages.

Since CDC began conducting annual flu vaccine effectiveness (VE) studies in 2004-2005, overall estimates for each season have ranged from 10 percent to 60 percent effectiveness in preventing medical visits associated with seasonal influenza illness. The MMWR report says this season’s vaccine offers reduced protection and this underscores the need for additional prevention and treatment efforts this season, including the appropriate use of influenza antiviral medications for treatment.

“Physicians should be aware that all hospitalized patients and all outpatients at high risk for serious complications should be treated as soon as possible with one of three available influenza antiviral medications if influenza is suspected, regardless of a patient’s vaccination status and without waiting for confirmatory testing,” says Joe Bresee, branch chief in CDC’s Influenza Division. “Health care providers should advise patients at high risk to call promptly if they get symptoms of influenza.”

Note that, in this case, effectiveness means "real world" effectiveness, as compared to efficacy, which refers to how well a treatment does in a randomized clinical trial. Naturally, this news has produced headlines such as:

Well, you get the idea.

So how did the CDC come up with this estimate? This MMWR study tells the tale.:

Patients aged ≥6 months were enrolled when seeking outpatient medical care for an ARI with cough at study sites in Michigan, Pennsylvania, Texas, Washington, and Wisconsin.* Study enrollment began once laboratory-confirmed cases of influenza were identified through local surveillance. Trained study staff members reviewed appointment schedules and chief complaints to identify patients with ARI. Patients were eligible for enrollment if they 1) were aged ≥6 months on September 1, 2014, and thus eligible for vaccination; 2) reported an ARI with cough and onset ≤7 days earlier; and 3) had not yet been treated with influenza antiviral medication (e.g., oseltamivir) during this illness. Consenting participants completed an enrollment interview. Nasal and oropharyngeal swabs were collected from each patient and placed together in a single cryovial with viral transport medium. Only nasal swabs were collected for patients aged <2 years. Specimens were tested at Flu VE laboratories using CDC's real-time reverse transcription–polymerase chain reaction (rRT-PCR) protocol for detection and identification of influenza viruses.

Participants were considered vaccinated if they received ≥1 dose of any seasonal influenza vaccine ≥14 days before illness onset, according to medical records and registries (at the Wisconsin site) or medical records and self-report (at the Michigan, Pennsylvania, Texas, and Washington sites). Vaccine effectiveness was estimated as 100% x (1 - odds ratio [ratio of odds of being vaccinated among outpatients with influenza-positive test results to the odds of being vaccinated among outpatients with influenza-negative test results]); odds ratios were estimated using logistic regression. Estimates were adjusted for study site, age, sex, race/ethnicity, self-rated health, and days from illness onset to enrollment. These early interim VE estimates for the 2014–15 season were based on patients enrolled through January 2, 2015.

To boil it down, the CDC enrolled patients who sought care for an acute respiratory illness with cough and determined what percentage of those who tested positive for flu and those who did not had been vaccinated. In essence, this is a prospective test-negative case-control design. Basically, controls were patients with a medically attended ARI who tested negative for the flu virus, while cases were patients who tested positive. The primary exposure of interest was, of course, the flu vaccine. The idea of this study is to determine whether a higher proportion of people who tested negative for the flu were vaccinated than those who tested positive.

One thing that should be pointed out. This study design does not answer the question that most people are interested in: If I get the flu vaccine, what by how much does it reduce my chances of getting the flu? Remember, this patient population consists of people who have an ARI with cough severe enough that they sought medical help for it. Of course, these people almost certainly don't have the flu, given that by definition they have no symptoms. On the other hand this study says nothing about people with the flu or ARIs who don't seek medical care, nor does it say anything about the ability of this vaccine to prevent severe disease, hospitalization, and death. Still, as a quick measure of effectiveness after the flu season is only halfway over, it's not bad. Think of it this way. To determine the absolute efficacy of the flu vaccine would require following a huge number of people, vaccinated and unvaccinated, given that in an average flu season around 2% of the population will get the flu, and seeing who does and doesn't get the flu that year. To reliably detect, say, a 50% reduction in flu would require thousands of people. Under the CDC's study design, of people showing up at a clinic during flu season, many more will actually have the flu; so a determination of flu vaccine effectiveness can be made with a lot fewer people.

In this case, there were 2,321 children and adults with AR enrolled in the various study sites, and 950 (41%) tested positive for the influenza virus by PCR. Among patients who tested positive for the flu, 49% had been vaccinated, while 56% of who tested negative were vaccinated. After the CDC adjusted for study site, age, sex, race/ethnicity, self-rated health, and days from illness onset to enrollment, the VE against medically attended ARI attributable to influenza A and B virus infections was 23% (CI = 8%–36%). Depending on age range, effectiveness ranged from 12% to 26%:

Another factor that influences how well the flu vaccine works is the age and health of the person being vaccinated. In general, the flu vaccine works best in young, healthy people and is less effective in people 65 and older. This pattern is reflected in the current season early estimates for VE against H3N2 viruses. VE against H3N2 viruses was highest -- 26 percent -- for children age 6 months through 17 years. While not statistically significant, VE estimates against H3N2 viruses for other age groups were 12 percent for ages 18 to 49 years and 14 percent for people age 50 years and older.

CDC recommends that people get a flu vaccine even during season’s when drifted viruses are circulating because vaccination can still prevent some infections and can reduce severe disease that can lead to hospitalization and death. Also, the flu vaccine is designed to protect against three or four influenza viruses and some of these other viruses may circulate later in the season.

The bottom line, obviously, is that so far this season the flu vaccine isn't the greatest. Still, it offers some protection and, given how safe the vaccine is, there's no compelling reason not to get it.

Clearly, what is needed is a much better flu vaccine. No one has ever denied that. However, the fact that some years the vaccine is not very effective because the strains used to make the vaccine don't match the circulating strains very well because the WHO guessed wrong last February doesn't mean that vaccines don't work. All it means is that this vaccine is not s good as many other vaccines because of the inherent difficulties involved in making an effective flu vaccine that are a consequence of its biology, specifically its ability to mutate that results in drift, such as what occurred this year.

It's only about a month until the WHO has to pick the strains to include in next year's flu vaccine. Certainly, the committee tasked with choosing the strains will include the drifted H3N2 strain, but who knows if that will be the dominant strain next year. One thing's for sure: I don't envy the people who have to come up with these recommendations. In years when the vaccine is good, there's no glory and the antivaccine movement attacks their vaccine, but in years when it's bad they can expect nothing but opprobrium directed at the vaccine without even the personal satisfaction of having produced a highly effective vaccine.

Certain viruses are just incredibly difficult to design effective vaccines against. Unfortunately, influenza is one of them. Just because its efficacy varies from year to year and this year appears to be significantly lower than average is not evidence that vaccines are harmful or don't work. It's just that this vaccine isn't the greatest, something that's always been acknowledged. Next year, it's likely to be much better. Until a vaccine is developed that targets a relatively invariant part of the virus and is thus universal, this is likely to be the case. Fortunately, scientists appear to be getting closer and closer to a universal flu vaccine.

Categories

More like this

If there's one thing about having a demanding day job, it's that the cranks usually have the advantage. They can almost always hit first when a news story comes out that they can spin to attack their detested science. On the other hand, it usually ensures that by the time I get home, have dinner…
We've discussed this already, but now CIDRAP News also has a story (which they got from AP) that this year's flu vaccine is not perfectly matched to the all the circulating viruses (of course we had it first, but hey, who's keeping track?). The data that are used to prepare vaccines for the next…
by revere, cross-posted from Effect Measure This year's flu season isn't over, but it's almost over, and it was fairly typical and much better than last year, which was nasty. It began at the end of September but didn't take off until early January, peaking in mid Februrary. New cases are still…
This year's flu season isn't over, but it's almost over, and it was fairly typical and much better than last year, which was nasty. It began at the end of September but didn't take off until early January, peaking in mid Februrary. New cases are still appearing but much less frequently and they are…

Well, a so-called "universal" flu vaccine that could offer wide-ranging protection for up to 20 years is going into clinical trials later this year (the linked story headline is somewhat deceptive, as "soon a reality" is just hype at this point).

http://abcnews.go.com/Health/universal-flu-vaccine-reality-scientists/s…

I was reading another abcnews.com story about the flu outbreaks so far this year, and came across a doozy of a link* at the end of the story, to "10 home remedies" that are supposed to help you beat the flu. #1 is garlic (well OK, I personally believe garlic is a terrific food/condiment that probably cures a bunch of things and eat a lot of it), but #10 is...a chiropractic adjustment!!!

Yep, you're supposed to see the chiro when flu symptoms strike, and after a little wrench here or there, flu goes bye-bye. And here I thought chiropractic was only supposed to _prevent_ flu:

http://www.hhchiro.com/chiropractic-care-colds-flu/

The limited effectiveness of flu vaccine this year should be a boon to quacks of all stripes. People are probably noticing that none of the people pushing these magical remedies are admitting to only 23% effectiveness, so why not give them a try?

*I would have provided the link, but it mysteriously vanished after I looked at the article. Either abcnews.com got complaints, or it just vanished into the void due to the Giant Pharma Lizard conspiracy.

By Dangerous Bacon (not verified) on 16 Jan 2015 #permalink

Dangerous Bacon, the idea of suggesting people go out while having the flu for anything other than something absolutely necessary horrifies me.

{Heck, when I was a grad student, flu seasons meant the Health Center had rules about handling respiratory infections so those students could get help while minimizing the chances they'd transmit whatever they had. One of those rules was a reminder of when to stay home.)

By Becca Stareyes (not verified) on 16 Jan 2015 #permalink

If non-seriously ill flu patients are going to go out seeking care, it's probably better for them to visit the chiro's office, as opposed to spreading the virus at emergency departments or physicians' offices.

By Dangerous Bacon (not verified) on 16 Jan 2015 #permalink

DB@3: I can't agree with you here. As tempting as it may be to give the chiropractic a taste of his own medicine, so to speak, there may be innocent parties there, too: kids who are accompanying a parent, kids who are themselves being treated (feel free to mock the parents in those cases, but leave the kids alone), and people who happen to be visiting the same office complex or strip mall where the chiro's office is located.

Remember also that some people live alone (I am one such) or with no other able-bodied adults in the house. If I need anything while I'm sick, I'm the one who has to go get it. And there are situations where a hospital visit wouldn't be needed if there were a second person available to care for the sick person, but the sick person can't handle it without such help. I hope anybody in the latter category gets the treatment they need.

By Eric Lund (not verified) on 16 Jan 2015 #permalink

If chiropractic adjustment is listed as a home remedy, does that mean that the chiropractor makes house calls?

By Mephistopheles… (not verified) on 16 Jan 2015 #permalink

@MO'B

Or it means you can do it yourself without the need for the chiropractor.

Hmmmm. I see you've been missing alot of work lately.

Well, I wouldn't exactly say I was missing it......

It's not a big TCM insert in Science, but I just ran across this item on GMOs:
Is Humanity Already Playing God With Its Precious Food Supply?
It's a page that comes up as the first post on Gawker — being sponsored — in this case by the SyFy channel to promote a show called Helix. In light of the whole AAAS custom publishing thing, I noted an apparent similarity in that SyFy seems to have hired something called "Studio At Gawker" to produce the piece. So maybe there's some larger collapse of editorial and advertising in publishing, and we'll be seeing a lot more nakedly 'bought' content. Or not. Anyway, here are the first two paragraphs, to show how the copywriters are trying to leverage the GMO controversy into viewership for the show:

There is no denying that we are in the midst of a global food crisis, and that more than ever, there is a great disparity between the way classes of people are able to feed their families. Is this a natural evolution — an unavoidable result of humanity's occupation of the earth? Or is human kind actually pushing itself further into a food crisis with the technologies they claim will help it?
In Syfy's Helix, a cult is playing God with the global food supply, with potentially catastrophic effects. Although the show looks at a very specific instance of food supply manipulation, it makes one think — are the things humanity is already doing to control and change the food supply actually going to help us?

@Todd #7:

DIY Chiropractic: Throw self down the stairs. That'll give the old spinal energy something to think about.

By Rich Woods (not verified) on 16 Jan 2015 #permalink

I'll consider consulting SyFy for advice when a sharknado takes out my "manipulated" crops...

By Richard Smith (not verified) on 16 Jan 2015 #permalink

Relatively minor quibble: the change in the virus is called antigenic drift. Genetic drift is something different.

Of course, quibbling right back at ya: Ya can't have antigenic drift without genetic drift. :-)

In any case, the CDC appears to use the two terms more or less interchangeably:

Increasing the risk of a severe flu season is the finding that roughly half of the H3N2 viruses analyzed are drift variants: viruses with antigenic or genetic changes that make them different from that season’s vaccine virus.

So you'll pardon my confusion. Also, who am I to argue with the CDC? :-) If I get a chance, maybe I'll go back and change "genetic" to "antigenic or genetic," to be in line with the CDC.

@sadmar: I don't watch much television, so I'm not familiar with Helix. I don't mind taking the stated premise as the basis for a science fiction story, as long as it's clearly labeled as fiction. After all, pursuing the consequences of a "what if" scenario is one of the things science fiction is for. But if Helix is supposed to be about factual science, or even informed speculation based on factual science, then yes, that show is problematic.

There are reasons to be concerned about what's happening to the food distribution network, and the possibility of deliberately induced famine. (We have seen examples of the latter already; the USSR in the 1930s is perhaps the most notorious example.) But that's a separate issue from genetic manipulation of food. It's also a matter of logistics and politics, not of science.

By Eric Lund (not verified) on 16 Jan 2015 #permalink

Helix is science fiction. Worrying overmuch about its premise as applied to the real world is rather like worrying about what The Andromeda Strain has to say about virology.

Yeah, the anti-vax crowd loves to hate on the flu vaccine as completely useless. They apparently can't read the research or accept the reality that for a virus that mutates so readily that we haven't yet nailed down a target that doesn't change. Universal vaccine has been in the works for a long time now and I am ever hopeful but I want to see the trial data before I get my hopes up too high. Yeah I feel for those with no family/friends to help them out during an illness. My sister got what was probably norovirus and her coworkers ended up dropping off a care package of Gatorade, soup, and crackers on her doorstep.

Speaking of Helix, I saw the first episode. It started with an interesting premise, but quickly got lost in a see of absurdity.

@Kiiri

Universal vaccine has been in the works for a long time now and I am ever hopeful but I want to see the trial data before I get my hopes up too high.

It will be available in "5 to 10 years".

Don't have the data they got it from, but at the news conference in my state about the state of the flu season there was a comment that very often the viruses as the season wears on tend to differ from the early season strains and those strains are currently still a match for the vaccine.

Well, antigenic drift is when the targeted antigen, hemagglutinin (HA), changes a bit, such that it may no longer be recognized by antibodies against the parental HA.

Genetic drift is (essentially) evolution at random, i.e. in the absence of natural selection. The CDC site mentions genetic changes, but that's not genetic drift.

I grant that antigenic drift might occur via genetic drift, but only when the changes in HA had no selective advantage for the virus. Certainly possible, but probably not the norm, in my guess. More likely is that changes in antigenicity spread in viral populations precisely because they do offer a selective advantage. they allow the virus to infect individuals whose immune systems would likely kill any virus with the previous version of HA. And if there's a selective advantage (or disadvantage), it's not genetic drift.

The point is that any TV show or commercial product is using the GMO controversy to promote viewership/sales, thus reproducing the 'ant-GMO tropes' to a wider audience.

Eric: I don't watch TV either, and certainly not the SyFy channel, but I'd doubt Helix makes any 'real' claims to be "informed speculation based on factual science" in other than a typical sci-fi way, or is in any way problematic.

It's not about anything on the network, it's about the advertising. Would you be unconcerned if a re-release of The Andromeda Strain was being promoted by advertorials spreading Ebola paranoia in general circulation publications?

(sigh)

So maybe there’s some larger collapse of editorial and advertising in publishing, and we’ll be seeing a lot more nakedly ‘bought’ content.

Yes, this is happening. See John Oliver's excellent, as always, video on native advertising.

I remember a stupid article that went viral on Facebook a while back claiming that Charles Shaw wine from Trader Joe's is so cheap because the mechanized grape harvesting process throws all kinds of rodents and birds and whatnot in with the wine. I spent a lot of time debunking this (don't nobody mess with my two buck Chuck) just based on its sheer implausibility before I noticed that it was actually "sponsored content" for a website called Quora that HuffPo had just decided to spew all over the Internet.

It's no longer online, but you can find a CNBC report on it here.

@kaymarie
Don’t have the data they got it from, but at the news conference in my state about the state of the flu season there was a comment that very often the viruses as the season wears on tend to differ from the early season strains and those strains are currently still a match for the vaccine.

That's what happened last year in NYS, with the matches being reversed. Started strong with H1N1, then that tapered, and in March the dominant strain picked up as Flu B, and actually rose to almost the same levels as H1N1. It maintained until the end of May, and I don't think it was that great a match. Our widespread flu season lasted until 6/8!

Already my county is seeing a drop in Flu As. Best case: flu season ends early-ish after the big punch. More likely case, particularly with the crap winter, it picks up again. Hopefully B is a better match if the switch happens.

By Frequent Lurker (not verified) on 16 Jan 2015 #permalink

I've already had the flu! I had a fever of 102 and extreme weakness and muscle aches. Fortunately I was vaccinated and recovered enough to resume normal (non-strenuous) activities in about a week. I'm looking forward to being immune for the rest of the season.

By Christine Rose (not verified) on 16 Jan 2015 #permalink

@Todd W. #17 Re the universal flu vaccine:

It will be available in “5 to 10 years”.

You reminded me of a joke from computer science.
"Artificial intelligence is just 10 years away, and has been so for ober 50 years."
:D

By Julian Frost (not verified) on 17 Jan 2015 #permalink

Orac writes,
"... given how safe the vaccine is, there’s no compelling reason not to get it."

In contradiction, the flu vaccine (i.e., Flucelvax) from Novartis has the highlighted warning, "The tip caps of the pre-filled syringes may contain natural rubber latex which may cause allergic reactions in latex-sensitive individuals."

In reality, you may be a latex-sensitive individual and you may get a flu-vaccine contaminated with natural rubber latex. Furthermore, you may not have latex-sensitivity but develop latex-sensitivity from this flu vaccine. In my opinion, these are compelling reasons not to get it.

By Michael J. Dochniak (not verified) on 17 Jan 2015 #permalink

Here comes the anecdote: I had the flu vaccine voluntarily for the first time this year and marched my husband and son for theirs too. Over Christmas, my entire work team went down like a sack of spuds and I was the only one left standing. I believe the office party had been the final opportunity to spread it around, what with all the exchange of fluids*. So I'm shilling for the flu vacc with extra enthusiasm now. Where's my shillbucks?
*no, not that - cigars were passed round.

By Charlotte (not verified) on 17 Jan 2015 #permalink

So Incoherent Troll cites Helix, then when Helix is is shot down, plays the old move-the-goalposts trick?

Please, proceed. You're not exactly covering yourself in glory, but you are giving the casual reader a picture of yourself that is too accurate to redound to your argument's benefit.

By Phoenix Woman (not verified) on 17 Jan 2015 #permalink

Todd @17, Julian @25: Another joke along those lines is that the universal flu vaccine is the vaccine of the future, and always will be.

We've never been as close as 5-10 years from having commercial fusion power, but it's been 20 years out since the 1950s.

By Eric Lund (not verified) on 17 Jan 2015 #permalink

@Michael #26

Then there's always other flu vaccines which do not have the latex issue - or even the nasal mist.

Again - still no reason to not protect yourself (and others) against the flu.

Darwy writes (#30),

"Then there’s always other flu vaccines which do not have the latex issue – or even the nasal mist."

Yes, for adults that will make the flu vaccine safer!

Furthermore, flu vaccines that have the label "not made with natural rubber latex" will show that pharmaceutical companies are striving towards "do no harm".

But, for some children this may not be enough due to their health (i.e., immune profile).

If a parent is concerned about giving their child a flu vaccine precautions such as a total white blood cell count can be evaluated to determine if their child's immune system is typical or atypical. See:

http://www.pageturnpro.com/Autism-Media-Channel/61559-AF58_AutismFile_O…

By Michael J. Dochniak (not verified) on 18 Jan 2015 #permalink

Great post @Orac. Regarding the (non-signficant) differences in VE with regards to age, could it be that children tend to get the live virus vaccine and adults receive the killed? Flu vaccines have always been problematic for those over 65 due to insufficient immune response, but that might explain the low VE for the 18-49 year olds.

The nasal mist is approved from the age of 2 up, and there are other flu vaccines on the market which do not contain any latex.

Darwy, did you see where the latex-fixated MjD provided a link to an article he wrote for the magazine published by Andrew Wakefield and Polly Tommey?

MjD's son's autism was not caused by latex in vaccine vial ports or latex in any medical product.

Bill Maher was stirring the shit again Friday claiming the flu the vaccine is a big pharma conspiracy for money. Even after the MD on his show politely but firmly disagreed and thoroughly refuted what he said, he continues to trumpet nonsense about the flu vaccine being ineffective, useless, and a big pharma scam.

By scourge99 (not verified) on 18 Jan 2015 #permalink

lilady writes (#36),

"...there are three inactivated flu shots which do not contain latex. One of those three vaccines is approved for pediatric use"

Thank you lilady for this vaccine safety information!

@lilady,

Why do so many vaccines still contain natural rubber latex even though the allergens within are known to be harmful?

By Michael J. Dochniak (not verified) on 18 Jan 2015 #permalink

"Yes, for adults that will make the flu vaccine safer!"

Our health department has been giving flu mist to kids at least since I started working there.

I hear it is 12% effective in those 18-49. LOL

By Sid Offit (not verified) on 18 Jan 2015 #permalink

My mother survived the Spanish flu and lived past 90 years old.
I have had flu about four times in my life but not since Bill Clinton still had his reputation intact.
I've never had a flu shot.
Maybe I just have a very high resistence to viruses.

PS
The word "verticle" means axis or hinge. An archaic term for suited to something that moves in a limited manner but goes nowhere .
Words like that pop up now and again.

By Greg Young (not verified) on 18 Jan 2015 #permalink

"I hear it is 12% effective in those 18-49. LOL"

Where did you hear about 2014-2015 seasonal influenza vaccine effectiveness for those ages 18-49 years old, Offal?

I don't understand why you think "LOL" appended to your comment is appropriate.

Greg Young: "My mother survived the Spanish flu and lived past 90 years old."

And at least one of my great-uncles did not. Thank you for minimizing the death of a child. Apparently you don't care about those whose death make them unable to say how illness affected them.

I have had flu about four times in my life but not since Bill Clinton still had his reputation intact.
I’ve never had a flu shot.
Maybe I just have a very high resistence [sic] to viruses.

Sounds like rather the opposite.

PS
The word “verticle” means axis or hinge. An archaic term for suited to something that moves in a limited manner but goes nowhere .

Not-PS: Go put it in the original comments where it has context, rather than demonstrating that you are immersed in some sort of weird megalomaniacal fantasy.

I hear it is 12% effective in those 18-49. LOL

Hey, Bob, I've heard that you're the author of two "soon to be published books, "LOL."

Quick, Bullethead: What would 12% do to the even-odds over-time window? I've got a 3.87 year bonus.

Oh, wait, you're a hit-and-run chickenshіt.

something that moves in a limited manner but goes nowhere .

Ah, a blog thread.

By herr doktor bimler (not verified) on 18 Jan 2015 #permalink

My mother survived the Spanish flu and lived past 90 years old.
I have had flu about four times in my life but not since Bill Clinton still had his reputation intact.
I’ve never had a flu shot.
Maybe I just have a very high resistence to viruses.

Maybe you are just a big talking, fantasizing idiot.

Oh.

"
Chris

January 19, 2015

Greg Young: “My mother survived the Spanish flu and lived past 90 years old.”

And at least one of my great-uncles did not. Thank you for minimizing the death of a child. Apparently you don’t care about those whose death make them unable to say how illness affected them."
Too bad about your uncle.
Survivors didn't always get off easy either, they left their stories so if you want to know how the illness affected them just look it up.
You shouldn't minimize the lifetime of a person because you envy their longevity and have no sense of history.

The Spanish Lady pretty much ended WW1. Too many young adults in Germany died from it and the Germans had little choice but surrender.
They should get their act together now before that sort of pandemic strikes again.
If they intended to manufacture vaccines then they should find a way to make sure they are going to work before flu season starts rather than end up with this sort of failure.
If you expect faith in a product you should earn that faith.

By Greg Young (not verified) on 18 Jan 2015 #permalink

1918 Flu Antibodies Resurrected From Elderly Survivors

Date:

August 18, 2008

Source:

Vanderbilt University Medical Center

Summary:

Ninety years after the sweeping destruction of the 1918 flu pandemic, researchers have recovered antibodies to the virus -- from elderly survivors of the original outbreak. In addition to revealing the surprisingly long-lasting immunity to such viruses, these antibodies could be effective treatments to have on hand if another virus similar to the 1918 flu breaks out in the future.

http://www.sciencedaily.com/releases/2008/08/080817223642.htm

I wonder what the odds are on a person who has never had a flu shot going two decades without contracting the more recent strains of flu?

By Greg Young (not verified) on 18 Jan 2015 #permalink

As for mucking about with the food supply.
I wonder who's bright idea it was to feed cattle animal tissue?
Before that you only heard of this sort of brain destroying disease among cannibals who ate uncooked human brains.

Probably the most pitiful way to die I've heard of, though I'm sure there are worse ways to go.

By Greg Young (not verified) on 18 Jan 2015 #permalink

Greg Young,

I wonder what the odds are on a person who has never had a flu shot going two decades without contracting the more recent strains of flu?

Since about 2-5% of people will get influenza each season you can expect to get it every 20-50 years on average. Going 20 years influenza-free is par for the course, especially when you had the disease unusually frequently in the past. Older people do sometimes have resistance to circulating strains, such as the H1N1 strain in 2009.

By Krebiozen (not verified) on 19 Jan 2015 #permalink

Probably the most pitiful way to die I’ve heard of, though I’m sure there are worse ways to go.

Such as choosing incorrigible ignorance as a hill to die on?

Ah, Donchiak is back with his idee fixe, is he? ;-) I needed a chuckle on this gray Monday.

Latex works very well. Very few people are allergic to it. Therefore, the sensible approach is to ask people about their allergies before giving vaccines, and providing a small supply of latex-free alternatives. This is similar to the strategy for egg allergy, which is far more common and far more relevant to flu vaccine given that it's usually grown in chicken eggs; people with egg allergy get the more expensive egg-free vaccines.

Not rocket science. Not brain surgery. This is easy stuff, but of course Donchiak struggles with it because somehow to him it validates his belief system. There are none so blind and those who tie the blindfold willingly.

By Calli Arcale (not verified) on 19 Jan 2015 #permalink

Greg:

I wonder who’s bright idea it was to feed cattle animal tissue?

I have no idea, but feeding animal tissue to livestock goes back quite a ways. The first person to think of it is surely long dead. The modern practice is more institutionalized, but comes from the same basic root: thrift. If you want cheap meat, then you need to find cheaper ways to produce it. Cattle do need a certain amount of protein and fat in their diets, and while they can get this from grazing, adding some of the bits that slaughterhouses can't sell otherwise efficient and economical. Thing is, this can transmit disease -- there's a reason cannibalism isn't smart -- but everybody figured there was no problem with that. Just don't feed them *raw* stuff. Cook it to kill all the pathogens. Nobody realized there was one that cooking would not destroy: prions.

So the people who started doing that weren't maniacs or crazy or careless. There was a whole class of pathogens they didn't know about, and certainly didn't realize would survive normal food safety protocols.

By Calli Arcale (not verified) on 19 Jan 2015 #permalink

"
Krebiozen

January 19, 2015

Greg Young,

I wonder what the odds are on a person who has never had a flu shot going two decades without contracting the more recent strains of flu?

Since about 2-5% of people will get influenza each season you can expect to get it every 20-50 years on average. Going 20 years influenza-free is par for the course, especially when you had the disease unusually frequently in the past. Older people do sometimes have resistance to circulating strains, such as the H1N1 strain in 2009."

Thank you , thank you very much (in my best Elvis Presley voice).
You've been helpful in the past which has not been par for the course for this site.
If any of my replies to you have seemed dismissive I want you to know it was unintended.

Would you agree that in this day and age with so many cases of TBI and PTSD due to the conflicts in Iraq and other places that most would prefer to show courtesy and understanding to an older person who suffers from Aphasia due to injury and complications from surgery?

I have noticed that some of the less civilized members here seem to use words as defined by "the Urban Dictionary". One even managed to misspell one of those made up words while ragging on about my misspellings of real words.

The level of ignorance shown by some of both world history and recent events was shocking. Guess the internet has not been as beneficial to them as they suppose.

By Greg Young (not verified) on 19 Jan 2015 #permalink

"
Calli Arcale

http://fractalwonder.wordpress.com
January 19, 2015

Greg:

I wonder who’s bright idea it was to feed cattle animal tissue?

I have no idea, but feeding animal tissue to livestock goes back quite a ways. The first person to think of it is surely long dead. The modern practice is more institutionalized, but comes from the same basic root: thrift. If you want cheap meat, then you need to find cheaper ways to produce it. Cattle do need a certain amount of protein and fat in their diets, and while they can get this from grazing, adding some of the bits that slaughterhouses can’t sell otherwise efficient and economical. Thing is, this can transmit disease — there’s a reason cannibalism isn’t smart — but everybody figured there was no problem with that. Just don’t feed them *raw* stuff. Cook it to kill all the pathogens. Nobody realized there was one that cooking would not destroy: prions.

So the people who started doing that weren’t maniacs or crazy or careless. There was a whole class of pathogens they didn’t know about, and certainly didn’t realize would survive normal food safety protocols."

Its refreshing you find that some here are interested in content rather than presentation. Thank you for your input.

As for vaccines, serums and immunization programs there were some real disasters in the early days of inoculation.
The Mandan people were pretty much wiped out by a misguided attempt to immunize them from Small Pox.
The serum was contaminated and deaths were from sepsis.

Thinking on the odds of infection it occurred to me that odds for an age group or a particular number of previous infections over many years doesn't really work when it comes to a particular individual.
Being exposed to infection every year it seems to me that statistics and/or percentages aren't really the answer.
There's the old saying about lies, damn lies and statistics.

I'm thinking it might be more on the line of a build up of long lasting antibodies or the ability of the body to produce anti bodies that work best against recent strains.

The anti bodies associated with Spanish Flu seem to be of limited use against less lethal strains.

By Greg Young (not verified) on 19 Jan 2015 #permalink

I just checked the CDC's weekly flu update, the pediatric deaths due to influenza this year is now up to forty five.

PROTIP: If your own communications are snarky and snide in tone, it does not reflect well on you to clutch your pearls and moan "Ohh, you brute, how can you be so mean to a veteran, and one in such fragile health??" when people respond in a similar vein.

By Antaeus Feldspar (not verified) on 19 Jan 2015 #permalink

"
Antaeus Feldspar

January 19, 2015

PROTIP: If your own communications are snarky and snide in tone, it does not reflect well on you to clutch your pearls and moan “Ohh, you brute, how can you be so mean to a veteran, and one in such fragile health??” when people respond in a similar vein."
I'm not a veteran, my TBI was not combat related.
The fact that so many of our troops have survived IED explosions has bought more attention to the effects.
As for my remarks ,Its called "responding in kind".
If one wishes to denigrate someone because they reveal that they have a disability, then prove that they have no room to talk despite the appearance of health, they soon learn they may have to take it as well as dish it out.

I make no bones about my combative nature, but I have made my bones.

An inadvertent misstatement, such as becoming momentarily confused about the titles of two movies that had some significant similarities, is not a lie. To claim someone is posting lies and that the institution he mentioned living near is fictional ignores the intent and belief at the time of the person they have insulted.

There are two living witnesses to my meeting Anthony Quayle at the country club after a play at the Clarence Brown Theatre and there is a photograph of Quayle and myself standing side by side , its hanging on the wall of the study of a close friend.
Quayle spent a good deal of time in Knoxville as producer and director of a number of plays, some that ended up on Broadway. I've noticed that this information does not appear on some websites that skim over his career.
He listened to my criticism of his performance intently and with grace, I really liked that guy, very professional.

I've found a witness to the SAT test scores. I remembered a class mate that took the same 300 question aptitude test and found a similar placement examine with 300 questions is still in use.
I also found that the program they had offered was similar to the program used during WW2 to fast track the commissioning of the "90 day wonders" as they were called back then.

Now back to your regular programing.
I've lost interest in this thread already.

By Greg Young (not verified) on 19 Jan 2015 #permalink

Ah. The heart of the problem is revealed.

Greg Young has a problem with certain people. These are people with whom he's had previous conversations about "Anthony Quayle" and SAT scores, and he has grudges against them, based on perceived slights these people offered him in previous conversations.

I do not know why he is under the impression that those people are to be found here, but apparently he is.

By Antaeus Feldspar (not verified) on 19 Jan 2015 #permalink

Calli Arcale writes about natural rubber latex in flu vaccines (#54),

"Not rocket science. Not brain surgery. This is easy stuff..."

@Calli,

What can happen to some children when the 60+ antigenic Hev-b proteins in natural rubber latex leach into vaccine solutions and bind to aluminum hydroxide?

Of course you know the answer Calli Arcale:

Allergy-induced regressive autism

By Michael J. Dochniak (not verified) on 19 Jan 2015 #permalink

Go away MjD. Your child's autism is not caused by a latex allergy or a vaccine.

So someone with the nym 'Phoenix Woman' shows up in this thread and posts #28:

So Incoherent Troll cites Helix, then when Helix is is shot down, plays the old move-the-goalposts trick? Please, proceed. You’re not exactly covering yourself in glory, but you are giving the casual reader a picture of yourself that is too accurate to redound to your argument’s benefit.

No. Retired media studies Professor (and former professional copywriter) cites an advertising campaign for Helix and then when careless readers assume that the text of Helix has anything to do with it, attempts politely to correct their cluelessness about how media campaigns work and why they matter. Anyone who would assert that #9 establishes Helix as a goal-post that has subsequently been moved is a blithering idiot or a troll. Anyone who would assert that #20 establishes a picture of anyone is a blithering idiot or a troll. Hmm, which is it? Let's see: "redound" and "casual reader" are in the same sentence, which claims I have failed to benefit my argument, after alleging I didn't have an argument because I'd moved the goal-post, which moving 'Phoenix Woman' could not have discerned if my comment had, in fact, been incoherent. So, I'm going with 'blithering idiot troll'!

Wait. How can anyone justify this epidemic in a highly vaccinated US population? The CDC's excuse is pure crapola. People should learn the horrible history of vaccination because it's bound to repeat.

H1N1 (which is old news): "How can anyone justify this epidemic in a highly vaccinated US population?"

