How epidemiologists try to fool us with flawed statistical practices

Reposted from Dr. Judith Curry’s Climate Etc.

Posted on May 17, 2021 by curryja 

by S. Stanley Young and Warren Kindzierski

Climate Etc. recently carried several insightful posts about How we fool ourselves. One of the posts – Part II: Scientific consensus building – was right on the money given our experience! The post pointed out that… ‘researcher degrees of freedom’… allows for researchers to extract statistical significance or other meaningful information out of almost any data set. Along similar lines, we offer some thoughts on how others try to fool us using statistics (aka how to lie with statistics); others being epidemiologists and government bureaucrats. 

We have just completed a study for the National Association of Scholars [1] that took a deep dive looking at flawed statistical practices used in the field of environmental epidemiology. The study focused on air quality−health effect claims; more specifically PM2.5−health effect claims. However, the flawed practices apply to all aspects of risk factor−chronic disease research. The study also looked at how government bureaucrats use these claims to skew policy in favor of PM2.5 regulation and their own positions. 

All that we discuss below is drawn from our study. Americans need to be aware that current statistical practices being used at the EPA for setting policy and regulations are flawed and obviously expensive. Viewers can download and read our study to decide the extent of the problem for themselves.

1. Introduction

Unbeknownst to the public and far too many academic scientists, modern science suffers from an irreproducibility crisis in a wide range of disciplines—from medicine to social psychology. Far too frequently scientists are unable reproduce claims made in research.

Given the irreproducible science crisis, we completed a study for the National Association of Scholars (NAS) in New York as part of the Shifting Sands project. The project—Shifting Sands: Unsound Science and Unsafe Regulation—examines how irreproducible science negatively affects select areas of government policy and regulation in different federal agencies.

Our study investigated portions of research in the field of epidemiology used for US Environmental Protection Agency (EPA) regulation of PM2.5. This research claims that particulate matter smaller than 2.5 microns (PM2.5) in outdoor air is harmful to humans in many ways. But is the research on PM2.5 and the claims made in the research misleading?

2. Bias in academic research

Academic researcher incentives reward exciting research with new positive (significant association) claims—but not reproducible research. This encourages epidemiologists – who are mainly academics – to wittingly or negligently use various flawed statistical practices to produce positive, but (we show) likely false, claims. 

There are numerous key biases that epidemiologists continue to unintentionally (or intentionally) ignore in studies of air quality and health effects. This is done to make positive, but likely false, research claims. Some examples are:

  • multiple testing and multiple modeling
  • omitting predictors and confounders
  • not controlling for residual confounding
  • neglecting interactions among variables
  • not properly testing model assumptions
  • neglecting exposure uncertainties
  • making unjustified interventional causal interpretation of regression coefficients

Our study focused on the multiple testing and multiple modeling bias to assess whether a body of research has been affected by flawed statistical practices. We subjected research claiming that PM2.5 is harmful to a series of simple but severe statistical tests. 

3. How epidemiologists skew research

Our study found strong circumstantial evidence that claims made about PM2.5 causing mortality, heart attacks and asthma are compromised by flawed statistical practices. These flawed practices make the research untrustworthy as it favors producing false claims that would not reproduce if done properly. This is discussed further below.

Estimating the number of statistical tests in a study – There is known flexibility available to epidemiology researchers to undertake a range of statistical tests and use different statistical models on observational data sets. The researchers then can select, use and report (cherry pick) a portion of the test and model results that favor a narrative.

One form of simple but severe testing we employed was counting. Specifically, we estimated the number of statistical hypothesis tests conducted in 70 different published epidemiology studies that make PM2.5−health effect claims. These results are presented in our study. The counting procedures are straightforward, and readers can learn and use them to count statistical tests in published observational studies. In our case, the median number of statistical tests performed in these 70 studies was over 13,000.

Epidemiologists typically use a Relative Risk (RR) or Odds Ratio (OR) lower confidence limit > 1 (or a p-value < 0.05) as decision criteria to justify a significant PM2.5−health effect claim in a statistical test. However, for any given number of statistical tests performed on the same set of data set, 5% are expected to yield a significant, but false result. A study with 13,000 statistical tests could have as many as 0.05 x 13,000 = 650 significant, but false results!

Given advanced statistical software, epidemiologists today can easily perform this many or more statistical tests on a set of data in an observational study. They can then cherry pick 10 or 20 of their most interesting findings and write up a nice, tight research paper around these findings—which are most likely to be false, irreproducible findings. We have yet to see an air quality−health effects study that reports as many as 650 results. How exactly is one supposed to tell the difference between a false positive or a possible true positive result when so many tests are performed and so few results are presented?

Diagnosing evidence of publication bias, p-hacking and/or HARKing – Publication bias is the failure to publish the results of a study unless they are positive results that show significant associations. P-hacking is reanalyzing data in many different ways to yield a target result. HARKing (Hypothesizing After Results are Known) is using the data to generate a hypothesis and pretend the hypothesis was stated first.

It is traditional in epidemiology to use confidence intervals instead of p-values from a hypothesis test to demonstrate statistical significance. As both confidence intervals and p-values are constructed from the same data, they are interchangeable, and one can be calculated from the other. 

We first calculated p-values from confidence intervals for data from meta-analysis studies that make PM2.5−health effect claims. A meta-analysis is a systematic procedure for statistically combining data from multiple studies that address a common research question—for example, whether PM2.5 is a likely cause of a specific health effect (e.g., mortality). We looked at meta-analysis studies claiming that PM2.5 causes: i) mortality, ii) heart attacks and iii) asthma. 

We then used a simple but novel statistical method—p-value plotting—as a severe test to diagnose evidence of publication bias, p-hacking and/or HARKing in this data. More specifically, after calculating p-values from confidence intervals we then plotted the distribution of rank ordered p-values (a p-value plot). 

Conceptually, a p-value plot allows us to examine a specific premise that factor A causes outcome B using data combined from multiple observational studies in meta-analysis. What should a p-value plot of the data look like?

  • a plot that forms an approximate 45-degree line provides evidence of randomness—supporting the null hypothesis of no significant association between factor A & outcome B (Figure 1)
  • a plot that forms approximately a line with slope < 1, where most of the p-values are small (less than 0.05), provides evidence for a real effect—supporting a statistically significant association between factor A & outcome B (Figure 2)
  • a plot that exhibits bilinearity—that divides into two lines—provides evidence of publication bias, p-hacking and/or HARKing (Figure 3)

Figure 1.  P-value plot of a meta-analysis of observational data sets analyzing associations between elderly long-term exercise training (factor A) and mortality & morbidity (injury) (outcome B); data points drawn from 40 observational studies.

