Antivaccine nonsense Bad science Medicine

Fun with Excel, or: Steve Kirsch is an antivax fool

Once again, Steve Kirsch has incompetently “analyzed” an Excel spreadsheet containing epidemiological data to claim that COVID-19 vaccines increase the chances of getting COVID. He desperately needs to take an Epidemiology 101 course.

Ever since I first discovered him in 2021 weaponizing the Vaccine Adverse Events Reporting System (VAERS) database to demonize COVID-19 vaccines, tech bro turned rabidest of rabid antivaxxers Steve Kirsch has been an all too frequent topic of this blog, if only for the simple reason that he is the living epitome of the arrogance of ignorance and black hole density Dunning-Kruger syndrome. Even back then, Kirsch’s most annoying characteristic was on display, an utter ignorance about the scientific topics that he was rambling on about combined with supreme overconfidence that led him not only to confidently proclaim that he knew better than experts who had dedicated their lives to a scientific discipline. I soon found that this arrogance of ignorance was coupled with a pugilistic streak that led Kirsch to go on the attack when criticized without the slightest self-reflection that might lead him to ask whether he might have made a mistake or not. This is yet another example of this tendency. It involves a FOIA request and an amusing (for a tech bro) inability to properly understand Excel spreadsheets.

He is also one of that most annoying variety of cranks, the “debate me, bro,” who seems to think that the only way to resolve scientific questions is through public “debates” on video or podcast. Over the last couple of years Kirsch has developed two signature go-to moves. One is to challenge science communicators to “debate” him and/or selected members of his posse of antivax pseudoscientists and quacks, offering a “reward” or just payment to anyone who takes him up on his offer and goes through with the “debate.” (That’s why I call him a “debate me bro” and his tactics, “debate me bro” tactics.) Kirsch’s second signature go-to move has been to challenge vaccine advocates to “bets” of up to $1 million if they can “prove him wrong.” Unsurprisingly, whenever anyone tries to accept these “challenges” and bets, things do not exactly progress in a straightforward manner, and those who have tried to “debate” him have generally found the usual result: A Gish gallop and refusal to admit defeat. One time, when a pseudonymous doctor on social media demonstrated that he was wrong and accused him of welching on his bet, Kirsch’s go-to move was to dox him and threaten to sue him.

Suffice to say, Kirsch never admits significant error. (At least if he ever has since the pandemic began, I haven’t seen it.) The most recent example is his hilariously off-base “analysis” of data in an Excel spreadsheet regarding COVID-19 and vaccination in Santa Clara County, California obtained, apparently, through a FOIA request:

Based on new data I just got from a FOIA request, it appears that the public health epidemiologists in Santa Clara County knew in January 2022 that the vaccines made people more likely to get COVID, but they remained silent.

I predict that there will be further silence on this matter: no accountability and no opportunity for public challenges. They will continue to push the shots as if nothing had happened and the mainstream media will ignore this important data.

Kirsch’s analysis is, like all his “analyses” of epidemiological data, total off-base, and unsurprisingly he comes to the wrong conclusion because that wrong conclusion aligns with his antivax bias. Let’s just put it this way. Epidemiologists in Santa Clara County obviously came to no such conclusion. That’s because they know how to analyze epidemiological data. Kirsch, who does not, concludes otherwise and, because of his arrogance and lack of introspection, doesn’t even entertain for a moment the possibility that he might be wrong. Instead, as all conspiracy theorists do, he concludes not that he is wrong, but that the epidemiologists in Santa Clara County “knew” that the COVID-19 vaccines didn’t prevent disease—and that they even increased the risk of disease!—but “remained silent,” presumably as part of a coverup.

So let’s see how Kirsch approached the data in this spreadsheet, which is surprisingly small. I downloaded it myself, and it’s only 7.8 MB. It has two tabs, the first one being a pivot table, the second being the data used to produce the pivot table of COVID-19 cases in Santa Clara County in January 2022:

Kirsch's pivot
This is the summary. This is what Kirsch thinks is a “smoking gun.” Note that N=unvaccinated; Y=vaccinated; U=vaccination status unknown.

Now, behold Kirsch’s genius “analysis”:

Santa Clara County is highly vaccinated (95%), but it isn’t that highly vaccinated!

The rows are 10 year age groups.