What epidemic? And exactly how is less that 50% coverage for influenza vaccine considered "highly vaccinated"?

"People should learn the horrible history of vaccination because it’s bound to repeat."

What particular horrible history?

Well, there was the Spanish Flu, which killed over 1 Million Americans (and tens of millions worldwide), because we didn't have vaccines at the time.

What epidemic? And exactly how is less that 50% coverage for influenza vaccine considered “highly vaccinated”?

Well, I've seen a lot of sick patients in my area who came down with influenza (lab-tested) were previously vaccinated against the flu, yet they still got the flu. What gives?

And then I found this:

The herd immunity thresholds in terms of prevalence of protected persons required to establish herd immunity (Ic) ranged from 33% to 73% for the first pandemic viruses and from 9% to 29% for the 2009–10 pandemic and the 2008–09 epidemic viruses (Table 1).
The vaccination coverage that would have been required to establish herd immunity for previous influenza viruses ranged from 0% to 100%, depending on vaccine effectiveness, setting and the prevalence of persons already protected in the population (Table 2).

The vaccination coverage required to establish herd immunity against influenza viruses

Methinks the vaccinated are the real source of this epidemic. Remember what happened to the measles vaccine and how the second dose came about? It's not because of the low primary immunization rate but rather the many vaccinated who got reinfected (vaccine failure), thus creating the epidemic.

"Well, there was the Spanish Flu, which killed over 1 Million Americans (and tens of millions worldwide), because we didn’t have vaccines at the time."

Wait. Spanish flu occurred in 1918 and the last time I checked transmission-based precaution was not established until 1970. Oh I am pretty those genius doctors then knew how to take care of flu patients.

Now hold on a second. Please elaborate.

It’s not because of the low primary immunization rate but rather the many vaccinated who got reinfected (vaccine failure), thus creating the epidemic.

This is unclear.
Do you mean that a number of people who considered themselves protected nonetheless got the flu?
Which is what "vaccine failure" means to me

Or do you mean that the vaccine was giving them the flu?
Which, considering that most flu vaccines (exception is flumist) only contain dead viruses, would be quite a feat. Even Jesus did't realize this type of resurrection.

In the first case, the exact same stuff would have happened if these people haven't been vaccinated.

By Helianthus (not verified) on 19 Jan 2015 #permalink

Wait - by transmission-based precautions? You mean, "quarantines" which have been in use pretty much since the dawn of human history?

Your statements also fly in the face of basic science.

What epidemic?

Influenza is annually epidemic.

Methinks

Given the catastrophe that followed, I suggest that you avoid such anachronisms.

transmission-based precaution was not established until 1970

The term "quarantine" is a bit older than the 70's. Or the 20th century.
So are other "transmission-based precaution".

those genius doctors then knew how to take care of flu patients.

That was a bit of a cheap shot.
They did that they can, based on that little knowledge and science available at that time.
You joke, but some 19th-century physicians and nurses were indeed geniuses, to manage to do so much with so little.

By Helianthus (not verified) on 19 Jan 2015 #permalink

This is unclear.
Do you mean that a number of people who considered themselves protected nonetheless got the flu?
Which is what “vaccine failure” means to me

Yep. Those previously vaccinated who are IgG positive for flu because of reinfection (secondary vaccine failure). This, however, does not include those who actually got the FLU from the vaccine, e.g primary vaccine failure, IgM positive. Vaccination is not protection but rather a hallucination.

Okay, please explain the biological process that allows a dead or attenuated virus to cause a primary infection......

Could H1N1 be one of our old "friends" making a comeback?

Th1Th2 - is that you?

Ah, so H1N1 believes in both the scenarios I was positing. It seems to boil down to the belief that vaccinating someone makes one more susceptible to catch the disease, if not outrightly giving one the disease.

If he could now explain to me how one could get a live pathogen from a dead vaccine...

@ Lawrence

It's starting to feel like it, isn't it?

By Helianthus (not verified) on 20 Jan 2015 #permalink

H1N1 stated

The CDC’s excuse is pure crapola.

The CDC's press release stated that

During seasons when vaccine viruses and circulating influenza viruses are well matched, VE between 50 and 60 percent has been observed. H3N2 viruses have been predominant so far this season, but about 70 percent of them have been different or have “drifted” from the H3N2 vaccine virus. This likely accounts for the reduced VE.

Could you explain why this is "crapola? Thanks.

By Mephistopheles… (not verified) on 20 Jan 2015 #permalink

A good point about false sense of security.
Some Mandan probably believed that the Small Pox infections were caused by the Army, then they'd have believed that fleeing the vicinity of the fort would mean they were safe. Unfortunately they fled to surrounding villages and isolated relatives and spread the infection to those who might otherwise have been safe in their isolation.

Quarantine stations were commonplace at sea ports involved in the slave trade.
Some outbreaks of Small Pox or Yellow Fever were likely caused by slave traders smuggling slaves past the quarantine stations in order to make a quick buck.

By Greg Young (not verified) on 20 Jan 2015 #permalink

Greg Young: What did that have to do with anything? At all?

Also, do you have any evidence for your relevant claims? You're just some random stranger on the Internet. We literally have no reason to trust you.

By Gray Falcon (not verified) on 20 Jan 2015 #permalink

#52

Gambler's fallacy.

Also consider Bayesian analysis.

And sample size.

(Cryptic comments intended to get people thinking about the issue in a 'scientific' way.)

@zebra- Actually, all it does is make it sound like you know the words, but not what they mean.

By Gray Falcon (not verified) on 20 Jan 2015 #permalink

#84

How do you know what effect it will have on other readers?

Are you psychic?

Perhaps other readers are more curious than you and interested in examining the original proposition in light of those words.

Perhaps they will discover that you are wrong?

Another factor that might be in play.
Growing up in the fifties and sixties with parents that lived through the Great Depression and WW2, any time you heard the words "we are from the government, we want to help, trust us we know what we are doing" you'd be very suspicious.
With all the paranoia about big government and encroachment on freedom of choice in many aspects of day to day life its no wonder some would hesitate to get with the program.
Someone mentioned people being charged with negligent child abuse if they did not have a child vaccinated. The 1st Amendment to the U S Constitution was in part copied from a colony/state constitution that had resulted from a law that anyone who did not have their child Christened was guilty of negligent child abuse.
Freedom of choice has always been a issue here.

I do remember that in the days before any flu vaccines were available , at least on any mass scale, the boss would call everyone's attention to his simple rule," if you got the flu don't come to work or I'll fire you on the spot". That was a not uncommon approach at the time.
Most families had a bread winner and a non working home maker, if kids were kept out of school it just gave them more time with a parent and a chance to catch up on home work.
When both parents worked it was not so easy.

"
Charlotte

UK
January 17, 2015

Here comes the anecdote: I had the flu vaccine voluntarily for the first time this year and marched my husband and son for theirs too. Over Christmas, my entire work team went down like a sack of spuds and I was the only one left standing. I believe the office party had been the final opportunity to spread it around, what with all the exchange of fluids*. So I’m shilling for the flu vacc with extra enthusiasm now. Where’s my shillbucks?
*no, not that – cigars were passed round."
Good story, good thing Bill Clinton wasn't at the party, I figure few would except a cigar he offered them.

Embrace the Chaos.

By Greg Young (not verified) on 20 Jan 2015 #permalink

@zebra- Or perhaps the readers already know what those words mean, and know you're just faking it.

@Greg- Please stay on topic.

By Gray Falcon (not verified) on 20 Jan 2015 #permalink

#87

If you are more comfortable at the Greg level of science, of course you are free to stay there. Less chance of disclosing your ignorance, and you get to feel superior.

#83 - If we assume that the act of catching influenza is governed by chance alone, that about 5% of the population gets influenza annually, and that the probability of catching it one year is independent of catching it another year, the probability of an individual going 20 years without catching influenza by chance alone is about 36%. The odds of not catching it for 40 years would be about 13%, so about 1 in 8 people should be able to say this.

Thus someone's claim to not have gotten influenza in 20 years is not remarkable by chance alone.

Of course, there are other factors which are best discussed by an epidemiologist to improve on this rather simple analysis.

By Mephistopheles… (not verified) on 20 Jan 2015 #permalink

#89

The hardest part is always, first, understanding the question. Then you have to deal with the other stuff, like what you should expect as an individual. And what we would consider remarkable.

***

Greg Young question:

"I wonder what the odds are on a person who has never had a flu shot going two decades without contracting the more recent strains of flu?"

Kreb answer:

"Since about 2-5% of people will get influenza each season you can expect to get it every 20-50 years on average. Going 20 years influenza-free is par for the course, especially when you had the disease unusually frequently in the past."

@zebra- Has obfuscating your points ever helped you? I have no idea what point you're trying to make!

By Gray Falcon (not verified) on 20 Jan 2015 #permalink

"Of course, there are other factors which are best discussed by an epidemiologist to improve on this rather simple analysis."
That's what I was wondering about, those other factors.

If you aren't exposed you won't get it but if exposed every year and you don't get it, what mechanism is at work there?
Not the numbers but the biology.

It comes into the body by way of the lungs, knocks out the first line of defense and the body has to produce a second line of defense as quickly as possible.
If the virus can't knock out the first line of anti-body defenses then you'd have less chance of an infection that produces noticeable symptoms.
Could be I still get the virus but it can't knock out my first line anti-body defense and take hold. Any exceedingly mild symptoms would not be noticed by me because I've always had sinus problems.

By Greg Young (not verified) on 20 Jan 2015 #permalink

#91

Try harder. It's actually pretty obvious if you read carefully.

Also, I'm not trying to 'help myself'. Is that what you are doing?

@zebra- Please tell me what you are trying to say. I work in computer programming. Do you know what happens when someone says something ambiguous, and a computer programmer has to try to puzzle it out? Bad things.

By Gray Falcon (not verified) on 20 Jan 2015 #permalink

Anyway, the seasonal epidemic threshold is the 95th percentile for mortality due to influenza and pneumonia.

zebra,

Gambler’s fallacy.

No it isn't. The gambler's fallacy would apply if immunity to influenza and the odds of contracting it remained constant, like throwing a die. In this example that is not the case.

Perhaps I need to explain my comment to you in simpler language.
1. "Since about 2-5% of people will get influenza each season you can expect to get it every 20-50 years on average". This, I hope, is self-explanatory and is based on clinical trials that show unvaccinated people getting influenza at this rate..

2. "Going 20 years influenza-free is par for the course", Given that 2-5% of unvaccinated people get influenza in any given influenza season, this is also a statement of fact. "Par for the course" is a term borrowed from golf, meaning, "as you would expect", in case you don't know.

3. "especially when you had the disease unusually frequently in the past". By this I mean that previous exposure to influenza increases the likelihood of immunity to a circulating strain in any subsequent year, thus reducing the chances of succumbing in that year.

4. "Older people do sometimes have resistance to circulating strains, such as the H1N1 strain in 2009". Anticipating someone objecting to 4, I provided evidence that this is sometimes the case.

Also consider Bayesian analysis.

I did. Prior probability - having had four cases in his youth makes GY's chances of getting it again later lower.

And sample size.

What is wrong with looking at average frequency of influenza in a very large population, comparing it to an individual's experience and concluding this is not unusual?

(Cryptic comments intended to get people thinking about the issue in a ‘scientific’ way.)

They may have been intended to do so, but instead I think you just demonstrated your ignorance of the subject matter.

By Krebiozen (not verified) on 20 Jan 2015 #permalink

^"Anticipating someone objecting to 3", not 4.

By Krebiozen (not verified) on 20 Jan 2015 #permalink

What can happen to some children when the 60+ antigenic Hev-b proteins in natural rubber latex leach into vaccine solutions and bind to aluminum hydroxide?
Of course you know the answer Calli Arcale:
Allergy-induced regressive autism

Citations needed, mjd: your evidence demonstrating the existence of a causal association between exposure to latex allergens as a consequence of vaccination and the development of regressive autism would be...what, exactly?

You do have some, right? Because, if not, your entire argument reduces to nothing other than "Latex--scary stuff!"

#96

Before addressing anything else, are you saying that the probability of getting flu for un-vaccinated people is 2-5%? In 1, you said "people" not "un-vaccinated people", and then changed it in 2.

zebra,

Before addressing anything else, are you saying that the probability of getting flu for un-vaccinated people is 2-5%? In 1, you said “people” not “un-vaccinated people”, and then changed it in 2.

I was referring to unvaccinated people throughout, since GY clearly stated that "I’ve never had a flu shot". I thought this was obvious given the context.

By Krebiozen (not verified) on 20 Jan 2015 #permalink

“Since about 2-5% of people will get influenza each season you can expect to get it every 20-50 years on average”. This, I hope, is self-explanatory and is based on clinical trials that show unvaccinated people getting influenza at this rate.

There's no need for a trial to extract the time frame; the 65.9% odds (1 σ) for 2%–5% annual risk are 53–21 years, respectively.

#100

OK. Let's use 4%. Could you explain how

"since 4% of unvax people will get influenza each season you can expect to get it every [20] years"

and

"No it isn’t. The gambler’s fallacy would apply if immunity to influenza and the odds of contracting it remained constant, like throwing a die."

are consistent with each other?

How do you both use and not use that 4%?

zebra- The gambler's fallacy only applies to independent events. When one gets the flu one year, it is possible that holdover immunity might protect the next year.

By Gray Falcon (not verified) on 20 Jan 2015 #permalink

When one gets the flu one year, it is possible that holdover immunity might protect the next year.

Or do the opposite.

#103

Read the question.

The Gambler's fallacy is inapplicable because what's being considered is the distribution of sequences in the population.

zebra,

OK. Let’s use 4%. Could you explain how
“since 4% of unvax people will get influenza each season you can expect to get it every [20] years”

I wrote (added emphasis), "Since about 2-5% of people will get influenza each season you can expect to get it every 20-50 years on average".

and
“No it isn’t. The gambler’s fallacy would apply if immunity to influenza and the odds of contracting it remained constant, like throwing a die.”
are consistent with each other?
How do you both use and not use that 4%?

Is there something about the term "on average" that you don't understand?

Influenza once every 20 years is not unusual in an unvaccinated person, and the paper I cited suggested that older people who had influenza when they were younger may be immune to currently circulating strains, reducing incidence even more.

By Krebiozen (not verified) on 20 Jan 2015 #permalink

zebra:

Try harder. It’s actually pretty obvious if you read carefully.

Y'know, that particular strategy always struck me a little bit like the Emperor's New Literature -- if we don't understand you, clearly it's because we're not being *careful* enough, since this is all so extremely *obvious*.

Or maybe it's just that your argument stinks and you can't be bothered to explain yourself. ;-)

By Calli Arcale (not verified) on 20 Jan 2015 #permalink

#107

My only 'problem' is that you didn't answer the question. You said

"since 4% of unvax people..."

I have no problem with the word "since"; it means that you used the 4% figure to arrive at the conclusion:

"...you can expect to get it every 20 years"

"On average", whatever you intend it to mean, doesn't change that, does it?

zebra,

My only ‘problem’ is that you didn’t answer the question.

Your question was based on a misunderstanding. You asked how my statement:

Since about 2-5% of people will get influenza each season you can expect to get it every 20-50 years on average.

is consistent with my other statement:

The gambler’s fallacy would apply if immunity to influenza and the odds of contracting it remained constant, like throwing a die.

correct?

Yet that first statement does not in any way suggest I have fallen for the gambler's fallacy, does it? What I believe you misinterpreted as the gambler's fallacy was my expansion of that statement (my emphasis):

Going 20 years influenza-free is par for the course, especially when you had the disease unusually frequently in the past.

There are two elements to what I originally wrote:

1. On average unvaccinated people get influenza every 20-50 years, so going 20 years without contracting it is not unusual, and is not evidence that GY has a turbo-charged immune system as he speculated.

2. Since GY stated that he had influenza four times previous to his 20-year influenza-free period, it is possible that he had developed immunity to some of the seasonal influenza viruses that he did not succumb to in subsequent years. I linked to a paper that found elderly people were considerably less likely than younger people to succumb to the influenza A (H1N1) strain, probably because of residual immunity after exposure in previous years and because there is less antigenic drift with this strain. Influenza A(H1N1) has made an appearance during several seasons over the past few decades.

In other words GY's previous bouts of influenza may well have left him with residual immunity to some strains of influenza, making his 20-year influenza-free period even less remarkable.

You appear to have either ignored or misunderstood the second element and jumped to the mistaken conclusion that I had become a victim of the gambler's fallacy.

Is that clear enough?

One last thing.

“…you can expect to get it every 20 years”
“On average”, whatever you intend it to mean, doesn’t change that, does it?

Yes it does. Getting something every 20 years implies regularity. "On average " qualifies that to mean that you might get it every year for three years, not at all for 50 years, or whatever, but if you look at a large enough group of people you will find that every 20 years is the average. My point was that not getting influenza for 20 years is unremarkable and is certainly not evidence or even suggestive that GY has a "very high resistence to viruses" [sic].

By Krebiozen (not verified) on 20 Jan 2015 #permalink

“On average”, whatever you intend it to mean, doesn’t change that, does it?

You seem to be having difficulty accepting the notion that there actually is such a thing as cumulative risk. Is the calculation that led to my #101 half-assed? Sure. But if you want a genuine "survival analysis," you're going to have to sharpen your own pencil (PDF).

^ Put another way, from the linked paper, cumulative hazard – the population concept – "will, in low-expected number situations, give a reasonable approximation to the risk" (emphasis in original). Risk is the term that applies to individuals, and at a 2%–5% annual cumulative incidence rate, the approximation holds.

GY: The level of ignorance shown by some of both world history and recent events was shocking. Guess the internet has not been as beneficial to them as they suppose.

The “events” you refer to were ones you pulled out of your rear, you didn't source them, and you also described combat moves that were not likely to work. Of course you got piled on. And no amount of four letter acronyms is going to win you any sympathy as long as you continue with this holier than thou thing or wave the iq test thing around like a male appendage.

And as long as we're on the topic of the 1918 flu, may I point out that survival was mostly a matter of luck? Young women who weren't pregnant tended to have the best chances of survival, but many still succumbed. Pregnant women often didn't survive or lost the baby.
Young men had one of the worst survival rates, possibly because many of them had been stuck in unsanitary conditions throughout the war. I don't think training camps or forts were very sanitary either.
Everything in the book at the time got thrown at the flu- one doctor in Italy prescribed alcohol for all his patients, garlic was a common remedy, both ingested and worn.

By Politicalguineapig (not verified) on 20 Jan 2015 #permalink

You also made fun of a lifeguard, Greg. And everything Anteaus said.

Apologies for the wall. Sciblogs wasn't working for me til now.

By Politicalguineapig (not verified) on 20 Jan 2015 #permalink

"GY: The level of ignorance shown by some of both world history and recent events was shocking. Guess the internet has not been as beneficial to them as they suppose.

The “events” you refer to were ones you pulled out of your rear, you didn’t source them, and you also described combat moves that were not likely to work. "

I described no "Combat moves". The method had been written of as one adapted to women's self defense against a rapist in very limited circumstances.
You made a claim that the U S had no involvement in Korea till after WW2 and I proved you wrong.
You made some remark about people being like cats and not fighting if they know they can't win, That's a shocking lack of understanding of how many people react when attacked and current events as well as world history have proven that a great many people will knowingly sacrifice their life just to get a piece of the other guy.

Also you used your "black belt" as a badge of authority. Around here they'd use that black belt to drag hang you after stomping a mud hole in you.

No one around here had heard of the method used to temporarily suppress the breathing reflex of a drowning person. That's not surprising since I found that rendering a drowning non swimmer unconscious has long been.

Near as I can figure is that the method was not intended for common surface rescues of drowning non swimmers.
I'd first heard of this at Cape Canaveral in the early 70's from a diver and lifeguard.
Rescue divers have to be able to bring a person out of a crashed vehicle, downed aircraft, even from a space capsule.

If you look up the Forensic pathology report on drowning you'll find that allowing the victim to inhale seawater can cause death at some point after breathing is restored.
http://www.forensicpathologyonline.com/E-Book/asphyxia/drowning
Fresh water is less destructive but death can follow days or weeks after the victim has been sent home.

I was interested in whether the vagus nerve was the key to how this could work.
I have found out a few things on this but found nothing useful from the replies I'd received.

I was hoping someone knew something about the area below the armpit and how the lymph nodes there could send signals that were involved in regulating autonomic involuntary respiration.
Apparently no one who replied knew a thing about the answer to the question yet they replied any way and every post they made only confused the issue further.
Rather than answering any of my questions they continued to ask me to speculate on how this worked even though if I had known I would not have asked the question to begin with.

You at least learned that the U S Navy had engaged in battle with Korean Warrior Monks lead by Warrior Priests in 1871 and had a on again off again presence in Korea through 1905.
Bet you didn't know that a Korean Assassin murdered an American diplomat because the U S hosted treaty negotiations between the Russians and Japanese that basically gave Japan free rein in Korea.
You pontificated on the U S militarie's state of knowledge of Martial arts before WW2 and you were proven wrong. Get over it.

Now outside of the effect of certain fluids in the lungs this has nothing to do with respiratory diseases and flu.

BTW
If anyone is interested in the use of pressure points in early surgery I have found an interesting old book on the subject.

By Greg Young (not verified) on 20 Jan 2015 #permalink

If anyone here does have actual information on how the autonomic nervous system regulates the breathing by means of pressure sensors in the chest I could use that information.
I did find that there are Lymph Nodes in the area below the arm pit and some are directly over ribs where the effects of external pressure would be magnified.

PS
Temporary interruption of respiration won't kill you, its the lack of oxygen in the blood. Victims of drowning can survive without brain damage for from 5 to 9 minutes (longer in very cold water) without drawing a breath of air.
If their lungs are full of seawater when brought up they may not survive even if resuscitated.

I'm no doctor, but I've read a lot of autopsy reports and they often tell us stuff you won't find any where else.

The effects of fluids in the lungs might be something to discuss as it relates to the flu.

By Greg Young (not verified) on 20 Jan 2015 #permalink

My post 116 shows several errors due to my aphasia kicking in, part of a sentence is missing though in my mind I had typed it in. That's how this condition affects communication skills sometimes.
Sometimes I use too few words and when I have to go back and explain exactly what I meant I end up having to type five times as much as I would have if I'd made myself clear in the first place.

Also the book I found is not a book on the pressure points only, its a book on the history of Anesthetics with some information on how surgery was performed before modern anesthetics became available.

By Greg Young (not verified) on 20 Jan 2015 #permalink

Greg, I think you need a bigger shovel.

That hole is getting rather deep.

I'm not even sure what Greg is trying to prove anymore - outside of his overblown attempts to show how smart he is.

#112 and #111

Narad, thanks for the reference. I know enough statistics to meet the standard of "knowing how little I know", and as for epidemiology I don't even know that much. (A bit more after reading the paper.)

It would be useful if you and Krebiozen and some others went back and carefully read my #83, where I said specifically that my purpose was to sharpen the discussion; I did not make any claim with respect to gambler's fallacy or anything else. I apparently did succeed somewhat, although Krebiozen is still not clarifying his reasoning.

So, I understand from the your reference that people have been using a formula for a long time without really understanding it, there's a new formula going around, there are different results, and the analysis depends on specific circumstance, and even terminology is not standardized. Sounds about right.

Let's see if we can reach some understanding for lay people like myself using K's numbers "you can expect to get it every N years".

Say our unvax subject is 76. (N=20) He "should expect" to have 4 cases of flu by his 80th birthday, right? So if he hasn't had any, should he "expect" to have a very unpleasant next 4 years?

Or, using the 50-year value. At 50, would one "expect" only one case of flu over the remaining years? Sounds like the anti-vax people may have a point after all, right-- why bother?

Perhaps there's a better way to communicate whatever it is y'all are trying to say?

@zebra:

Perhaps there’s a better way to communicate whatever it is y’all are trying to say?

Perhaps you should follow your own suggestion, o ambiguous one.

By Julian Frost (not verified) on 21 Jan 2015 #permalink

"Allergy induced autism"

Latex allergy in the population is at around 1%. Sorry, but that's not a straw you can grasp at to point fingers at autism.

"Say our unvax subject is 76. (N=20) He “should expect” to have 4 cases of flu by his 80th birthday, right? So if he hasn’t had any, should he “expect” to have a very unpleasant next 4 years?"

That actually *is* the Gambler's Fallacy, the notion that the odds will "catch up" with you. They don't.

#125

OK, we're moving towards my goal of sharpening the discussion. You come down on the side that Krebiozen *is* engaging in a form of Gambler's Fallacy, then?

"
NightShadeQueen

January 21, 2015

Greg Young 117

I don’t know how the heck you got from monitoring pressure in the lungs to lymph nodes of all things. I’d suspect stretch sensors in the lungs themselves first.

If their lungs are full of seawater when brought up they may not survive even if resuscitated.

Surfactants are important you don’t say."
Thanks for the links.
Best I could do in 1970 was 5min at 25 ft.
So you at least understand the importance of preventing a submerged non swimmer from taking in a lung full of salt water. No one else seemed to get that, or admit it if they did.
With seawater its not just the surfactants its the salinity and the effects listed on the Forensic pathology site I linked to.

Pressing on the nodes to trick the autonomic nervous system bypasses the stretch sensors involved in normal breathing.
It makes breathing a choice rather than an irresistible impulse.
If the subject does not know that his autonomic reflex to inhale has been interrupted then he won't know that he should inhale.

The previous discussion that I found so vexing was on a thread about traditional medicine. The thread starter of this thread contains this
"I was reading another abcnews.com story about the flu outbreaks so far this year, and came across a doozy of a link* at the end of the story, to “10 home remedies” that are supposed to help you beat the flu. #1 is garlic (well OK, I personally believe garlic is a terrific food/condiment that probably cures a bunch of things and eat a lot of it), but #10 is…a chiropractic adjustment!!!

Yep, you’re supposed to see the chiro when flu symptoms strike, and after a little wrench here or there, flu goes bye-bye. And here I thought chiropractic was only supposed to _prevent_ flu:

http://www.hhchiro.com/chiropractic-care-colds-flu/
"
Anyway I see no further reason to discuss this, it simply takes up space.
Though I'm sure your links will be of help.

Check this out you might find it of interest
http://www.forensicpathologyonline.com/E-Book/asphyxia/drowning
The Forensic pathologist is often the only spokesman for the dead.

By Greg Young (not verified) on 21 Jan 2015 #permalink

zebra:

I bet a friend five bucks that if I roll a fair-sided die six times I will see at least one six. If I assert that I have a 1-(probability of not getting a six)^6=1-(5/6)^6=~66% chance of winning, am I falling under the Gambler's Fallacy, and if so, how do you propose statistics deal with multiple events?

By NightShadeQueen (not verified) on 21 Jan 2015 #permalink

#128,

Not GF. What's the connection to what I asked?

@Zebra- The connection is simple: You don't get to invoke the term "Gambler's Fallacy" unless it actually applies to the question at hand.

By Gray Falcon (not verified) on 21 Jan 2015 #permalink

#130

Still having trouble with reading comprehension?

Recall the definition of the gambler's fallacy, zebra: when you mistakenly believe that the outcome of an independent trial is dependent on the outcomes of previous trials (for example, if after flipping a coin 7 times and getting tails you beleived the odds of getting a heads on the next flip was greater than 50%.

So no, you would not be falling under the gambler's fallacy (unless after having rolled the die multiple times without getting a six youbelieved the odds of getting a 6 on the next roll were greater than 1 in 6.)

zebra, as it's been explained to you about a million times at this point:

If you assume the probability of getting the flu one year is independent from the probability of getting the flu any other year, the probability of not getting the flu for twenty years would be (assuming a 5% chance of getting the flu each year) .95^20 = ~.36 which is not all that improbable.

By NightShadeQueen (not verified) on 21 Jan 2015 #permalink

Well it looks like asking about the odds wasn't going go anywhere either.
Thanks for considering the question. Some can handle numbers better than others but often the numbers don't really tell the story.

Could be that having the Asian flu as a child in 1957 or there about is why the Hong Kong Flu in 1968 or 69 had so little effect on me.
Perhaps the earlier infections have resulted in a higher overall resistance to the flu.

By Greg Young (not verified) on 21 Jan 2015 #permalink

#132: That's what I said. Not GF. (#129)

#133: You are not explaining the connection of your #128 to #121. I thought that's what you were referring to?

@zebra- How is one the gambler's fallacy, and not the other?

By Gray Falcon (not verified) on 21 Jan 2015 #permalink

Say our unvax subject is 76. (N=20) He “should expect” to have 4 cases of flu by his 80th birthday, right? So if he hasn’t had any, should he “expect” to have a very unpleasant next 4 years?

At no time did Krebiozen say this or even hint at that. Krebiozen clarified

Yes it does. Getting something every 20 years implies regularity. “On average ” qualifies that to mean that you might get it every year for three years, not at all for 50 years, or whatever, but if you look at a large enough group of people you will find that every 20 years is the average.

Thus saying that on average if you flip a coin 50 times you should come up with 25 heads and 25 tails does not imply that if you end up with 25 heads and 15 tails at flip 40 that you will most likely get 10 heads in the remaining flips.

By Mephistopheles… (not verified) on 21 Jan 2015 #permalink

zebra,

It would be useful if you and Krebiozen and some others went back and carefully read my #83, where I said specifically that my purpose was to sharpen the discussion; I did not make any claim with respect to gambler’s fallacy or anything else.

So when you wrote

#52 Gambler’s fallacy. [...] (Cryptic comments intended to get people thinking about the issue in a ‘scientific’ way.)

you weren't suggesting that in my comment at #52 I had fallen for the gambler's fallacy, and that I wasn't thinking about it in a scientific way? How foolish of me to read it that way.

If I were less charitable I might think that you made a dumb mistake and now, instead of just admitting it, you are trying to pretend you didn't. Deliberate vagueness so you can later deny that you meant what you said does appear to be your favored strategy, or perhaps I misunderstand you.

I apparently did succeed somewhat, although Krebiozen is still not clarifying his reasoning.

I don't know how I can explain it more simply, but I'll try:

a) It's a fact that unvaccinated people get influenza every 20-50 years on average - I can dig out some clinical trials if you dispute this. This means that going 20 years without getting influenza is nothing unusual.

b) If you had influenza several times when you were younger you may have residual immunity to strains circulating in more recent years, meaning that a 20-year influenza-free is even less unusual. I already posed a link to a paper supporting this.

What requires clarification? What specifically don't you understand about that?

Say our unvax subject is 76. (N=20) He “should expect” to have 4 cases of flu by his 80th birthday, right? So if he hasn’t had any, should he “expect” to have a very unpleasant next 4 years?

No. That is the very opposite of what I originally wrote, and as LW pointed out that really is the gambler's fallacy.

Or, using the 50-year value. At 50, would one “expect” only one case of flu over the remaining years? Sounds like the anti-vax people may have a point after all, right– why bother?

I think it's more useful to think of it as your risk of getting influenza in any given year, which is somewhere in the region of 1 in 20 if you are unvaccinated. In a good year a vaccine can reduce this risk to 1 in 60 or less. Using an extremely safe intervention that dramatically reduces your chances of getting an infection that could easily lead to hospitalization and kill you seems a good idea to me. There is also the matter of becoming a vector for the disease and passing it on to others, especially if one happens to work in health care.

Even if those are not reasons enough, as with any public health issue it's really about large numbers of people. Small individual risks add up to large numbers of sick or dead people in a large population. Thousands of people are hospitalized and hundreds (at least) die every year and we can expect increasing vaccine uptake to reduce those numbers.

Relative risk is widely misunderstood. There are many areas where we take measures to reduce small risks even further. For example, we can only 'expect' to have a fatal car accident every 79 million miles we drive, yet we still take measures to prevent that, by wearing a seat belt and having air bags fitted. By your logic, why bother?

A driver with an alcohol concentration of 0.15 or greater (that's impaired reflexes, reaction time, and gross motor control, staggering and slurred speech drunk) is about 25 times more likely to be involved in a fatal car accident than a sober driver. So a drunk driver can 'expect' to 'only' have a fatal accident every 3.1 million miles they drive, yet I'm guessing you wouldn't condone driving a few miles drunk because the chances of a fatal accident are less than 1 in a million.

Perhaps there’s a better way to communicate whatever it is y’all are trying to say?

I taught similar concepts for years without anyone struggling with them, but perhaps you don't have the education in science and statistics those students had.

You come down on the side that Krebiozen *is* engaging in a form of Gambler’s Fallacy, then?

[Bangs head on keyboard repeatedly]

By Krebiozen (not verified) on 21 Jan 2015 #permalink

Dammit. Blockquote is, I hope, decipherable.

By Krebiozen (not verified) on 21 Jan 2015 #permalink

#137

There, K was clarifying that he *didn't* mean regularly every 20 years, which of course I never thought he meant.

But yes, he does mean that over the course of 80 years "you should expect" to get it 4 times.

The use of "on average" all the time is very confusing, whether it is an intentional fudge or not. But the meaning in this case is clear.

#138

How is "cryptic comments intended to get people thinking in a more scientific way" vague? Obviously I think there is room for improvement in how you answered Greg, and obviously I offered no hypothesis about GF. The point was to get the discussion started with GF and other things in mind-- and that's what happened.

The problem is, you keep repeating exactly the same phrases and thinking you are 'simplifying'.

Are you *not* saying that over the course of 80 years, an individual should expect to get flu 4 times?

Based on the numbers, if you take a large enough sample of the unvaccinated population you'll find that when you calculate the distribution for the number of times people caught influenza between 1 and 80 - the mean will about 4. Some people will never have contracted influenza, some will have contracted it many more than 4 times. But the mean (or perhaps the median) will be 4.

By Mephistopheles… (not verified) on 21 Jan 2015 #permalink

And I meant between the ages of 1 and 80.

By Mephistopheles… (not verified) on 21 Jan 2015 #permalink

#142

Now we're getting somewhere.

GY: Comic books are not history. I actually did study Asian history.
Finally, where the heck do you live, Batman's Gotham? You have a severe misunderstanding of gangs and people. Most criminals are lazy. Someone who is walking around, alertly, keeping an eye on their environment is going to be a less attractive target than someone engrossed in a cell phone. Furthermore, few gangs will be unarmed and they don't usually go around randomly targeting civilians. The few who do make those mistakes will disengage if their target fights back. Most of the young men who join gangs can barely manage to shoot straight, let alone fight.

By Politicalguineapig (not verified) on 21 Jan 2015 #permalink

Also, you appear to misunderstand the nature of a virus. Every single flu virus is different than the one that came before. Sometimes immunity carries over, when two strains have enough in common. Mostly, they don't. It's pretty much down to chance and personal factors in whether or not someone gets the current strain in circulation.
As far as life-guarding goes, what do you think the whole thumping on the chest or back is about? Clearing the air way " genius." (That was in quote marks, not a compliment.)