Figure 2.  P-value plot of a meta-analysis of observational data sets analyzing associations between smoking (factor A) and squamous cell carcinoma of the lungs (outcome B); data points drawn from 102 observational studies.

Figure 3.  P-value plot of a meta-analysis of observational data sets analyzing associations between PM2.5 (factor A) and all−cause mortality (outcome B); data points drawn from 29 observational studies.

We show over a dozen p-value plots in our study for meta-analysis data of associations between PM2.5 (and other air quality components) and mortality, heart attacks and asthma. All these plots exhibit bilinearity!

This provides compelling circumstantial evidence that the literature on PM2.5 (and other air quality components)—specifically for mortality, heart attack and asthma claims—has been affected by statistical practices that have rendered the underlying research untrustworthy.

Our findings are consistent with the general claim that false-positive results from publication bias, p-hacking and/or HARKing are common features of the medical science literature today, including the broad range of risk factor−chronic disease research.

4. How government bureaucrats skew policy

The process is further derailed with government involvement. The EPA have relied on statistical analyses to show significant PM2.5−health effect associations. EPA bureaucrats who fund this type of research depend on regulations to support their existence. The EPA has slowly imposed increasingly restrictive regulation over the past 40 years. 

However, the EPA appears to have acted selectively in its approach to the health effects of PM2.5. This has been done by paying more attention to research that supports regulation (i.e., shows significant PM2.5−health effect associations) and ignoring or downplaying research that shows no significant PM2.5−health effect associations. This latter research exists, it is simply ignored or downplayed by the bureaucrats! Nor are the researchers finding negative results funded.

It is apparent to us that bureaucrats lack an understanding of, or willfully ignore, flawed statistical practices and other biases identified above in PM2.5−health effects research. They, along with environmental activists, continuously push for tighter air quality regulation based on flawed practices and false findings.

5. Can this mess be fixed?

Epidemiologists and government bureaucrats collectively skew results of medical science towards justifying regulation of PM2.5, while almost always keeping their data sets private. Far too many of these types, and a distressingly large amount of the public, believe that academic (university) science is superior to industry science. However, as epidemiology evidence is largely based on university research, we should treat it with the same skepticism as we would industry research.

Mainstream media appear clueless and uninterested in glaring biases in epidemiology research that cause false findings—flawed statistical practices, analysis manipulation, cherry picking results, selective reporting, broken peer review.

Epidemiologists, and government bureaucrats who depend on their work to justify PM2.5 regulation, proceed with far too much self-confidence. They have an insufficient sense of the need for awareness of just how much statistics must remain an exercise in measuring uncertainty rather than establishing certainty. This mess plagues government policy by providing a false level of certainty to a body of research that justifies PM2.5 regulation.

In our study we make several recommendations to the Biden administration for fixing this mess. However, we do not hold our breath that they will be considered. Some of these include:

  • the administration needs to support statistically sound and reproducible science
  • unsound statistical practices silently supported by the EPA need to stop
  • the building and analysis of data sets should be separately funded
  • these data sets should be made available for public scrutiny

Most importantly, Americans need to be aware that current statistical practices being used at the EPA for setting policy and regulations are flawed and obviously expensive.

S. Stanley Young (genetree@bellsouth.net) is the CEO of CGStat in Raleigh, North Carolina and is the Director of the National Association of Scholars’ Shifting Sands Project. Warren Kindzierski (warrenk@ualberta.ca) is an Adjunct Professor in the School of Public Health at the University of Alberta in Edmonton, Alberta.

[1] Young SS, Kindzierski W, Randall D. 2021. Shifting Sands: Unsound Science and Unsafe Regulation. Keeping Count of Government Science: P-Value Plotting, P-Hacking, and PM2.5 Regulation. National Association of Scholars, New York, NY. https://www.nas.org/reports/shifting-sands

4.9 36 votes
Article Rating
84 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
ResourceGuy
May 18, 2021 6:08 am

And the Asthma Cherry Picker Award goes to EPA and its political handlers–who else?

Joel O’Bryan
Reply to  ResourceGuy
May 18, 2021 8:32 am

Over the past 50 years, US Asthma rates have been steadily increasing while all air quality measures have improved. Without using fancy statistics, that is the strongest piece of evidence of no linkage between the two. Something is driving the rates rates upwards. But that doesn’t stop EPA fraud on pm2.5 because power and funding is on the line.

Joseph Zorzin
Reply to  Joel O’Bryan
May 18, 2021 1:25 pm

Lots of pollutants in family homes- cigarette smoke and dirty homes.

Bob boder
Reply to  Joseph Zorzin
May 18, 2021 3:33 pm

Or is it people spending all there time in their house and never going outside.

anthropic
Reply to  Joseph Zorzin
May 18, 2021 9:38 pm

Smoking, inside or out, has declined. A lot. And study after study has shown that children exposed to dirt do better health wise than those who are ensconced in very clean environments.

Bill Treuren
Reply to  Joel O’Bryan
May 18, 2021 8:40 pm

Absolutely Joel. I find it amazing here in NZ people tell me that 60,000 people die a year from car pollution and without a blink ask where the millions who died when pollution was dozens of times higher in the past. Yes, silence.

Sara
Reply to  ResourceGuy
May 18, 2021 12:46 pm

If air quality has measurably improved but asthma rates have risen, then the entire thing is skewed to get the grant money and go on to the next bit of hocus-pocus magic numbers that are also totally bogus.
If asthma rates are up, and asthma is a chronic condition, but air quality has improved at the same time, plainly, the study is bogus, twerked toward justifying more grant money for more studies.

Meisha
Reply to  Sara
May 19, 2021 5:50 am

Asthma is an immune system disregulation. Could it be the cleaner our homes/ environment are while we are in utero and in the first few years of life has resulted in immune systems not developing properly because they have not been exposed to irritants at a time when they are “learning” to deal with them?

Or, could it be in decades past far fewer children with asthma would not survive to reproductive age because any disease with pulmonary implications (now prevented with vaccinations) would be more fatal to them?