So the percentage of people who were diagnosed with COVID (98% or more) was higher than the percentage of people who got the vaccine (under 95%).

In other words, the vaccine made you more likely to get COVID instead of 10X lesslikely that they claimed in the clinical trials.

I haven’t even gotten to the fun with Excel spreadsheets yet, and I was seriously facepalming at this.

Godzilla facepalm
This is how I generally end up reacting to Steve Kirsch’s “analyses.”

Maybe, with the release of Godzilla x Kong: The New Empire this weekend, I can find a King Kong facepalm to go along with Godzilla. Oh wait, never mind:

King Kong facepalm
Now, both Godzilla and King Kong agree that Steve Kirsch is facepalm-worthy.

A little sarcastic fun aside, though, Kirsch seems to think that these data somehow demonstrate that COVID-19 vaccines increased the risk of getting COVID:

The numbers here are highly statistically significant. It appears that the Santa Clara County Health epidemiologists knew something was wrong by January 2022, but instead of warning people, they kept their mouths shut about it. There was no public admission of this, no public warning. I predict that there will NEVER be any public accountability of what happened because public officials never like to admit they were responsible for killing people with these useless and deadly vaccines.

And I predict that, after looking at my discussion and the two deconstructions of where he went wrong that I cite, Kirsch will never take public accountability for doing seriously bad analyses and making rookie mistakes, but he will keep saying stupid things like this:

The COVID vaccine trials were fraudulent. There is no possible way they got 90% efficacy (a 10X reduction in infection risk). They did it through deception as described here. The vaccines actually made you more likely to get COVID as we learned from the Santa Clara data, the Cleveland Clinic study, and numerous other sources (see this article for example).

Peter Doshi. He’s citing Peter Doshi, The BMJ‘s resident antivax-adjacent (if not outright antivax) “senior editor,” one responsible for a truly awful and deceptive “reanalysis” of the original Pfizer data from its phase 3 clinical trial used to win emergency use authorization (EUA) for its COVID-19 vaccine, his own misinterpretation of a Cleveland Clinic study, and another of his articles misinterpreting data. Basically, Kirsch is making the same rookie error (or intentionally misinterpretation) that antivaxxers used to do routinely during measles outbreaks when they would misleadingly claim that more vaccinated children got measles than unvaccinated. The error? When there are way more vaccinated people in a population than unvaccinated, even in a highly vaccinated population with an effective vaccine, it can end up that in terms of raw numbers there will be more cases of the disease among the vaccinated than the unvaccinated. What you have to look at is the attack rate, the risk of catching the disease, normalized to the number of individuals in the population who are unvaccinated versus those who are vaccinated.

Dr. Vincent Iannelli and one of our regular readers, epidemiologist René Najera, did just that and took a look at the Santa Clara data, to explain why Kirsch is making a rookie mistake. (Either that, or he is intentionally misleading. Take your pick.) Personally, I don’t think that Kirsch is smart or clever enough to intentionally mislead with statistics. He’s really that clueless and thinks his analysis is really that slam-dunk convincing. When Dr. Iannelli asks the question Do COVID Shots Increase Your Risk of Getting COVID?, the answer is, of course, no. The real explanation is two-fold. First, Dr. Iannelli:

If you look through the extensive data available in the Santa Clara County Health Department COVID dashboards, you will see that while most folks got the primary series, far fewer got the booster doses or updated COVID vaccine.

So if more vaccinated people got COVID, it was simply because there were more vaccinated people than unvaccinated people to get COVID – “vaccinated people” who were not necessarily up-to-date with the latest COVID vaccines.

Looking even more closely at the data from Santa Clara County though, despite how Kirschanalyzed” the data, you can easily see that unvaccinated people were more likely to get sick with COVID.

case rates Santa Clara County
The average case rates are highest in those who are unvaccinated in Santa Clara County – and everywhere else. Thanks to René F. Najera, MPH, DrPH for help creating this data visualization!

Same as it ever was. Kirsch keeps making rookie mistakes. Hell, not even rookie mistakes. Rookies, at least, generally have enough knowledge to be rookies in their field. Kirsch has less knowledge than that.