By Politicalguineapig (not verified) on 21 Jan 2015 #permalink

#144 - based on history, I assume that statement is sarcastic. However, I do not plan to revise or discuss my #144 before receiving some sort of discussion of why it is either a) incorrect or b) unhelpful.

By Mephistopheles… (not verified) on 21 Jan 2015 #permalink

With every post that zebra makes, he reminds me more and more of Humpty Dumpty.

"When I use a word," Humpty Dumpty said, in rather a scornful tone, "it means just what I choose it to mean—neither more nor less."

That combined with those annoying Flat Earth Society members who are deliberately obtuse, purportedly for the goal of "stimulating better scientific discourse".

#147

Not at all. Completely serious.

The point being, *we have the freakin data*, otherwise we couldn't test our actuarial models, right?

How about one of the putative experts here lets the rest of the world in on what the distribution looks like. Mean, median, and much much more.

Greg is x years old. What percentage of the population in his age cohort have had the same number of flu incidents?

Then maybe we can have a real discussion.

"GY: Comic books are not history. I actually did study Asian history."
Then how could you have claimed there were no Warrior Monks in Korea and that the U S had no involvement in Korea till after WW2?
If you ever studied Asian History the course you took was sadly lacking.
And how could you make any claim to maturity with remarks like that after you have been proven beyond doubt to be wrong?
I suspect too many young people get their knowledge of history from politically motivated pseudo historians like Howard Zinn and proven frauds like Ward Churchill.

As for where I grew up It was a syndicate controlled town that did in fact resemble Gotham city in some ways,
The level of violence was comparable to the worst inner city violence you read of today. There have been three mass murders at homes within sight of my front door, and two serial killers lived within a two block radius of my house.
When a young fellow known to have a compulsion to drink human blood was arrested for the murder of his mother I looked into it and convinced the police that he was an unlikely suspect and that the details of the crime suggested it was connected to the Zooman killings.
I knew the kid well enough that his hemophagia was not a sign of homicidal tendencies.
In the next county during an FBI investigation that resulted in indictments against 35 of 95 county sheriffs a body dump containing skeletal remains of hundreds of missing persons was discovered. I won't even go into the Red Lantern shoot out.
If you look up crime in East Tennessee you'll find that this area was a major trans shipment point for drugs brought into the U S by aircraft.
Knife toting black leather jacket wearing thugs hopped up on goof balls was not something from the movies it was an everyday thing in the 50's and early 60's.
By mid 70's the very mention of a black belt was the same as asking for a fight, and no amount of eastern philosophy can get you out of that sort of fight.

If you are still trying to salvage some reputation as an authority on martial arts don't bother, anyone can look into your claims and see you don't know diddly about the history of U S Military Hand to Hand Combat training and your claim that Europeans had no martial arts was just plain silly.
...........................................................................................
The page I looked up on the Hong Kong flu stated the connection between prior infection with the 1957 Asian flu virus and increased resistance to the Hong Kong virus.
I don't know much about viruses otherwise I would not bother asking questions about viruses.
Seems like you didn't bother to look up the Hong Kong virus.
Living through infections of known deadly pandemic flu strains on more than one occasion has given me some experience with the effects.

By Greg Young (not verified) on 21 Jan 2015 #permalink

Okay, I think Mr. Young has gone completely off the deep end with that last post....

zebra,

Obviously I think there is room for improvement in how you answered Greg,

Such as introducing irrelevant concepts like the GF? Very helpful.

and obviously I offered no hypothesis about GF.

Obviously? I find it hard to see any other way of interpreting your comment. Even assuming you did not, why bring it up at all?

The point was to get the discussion started with GF and other things in mind– and that’s what happened.

GF has nothing whatsoever to do with this.

The problem is, you keep repeating exactly the same phrases and thinking you are ‘simplifying’.

Which phrases do you want me to simplify? If you explain precisely what is it you do not understand about what I wrote I would be happy to explain.

But yes, he does mean that over the course of 80 years “you should expect” to get it 4 times.

I don't like the use of "expect" in this context because there is clearly great variation. Some people might not get it at all, others might get it many more times. I certainly do not mean that if you don't get it in the first 40 years of your life you can expect to get it four times in the last 40 years to even out the odds.

The use of “on average” all the time is very confusing, whether it is an intentional fudge or not. But the meaning in this case is clear.

Confusing to you, clearly. 'On average' has a clear meaning so I don't know why you find it so confusing. I think perhaps I'm assuming a level of scientific education here that you simply do not have. That's not intended as an insult, as I'm sure you have areas of expertise in which I am ignorant, and in which I would not presume to try to educate you, with cryptic comments or otherwise.

How is “cryptic comments intended to get people thinking in a more scientific way” vague?

Is that a serious question? Cryptic literally means, "having a meaning that is mysterious or obscure". How is incorrectly accusing me of committing bringing up the gambler's fallacy intended to get anyone thinking in a more scientific way?

How about one of the putative experts here lets the rest of the world in on what the distribution looks like. Mean, median, and much much more.

Greg is x years old. What percentage of the population in his age cohort have had the same number of flu incidents?

We don't really need to know those to reject the hypothesis that 20 influenza-free years is evidence of a super-charged immune system. Ball-park figures suffice, in my opinion. Still, is this what you had in mind? See in particular Figure 1 which supports my suggestion that the incidence of influenza falls with increasing age, presumably due to the " positive correlation between age and seroprotection from influenza".

By Krebiozen (not verified) on 21 Jan 2015 #permalink

"We don’t really need to know those to reject the hypothesis that 20 influenza-free years is evidence of a super-charged immune system. "
I wouldn't consider it "supercharged" though thinking back on it I've never contracted any sort of viral infection other than those four cases of the flu.
I've never contracted an STD which is a minor miracle in itself.
As for germs
I've only taken anti-biotics after surgery or tooth extractions, and cuts or other injuries have never gotten noticeably infected even if not treated promptly or not treated at all.

What I'd remembered as a post operative infection may instead have been drainage of CSF from an Arachnoid Cyst.
My memory of those days is very limited. I do remember something working its way into my ear and pushing past the ear drum. It appeared to be part of some sort of shunt. The material was very degraded and resembled a callus .
I asked a doctor about that and he said it should not have been implanted there.
I've been contacted by a organization that handles claims against Dow Corning for a class action law suit over some material they manufactured ( a Mesh used to allow bone regrowth) used in surgeries at that time. Unfortunately there's a clause that excludes surgeries that were intended to restore sight.
From what I found on the Arachnoid cyst it was probably already there and it only ballooned up as a result of the previous surgery.
They put off surgery till a large amount of blood tinged CSF came out of my left ear.
From what they say when such a cyst forms inside the skull it most often presses on the left temporal area. That's pretty much what I figured. The other factors are the contra coup injury to the upper right portion of the brain, I remember seeing those pictures. I'll have to look up just what that area controls.

I'll try to remember the time frame of the other two flu infections.

Like I said it could be that I get the flu but with symptoms so mild I don't take any notice of it.
Not an immunity but rather my body not reacting in the same manner as would be expected.

By Greg Young (not verified) on 21 Jan 2015 #permalink

Krebiozen, # 56, which prompted my original comment about being more scientific:

"Since about 2-5% of people will get influenza each season you can expect to get it every 20-50 years on average."

Krebiozen, #152:

"I don’t like the use of “expect” in this context because there is clearly great variation"

And they say Greg is crazy...

@zebra

Krebiozen, # 56, which prompted my original comment about being more scientific:

“Since about 2-5% of people will get influenza each season you can expect to get it every 20-50 years on average.”

Krebiozen, #152:

“I don’t like the use of “expect” in this context because there is clearly great variation”

And they say Greg is crazy…

And what definition of "context" are you using today, zebra? Because it certainly differs from the way that Krebiozen and everyone else uses it.

Almost forgot.
My mother never caught the flu that I know of other than the Spanish lady infection as a child. Young adults were more likely to die from Spanish Lady than the elderly or children.
The cause of death being the stronger immune system to more radical the reaction to the infection.
She took flu shots regularly after surviving cancer in her 80's.
Her exact age was unknown because birth records of the time were destroyed in a courthouse fire before they could be recorded by the state. Officially she was born in 1916 but she always told us that her mother said she was born in 1913.

When I had the Asian flu she made a poultice for my chest that was mostly made from onions. It did help a lot.
Her family were fairly well to do for those days but the community was and is secluded deep in the mountains.
Those were the real horse and buggy days.

By Greg Young (not verified) on 21 Jan 2015 #permalink

#149

Not at all. Completely serious.

My sincere apologies for doubting you.

By Mephistopheles… (not verified) on 21 Jan 2015 #permalink

GY: Sigh. I still don't see a cite, like actual books, reports, etc. And everyone knows that Zinn was exclusively concerned with American history. Not Korean, not Chinese, not Japanese. He had no interest in anything that was beyond the borders.
One of my main textbooks in the second course was In the Land of a Dying Emperor, which would probably have mentioned the Korean warrior monks since Korea was one of the first places Japan took over. The other one that I read was a memoir of the Cultural Revolution, though I don't remember the title. I don't remember what textbooks I had in the first semester, but I still have my old notes. Don't know why I should dig them out for a slowcoach like you.

As for the second part of the comment, nice try. You have obviously confused comic books and so called 'true crime' shows with your reality. And I repeat that you have never, ever stepped inside a dojo, since you seem to think it's merely a philosophy class, rather than a place to work out. I bet you would collapse after a single whitebelt class, let alone four hours.

Finally quarantine used to be one of the most effective flu-control tactics. You obviously live in self-imposed quarantine, so not having the flu for a long time is less a product of a good immune system and more of a declaration that you can't get along with your fellow humans. I suppose that's good for all the people around you, who don't have to deal with your whines and paranoia.

By Politicalguineapig (not verified) on 21 Jan 2015 #permalink

#155

Huh? What context?

"Most criminals are lazy. Someone who is walking around, alertly, keeping an eye on their environment is going to be a less attractive target than someone engrossed in a cell phone. Furthermore, few gangs will be unarmed and they don’t usually go around randomly targeting civilians. The few who do make those mistakes will disengage if their target fights back. Most of the young men who join gangs can barely manage to shoot straight, let alone fight."
Wow have you got a lot to learn, I hope the lessons don't cost you your life.
Look up "the Latin Kings" for a start.
In Florida the gang initiation in the 70's and 80's included targeting a random pedestrian and killing them just to prove you would kill on command.
Now days you'd hear about the "knock out game"
Just pulling into a parking lot at the wrong hour of the night could cause your death. Just in case you saw something they didn't want you to see.

We had Mafia try to muscle in on syndicate territory here in the late fifties and early 60's, they left well enough alone after they lost more wise guys and lower tier employees than they could afford. Most ended up in old mine shafts or rock quarries, a few were fed to pigs. "Accidental" drowning with the victim being found pressed into the mud with boot prints on his back were a warning.
When the Mexican cartel tried to move in a lot of healthy young Mexicans seem to have drowned while swimming fully clothed. They deserved it of course after they had beheaded several locals and cut one old guys fingers off with bolt cutters.

In the 50's and 60's street gangs were punks right enough, but very dangerous punks just the same. The real danger was in crossing some one with syndicate connections.
The local mob boss was a close friend of my older brother since they were teen agers, his enforcer was also a family friend and always looked kindly on me when I was a youngster. That guy looked a lot like Lucco Brazzi.

It would appear that it was compartmentalized.
Those involved in trans shipment of drugs usually kept a low profile.
Those involved in prostitution and drugs eliminated competition in a very splashy manner.
"The Florida Boys" hit squad was pretty well known in Knoxville at one time.
One of my doctors went to prison for his involvement in a drug ring, several other prominent doctors were part of that ring.

As for shear crazy stuff they found a guy near here carrying a severed head in a bucket, showing it off to passers by.
It was the head of a man who had been making inquiries in to the crazy guy's wife's disappearance. When the cops went through the madman's personal effects they realized the fob of his key ring was carved from a human kneecap.
on finding his house they found art objects made from human bone, The remains of the missing woman's skeleton were found in a well out back.
I suspect that the horror film "Motel Hell" was inspired by an all too similar case at a road side dinner in the next county. Only instead of making sausages of the customers the real culprit fed the customers to the hogs then made sausages from the hogs and fed those to customers.
Could be something in the water I suppose.

Don't be over confident, and don't underestimate some of the street gangs of today, even though I consider them punks compared to some of those that were around in the 50's and 60's.

Heck you might not even survive the traditional "on the field or off" fight after every important ball game.

Mentioning Batman reminded me of my favorite pastimes in those days, traveling by way of the roof tops. There's a sport that's very similar to this these days, moving swiftly as possible by way of rooftops and fire escapes.
My sis had mentioned that the Gotham TV series reminded her of this town in the 50's and 60's. The architecture is very similar. A mix of world of tomorrow and post Civil War buildings with the look of castles. One old grade school is now a museum and people sometimes call it Hogwarts. Pretty cool old buildings are still standing in some places.

Things aren't as bad as they were, but could go south in a hurry if conditions were right.

By Greg Young (not verified) on 21 Jan 2015 #permalink

A lot, if not all, of those "gang-killing" initiation stories are urban legends.....Snopes.com is your friend Greg.

"One of my main textbooks in the second course was In the Land of a Dying Emperor, which would probably have mentioned the Korean warrior monks since Korea was one of the first places Japan took over. "
Well then are we to take that as your belief on the subject of Warrior Monks?
If you insist on cites the list of Marines decorated in the action at the Han River fortresses is proof that the U S Navy did indeed engage the Koreans in 1871.

Korean Warrior Monks fighting the Japanese
http://en.wikipedia.org/wiki/Battle_of_Cheongju
Warrior Monk traditions
http://www.oxfordbibliographies.com/view/document/obo-9780195393521/obo…

http://en.wikipedia.org/wiki/Pohyonsa

"Many Buddhist monks were so accomplished as martial artists that they were occasionally called upon during national emergencies to fight in battles by forming unprecedented armies of warrior monks."
http://www.kuksoolwon.com/site/history

Can't open this file for some reason you'll have to read it yourself to find out what's in it.
https://www.google.com/url?q=http://www.raskb.com/transactions/VOL37/Vo…

Now that's after a very few minutes of google fu.
I can find more cites if you like, but since I knew of Korean Warrior Monks and the battles of the Fortresses on the Han river decades before Al Gore invented the internet I can't link you to a history book that would if any copies still exist be collectors items.

" I don’t remember what textbooks I had in the first semester,"
And some here believe I should remember the date and title of a magazine article I read over 40 years ago.

The ROK teach Karate for combat not putting on a show.
Part of what I was taught are common Karate kicks and how to counter them. Last time anyone tried a Dojo taught Karate kick against me I danced him around in circles and threw him in front of a passing car.
When you are taught to expect a specific reaction from an opponent and that opponent does the exact opposite you are screwed.
That's a philosophy that comes straight from the Plains Indian "Contraries".

Ordinarily I'd really love discussing history, but its wasted on you. Your book who's title you can't remember didn't mention something so you assumed that it didn't exist and made a royal ass of yourself by continuing to insist that it didn't exist rather than simply looking it up.

By Greg Young (not verified) on 21 Jan 2015 #permalink

Not to mention the "Knockout Game" being a totally made-up Faux Noise meme aimed at even further inflaming the bigotry of their target demographic.

Seriously, GY: if you really believe any of these movie tropes actually happened to you, you're suffering from Reagan's Disease.

By The Very Rever… (not verified) on 21 Jan 2015 #permalink

That ?H1N1? strain was almost the death of me. Please tell me that the strain I am not inoculated against this year isn't going to be as bad.

By Colonel Tom (not verified) on 21 Jan 2015 #permalink

Latin kings

"he quit the gang shortly after doing the shooting which was a part of his initiation ..."
https://www.google.com/url?q=http://www.ngcrc.com/ngcrc/page15.htm&sa=U…

Its the "not all" part you have to think about.
"His attackers, police say, were members of the Latin Kings, the gang that ruled his Little Village neighborhood. They grabbed the quiet teen, who weighed barely 100 pounds and stood 5 feet tall, outside his family's front door on a cold evening in March. They told him this was his initiation, police say. They were going to make him a King.

The boy said no, according to police; he would not join the gang.

In the hours that followed, the three gang members allegedly beat him, stabbed him in the neck and dumped him into the garbage can."
https://www.google.com/url?q=http://articles.chicagotribune.com/2014-06…

"Dad of Toddler Shot Six Times as Part of Gang Initiation, Prosecutor Says "
http://www.dnainfo.com/chicago/20130728/little-village/man-shoots-fathe…

"A 14-year-old boy being initiated into the Latin Kings gang was told to drop a 60-pound steel mesh trash basket from the roof of a housing project onto a police officer, prosecutors said yesterday. "

I was in Florida when a string of this sort of killings was going on.
http://www.nytimes.com/1995/11/04/nyregion/gang-tried-to-kill-officer-a…
You can call it an urban legend if you like but the victims are still dead.

Look up "the Zebra Killings", some think that was an urban legend. Hard to say after all these years.
I don't think they ever found a convincing alternative explanation.

By Greg Young (not verified) on 21 Jan 2015 #permalink

http://en.wikipedia.org/wiki/Knockout_game

"The Guardian Angels, an anti-crime group, posted warnings in Brooklyn this week about the Knockout Game, in which teenagers try to knock a stranger out with one punch. "
The Guardian Angels are a very serious Organization with a very long history of protecting the public. If they take it Seriously I'll take it seriously.
http://www.nytimes.com/2013/11/23/nyregion/knockout-game-a-spreading-me…

Since the victims are real and a number of these attacks have been caught on camera and some who were caught confessed that this was exactly what they were doing, what's the supposed controversy?
Is it some detail of why these punks are doing this that's in question?
People were doing the same stuff in the 50's and 60's with no racial motivation. So perhaps its the racial angle you are questioning.

By Greg Young (not verified) on 21 Jan 2015 #permalink

The Guardian Angels are well-meaning but have limited resources and no particular training.

The New York Times is the same newspaper that uncritically repeated all the Bush administration's claims about "weapons of mass destruction" before the second Iraq war, and only several years later admitted that maybe they should have done some fact-checking. So they'll probably retract the story about the "knockout game" around the time of the next presidential inauguration.

The supposed Head light initiations were a myth, not the Latin kings initiations.
http://www.ngcrc.com/ngcrc/page15.htm

https://www.google.com/url?q=http://www.dnainfo.com/chicago/20130728/li…

https://www.google.com/url?q=http://www.nytimes.com/1995/11/04/nyregion…

There are more news stories on known Latin Kings initiation killings.

You guys are confusing the few urban myths about such initiations with the real known killings.

The Knock Out game is the most recent label for this sort of random attack, Its been called by many names in several countries over many years.
Call it what you will its been going on off and on so long as there have been street punks.

The tendency to label something you don't understand as a myth has also been going on as long as there have been differences of opinion.

By Greg Young (not verified) on 21 Jan 2015 #permalink
By Greg Young (not verified) on 21 Jan 2015 #permalink

For some reason, probably the length of some of my posts the post seemed to disappear while awaiting moderation so I posted the same information in another post thinking the first had been lost.

It occurs to me that some here are more paranoid than they would like to believe of me.
You label something as a myth because one political party tries to make hay of it, and another party tries to minimize it for whatever reasons they may have.

Also if you hear of gang initiation killings of random victims and find a particular urban legend about gang members choosing victims by flashing their head lights you look no further.
There's a lot more going on out there than the few urban legends you may have heard of.

I'm beginning to suspect a few of you guys are "home schooled'.
I'm old enough and stove up enough to have some excuse for spending too much time on the net. Those younger and in better health should stop wasting their youth and go out and see the world before its too late.

By Greg Young (not verified) on 21 Jan 2015 #permalink

Greg continues to be Fractally Wrong

Remember, Fox News went to court for the right to lie in newscasts.

By Militant Agnostic (not verified) on 21 Jan 2015 #permalink

Most of us are educated enough to know you are full of crap.

Greg, I have seen enough of the world to recognise a blowhard when I hear one.

GY: So, you must live in Colorado. You are aware that 8 out of 10 white men there are insane, five are high, and the others are air force?

Oh and, you mean these Red Lanterns: http://dc.wikia.com/wiki/Red_Lantern_Corps

By Politicalguineapig (not verified) on 21 Jan 2015 #permalink

So you boys continue to ignore the facts in order to hold onto your politically correct delusions, that's fine with me.

https://www.google.com/url?q=http://articles.philly.com/1993-03-14/news…

http://abc7chicago.com/news/police-several-u-of-i-students-attacked-by-…

Funny thing is I never watch Fox News, Only time I hear anything about Fox News is when John Stewart goes on a tear, or when Colbert mentions something they might have said.
I don't watch the Daily Show as a news source either.
I'm an Independent Centrist.
I've tore into Birthers a few times because of their blind ignorant rants about Obama's birth certificate.
I'm no blind follower but I believe in giving the Prez a certain amount of respect no matter which party he is from.

I'm not at all surprised to find a few here are just as blinded to fact by their political filter as the Birthers.

Get real boys.

Now lets get back to the subject of Flu and Vaccines. It was just beginning to get interesting.
Seems like some want to politicize that as well.

By Greg Young (not verified) on 21 Jan 2015 #permalink

And what, pray tell, is wrong with being Air Force?

By Colonel Tom (not verified) on 21 Jan 2015 #permalink

@Colonel Tom

Ignore Political Guineapig. She habitually makes sweeping generalizations that are derogatory of anyone not like her. We've tried correcting her numerous time, but it never quite sinks in.

Oh, and as to this year's flu strains, the dominant one this year is an A-H3N2 variety. IIRC, A-H3N2 is actually associated with more severe course of disease than other strains. But take that with a grain of salt, as I'm working from memory.

@Todd W. O.Crappers. The last flu I had, which was almost certainly H1N1 was almost the death of me. I had minor to moderate heart failure at the time (Ejection Fraction of 25-30), and I went down greatly with that flu. Of course, I got it before the supplemental vaccine.

They don't tell you before you become a parent that your child become this additional vector to expose you to all these additional risks.

By Colonel Tom (not verified) on 21 Jan 2015 #permalink

Todd W.: "Yes, children are wonderful little disease factories, aren’t they?"

My most memorable kid vector illnesses were rotavirus (I had to borrow some of the kid's diapers!) and a few years later pink eye.

Colonel Tom, what you went through sounds a thousands times worse.

PGP:

Please stop using "mentally ill" as an insult. Some wonderful, decent people I know are mentally ill; some selfish people appear not to be. ("Appear" only because I wouldn't necessarily know if they were.)

@Chris, I was the last in the family to get the flu, and it was typed when my daughter contracted, so I assume there is like a 99 chance I had the same strain. They suffered a mild case, and I thought I was too, except long after the seven sick days I relapsed and lingered. I then began to retain fluids, and of course I had to be atypical showing no swelling of the ankles but putting pressure on my breathing and stomach. Took months before anyone thought to put me on diuretics. Happily, I got better. Ejection fraction back up to 35 percent now, life is good.

Until the next killer flu (Dun, Dun, Dunnnn)

By Colonel Tom (not verified) on 21 Jan 2015 #permalink

One of the worst things my kids passed on to me was conjunctivitis.

It wasn't till I had to put the drops into my own eyes that I understood why the kids carried on so much over them. Bloody hell they stung.

I suspect that the horror film “Motel Hell” was inspired by an all too similar case at a road side dinner in the next county.

Oh, G-d, not this again.

zebra,

Krebiozen, # 56, which prompted my original comment about being more scientific:
“Since about 2-5% of people will get influenza each season you can expect to get it every 20-50 years on average.”
Krebiozen, #152:
“I don’t like the use of “expect” in this context because there is clearly great variation”
And they say Greg is crazy…

The context I was referring to is the sentence the word "expect" was used in. I wrote, “you can expect to get it every 20-50 years on average”, whereas you wrote “you should expect to get it 4 times" (inaccurately quoting me, as it happens, I never wrote "should expect").

The two phrases are not equivalent. It was the latter unqualified use of "expect" I objected to.

In retrospect I probably shouldn't have used the word "expect" at all. I didn't anticipate someone nit-picking what I wrote in increasingly desperate and unsuccessful attempts to show I made a mistake.

By Krebiozen (not verified) on 22 Jan 2015 #permalink

#186

should expect v can expect--- yes, of course, I'm the one nitpicking

So I "can expect" to get flu every 20 years, on average, but that doesn't mean I "can expect" to have 4 cases of flu over the course of 80 years. Sure.

Different context? Sounds like wacky math to me.

Or maybe you just got the approach wrong and the discussion I prompted with my 'vague' question clarified things a bit?

Of you aren't dealing with someone determined to misinterpret you, though, "expect" is a good word because it is forward-looking: you expect things about the future based on statistics. That's no guarantee; as others have pointed out, with a fair die you expect about one throw in six to be a one, but it's no refutation of statistics if you throw twelve times and get three or four ones. It's just less likely, and it doesn't change the fact that we continue to expect, as you prepare to throw again, that you will throw a one about once in every six throws.

But zebra is determined to twist your words so it's probably impossible to communicate the concept in a way that s/he cannot distort.

@zebra, I am not Krebiozen but I will take the liberty of trying to clarify. Let's look at a newborn baby. You say, "assuming this child lives eighty years and is never vaccinated (and that nothing else changes like an effective campaign that eradicated flu as smallpox was eradicated), how many times should this child expect to get the flu in those eighty years?" I say, "based on probabities computed on the whole population, about four times".

We hop in our time machine, go forward seventy-six years, and talk to the man whom the baby grew into. He says, "I have never had the flu. How many times am I going to get the flu before I reach eighty?"

You say, "Krebiozen claims you will get the flu every one of the next four years because you're expected to get it four times in eighty years", but that's because you're false imputing the Gambler's Fallacy to Krebiozen.

I, on the other hand, say "you have about a 2 to 5% chance of getting the flu each year for the next four years, just as you had for each of the prior seventy-six". That's because I *don't* fall for the Gamblers Fallacy.

By the way, you want to know the probability that that baby makes it to seventy-six without every having the flu, assuming a constant 4% chance of catching the flu each year? Almost 4.5%. Unlikely, but by no means a miracle.

Ahem, @LW, not to be too outside of my field, but is it a fair di? I always had the impression that probability of catching flu was greatly increased for the young, and the very old. This being a combination of both immunity and exposure opportunity.

By Colonel Tom (not verified) on 22 Jan 2015 #permalink

@LW

I propose that zebra change his name to "Humpty".

@Colonel Tom, it's unlikely to be a fair die, certainly. But these are back of the envelope estimates of probability, not expected (there's that word again) to perfectly reflect reality. I was trying to illustrate what was meant by expecting to get the flu about every twenty years, and clarifying the meaning of the Gambler's Fallacy in this case.

#189

It seems that there are many hammers here looking for a nail.

I never said anyone was 'guilty of the gambler's fallacy'. Poor communication, certainly. But beyond that, the reasoning is questionable.

I will reprise my comment at 149:

It should be a simple matter for one of the many 'experts' here to provide actual data that illustrates the probability distribution for flu cases, for some arbitrary age cohort. (e.g. 70yr old males who think they are kung-fu masters, controlled of course for confounding variables.)

It could simply be a list of incidence by percent of population for 0,1,2,3,....9, >9. Here's the problem for me:

1) Greg asks whether his individual experience is the result of some particular cause.

2) You say "no, see, here are the probabilities I calculate by assuming randomness."

?

zebra,
This discussion is getting increasingly stupid.

should expect v can expect— yes, of course, I’m the one nitpicking

It was "can expect, on average" that you compared to an unqualified "should expect". They do not mean the same thing.

My secondary point was that if you quote someone using quotation marks it is customary to put the words the person used inside those quotation marks, not some words with a similar but not identical meaning you have made up. Even without the qualifier "on average", I don't think "can expect" and "should expect" mean exactly the same thing, since "should" is a stronger word "than "can", but it doesn't affect this discussion greatly.

So I “can expect” to get flu every 20 years, on average, but that doesn’t mean I “can expect” to have 4 cases of flu over the course of 80 years. Sure.

I'm puzzled by your apparent inability to understand simple English, particularly given your frequent complaints about other people's lack of reading comprehension. "Every 20 years" does not mean the same thing as "every 20 years, on average". "Four times in 80 years" does not mean the same thing as "every 80 years, on average".

To co-opt an old joke, does "Canadians have one testicle each, on average", mean the same thing as "Canadians have one testicle each"?

Different context? Sounds like wacky math to me.

What "wacky math" have I used? Please be specific.

Or maybe you just got the approach wrong and the discussion I prompted with my ‘vague’ question clarified things a bit?

I don't think your contribution has clarified things at all, quite the opposite anything. I think you have succeeded in making yourself look like a bit of a twit, but are amusing me with your contortions, so thanks for that.

By Krebiozen (not verified) on 22 Jan 2015 #permalink

zebra,

I never said anyone was ‘guilty of the gambler’s fallacy’.

Then why did you bring it up at all? What did you mean by "#52 Gambler’s fallacy"?

Poor communication, certainly. But beyond that, the reasoning is questionable.

The only person who seems to have had any problem understanding what I meant is you. I am amused that you complain about my poor communication when you have stated that your poor communication (in the form of "cryptic comments") was deliberate, as some sort of didactic strategy. How's that working out for you?

As for the reasoning, I disagree that it is questionable.

It should be a simple matter for one of the many ‘experts’ here to provide actual data that illustrates the probability distribution for flu cases, for some arbitrary age cohort. (e.g. 70yr old males who think they are kung-fu masters, controlled of course for confounding variables.)

I suppose you could do that, if you wanted to make a simple problem unnecessarily complex.

1) Greg asks whether his individual experience is the result of some particular cause.
2) You say “no, see, here are the probabilities I calculate by assuming randomness.”

I'm not assuming randomness, I am pointing out that GY's experience is consistent with randomness, and that therefore it is an error to think it is evidence of causality. I'm not suggesting it is evidence against causality either.

By Krebiozen (not verified) on 22 Jan 2015 #permalink

How common is it for someone to have shown no reaction at all to a Small Pox vaccination?
I just remembered getting that vaccination and the people in charge making remarks about there being no visible reaction and no scarring or any other signs showed on the skin. I don't have the round scar on the arm that almost everyone I know in my age group has.
Apparently they stopped vaccinating against small pox in 1972.
I do have pretty thick skin though. Once I slipped with a razor edged knife and cut entirely through all layers of skin and whatever underlying tissue on the back of my thumb to expose the muscles but there was no bleeding at all. The flesh sort of looked like an orange rind.
I just pressed the edges tight and a substance that looked like brown glue seeped out along the edges and after holding the edges closed for awhile the cut had sealed up like it had been super glued. It healed without bandaging.
Other surface cuts to the arms have bled very little.
I seem to heal without much scarring and after a while even the larger wounds that had formed scars no longer show any signs other than a nearly invisible lighter colored line in the skin.

I once slipped with a medium sized Phillips head screwdriver while installing a door knob and drove the screwdriver blade almost completely through my hand. I could see the impression of the cross shaped tip on the back of my hand.
That wound healed very quickly without a mark.
I wonder what sort of medical condition or genetics could allow for this sort of quick healing without scars.

If I take Tylenol even a tiny nick will bleed for quite awhile, but there's no such bleeding when I take aspirin as a blood thinner.

BTW
Who was Phillips and what did his head look like that they'd name a screwdriver after it?

By Greg Young (not verified) on 22 Jan 2015 #permalink

Failure to respond to the smallpox vaccine seems (anecdotally) surprisingly common. Both I and my best friend were vaccinated as children, and neither of us has the slightest trace of a scar. My sister does have a scar though.

Greg, I also had the smallpox vaccine but have no scar.

If you're wondering about medical or genetic conditions, you need to talk to your own doctor or a trained genetic counselor: even the actual doctors here are not your doctor.

You can google Phillips head as well as I can, if you want his bio.

#195

"I’m not assuming randomness, I am pointing out that GY’s experience is consistent with randomness, and that therefore it is an error to think it is evidence of causality. I’m not suggesting it is evidence against causality either."

You are obviously assuming randomness; how else could you determine what Greg's experience is consistent with?

"You can google Phillips head as well as I can, if you want his bio."
That was an old mechanics joke, like the left handed crescent wrench.

"
LW

January 22, 2015

Failure to respond to the smallpox vaccine seems (anecdotally) surprisingly common. Both I and my best friend were vaccinated as children, and neither of us has the slightest trace of a scar. My sister does have a scar though."

That's good to know. I was wondering if not having the scar meant the vaccination didn't take and I might not have had protection against small pox.

I got to thinking that with my looking not much more than half my age and having no vaccination scar that if in an accident they'd think I was much younger and that would make it very difficult to identify me, or my remains. Stuff like that crosses my mind now and then.

Remembering the incident where that brown gluelike substance glued the edges of the wound together also reminded me that a Shar Pei has a similar substance between the layers of its skin. It does the same thing in gluing together torn skin if they are injured.
I treated a Shar Pei for a skin condition that caused this substance to come to the surface and form an asphalt like appearance, they called it "Elephant Skin". I think they are the only breed of dog that has this stuff under their skin.

By Greg Young (not verified) on 22 Jan 2015 #permalink

zebra,

You are obviously assuming randomness; how else could you determine what Greg’s experience is consistent with?

In what data am I assuming randomness i.e. a random distribution?

If I throw a die 60 times and get a six ten times, it is consistent with random chance, even if the sixes all occurred in the first 30 throws. If someone tells me that this is evidence that the die is rigged, I am justified in disagreeing. No assumption of randomness is required, apart from the assumption that in a non-rigged die we would expect the distribution of numbers to tend towards evenness as the sample size increases.

If you mean I am assuming that influenza strikes people at random, that is because this is what large epidemiological studies tell us. There are variations in how severe influenza is, and in the incidence of complications, but not, as far as I am aware, in who contracts the virus (aside from age as I discuss below). If you are not immune to an influenza virus and you are exposed to it you will very likely succumb (greater than 80% in this review).

The study I linked to at #152 (here's the relevant figure from that study) shows us that older people (over 35) are generally less likely to get influenza than younger people, as I suggested in my original comment. This is consistent with GY's reported experience: four bouts of flu when he was younger and none during the last 17 years.

It could be that GY now has a cast-iron immune system, as this is also consistent with his reported experience. However the principle of parsimony (Occam's Razor if you prefer) tells us to choose the simplest explanation, which is random chance, along with perhaps residual immunity from previous infections, as I suggested.

By Krebiozen (not verified) on 22 Jan 2015 #permalink

#201 (a)

"If I throw a die 60 times and get a six ten times, it is consistent with random chance, even if the sixes all occurred in the first 30 throws. If someone tells me that this is evidence that the die is rigged, I am justified in disagreeing. No assumption of randomness is required, apart from the assumption that in a non-rigged die we would expect the distribution of numbers to tend towards evenness as the sample size increases."