2hotel9
May 18, 2021 6:21 am

“epidemiologists” are no longer scientists or doctors, they are simply political tools of the Democrat Party and all the other leftist America haters. Screw them all. Stop lying and do real medical science, scumbags.

Doug Huffman
Reply to  2hotel9
May 19, 2021 4:26 am

Medical-science is an oxymoron.

Medicine is an art and a technology validated by repetition and statistical proof. Science is falsifiable, falsifiability the boundary demarcation of science from nonsence non-science. See Karl Popper’s The Logic of Scientific Discovery. See E.T. Jaynes’ Probability Theory: The Logic of Science.

Learn Bayesian inference, the bugaboo of MDs trying to pass as epidemiologists.

2hotel9
Reply to  Doug Huffman
May 19, 2021 7:05 am

All that said they are, as a lot, liars spewing politically motivated lies whilst claiming to be scientists.

Doug Huffman
Reply to  2hotel9
May 19, 2021 12:22 pm

Self-assigned epithet ‘scientist’ is sufficient impeachment.

jim hogg
Reply to  2hotel9
May 19, 2021 7:07 am

Yeah. Definitely. Lets do real science that skews the stats to suit the prejudices of the NAS, which will clearly be a better form of skewing than that of the left. The irony just flies over the protesting right wing heads as much as does the delusion of objectivity that plagues and deceives the hubristic left. Why not wake up and recognise that we all see the world through the prism of our prejudices, and that we skew the data and interpretation thereof accordingly. From that honest and yes, very unwelcome, position we could then strive – and it’ll take a great deal of determination and brutal honesty to fully swallow acceptance of the bitter pill of our own entrenched and incredibly resilient biases – for objectivity with accuracy in all things as our irreducible primary. I’m not holding my breath, I’m running out of time, and have even less optimism.

Tom Halla
May 18, 2021 6:22 am

This is an interesting tool to detect cherry picked results.
Notably, the Trump administration tried to ban “secret science”, and drew fire from the activist community.
Where this sort of malfeasance occurred earlier was the dispute over the health effects of power lines. The Swedes ran a study corelating the health effects of proximity to power lines and certain diseases, using several hundred districts. Of course, they found several districts with a “significant” correlation, as they acted as if they had separate studies.

Joel O’Bryan
Reply to  Tom Halla
May 18, 2021 8:37 am

Enemy fire is always strongest when the bombers are nearest an important target. And being able to employ “secret science” is a very important tool on the path to power for Dems.

Caligula Jones
May 18, 2021 6:44 am

Hope they go after the lunatics calling themselves “pandemic epidemiologists” next, particularly here in Ontario, Canada.

They can start with that fraud who turned a career in food science into a media celebrity sensation…

dk_
Reply to  Caligula Jones
May 18, 2021 11:15 am

“…fraud…into a media celebrity…”
C. Jones, you may have to be a little more specific. There’s a host of the type. While “string ’em all up, we’ll worry about the charges in the morning” seems quite attractive, it is bound to get us noticed.

Sunny
May 18, 2021 6:53 am

Goodness me this is a bbc article, We need the truth be told, as the greens abd climate scam club will ruin our way of life..

https://www.google.com/amp/s/www.bbc.co.uk/news/science-environment-57149059.amp

Richard Page
Reply to  Sunny
May 18, 2021 7:32 am

For ‘ruin our way of life’ read ‘lead to the death’s of thousands every year’.

dk_
Reply to  Richard Page
May 18, 2021 11:18 am

Richard and Sunny — good spotting those “tags.” Those phrases are among those that are used to make an hysterical threat seem real. Meaningless words by themselves, cribbed from other contexts. A huge propaganda tool, pulled out of Animal Farm and the 50’s Red Scare.

Sara
Reply to  Sunny
May 18, 2021 12:51 pm

Gee, isn’t that sweet? They really do want to return to the late 18th – early 19th century, where you cooked food in the fireplace or over an open stove instead of a closed stove and had no temperature controls on anything.
Perhaps we should return to wearing corsets and hoops or bustles, too? And how about we start shipping goods by sail instead of by freighters, which run on petroleum derivatives?
I think I could handle that stuff, but then, I don’t take modern technology for granted. It is far too ephemeral for me.

dk_
Reply to  Sara
May 18, 2021 2:14 pm

Sara
“return to wearing corsets and hoops or bustles”
Did you ever notice how many historical and dystopian cosplay hobbyists are also eco warriors?
Did you ever think on the potential for survival any of either sort would have in the wilderness? Or in 1950’s rural U.S.A?

May 18, 2021 6:54 am

Yes, we are fooled, day over day.
They tell us we should not have any hope for an end of lockdowns, because there is an Indian mutant, worse then all we have had ’til today.
One the other hand, we are drifting in a two class society, vaccinated and the other.
The vaccinated get privileges, where the “privlege” is the normal life.

They “know”, a natural immunity doesn’t exist over more than 6 moth. An antibody test doesn’t give you the status of a healed person, you are only healed aftter positive PCR test and following quarantine.
We have no compulsory vaccination in Germany per law, but looking after facts, we have.

dk_
Reply to  Krishna Gans
May 18, 2021 11:22 am

Krishna G. To be in power, it is good to become or be seen as the one who can grant privelege to those who are worthy, and deny everything to those who are not. Again, Orwell’s Animal Farm — who is more equal?

Reply to  Krishna Gans
May 19, 2021 6:00 am

The Covid-19 lockdown fraud is deep and sinister – a Great Reset power play:
 
The following credible paper states that USA deaths attributed to Covid-19 were 16 times higher than actual. Covid-19 was a false crisis.
 
I published a similar comment many months ago in November 2020, by comparing the per capita death rate attributed to COVID-19 in Alberta, versus the reported per capita death rate in the United States, which was about 10 times higher. 
 
Covid death statistics in the USA are false and fraudulent – not ~500,000 Covid deaths, but 30,000 to 50,000 – similar to a typical flu year in the USA.
 
Furthermore, in my recent paper, excerpted below, note that there is no “death bump” to mid-2020 in either Alberta or Canada – that is, no significant increase in total deaths from all causes over the trend of the previous seven years.
 