But what about the fun with Excel spreadsheets that I mentioned. For that, I have to give a massive hat tip to Dr. Najera, whose post A Second Look at Santa Clara’s COVID-19 Vaccination Data: Reevaluating Vaccine Risk Claims has the amusing blurb “Or why using spreadsheets to do an epidemiological and statistical analysis can cause misunderstandings.” I give Dr. Najera full credit because, although I immediately spotted the same error that Dr. Iannelli did, having dealt with antivaxxers making this error going back as long as I can remember, Dr. Najera spotted something that I didn’t, possibly because I’m not that great with Excel pivot tables myself:

Mr. Kirsch linked to the data in an Excel spreadsheet, and I downloaded it. The data contain 117,839 records of people diagnosed with COVID-19 in January 2022. The pivot table Mr. Kirsch shows on his post shows 83,104 records:

Kirsch vs pivot tables

He writes that the codes for the “NCOVPUIVaxVax” variable are “N” for unvaccinated, “Y” for vaccinated, “U” for unknown if they were vaccinated, and a blank for unknown if they were vaccinated. That doesn’t account for the missing data.

Here’s the thing about pivot tables in Excel… They don’t count blank cells. Take 117839 and subtract 83104, and you get 34735, which is the number of records with blanks.

Yes, he seems to have forgotten to include the blanks. Or maybe he did not understand his underlying data enough? Because, if I see my data set is almost 118-thousand records, I’m going to try to find out why I only have a “Grand Total” of about 83-thousand in my pivot table.

There’s another appropriate meme for this error:

Steve Kirsch is Homer Simpson.
“D’oh!” indeed.

Being an epidemiologist, Dr. Najera then did what epidemiologists do. He did a real analysis, unlike Mr. Kirsch’s uninformed and incredibly simplistic analysis. He fired up R Studio and went to work, because real epidemiologists use real statistical software. First, though, he used the Santa Clara open data portal to look at the data and generate a graph like the one that Dr. Iannelli cited and I reproduced above, only more detailed:

Santa Clara analysis
Notice how the unvaccinated were more likely to get COVID-19 in all age ranges? Mr. Kirsch didn’t. Notice, also, how COVID-19 cases are normalized to cases per 100,000 population, rather than expressed as raw case counts? That’s how you do it, but Mr. Kirsch is too clueless to realize it.

Having once again demonstrated a finding that we’ve been demonstrating to antivaxxers for at least two decades for every disease from measles to pertussis and beyond, namely that, regardless of raw numbers of cases, it’s the risk of disease that matters and that the unvaccinated are always at a higher risk of catching the disease vaccinated against, Dr. Najera continued, firing up R Studio to do analyses of the January 2022 data using three different scenarios:

To analyze his January 2022 data, I used R Studio and the following three assumptions:
  1. The unknowns are all vaccinated.
  2. The unknowns are all unvaccinated.
  3. The unknowns are vaccinated/unvaccinated in the proportion of vaccinated/unvaccinated in those for whom the data show their vaccine status.
Then I integrated those numbers into a table, like Mr. Kirsch attempted.

You can read his analysis himself for the details. I’ll just summarize his findings under these scenarios:

Under scenario #1, “the vaccinated are 12 times more likely to be in the dataset of cases than the unvaccinated. Again, this assumes that those for whom their vaccine status is unknown were all vaccinated. That is, the missing/unknown data is not at random.”

Under scenario #2, assuming that all the unknowns are unvaccinated and using the same denominator data, “we have 5.1% of cases vaccinated and 14% of cases unvaccinated. In this scenario, the vaccinated are about two-thirds less likely to be cases than those unvaccinated. Again, the assumption here is that the missing/unknown data is not at random.”

Under scenario #3, Dr. Najera knew that approximately 98% of the cases for whom a vaccine status is known are vaccinated. Using that knowledge, he then took the unknown/blank cases and randomly assigned them to the “vaccinated” or the “unvaccinated” category in a 98-to-2 proportions noting that in this case the “data are missing at random” and that we “make no assumption why the data are missing.” This analysis, of course, ends up coming up with a very similar number to scenario #1 because it is assumed that only 2% of the missing cases were unvaccinated.

There is another issue, of course. As Dr. Najera noted, missing data in a dataset are rarely missing at random, or, as he put it, “To have that many records missing vaccine status at random is highly improbable.” Usually there’s a reason, a mechanism, and you can’t know the reason and/or mechanism behind the missing data unless you have access to the detailed at a, as a county health department would.