This has nothing to do with anything we are discussing, except to demonstrate that your thinking is disordered. You making up such unrelated imaginary problems doesn't help with that.

Goodness. zebra, it's called an example. You were complaining about "wacky math" so he tried to explain how probability works.

Tell me, do you think math works differently if different things are being counted?

By Calli Arcale (not verified) on 22 Jan 2015 #permalink

#201 (b)

"If you mean I am assuming that influenza strikes people at random, that is because this is what large epidemiological studies tell us."

Really? I thought large epidemiological studies find that only 2-5% of the population contracts flu every year. Wouldn't that contradict your statement:

"There are variations in how severe influenza is, and in the incidence of complications, but not, as far as I am aware, in who contracts the virus"

What do you even mean by "in who" contracts the virus? Is this one of your scientific expressions like "par for the course"?

Humans are neither unstable nuclides nor gas molecules; "who" is not a valid characterization. People get sick due to a combination of discrete causes and conditions that we know about.

Zebra, what's your point? This whole argument is dumb, and you're just being purposefully obtuse and willfully misunderstanding statements by Krebiozen that even I, a mere literature PhD student, can follow perfectly well. To what end, I have no idea.

@Calli Arcale

Perhaps Humpty is trying to get us into a discussion of chaos theory. Apparently, approximations and examples are lost on him. He wants the minutiae, somehow thinking that getting into the weeds will make things clearer.

#201 (c)

"It could be that GY now has a cast-iron immune system, as this is also consistent with his reported experience. However the principle of parsimony (Occam’s Razor if you prefer) tells us to choose the simplest explanation, which is random chance..."

"Random chance" is not more parsimonious than "immunity".

You are inserting your preconception into your reasoning. If Greg had been having flu shots every year over that period, would you also say that "random chance" is "simpler" than "immunity"?

#205

I've explained my 'point' multiple times-- the answer to Greg was flawed. It's ok if you don't care; go back to reading Greg's dissertation on armpit strikes and mystical Korean warriors.

I don't see anyone rushing to Krebiozen's defense on the science; just through editorializing like you.

I'd rather take Bayesian statistics, view the outcome of getting X number of sixs out of sixty roles.

Or as I learned in graduate school, the odds of rolling two sixes is 1/36. The odds of someone you see walking home without his shoes having rolled boxcars is MUCH higher.

By Colonel Tom (not verified) on 22 Jan 2015 #permalink

#203

"Tell me, do you think math works differently if different things are being counted?"

In science, we use different math depending on which different things are being measured. Math and science are not the same thing.

zebra- Krebiozen had a legitimate point. Not getting the flu twenty years in a row is certainly possible by random chance, and he had the mathematics to prove it.

Now, why did you think it was a good idea to obfuscate your argument?

By Gray Falcon (not verified) on 22 Jan 2015 #permalink

#211

GF,

Seriously, were you diagnosed with some kind of reading issues when you were a kid? Nothing wrong with that, but you should take advantage of interventions that are available.

I don’t see anyone rushing to Krebiozen’s defense on the science; just through editorializing like you.

I haven't seen you actually rush to the offense against Krebiozen on any actual math and science; only by making cryptic statements and claiming not to understanding the meaning of pretty basic English words like "who."

@Greg Young, #86

"Someone mentioned people being charged with negligent child abuse if they did not have a child vaccinated."

I don't see that anyone here said that. There are legitimate reasons not to vaccinate some children.

I know that some people have said some things along those lines elsewhere. There's always a dispute over whether or not people have the right to kill or torture their children by making them ill, letting them remain ill or by subjecting them to quack treatments. Vaccines go beyond this because it also allows ideologues to kill and torture other people's children.

Personally I can't a imagine a worldview so compelling that I would hang onto it rather than save the life of a child.

By ChristineRose (not verified) on 22 Jan 2015 #permalink

zebra, here's you first comment here:

(Cryptic comments intended to get people thinking about the issue in a ‘scientific’ way.)

That was deliberate obfuscation, and you actively admitted to it.

By Gray Falcon (not verified) on 22 Jan 2015 #permalink

For another example of obfuscation:

In science, we use different math depending on which different things are being measured. Math and science are not the same thing.

No details, no examples, no sources, just a declaration we are expected to simply nod and accept as true.

By Gray Falcon (not verified) on 22 Jan 2015 #permalink

#215 and 216

You seem to have unrealistic expectations of how others should communicate. That's why I am guessing that you have had some issues in your educational experience.

If you want me to explain something you have to do more than say "I don't get it because you aren't being clear."

I made a simple statement of fact: Math and science are not the same thing. Science uses different kinds of math in different applications. Anyone who has studied math, science, or engineering, is quite aware of this.

#213

If you could tell me what "actual math and science" would be like in your opinion, I would certainly take you more seriously.

zebra,
I'm finding this increasingly bizarre exchange hilarious.

This has nothing to do with anything we are discussing, except to demonstrate that your thinking is disordered. You making up such unrelated imaginary problems doesn’t help with that.

It's a perfectly apposite example. How is my thinking disordered? Do please explain how you came to that conclusion.

“If you mean I am assuming that influenza strikes people at random, that is because this is what large epidemiological studies tell us.”
Really? I thought large epidemiological studies find that only 2-5% of the population contracts flu every year. Wouldn’t that contradict your statement:

How does 2-5% of the population contracting influenza contradict it being a random event? If someone inhales enough of an aerosol containing enough of a virus they have no immunity to they will very probably become infected. Whether a specific person is immune or not and whether they inhale enough viral particles is random and unpredictable.

“There are variations in how severe influenza is, and in the incidence of complications, but not, as far as I am aware, in who contracts the virus”
What do you even mean by “in who” contracts the virus? Is this one of your scientific expressions like “par for the course”?

Are you really unfamiliar with the word "who"? It means, "what or which person or people". Very young and very old people are more likely to suffer complications, but anyone of any age, state of health proficiency in kung fu will very probably contract influenza if they are not immune and they are exposed. Influenza infection is not more or less common in kung fu experts or in beekeepers, it is unpredictable, random.

Humans are neither unstable nuclides nor gas molecules;

Maybe not, but human behavior and disease can be measured and subjected to statistical analysis just the same. Every year 2-5% of people will contract influenza and we cannot predict which of us will get it, just as we cannot predict an individual throw of the die, but we can accurately predict the outcome of a large number of throws.

“who” is not a valid characterization.

I don't know what you mean by this. By "who" I meant "which people" which I think is obvious to everyone, apart from you, evidently.

People get sick due to a combination of discrete causes and conditions that we know about.

Unstable nuclides and gas molecules also behave as they do due to causes and conditions we know about, but their behavior is still random. In humans those "discrete causes and conditions that we know about" occur randomly.

Exposure to viruses is random, for all intents and purposes, if you mix with other humans and inhale the same air, as almost all of us do. Having immunity to a specific virus is random (unless you are vaccinated), because it depends on the random combination of exposure to a particular virus and whether or not you were infected in the past.

“Random chance” is not more parsimonious than “immunity”.

Of course it is. Why ascribe to a specific cause something that is equally attributable to chance? We don't do a scientific experiment, find that the results are consistent with chance and conclude that causality is involved.

You are inserting your preconception into your reasoning. If Greg had been having flu shots every year over that period, would you also say that “random chance” is “simpler” than “immunity”?

Yes, I would, since GY's incidence of influenza is what you would expect from chance in an unvaccinated or an vaccinated person, it would be inaccurate to claim that it was evidence of the efficacy of vaccination. We know that getting vaccinated reduces one's chances of getting influenza from studies of large numbers of people. A single person going 20 years without influenza is not remarkable even if they are unvaccinated.

I’ve explained my ‘point’ multiple times– the answer to Greg was flawed.

You haven't explained how it was flawed in any remotely coherent way.

I don’t see anyone rushing to Krebiozen’s defense on the science; just through editorializing like you.

I suspect most people are as bemused as I am by your disagreement with what is a very simple interpretation of the epidemiological facts.

Does anyone else disagree with my original statement? That not having influenza for 20 years is not remarkable given the percentage of people that get it every year?

By Krebiozen (not verified) on 22 Jan 2015 #permalink

@zebra

You seem to have unrealistic expectations of how others should communicate.

You just blew another of my irony meters, Humpty.

@zebra

You seem to have unrealistic expectations of how others should communicate.

I work in the computer industry. If you do not communicate what you want to say clearly and completely, bad things happen. I am being very realistic in my expectations.

By Gray Falcon (not verified) on 22 Jan 2015 #permalink

What do you call the point of discussion where everything is being said is so unrelated to the original topic as to unrecognizable?

By Colonel Tom (not verified) on 22 Jan 2015 #permalink

I should probably correct a typo in #219 that zebra will likely seize upon as evidence of my mental incapacity: it should read "anyone of any age, state of health or proficiency in kung fu will very probably contract influenza".

By Krebiozen (not verified) on 22 Jan 2015 #permalink

Apollo is not carefully selecting which people will get sick: it really is mostly random. Things like geography affect the odds of exposure (working in a day care center will also increase them); life history, including vaccination and previous exposure to the same or a similar pathogen, affects whether an exposure will actually make you sick.

You can model a pathogen as "wanting" to infect people and reproduce (though bacteria and viruses don't actually have a brain or desires), but it's not useful to model it as specifically wanting to infect Jim and not Greg, or your obnoxious cousin, or people with French accents. H1N1 didn't "want" to infect people who lived in New York City; it just happened to spread there early in the 2009 outbreak.

re 2009 NYC H1N1

Tell me about it!
My younger relative, a student then / now an architect, got infected and had a difficult time- her mother had to come into the city to help her get to the doctor, care for her apartment etc. She missed important class time and had to catch up on her own. She was/ is healthy, athletic and very independent minded. Well, not for those 2-3 weeks.

Her mother and I, who are of similar age, got absolutely no flu. I received the early non-H1N1 vaccine and left the designated one for people who have more chronic illness/ higher risk.

By Denice Walter (not verified) on 22 Jan 2015 #permalink

@Denice LOL, I can go one better, I had to get a pacemaker/defibrillator after a bout with the stuff. Wife and child had a very mild time.

Although, my medtronics device had done such a job of improving my heart it might have been a blessing in disguise.

I too had gotten the non-HINI vaccine.

By Colonel Tom (not verified) on 22 Jan 2015 #permalink

#219

OK, we've established that you do not understand the differences among the types of unpredictability that are colloquially termed "random"-- nuclear decay is indeed acausal, and your claims that humans and gas molecules are equivalent is just silly. Human immune systems and their response to challenge are not identical across individuals. Gas molecule's response to being bumped is.

But you've finally gotten to the flaw in your reasoning. A flaw if we accept your assumption for which you argue so vehemently.

If you believe getting the flu is completely 'random', there's no case where you can conclude *anything* about Greg. Your little calculation is bogus sleight-of-hand.

We couldn't use Greg's experience to make the case for an iron-clad immune system *even if he were the only person in the world* to go 20 years without the flu.

If I have a population of those unstable nuclides, it doesn't matter what the half-life is. The last one to decay, however long it takes, is... not remarkable.

And in closing, for the peanut gallery-- I suggested the way out a couple of times. Look at the actual distribution, which will be expressed discretely. Maybe, we could learn from that and draw some conclusions, or not.

But hey, you guys know your math and stuff so why should you listen.

If you believe getting the flu is completely ‘random’, there’s no case where you can conclude *anything* about Greg.

I haven't however seen anyone arguing that getting the flu is completely random. I have instead seen them note that it isn't statistically remarkable for someone (Greg or anyone else) to go a couple of decades without becoming infected by seasonal influenza.

If you believe getting the flu is completely ‘random’, there’s no case where you can conclude *anything* about Greg. Your little calculation is bogus sleight-of-hand.

Once again, you're merely demonstrating that your head is impervious to the notion of survival analysis. If you actually read the paper linked in #112 (as you claim), it's clear that you didn't understand a bit of it.

^ And, in fact, uniform mixing is an entirely reasonable assumption across a complete flu season.

And here we come to the heart of the matter.

Humans are neither unstable nuclides nor gas molecules; “who” is not a valid characterization. People get sick due to a combination of discrete causes and conditions that we know about.

And:

Human immune systems and their response to challenge are not identical across individuals.

This is unreconstructed Bill Maher.

^ Blockquote fail should be obvious.

The results of any study on how many times people have the flu in their lifetime could be skewed by the simple fact that for many a mild case of flu would be easily forgotten.
For others they might remember a simple but long lasting sinus problem or allergy as flu.

When trying to remember if I'd had the measles I remembered that when a child every time I ate strawberries I'd get tiny re bumps on my face and a rash, Of course a kid who loves strawberries isn't going to let that stop them.
Not sure if I ever had measles.
At 19 I had what I and others thought to be the Mumps, swelling was on one side only.
Turned out I only had an impacted wisdom tooth. Until I remembered the visual image of the dentist holding the X-ray I only remembered it as having the mumps.
If you only went by those who have medical records that prove whether or not they had an infection that would not give you a full over all picture since not that many people seek a doctor's care for non life threatening cases of the flu, and if you didn't you'd have to rely on too often faulty memories of older persons.

As for numbers. Sometimes you need a rocket scientist and other times you need a brain surgeon, but there are those rare times that you'll really need a rocket surgeon.
Sometimes methods associated with one field can be adapted to problems in another field.

These days I have nearly zero math skills, but at one time I could do fairly complex equations in my head.
Now days even trying to remember a phone number can bring on a blinding head ache.
The skills may still be there, but if every time you tried to open a door you got a powerful electric shock you'd soon stop trying to open that door.

By Greg Young (not verified) on 22 Jan 2015 #permalink

Human immune systems and their response to challenge are not identical across individuals. Gas molecule’s response to being bumped is.

False, BTW, as molecules are not in fact billiard balls.

The results of any study on how many times people have the flu in their lifetime could be skewed by the simple fact that for many a mild case of flu would be easily forgotten.

Then it's a good thing that that's not something anyone is talking about.

Here's something that might be a factor
https://www.google.com/url?q=http://www.cancer.org/treatment/treatments…

"Hero, along with scientists from Duke University Medical Center and the Duke Institute for Genome Sciences & Policy, studied participants’ gene expression to watch how the immune system reacted to the flu virus. The analysis reviewed 22,000 genes and 267 blood samples, and used a pattern recognition algorithm and several other methods to discover the genomic signatures associated with the immune response in people who get flu symptoms and those who do not.

They found significant and complex immune responses in the people who got sick and the people who didn’t. Scientists noticed changes in their blood 36 hours before some people actually felt sick. Although they understand that some people’s immune systems resist the virus, they still don’t know how or why that happens.

“There is a behind the scene active immune response even when you don’t get sick,” Hero says. “What we found were differences in their biological metabolism and gene expression. These differences had to do with antioxidants.”

"“It’s certainly possible that people who came in had a very high level of antioxidant precursors in their blood, and this may what protected them, but we’re not saying that because we don’t know. You can’t go beyond the data to make these hypotheses.”
http://www.nbcnews.com/id/44275043/ns/health-cold_and_flu/t/why-some-pe…

Since I've used Melatonin off and on for around ten years and have made it a point to eat plenty of fresh fruits and drink fruit juices for more than 20 years, perhaps the anti oxidants do provide some protection.
The only times after 1968-69 that I caught the flu I was not eating regularly (usually only two meals per day) and for the most part only fast foods.

By Greg Young (not verified) on 22 Jan 2015 #permalink

Study
"5.

research or a detailed examination and analysis of a subject, phenomenon, etc.: "

"6.

a written account of such research, examination, or analysis:"

"21.

to examine or investigate carefully and in detail: "

By Greg Young (not verified) on 22 Jan 2015 #permalink

I went at least 30 or 40 years, - unvaccinated, - without coming across the flu. Then in 2009 Swine flu rolled up that particular little accomplishment and jammed it up my whole family's a$$.

By janerella (not verified) on 22 Jan 2015 #permalink

Study
“5.

research or a detailed examination and analysis of a subject, phenomenon, etc.: ”

Don't quote the fυcking dictionary at me, dipshіt. You were talking about fυcking recall bias, which is apropos of fυcking nothing.

@Narad not to mention that many of us do not and can not tell the difference between the flu and a cold. As far as I know,

I have always found the analogy of the spread of flu like nuclear fission. A unvaccinated person is an isotope that when struck by a neutron, and vaccinated one is an inert. The airbourne virus is the neutron. Throw some neutrons into a mass of pure fissionable material and you get an explosion, throw a mass of flu/neutron into a mass with enough inert material and you don't get a self-sustaining reaction.

By Colonel Tom (not verified) on 22 Jan 2015 #permalink

"You were talking about fυcking recall bias, which is apropos of fυcking nothing."
If the supposed odds are going to be established by calculations based on studies that include surveys and questionnaires etc. its relevant.
If you are basing your initial figures on a simple WAG you are wasting your time from jump street. GIGO.

The supposed figures for infections over the course of a lifetime are not reliable because there's simply no way of knowing if those responding to a survey are giving accurate information to begin with .
Since hospital or other medical records can't tell you anything about those who are actually infected during any particular year and did not seek treatment or call in sick you are back to relying on estimates based on WAG rather than even an educated guess.

Careful now your childish streak is showing again.

By Greg Young (not verified) on 22 Jan 2015 #permalink

@Colonel Tom:

@Narad not to mention that many of us do not and can not tell the difference between the flu and a cold. As far as I know

I'm pretty sure that that's a one-way street, with a bout of influenza's being distinctly recognizable as Not A Cold, but I still don't see any relevance.

If the supposed odds are going to be established by calculations based on studies that include surveys and questionnaires etc. its relevant.

Thank you for conceding that your remark was not only completely irrelevant but based on some bizarre misunderstanding of what's actually under discussion. Allow me to remind you of it, as you seem to have a habit of pretending that such things cease to exist when they prove to be inconvenient:

The results of any study on how many times people have the flu in their lifetime could be skewed by the simple fact that for many a mild case of flu would be easily forgotten.

Now,

The supposed figures for infections over the course of a lifetime are not reliable because there’s simply no way of knowing if those responding to a survey are giving accurate information to begin with .

And where, exactly, are these "supposed figures"?

Since hospital or other medical records can’t tell you anything about those who are actually infected during any particular year and did not seek treatment or call in sick you are back to relying on estimates based on WAG rather than even an educated guess.

I take it that your undefined acronym denotes "wild-ass guess(es)," which makes this bit of stupidity, which was advanced first – viz., begging the question – burn even brighter:

If you are basing your initial figures on a simple WAG you are wasting your time from jump street. GIGO.

I'll throw you a couple of meaty bones: Tell me (1) precisely how annual influenza prevalence is estimated, (2) what a "validated instrument" is, and (3) what the relationship between Nos. 1 and 2 is.

@Colonel Tom

@Narad not to mention that many of us do not and can not tell the difference between the flu and a cold. As far as I know,

Oh. Man colds.....
Aka can pretty safely assume from that statement that you've never had the flu.

By janerella (not verified) on 22 Jan 2015 #permalink

zebra, I have no idea what you are talking about- and neither do you.

You make about as much sense as Greg.

Seriously, could we please have some better trolls?

"I’ll throw you a couple of meaty bones: Tell me (1) precisely how annual influenza prevalence is estimated, "

You are the one doing the calculations, you tell us.
If you are trying to tell us that your baseline figures are trust worthy then justify that.

If all you are interested in is what's the best method to use to come to an unreliable conclusion rather than the original question so be it.

By Greg Young (not verified) on 22 Jan 2015 #permalink

zebra,

zebra
OK, we’ve established that you do not understand the differences among the types of unpredictability that are colloquially termed “random”

No. Reading from the sidelines, what you've managed to establish is a portrayal of a whiny troll unwilling or unable to grasp simple concepts like "random", "who" or "context".

But you’ve finally gotten to the flaw in your reasoning. A flaw if we accept your assumption for which you argue so vehemently.

If you believe getting the flu is completely ‘random’, there’s no case where you can conclude *anything* about Greg. Your little calculation is bogus sleight-of-hand.

We couldn’t use Greg’s experience to make the case for an iron-clad immune system *even if he were the only person in the world* to go 20 years without the flu.

Well, if it was for example that Greg has mastered his superior kung fu qi so well he had attained inedia and had survivied without any kind of food for 20 years, we could argue that 20 survival without food was unexpected, statistically speaking.

But seriously, why couldn't we compare Greg's personal anecdote against statistics and data collected from tens/hundreds of thousands/millions of Greg's contemporaries and peers? We can see from that data that approximately one third of that pool of people could give a similar anecdote, making Greg's special hyper-immunity seem less likely explanation.

And in closing, for the peanut gallery– I suggested the way out a couple of times. Look at the actual distribution, which will be expressed discretely. Maybe, we could learn from that and draw some conclusions, or not.

I'll suggest a way out for you too. Please type the response Krebiozen should have originally made that you wouldn't have objected to - one that portrayed math and science true, and gave Greg that individuality you seem to feel a random world would rob him of.

Also, could you explain what gambler's fallacy had to do with Krebiozen's comment in #52.

And maybe use complete sentences rather than unnumbered lists or word pairs, and please no intentionally cryptic hints. Surely you can write it so there's no risk of misunderstanding.

If you had all the records of every case of the flu that required a visit to a health care provider and made a list of all the social security numbers of patients then each and every year you made a list of those social security numbers that repeatedly appeared you might get an idea of how many subjects contracted cases of the flu over the years that required medical treatment.
Of course since in the old days you didn't get assigned a social security number till you applied for one any childhood illnesses couldn't be traced by the SS number.
And it would still not reveal any information on cases of the flu that required no medical treatment.

If only cases that required a visit to a healthcare provider counted then I would be nowhere to be found in any such records.
Then the original question would have to be altered to what are the odds of a mid sixties man who has never had a flu shot and never had a case of the flu that required treatment could go for 20 plus years without catching the flu.

If you have to alter the question to fit the possible answers derived from calculations based on incomplete data from a very limited source that never addressed the question in any way shape or form to begin with then the whole exercise is a waste of time.

By Greg Young (not verified) on 22 Jan 2015 #permalink

If you have to alter the question to fit the possible answers derived from calculations based on incomplete data from a very limited source that never addressed the question in any way shape or form to begin with then the whole exercise is a waste of time.

Mr. Young, the only reason you and your random burblings haven't been added to my killfile is because I was considering the finishing things up on the now moribund thread that you've oozed out of.

I'm wholly unsurprised at your inability to muster even a semblance of a coherent response to the three hints that I offered you, but I must ask what the fυck you think is being "altered."

Why would it be relevant whether they did or didn't go to health care provider? The question was what are the chances of somebody catching or not catching flu (in a given time frame). Why would the original question have to be altered?

"Why would it be relevant whether they did or didn’t go to health care provider?"
Because the CDC gathers information on flu infections from reports by health providers.
Those who have never required medical attention for a case of the flu would not leave any record of having had the flu.
How can they begin to guess how many unreported cases there are if these cases were never reported?

The percentage of American Citizens who have received vaccinations is determined by surveys. Those at least have an actual yes or no answer to go by.

With no way of knowing how many unreported cases there are every year, much less how many unreported cases there were in the past we might as well say that anyone who has ever had the flu and never reported it has a fifty-fifty chance of catching the flu and not reporting it or not catching it at all in any particular year.
Since they aren't going to leave a record of it, any case of the flu they do catch can have no bearing on calculating the odds of catching it in later years.
No one who has never reported a case is of any use in trying to guesstimate the odds for any other person who never reported a case.
The medical records of anyone who has had a case severe enough to require medical attention is not relevant to calculating the odds for those who have never had a case that required medical attention.

Its not apples and oranges but more like naval oranges vs tangelos.

Embrace the Chaos.

By Greg Young (not verified) on 23 Jan 2015 #permalink

What I meant to ask was how - even if you tweak your story about having had the flu - we now "have to alter the question".

How do you know you had the flu for times of you never confirmed it? And how do you know you haven't had a mild one* in the last 17 years?

* admittedly, mild ones are usually not flu but something else with flu - like symptoms.

#247

"But seriously, why couldn’t we compare Greg’s personal anecdote against statistics and data collected from tens/hundreds of thousands/millions of Greg’s contemporaries and peers? We can see from that data that approximately one third of that pool of people could give a similar anecdote, making Greg’s special hyper-immunity seem less likely explanation."

No, it wouldn't make it seem less likely. Because maybe the one-third with the same experience has the same hyper-immunity. Duh.

Did you read my last sentence in the last comment, or the previous times I suggested that we do exactly what you said?

Look. At. The. Data.

There is no guarantee that it will tell us something useful, but to the extent that we eliminate variables to get closer to Greg, it might.

"I’ll suggest a way out for you too. Please type the response Krebiozen should have originally made that you wouldn’t have objected to – one that portrayed math and science true, and gave Greg that individuality you seem to feel a random world would rob him of."

If I believed, as K maintains, that Greg is in the class of entities for which getting the flu is truly random, I would say:

"Since *nothing* can affect the outcome, it doesn't matter how strong your immune system is."

Same thing I would say to an unstable nucleus that had survived 10 times its half-life.

Now, if you are something other than a whiny drive-by troll, let me know if you understand, or explain why you disagree.

And that goes for Narad as well.

Do you think one third of people have the "hyper-immunity" to flu? Or some other percentage?

Because if that were the case, there should - no matter the time frame - be around the same number of people who never get the flu, right? In addition to those gals and guys who just did not catch it that year.

But if you extend the time frame beyond the 20 years Krebiozen choose, the approximately one third gets smaller and smaller. So unless you can give solid enough reason why hyper-immunity vanes over time at the same rate simple chance does, I personally feel the probabilities make more sense.

If I believed, as K maintains, that Greg is in the class of entities for which getting the flu is truly random, I would say:
“Since *nothing* can affect the outcome, it doesn’t matter how strong your immune system is.”

But that's not what Krebiozen said. He said Grey's experience was consistent with not having special super immunity and his close to 20 year flu-free steak was not so unbelievable, with relevant numbers.

And if Krebiozen would have written that, you wouldn't have argued? Really? And you still haven explained why you brought up gambler's fallacy in the first place.

zebra,

OK, we’ve established that you do not understand the differences among the types of unpredictability that are colloquially termed “random”–

I don't think so. What we colloquially call "random" is the result of so many variables it is impossible for us to calculate them. The throw of a die is not, strictly speaking, random if we know its angular momentum, the friction of the surface it is rolling on , air resistance etc., but for all practical purposes we can act as if it is. The same goes for large populations of people and phenomena, like contracting influenza, that are the result of many incalculable variables.

nuclear decay is indeed acausal,

I'm not an expert in physics, but as I understand it it may be unpredictable (unless you know the value of that elusive and possibly imaginary hidden variable), and we do not yet know precisely why an atom decays at a specific moment, but I doubt it is acausal. Is anything in the universe acausal? But I digress.

and your claims that humans and gas molecules are equivalent is just silly.

I didn't say they were equivalent, I said that the factors that determine whether or not a person gets influenza are for all intents and purposes random. It's about what models we use.

Human immune systems and their response to challenge are not identical across individuals.

That is true, but by looking at large numbers of people we can use statistics to estimate probabilities for an individual.

Gas molecule’s response to being bumped is.

Not really, since you don't know its original velocity or direction, but I see what you mean.

But you’ve finally gotten to the flaw in your reasoning. A flaw if we accept your assumption for which you argue so vehemently.

I am filled with anticipation.

If you believe getting the flu is completely ‘random’, there’s no case where you can conclude *anything* about Greg. Your little calculation is bogus sleight-of-hand.

That betrays a breath-taking ignorance about statistics. If we apply that to throwing a die, if it is completely random we cannot conclude anything about the die even if we throw one hundred sixes in a row. That clearly isn't true. Throwing a hundred sixes in a row is so unusual we would be justified in concluding that that the die was rigged even if it is theoretically possible by chance.

We come to conclusions like this all the time based on how likely it is that something is due to chance. Finding DNA identical to a suspect's at a crime scene may theoretically be due to someone else having identical DNA by chance, which is why such things are expressed as probabilities, but it is vanishingly unlikely.

We can also say if GY's experience with regard to influenza is unusual, very unusual or not unusual at all, as in the case of not getting it for 20 years.

We couldn’t use Greg’s experience to make the case for an iron-clad immune system *even if he were the only person in the world* to go 20 years without the flu.

I'm glad you finally agree with me that Greg was mistaken in thinking his 20 influenza-free years is evidence of his iron-clad immune system. My point was that not only is it not evidence of unusual immunity, it is not unusual given the percentage of people who get influenza annually.

If I have a population of those unstable nuclides, it doesn’t matter what the half-life is. The last one to decay, however long it takes, is… not remarkable.

It is remarkable from a statistical point of view; if you picked a specific atom at the beginning, its odds of being the last to decay are very small. However it is not evidence that there is something remarkable about that particular nuclide, because, as you pointed out, as far as we know they are all the same.

Humans are not all the same, and by using this sort of modelling we can determine the probability that a specific human is part of a group of people that has, for example, immunity to a particular strain of influenza, or is part of a group that does not. A deviation from what we predict statistically is not absolute proof that that that specific human is unusual, but it is suggestive. That's how reference ranges work, a patient may have a blood test result outside the normal range by chance, since we use the mean plus or minus 2 standard deviations to estimate that range. By your argument no blood test result could ever be evidence of illness since a small number of people will always have that result by chance.

I am not arguing that GY is remarkable, I am arguing the opposite, and you now seem to have backed yourself into a corner in which you agree with me.

And in closing, for the peanut gallery– I suggested the way out a couple of times. Look at the actual distribution, which will be expressed discretely. Maybe, we could learn from that and draw some conclusions, or not.

Again, you seem to be agreeing with me that GY was mistaken in taking his experience as evidence that he is unusual. I agree that if we had more information about GY, his immune system, circulating antibodies, ethnic origin, the number of people he comes into contact with or whatever, we might be able to make a more accurate estimate of his risk of getting influenza in any given year, but that wasn't the problem we were addressing. The question was whether GY was correct in thinking that his 20 influenza-free years is prima facie evidence of an unusually robust immune system. It isn't.

But hey, you guys know your math and stuff so why should you listen.

This sort of modeling and the assumptions inherent in it are very widely used in medicine and epidemiology because they work. I spent decades working directly with relative risk calculations not too dissimilar to this, though sometimes much more complex. For example, I was involved in a prenatal screening program for Down's syndrome that involved multivariate risk analysis of blood levels of biochemical markers, gestational age, maternal age and weight to generate a single risk factor for Down's. That risk factor was based on an individual's deviation from the average, just as my estimation of how unusual GY's experience with influenza was based on whether it deviated from the average.

By Krebiozen (not verified) on 23 Jan 2015 #permalink

#254

You've already forgotten what your question was?

You asked "what if the data showed one third of people had Greg's experience of no flu for 20 years", right?

The answer stands-- the data doesn't tell me anything about why, and having hyper-immunity might be the reason.

And yes, if K had said that, I probably wouldn't have been as disturbed by the comment, and let it slide.

From your response, you obviously still don't understand what true randomness means, and why the numbers are *not* relevant. Have you heard of nuclear decay and half-life?

You asked “what if the data showed one third of people had Greg’s experience of no flu for 20 years”, right?

If you're going to use quotation marks to mark which bit of my (or anybody elses) text you are referring to, can you, you know, actually quote? Not paraphrasing or rephrasing, just copy-paste the actual bit of text.

why couldn’t we compare Greg’s personal anecdote against statistics and data collected from tens/hundreds of thousands/millions of Greg’s contemporaries and peers? We can see from that data that approximately one third of that pool of people could give a similar anecdote, making Greg’s special hyper-immunity seem less likely explanation.

I think this is what you referred to. And true, just the number of people who catch the flu on average doesn't tell us why, or why not. But that also means you can't just dismiss chance as a possible answer.

And if you are not going to test the hypothesis, just saying "hyper-immunity!" doesn't solve or mean anything. Would a study showing that while around 36% didn't catch a single flu during 20 years, only around 13% didn't have any flu during another test which span 40 years. (calculated with 5% chance of catching the flu).

As a thought experiment, if there really were such studies, would that be evidence against such hyper-immunity hypothesis? If not, why?

And yes, if K had said that, I probably wouldn’t have been as disturbed by the comment, and let it slide.

That's not the same as not objecting to, now is it?. I asked you what kind of answer you wouldn't have objected to. But that's besides the point...

This is the comment you were disturbed with enough to object: "Since about 2-5% of people will get influenza each season you can expect to get it every 20-50 years on average. Going 20 years influenza-free is par for the course, especially when you had the disease unusually frequently in the past. Older people do sometimes have resistance to circulating strains, such as the H1N1 strain in 2009."

This is how I and it seems to me almost all other commenters took it: "He said Grey’s experience was consistent with not having special super immunity and his close to 20 year flu-free steak was not so unbelievable, with relevant numbers."

I still think my summary is apt. You argue my interpretation is drastically different from the original, however, and I still can't see how. Care to elaborate? Especially since it's just you arguing it was meant differently.

Also, for the third time (just me), and for the nth time (me+everybody else): Could you explain what gambler’s fallacy had to do with Krebiozen’s comment in #52.

FWIW, there is probably a intersection of several "chance" variables that determines which people like me have not had influenza (although plenty of other viral ills) for some 40 years now. I got it once, in middle school.

And I know people in my immediate circle that seem to get it at least once every 5 years, and pretty much every time the flu vaccine selections don't match the ones circulating by the time the vaccine is produced.

Some of that may be the chance of happening to be near someone when they are shedding virus.

Some may be which particular set of immune system genes you got that may alter just how likely you are to get it or not (but again not something I expect some potion that claims to create super immunity will change and they resort every generation so you can't put an order in to get what you want).

Some may even be what particular set of antigens you've been exposed to in your life by chance that may cross react so help you tamp it down before it fires up.

But again, not anything one really has much control over and certainly not anything that makes me attribute super human body parts to myself, or say some behavioral pattern (other than once the asthma got bad I do get a flu shot every year as the mildest of colds sets off a moderately severe exacerbation and the asthma, it can kill).

I can't come up with any reason (don't do any super special immunity rituals other than my daily sneezing from severe dust allergies) for not rarely getting influenza (other than I'm also very clear on the difference between a bad cold and even a mild case of the flu). I've gotten plenty of colds, especially back when I worked in a building with exposure to a lot of people with international travel.

#257

I usually use '....' when I am paraphrasing; that was just a typo.

"And true, just the number of people who catch the flu on average doesn’t tell us why, or why not. But that also means you can’t just dismiss chance as a possible answer."

I didn't dismiss it. Why don't you just admit that I answered your point and try to move on and learn something. Without making up more hypotheticals to mask your lack of understanding. Let's stick to one thing at a time to avoid confusion.

If you don't understand the concept of half-life, just say so. Repeating over and over what K said doesn't help, because what he said is just wrong:

If the phenomenon is truly random, then the half-life is irrelevant to the experience of the individual.