That means there was no significant deadly Covid-19 “epidemic” in Alberta or Canada to mid-2020, and no justification for the panic, the lockdowns of the workforce and students, and the destruction of our economy – just as I correctly published more than one year ago, on 21&22 March 2020:
 
21March2020 – Allan MacRae
LET’S CONSIDER AN ALTERNATIVE APPROACH: Isolate people over sixty-five and those with poor immune systems and return to business-as-usual for people under sixty-five. This will allow “herd immunity” to develop much sooner and older people will thus be more protected AND THE ECONOMY WON’T CRASH.
 
22March2020 – Allan MacRae
This full-lockdown scenario is especially hurting service sector businesses and their minimum-wage employees – young people are telling me they are “financially under the bus”. The young are being destroyed to protect us over-65’s. A far better solution is to get them back to work and let us oldies keep our distance, and get “herd immunity” established ASAP – in months not years. Then we will all be safe again.…
 
All we really needed to do was over-protect the very elderly and infirm – the high-risk population – which we failed to adequately do. What a debacle!
_______________________
 
FAULTY COVID DEATH NUMBERS EXPLAINED
The American Thinker, April 9, 2021
By Dennis McGowan
 
For many of us who have had ties to the scientific standards and procedures connected to recording fatalities there has been a serious sense of doubt about the numbers of Covid 19 deaths reported by the media, courtesy of the Centers for Disease Control and Prevention. The numbers have appeared to be somewhere between marginally overstated and grossly exaggerated. Finally, these instincts have been supported by a peer-reviewed scientific paper.
On October 12 of last year, a 25-page paper in Science, Public Health Policy and The Law was released that explained, in detail, the foundational reason for the publicly announced fatality numbers and the mechanism by which they were derived. This paper, authored by ten members of the scientific community is titled “Covid-19 Data Collection, Comorbidity & Federal Law: A Historical Perspective.” At the heart of the issue is the CDC and its methods for collecting and reporting the data, a model which was changed radically in the face of the current crisis.
In 2003, the CDC authored and released guidance documents used by the forensic community titled Medical Examiners’ and Coroners’ Handbook on Death Registration and Fetal Death Reporting along with “Physicians’ Handbook on Medical Certification of Death.”  These have been the standard for the certification of fatalities, nationwide, for seventeen years. However, in March of 2020 things changed.
The National Center for Health Statistics released Covid-19 Alert No. 2 which changed the way deaths with connections to Covid 19 were reported and tabulated. The revealing line in the alert is in the last paragraph: “Covid-19 should be reported on the death certificates for all decedents where the disease caused or is assumed to have caused or contributed to death.” [emphasis added] This changed the parameters for the inclusion of deaths from Covid, raising the numbers substantially.
A table in the October study titled “Comparison of Total Covid-19 Fatalities Based Upon Different Reporting Guidelines” demonstrated that deaths through August 23rd of 2020 were higher by over 16 times as compared to the traditional definition. If the reporting of these deaths followed the CDC guidebook from 2003, the number of Covid deaths would have been 9,684. However, utilizing this new reporting and classification method that exclusively applied to Covid-19, the number of deaths is 161,392.
file:///C:\Users\Owner\AppData\Local\Temp\msohtmlclip1\01\clip_image001.gifThe paper delves into a variety of other topics, some legal and some statistical, that are all intrinsically functions of the change in death reporting parameters initiated by the March 2020 alert. However, for so many of us who have had nagging doubts about the actual numbers, knowing that the calculus for recording these deaths had been replaced is reassuring. Having spent a year looking at Covid fatality numbers and assuming that the true count was more likely half or a third of what was being reported, this new report is both satisfying and startling. None of us would have guessed that the actual disparity would be that the number of Covid deaths, according to this study, is a bit less than 6% of the numbers reported by the media.
________________________________
 
NO “DEATH BUMP” MEANS NO REAL DEADLY COVID-19 PANDEMIC.
My recent paper:
In 2020 there were NO excess deaths in Alberta or Canada – no total death bump means no deadly pandemic.
Average age of Covid-19 deaths in Alberta was 82 – four years longer than average life expectancy of 78.
 
WHERE IS THE COVID-19 PANDEMIC? WHY THE FULL LOCKDOWN?
ANNUAL TOTAL DEATHS IN ALBERTA AND CANADA SHOW NOTHING UNUSUAL TO 30June2020
by Allan MacRae, B.A.Sc., M.Eng., April 3, 2021

From the total deaths plotted for Alberta and Canada, there were NO significant excess total deaths to 30June2020, and so there was NO justification for the incredibly costly Covid-19 lockdown, which is estimated to have caused 10 to 100 times more current and future harm than the Covid-19 illness. 
 
The important question is why the lockdowns were ever enacted, and why the tried-and-tested Alberta Emergency Management Plan was tossed out and a young medical officer given the impossible task of managing the alleged pandemic – that in the first half of 2020 was a hugely exaggerated, false crisis.
 
In addition to needlessly destroying the economy, the impacts of the lockdown continue to cause harm:
– Hospitals were emptied for ~2 months to make room for a “tsunami of Covid-19 cases” THAT NEVER HAPPENED;
– Cancer tests, surgeries and other needed medical procedures were delayed and backlogged;
– Deaths from opioid overdoses more than doubled, resulting in an increase of tens of thousands of years-of-life-lost;
– Societal problems including substance abuse, family violence, poverty, and mental illness all increased;
– The education system was disrupted and the harm to students will continue for years.
 
There is ample evidence that lockdowns and masking had little impact on Covid-19 mortality. Sweden and South Dakota that did not lock down had similar or better mortality outcomes to those that did.
 
In fact, the lockdowns encouraged the longevity of the Covid-19 illness and the development of more deadly variants. Most flu’s die out in the summer season, but lockdowns and masking allowed Covid-19 to survive through the summer. The Covid-19 problems since 30June2020 with renewed contagion and more dangerous variants were enhanced by the lockdowns – the lockdowns were a total debacle.
 
Furthermore, the incessant testing and reporting of positive PCR tests as “cases” is harmful nonsense, needlessly creating fear and bad policy. A “case” exists NOT from a positive test, which is often asymptomatic, but from a real illness that requires treatment.
 
I correctly concluded that the Covid-19 lockdowns were a huge error in early March 2020, and published that conclusion on 21&22March2020. All we needed to do, which I published at that time, was over-protect the very elderly and infirm (which we failed to do) and get everyone else back to work and school. The data shows that the lockdowns were not justified and were hugely harmful.
 