Dr. Najera tends to be far less…Insolent…than I am and yet even he managed to get a bit spicy in the last part of his post:

A lot of work would have to be done to review the health records of over 34,000 cases to determine their true vaccination status. And that is exactly what the health department seems to have been doing, as shown in the rate of disease between vaccinated and unvaccinated in the open data portal. Month after month, the rate of disease in unvaccinated people has been consistently higher.

Finally, don’t forget his pivot table failed to include about 34-thousand cases. So throw out that table. My recommendation is to use a statistical analysis package like R or Python to do it. Or hire someone to do it for you.

Personally, when confronted with data like these, I would find an epidemiologist and statistician to help me with the data. I can do fairly basic analyses using commercial statistical packages like Prism or SPSS, but I’ve never had the time to learn R or Python. R, for instance, is a multiplatform language and environment for statistical computing and graphical data presentation. It has a high learning curve for someone like me without more than fairly basic biostatistics training, as I’ve discovered every time I’ve downloaded it and played around with it. That is, of course, why I rely on statisticians for my research and, where appropriate, epidemiologists. Unlike Mr. Kirsch, I know my limitations, kind of like Dirty Harry in Magnum Force, although I am nowhere near as tough:

Mr. Kirsch should take a cue from “Dirty” Harry Callahan.

OK, I’ll give Dr. Najera credit for this, too, which is almost Oracian:

Oh, and take some time to go to epidemiology and biostatistics school. Entire lectures are dedicated to what to do with missing data, and how to make the most informed decision when so much of it is missing. It may save you the embarrassment of not including 34,000+ cases in your analysis. I give out failing grades for less than that.

Unfortunately, Mr. Kirsch appears incapable of embarrassment, as soon became apparent on the hellscape formerly known as Twitter but now rebranded as X:

How amusing. Mr. Kirsch thinks that he is entitled to a “public recorded conversation” with an actual epidemiologist who saw his rookie mistake and corrected it publicly on his Medium blog.

So upset by Dr. Najera’s understandable lack of desire to interact with him was Mr. Kirsch that he edited his post to show that he clearly hasn’t learned a thing:

In fact, one of my critics (who I challenged to discuss his criticism with me), pointed out that at the time, the vax rate was a measly 86%! He wrote, “On January 31, 2022, the county reported 260,861 unvaccinated residents and 1,595,689 vaccinated residents.”

René F. Najera, MPH, DrPH blocked me on both X and Medium after I openly challenged his article and his analysis. These people who attack my work are like cockroaches… when you turn on the light, they run for cover instead of defending their work.

Seriously, Mr. Kirsch. You’re just digging yourself in deeper.

More likely, Dr. Najera knew from having been a longtime reader of this blog that it is utterly pointless to have a “public conversation” or “debate” with Mr. Kirsch, as observed here:

Indeed he is. As am I. Of course amusingly, Mr. Kirsch blocked me long ago, after which I just blocked him and was done with it. He’ll have to email me if he wants to whine about this post and how mean I am to him. He has on occasion done that. I either ignore him or respond briefly and sarcastically.

And here:


Seriously, the level of delusion exhibited by Mr. Kirsch in his interaction with Dr. Najera is hard to fathom. Instead of wondering whether he had made a massive mistake, which would have been appropriate after an actual expert described why your analysis was wrong, he lashed out, wanting a “public conversation” in which he can Gish gallop to his heart’s content. Remember, Mr. Kirsch is a man who only appears “reasonable” when teamed up with a crank like Denis Rancourt, who “doubts” that SARS-CoV-2, the coronavirus that causes COVID-19, even exists. In other words, the only way Mr. Kirsch appears even slightly non-delusional is when compared to virus deniers.

Again, Mr. Kirsch is man with no training in epidemiology and statistics, and it shows. He doesn’t even know how pivot tables work. In fairness, I’m not that great with pivot tables either, but I do know enough statistics and epidemiology to know that (1) it’s not the raw numbers of cases that matter, but the number of cases normalized to the proper denominator, and (2) it takes a lot of statistical knowledge and training to know how to properly deal with a large number of missing pieces of data.

Too bad Mr. Kirsch is persistently immune to Dirty Harry Callahan’s cautionary saying. Truly, a man’s got to know his limitations. (This applies to all genders as well, of course.) Quite simply, Mr. Kirsch does not know (or does not acknowledge) his very obvious limitations. As a result of this lack of self-awareness, he publicly embarrasses himself again and again and again, and then comes back for more because he is apparently incapable of embarrassment.