If the phenomenon isn't truly random, then my suggestion of looking at actual data is the correct path.

I would explain the latter process to Greg, and explain how much work would be involved. That would be the 'least objectionable' answer; I assumed you could see that from the multiple times I have proposed it.

@Narad and @janerella Yes, I have had the flu. When the family had H1N1 my daughters and wife's symptoms did not express as anything much more than as a bad cold.

That was my musing to people to say they not gotten flu in 30 or 40 years. That was my observation from when the doctor took the swap and did the fast testing to confirm it was a case of the flu when my daughter first presented.

Add into this the four (?) families of virsuses that can cause a cold, the wildly different expressions of a cold, then I take with a grain of salt the statement that anyone makes that "I've never had the flu".

By Colonel Tom (not verified) on 23 Jan 2015 #permalink

Would the two studies and others like them I mentioned be evidence against your hypothesis of hyper-immunity? If not, why not? It's no more hypothetical than your personal definition of "random", for example.

I take your whining to mean you can't explain why my interpretation of Krebiozen's message was incorrect.

But sure, let's stick to the original messages. Why do you seem to be unable to explain why you invoked the gambler's fallacy originally, and what does it have to do with what Krebiozen said?

Prove me wrong and stop whining about other people's understanding.

zebra,
Gaist wrote:

But seriously, why couldn’t we compare Greg’s personal anecdote against statistics and data collected from tens/hundreds of thousands/millions of Greg’s contemporaries and peers? We can see from that data that approximately one third of that pool of people could give a similar anecdote, making Greg’s special hyper-immunity seem less likely explanation.

You replied:

No, it wouldn’t make it seem less likely. Because maybe the one-third with the same experience has the same hyper-immunity. Duh.

We have no evidence to suggest that such a state of hyper-immunity even exists. Inventing a phenomenon to explain data that is entirely consistent with chance really does merit a "duh". Maybe it was Russell's Teapot or the Flying Spaghetti Monster that stopped GY from getting the flu all those years.

Did you read my last sentence in the last comment, or the previous times I suggested that we do exactly what you said?
Look. At. The. Data.

The data tell us the probability of an individual contracting influenza in a given year, they tell us that almost every non-immune person will develop influenza when exposed, they tell us that influenza viruses mutate and that no one will be immune to every newly mutated virus, and they also tell us that older people are even less likely than average to contract influenza in a given year. The data are what I based my original comment on.

There is no guarantee that it will tell us something useful, but to the extent that we eliminate variables to get closer to Greg, it might.

That may be true, but it doesn't address the original question, which I am not going to reiterate yet again for fear of driving some poor observer insane.

If I believed, as K maintains, that Greg is in the class of entities for which getting the flu is truly random, I would say:
“Since *nothing* can affect the outcome, it doesn’t matter how strong your immune system is.”

No one has suggested that nothing can affect the outcome. The whole point of constructing a model that assumes that getting influenza is random is to see if an individual case is consistent with that model. If it is not consistent with a random model, we have grounds for suspecting that something other than chance has affected the outcome. If it is consistent with the random model we do not have grounds for suspecting that something other than chance has affected the outcome.

If we look at people who have been vaccinated against influenza, for example, we find that their incidence of infection deviates significantly from the random model, also known as the null hypothesis. From that we conclude that the influenza vaccine is effective and we can estimate just how effective from year to year.

Same thing I would say to an unstable nucleus that had survived 10 times its half-life.

Are you seriously suggesting that we cannot come to any conclusions about an event based on how likely it is? If so we have to discard all fingerprint and DNA evidence, and we can abandon all clinical trial results as they might all be due to chance, to name just a couple of areas that we depend on probabilities to interpret.

The macro world is full of incalculable variables, and in many areas all we can do is work with imperfect mathematical models, approximations and probabilities. Real life is messy. We have to make do with what we have, and that includes not making unfounded assumptions and inventing entities when chance is an adequate explanation..

Now, if you are something other than a whiny drive-by troll, let me know if you understand, or explain why you disagree.

You really are adorable.

And that goes for Narad as well.

Good luck with that.

By Krebiozen (not verified) on 23 Jan 2015 #permalink

@KayMarie You are now getting to a subject near and dear to my heart. To expand upon your topic you need to get fuzzy. Say you have the flu and sneeze. I am in the same room. There is (a1) chance I get your virus from an air borne droplet of your mucus. There is an (a2) chance that I touch an settled droplet. However to event (a) to lead to flu I now have to have behavior where I touch your mucus and then touch my eye, (b2) or my lips (c2). So now, the chance for you giving me the flu are (a1)+(a2)*(b2+c2).
Now, am I one of those people that is always touching my eye or lips, that chances the factors (c2) and (b2). Are the surfaces regularly cleaned, (reduces (a2)) Do I have the physical structure to keep the virus away from a mucus membrance in my nose , reduces a1.

It is this delightful combination of unavoidable risks, inherent behaviors and biological defenses. You can reduce your risks through those factors you can change, avoiding close contact, hand washing, etc, but you are lowering not reducing your risk.

At the beginning of my civilian career I worked with a lot of bright people with Master's in Public health. About a year into our work, one of the good ones dropped dead of a heart attack at 32. I heard at his wake, "he ate right, he exercised, he did everything right. He wasn't suppose to die".
No he had reduced the factors that could cause death, he decreased a lot of fuzzy variables that lead to terminal outcomes. The irony of this, is this type of fuzzy math/risk is what we were launching, how I made my wampum . Yet, still, people fall back upon the simple Calvinistic belief that things happen for a reason.

Flu is like that, you do everything right and you still might catch it, but you've really improved your odds that you won't.

By Colonel Tom (not verified) on 23 Jan 2015 #permalink

Ahem, "Lowering not eliminating your risks"

By Colonel Tom (not verified) on 23 Jan 2015 #permalink

It's kind of sad, really. All these people who feel superior to people like Greg, and they don't have the stones to take a yes or no position on this simple statement:

"If the phenomenon is truly random, then the half-life is irrelevant to the experience of the individual."

Depends on the phenomenon. Can you rephrase your question to logically and rationally involve probability of catching the flu in a given year?

Riddle me this: Is it necessary for there to be absolute randomness in order for there to be probabilities? A simple yes or no suffices.

Also, could you give me a direct quote, please, of anybody here (apart from you) invoking pure randomness.

#266 and 267

If one doesn't invoke randomness, how does one claim to calculate the probability of not having the flu for 20 years, based on the yearly incidence?

(answers both comments)

"If the phenomenon is truly random, then the half-life is irrelevant to the experience of the individual."

Why do you keep going on about half-lives? For any individual atom, the half-life allows us to compute the probability that the atom will split in any period of time. This is similar to the statistics that tell us the probability that an given individual will contract flu in any period of time.

@zebra-
By assuming that the universe is consistent. Which is, incidentally, one of the foundation principles of science.

By Gray Falcon (not verified) on 23 Jan 2015 #permalink

#269

Spot on.

And if an individual atom asked me the same kind of question that Greg asked of Krebiozen:

"Hey boss, I've gone a long time without decaying, do you think that's evidence that I have a particularly robust nuclear structure?"

And I answered:

"No, because the probability of your decay is only 50%-- look around, half your fellows are still around too.

Would my answer be correct? My exact answer, not rephrased.

(Please look back to see what I consider the correct response, #253, and explain why you would disagree if you do.)

So about half-lifes, you could say a certain material has a half life of a year, then take a kilo of that material and in one year you'd expect half of it to have decayed into its normal decay products.

Unless, of course, you'd subjected it intense bombardments of neutrons or alpha particles or what ever you need to burn the material.

Are you saying that because we can manipulate a material in such a manner that the concept of half-lives no longer has meaning? I hope not, because that would be right silly.

By Colonel Tom (not verified) on 23 Jan 2015 #permalink

You know, ordinarily it is one of the joys of my life to wonk out about probability and related matters, but it's not so much fun when it involves trying to decipher the pointlessly cryptic utterances of someone who by their own admission is deliberately being cryptic.

Perhaps if zebra actually wants to communicate eir beliefs, E might express them more clearly in the form of syllogisms. I can demonstrate how something like this would work:

1. Greg Young's experience with the flu is well within the bounds of normal chance.
2. Science does not go looking for or accept extraordinary explanations like "Greg Young has a supah-l33t immune system with KUNG FU GRIP" for phenomena that are explained just as well by normal chance.
3. Therefore science would not support Greg's belief in his Kung Fummunity.

Now, if a Korean Warrior Monk were to pop up and say "Hold on! I happen to know that a syllogism is only sound if the two foundational premises are in fact true! What if I don't accept the premise 'Greg Young's experience with the flu is well within the bounds of normal chance' - what if it seems pretty amazingly against the odds to me??"

Then we'll go through another syllogism.

1. Something which happens with a probability of 5% or higher is well within the bounds or normal chance.
2. For a person to go 20 years without a case of the flu (i.e. Greg Young's experience with the flu) has a probability higher than 5% of happening.
3. Therefore Greg Young's experience with the flu is well within the bounds of normal chance.

Now one of the Latin Kings takes a break from eating a hearty breakfast of sausages made from pigs that were fed on humans and exclaims "Wait a minute! Just how on Earth did you get the calculation that the probability of going 20 years without a case of the flu has a probability greater than 5%??"

Well, I could continue with piece by piece syllogisms, but that could take quite a while, so I am going to compress it thusly:
* On average, about 2-5% of the population in a year gets the flu.
* By the law of large numbers, this means that a normal immune system has about a 95% chance of NOT contracting the flu in a given year. If the "normal" immune system did not have such a good chance of not contracting the flu, then the percentage of people who contracted flu each year would be higher.
* If the odds of A not happening in *one* year is X, the odds of A not happening in B consecutive years is A^B, or A to the power of B. Thus, the odds of a normal immune system not contracting the flu any year for twenty consecutive years is (0.95)^20, which comes out to roughly 0.36 .
* 0.36 is a considerably higher probability than 0.05.

Now if I read right, Greg Young is now alleging that the flu is underreported and the actual incidence of the flu is much higher than 2-5%. Let's go along with that purely for the sake of argument. Suppose that we guesstimate the real incidence of flu to be 12.5% of the population - that on average, one out of eight people gets the flu in a typical year.

That makes the chance of avoiding the flu in a given year 0.875. 0.875^20 = 0.069, which is STILL greater than 5%. Even if we more than DOUBLE our estimates of how frequently the average immune system succumbs to flu, that still doesn't make avoiding flu for 20 years an improbable enough event that we need to invoke Super Kung Fu Immune Systems as an explanation.

And if you want to argue that "yes it is! It totally is!!" please present your argumentation in syllogism form.

By Antaeus Feldspar (not verified) on 23 Jan 2015 #permalink

The Rabbit.

Many years ago, Robert Nerem had directed one of his researchers to repeat an earlier experiment done at another campus. Rabbits would be feed a certain diet, then sacrificed and analyzed to determine the amount of cholesterol deposits in the coronary arteries . The previous study had noted a strong effect, when recreated under Dr. Nerem's team no such effect was noted.

Upon investigation, it was noted that one of the researchers in the lab had a pet rabbit. In fact she liked rabbits, as do I. During the day to day treatment of the experimental animals she'd talk, occasionally pet, and give a little comfort to an animal facing a terminal experiment.

So the experiment was repeated, with Maria carefully behaving normal around one group of experimental rabbits, and being cold around the others. The experiment was careful (her words, not mine) to ensure that there was no additional exercise or activity between the experimental population and the control population.

As you surely expect, results showed arterial deposits in the the "cold" group and no statistical increase in the "petted" group. This is when serious biomedical research gets reprinted in Reader's Digest.

So conclusion, the original results were validated because the dietary changes did caused increased deposits, The removal of stress was an independent variable that depressed the observed effect.

By Colonel Tom (not verified) on 23 Jan 2015 #permalink

@Antaeus Feldspar

That, my good sir, was a thing of beauty.

Before my last few posts I had been looking back to how I caught the flu at different times. The circumstances of my infection in 1969 were so entirely improbable I can barely believe it myself and at the time my reaction was "you better get off me woman or I'll knock you out".
Being accosted at random by a drunken woman who was dressed like a socialite and who came out of nowhere and grabbed my jacket like she was holding on for dear life and began firing questions about philosophy while blubbering and spraying gin and mucous with every word. That may happen every day for some people, but it seemed pretty out of the ordinary to me.
Apparently Gin has no effect on the Flu virus.
I've seen estimates of up to 11-12 million undocumented immigrants in the U S at anyone time, with WAGs of up to 20 Million. How the heck could you trust figures like that?
By definition there would be no documentation. They just extrapolate from other sources.
How many illegal immigrants ever sought healthcare at any facility that kept records during the 50's ,60's and 70's?
Some may have been hauled into the emergency room by police if found blacked out in a parking lot but few would have risked deportation over a simple case of the flu.

How can you check the records if there are no records. Un reported cases are by definition unreported.

Another time I was warm and cosy in my home with the heat turned up and reading a book when I remembered I had not collected the mail from the day before.
I went to the mail box, the sky clear and cold and suddenly a veritable cloud of the unmistakable odor of flu infection drifted by surrounding me. There was no one in sight anywhere.
I Suspect there was some sort of venting of sewer gas upwind.
The moment I realized what the odor was I knew I was in for a case of the flu.

On the other hand I've visited a badly run overcrowded nursing home where the odor of flu was so thick you could cut it with a knife, and got no infection that year.
I've transported injured people to the emergency room on several occasions when the place was packed with flu sufferers hacking and puking and got no infection from any of those exposures.

While the Universe may be constant human behavior is not.
If anyone could accurately predict human behavior there would be few wars and riots but the world would be pretty boring.

By Greg Young (not verified) on 23 Jan 2015 #permalink

“No, because the probability of your decay is only 50%– look around, half your fellows are still around too.

Would my answer be correct? My exact answer, not rephrased.

Firstly, you have rephrased everybody else's answers and questions to suit you, so it is rather hypocritical to demand fidelity from others.

Secondly, weren't you the one arguing (the straw man that) atoms and humans weren't the same thing?

But here goes. I think your hypothetical (thirdly, now who's using hypotheticals to mask your lack of understanding) boss would be wrong to attribute the atom's survival to a super-atomic structural integrity, if there was nothing out of the ordinary. That would be equivalent to assuming my gender based solely on the fact I'm human.

Why did you invoke the gambler's fallacy originally?

#277

" if there was nothing out of the ordinary."

And you would be completely wrong. Why don't you do a little research on radioactive decay. Then, if you still think that way, write a paper explaining why and submit it to a physics journal.

(You would have a better chance of acceptance if you could take a time machine back to the 19th century, though.)

Such telling silence on the first, second and third points I made, as well as on the gambler's fallacy.

So you think the boss should attribute the atom's survival to super-atomic structural integrity, based solely on the atom's word?

And care to explain in detail how Greg is an atom and Krebiozen an atomic boss?

@gaist (#279):

So you think the boss should attribute the atom’s survival to super-atomic structural integrity, based solely on the atom’s word?

Well, come on! Of course he's going to think it's a special atom! I mean, it's talking!

By Richard Smith (not verified) on 23 Jan 2015 #permalink

Greg:

If I wanted to know how reliable the estimates of the number of undocumented immigrants in the U.S. are, I would look into how the statisticians come up with those numbers. Ditto if I wasn't sure I trusted the numbers on flu infection. You don't have to choose between "we know nothing about this" and "the CDC numbers must be exactly accurate because they're professionals." In 2009, the New York City Department of Health estimated that about a million New Yorkers had been infected with H1N1, and that a lot of people's immune systems had dealt with it without the person noticing, and some others had felt "not well" but didn't realize it was the flu. They said how they'd gotten that estimate; I don't remember, but it wasn't a secret, it was in the local newspapers.

zebra:

“Random chance” is not more parsimonious than “immunity”.

Something tells me you don't know what the word "parsimonious" means if you think the explanation that requires more conditions is the parsimonious one.

By Calli Arcale (not verified) on 23 Jan 2015 #permalink

The concept of "parsimony" is an interesting one. I remember a Japanese comic where the heroes arrived at an island after hearing a distress call on the radio. Said caller was complaining about the severe cold. The island in question was on fire. The captain gave the most parsimonious explanation for this: "He's an idiot!"

By Gray Falcon (not verified) on 23 Jan 2015 #permalink

zebra,

It’s kind of sad, really. All these people who feel superior to people like Greg, and they don’t have the stones to take a yes or no position on this simple statement:
“If the phenomenon is truly random, then the half-life is irrelevant to the experience of the individual.”

Whether or not the phenomenon is truly random (it isn't, obviously, see below), I cannot see the relevance of the concept of a half-life to whether or not an individual gets influenza in any given year. Unlike a decaying nucleus, an individual can get influenza in multiple years, making half-life an inept analogy at best. Perhaps you could explain why you thought it was useful to introduce half-life to this discussion.

As I wrote at #255, what we colloquially call a “random” event is the result of so many variables it is impossible for us to calculate them. Of course who gets which particular people get influenza in any given year is not truly random. It depends on a huge number of variables that we could, in theory, use to determine which individual will get it if we knew their values, just as we could predict the fall of a die or the spin of a roulette wheel if we knew all the variables involved. But we don't, and we have no practical way of determining them.

That's why we have developed statistics using large numbers of people to estimate risks and probabilities for individuals. Our models are constructed as if these events are truly random, and although they are not, the models still prove to be very useful.

I still don't see why my original comment would have disturbed anyone. I strongly suspect you mistakenly thought I had fallen for the gambler's fallacy, and all this hot air you have emitted since has been a desperate attempt to distract from the fact that you made a mistake. That really is kind of sad.

By Krebiozen (not verified) on 23 Jan 2015 #permalink

Take Charlotte's story
Charlotte

UK
January 17, 2015

Here comes the anecdote: I had the flu vaccine voluntarily for the first time this year and marched my husband and son for theirs too. Over Christmas, my entire work team went down like a sack of spuds and I was the only one left standing. I believe the office party had been the final opportunity to spread it around, what with all the exchange of fluids*. So I’m shilling for the flu vacc with extra enthusiasm now. Where’s my shillbucks?
*no, not that – cigars were passed round.”

There we have people partying down and swapping spit, and probably leaving butt prints on the photo copier. Of course they are going to be at higher risk of catching the flu.
How do you factor that sort of all too Human behavior into a study on prevalence of flu among white collar workers?
Before the party perhaps only one or two had the begining of a case of the flu. After the party every man jack of them who wasn't vaccinated had the flu.

I've had to peel a few drunk chicks off me at parties, but a broad in a stylish fur trimmed coat roaming the streets grabbing total strangers hoping they could tell her the meaning of life is not something I was prepared for.
I was PO'ed because I had places to go and people to see and this batty broad had left huge smears of make up, tears and mucous on my brand new jacket along with enough of the odor of Gin to get me arrested on suspicion of PD.
If not for the crooked wig and smeared make up and wobbling along on high heels she would have been a real looker.
I've heard that Hong Kong flu caused some to hallucinate, I'm pretty sure she was more effected by the flu than by the alcohol.

By Greg Young (not verified) on 23 Jan 2015 #permalink

@Greg Young (#286):

I’ve heard that Hong Kong flu caused some to hallucinate,

Maybe that was flu #5 you had then..?

By Richard Smith (not verified) on 23 Jan 2015 #permalink

"Maybe that was flu #5 you had then..?"
Nope, I had Hong Kong flu in the 68-69 outbreak. I finally narrowed it down to early 69. That was the second time I caught the flu.
I had Asian flu as a youngster around 57-58.

Not sure about the years of the other two except it was before Clinton left his DNA on that blue dress.
What kind of skank would keep a nasty thing like that as a memento ?

Only recent problems I've had is the occasional pressure in my head that causes problems with spelling and math.
The fluid began draining normally again a few days back.
I may need to have a replacement shunt implanted to avoid this in the future.

Funny thing though the drainage didn't start till I tried a Yoga method to improve breathing.
I checked several sources and found that excess CSF does often drain through the nostrils and is mistaken for post nasal drip. With my chronic sinus problems recognizing what was going on took some research.

PS
Narad can't trust anything found on the CDC site anyway.
On one search page I found that someone had spelled "Flu" as "Flue". By Narad's reasoning any misspelling makes the information suspect.

By Greg Young (not verified) on 23 Jan 2015 #permalink

I’ve heard that Hong Kong flu caused some to hallucinate

I've only just started reviewing what I've missed, yet I'm already here.

I'll put it down to anti-oxidants and melatonin and my immune system rapidly producing anti-bodies that help suppress an infection to the point that if I did get the flu in the last 20 odd years It would seem to be no more than a sinus problem.

As for odds. You could calculate the odds of surviving 100 rounds of the game of Russian Roulette but each and every time you spun the cylinder and put the gun to your head the odds would still be one in six of scattering your brains on the wall.

Also I checked with a few friends in my age group to ask if they had taken their flu shots. Each of them said that their doctors had advised that they not take a flu shot because of their medical conditions. In fact they had not taken flu shots for several years in a row while under a doctor's supervision and monitoring of their condition. Apparently some doctors feel there could be adverse reactions to medications they take for their conditions, or perhaps some other factor is at work there.

By Greg Young (not verified) on 24 Jan 2015 #permalink

#283

It might be interesting to further explore your point, but there's a problem.

I've come to realize that I've discovered a population here that has the characteristics of:

1. Being critical of those who reject certain aspects of biological science, and mocking them for believing in 'woo' like "chi" and "chakras" and so on, while at the same time...

2. Themselves rejecting certain aspects of modern physics, and demonstrating a belief in 'physics woo'-- like "hidden variables".

So, unless someone can explicitly, without filibustering and deflection and rambling on and on, just agree with the statement:

"“If the phenomenon is truly random, then the half-life is irrelevant to the experience of the individual.”

...it is kind of pointless to continue.

You guys seem to take some perverse pleasure in going round and round with Greg, but I long ago stopped trying to convince the climate denialists and creationists and so on that their reasoning is faulty. On a physics blog, Krebiozen would be Greg, and everyone would be rolling their eyes.

Y'all sure do yap alot.

Seems to me, that this fight could be ended real quick. Let all of those that believe they have some sort of superimmunity and don't get vaccinated raise their hands. We'll then get all of this years major strains, rub a little in their noses, a little in their eyes. We'll then see what happens.

By Colonel Tom (not verified) on 24 Jan 2015 #permalink

" We’ll then get all of this years major strains, rub a little in their noses, a little in their eyes."
Or let a drunken broad try to crawl inside their jacket while spewing flu virus like a bio-warfare warhead.

What I'd said was
"Maybe I just have a very high resistance to viruses."
Nothing "Cast Iron" or "Hyper" or "Super" just "very high".
The other terms seem to me to be exaggerations. Exaggerations don't help a calm discussion, once they start they grow.
The page I found on the Hong Kong flu did say there was a connection to the Asian flu, those who'd had Asian flu were likely to be less effected by the Hong Kong flu. So there was a heightened resistance not some sort of full immunity.

Could be my immune system has grown stronger with each exposure, whether the exposure resulted in a true recognizable infection, and with the Melatonin stimulating the immune system even more the effect is very nearly that of having had the vaccine.

With my unusual medical situation its best I don't risk some unknown side effect from receiving the vaccine.

With the pressure lessened I can think more clearly.
I can also see that you guys will some times ride each other when one or more don't make themselves clear. I feel a bit better having seen this.

For those other visitors who thought that just because something has been depicted in a movie or a novel it can't have ever existed in real life, I think we've settled that.
Same goes for "urban legends" the legend may not be true but that doesn't make the real incidents that suggested it any less real.

It was pretty cool finding that Yoga method that actually worked to relieve pressure and swelling to let the CSF drain as it should. I was amazed at how quickly it worked. The benefits were much greater than simply making it easier to breathe.

I think I'm done with this thread see you all later.

By Greg Young (not verified) on 24 Jan 2015 #permalink

zebra,

““If the phenomenon is truly random, then the half-life is irrelevant to the experience of the individual.”

Can you explain what possible relevance this has to anything we have been discussing here?

By Krebiozen (not verified) on 24 Jan 2015 #permalink

Still waiting.

And we're still waiting on your explanation why you tried to invoke gambler's fallacy originally.

Along with all the other questions you've evaded thus far.

If I have a population of those unstable nuclides, it doesn’t matter what the half-life is. The last one to decay, however long it takes, is… not remarkable.

I take it you never asked yourself why 5 σ is used as a discovery threshold in HEP.

Apparent stability of the proton? Not remarkable.

Contracting influenza four times in a single year? Not remarkable.

Failure of plant viruses to infect humans? They just have a really long "half-life."

Nope, no conclusions whatever can be drawn about anything.

Allow me to hammer on this bit of stupidity a little bit more for the benefit of Mr. You People Are "Rejecting Certain Aspects of Modern Physics":

If the phenomenon is truly random, then the half-life is irrelevant to the experience of the individual.

This statement, by definition, applies to every individual, any specific decay time of which you have asserted is "not remarkable." Ergo, there is no such thing as half-life.

^ For complete clarity, replace "individual" with "particle in the ensemble."

Perhaps the most parsimonious explanation (based on his commenting record) is that Greg is full of shite. He reminds of another self proclaimed Mensan who told a bunch of road racing gearheads the old BS story about the line to the oil pressure gauge breaking and spraying oil all over his girlfriend's white dress (it is always a white dress). Leaving aside the question of why the oil pressure gauge was on the passenger's side, the vehicle in question was a Sunbeam Tiger with a 327 chevy motor. Virtual everyone hearing this was well aware that the Tiger (a little British sports car with an American V8 shoehorned into a space meant for 4 cylinder engine) had either a Ford 260 or 289 and that the bulkier small block chev wouldn't fit. What do people get out of BSing people in domains where half their audience has the expertise to know they are lying and the other can clearly smell the bovine byproduct from all the usual self aggrandizing tells?

I expect zebra's knowledge of "modern physics" is equivalent to their knowledge of statistics - all smug superiority and no substance. Xer blatherings about half life seem to support this.

By Militant Agnostic (not verified) on 24 Jan 2015 #permalink

Narad

Ergo, there is no such thing as half-life.

But there certainly is such a thing as a half-wit and boy howdy do we have a pair of them on this thread.

By Militant Agnostic (not verified) on 24 Jan 2015 #permalink

“Why would it be relevant whether they did or didn’t go to health care provider?”
Because the CDC gathers information on flu infections from reports by health providers.

Oh, G-d, my head.

1. Do you think this hasn't occurred to anybody before?

2. Are you confusing this with this (PDF)?

3. I offered you #243 on the off chance that you might be led to the fact that BRFSS has an ILI module, which directly leads to the question of follow-up of such data. I'll concede without reservation that this was written after spending 20 minutes of invoking such obscure search terms as "influenza," "prevalence," and "survey," though, which – as Z. has demonstrated as surely as the fact that sometimes, all triangles are in fact isosceles – makes any attempt to generalize from the experience metaphysically null and void.

@Narad, we can work on some meditation techniques for that headache of yours, although I might suggest that arguing with crazy people is high upon the list of things to avoid. After my death I found I really needed to work on staying calmer.

The idea that the CDC does not have the knowledge and experience to correlate between health care reported incidents and total incidents is mind staggering, how to put this politely, wonkledoodle. After all, you'd think an agency who's mission it is to accurately determine such things would actually hire a couple of good epidemiologist and statisticians to work on the problem. I mean, if I could do it you'd think they'd have thought of it too.

By Colonel Tom (not verified) on 24 Jan 2015 #permalink

@Narad, we can work on some meditation techniques for that headache of yours

It's OK, thank you. I'm not going start in with the Ch'an primary sources when this is to hand.

Militant Agnostic, January 25, 2015, #302:

Narad

Ergo, there is no such thing as half-life.

But there certainly is such a thing as a half-wit and boy howdy do we have a pair of them on this thread.

Which observation leads me to wonder whether half-wits combine additively or multiplicatively. In the former case, we would have a whole wit; in the latter, we have a quarter-wit.

I don't see a whole wit in this thread's trollery, so I conclude that half-wits, at least in blog threads, combine multiplicatively.

By Bill Price (not verified) on 24 Jan 2015 #permalink

#298, 300

Narad appears to be a master of woo physics. Perhaps he could resolve all this with a simple analysis.

If we follow the assumptions and calculation Krebiozen and others have been using to determine the probability of someone going 20 years without the flu (something like 1/3 as I recall), could you use your statistical expertise to tell us:

How long would Greg have to go without the flu in order for his case to be remarkable?

[Cue either silence or filibuster.]

Zebra @308
Given that Greg has demonstrated that he is a compulsive Bullshitter, I would say that the heat death of the Universe would occur first.

If we set a p value of 0.01 as remarkable, such a statistical genius as yourself should able to do the math and show their work.

By Militant Agnostic (not verified) on 25 Jan 2015 #permalink

zebra,

As you're so keen on absolute definitions for words, define remarkable.

Two standard deviations from the norm? Three? Or maybe simple, unmistakable numerical division? One in a hundred, while rare, doesn't strike me as all that remarkable. One in a thousand? One in ten thousand? One in a million?

Or to rephrase your question back at you: Do you think every one out of three is remarkable?

And how many simple questions are you evading while complaining about others filibustering?

Oh lookie!
RI has ANOTHER Greg!
Remember the chiro/ natural hygienist, Fitz______, from Oz ( a/k/a Emily, Peg etc)?
And Greggeh, the anti-vaxxer who supposedly worked with autists? Now appearing @ AoA.
Possibily another one briefly?
Happy days are here again!
( Maybe not, just loads of electrons wasted in cyberspace arguing endlessly with entertaining repartee from the regulars above).

By Denice Walter (not verified) on 25 Jan 2015 #permalink

zebra: If he's one in a million, there are 15 or 20 in the greater New York area, and more than 350 north of the U.S./Mexico border.

The statistical universe here is large--any human being who can read and write English and has access to the internet, plus potentially their friends and relatives who can't read, do so only in a language other than English, or lack internet access.

Keep that in mind when you think about what would make a poster here "remarkable" in this context. I used to know someone who was a genuine statistical anomaly, a person who had a dangerous reaction to ibuprofen. She was one of the first dozen people this happened to, and the further-unusual one of those twelve who had survived the experience. It was interesting that I knew her, but no more surprising than that my mother and brother didn't.

I have been trying to understand what zebra means. I think he is suggesting that because everything in a truly random system is random, nothing in that system is remarkable (i.e. non-random), no matter how unlikely. He'll get no argument from me there; even tautologies are true.

However he seems to argue from this that we can never deduce anything about anything, no matter how unlikely an event, even in a non-random system where non-randomness is a possibility.

I'm not sure if that's a category error, simple circular reasoning or both. Or perhaps I misunderstand him, which hopefully he will use his superior communication skills to explain what he does mean.

Otherwise, I think I'll stick with my disordered thinking.

By Krebiozen (not verified) on 25 Jan 2015 #permalink

#312, Vicki and some others who haven't been following the conversation-- "remarkable" refers to the original usage by Krebiozen, in reference to Greg's incidence of flu. Those of us who have been engaging back and forth understand what it means; either read old comments or try to catch up by listening.

#313

K,

Good on you for actually trying. Your first paragraph is correct, and since I've said exactly that more than once, I can't understand why it took you so long. Maybe it's disordered reading?

So you now agree that your original reasoning was flawed?

@ Kreb:

Right. About as 'disordered' as a fox.

And thanks for your efforts- at least you may be getting through to others.

By Denice Walter (not verified) on 25 Jan 2015 #permalink

@ Krebiozen.
1. I have Hypercholesterolemia.
2. Some people with hypercholesterolemia do not develop heart disease.
3. I should not be concerned about developing heart disease.

Is that pretty much the argument?

By Colonel Tom (not verified) on 25 Jan 2015 #permalink

Denice,

Right. About as ‘disordered’ as a fox.

:-) Speaking of which, a fox appeared on our kitchen windowsill the other night, to our surprise.

And thanks for your efforts- at least you may be getting through to others.

Thanks, I do try, and I often learn something useful myself, even when I think the person I am arguing with is talking utter bollocks.

By Krebiozen (not verified) on 25 Jan 2015 #permalink

zebra,

Good on you for actually trying. Your first paragraph is correct, and since I’ve said exactly that more than once, I can’t understand why it took you so long. Maybe it’s disordered reading?

In a random system everything is random? I understood that the first time around. That's not the part of your argument I have a problem with.

What you do not appear to understand is that we are comparing an observation (in this case an individual's reported experience) to see if it is consistent with what we expect to see in a truly random system, or not. If the observation is consistent with randomness, we are not justified in claiming it is evidence of non-randomness. If the observation is unlikely to be due to randomness, which in medical science is, by convention, anything that has a probability of less than 5%, we can agree that something non-random may be occurring.

If GY had gone 60 years without getting influenza I would agree with him that this is unusual and that perhaps something interesting was going on. Is it impossible that someone might go 60 years without getting influenza by pure chance? Yes, of course, but it is very unlikely.

This is about probabilities and making deductions about what is likely to be true. It isn't about syllogistic logic and deducing what must be true from a set of premises.

So you now agree that your original reasoning was flawed?

No I do not.

By Krebiozen (not verified) on 25 Jan 2015 #permalink

Colonel Tom,
Joining in the yapping?

1. I have Hypercholesterolemia.
2. Some people with hypercholesterolemia do not develop heart disease.
3. I should not be concerned about developing heart disease.
Is that pretty much the argument?

Not my argument, but if you take zebra's argument to its logical conclusion I think you end up somewhere similar.

Large studies tell us that those with an elevated cholesterol are more likely to develop cardiovascular disease. They also tell us that reducing an elevated cholesterol by any means reduces the risk of CVD. Therefore by reducing your cholesterol you are likely to reduce your risk of developing CVD.

How do we determine what is a normal range for cholesterol? This is something I happen to know a bit about. We take a number of normal healthy people without CVD, the more the better, and measure their blood cholesterol levels. If we plot these values on a graph of cholesterol against frequency we find they fall into what we call a normal distribution, a bell-shaped curve. We then calculate the mean and standard deviation and define the normal range as the mean plus or minus two standard deviations. We know that 95% of normal healthy people will have a cholesterol level within this range, by definition.

This tells us what cholesterol levels we see in people who are currently healthy. To predict who will get CVD decades in the future we have to do large prospective studies to see how likely people with different cholesterol levels are to develop CVD, and adjust the normal ranges accordingly. That's why the upper limit for a healthy cholesterol has fallen over the past few decades, as we get more and better data.

Another way of looking at this is to think of the population as being divided into two groups: those who will develop CVD in the future and those who will not. How do we figure out if an individual belongs to one group or the other? If we plot two curves of cholesterol against frequency for these two groups we find that they are different, that the mean cholesterol in the former group is higher, but the two curves overlap significantly. The overlap includes those people with elevated cholesterol who do not develop CVD that you mentioned, as well as those with normal cholesterol who develop CVD. By measuring an individual's cholesterol we can estimate how likely it is that they belong to the group of people that will develop CVD in the future, or not.