Alberta and Canada are now in a far worse situation than if we had done nothing – no lockdowns, no masking, etc. How do we get out of the needless mess our governments have created? First, stop reacting in panic to overblown test results and other scary propaganda. Adults should take 4000IU of Vitamin D3 daily. Cease all lockdowns, masking and distancing measures now. Get everyone back to work and school.
 
Let’s walk out into the sunshine and get back to enjoying our lives.

Last edited 2 months ago by ALLAN MACRAE
dk_
Reply to  ALLAN MACRAE
May 19, 2021 12:17 pm

Off topic, and mostly unreadable and little to do with the site. Recommend you approach WUWT for publishing a separate piece. You seem to link to commercially suspicious sites and URLs to me. And you seem to post randomly. I don’t find it useful.

Reply to  dk_
May 19, 2021 8:48 pm

Listen dk:

The thread is about flawed statistical practices used by epidemiologists.

I have pointed out perhaps the two most significant statistical scams in the Covid-19 lockdown, the greatest and most costly fraud in history.

  1. Covid deaths in the USA were over-estimated by ~16 times, based on a March 2020 re-definition of how deaths are coded.by The National Center for Health Statistics in Covid-19 Alert No. 2. Be very afraid!
  2. To mid-year 2020 there was no total death bump (increase) vs past years in both Alberta and Canada. Hence no deadly pandemic. More false alarm, possibly similar also in the USA.
  3. The above two observations point to a global-scale lockdown fraud, driven by political objectives, not health concerns..

It is possible that New York State showed a total death bump because of the “Cuomo Cullings” of the elderly and infirm.
______________________________

Was that too complicated for you?

dk_
Reply to  ALLAN MACRAE
May 21, 2021 11:36 am

Alan,
I did apologise on another posting in another thread, probably to blame on my method of reviewing comments on comments on WUWT. I will also take any blame for the way and order in which I encountered your replies. But I won’t repeat the apology.

I can understand this second post. It is not too complicated for me. Thank you.

The original post, in this thread is still not readable. Still reflects work not your own. Still is nearly the length of the original article at the heading. And contains browser links that do not go anywhere. It seemed then and upon re-reading now, quite angry and almost unintelligible.

Your second post did explain it, and make it more understandable, and I will accept as deserved the two intentional or implied insults, I probably earned them.

I still wish you’d formatted the post more readably, and possibly in better context. I also wish you’d not linked the other article, then pasted it in full in your text. I am reading it and agree that it is helpful from your later context. I suggest your work and the McGowan piece, if not already re-published as separate articles or features on WUWT would be useful on this site for other readers. I’m almost certain that they have not been so featured, or you might have just linked them to begin with.

This is probably wasted effort for me on a dead or dying thread, that you might not even see. I too, obviously, get quite verbose. I do err, frequently. I often do not correct appropriately. I wish to make this interchange constructive, because I really think that I can learn something from you.

But the abuse stops here. At best, it is accidental friendly fire from a misunderstanding between people on the same side of an issue. I will assume until proven otherwise that the best is the correct interpretation.

Last edited 2 months ago by dk_
H. D. Hoese
May 18, 2021 6:55 am

If I remember correctly the discipline of statistics can be traced back over two centuries, certainly no excuse for misuse for at least a century. American Statistical Association not long ago came out with a statement about misuse of p for causation. Don’t have the link handy.

Gary Pearse
Reply to  H. D. Hoese
May 18, 2021 10:23 am

Statistics was basically developed by matematically gifted gamblers, mostly French I believe, in 17th – 18th Centuries.

Kevin kilty
Reply to  Gary Pearse
May 18, 2021 1:14 pm

Probability…statistics is a younger science.

niceguy
May 18, 2021 7:32 am

Epidemiology is in principle the only truly scientific part of biomedicine but its the most buffoonish one!

dk_
Reply to  niceguy
May 18, 2021 11:31 am

niceguy: Perhaps it is an area worthy of study which does not provide the student any useful skills or principles in public administration or policy. Self destruction might come from an epidemiological study on prevention of the spread of hubris among specialists.

Michael in Dublin
May 18, 2021 7:35 am

Douglas Altman – a professor of statistics in medicine at the University of Oxford, who died in 2018 was deeply concerned about the use of statistics in medical research.
The majority of statistical analyses are performed by people with an inadequate understanding of statistical methods. They are then peer reviewed by people who are generally no more knowledgeable. Sadly, much research may benefit researchers rather more than patients, especially when it is carried out primarily as a ridiculous career necessity.

May 18, 2021 7:39 am

Follow the science used to mean adhere to the scientific method. It now means establish what your politics wants the science to be, create a model to ‘prove’ your presumption, bolster the fantasy with misleading statistics and esoteric language and then apply extreme self righteousness by denigrating anyone who who has the gall to think otherwise.

This methodology has worked so well for shaping climate science to pursue political goals, is anyone surprised that it’s found its way into other fields of science that unfortunately also intersects with politics?

fretslider
May 18, 2021 7:50 am

This is very much the age of the ego.

n.n
Reply to  fretslider
May 18, 2021 7:55 am

Narcissistic delusion is a common condition driven by individual and social ego, sympathetic and empathetic appeal, as well as gnorance and malice.

n.n
Reply to  fretslider
May 18, 2021 7:57 am

… and, of course, both positive and negative secular incentives.

fretslider
Reply to  n.n
May 18, 2021 8:12 am

Secular? I was under the impression that one has to believe…

Danley Wolfe
May 18, 2021 8:02 am

Many or most of WUWT followers also subscribe to Judith Curry’s Climate etc. WUWT used to have much less reposting in the past. I vote for that.

Ron Long
May 18, 2021 8:10 am

The epidemiologists favorite tool is the “Linear-NO Threshold” trick, which utilizes decreasing probability of bad effect but over larger and larger segments of the population. This is used to justify avoidance of nuclear reactors. AT Fukushima one (1) death of a worker has now been reported. Contrast this with more than one thousand (+ 1,000) who died during the panic evacuation.

cat
May 18, 2021 8:26 am

I’d like to read the study but the link says page not found. Does anyone have a working link? Thanks.