By Orac

Orac is the nom de blog of a humble surgeon/scientist who has an ego just big enough to delude himself that someone, somewhere might actually give a rodent's posterior about his copious verbal meanderings, but just barely small enough to admit to himself that few probably will. That surgeon is otherwise known as David Gorski.

That this particular surgeon has chosen his nom de blog based on a rather cranky and arrogant computer shaped like a clear box of blinking lights that he originally encountered when he became a fan of a 35 year old British SF television show whose special effects were renowned for their BBC/Doctor Who-style low budget look, but whose stories nonetheless resulted in some of the best, most innovative science fiction ever televised, should tell you nearly all that you need to know about Orac. (That, and the length of the preceding sentence.)

DISCLAIMER:: The various written meanderings here are the opinions of Orac and Orac alone, written on his own time. They should never be construed as representing the opinions of any other person or entity, especially Orac's cancer center, department of surgery, medical school, or university. Also note that Orac is nonpartisan; he is more than willing to criticize the statements of anyone, regardless of of political leanings, if that anyone advocates pseudoscience or quackery. Finally, medical commentary is not to be construed in any way as medical advice.

To contact Orac: [email protected]

35 replies on “Fun with Excel, or: Steve Kirsch is an antivax fool”

Maybe not when I was younger, in my late teens and early twenties, but definitely now… When someone points out something wrong in my data analysis or manuscripts, I do not do the following: four angry text messages to their personal phones at midnight on a Friday, one angry email challenging them to a broadcasted debate to their work email, numerous phone calls to their personal cell phone.
That is what I’ve been dealing with since I wrote that.
As a professional, and an adult with a fully developed prefrontal cortex, my reaction is not so much apparent anger that I harass the person critiquing me in their private space. I’ve had some “mean” comments on papers I’ve submitted for publication. I’ve had a professor call my dissertation work “incompetent” and “unnecessary.” Never did I think of challenging them to a debate, or contacting them in every way I could to call them cockroaches.
As a professional, I take the objective parts of their critique and improve my product or give them back feedback on why they may have misunderstood what I wrote. Most of all, if I make a mistake, I own it.
Being a software developer, Mr. K should know about data debugging and the importance of version control. You document what the mistake or bug was so others working after you can make something better.
As for offers of cash for debating… Good luck. I grew up in a dictatorship. I’ve seen what angry and powerful men can do to people who challenged them. And I did not back down. There is no existing amount of money that would convince me to a debate. Because my professional reputation is priceless.
It’s sad, though. With all that money and power, he could be remembered as a great philanthropist. Instead, all I’ll ever know him for is an inability to use spreadsheets. Something he now shares with my sharecropper ancestors. But at least my ancestors did their part to eradicate smallpox.

Thanks to Drs Ianelli and Najera.

Mr Kirsch “is a man with no training in epidemiology and statistics, and it shows” and he cherry picks**

I looked at his Wikipedia entry – he has a BS and MS in electrical engineering from MIT which does not exactly translate to med research BUT shouldn’t he AT LEAST know that if an event follows another event it doesn’t necessarily mean that it was CAUSED by it!: Isn’t that independent of number skills?

Kirsch gives his anti-vax position away immediately such as asserting that epidemiologists in Santa Clara County “knew” but “kept their mouths shut”, never warning the public or admitting their “error”. Usually, anti-vaxxers suggest that professionals in epi/ SBM are part of a deep rooted conspiracy that harms the general public whilst he certainly will enlighten and protect them.

AND shouldn’t missing data be a clue that the data are not as straightforward as he thinks?

** too obvious pun

he has a BS and MS in electrical engineering from MIT which does not exactly translate to med research

Having this background is a hobble in some cases. He’s so used to success that he can’t imagine himself as wrong about something. I’ve bumped into this with physicists more than I would like to believe –being really smart and really confident about that increases one’s capacity for sustaining and rationalizing cognitive dissonance.

Going back to my engineering education the sophomore statistics class was fairly weak. It did not cover much on sample size, etc. Later I took some classes at a community college, one was in statistics. It showed much of what I had missed (or forgotten) from the class I took decades before. One big thing was the importance was the quality of the data… like do not rely on self-selected surveys.

I don’t think he has the humility to admit he does not know what he is doing.