Assuming that things work the same way in your part of the world as they do in mine, when you see your GP for your regular cholesterol check and he tells you your chances of developing CVD in the next ten years are X% he is using an algorithm based partly on your cholesterol level to estimate this. By reducing your cholesterol and by changing other modifiable variables, such as losing weight, you can reduce your risk of developing CVD i.e. you move yourself from probably being in one group to probably being in the other.

By Krebiozen (not verified) on 25 Jan 2015 #permalink

@Krebiozen, yes I meant the argument that you are countering, and doing a fine job. I opinion that your opponents arguments, when reduced, are as ludicrous as my cholesterol argument. While one should consider if they are reducing to the ludicrous extreme, I don't think I am.

As long as we are using the cholesterol example, I could even attempt to divert and confuse with my "Niacin" example. I took niacin both over the counter and prescription for a numbers of years. Given my bad numbers, baseline above 550 and an LDL/HDL ratio above 50, it seems like an appropriate approach, best science of the day. Even though current best science is that I was wasting my money and experiencing bad "flushing" for no health gain, that does mean that my physician was wrong to prescribe niacin.

By Colonel Tom (not verified) on 25 Jan 2015 #permalink

Crappers, I meant "that does NOT mean my physician was wrong"

Anymore than my pointing out that your advice is of limited use for those of us with familial hypercholesterolemia, A fairly rare condition, that does not invalidate advice for the general population.

By Colonel Tom (not verified) on 25 Jan 2015 #permalink

#319

"What you do not appear to understand is that we are comparing an observation (in this case an individual’s reported experience) to see if it is consistent with what we expect to see in a truly random system,"

An individual not having flu for 60 years is not inconsistent with a truly random system. Why would you think that?

zebra, maybe read the rest of the paragraph.

If the observation is unlikely to be due to randomness, which in medical science is, by convention, anything that has a probability of less than 5%, we can agree that something non-random may be occurring.

#324

So getting the flu each year is not random, because the probability is less than 5%?

Sorry, I can only deal with one crazy statement at a time. Why don't you let K answer the first one first.

You really have difficulties with probabilities. And reading comprehension. The quoted bit was by Krebiozen, so it

Not catching the flu in a given year is about 95% likely. There are variations of course, but on a population wide scale and on average. So not catching it wouldn't be very uncommon.

But, as catching flu one year doesn't prevent someone from catching (or not catching) it again next year, the probability remains pretty much the same 95%. In fact, I all else being equal it might be harder for you to catch it because you may be resistant to the station that infected you.

So, nor catching the flu two years in a row is also not that uncommon. Nor is not catching it three years straight and so on and so forth.

Until one doesn't catch the flu for enough years that it becomes uncommon enough to potentially warrant investigation. In medical science this seems to be 5%, usually. The 5% limit doesn't say anything about the phenomenon, whether it is random or not, just that it is curious enough for a closer look. Is all this really that difficult, that people have to keep explaining it to you time and time again?

And before you confuse yourself more with obvious statements, how about you answer those questions people here have been asking you.

Preview would be nice.

The bit I quoted is part of the same statement - you can't just select a bit too argue against and ignore all the aid that elaborates and explains the premise further. That's the bit that seems to have been cut off from my previous post, paraphrased.

Damn auto correct.
"You can’t just select a bit to argue against and ignore all the rest that elaborates on and explains the premise further."

So, zebra, are you afraid to explain exactly what you're arguing and why in syllogism form, or did you just assume the invitation needed to be engraved?

By Antaeus Feldspar (not verified) on 25 Jan 2015 #permalink

OK, some numbers for zebra:

if the chance of getting the flu in any given year is 5%, the chance of not getting the flu that year is 95%. So if there's no connection between getting the flu in different years, your chance, or Greg's, or anyone's of not getting the flu for a 20-year period is 35%. Something that applies to about a third of the population isn't rare. With the same probabilities, the chance of not getting it for 60 years is about 4%; so starts to be noteworthy, but it's still 4 people out of every hundred.

Now, let's look at the other end of the range, where 2% of people get the flu a year. That would give you two chances in three of not getting the flu for a twenty-year period, meaning that getting the flu in that long would be more unusual than not. That's about a 30% chance of someone not getting the flu for sixty years.

Split the difference, with a 3.5% chance of the flu in a year, and you get a 49 percent chance that any given person won't have had the flu in twenty years, and slightly more than that chance that they have. Those same numbers give about one chance in eight of not getting the flu for sixty years.

This might be an argument that you don't need a flu shot (though I think it's worth it to get myself closer to a 2% than a 5% risk), but none of those numbers make it remarkable or in need of explanation that there's one person here who didn't get the flu for twenty years.

Also, there are things other than vaccination that are known to affect those odds: do you (does Greg) spend much time around small children? Are you getting enough sleep? Are you careful to wash your hands? How often do you touch your face? You may have more control over some of those variables than others; it's easier to improve your hand-washing techniques than to get more sleep if you are chronically overworked than insomniac. But look at the known variables before you assume that there must be something else going on.

I think zebra's getting stuck on the word "random". To him, it means (and, it seems, only means) completely, 100% random and without order or causal factors. In contrast, everyone else around here is using the term "random" to refer to events which have so many variables that make factoring in every single one of them in a causal analysis so burdensome that we can, for all intents and purposes, classify their contribution as "randomness".

Or, as Krebiozen put it much more succinctly, "What we colloquially call “random” is the result of so many variables it is impossible for us to calculate them."

This has been explained to Humpty Dumpty already, yet he seems to persist in his penchant for defining words solely as he wishes to use them.

Narad appears to be a master of woo physics.

Peaches, my degree may be dated, but it beats the fυck out of trying to lecture people about GR by copying and pasting from W—dia.

You have not just pathetically tried to evade the very comments you're "responding" to ("#298, 300"), but you're so fυcking desperate and enamored of yourself that you capped off the wretched display with this staggering irony:

[Cue either silence or filibuster.]

No, you do not get to change the subject. The outcome of your idiotic pronouncements about "modern physics" – and every fυcking thing you have to say about stochasticity – is that half-life doesn't exist, because every particle "gets a vote" and the polls are open forever, or some such shіt that that you're too fυcking stupid to understand represents an internal contradiction.

So go sort that the fυck out rather than calling in a fυcking backhoe.

@Narad
Never attempt to teach a pig to whistle, as it wastes your time and annoys the pig.

P.S. I did not understand the cartoon with the dirty dishes.

By Colonel Tom (not verified) on 25 Jan 2015 #permalink

P.S. I did not understand the cartoon with the dirty dishes.

I see. Mr. Natural is an R. Crumb creation, the hard-pressed guru/nemesis of Flakey Foont. And so on.

So getting the flu each year is not random, because the probability is less than 5%?

Sorry, I can only deal with one crazy statement at a time.

That's the canonical p-value for significance in the medical literature. Are you really so freaking dense that that had to be spelled out for you? Never figured out that 5 σ bit?

In other shocking news, yes, Virginia, life expectancy decreases with age. Older people, "on average," can "expect" to not have as much time left on the planet as younger people. It's insane!

@ Narad

life expectancy decreases with age. Older people, “on average,” can “expect” to not have as much time left on the planet as younger people.

Careful, the second part may not be true for teenagers, from a mathematical standpoint (but it's true from a layman, generally-speaking standpoint).

There is this little bump in mortality when they discover the joy of driving, drinking and other risky behaviors. A few genetic or teratogenic malformations may also be prone to claim their victims before they reach adulthood.
When life expectancy is reported as an average, teenagers have less time left on the planet than a somewhat 50-year old, because a good number of them unfortunately have a very early rendezvous with the Grim Reaper.
As I like to say, biology is complicated, and numbers are fun.
(but not to the point of being unable to devise models and calculate predictive values)

At some point in the past, there was also another big bump of mortality for infants. This Wikipedia article reports how Old Roman children had an average life expectancy of 21 after birth, but of 42 after becoming 5-years old.
Which would send us to the "infant diseases are mostly harmless" topic...

**OK, nitpicking mode off**
(until it auto-starts itself again)

By Helianthus (not verified) on 26 Jan 2015 #permalink

#326

The statement gaist is defending:

K:

"If the observation is unlikely to be due to randomness, which in medical science is, by convention, anything that has a probability of less than 5%, we can agree that something non-random may be occurring."

gaist's defense of K's statement:

"The 5% limit doesn’t say anything about the phenomenon, whether it is random or not,"

Huh. Maybe you are too interested in proving zebra wrong to help out your friend Krebiozen, who is mixing up certain concepts.

#333

So, Narad, let me suggest an experiment.

We set up our lab with E and M fields, and lasers, and trap a single atom in an otherwise evacuated chamber. We open a bottle of wine, and wait for the inevitable pre-5 sigma decay which we have calculated from the half-life.

But, it doesn't happen!

What would be the parsimonious course of inquiry to follow?

To all:

I originally said that I thought the reasoning behind K's comment was flawed.

And then, someone (properly) asked me what I would have said to Greg, and I provided that.

What's fascinating is that not one person has addressed my answers, which it seems to me would be the parsimonious and symmetrical way to 'figure out what zebra means'.

Just sayin'

zebra, once again reading seems to have been to hats for your confused brain.

“If the observation is unlikely to be due to randomness, which in medical science is, by convention, anything that has a probability of less than 5%, we can agree that something non-random may be occurring.”
gaist’s defense of K’s statement:
“The 5% limit doesn’t say anything about the phenomenon, whether it is random or not,”

How do you not understand what may or unlikely mean?

And you're still not answering any of the multiple questions you've been asked. Why is that? Do they also have big intimidating three letter words too hard for you?

To hats = too hard. Thank you auto correct.

zebra,

An individual not having flu for 60 years is not inconsistent with a truly random system. Why would you think that?

Perhaps my use of the word "consistent" is confusing you. I don't mean it in the logical sense, I mean it in the sense of "agreeing with", as it is commonly used in medical research, for example here (PDF):

p value: Probability that the observed results could arise by chance alone.
if p > 0.05, we say "consistent with chance", "not statistically significant".
if p < 0.05, we say "not consistent with chance", "statistically significant

When we are presented with a phenomenon and want to decide if it is likely to be due to chance or not, we use probability. The more unlikely it is that the phenomenon is due to chance, the more likely it is that a non-random factor is responsible. If something is extremely unlikely I might state that it is inconsistent with chance, even though from a strictly logical perspective, it is.

I know your single remaining undecayed nucleus is scratching his little anthropomorphic head at this point, but in its purely random system there is no other explanation for his survival when all ten billion of his friends have decayed. If another explanation existed for his survival it would be appropriate to speculate whether that was responsible instead of chance.

When DNA matching mine is found at a crime scene, a forensic scientist will give a probability that the DNA does not belong to me, and if that probability is 1 in ten million she might state that the evidence is consistent with me having been at the crime scene. The evidence is still consistent (in the logical sense) with me not having been at the crime scene, as it is possible that someone with very similar DNA left the DNA instead. Presumably there is a cut-off point where the probability of DNA belonging to an individual becomes legally compelling, just as there is in medicine where we start looking at causality instead of random chance as an explanation for a phenomenon.

To use another example from my own field, when we construct a normal range for fasting blood glucose, we use the mean glucose for normal healthy non-diabetic people plus or minus two standard deviations. If my blood glucose is two standard deviations above the mean there is a 95% possibility that this is due to chance, and this will still be considered normal. If it is six standard deviations above the mean there is only a very small probability that this is due to chance, and my doctor will suspect a non-random factor, which may be that I have developed diabetes. If a repeated test gives a similar result it is even less likely to be due to chance and, depending on local guidelines, I may acquire a diagnosis of diabetes, despite the theoretical possibility that those results are due to chance.

Does that help?

By Krebiozen (not verified) on 26 Jan 2015 #permalink

zebra,

gaist’s defense of K’s statement:
“The 5% limit doesn’t say anything about the phenomenon, whether it is random or not,”

By selective quoting you have distorted what gait wrote by omitting the rest of the sentence, which reads in full:

The 5% limit doesn’t say anything about the phenomenon, whether it is random or not, just that it is curious enough for a closer look.

I wrote:

If the observation is unlikely to be due to randomness, which in medical science is, by convention, anything that has a probability of less than 5%, we can agree that something non-random may be occurring.

Those two statements seem to me to be in complete agreement. Perhaps you could explain why they are not.

Krebiozen [...] is mixing up certain concepts.

Which concepts are those?

The only person I can see mixing up concepts here is you. You appear to think that if something could have happened by chance it must have happened by chance.

Do you think we should abandon diagnosing diabetes on the basis of an elevated fasting blood glucose, on the grounds that we can expect to see elevated blood glucose levels in non-diabetics purely by chance?

By Krebiozen (not verified) on 26 Jan 2015 #permalink

Sorry, typo, gaist, not gait.

By Krebiozen (not verified) on 26 Jan 2015 #permalink

#344

I appreciate your clarification of your use of an obscure term of art-- perhaps now ToddW will start calling you Humpty as well?

Your answer is well organized, and it does help when you say this:

"If another explanation existed for his survival it would be appropriate to speculate whether that was responsible instead of chance."

The thing is, if another explanation existed, then we wouldn't think the outcome is the result of chance in the first place. And that's why my answers to Greg are correct, and your answer is wrong.

In fact, when I suggested that Greg's conjecture-- that his immune response was different, was a reasonable one, I was attacked as being an anti-vaxer-- Narad I believe compared me to Bill Maher.

So, I await Narad's response to my thought experiment.

@Narad, I AM familiar with the work of Crumb. Never saw the movie, but have read some of his work. I just don't get the joke, but that is ok. Just because I've never smoked weed it does not mean I don't find much of Crumb's work funny. Just I don't get that one.

By Colonel Tom (not verified) on 26 Jan 2015 #permalink

zebra,
I'm intrigued by your apparent inability to understand this. I have taught similar concepts to dozens of students over the years without any problems. You are clearly not unintelligent, so I suspect you are simply unfamiliar with thinking about things in terms of probabilities. Hard sciences like math and physics deal either with logical certainties or with probabilities that are often greater than 99.99%. When dealing with biological systems, such as people, things get much vaguer, and we have had to adopt a number of assumptions and rules of thumb to deal with this. I cannot provide a logical proof that using 95% confidence limits is valid, for example, I can merely state that this has proven to be a useful cut-off point for practical purposes. Some people have difficulties dealing with uncertainty and ambiguity and I am wondering if you are one of them.

I appreciate your clarification of your use of an obscure term of art– perhaps now ToddW will start calling you Humpty as well?

I'm not sure what you mean. Is "consistent" an obscure term of art? I used the term in a discussion about a medical problem in the way it is regularly used in medicine (I even offered an example). I clarified the way I was using it because it occurred to me that you might be unfamiliar with this usage. Why would that warrant Todd comparing me to Humpty Dumpty who notoriously claimed that a word means whatever he wants it to mean?

The thing is, if another explanation existed, then we wouldn’t think the outcome is the result of chance in the first place. And that’s why my answers to Greg are correct, and your answer is wrong.

So if there is any possible explanation for a phenomenon other than chance we automatically adopt it? That's ridiculous.

There are (at least) two possible explanations for Greg's reported experience: random chance and super-immunity. So how do we decide which of these explanations is more likely? Again the principle of parsimony suggests we choose chance as an explanation unless that is very unlikely, which in this example it is definitely not.

We are often faced with similar decisions in medical science. For example, let's say that in a clinical trial those given a new drug designed to treat hypertension have a greater decrease in blood pressure than those given a placebo. If a statistical analysis of these results tells us that p = 0.08, we do not simply assume that the drug works, because we know that there is an 8% probability that these results are due to chance, and we are looking for a 5% probability of chance or less.

In other words we do not assume that a possible explanation for these results (i.e. effects of the drug) must be the explanation because the possibility that chance is responsible is too high. The results may be regarded as promising, depending on the effect size, but the drug would certainly not be approved on the basis of these results.

In fact, when I suggested that Greg’s conjecture– that his immune response was different, was a reasonable one, I was attacked as being an anti-vaxer– Narad I believe compared me to Bill Maher.

Greg's original suggestion was that his experience was evidence for a different immune response. I don't think it is reasonable to take Greg's experience as evidence that his immune response is different when chance is a likely explanation. It is reasonable to state that a different immune response is a possible explanation for his experience, but that is not the same thing as a likely explanation, especially when we have no reason to think a state of super-immunity to influenza even exists.

If someone throws a die twelve times and gets three sixes instead of the expected two, is it reasonable for them to suggest this is evidence that the die is rigged? Or is it more reasonable to think these results are very likely due to chance?

An old medical adage is, "when you hear hooves, think horses, not zebras", which is particularly appropriate given your pseudonym.

By Krebiozen (not verified) on 26 Jan 2015 #permalink

Krebiozen

By Colonel Tom (not verified) on 26 Jan 2015 #permalink

Krebiozen, I believe that between 6 and 8 passes they will switch the dice out on you. When you've turned $200 at the blackjack table into $35,000, they very politely request that you call it a night. They hand you some buffet coupons and a room voucher.

By Colonel Tom (not verified) on 26 Jan 2015 #permalink

#349

"So if there is any possible explanation for a phenomenon other than chance we automatically adopt it? That’s ridiculous."

No, that's a strawman.

Either your sincerity is in question or you are indeed confused.

I said we would not assume that it is the result of chance.

I said Greg's idea was reasonable, as a conjecture. I based that on the assumption that immune response varies among individuals, not on any numbers. But I see that you are still unwilling to take a position on that question. (And by a position, I don't mean another long filibuster with examples about everything but the issue at hand.)

I said Greg’s idea was reasonable, as a conjecture.

Maybe. It could be something genetic.
So was Krebiozen's conjecture. It could be chance alone, a look at the general population give us numbers which support it.

So what?

By Helianthus (not verified) on 26 Jan 2015 #permalink

@zebra

I said Greg’s idea was reasonable, as a conjecture. I based that on the assumption that immune response varies among individuals, not on any numbers.

In a void, sure, we can assume that Greg's idea is just as likely as chance. However, we are not operating in a vacuum. We have some preexisting knowledge about basic human physiology which makes one idea more likely than the other. With that preexisting information, we can conclude that the most likely explanation, barring a result or additional information well outside the parameters of that prior knowledge, we can rightly conclude that the most likely explanation is that Greg's experience is consistent (in the medical and/or generally understood meaning of the word, not Humpty's definition of the minute) with chance. The alternative requires greater levels of evidence to be assumed to be a valid explanation.

#353, 354

Mercifully short, but still filibuster.

"I based that on the assumption that immune response varies among individuals"

Very easy to understand. Does Krebiozen use that assumption in responding to Greg?

I based that on the assumption that immune response varies among individuals, not on any numbers. But I see that you are still unwilling to take a position on that question.

*Koff*

If you mean I am assuming that influenza strikes people at random, that is because this is what large epidemiological studies tell us. There are variations in how severe influenza is, and in the incidence of complications, but not, as far as I am aware, in who contracts the virus (aside from age as I discuss below). If you are not immune to an influenza virus and you are exposed to it you will very likely succumb (greater than 80% in this review).

I said Greg’s idea was reasonable, as a conjecture.

What, this? More than 100 comments after you chimed with this?

Gambler’s fallacy.

Also consider Bayesian analysis.

And sample size.

Is there someplace else that you put this item, or are you simply trying to retcon your performance?

A wonderful article that seems germane

http://jrs.sagepub.com/content/108/1/4.full

Strangely, I came across this gem while defending the possibility of the autism/circumcism linkage. That theory was also published in JRSM.

By Colonel Tom (not verified) on 26 Jan 2015 #permalink

Narad

In other shocking news, yes, Virginia, life expectancy decreases with age. Older people, “on average,” can “expect” to not have as much time left on the planet as younger people. It’s insane!

and totally unfair.
By Militant Agnostic (not verified) on 26 Jan 2015 #permalink

If I believed, as K maintains, that Greg is in the class of entities for which getting the flu is truly random, I would say:

“Since *nothing* can affect the outcome, it doesn’t matter how strong your immune system is.”

Same thing I would say to an unstable nucleus that had survived 10 times its half-life.

"Its half-life"? This sort of gibberish is why it's blindingly apparent that you have no understanding of that about which you're really, tediously pretending expertise.

The decay of an ensemble of such nuclides is Poissonian.* I nearly enshrined your entire comment in the blockquote as a monument to Assbaggery, but "Look. At. The. Data." needs to be fished off the cutting-room floor here.

(This may prove to be a markup catastrophe.)

The data are the count rates. The probability of observing k events with a half-life T is

P(k) = [(λt)ᵏ exp(−λt)]/k! ,

where λ ≡ T₀/(ln 2) with T₀ the half-life is the decay constant.

Now suppose that someone sells you what is supposed to be a 100% pure sample of unobtanium-42** and you assay a sample. In exquisite detail. Given N particles of ⁴²Un, consider the probability of observing k = N−1 decays in a time t = 10T₀.

Trivially, λt = 10 ln 2, so just to simplify the notation, with N−1 ≡ W (because Unicode is a disaster and I need another superscript), one has

P(W) = 2[(ln 2)ᵂ/W!].

If there is an error in that derivation, I'm sure someone will point it out, and I'm equally as certain that it's not going to be you.

In the meantime, you get yourself some graph paper or something and tell me exactly what a 10T₀ holdout means about whether you got a good deal on the unobtainium and how much of a sample you would need to test.

Now, if you are something other than a whiny drive-by troll, let me know if you understand, or explain why you disagree.

And that goes for Narad as well.

Well, see, I can't speak for gaist here, but I would offer to a plebiscite that you're well into FOADIAF territory with this one.

* Influenza statistics are not.
** Pause before complaining that it's even.

^ Aside from the unclosed italics affecting "particles of ... observing," it could have been worse.

^^ "in at a time t = 10T₀."

I based that on the assumption that immune response varies among individuals

Just hitting me. But is this circular reasoning?
The key word being "assumption".

1 - immune response varies among individuals
2 - Greg reports an unusual lack of flu in his life *
3 - assumption based on 1 and 2: Greg must have a overcloking immune system
4 - conclusion: Greg has a overcloking immune system

The circular part being between 3 and 4. Why does Greg didn't have flu? Because of his strong immune system. How do you know he has a strong immune system? Because he didn't catch the flu.

What I mean is, #3 is an assumption you can make as an hypothesis, but not one you can assume to be right.

*as Krebiozen and other has been saying again and again (e.g. #330) , a look at epidemiology tell us that this may not be that unusual. Something with a 30% chance of happening isn't generally a rare event.

By Helianthus (not verified) on 26 Jan 2015 #permalink

#361

Narad, I guess you don't want to answer 340 so you are taking the filibuster route?

I provided an example that I believe every reader can follow, and I would bet some percentage have figured out the right answer. Are you so insecure that you have to throw around numbers and terms to impress and confuse people so they don't question your position?

The only answer I can dig out of your ranting is that we should verify the identity of the nuclide-- is that it? Well, good, I may turn you into an experimentalist yet.

But then what?

To all:

As I said before, someone (gaist, actually) asked the proper question. How would zebra answer Greg?

My answers were (paraphrasing):

case 1 (If I believe that there is no variation in immune system, so that everyone's chance of getting the flu is equal):

"Greg, it doesn't matter if you have gone 20 years or 60 years."

Now, I think K has actually agreed with this ['random is random'], but you never know with K.

case 2 (I believe there is some variation in immune response-- for example, if the study referenced in #357 says that 20% exposed to the virus in controlled circumstances didn't get it, then there would be all these people with degrees of 'superimmunity' [note scare quotes].):

"Greg, we can't say. We would need to look at the original data, and narrow it down to more closely match your characteristics-- for example by age cohort-- and eliminate variables that might affect the outcome. Just knowing how many people get flu each year doesn't help."

So, again, this is a simple case that most people should be able to think about-- it isn't about anything but the reasoning in this specific instance.

It does require, as I said very early on, understanding what the question is.

zebra,

“So if there is any possible explanation for a phenomenon other than chance we automatically adopt it? That’s ridiculous.”
No, that’s a strawman.

So how else am I supposed to understand the following?

The thing is, if another explanation existed, then we wouldn’t think the outcome is the result of chance in the first place.

A gambler wins repeatedly on roulette and claims that he has telekinetic powers. This is another explanation - though a vanishingly unlikely one - does that mean we wouldn't think the outcome was the result of chance?

Either your sincerity is in question or you are indeed confused.

I'm certainly confused by whatever it is you think you are arguing.

I said we would not assume that it is the result of chance.

You wrote "think", not "assume", which I took to mean "believe". Maybe that's where our differences originate. I would assume that an outcome is the result of chance unless this was very unlikely, or there was good evidence for it being otherwise.

I would not assume that super-immunity, or anything else, is the explanation for someone not getting influenza in a particular year, I would assume that chance is the explanation unless I had good reason to think otherwise. This may not be the perfect way of approaching a problem, but it is the basis of a great deal of scientific investigation.

Incidentally, you mentioned the possibility of a Bayesian approach to this problem. Given a lack of any evidence for super-immunity to influenza, and evidence that more than 80% of healthy people contract influenza after a single exposure, I think it is fair to say that the prior probability of GY's 20-year influenza-free period being due to super-immunity is very low.

By Krebiozen (not verified) on 27 Jan 2015 #permalink

Zebra:

The thing is, if another explanation existed, then we wouldn’t think the outcome is the result of chance in the first place. And that’s why my answers to Greg are correct, and your answer is wrong.

I had a difficult time connecting to the Internet this morning. It's possible that, by chance, I was assigned the same IP address as another computer on my network. But I can also say that you placed a curse on me. Which is the more reasonable explanation?

Remember, you deliberately refused to communicate clearly. You have nobody to blame but yourself when people don't understand you.

By Gray Falcon (not verified) on 27 Jan 2015 #permalink

#368

I explained very clearly at 366. I've said the same thing in slightly different ways several times. That there is no response is more an indication that some people understand very well what I'm saying, but can't bring themselves to acknowledge the validity of my point.

K's response to Greg was wrong. My responses are correct. That's all there is-- we are talking about this very narrow question, or at least I am, since it makes you uncomfortable.

Narad, I guess you don’t want to answer 340 so you are taking the filibuster route?

Your desperate attachment to the word 'filibuster' is telling. This isn't a command performance, Peaches. I was responding to #253, as I clearly indicated.

I provided an example that I believe every reader can follow, and I would bet some percentage have figured out the right answer. Are you so insecure that you have to throw around numbers and terms to impress and confuse people so they don’t question your position?

One might note that explaining a Poisson distribution goes directly to what you're still failing to get through your head, viz.,

case 1 (If I believe that there is no variation in immune system, so that everyone’s chance of getting the flu is equal):

“Greg, it doesn’t matter if you have gone 20 years or 60 years.”

Now, I think K has actually agreed with this [‘random is random’], but you never know with K.

Are you so insecure that when you can't understand something you turn accusatory?

The only answer I can dig out of your ranting is that we should verify the identity of the nuclide– is that it?

No, it is that your remark that I quoted was phenomenally stupid.

Well, good, I may turn you into an experimentalist yet.

Given that your (non-) response to #298 was "Narad appears to be a master of woo physics," I suggest that you park your delusions of grandeur. Effectively every comment you've made about physics has been the equivalent of dumping a pail of garbage over your head.

It's odd that your "experimentalist" credentials didn't allow you to figure out the answer to your own question:

How long would Greg have to go without the flu in order for his case to be remarkable?

Didn't understand what 5 σ means? Tried and failed? Truly impressive, given that the answer is basically in #101, but then again, you're a painfully obvious poseur. Then again, it might as well be tuned up:

Bernoulli trials aren't normally distributed, so let's relax the usual Gaussian 3 σ requirement of 99.7% to 98% for capture of a noteworthy outlier. Hell, let's even ignore the fact that that's two-tailed, since you seem to consider actually working things out to be "filibustering."

Me makee easy for zebra, show zebra picture.

Hey, remember this one, Perfesser?

I’ve come to realize that I’ve discovered a population here that has the characteristics of ... rejecting certain aspects of modern physics, and demonstrating a belief in ‘physics woo’– like “hidden variables”.

Now, that's a fυcking knee-slapper in this context (and no, there is no chance that I'm going to refrain from letting you fail to figure it out for yourself).

That there is no response is more an indication that some people understand very well what I’m saying, but can’t bring themselves to acknowledge the validity of my point.

K’s response to Greg was wrong. My responses are correct. That’s all there is– we are talking about this very narrow question, or at least I am, since it makes you uncomfortable.

Oh, sweet Jesus, you're desperate.

That there is no response is more an indication that some people understand very well what I’m saying, but can’t bring themselves to acknowledge the validity of my point.

I can think of several other explanations for this supposed 'silence'. Why do you think your version is the most probable? Maybe it's random.

Also, your insistence on narrowing the discussion into one facet contradicts several of your own posts, and seems only a gish galloping exercise trying to avoid having to answer the multitude of questions waiting and to allow yourself to ignore reasonable counter arguments.

And for
your two answers - I could quote your own comments to argue what you now claim was your original. And besides, how come it took you so many posts to actually figure you ought to perhaps explain yourself rather than whine mock-cryptically and insist on your definitions of words above what the rest of us understand them to mean?

But whatever, now that you've silenced is all with your mighty intellect, how about answering those questions?

Pistols
20 paces
Dawn.

By Colonel Tom (not verified) on 27 Jan 2015 #permalink

zebra,

I explained very clearly at 366. I’ve said the same thing in slightly different ways several times.

Your explanation at #366 is not an explanation at all, it just states what you would tell Greg depending on your pre-existing beliefs. Instead of trying to find a way of answering the question, you have merely restated it. If you believed it was random chance you would tell him it was random chance, and if you believed it might not be random chance you would tell him it might not be random chance, but you would need more information to be certain.

You offer no method of assessing which answer might be right, given the available evidence, just that the question cannot be answered definitively without further data, which is blindingly obvious. That really isn't very helpful.

To resort to yet another example you will probably fail to understand, say there are 30 people in a room and they discover that two of them share a birthday. One of them, Fred, suggests that this might mean something remarkable is going on. Would you tell him the following?

Case 1 (If you believe that there is nothing strange going on):
“Fred, it doesn’t matter if two of you or all of you share a birthday.”

case 2 (You believe that something other than chance is at play)
“Fred, we can’t say. We would need to question each of the people here to find out what led to them being here, and eliminate variables that might affect the outcome. Just knowing how many people share a birthday doesn’t help.”

I wouldn't, I would tell Fred that the probability of two people in 30 sharing a birthday by chance is about 70%, making chance the most likely explanation.

That there is no response is more an indication that some people understand very well what I’m saying, but can’t bring themselves to acknowledge the validity of my point.

No one agreeing with you is evidence that some people agree with you? I don't think so. I think most people following this will have realized by this point that you don't understand probability or hypothesis testing, but for some reason you are unable to admit this.

K’s response to Greg was wrong. My responses are correct.

I just love it when people declare victory by fiat. There should be an internet Law about this.

That’s all there is– we are talking about this very narrow question, or at least I am, since it makes you uncomfortable.

Really? You demonstrate, in rather spectacular fashion, that you do not understand a number of basic scientific concepts and then declare victory, claiming that no one agrees with you because it makes them feel uncomfortable? I can almost hear you petulantly stamping your feet, "I'm right and you're wrong, and that's all there is!" Ever hear of the Dunning-Kruger effect zebra?

By Krebiozen (not verified) on 28 Jan 2015 #permalink

Zebra - remind us what ipse dixit means in the context of "My responses are correct."

By Mephistopheles… (not verified) on 28 Jan 2015 #permalink

#374

"You offer no method of assessing which answer might be right, given the available evidence, just that the question cannot be answered definitively without further data, which is blindingly obvious"

And yet you answered it?

zebra: I was late for work today because of traffic, and the elevator at my office was not working. Does this or does this not support my hypothesis that you placed a curse on me? And I need a real answer, not you acting as if it were obvious.

By Gray Falcon (not verified) on 28 Jan 2015 #permalink

And yet you answered it?

I pointed out that Greg's experience is consistent with random chance and that further speculation about putative causes is unnecessary. That's because I understand it is possible to say more about a proposition than that it is true or false. You, evidently, do not.

By Krebiozen (not verified) on 28 Jan 2015 #permalink

I was going to string K along some more but I just decided this is getting boring again, and certainly with all the hostility there is zero probability that someone will follow up on the hint. So I will give the answer. The answer is: What's the question?

Before you reflexively pull out your calculator to find the p-value, science demands that you be able to state the *hypothesis* you are *testing*. So far, unless I missed it in all the comments, not one person has addressed this.

K's error is this: He is testing the hypothesis...

"Greg has super-immunity"

...against the general population.

But the only way we can test that hypothesis is against *the part of the population that has super-immunity.*

It's right there in my two answers (which I *did* give in response to gaist way back in the thread,) if you have an open mind.

But, instead of having a scientific inquiring mind, we have Narad shouting vaxer! vaxer! when I suggested that there was variability in immune response. I can't say, scientifically, whether that is exactly the case, but it is not *unreasonable* to think it. And, it is *necessary* to allow it to properly set up the analysis.

If there is no such thing as super-immunity, case 1.

If there is such a thing, K's answer is still wrong, because the p-value he calculates is *determined* in part by the people with super-immunity. Like Greg, maybe. Maybe (and I also said something like this way back) super-immunity manifests itself as getting flu every 20 years, and some other factor balances it out in the general population.

So, do I think Greg has SI? Nope. But I would not tell him that his 20 years flu-less is irrelevant because of its probability in the general population.

Narad,

The answer is: If we determine that some (correctly identified nuclide) individuals are surviving an interval inconsistent (by your standard) with the established half-life, we......recalculate the half-life.

What would you do, flush the Standard Model? Postulate some chi-like hidden variable?

A wise man once said: "The value of the half-life is irrelevant to the experience of the individual."

If we determine that some individuals are surviving an interval inconsistent with the established half-life, we……recalculate the half-life.

No.
Any population has outliers. You don't go around changing all the numbers because you found one or two individuals outside of the expected parameters.
If you are really concerned, you do a study on a large subset of the population to confirm or correct your numbers.

K’s error is this: He is testing the hypothesis…
“Greg has super-immunity”

No.
He was testing the hypothesis "Is Greg's experience unexpected?" and he was right to do it against the general population.

In science, it's the null hypothesis. Before we go all worked-up, is there any reason to believe that what we are seeing is out of the ordinary, instead of due to chance alone?

"Does Greg has a super immunity?" is an hypothesis to explore if the first hypothesis is rejected, i.e. if Greg was found to not be part of the normal population (no offense intended, Greg - I mean "normal" as in "ordinary")

Actually, your two last posts contradict one another. In one, you advocate that when faced with some odd event, we should not make a comparison with the whole population. In the last one, you wrote that, when faced with an odd event, one should redo the calculation so that it's absorbed by the whole population.