Reply to  cat
May 18, 2021 8:37 am

Seems not to exist somewhere

dk_
Reply to  cat
May 18, 2021 12:04 pm

Cat: Me too. https://www.nas.org/reports/shifting-sands is broken. Searched for authors and title on NAS site. Tried from Climate.etc and HootNeoos. Couldn’t locate it. “Shifting Sands” seems to be a separate project on NAS, but didn’t contain this article.
Perhaps the project launch date of 20 May is the issue? Might this be released with other articles on the same day? See https://www.nas.org/blogs/event/event-shifting-sands-launch

Last edited 2 months ago by dk_
fretslider
May 18, 2021 8:36 am

flawed statistical practices

Not to mention flawed models…

The source code behind the Ferguson model has finally been made available to the public via the GitHub website. Mark E Jeftovic, in his Axis of Easy website, says: ‘A code review has been undertaken by an anonymous ex-Google software engineer here, who tells us the GitHub repository code has been heavily massaged by Microsoft engineers, and others, in an effort to whip the code into shape to safely expose it to the public. Alas, they seem to have failed and numerous flaws and bugs from the original software persist in the released version. Requests for the unedited version of the original code behind the model have gone unanswered.’

https://www.technocracy.news/neil-fergusons-computer-model-is-ripped-to-shreds/

The example of Neil Ferguson shows, the more hopelessly wrong you are the more eager they are to listen.

StevenF
May 18, 2021 8:53 am

I am an epidemiologist, among other things. I am also a physician. There are many who call themselves epidemiologists but are really just statisticians. The big difference being that an epidemiologist must tie the statistics back to principles and an understanding of physiology and medicine. Statisticians only look at the numbers.

It is much like looking at labs and making a diagnosis without examining the patient and understanding the context of their condition.

This was a very good article and touched on many of the problems that I see with epidemiology and medical statistics in general

Reply to  StevenF
May 21, 2021 3:47 am

Hi Dr Steven,

Last night I sent the following to two of my physician friends for comments – now awaiting replies. .

Nobel Prize winner: Mass COVID vaccination an ‘unacceptable mistake’ that is ‘creating the variants’
In every country, ‘the curve of vaccination is followed by the curve of deaths,’ the famous virologist said.
Wed May 19, 2021 – 6:59 pm EST
https://www.lifesitenews.com/news/nobel-prize-winner-mass-covid-vaccination-an-unacceptable-mistake-that-is-creating-the-variants
French virologist and Nobel Prize winner Luc Montagnier called mass vaccination against the coronavirus during the pandemic “unthinkable” and a historical blunder that is “creating the variants” and leading to deaths from the disease.

My own year-long opinion is that the Covid-19 “vaccines” are high-risk treatments that should not be given to the low-risk population, especially children and young adults. That was clearly true up to ~1Dec2020 before the variants appeared – the risk of death in Alberta for under-65’s was 1 in 300,000.

I have not researched the variants. The death toll from the Covid-19 injections is quite high, officially more than 4000 in the USA – reportedly more than the deaths from all previous vaccine injections in USA history. In the past, such a deadly vaccine would have been withdrawn by now.
 
I can see more than one possible conclusion from the data provided in the above article – more study required:
1.  As suggested by Dr Montagnier – the vaccinations are causing the variants to occur and thus increased deaths – “survival of the fittest”..
2.  The vaccinations are causing the deaths directly, and do not significantly cause the variants.
3.  The “second wave” is primarily seasonal; winter flu season is the primary cause.
4.  The second wave and variants were caused by the sanitizing, masking and distancing, which enabled the Covid-19 virus to survive through the summer and gave it the time and gradient to mutate into the variants – again, “survival of the fittest”.

Yours and others’ thoughts are welcomed.
 
Regards, Allan

Mike S
May 18, 2021 8:58 am

Not surprising. I taught university-level introductory statistics for a couple of years back in the 80s, and one of the chapters in the textbook we used was titled “How To Lie With Statistics”. I devoted one class session each term to that chapter, showing the students how to spot the most common ways people would misuse statistics to try to mislead them. It’s only gotten much, much worse since then.

Jim Gorman
May 18, 2021 9:01 am

“They have an insufficient sense of the need for awareness of just how much statistics must remain an exercise in measuring uncertainty rather than establishing certainty. ”

Truer words were never spoken. Climate science is awash in this. Uncertainty in measurements, precision, anomalies, on and on are ignored or worse, no knowledge of the underlying issues. Too many statisticians/mathematicians with no concept of the uncertainties of physical measurements. They have no recognition that climate parameters are a continuous function. Temperature measurements are points on a time series or a continuous wave but statisticians treat them as a collection of individual physical characteristics.

dbidwell
May 18, 2021 9:05 am

Maybe I missed it in the article, but if the data sets are private how were you able to a meta-analysis on it to conclude some studies are bilinear? The plot data comes directly from the articles and do not need raw data?

n.n
Reply to  dbidwell
May 18, 2021 2:33 pm

The plot is of p-values.

Conceptually, a p-value plot allows us to examine a specific premise that factor A causes outcome B using data combined from multiple observational studies in meta-analysis. What should a p-value plot of the data look like?

a plot that exhibits bilinearity—that divides into two lines—provides evidence of publication bias, p-hacking and/or HARKing (Figure 3)

May 18, 2021 9:06 am

During the time discussing Didiers Raoults HCQ work, for the first time I realised “France Soir Online” – the online presence of a French newspaper – I realised the very realistic view they present.
I found a German post based on the translation of an article published in French at “FSO” very interesting and clarifying:

57 scientists and doctors call for immediate halt to all Covid-19 “vaccinations”
A group of 57 leading scientists, physicians and political experts has released a report calling for questioning the safety and effectiveness of current COVID-19 “vaccines” and is now calling for an immediate end to all immunization programs. Among them is geneticist Alexandra Henrion-Caude.
The therapies used called “vaccines” do not meet the definition of the word vaccine and it would be more appropriate to name them gene therapies or vaccine-vector therapies.

Worth reading !

Last edited 2 months ago by Krishna Gans
KevinM
Reply to  Krishna Gans
May 18, 2021 2:27 pm

A group of … and political experts”?

Reply to  KevinM
May 18, 2021 2:40 pm

All authors are listed, first by name and index number, than inidex number and function.