Although engineers-physicists may be especially prone to this syndrome, I think that there is also something fundamentally missing in these particular guys: kids eventually learn to self-evaluate their own skills throughout adolescence. Other developing skills include self-perception, person perception, metacognition, social cognition, self-programmed learning, abstraction etc.

Usually, they learn that they have strong and weak points- whether these are in physical skills or in school subjects AND they compare themselves to others. Adults who are not ‘compromised’ and who are reasonably well educated, should be somewhat realistic about what they understand/ can do and what they don’t understand/ can’t do.

Psychologists who study CT believers and anti-vaxxers found that certain personality factors influence belief: narcissistic and paranoid features as well as rejection of hierarchies of expertise. Know-it-all ism.

Kirsch sounds like he doesn’t ( as Orac says) know his limits. Is this because he succeeded as an engineer, solving complex problems in a certain way? Does he also imagine himself to be competent in areas beyond engineering and his current interest in vaccines/ viruses?

I venture that there is something wrong with his self-perception. If you see yourself as superior to experts despite lack of education and training, you are probably being unrealistic. Expertise is subtle and not achieved by reading papers for a few weeks or months.
Some other examples: Naomi Wolf – who studied poetry and wrote a few books misquoting. misunderstanding data. Also Jennifer Margulis
RFKjr- who studied law but pontificates about bio, physio and psych
All the rest of alt med/ anti-vax, woo-meisters, Igor.

I suppose we could ask if some of them are merely presenting scenarios for their enraptured fans and don’t believe their own sales pitch at all.

Mind set on things can overrule a great deal of education. I’m thinking of any schlafly, the guy who started conservapedia.

“ After graduating from Princeton, Schlafly briefly worked as a device physicist for Intel in Santa Clara, California until 1983, when he became a microelectronics engineer at the Johns Hopkins University Applied Physics Laboratory. later worked for Bell Labs before enrolling at Harvard Law School.”

Despite that he maintained for years in conservapedia that complex numbers were not needed for any real application and were, at their core, some sort of attack on culture by liberals.

Apparently kirsch never learned, or doesn’t care, that if you need good statistical analysis you don’t use Excel, or that using Excel means you won’t get good statistical analysis.

It’s also possible, as noted, that he doesn’t have any clue how to perform any statistical analysis at all. That seems to be common with the anti-science crowd.

It’s kind of fun to have providers complain about statistics; since they were the ones taking such liberties a short time ago.


No it’s not.

Oh, complaining about maths now? How fun.

No you dunce, complaining about your ability to DO maths.

You’d have to provide proof of your assertion JLB. Since you’ve repeatedly bailed on supporting your oft-stated “studies were designed to fail” comment none of us will expect that you will provide proof that “providers were taking liberties with statistics a short time ago” comment.

For one thing, you don’t seem to know any more about statistics than Igor, and he doesn’t know jack shit about statistics.

Do you know what a denominator is? And why it is important? Do you know what it means to normalize data?

Here is something I created to explain why more vaccinated get a vaccine protected disease than the unvaccinated. First vaccines are not perfect and there are simply more vaccinate folk than unvaccinated (because of the dreaded denominator):

Some community immunity arithmetic:

Take 1000 people (ignoring the infants under 2 months who cannot be vaccinated, or babies under a year who can only be partially vaccinated), if 5% refuse vaccines then the numbers are:

950 vaccinated persons (assuming full schedule)
50 unvaccinated persons

The pertussis vaccine is actually only 80% effective at worse, so the numbers are:

760 protected persons
190 vaccinated but vulnerable persons
50 unvaccinated persons

There is an outbreak and it gets spread to 20% of the population, then:

760 protected persons without pertussis

38 vaccinated persons get pertussis
152 vaccinated person who may still get pertussis

10 unvaccinated persons get pertussis
40 unvaccinated persons who may still get pertussis.

This is how more vaccinated persons get the disease than unvaccinated. Even if the infection rate was at 100%, there would still be more of the vaccinated getting the diseases because there are more of them!

I have a hard time believing this is the same person who developed the optical mouse as his own original concept

In another tweet Mr. Kirsch claimed he knew about the missing data and just assumed they were in proportion to the data he did have.

“ You don’t know what you are talking about Dorit.

His only point of any significance is that there is a high proportion of missing data in the ‘NCOVPUIVaxVax’ field. BFD. I knew that all along. No surprise.