Why can I not propose the hypothesis that this long-living nucleid has a super-immunity to decay? Maybe one of its neutrons has a special flavor of boson.

By Helianthus (not verified) on 28 Jan 2015 #permalink

#381

At least you are trying to reason.

"Why can I not propose the hypothesis that this long-living nucleid has a super-immunity to decay?"

Because we have an established Theory (note capitalization,) and Parsimony operates at the level of Theory (not Hypothesis).

"“Does Greg has a super immunity?” is an hypothesis to explore if the first hypothesis is rejected, i.e. if Greg was found to not be part of the normal population (no offense intended, Greg – I mean “normal” as in “ordinary”)"

Here's what I said: "Maybe (and I also said something like this way back) super-immunity manifests itself as getting flu every 20 years, and some other factor balances it out in the general population."

So, again, I appreciate that you are trying to reason, but you are not correct that Greg's experience tells us that he is 'normal', and therefore not a member of the SI class.

What we need is a p-value test with respect to the experience of the SI population.

It isn't a p-value test to say Greg's experience is 'unusual' or 'not unusual'. It's calculating the value; there is no 'test'.

#382, to be 'clear'

"It isn’t a p-value test to say Greg’s experience is ‘unusual’ or ‘not unusual’ [with respect to the general population]. It’s calculating the value; there is no ‘test’.

Here’s what I said: “Maybe (and I also said something like this way back) super-immunity manifests itself as getting flu every 20 years, and some other factor balances it out in the general population.”

It's possible, but is there anything in the evidence that prompts us to consider this as a possibility? And wouldn't large, population-level studies that show that on average people go about 20 years without getting the flu show those outliers who might have super-immunity (and, conversely, those who got flu more frequently)? If the distribution were such that 20 years flu-free were on the extreme end of the curve, maybe we would consider it possible that he has super-immunity. But if he fits well within the distribution of "normal", there's nothing to make us suspect, and therefore use, super-immunity as an explanation. We don't need to.

The answer is: If we determine that some (correctly identified nuclide) individuals are surviving an interval inconsistent (by your standard) with the established half-life, we……recalculate the half-life.

What would you do, flush the Standard Model? Postulate some chi-like hidden variable?

This is quite possibly the most shockingly stupid thing that you've said yet. The Standard Model? Go learn what a G-ddamned Poisson process is.

But, instead of having a scientific inquiring mind, we have Narad shouting vaxer! vaxer! when I suggested that there was variability in immune response.

WTF are you talking about? You still don't understand the Maher reference? Let me remind you of the comments that made clear the comparison:

People get sick due to a combination of discrete causes and conditions that we know about.

(Note that what they are was never stated.)

Human immune systems and their response to challenge are not identical across individuals.

These are nothing other than the assertion that some people, like Bill Maher, are less vulnerable to the flu.

Before you reflexively pull out your calculator to find the p-value, science demands that you be able to state the *hypothesis* you are *testing*.

You didn't look at the picture, did you? Or didn't you understand it? There's no particular reason for an alternate hypothesis until he becomes an outlier, that is after 96 years without the flu.

K’s error is this: He is testing the hypothesis…

“Greg has super-immunity”

…against the general population.

But the only way we can test that hypothesis is against *the part of the population that has super-immunity.*

I was clearly mistaken in #385. This takes the fυcking cake.

^ You're free to try to weaken it with Chebyshev's theorem, though, assuming that (1) you can justify applying it, (2) specify what you're capturing, and (3) figure out what this means.

#384

Good question. I guess you haven't noticed that I've asked multiple times if anyone can provide.........actual data. (Maybe even in response to you.) If we could look at a distribution we might be able to draw some conclusion, or not.

But, calculating the distribution by *assuming randomness* is not actual data. It's case 1.

Narad,

The only thing I knew about Maher was that he was some celebrity. Now I know that he believes he is immune to flu (?).

Why don't you cowboy up and tell us-- unequivocally-- if you think that there is no such thing as SI, and we can work from there.

Why don’t you cowboy up and tell us– unequivocally– if you think that there is no such thing as SI, and we can work from there.

"We"? I have no reason to suppose that there is, and you've offered nothing even vaguely responding to #201 or #356, and I – like everyone else – have wasted far more time than you deserve pointing out where a threshold for speculating that something other than routine counting statistics were in play. It has nothing to do with "belief."

Instead, you've treated your audiences to such idiocies as this:

A wise man once said: “The value of the half-life is irrelevant to the experience of the individual.”

There's nothing going on between your ears to "work" with.

If we could look at a distribution we might be able to draw some conclusion, or not.

Gee, I guess it's too fυcking bad that you've been presented with not just one, but two, not to mention #304. Oh, but that's "filibustering" in the Land of the Self-Proclaimed Wise Man.

I should note that using the wrong statistical distribution is an excellent way to skew results in the way you want them to be skewed.

By Gray Falcon (not verified) on 28 Jan 2015 #permalink

Narad

If your working Theory is that there is no such thing as SI, then we have case 1. If Greg lived to 96 (or even 110) never having had the flu, Parsimony would suggest we look for an explanation that comports with the entities already present in your Theory-- perhaps Greg is an OCD compulsive hand-washer, or a bubble boy.

We could only do a p-value test with respect to SI *if we had some freakin SI people with which to compare him*!! They don't exist in your Theory!!

How is that 'stupid'??

We could only do a p-value test with respect to SI *if we had some freakin SI people with which to compare him*!! They don’t exist in your Theory!!

How is that ‘stupid’??

1. I don't have a "Theory."

2. There is no such thing as "a p-value test." I strongly suspect that you don't even know what p-values are supposed to constrain.

3. Your insistence that "the only way we can test that hypothesis is against *the part of the population that has super-immunity" is jaw-droppingly ignorant. It is utterly senseless, and you have made not the slightest effort to justify it.

We could only do a p-value test with respect to SI *if we had some freakin SI people with which to compare him*!! They don’t exist in your Theory!!

Null hypothesis, dude (or lass). Seriously, read about it.
It's perfectly possible to compare an individual to a whole population and figure out it is not like the others.

Saying that someone is special is not an hypothesis. Since you don't define your "superimmune" people (no, saying they may exist is not defining them), there is no way to compare Greg to them. They may not exist in "our" theory, but they don't exist in yours either.

Imagining the existence of special snowflakes who only get sick every 20 years is not being scientific or open-minded. It's just mangling the data so it fits your preconceptions.

Science is not about imagining a fancy story about how the world works. It's about imagining a fancy story and then doing one's best to punch holes through it.
If your story survives, you may be onto something. If not, put it on a shelf and search something else to be interested at.

By Helianthus (not verified) on 28 Jan 2015 #permalink

Zebra, I have a super-duper car I have invented that outperforms all other cars in the world. How do I know? Because it has five wheels and made from ferrocement , and therefore it can only be compared to other five wheeled ferrocement cars and therefore it is probably but unfalsiably superior in relation to cars arbitrarily similar to itself, because otherwise the universe would be so random it might seize to exist at any moment. And if you disagree it is just your inferior and limited Theory that blinds you.

Is this really your argument?

And in all those posts, you couldn't have answered any one of the many questions you keep actively evaded? Your silence on the issue speaks volumes.

G*d damn cellphone and not having had coffee yet. Previous post written by gaist, not zebra...

On the hope it gets deleted, here it is again.

Zebra, I have a super-duper car I have invented that outperforms all other cars in the world. How do I know? Because it has five wheels and made from ferrocement , and therefore it can only be compared to other five wheeled ferrocement cars and therefore it is probably but unfalsiably superior in relation to cars arbitrarily similar to itself, because otherwise the universe would be so random it might seize to exist at any moment. And if you disagree it is just your inferior and limited Theory that blinds you.

Is this really your argument?

And in all those posts, you couldn’t have answered any one of the many questions you keep actively evaded? Your silence on the issue speaks volumes.

I suppose I skipped this:

If Greg lived to 96 (or even 110) never having had the flu, Parsimony would suggest we look for an explanation that comports with the entities [sic] already present in your Theory [sic]– perhaps Greg is an OCD compulsive hand-washer, or a bubble boy.

You don't know what the fυck 'parsimony' means, either, so you might as well skip the personification.

More importantly, you don't understand a G-ddamned thing about statistical inference, despite having the unmitigated gall to have been dripping with condescension from the moment you came knocking.

I told you what what it would take for someone to represent – roughly speaking – a "3 σ outlier." I didn't say that a single such outlier would be in the slightest bit interesting.

Do you know how many times I've had the flu sans vaccination? Three. Do you know what their intervals were, as best I can reckon? Nineteen and 20 years.

Allow me to remind you of Greg Young's original comment:

I wonder what the odds are on a person who has never had a flu shot going two decades without contracting the more recent strains of flu?

Answer: That's pretty fυcking typical.

Yet, for nine days just here, having oozed out from this debacle, you have been demonstrating nothing other than that the more effort invested in replying to you, the greater the jackass you become.

#396

Sorry, that's gibberish.

SI people do exist in my case 2. Did you miss that?

#395

"I don't have a Theory"

Sure you do-- in your Theory, there's no such thing as SI, so everyone has an equal chance of getting the flu with respect to immunity.

That's case 1.

In case 2, the Theory would say there *are* SI people. And so you couldn't calculate the probabilities assuming randomness with respect to immunity using the entire population.

#398

Gaist, that's a terrible analogy. Not even close.

If anything, I would be the one saying: You can't make *any* claim about the car, good or bad, unless you have others to compare it to. But it is too much of a stretch to try to fix it (the analogy).

zebra,

Remember the person in my example who finds that two people in a room of 30 share a birthday? What would you tell him when he asks you if this is evidence that something unusual is going on?

By Krebiozen (not verified) on 29 Jan 2015 #permalink

Narad,

I had a chance to read your comments more closely, and I see that you believe "personal anecdote" = "data". But that's OK, because you are helping make my point.

Super-immunity aside, had K told Greg that his experience was unremarkable because we had asked a representative sample of the population how often they had had flu, and we were confident because of the sample size that the numbers applied to the population as a whole, that would at least start K off on the correct foot.

But that's not what he did-- he said "since 2-5% of the population gets flu every year..."

And wait-- if you say that going 96 or 110 years would be unremarkable, aren't you agreeing with me? (Rustle of pig's wings.)

Krebiozen,

" he asks you if this is evidence that something unusual is going on?"

K, "something unusual" is the same as "low probability".

There's no hypothesis being tested here. He's just asking what the probability of the outcome is.

We're back to this conceptual trap-- you seem to be confusing

1) "the probability of y happening at random *rather than* because of x"

with

2) "the probability of y happening at all".

We use an arbitrary number (e.g. 5%) to say that in (1), if the chance probability is < or = to 5%, then the probability is (in effect) zero that it is chance *rather than* x.

But you don't have a "rather than" in your question, right? So you just have my "case 1".

zebra,

Super-immunity aside, had K told Greg that his experience was unremarkable because we had asked a representative sample of the population how often they had had flu, and we were confident because of the sample size that the numbers applied to the population as a whole, that would at least start K off on the correct foot.
But that’s not what he did– he said “since 2-5% of the population gets flu every year…”

I did offer to dig out the clinical trial data I was basing that on at #138 above, but you didn't take me up on my offer. For example, this Cochrane review of controlled randomized clinical trials that included over 70,000 people:

15.6% of unvaccinated participants versus 9.9% of vaccinated participants developed ILI symptoms, whilst only 2.4% and 1.1%, respectively, developed laboratory-confirmed influenza.

That's 2.4% of unvaccinated individuals who got laboratory-confirmed influenza in these studies. Is over 70,000 a large enough sample size for you?

By Krebiozen (not verified) on 29 Jan 2015 #permalink

zebra,

” he asks you if this is evidence that something unusual is going on?”
K, “something unusual” is the same as “low probability”.

I struggle to see any substantive difference between this and Greg's suggestion that not getting influenza for 20 years suggests an explanation other than chance. In both cases we are asking if a particular observation is unusual enough for us to construct a hypothesis other than random chance to explain it. In both cases the answer is "no".

There’s no hypothesis being tested here. He’s just asking what the probability of the outcome is.

His hypothesis is that there is something other than chance at play, just as Greg's is, based on an experience that is consistent with random chance, just as Greg's is.

We’re back to this conceptual trap– you seem to be confusing
1) “the probability of y happening at random *rather than* because of x”
with
2) “the probability of y happening at all”.

This does seem to be where your confusion arises. When we have no reason to believe x even exists, and the individual's experience is exactly what we would expect from chance, why construct an alternate hypothesis at all?

We use an arbitrary number (e.g. 5%) to say that in (1), if the chance probability is < or = to 5%, then the probability is (in effect) zero that it is chance *rather than* x.

I'm glad to see you have learned something from this exchange, but you seem to have forgotten that p in Greg's case is at least 0.35, so we can comfortably say that his experience is consistent with chance, and an alternative explanation "x" is not required. That really is "all there is", apart from what looks to me very like desperate special pleading.

But you don't have a "rather than" in your question, right? So you just have my "case 1".

There is a "rather than" in my question, in the form of "something unusual". We can make it something specific if you prefer, such as it being a club of people who belong to the same sign of the zodiac, or some sinister person behind the scenes who has the unusual hobby of manipulating things to get people with the same birthday together. It doesn't really matter what the cause other than random chance is.

Your "case 1" takes as one of its premises that there is no possible explanation other than chance:

If I believe that there is no variation in immune system, so that everyone’s chance of getting the flu is equal

When one of the premises of an argument includes its conclusion it is circular reasoning. It is of no use in this problem, it's a red herring, as it has been all along.

Your "case 2" takes as a premise that there is variation in immune systems:

I believe there is some variation in immune response

Since the problem we are addressing is whether Greg's experience is unusual enough to warrant positing "some variation in immune system", or not, this is also a form of circular reasoning, since your conclusion is also one of the premises of your argument.

Look at it this way. We assume that the 2-5% of people who contract influenza in any given year are selected randomly, for all practical purposes. This is a reasonable assumption given the epidemiological evidence we have about this illness.

Let's also posit a group of people who have an increased immunity to influenza. We don't know how many of them there are, or anything else about them, but we can deduce that fewer than 2-5% of them will contract influenza in any given year, this being an expected consequence of increased immunity to influenza.

The question is whether Greg's experience suggests that he belongs to the former group, or if his experience is unusual enough for us to posit the existence of the latter hypothetical group and his membership of it.

Very clearly (p>0.35) random chance is a perfectly adequate explanation and we do not have to invent "super-immunity" as there is nothing unusual to explain, any more than the two people sharing a birthday is unusual.

By Krebiozen (not verified) on 29 Jan 2015 #permalink

Can we now discuss this 2 month SUPER cough?? WTH!!! Into the doctor with 2 of my kids 3 times in a 2 mth period with a cough that leaves you literally breathless in between... it's ridiculous. They sent my kids home with inhalors and called it bronchitis... I ended up getting this same cough finally. I have gotten bronchitis since I was 12 yrs old every year of my life and this was NOT bronchitis.
Really, these doctors need to start re-visiting the old whooping cough and other things of the past because something is up, and with all of the brushing off of this stuff, everyone is getting sick because it just keeps going and going and going.

chays:

Part of what's up is that the current acellular whooping cough vaccine is less effective than the previous one; this is a tradeoff for fewer side effects. Adults often don't get the vaccine booster that they ought, or don't get it often enough: it's bundled with the tetanus vaccine, and many people will let that slide until/unless they have a puncture wound.

(I saw speculation a while back, don't think it was on this blog, that the tendency to prescribe antibiotics for viral infections had been limiting the spread of pertussis, by killing the bacteria in people who hadn't had symptoms.)

I am not a doctor, but in your shoes I would be pressing to have myself and my children tested for pertussis: http://www.cdc.gov/pertussis/clinical/diagnostic-testing/diagnosis-conf…

And wait– if I say that you say that going 96 or 110 years would be unremarkable, aren’t you agreeing with me?

Does that help?

(Rustle of pig’s wings.)

You're in bad enough shape already to make weather updates from your pants are an unwise tactical maneuver.

Go back to what you're "replying" to, pay special attention to the paragraph immediately following "personification," and then try to frame a response that suggests you have learned a single thing. It may help if you review the record while also attending to the words "discovery threshold," "process," and "relax."

^ "are an"

#407

Someone criticized me earlier for using the term filibuster, but I can't think of anything else to describe what you are doing here.

At 403, you proposed a nice simple case, which I assume you chose because it would demonstrate how my reasoning is incorrect. As soon as I answered, though, you were off and running with this long screed about Greg's case. If that's not filibustering and obfuscation, I don't know what is.

I am willing to discuss the case you presented-- remember, *you* chose it; we are playing on *your* pitch. However, I am not going to be distracted trying to answer the errors you present about the Greg case before settling this one. If you are abandoning 403, I can only assume that you see it is a lost cause. You should admit that.

So, the only scientific way to state the hypothesis of 403 is:

a) "The group of 30 was not chosen at random."

This has the antithesis:

b) "The group was chosen at random"

We accept b if your statistical test falsifies a.

Now be my guest-- show how calculating the probability of 2 identical birthdays (IB) in a group of 30 falsifies a.

Then we can discuss your errors with respect to Greg.

I think someone needs to learn what the term "null hypothesis" means and how it applies to a) and b).

#413

*yawn*

You presume to teach the master of using cryptic comments to draw people into debate how to suck eggs?

sad...

Oh, this is just too tempting.

So, zebra, are you saying you're a master baiter?

zebra,
I was tending toward the idea that you are deliberately being obtuse, but you really still don't get it do you? This isn't about definitively proving what is true or false, or falsifying a hypothesis. It is about assessing whether we require a hypothesis other than random chance to explain an observation.

If scientists looked for alternative hypotheses to explain every observation that was consistent with random chance they would get nowhere.

By Krebiozen (not verified) on 30 Jan 2015 #permalink

How is null hypothesis a cryptic reference when it is both basic statistical concept and explained at least three times already in these comments?

So, the only scientific way to state the hypothesis of 403 is:

a) “The group of 30 was not chosen at random.”

This has the antithesis:

b) “The group was chosen at random”

We accept b if your statistical test falsifies a.

Holy fυck, the Bunker of Utter Cluelessness truly is impenetrable.

This isn’t about definitively proving what is true or false, or falsifying a hypothesis.

Yah, but H1 is terrific coming from Mr. "Also consider Bayesian analysis": Someone has secretly chosen everyone in the room.

You call that mastery, or "drawing people into debate"?

For example, this Cochrane review of controlled randomized clinical trials that included over 70,000 people

I never quite finished a comment on this that I started at the time, but given that zebra's inverse S/N promises to keep the meters pegged until their coils burn out....

It's not clear to me that that review was geared so as to allow meaningful population-wide estimates of that spread. I can't see the full text of this, but it suggests that clinical diagnosis ain't half-bad.

Anyway, I ran across that while looking for influenza recurrence data, specifically, here (I'm pretty sure their Reichert et al. reference 49 should be to PMID 11259722).

WTF,* France? This is from the best-fit model, but still:

"The cumulative average number of influenza infections was 2·1 (95% CI 1·4–3·2), 3·9 (95% CI 2·6–6·0), 6·4 (95% CI 4·3–10·5) and 8·7 (95% CI 5·8–14·9) in individuals of ages 10, 20, 40 and 65 years respectively. The first episode of influenza occurs at a mean age of 3·3 years (95% CI 1·9–5·4) and the second at a mean age of 7·9 years (95% CI 5·0–12·6). At age 5 years, 28% (95% CI 12–46) of children remain free of influenza since birth."

* The same goes for Reichert's affiliations, but I'll defer to NEJM here.

As much as I love biostatistics, there comes a point where they reach their limits. There is nothing that can prove that a person has never gotten the flu has or does not have a magical immunity, all that statistic prove, (AND THEY DO) is that in a large population size that someone that does not get the flu would be expected by random chance.

This fact should be acknowledge by all.

Now, we're at the point to either do a far larger sampling or perform experimental validation. Since it is a simple validation and no more dangerous than the cold water survival gear we tested for pilot rescue, we puff some virus up noses.

We also can agree that some people reach the limit of their intellectual and emotional capacity and further debate becomes,,,,,,,,,,,,,

By Colonel Tom (not verified) on 30 Jan 2015 #permalink

There is nothing that can prove that a person has never gotten the flu....

Ah, but one can disprove it, so "zebra wins," as Nature has decreed a constant of the motion.

Well, then I am going to mention that I want to go back to first principles. After reading the CDC's webpages, I felt my observation validated that not everyone that gets the flu even knows they have the flu. Therefore, I'm proposing that there is a third equality valid hypotheses, at least as valid as the kryptonian immunity hypotheses, that some people got the flu and were just really unobservant.

By Colonel Tom (not verified) on 30 Jan 2015 #permalink

o.O
I last looked at this maybe 250 comments ago. I cannot help thinking of what my friend calls a "wrong onion." You peel off the outer later of wrongness, and instead of finding something right, there are just more nested layers of wrong lurking below. It's 25 years since I took logic and I dropped the math major before getting to statistics, but even I can spot the zebra-onion.

By Emma Crew (not verified) on 30 Jan 2015 #permalink

Krebiozen:

"This isn’t about definitively proving what is true or false, or falsifying a hypothesis. It is about assessing whether we require a hypothesis other than random chance to explain an observation.

If scientists looked for alternative hypotheses to explain every observation that was consistent with random chance they would get nowhere."

Krebiozen, meet Victor:

"Certain climate scientists have proposed that increased CO2 emissions are causing the planet to heat up in a drastic and dangerous manner. As with the proponents of any theory, it is up to them to provide evidence to back it up. Skeptics have no such burden, since they are proposing no theory, only questioning one.

Since, as we know, the Earth has both heated and cooled many times in the past, sometimes by drastic amounts, there is really no need to come up with a theory to explain why the atmosphere seems to have warmed considerably over a 20 year period in the recent past (1979 through 1998)."

(Comment 204 on the most recent post on RealClimate blog.)

I didn't bother to find quotes, but I can attest that Victor and others like him also invoke "The Null Hypothesis", and tell us, as Krebiozen does, that it is "Parsimonious" to accept that explanation (chance) rather than a cause.

Maybe you guys should hook up?

On the 'null hypothesis', a couple of exercises:

1) J, a comely young student, is standing on the steps of the math building after her stats 101 class. K, a slightly older sociology student, who has a crush on J, comes up and says:

"Listen, if you have any problems with statistics, why don't you come over to my place and I can tutor you. I'm pretty good; let me demonstrate. I predict that if I ask any 30 students walking across the quad, two of them will have identical birthdates. Wait here"

K takes a clipboard, stops 30 students, and writes something down for each one. He comes back and shows the list to J, and sure enough there are two entries with IB.

But J says:

"Look, my momma told me I could date cowboys and truckers and bikers and such, but I should never trust a social scientist-- they are always misleading people in their experiments. You could just as easily have written down the second IB, so why should I believe you?"

So: What is the null hypothesis, and what is the p-value?

2) Back in the beginning of physics, the very first experiment with a pendulum is this:

a) Measure the period when the length is L.
b) Measure the period when the length is 2L.

What is the null hypothesis? What is the p-value?

zebra,
You quoted an AGW denialist:

"Since, as we know, the Earth has both heated and cooled many times in the past, sometimes by drastic amounts, there is really no need to come up with a theory to explain why the atmosphere seems to have warmed considerably over a 20 year period in the recent past (1979 through 1998).”

As NASA points out, "The globe is warming at a faster rate than it ever has before" (my emphasis). Random chance is not a plausible explanation for the unprecedented warming we have seen since in recent decades. That is especially true since mathematical models of increasing CO2 levels due to human activities leading to global warming are consistent with our observations, and models of random chance are not.

I didn’t bother to find quotes, but I can attest that Victor and others like him also invoke “The Null Hypothesis”, and tell us, as Krebiozen does, that it is “Parsimonious” to accept that explanation (chance) rather than a cause.

So when one person wrongly suggests random chance as an explanation for something that the vast majority of climate scientists tell us is not consistent with chance, it means I am wrong for suggesting random chance is the likely explanation for something that indisputably is consistent with chance? We should abandon the concepts of the null hypothesis and parsimony because they have been used to argue for nonsense?

You are even more confused than I thought.

By Krebiozen (not verified) on 01 Feb 2015 #permalink

zebra, despite your cute little backstories and soap opera plot twists, the fact remains that if you pick 23 people without any regard for their birthdays, there is a 50% chance that two of them share the same birthday. 30 people, and it's 70% chance. 70 people and it's 99,9% chance. No ulterior motives or shenanigans required.

Why is this so hard a concept to grasp?

And why not - instead of trying your hand at witty script writing - just answer the multitude of questions you've so far kept avoiding?

Krebiozen - Thank you for all your patient explanations. They've been very helpful.
Can you recommend any sources that would be especially good for beginners to this kind of probability & statistics? I clearly have a lot to learn.

2) Back in the beginning of physics, the very first experiment with a pendulum is this

It is so cute when you demonstrate that you've never had even the most rudimentary exposure to a physics curriculum and, most importantly, can't conceive of what the lab component is for.

Mel,

Krebiozen – Thank you for all your patient explanations. They’ve been very helpful.

I'm happy to be of assistance. I enjoy playing with this stuff, as you probably gathered.

Can you recommend any sources that would be especially good for beginners to this kind of probability & statistics? I clearly have a lot to learn.

It's been a few years since I last taught this material, and that was in the context of clinical biochemistry; reference ranges, mutivariate analysis and that sort of thing. I have heard good things about this book, which teaches statistics in a humorous way. I know that many people find statistics intimidating, so anything that helps put people at their ease must be a good thing. There are also some excellent resources on-line, such as this website.

By Krebiozen (not verified) on 02 Feb 2015 #permalink

zebra,

On the ‘null hypothesis’, a couple of exercises:

Oh great, I love it when you play professor.

1) J, a comely young student, is standing on the steps of the math building after her stats 101 class. K, a slightly older sociology student, who has a crush on J, comes up and says:

What possible relevance does this weirdly detailed and elaborate scenario have to... well, anything at all? Is the use of "K" significant? Are you implying I'm a sociology student? I haven't been a student for almost 30 years and have never studied sociology. Since you made allegations of a lack of a comprehensive education and reasoning skills on another thread, I might add that I was taught mathematics and probability by a man who was at Princeton with Einstein, and who worked with Alan Turing at Bletchley Park. He seemed to know what he was talking about.

[Essentially the same problem I presented earlier of what the odds are of two people in a group of 30 sharing a birthday snipped.]

But J says:
“Look, my momma told me I could date cowboys and truckers and bikers and such, but I should never trust a social scientist– they are always misleading people in their experiments. You could just as easily have written down the second IB, so why should I believe you?”

Sometimes our data may be unreliable, or may be deliberately manipulated to mislead us. What do cowboys, truckers, bikers or social scientists have to do with this? Or J's mother's advice? Why not throw in a violin-playing kangaroo and a passing Catholic priest on a unicycle for good measure?

So: What is the null hypothesis, and what is the p-value?

The null hypothesis, such as it is, is that K recorded the birthdays accurately and that the students walking across the quad were selected randomly. In this case p is still about 0.7 for two birthdays being the same. However, in an experiment where we have good reason to suspect the experimenter of serious bias, does it really matter? I suppose a Bayesian approach would take K's motivation to impress J into account, increasing the prior probability of K faking the data. The odds of him getting the results he claimed by chance remain the same.

Coming up with bizarre examples where the concept of a null hypothesis and p values are of limited use does not demonstrate they are of no use in other situations.

By Krebiozen (not verified) on 02 Feb 2015 #permalink

zebra,
Moving on to your second 'exercise':

2) Back in the beginning of physics, the very first experiment with a pendulum is this:
a) Measure the period when the length is L.
b) Measure the period when the length is 2L.
What is the null hypothesis? What is the p-value?

That kind of statistical analysis is not appropriate for this kind of experiment. We are not attempting to figure out whether our results are likely to be the result of chance or of something else.

You could, I suppose, argue that the null hypothesis is that there is no relationship between L and the pendulum's period, make multiple measurements of the period with lengths L and 2L, and calculate a standard deviation from the mean in each case. Assuming the amplitude is kept low enough I would expect p to be very low indeed in this scenario, as all you are doing is estimating the errors in the system. Chance plays no part in this experiment, apart from variations in measurements and in the experimental setup.

By Krebiozen (not verified) on 02 Feb 2015 #permalink

There's a dead link in #433 but it is of no importance.

By Krebiozen (not verified) on 02 Feb 2015 #permalink

Assuming the amplitude is kept low enough I would expect p to be very low indeed in this scenario, as all you are doing is estimating the errors in the system.

Not that p-values by themselves have much of a role in physics in the first place. Perhaps zebra could try to make sense of this. (D'Agostini looks to have quite a bit of good stuff over here.)

#436

Ummm.... D'Agostini is pretty much saying what I'm saying. I was even going to reference the Wikipedia article myself, but I know some people question Wiki's validity.

I have to do some real work so I will offer my little intro lecture for any hypothetical unbiased readers who are just beginning to learn this stuff, and go further when I can get back.

Exercise 2: A likely source of confusion for K and others (evidenced by his answers) is the distinction between

a) "The null hypothesis for this experiment", and

B) " *The* Null Hypothesis".

When we design an experiment, we require a working hypothesis to justify what we are doing. By convention, we state the hypothesis as a negative, although, as you will see, the probative value of doing that is limited. So, keeping in mind that the experimenter (e.g. a caveman with a rock tied to a piece of string) knows nothing about pendulums yet:

1) "Doubling the length of a pendulum will have no effect on the period."

2) "Doubling the mass of a pendulum will have no effect on the period."

Are both examples of "A" null hypothesis, but they are not *The* Null Hypothesis; they are "the null hypothesis for this experiment."

TNH is a concept exclusive to statistics, that has a specific and narrow application, where the p-value is used. It is a (useful) tool when it is applied to data we might collect in an experiment, but it is widely misconstrued as having a broader application as a principle of epistemology.

Sometimes, also, as I think happens with Krebiozen, there is the hammer/nail problem, where he approaches every case looking to use what he is familiar with:

"The null hypothesis, such as it is, is that K recorded the birthdays accurately and that the students walking across the quad were selected randomly. In this case p is still about 0.7 for two birthdays being the same."

Note how he interjects "In this case p is still about 0.7 for two birthdays being the same." even though it is redundant and irrelevant. When is it *not* 0.7, if the selection is random?

More to follow on the more important issue of Krebiozen's agreement with the conceptual basis of much of climate Denialism at #428 and 416, unless he wishes to reconsider if he means what he said literally.

Even my worst teachers actually answered questions posed by the audience, as they came up. You, however, seem unable to do even that, even after dozens of posts.

zebra,
At this point I am seriously wondering what is wrong with you.

Exercise 2: A likely source of confusion for K and others (evidenced by his answers) is the distinction between
a) “The null hypothesis for this experiment”, and
B) ” *The* Null Hypothesis”.

Word salad. The term "the null hypothesis" only makes sense in the context of a specific experiment.

1) “Doubling the length of a pendulum will have no effect on the period.”
2) “Doubling the mass of a pendulum will have no effect on the period.”
Are both examples of “A” null hypothesis,

Apart from adding mass as a variable, how does this differ from my statement that, "You could, I suppose, argue that the null hypothesis is that there is no relationship between L and the pendulum's period"?

but they are not *The* Null Hypothesis; they are “the null hypothesis for this experiment.”

More word salad. In what possible context could we refer to "*The* Null Hypothesis" except in reference to a specific experiment, real or hypothetical? Do you not understand "a" and "the" as well as the other simple English words we have established you struggle to understand?

Only the experimenter knows what the null hypothesis is for a specific experiment, so unless she tells us we can only speculate what it might be, as I did.

TNH is a concept exclusive to statistics, that has a specific and narrow application, where the p-value is used. It is a (useful) tool when it is applied to data we might collect in an experiment, but it is widely misconstrued as having a broader application as a principle of epistemology.

No, the null hypothesis is a concept commonly used in different areas of science. The p-value is a concept exclusive to statistics, and is not appropriate for this kind of experiment, as I stated.

Sometimes, also, as I think happens with Krebiozen, there is the hammer/nail problem, where he approaches every case looking to use what he is familiar with:

You asked, "What is the p-value?" for the pendulum experiment, and I pointed out that this isn't the right tool for the job. Now you accuse me of using the wrong tool for the job? I used the right tool for the job in the first instance, and now you are apparently attempting to prove I was wrong to do so by coming up with examples of situations when it isn't the right tool.

I am familiar with a wide range of basic statistical tests such as t-test, f-test, chi-square, correlation, linear regression, multivariate analysis and others, and I know how to decide which is the right tool for a specific job. I used these statistical tools on a regular basis for decades, generating results that had life-changing consequences for patients, so please don't try to tell me I don't know how or when to use them. I have also played with more sophisticated statistics but I'm by no means an expert.

“The null hypothesis, such as it is, is that K recorded the birthdays accurately and that the students walking across the quad were selected randomly. In this case p is still about 0.7 for two birthdays being the same.”
Note how he interjects “In this case p is still about 0.7 for two birthdays being the same.” even though it is redundant and irrelevant. When is it *not* 0.7, if the selection is random?

Obviously p is not 0.7 when the selection is not random, which is the possibility that I assume your painfully clumsy example was intended to raise. Sometimes it is necessary to state the bleeding obvious, especially when one is dealing with someone who is apparently struggling to understand the subject.

More to follow on the more important issue of Krebiozen’s agreement with the conceptual basis of much of climate Denialism at #428 and 416, unless he wishes to reconsider if he means what he said literally.

At #416 I pointed out that we don't generally investigate the cause of phenomena that are consistent with our current models, e.g. chance. I don't think I need to reconsider that. At #428 I pointed out that global warming is an unprecedented phenomenon that does require an explanation other than chance or natural variation, which is why climate scientists have constructed mathematical models that explain it. I don't think I need to reconsider that either. On what planet does this make me a supporter of climate Denialism? The same planet where jokingly referring to the hidden variable as "possibly imaginary" makes me a believer in physics woo? Your desperation to prove me an idiot is extraordinary.

By Krebiozen (not verified) on 04 Feb 2015 #permalink

Isn't THE Null Hypothesis a show on PRN?

By Mephistopheles… (not verified) on 04 Feb 2015 #permalink

Krebiozen,

Sorry for the delay in replying.

I'm not trying to prove that you are an idiot. At this point I think we are making progress in clarifying the issues that originated with your initial comment and my reply.

I don't think it has been clear at all what you understand, and certainly it doesn't hurt for readers who don't have your experience to have things explained in a simple and coherent manner.