Michael E McHenry
May 18, 2021 9:25 am

I didn’t see anyone mention one of the most fraudulent tools of the EPA “dose extrapolation”. I first encountered it in the late 1980’s when everyone in my area of New Jersey was in a tizzy about radon seeping into houses due radioactive decay of uranium in the underling rock. The highest levels in basements. I measured my basement with a kit and found 11 pico curies the EPA’s recommended limit was 4pc. I investigated where this number came from before making an expensive in ventilating.. The EPA came up with the number by dose extrapolation of mine workers exposed to 20,000 to 50,000 pc!

rd50
Reply to  Michael E McHenry
May 18, 2021 2:03 pm

Yes, I remember this.

Frank from NoVA
May 18, 2021 9:26 am

Steve Malloy’s book “Junk Science Judo” was a good introduction to alarmism many years ago.

May 18, 2021 9:56 am

A new view on Cov-19

Day 8 therapy for COVID-19
This article focuses on the concept of “8th Day Therapy” developed by Dr. Shankara Chetty of South Africa, who treated some 4,000 COVID-19 patients and studied the pathogenesis of the disease and refined its treatments at the same time.

No idea, how serious that is, but an interesting view.

Reply to  Krishna Gans
May 18, 2021 11:02 am

If true, it’s a complete new aspect with regard to vaccinated people.

stinkerp
May 18, 2021 9:57 am

To simplify it even further…

If it’s a “meta study” or a “meta analysis”, which is a study of other studies, you can be sure its conclusions are worthless. All the flaws that may exist in the underlying studies are simply propogated and amplified in the meta study. Garbage in, garbage out.

If it makes any claim about air pollution, particulates, or secondhand smoke causing deaths each year, it’s junk science. These are not categories on a death certificate. They are simply the worst kind of data torture, connecting the often shorter lifespan of people with chronic pulmonary problems like asthma and emphysema to the idea that particulates and other pollution may exacerbate the condition and perhaps shorten life; ignoring the fact that particulates in the form of dust and pollen are a fact of everyday life and impossible to avoid.

Smart Rock
May 18, 2021 9:59 am

It is indeed a sorry state of affairs where the researchers’ own career advancement is more important (to the researchers) than reaching robust conclusions. “Robust” – in this situation – means capable of standing up to rigorous scrutiny. Especially when robust conclusions wouldn’t lead to public concern, calls for new regulations or (crucially!) new research grants for further study. And might have difficulty getting published, and definitely wouldn’t get media attention.

There’s not much career advancement there, even though they would be doing their jobs.

It’s not hugely different from the “bad news factory” of climate-related research. Harking would actually be a step up for most climate-related studies. Their conclusions are reached BEFORE the study starts. And of course they use model outputs as data inputs into their own modelling projects where they will yet again demonstrate that the subject under study will be “worse than we thought” due to human emissions. Climate Science is many things, but science is not one of them.

May 18, 2021 10:06 am

Cool! Now do climastrology. Do climastrology. C’mon, do climastrology next. You know, like this one, but climastrology….do climastrology next….

“…statistics must remain an exercise in measuring uncertainty rather than establishing certainty.” That’s poetry, that!

Gary Pearse
May 18, 2021 10:10 am

Interestingly, 2.5 microns is the diameter of cigarette smoke particles. Also, it is well known physiologically that ‘cilia’ (motile hairs in the lungs) very efficiently move particles of this size upwards in the lungs through the thin mucus coating and into the bronchial tubes where they can be coughed out.

Even a smoker gets good service for many years, but the constant workload of the cilia eventually wears them down resulting in loss of cilia motility and emphysema eventually results.

Now this physio fact is known to epidemiologists (and compulsive trivia junkies like me) but maybe not to many stat-math professionals. I am blown away that just doing the rigorous statistical retesting of data sets from a number of flawed studies essentially tells us there is something grossly wrong with the characterization of 2.5 micron particles as dangerous to health! Moreover, the gov regulators have ‘logic’ on their side because most people, even well educated ones would grant that breathing this stuff in can’t be doing anyone any good.

I love statistics even though my skills with it aren’t anywhere on the same planet as that of Drs Young and Kindzierski. But I revere it all the more because of this analysis.

Gary Pearse
Reply to  Gary Pearse
May 18, 2021 1:35 pm

Of course, the chemical composition of the PM2.5 must also be considered. Inert particles are easily expelled by the lungs in the concentrations usually regulated. Sheesh, this is so easy to investigate by actual experiment.

Silica (quartz – SiO2) dust is dangerous because it is sufficiently soluble and solidifies as hyaline (opal) that repeated exposure builds up causing a terminal illness
known as silicosis, once common among gold quartz-vein miners.

Robert of Texas
May 18, 2021 10:12 am

Great posting. I love this p-hack test – very easy to understand and follow.

The EPA is a good example of how propaganda is turned into pseudo-science in order to convince the general population that a change is needed. Start with the end-result, make up some sort of pseudo-science to support it, and then use cherry-picked statistics to “prove” you are right.

Mike Dubrasich
May 18, 2021 10:16 am

Not to steal the authors’ thunder, but PM25 “research” is much worse than described. There is no measurement of exposure, only proxies such as distance from a home to a highway. There is no measurement of other variables, especially potentially alternate causes for the selected diseases, which include Parkinsons and Alzheimers (they are NOT lung diseases). No one knows who got exposed to what or how dust could cause the diseases. It’s completely phony junk science.

All the PM25 research comes from one institution, the Harvard T.H. Chan School of Public Health. It’s their bread and butter. Nobody else gets funded to do this phony research, or if they do it’s a side channel scam with the money run through Chan.

The EPA is a political animal, a bureau-jackal, that hunts prey. They do nothing to stop pollution; in fact they cause it [here]. The PM25 scam is just another panic fad ala climate/covid/cholesterol/whatever is handy.

https://www.newsweek.com/epa-causes-massive-colorado-spill-1-million-gallons-mining-waste-turns-river-361019

Gary Pearse
May 18, 2021 10:33 am

[Mods please don’t quarantine my excellent apropos comment too long]

dk_
May 18, 2021 11:08 am

In 2014 I “took” a non-credit Coursera on-liine class on “R” programming, mostly out of curiosity and an excess of free time. It was taught by a professor from Johns Hopkins, a computer scientist specializing in data modeling and statistics. Several lesson blocks used datasets made available through Johns Hopkins, often and unsurprisingly dealing with one or several various physio/medical and environmental data collection projects from University or government web sites.

I didn’t keep up with the R programming tool after the course, but I found it a very good statistical computation tool with great tools for data I/O and presentation.