This creates some uncertainty in the estimate, but the ‘usual’ method for dealing with this in epidemiology is to assume the missing data is in the same proportion as the known data which changes nothing.

If I draw 200 balls out of a bag, and the first 50 are red and 50 are blue, what do you think the stats on the other 100 will be?”

(I answered there).

In another tweet Mr. Kirsch claimed he knew about the missing data and just assumed they were in proportion to the data he did have.

In other words, René’s Scenario #3. So Kirsch is still wrong.

As long as Kirsch fails to see the (colossal) error of his ways, I suggest that any comment, message or challenge to ‘debate’ coming from him is always answered with the following simple map: You are Here

R, for instance, is a multiplatform language and environment for statistical computing and graphical data presentation. It has a high learning curve for someone like me without more than fairly basic biostatistics training,

As a long time, if not terribly competent, user of R I think I can assure you that R is totally “New user hostile”. Especially if you are an SPSS or SAS user. I swear I thought my brain was twisting in my skull when I did one or two R things that I now think are totally normal.

R’s major strength is that there are libraries that supply framework for pretty much anything you want to do. A down side is that there are many that do essentially the same thing, have names that don’t indicate what they do, and are just enough different that determining which to choose and use as it best suits a particular need is quite challenging.

All in all though, I much prefer teaching with it over SPSS.

Fun fact (which you may already know: if so, sorry for putting it here): Every new release of R has a name that comes from a Peanuts cartoon or movies. Release 2.14.0 in 2011 was Great Pumpkin.

To me I think of R as fundamentally different from Prism or JMP (subset of SAS) or any of the other stats programs because when I learned R it was taught as a coding language, rather than a “program”.

So I approached it like I approached learning Java or Python, rather than how I approached learning MiniTab.

If I had a reason and a lot of spare time I’d learn R, but I’m happy with JMP (and Matlab can just go away and never, ever come back).

“when I learned R it was taught as a coding language, rather than a “program”.”

That’s sort of the way we do it:
– you know what you want to do
– you find a library that has the tools to use
– it’s up to you to write the code to use those tools to get the analysis and format everything

One of the great things about the advancements in availability of computing power and software is that those things have opened up the way for more people to do statistical analysis.

That ability is also one of the downsides of the advancements in availability of computing power and software.

Given the heinous crimes against statistics I’ve seen people do with Excel (standard error of a single point should just generate an error, not a value!), making stats programs a little bit difficult to use isn’t a bad thing in my book.

(And no, you can’t use a student’s T test to say two things are the same!)

I mean – Excel does count blank fields, IF tell the program to do so somehow (“Null,” or “N/A” or “blank”, whatever you want).

One wonders why Kirsch bothers with faux epidemiology/ anti-vax in the first place!
He has a REAL lot of money : he can do things, go places, start new businesses or foundations. So much more is open to him than even what most people could imagine.
Could it instead be easily accessed adulation in the comments section? SRSLY
A saviour complex? A need to be seen as not purely self-serving ( though he is)?

-btw- he supports RFKjr. Birds of a feather. Save the world from vaccines and toxic chemicals and elevate yourself in the process.

There are real health threats and lacks of education in both the developed and developing world:
food, meds and schools cost money. Wake up, Steve. Kids, especially girls, need your help all over the globe: their eventual actual contributions will far outweigh any momentary buzz you get from your devoted followers.

“He’s not the Messiah.”
– Life of Brian

Seriously, a case of audience capture.

On a happier note, I just got the Novavax covid vaccine. Minimal arm pain. Collect them all!

That’s the conference that featured a masquerade ball (cranks disguised as scientists?) and the premiere of Plandemic: The Musical.*

The activities of Kirsch and like-minded folk have always been heavily geared towards making money. The purported free speech angle is about ensuring a narrative of persecution and panic that will let them harvest more $$$ from the gullible. It’s similar to the scam that alties pulled off some years back when warning that mysterious foreign regulatory bodies were about to ban all supplements (curiously, no such thing occurred).

It’s totally unsurprising that the Vegas crowd would willingly associate with the worst elements in society. Any allies that support the cause and enable you to suck in cash are acceptable.

*I heard that Plandemic: The Musical had some catchy tunes.

“There’s no business like grift business like no business I know
Ripping off the marks is so appealing
Everything the traffic will allow
Nowhere could you get that happy feeling
when you are stealing their dough”

I remember the scam that all supplements were about to be banned (I’d support REGULATION of them, and a few deserve banning, at least). The guy who first told me about them suddenly died of a heart attack one day (years before 2020). Didn’t trust doctors and didn’t get checked out. RIP.

No comment on Steve’s analysis.

However, Santa Clara County did indeed engage in some hanky-panky with data and produced fake “cases by vaccination status” charts. I wrote two posts about it. The short of the story is that SCC understated the percentage of unvaccinated people. It was very easy math to do requiring zero statistical tools.

All that was needed is to calculate implied percentage of vaccinated people based on

1) Total case rate
2) Vaccinated Case rate
3) Unvaccinated case rate

(this is similar to algebra “mixture word problems”)

The resulting percentage of unvaccinated people was absurdly low and contradicted even other reports by Santa Clara County.

This approach (undercounting the unvaccinated and counting some vaccinated cases as unvaccinated) was also used by the CDC to produce absurd charts about how the unvaccinated people are so many more times likely to suffer various negative outcomes.

CDC itself conducted so called VISION network study that used correct vaccination status known from health providers. That yielded a very modest severed disease protection from Covid shots, completely contrasting the fake calculations based on undercounting the unvaccinated.

re Kirsch, Wolf, Kennedy et al

Although they present as investigative reporters uncovering hidden ‘truths’, the only fact they reveal is their own lack of expertise in the fields they discuss:
vaccines cause autism, illness/ “turbo cancer”; aids is not caused by a virus but malnutrition etc.
They are truly abysmal investigators because they miss mountains of evidence acquired over decades in diverse areas of inquiry that show
— autism develops largely pre-natally
— vaccines are safer than the illnesses they prevent/ diminish
— cancer takes longer to develop even after atomic bombs or reactor meltdowns
— aids is caused by hiv

So what does that tell us about these incredible detectives?
— they don’t know what they’re talking about
— they don’t realise their error
— they are misleading their followers
— they may be lying to gain attention/ followers/ money
— they steer followers away from more realistic sources

Why would someone pontificate about scientific material they barely understand?**

I toy with the idea of an alternate theory of tectonic plates: they don’t exist at all but were invented by insurance companies- they knew that artists and rich people enjoy dramatic landscapes so they spread the myth that these are fault zones are where earthquakes are common so they can raise policy rates on buildings there. This happens in California and Japan thus, an international cartel of insurance companies!
Prove me wrong.

** unlike the aforementioned I actually studied geology . Tectonic plates are real.

“They are truly abysmal investigators because they miss mountains of evidence acquired over decades in diverse areas of inquiry…”

I think “miss” should be replaced by “intentionally ignore”

“they may be lying…”

Replace “may be” by “are”?

Just as much of the media gave and is giving trump too much of a break by failing to mention everything he’s done and says he will do, saying these anti-vaxx folks MAY be doing X, Y, or Z out of simply not understanding things doesn’t cut it. They need to be called out for what they are and what they’re doing: lying to draw attention, and money, to themselves, without caring at all that what they say would, for anyone who follows their advice, put those people at risk.

I think the vast majority of them are lying but I’m not sure that they know it themselves: is it lying, delusion or purely ignorance?

However, we can point out how they misinform: what they say, how they distort, what they leave out, what they achieve ( money, fame, attention) which informs and arms their audiences against them. In certain cases, we can even count their earnings or show their estates.

[…] Iowa Fertilizer Spill Kills Nearly All Fish Across 60-Mile Stretch of Rivers. Officials in Iowa and Missouri estimated that nearly 800,000 fish had died in waters that flow into the Missouri River. (gift link) Erratic Fish Behavior In Florida Prompts ‘Emergency Response’ From National Agency The Race to Reinvent CPR (gift link) Do COVID Shots Increase Your Risk of Getting COVID? Multispecies interactions and the community context of the evolution of virulence Fun with Excel, or: Steve Kirsch is an antivax fool […]

Want to respond to Orac? Here's your chance. Leave a reply! Just make sure that you've read the Comment Policy (link located in the main menu in the upper right hand corner of the page) first if you're new here!

This site uses Akismet to reduce spam. Learn how your comment data is processed.


Subscribe now to keep reading and get access to the full archive.

Continue reading