Let me offer a couple more cases, beginning with this from climate science (temps refer to global mean):

We have a nice recent study of paleoclimate that one can use to make a good case that there has not been a rise of .6C over a period of 100 years during the preceding 10K years. So, yes, this offers support for the claim that "something is going on" when we observe it now.

Now, here's your statement:

"If scientists looked for alternative hypotheses to explain every observation that was consistent with random chance they would get nowhere."

Let's apply it to this hypothetical case (assume we are not aware of the greenhouse gas effect):

The same study in fact finds that there *have been* ten instances of .6C/century over that 10Kyear period, at various intervals.

Do you seriously argue that it would be a waste of time to investigate the most recent instance, since it comports with 'chance'?

My issue is that you appear to be doing exactly what I said, which is using "the null hypothesis" as "the" default epistemological concept, and invoking Parsimony where it doesn't apply. And you don't answer that with 439.

If scientists followed your advice, little science would ever get done. Speculating about causes is what scientists *do.*

Ummm…. D’Agostini is pretty much saying what I’m saying.

I'd be fascinated to hear how you arrive at this conclusion.

I was even going to reference the Wikipedia article myself

Particularly given this demonstration of apparently leaping to one.

Let’s apply it to this hypothetical case (assume we are not aware of the greenhouse gas effect)

Jesus, it's like the drunkard's search. What part of "phenomena that are consistent with our current models, e.g. chance" do you not understand?

Ah, I see where "I was even going to reference the Wikipedia article myself" comes from: you seized upon pp. 6–7. This is quite impressive, given that your stock in trade is tediously failing to understand prior probabilities.

zebra, you seem to have missed the following bit from Krebiozen's post you argue is promoting climate chance denialism...

“The globe is warming at a faster rate than it ever has before” (my emphasis). Random chance is not a plausible explanation for the unprecedented warming we have seen since in recent decades.

It's pretty disingenious to try to argue that statement is "in conceptual agreement with much of climate Denialism" as you put it.

And if you intend to counter with this,

assume we are not aware of the greenhouse gas effect):

Why?, and no.

You insisted on delving into this side-track issue of global warming denialism, and now you need to move goal posts to keep your argument straight? It doesn't work like that. And even ignoring the moving goalposts, your argument basically boils down to "in a hypothetical situation where climate scientists wouldn't know anything about greenhouse gases, I postulate that they would not understand global warming, therefore I am correct." which is mind-numbingly, ridiculously faulty reasoning.

And you still haven't answered any of the simple questions posed to you, some hundreds of posts ago. Trying to chance the subject (and especially with such "logic") doesn't work.
Answer the questions finally, please.

zebra,
I see Narad and gaist have already made some of the following points, but anyway....

At this point I think we are making progress in clarifying the issues that originated with your initial comment and my reply.

Let's remind ourselves of those, shall we? Greg's original question was at #49, referring to his personal experience:

I wonder what the odds are on a person who has never had a flu shot going two decades without contracting the more recent strains of flu?

To which I replied at #52:

Since about 2-5% of people will get influenza each season you can expect to get it every 20-50 years on average. Going 20 years influenza-free is par for the course, especially when you had the disease unusually frequently in the past. Older people do sometimes have resistance to circulating strains, such as the H1N1 strain in 2009.

It seems clear to me that I was making a number of assumptions to roughly estimate the odds of getting influenza based on the number of people who get it each season. That rough estimation tells me that "going two decades without contracting the more recent strains of flu" is nothing unusual. I also speculated that since Greg reported having influenza four times when he was younger he might have developed a resistance to circulating strains. This all seems perfectly reasonable, simple and coherent to me.

Your response to this was at #83:

Gambler’s fallacy. Also consider Bayesian analysis. And sample size.

I still don't see how that was at all helpful in clarifying what I meant, or how it would "get people thinking about the issue in a ‘scientific’ way".

Your "gambler's fallacy" remark was clearly a mistake, whether you acknowledge it or not.

The reference to Bayesian analysis is unhelpful. What prior probability of getting influenza can we attribute to Greg, given the available information?

As for sample size I based this estimate on various clinical trials with sample sizes in the thousands, as I have since explained.

Whether we put it in terms of a null hypothesis and parsimony or not, the bottom line is that in the absence of any reason to think Greg is in any way unusual, it is not reasonable to conclude that he is.

I don’t think it has been clear at all what you understand, and certainly it doesn’t hurt for readers who don’t have your experience to have things explained in a simple and coherent manner.

What precisely have I not explained clearly and coherently? Is anyone but you struggling to understand what I have explained?

Let me offer a couple more cases, beginning with this from climate science (temps refer to global mean):

The Earth's climate is a hugely complex system with one of a number of endpoints (in this case average global temperature) dependent on an enormous number of variables fluctuating greatly over millennia. The concerns over climate change are expressed by the IPCC as probabilities of different degrees of catastrophic warming over the coming years, based on variations in the mathematical model(s) climate scientists have constructed. As I understand it the probability of the observed changes being due to nothing but chance are vanishingly small.

With the Greg/influenza question we are looking at the reported experience of one individual and comparing that to a simple mathematical model that assumes that whether or not an individual contracts influenza in a given year is determined by chance, and that this follows a Gaussian distribution. In the absence of anything to make us believe otherwise I think the assumptions I made are reasonable. I see no reason to postulate super-immunity to influenza or anything else to explain Greg's experience, which is entirely consistent with a combination of chance and perhaps increasing immunity to some strains of influenza with age. Nothing you have written since has changed my mind about this.

How does comparing these two scenarios clarify anything?

Also, climate change is a subject I know only a little about, so I'm in danger of saying something stupid, but I'll run with it anyway.

Now, here’s your statement:
“If scientists looked for alternative hypotheses to explain every observation that was consistent with random chance they would get nowhere.”
Let’s apply it to this hypothetical case (assume we are not aware of the greenhouse gas effect):
The same study in fact finds that there *have been* ten instances of .6C/century over that 10Kyear period, at various intervals.
Do you seriously argue that it would be a waste of time to investigate the most recent instance, since it comports with ‘chance’?

Firstly, why should we assume that we are not aware of greenhouse gas effects? They are important mechanisms in the models climate scientists have made to explain changes in temperature. You can't look at the history of our understanding of climate change in isolation from CO2, water vapor and methane levels as these were and are major factors.

Secondly, are you seriously suggesting that these previous increases in the Earth's temperature were due to random chance? As I understand it they had causes, whether greenhouse gases, variation in solar output, ocean current changes, volcanic activity or whatever. The Medieval Warm period, for example, was likely "mainly a regional phenomenon caused by altered heat distribution rather than a global phenomenon". If there is a climate model that convincingly explains the observations of these temperature changes as the results of random chance, I am not aware of it.

My issue is that you appear to be doing exactly what I said, which is using “the null hypothesis” as “the” default epistemological concept, and invoking Parsimony where it doesn’t apply. And you don’t answer that with 439.

Taking the null hypothesis as the default epistemological concept is the basis of experimental science, is it not? In what circumstances does parsimony not apply?

Our mathematical model predicts that the odds of X happening to an individual by chance is approximately even (see Vicki's calculation at #330). We observe X happening to an individual (Greg's reported experience). The null hypothesis, in this case random chance, perfectly explains our observation. I see no reason to assume there is anything unusual about that individual or that anything other than chance is involved. Is there another hypothesis that perfectly explains our observation that does not involve the invention of additional entities such as super-immunity? Why would parsimony not apply here?

If scientists followed your advice, little science would ever get done. Speculating about causes is what scientists *do.*

Scientists speculate about the causes of observations that are not consistent with existing models. They also construct experiments to test a hypothesis against a null hypothesis; if their experimental results are consistent with the null hypothesis they reject the hypothesis. You appear to be rejecting the entire epistemological basis of scientific experiment, which seems a bit ambitious, and frankly I don't understand your arguments for doing so. I'm not convinced you understand them yourself.

By Krebiozen (not verified) on 08 Feb 2015 #permalink

Krebiozen,

"Taking the null hypothesis as the default epistemological concept is the basis of experimental science, is it not?"

Well no, it isn't. This is the problem I've been describing. You think that because you have used it in your statistical testing, and you don't have a broader vision.

Please explain the difference between these two hypotheses:

a) "Doubling the length of the pendulum will have no effect on the period."

b) "Doubling the length of the pendulum will change the period."

Why is one a 'better' assumption than the other? Why is (a) the default and "the basis of experimental science"?

Each hypothesis is the inverse of the other, naturally.

Which is the better assumption than the other depends on previous observation. If one has noticed that longer pendulums appear to have a different period than shorter ones, then hypothesis B is the better starting point. However, if one has not noticed a difference, or has made no observations at all, the better assumption is A.

Why is A better in the absence of observation to the contrary? Because the set of things that don't affect the period of a pendulum is many orders of magnitude larger than the set of things that do. If the starting assumption was that everything affected the period of a pendulum, then it would be nearly impossible to ignore factors we now know have no impact, such as temperature, time of day, alignment of the planets, and, interestingly, the mass of the pendulum.

Of course, once someone has observed that two pendulums of different lengths have different periods, one discards hypothesis A as disproven.

By Mephistopheles… (not verified) on 08 Feb 2015 #permalink

it would be nearly impossible to ignore factors we now know have no impact, such as temperature

Ahem.

frankly I don’t understand your arguments for doing so

The reason, on the other hand, seems plain.

Narad - I see you caught my discussion point.

By Mephistopheles… (not verified) on 08 Feb 2015 #permalink

#448

zebra:

"Please explain the difference between these two hypotheses:"

O'Brien:

"Each hypothesis is the inverse of the other, naturally."

*sigh*

That would make them identical, wouldn't it?

And your argument goes downhill even from that; it is not remotely relevant-- we are only talking about *one* variable, and the question I asked is why the null has some privileged place in Krebiozen et al's epistemology over the affirmative.

The correct answer, as everyone should know, is that *there is no privileged hypothesis*-- we use the null form by convention, and not everyone even does that.

Now if Krebiozen would just agree to this obvious fact, I can explain how he is misapplying Parsimony.

With the Greg/influenza question we are looking at the reported experience of one individual and comparing that to a simple mathematical model that assumes that whether or not an individual contracts influenza in a given year is determined by chance, and that this follows a Gaussian distribution.

Binomial; for a 4% chance, Greg Young would need to be about 245 years old for the approximation of normality to hold, by my W—reckoning.

I can't believe I missed this gem:

When we design an experiment, we require a working hypothesis to justify what we are doing.

Did you pick up your comically grandiose "understanding" of the subject at the Exploratorium or something?

The correct answer, as everyone should know, is that *there is no privileged hypothesis*– we use the null form by convention, and not everyone even does that.

Hey, that reminds me of something.

Yet, some 150 comments later, you're doing the same thing.

*there is no privileged hypothesis*

I suddenly have the sinking feeling that Zebra belched up GR for a painfully sophomoric reason.

“Each hypothesis is the inverse of the other, naturally.”

*sigh*

That would make them identical, wouldn’t it?

No, that would make them opposites.

By Mephistopheles… (not verified) on 08 Feb 2015 #permalink

Somebody (*cough*zebra*cough*) apparently doesn't understand the meaning of the word "inverse," and didn't bother looking it up upon encountering it.

Hey, kids! If you don't know, look it up!

zebra, I think it's a bit disingenuous to ignore all the valid criticism in Krebiozens, Narads and others posts and just pick a sentence or two to jump upon - especially when you can't even manage that without having to ignore basic principles.

I think in my first post in this thread I described you as something of a whiny troll. I haven't had reason to change my estimate, except to perhaps precede it with insistent and follow up with who can't deal with being wrong.

Answer the questions asked of you, and answer these as well:

Why bring up (imagined) simlarities between global warming denialism and Krebiozen if even in your head for it to be true we'd have to ignore central, basic principles of climate science?

Why not address the criticism when it was pointed out to you?

In what way are opposites identical?

And if you reject unequivocally priviledged null hypotheses, (and not just in your made-up arbitrary thought experiments) who do you think in a scientific debate has burden of proof? No-one? The group accepting previous validated tests or the group challenging previously accepted tests?

For f*ks sake... I'll never post via this phone again.

$459 is by me, not zebra. Below is the message again, in the hope that the previous one gets deleted...

-

zebra, I think it’s a bit disingenuous to ignore all the valid criticism in Krebiozen's, Narad's and others' posts and just pick a sentence or two to jump upon – especially when you can’t even manage that without having to ignore basic principles or resort to frantic hand waving.

I think in my first post in this thread I described you as something of a whiny troll. I haven’t had reason to change my estimate, except to perhaps precede it with insistent and follow up with who can’t deal with being wrong.

Answer the questions asked of you, and answer these as well:

Why bring up (imagined) similarities between global warming denialism and Krebiozen if even in your head for it to be true we’d have to ignore central, basic principles of climate science?

Why not address the criticism when it was pointed out to you?

In what way are opposites identical?

And if you reject unequivocally privileged null hypotheses, (and not just in your made-up arbitrary thought experiments) who do you think in a scientific debate has burden of proof? No-one? The group accepting previous validated tests or the group challenging previously accepted tests?

In what way are opposites identical?

There aren't any opposites, you ninny. Nonetheless, it appears that the zorse is only interested in trying to hump Krebiozen at this point.

I will respond to zebra when I'm a little less busy.

By Krebiozen (not verified) on 09 Feb 2015 #permalink

FWIW, every word in my comment 448 was in response to the questions in 447 as asked. My points were relevant, well reasoned, and insightful.

I don't usually like to point that out, as I believe it should be obvious and it offends my natural modesty to do so.

By Mephistopheles… (not verified) on 09 Feb 2015 #permalink

zebra,

“Taking the null hypothesis as the default epistemological concept is the basis of experimental science, is it not?”
Well no, it isn’t.

Practically any text book on empirical science will tell you that it is. Do you have a citation to support your assertion that it is not?

This is the problem I’ve been describing. You think that because you have used it in your statistical testing, and you don’t have a broader vision.

Well, I took three years out of my career as a biomedical scientist to study social anthropology, a great deal of which involved discussing epistemology, especially the basis of 'western' scientific thought which, as you might imagine, came up a lot given my background and the popularity of post-modernism at the time, so I dispute your claim that I lack a broader vision. This is something I have spent a lot of time reading about, thinking about and discussing with a wide range of people.

Please explain the difference between these two hypotheses:
a) “Doubling the length of the pendulum will have no effect on the period.”
b) “Doubling the length of the pendulum will change the period.”
Why is one a ‘better’ assumption than the other? Why is (a) the default and “the basis of experimental science”?

There are several different ways to answer your question.
I cannot demonstrate logically that assuming (a) is the best way of doing things, any more than I can offer a logical proof of Occam's Razor or parsimony in general. I can only appeal to pragmatism and point out it is a convention that has been tried and tested and has proven very effective. However, I think there are a number of reasons that it has proven such a useful approach, such as:

1. Hypothesis (a) is precise whereas hypothesis (b) is not. Doubling L might double P or halve it, or relate to it in an infinite number of other ways. Adopting (a) as our default hypothesis gives us a baseline to compare our experimental observations to, whereas hypothesis (b) does not. This applies to any null hypothesis, which essentially states that there is no relationship between two variables, as opposed to an alternative hypothesis that there is some, as yet unknown, relationship between them.

This allows us to establish a mathematical model for our null hypothesis, by using a control group or otherwise, allowing observations of our active group to be compared statistically to that model. It is, obviously, impractical to construct an infinite number of mathematical models to describe all possible ways in which the null hypothesis might be false.

2. In this case we are attempting to construct a mathematical model of how pendulums behave on Earth. Assuming that we know nothing, i.e. that nothing is related to anything else, is a good starting point, allowing us to add bits to our models as we generate and test hypotheses. It is a systematic and methodical way of building an accurate model of reality that allows us to make accurate predictions.

3. This approach helps us to avoid cognitive biases, such as confirmation bias. If we adopt a hypothesis and then attempt to prove it, we can very easily fool ourselves. Adopting the null hypothesis in each experiment and only accepting a competing hypothesis if the experimental data support it protects us against this.

I'm sure there are other reasons that this approach has proven so successful, but these are the ones that immediately occur to me.

I could continue with a general discussion of parsimony and why it is a useful rule of thumb, but I won't. Others have done it much better than I could.

This is all very interesting but has little relevance to the original question of Greg's influenza, where we already have a well-established model of reality that we can use to make predictions. That model is based on our understanding of viral infections, how they are spread and immunology in general. That model, and various experimental data, tells us the approximate number of people who get influenza each year, it tells us that (as far as we know) no one is immune to influenza unless they have been exposed to that strain of the virus or have been vaccinated against it, and that while susceptibility to complications may vary between groups, susceptibility to infection does not.

Given this, I think it is reasonable to assume a binomial distribution (approximating to a Gaussian distribution - thanks Narad) of the period that people go between bouts of influenza, with a very few getting it twice in rapid succession, a very few going their entire lives without getting it at all and the vast majority somewhere in-between. Comparing Greg's reported experience to that model we see that it is consistent with it, and that we can expect a sizable proportion of the unvaccinated population to have had a similar experience to his. My conclusion that Greg's experience does not suggest he possesses super-immunity to influenza is valid.

How this fairly simple practical application of a simple mathematical model has led to you challenging the validity of the concept of the null hypothesis and parsimony beats me.

By Krebiozen (not verified) on 09 Feb 2015 #permalink

#465

"How this fairly simple practical application of a simple mathematical model has led to you challenging the validity of the concept of the null hypothesis and parsimony beats me."

I am not questioning the validity of either, only the way you have invoked the terms in trying to justify your initial comment, which I believe is in error due to your misuse of those concepts.

(Let's remember that I very clearly stated that I was not questioning the conclusion that Greg did not have SI, only your reasoning-- as expressed-- for arriving at that conclusion.)

The issue is that you are conflating what you call a model (or theoretical construct, or Theory) with "hypothesis". We can falsify a hypothesis, but we can not 'falsify' a model.

An experiment is just an experiment, and it it tests its hypothesis-- that's it. But I give you this simple example, and you once again go off on a long-winded exposition on the implications for Theories not yet even imagined by my caveman with a rock on a string. That you have done this more than once further convinces me that you are not getting the difference. And I mean when you apply it, whatever you might say theoretically.

We apply parsimony (OR) to Theories. It is not a test of truth; Theories are not falsified, they are replaced. So, we choose between Theories that yield identical results, based on the profusion of entities, in particular entities that require themselves further examination. That is literally pragmatic parsimony, if you think about it.

So, what about Greg?

Once again, if you say:

"This is all very interesting but has little relevance to the original question of Greg’s influenza, where we already have a well-established model of reality that we can use to make predictions. That model is based on our understanding of viral infections, how they are spread and immunology in general. That model, and various experimental data, tells us the approximate number of people who get influenza each year, it tells us that (as far as we know) no one is immune to influenza unless they have been exposed to that strain of the virus or have been vaccinated against it, and that while susceptibility to complications may vary between groups, susceptibility to infection does not."

If you are working from this "well established model", then, the correct answer to Greg is: "Greg, there's no such thing as SI; *it doesn't matter if you go 20 years or 100 years* without catching the flu, your experience is unremarkable."

That would be the correct application of your Theory; random is random. *Calculating the probability* using the assumption of randomness is unnecessary and meaningless.

#460

gaist,

I'm happy to answer questions that I consider serious and not attempts at distraction.

So, let's look at what K said:

" and various experimental data, tells us the approximate number of people who get influenza each year, it tells us that (as far as we know) no one is immune to influenza unless they have been exposed to that strain of the virus or have been vaccinated against it, and that while susceptibility to complications may vary between groups, susceptibility to infection does not. "

And, he has cited actual experimental studies that involve exposing people to determine the extent of immunity.

So, can you tell me why they did those studies in the first place? Why, if "random" is always the best answer, would these people who specialize in the field *look for* something like what we are calling 'super immunity'? The answer is, *it's what scientists do*. They look for stuff they don't already know about.

My climate example was simply a version of this. Even if there was evidence of rapid temperature changes in the past, and even without knowing about GHG, climate scientists would surely be interested in the current phenomenon. They wouldn't just say "well, it's happened before so humans couldn't be the cause; let's just assume it's internal system variation."

zebra,

I am not questioning the validity of either, only the way you have invoked the terms in trying to justify your initial comment, which I believe is in error due to your misuse of those concepts.

A distinction without a difference, I believe that is called, given your evident ignorance of how these concepts are used.

(Let’s remember that I very clearly stated that I was not questioning the conclusion that Greg did not have SI, only your reasoning– as expressed– for arriving at that conclusion.)

That does make me wonder what the point of your comment at #83 was.

The issue is that you are conflating what you call a model (or theoretical construct, or Theory) with “hypothesis”.

No, a model is a representation of a part of the universe. It can be a hypothesis, if it is tentative and unproven, or it can be a scientific theory if it is well-established and supported by lots of evidence. A model can be plain wrong, like the geocentric model of the solar system. Take a look at the Wiki on statistical modeling, especially the parts about evaluating and comparing models, which explains the basics pretty well.

We can falsify a hypothesis, but we can not ‘falsify’ a model.

Nonsense. If we find data that does not fit our model we have falsified it. A hypothesis can be in the form of a mathematical model, for example "doubling the length of a pendulum will double its period" is a description of a simple mathematical model that is also a falsifiable hypothesis.

An experiment is just an experiment, and it it tests its hypothesis– that’s it. But I give you this simple example, and you once again go off on a long-winded exposition on the implications for Theories not yet even imagined by my caveman with a rock on a string.

You asked, "Why is one a ‘better’ assumption than the other?" When I try to explain you complain about a long-winded exposition? Is your inability to understand even the simplest scientific concepts somehow my fault? If you showed some sign of understanding this stuff I wouldn't have to explain it in painful detail. I find it extremely peculiar that you take this pompous didactic stance when you clearly don't have a clue.

That you have done this more than once further convinces me that you are not getting the difference. And I mean when you apply it, whatever you might say theoretically.

No, I'm not getting the difference. I suspect that's because you are talking gibberish.

We apply parsimony (OR) to Theories.

We can apply parsimony, as a general principle, wherever we damn well like. Why would anyone think it is restricted to scientific theories? It's a useful general principle in multiple contexts.

Why do you randomly capitalize some terms, like 'parsimony' and 'theories', by the way?

It is not a test of truth;

No one has suggested that it is. In fact I have repeatedly explained that it is not.

Theories are not falsified, they are replaced.

If data emerges that contradicts an existing theory, it is indeed falsified, and then eventually replaced with a better theory.

So, we choose between Theories that yield identical results, based on the profusion of entities, in particular entities that require themselves further examination. That is literally pragmatic parsimony, if you think about it.

That is one application of parsimony, yes. I find it curious that you evidently cannot conceive of parsimony being useful in other contexts.

If you are working from this “well established model”, then, the correct answer to Greg is: “Greg, there’s no such thing as SI; *it doesn’t matter if you go 20 years or 100 years* without catching the flu, your experience is unremarkable.”

No, because we do not know for certain that "there's no such thing as SI", which is why I have stated it as, "as far as we know". SI is a viable hypothesis. Also, I would say that an unvaccinated person going 100 years without a single bout of influenza is certainly remarkable enough to warrant my curiosity.

Anyway, this isn't an examination; there is no "correct answer". Stating that an individual's experience is consistent with a model of randomness, and is one shared by a substantial proportion of the population purely by random chance is not incorrect, it is perfectly true.

That would be the correct application of your Theory; random is random. *Calculating the probability* using the assumption of randomness is unnecessary and meaningless.

Good grief, you are horribly confused about this. If someone had thrown a die three times, got three sixes and speculated that this meant the die was rigged, would you respond, "random is random"? Or would you point out that the odds of throwing three sixes is approximately 0.17, well within the range we can expect from chance? If they threw 100 sixes in a row, would you respond the same, "random is random", and insist that rigged dice do not exist? Can nothing ever happen that is extraordinary enough for you to think that something other than chance might be operating?

Greg speculated, "Maybe I just have a very high resistence to viruses" [sic], and specifically asked, "I wonder what the odds are on a person who has never had a flu shot going two decades without contracting the more recent strains of flu?"

To answer that question we have to estimate those odds, don't we? As Narad kindly told us at #101, "the 65.9% odds (1 σ) for 2%–5% annual risk are 53–21 years, respectively". If the odds of going at least 21 years without getting influenza are 0.66, I think we can safely conclude that 20 years is unremarkable.

I keep wondering if you really have some salient point to make, but I am being forced to the conclusion that for some reason you are simply unable to understand this.

By Krebiozen (not verified) on 10 Feb 2015 #permalink

I’m happy to answer questions that I consider serious and not attempts at distraction respond when I can read from my confused script with Random capitalization.

FTFY.

If you are working from this “well established model”, then, the correct answer to Greg is: “Greg, there’s no such thing as SI; *it doesn’t matter if you go 20 years or 100 years* without catching the flu, your experience is unremarkable.”

That would be the correct application of your Theory; random is random.

And, again, half-life doesn't exist. Your ability to combine pomposity with rank ignorance is an amazing thing to behold.

The answer is, *it’s what scientists do*. They look for stuff they don’t already know about.

This is a choice bit of self-contradiction when paired with this:

When we design an experiment, we require a working hypothesis to justify what we are doing.

The problem, of course, is that Zebra has no understanding whatever of experimental science. There are plenty of experiments with no "working hypothesis" (ironically including everything encompassed by "back in the beginning of physics, the very first [sic]* experiment," i.e., the whole of pedagogical lab sections), unless one wants to strip the term – as it were – of any semantic value, which would make Zebra's tediously evasive prattling even worse.

Sometimes, the experiment is measurement. There is no meaningful "hypothesis" needed to "justify," say, refining the present** value of the fine-structure constant.

* No, introductory mechanics doesn't start with SHOs, but I've already pointed that out.
** One can always hope.

Just as an aside, I made a fair amount of money on software that looked at big data to find relationships. The seatbelt usage to cigarette smoking strong statistical relationship being the most famous that I never acted upon. Finding statistical significance without hypothesis is still "science". It just means after finding a relationship, it is up to science to investigate and determine if there is causality.

We linked poor dental health to coronary events years before anyone discovered the mechanism.

By Colonel Tom (not verified) on 10 Feb 2015 #permalink

I’m happy to answer questions that I consider serious and not attempts at distraction.

I can sort of understand why someone might think admitting their own mistakes would feel distracting, but it's still disingenuous debating tactics. Especially if one is wont to repeatedly (if erroneously) claim "victory" while harping on what one views as errors in other people, like you have.

And most of the time you've addressed even if not answered outright the questions by further insisting on your point (and claiming victory in the process), digging yourself deeper, until you've realized there's no way out except by gish galloping elsewhere or by moving the goal posts. So it's not like you regarded them as distracting side issues when you thought you could "win" them.

Also, even apart from not answering questions you think are beneath the issue at hand, your practice of ignoring all the valid, painstakingly explained criticism and just repeatedly repeating the same claims, is - while admittedly not distracting - also not useful. If you think nobody else here reading this back-and-forth won't pick up on the criticism and your failure to address it, and indeed your active evasion of it, and come to the obvious conclusions, you are fooling yourself. If you think they'll proclaim you victor and with online high fives and brofists cheer and congratulate you on winning the internetz despite your obvious evasions and gish gallops, you're fooling yourself even more.

If you are interested in anything else than trying to hump Krebiozen, you should try to act like you were at least trying to learn from your numerous mistakes.

#468

Let's review what you said again:

"we already have a well-established model of reality that we can use to make predictions. That model is based on our understanding of viral infections, how they are spread and immunology in general. That model, and various experimental data, tells us the approximate number of people who get influenza each year, it tells us that (as far as we know) no one is immune to influenza unless they have been exposed to that strain of the virus or have been vaccinated against it, and that while susceptibility to complications may vary between groups, susceptibility to infection does not.”"

And my reply:

"If you are working from this “well established model”, then, the correct answer to Greg is: "

It is clear that you and everyone else here has a strong conviction that the model is correct-- that there is no such thing as SI. In fact, suggesting otherwise (I assume that's what Maher does?) provokes universal outrage and indignation amongst the Minions.

So, I say *IF you are working from this "well established model" (about which you have a strong conviction)*, your answer to Greg should be as I described. There is no SI, whatever length of time he has gone without flu, so he *can't possibly* have SI.

Now, if you are saying that Greg going 100 years would shake your conviction about the non-existence of SI, that's fine. But that's not the question we are answering-- it is whether Greg has SI or not.

So, (and here's the significance of all the scientific theory [Theory] v hypothesis, and OR, and null, definitions), you would have to first *replace* your no-SI Theory with one in which SI and its prevalence are characterized. Only then could you determine if Greg has it or not.

But it is pretty obvious that nobody here would abandon the no-SI model based on one person going 100 years without the flu. That's the kind of anecdotal argument someone like Maher would use, right?

To me, this is all trivially obvious, and it has been stated as clearly as it can be.

I think it has been worthwhile to have the discussion about scientific reasoning; anyone who wants to look at it some more should read the references Narad gave. You will see that there is much disagreement about statistical interpretations, and one should always examine the underlying physical and logical structures before reaching for the calculator.

bye

zebra,

It is clear that you and everyone else here has a strong conviction that the model is correct– that there is no such thing as SI. In fact, suggesting otherwise (I assume that’s what Maher does?) provokes universal outrage and indignation amongst the Minions.

I would never exclude the possibility that some lifestyle, supplement, genetic trait or whatever might make someone unusually resistant to viral infections; I just haven't seen any convincing evidence that this is so. It is making claims that such a state of super-immunity definitely exists (often accompanied by claims that a specific supplement or diet will confer it) that I object to, not proposing it as a testable hypothesis.

So, I say *IF you are working from this “well established model” (about which you have a strong conviction)*, your answer to Greg should be as I described. There is no SI, whatever length of time he has gone without flu, so he *can’t possibly* have SI.

Such absolutism would put an end to science. As Dara O'Briain put it, "if science knew everything, it would stop". There are always new hypotheses to consider, and it is not a scientific approach to reject one out of hand. Is all the data compatible with our new hypothesis? Does it explain the data more elegantly than our current model? Does this specific observation fit with our current models? If not, is it unusual enough to warrant proposing an alternate to the currently accepted model or to otherwise investigate further? There are questions that any scientist should be asking continually.

Now, if you are saying that Greg going 100 years would shake your conviction about the non-existence of SI, that’s fine.

It wouldn't shake my conviction, it would make me go "hmm", and ask more questions.

But that’s not the question we are answering– it is whether Greg has SI or not.

Again you display a curious insistence on deductive certainty. What I am describing is inductive reasoning, which tells us which conclusion is most likely. We clearly do not have enough information to conclude categorically whether or not Greg has SI. Greg proposed an alternate hypothesis: "Maybe I just have a very high resistence to viruses" [sic]. I assessed whether that hypothesis is worth considering given the available information.

So, (and here’s the significance of all the scientific theory [Theory] v hypothesis, and OR, and null, definitions), you would have to first *replace* your no-SI Theory with one in which SI and its prevalence are characterized. Only then could you determine if Greg has it or not.

You do seem to be very confused between the difference between a hypothesis and a scientific theory. That's ironic for someone who insists on capitalizing 'theory' to denote, I assume, the distinction. I also note that your use of the word 'determine' further supports my suspicion that you are stuck in deductive reasoning, which is not what this discussion has been about, though you seem unable to grasp this. Anyway...

Firstly, I don't have a no-SI scientific theory (I can't bring myself to capitalize 'theory'). I certainly would not elevate the idea that SI does not exist to that level of certainty. It's a working hypothesis that is open to challenge, like many things in biology and other complex systems. It certainly isn't impossible that there is a range of susceptibility to viruses, determined by genetics and/or lifestyle, and FSM knows what else, indeed it is quite likely that there is some degree of variation. The mouse paper Narad cited at #356 suggested as much. We just don't yet know for sure if that is the case in humans and, if it is, what the degree of variation might be.

I will note in passing that the fact that Greg reported having had influenza four times does argue against an innate, genetic SI, but not against an acquired SI.

Secondly, as we have discussed before, you don't have to have an explicitly characterized competing hypothesis to challenge an existing one. The hypothesis, "all swans are white" is falsified equally by a black swan, by a green swan and by a polka-dot swan. I do not have to propose the existence of black swans, or their prevalence. The hypothesis "changing the length of a pendulum will not affect its period" is falsified by any consistent relationship observed between length and period - plotting length against period and finding the points fall on a line or curve will suffice, and any hypothesis will follow from that correlation. Reliable records of a family that has never contracted influenza over several generations would be enough to challenge the hypothesis that SI to influenza does not exist, without any sort of characterized hypothesis required.

As a matter of fact, even if "SI and its prevalence are characterized", the two frequency distribution curves must overlap, since it is possible for someone to never get influenza by pure chance. So we still couldn't categorically say whether Greg belongs to the non-SI or the SI group. We would still only be able to express this as a probability (the screening for Down's Syndrome I was involved with used an algorithm to calculate which curve, Down's or non-Down's, a specific patient was most likely to belong to).

But it is pretty obvious that nobody here would abandon the no-SI model based on one person going 100 years without the flu. That’s the kind of anecdotal argument someone like Maher would use, right?

No one here has proposed abandoning any hypothesis (or scientific theory for that matter) based on a single anecdote. What I have proposed is that a single anecdote can sometimes be hypothesis-generating, but in this case, since the evidence is consistent with our existing model, it isn't.

To me, this is all trivially obvious, and it has been stated as clearly as it can be.

Based on what you have written here, I don't believe any of this has been the slightest bit obvious, trivially or otherwise, to you. I could be wrong, of course.

I think it has been worthwhile to have the discussion about scientific reasoning;

Perhaps, though I still think you have been insisting on using deductive reasoning where inductive reasoning is more appropriate/useful. I wish I had picked up on that earlier.

anyone who wants to look at it some more should read the references Narad gave.

Always good advice.

You will see that there is much disagreement about statistical interpretations, and one should always examine the underlying physical and logical structures before reaching for the calculator.

Yes there is some disagreement, but I don't think many statisticians would disagree with my approach to this particular problem. I still maintain that quick and dirty back-of-an-envelope estimations, of the sort I made in my original comment, can be very useful. We don't always need to be able to calculate exact values to decide whether evidence challenges existing models or not. A rough estimate to the nearest degree of magnitude will often suffice. This comes up an awful lot in practical scientific endeavors, and I make a habit of using rough estimates to confirm more precise calculations.

bye

Bye. I do hope this exchange has been of some use to you and others.

By Krebiozen (not verified) on 11 Feb 2015 #permalink

You will see that there is much disagreement about statistical interpretations, and one should always examine the underlying physical and logical structures before reaching for the calculator.

It's fitting that you started out with an aimless recitation of mantras and ultimately wound up demonstrating that everything that followed collided in perfectly inelastic fashion with your head.

bye

And that your capping item is neither more nor less than an etymological failure.