One course block used a sample data set for 2.5 ppm data and corresponding data for incidence of several sorts of lung disease and respiratory disorders in the same metropolitan areas. If done right, per the professor’s instructions, the student was not supposed to be able to obtain a significan correlation between the particle concentration and respiratory disease. These data somehow had spurious correlations in them that should be filtered out by proper programming.

This is all subject to my often faulty memory, of course. I didn’t keep up my password and git access to my results, and I’ve moved on to other interests. I did recall the lesson when reading this particular piece. Is the EPA using different data sets? Or did someone innocently or maliciously exploit easy correlations out of incompetance or to meet preselected conclusions or support an agenda? Was my course lesson incorrect, and there should have been correlations found in these data?

My little story really isn’t pertinent to the authors’ points – I can’t verify my own results from seven years ago, and don’t really care to renew my aquaintance with or access to any old research or my rapidly obsoleting and superficial data extraction skill set. But it is that one little bit of personal experience that makes me tend to believe them and support the authors’ conclusions.

Discussions on WUWT seem often recently to have been about analysis of other press releases, “news”, government policy as propaganda. To me, this article is convincing.

Last edited 2 months ago by dk_
chris
Reply to  dk_
May 18, 2021 1:39 pm

so
(a) “My little story really isn’t pertinent to the author’s points”
Agree completely
(b) “I can’t verify my own results from seven years ago, ”
ok, they are way out of date WRT modern hypothesis testing (which now-a-days eschews p-value testing and is mostly Bayesian)
yet
(c) “But it is that one little bit of personal experience that makes me tend to believe them and support the authors’ conclusions.”

wow. You also might want a refresher on Freshman-level Logic.

🙂

dk_
Reply to  chris
May 18, 2021 2:08 pm

Chris,
(a)good
(b)agreed
(c)right — persuasion isn’t logic. I found it persuasive while acknowledging (verbosely) that I couldn’t honestly evaluate the science. I obviously failed to put that in context. My bad, but it was a shot.
To me it seems that this piece is an introductory article meant to bring attention to the authors’ program at NAS. Not a scientific paper itself. IMO, evaluating it as persuasion seems more appropriate for an acknowledged, attention-limited layman.

Petras
May 18, 2021 1:47 pm

Typo in Figure 1 title. Based on the x-axis (Rank Order), there are 69 (maybe 70) studies not 40 in the meta analysis.

Izaak Walton
May 18, 2021 2:31 pm

What a surprise. Another right wing lobby group is arguing for more pollution and unclean air.

chickenhawk
Reply to  Izaak Walton
May 18, 2021 4:00 pm

I commend you comrade. You strike decisive blow to win the day.

dk_
Reply to  Izaak Walton
May 18, 2021 4:04 pm

What a surprise: another illiterate eco-terrorist.

May 18, 2021 2:53 pm

‘UK Government’ and ‘Corrupt Scientific Advisors’ Are to be Tried for ‘Crimes against Humanity’ and ‘Genocide’
As the mainstream media are remaining silent on the subject, it may surprise you to discover that papers have been laid to start two separate legal proceedings against the UK Government and their corrupt scientific advisors for genocide and crimes against humanity.

The first is described in the following press release from attorney, Melinda C. Mayne, and Justice of the Peace, Kaira S. McCallum who has presided as a JP in Central London Magistrates and Crown Courts for the past twenty years, who also used to be a highly qualified pharmacist.

Also the London Times

Last edited 2 months ago by Krishna Gans
Alan Watt, Climate Denialist Level 7
May 18, 2021 7:20 pm

William Briggs had some interactions on PM2.5 studies with the California Air Resources Board (CARB) who wanted to impose some new regulations. I don’t have the link but as I recall Briggs pointed out several errors in the statistical methodology used and the response was that since it the same methodology was used in earlier studies behind previous regulations, they were going to continue to accept it.

Ed Bo
Reply to  Alan Watt, Climate Denialist Level 7
May 18, 2021 7:57 pm

The link is here: Criticism of Jerrett et al. CARB PM2.5 And Mortality Report – William M. Briggs (wmbriggs.com)

You can search his website for “PM2.5” and find many other similar posts.

Alan Watt, Climate Denialist Level 7
Reply to  Ed Bo
May 19, 2021 11:20 am

Thank you Ed; I was rushed lazy

Tombstone Gabby
May 18, 2021 7:52 pm

Looking at the playing card in the lead photograph. The first time I’ve seen a card that apparently has indices in all four corners. The last time I handled a deck of cards was a faro deck, no indices at all. Nothing at all to do with climate, but some interesting history, for folks who are involved in recreating the 1880’s for visitors

https://www.vanishingincmagic.com/playing-cards/articles/how-did-playing-cards-get-their-symbols/

Bob
May 18, 2021 10:38 pm

Why not require a P value plot for all published work used to support more regulation?

Alan the Brit
May 18, 2021 11:19 pm

PM2.5? That’s the new local scary story developed by the same crowd who were promoting diesel over petrol a few years back! Yes I’ve heard of this before & my opinion hasn’t changed! Throughout my life I have listened to many a bureaucrat who seeks some form of power & control with the view to manipulate, seize upon a technical term, then bandy it around with great authority as they were of a technical mind, to appear knowledgeable to those around them, & for the moment, PM2.5 suits the current scare story in the wings, ready for perhaps when the CAGW scare has run its course, but more likely to heap ever increasing fear upon the population to effect overload & panic, by keeping the pressure up, “the end of the world is nigh you must listen to me”, (love that, I bags the title on that one as nobody else has ever claimed that before for certain!) 😉

I still love the good old Penn & Teller scare gag where they employed a few pretty young ladies (sorry I’m old) to go around a park on a sunny weekend day, each with a clip board & form, for people to sign a petition in support of getting guvment to ban Dihydrogen-Monoxide, because big oil, nuclear, food companies, & even drinks companies, were adding this toxic stuff to everything, they got hundreds of signatures in favour of the ban, only none of the signatories knew they were being asked to ban “water” !!!! Go figure!!!

TonyG
Reply to  Alan the Brit
May 19, 2021 7:22 am

The DHMO thing was also tried at an international environmental conference and got tons of signatures too. And there was at least one town that almost voted to ban it – they were set to vote on the resolution before someone finally realized what was going on.

D Boss
May 19, 2021 5:06 am

There are Lies, Damned Lies, and Statistics!

Lies.jpg
%d bloggers like this: