The Unz Review: An Alternative Media Selection
A Collection of Interesting, Important, and Controversial Perspectives Largely Excluded from the American Mainstream Media
 TeasersiSteve Blog
The Privacy Problem with Data-Mining for Coronavirus Clues Explained
🔊 Listen RSS
Email This Page to Someone

 Remember My Information



=>

Bookmark Toggle AllToCAdd to LibraryRemove from Library • BShow CommentNext New CommentNext New ReplyRead More
ReplyAgree/Disagree/Etc. More... This Commenter This Thread Hide Thread Display All Comments
AgreeDisagreeThanksLOLTroll
These buttons register your public Agreement, Disagreement, Thanks, LOL, or Troll with the selected comment. They are ONLY available to recent, frequent commenters who have saved their Name+Email using the 'Remember My Information' checkbox, and may also ONLY be used three times during any eight hour period.
Ignore Commenter Follow Commenter
Search Text Case Sensitive  Exact Words  Include Comments
List of Bookmarks

I have been pointing out for awhile that we have giant databases of patient medical records (e.g., Kaiser Permanente has about 12 million patients) that could be subjected to Big Data analyses to get a better feel for the relative risks of coronavirus infection in different workplaces. For example, if schools were reopened, how much at risk would teachers be? How does the risk of mass transit compare to the risk of flying, as judged by infection rates of bus drivers and stewardesses?

A reader explains the subtle ways privacy rules hamstring massive data-mining of medical records for clues about how best to fight the epidemic:

I work in an obscure branch of healthcare IT, and I am pretty familiar with HIPAA and HITECH provisions. HIPAA has an explicit provision for de-identification, but here’s why it doesn’t help. There are two different ways to de-identify. The easiest is the so-called “safe harbor” provision. It’s so straightforward, anyone can understand it. But part of requires all dates (including disease onset, diagnosis, treatment, etc.) to be rounded to the nearest year. So kind of useless for what you want it for. Also, the government reserved the right to come after you, even if you use the safe harbor. There’s some fine print that says that if you have reason to believe that the data can be re-identified in any way, you can’t use the safe harbor (like if your dataset includes the diagnosis V95.43XS: Spacecraft collision injuring occupant). Entire books have been written on the perils of de-identification, so it ought to make potential anonymizers nervous.

The second method is the “expert determination” method. Basically it means hiring a statistician familiar with these kinds of protocols to figure out how you can get what you want while maintaining anonymity. It’s a small fraternity, and they are booked months and years out by data miners, pharmaceuticals, etc. It becomes a negotiation in which you give some bit of interesting data in order to keep some other bit that’s even more interesting. The expert has to build and test models of your special method and estimate re-identification risk. It takes months, and there is no guarantee you will even get to a satisfactory solution. There is no official certification for these guys, so the government can always come back and claim your expert was not expert enough. And there’s no objective threshold for acceptable risk, either.

Every time you mention Chetty, I cringe. I have looked in the past, but I can’t find any detailed documentation of the anonymization protocols the IRS used. Here is the closest thing I could find, but it doesn’t lay out the protocols. I’m going to make a bold claim: there was a massive risk of re-identification in the datasets Chetty got his hands on. It’s naive to think that stripping out the obvious personally identifying data is sufficient. There are massive publicly available consumer datasets out there that can be correlated with your “anonymized” data to re-identify individuals.
I presume the guys working for Chetty at Harvard and Stanford on his Equality of Opportunity project are extremely good at what they do, so that if they felt like identifying Trump’s or Gates’ or Bezos’ 2012 tax return, they probably could. On the other hand, it would be a big faux pas for their careers for this to leak out, so perhaps they’ve all chosen to be above reproach.
Anyway, my point is that Chetty and Co. have had access to information since 2013 that nobody before them ever had the gall to imagine they could get … and the world hasn’t come to an end because of it. Maybe somebody could do something similar to help figure out how to get out of this medical and economic hole we’re in?
 
Hide 49 CommentsLeave a Comment
Commenters to Ignore...to FollowEndorsed Only
Trim Comments?
  1. And most importantly, we don’t want the wrong facts discovered.

    • Agree: Meretricious
  2. In this case, Being beyond reproach means being beyond human nature!

    (And thanks to your anonymous “reader” for shedding professional insight into this arcane topic!)

  3. Besides terrorizing the Wisconsin Northwoods, in the last few years I have been working on the law side of data privacy.

    The reader is correct, there is a huge risk of reidentification if any of this gets published. It is a problem. BUT, we are in a plague. Keep the data private and don’t publish the data, just the results. But have the analysis done.

    • Replies: @Jonathan Mason

    The reader is correct, there is a huge risk of reidentification if any of this gets published.
     
    Not if you anonymize it properly.

    Patient X who was treated with C drug was prevented from carrying out his duties as Prime Minister of an island nation was missing in action for 20 days, while patient Y who was in the control group was able to return to his position as Prime Minister of a NAFTA member country within 6 days.
  4. Or you could focus on how to get quality test kits produced early enough to do a targeted quarantine. Of course that would mean the epidemic wouldn’t be rampant enough to indicate which workers were more vulnerable. Thus, Sailer and co. wouldn’t get payed.

  5. HIPPA confidentiality rules are quite ridiculous. For example in the prison for young men where I once worked, the medical department would have considered it to be a breach of HIPAA confidentiality laws if the diagnosis of a young man receiving a cream for acne om his face was accidentally revealed to his bunkie.

    I totally understand that medical information needs to be kept confidential, because you don’t want your neighbors or employers knowing that you have some embarrassing complaint that is none of their business, and you don’t want to get on commercial mailing lists sending out envelopes addressed to ” THE HIV PATIENT AT APARTMENT #10″.

    But the HIPAA legislation, which was originally intended to make it EASIER for health care providers to share important information such as “do not prescribe narcotics to this addict, because he will sell it on the street” has actually become a massive burden preventing sharing of health information that would benefit patients and makes the provision of health care more expensive.

    At least that is my opinion.

  6. @Hodag
    Besides terrorizing the Wisconsin Northwoods, in the last few years I have been working on the law side of data privacy.

    The reader is correct, there is a huge risk of reidentification if any of this gets published. It is a problem. BUT, we are in a plague. Keep the data private and don't publish the data, just the results. But have the analysis done.

    The reader is correct, there is a huge risk of reidentification if any of this gets published.

    Not if you anonymize it properly.

    Patient X who was treated with C drug was prevented from carrying out his duties as Prime Minister of an island nation was missing in action for 20 days, while patient Y who was in the control group was able to return to his position as Prime Minister of a NAFTA member country within 6 days.

    • Replies: @Charles Erwin Wilson's phone
    Not if you properly deidentify? If you deviate from the narrative you will be punished. It's a certainly.
  7. I wonder if the CDC observes the same HIPAA guidelines. On April 17th they issued a weekly mortality and morbidity report describing how one Chicago man with mild respiratory symptoms infected 16 people over the course of a few days, three of them fatally.

    There were no names in this report, just designations like a1.1, b2.3, etc. But it was very obvious who a2.1 was. Only one Chicagoan died of COVID19 on the date CDC gave for a2.1’s death, and her name was widely publicized. So much for anonymity. If the CDC can get away with it, maybe they’re the ones who should do the data mining.

    That said. it’s really an interesting report. I would like to see more case studies like this from the CDC.

    https://www.cdc.gov/mmwr/volumes/69/wr/mm6915e1.htm?s_cid=mm6915e1_w

  8. @Jonathan Mason

    The reader is correct, there is a huge risk of reidentification if any of this gets published.
     
    Not if you anonymize it properly.

    Patient X who was treated with C drug was prevented from carrying out his duties as Prime Minister of an island nation was missing in action for 20 days, while patient Y who was in the control group was able to return to his position as Prime Minister of a NAFTA member country within 6 days.

    Not if you properly deidentify? If you deviate from the narrative you will be punished. It’s a certainly.

    • Replies: @Reg Cæsar

    Not if you properly deidentify?
     
    We could use some de-identity politics at the moment.

    Wait... I take that back. Not if it leads to President Biden.

    How's the Draft Dukakis movement faring? "We could use someone born in the Thirties!"
  9. I’m not sure how effective it would be to further quantify risk, Steve. The “if it saves one life” fallacy seems to be ruling everyone’s decision-making right now.

    I read an apparently serious article in my local paper yesterday bemoaning the death from Coronavirus of a 93-year-old with “no underlying conditions” (!). We are clearly no longer able to think rationally about this illness.

    We may just have to wait for the panic to subside. Remember backwards messages on heavy metal records? Satanic daycares? Lobotomies for unruly teens? #metoo? Society must shudder, sweat this thing out, and eventually the fever will break.

  10. Under Obamacare all group health insurance plans with under 100 employees are fully insured. They are lumped in a risk pool with all the other small employers in the state and can only discriminate on price by looking their geographic zone and age. But once they go over 100 and become a “large” group, the insurance company can use loss experience to price the group. Each year they have to produce detailed loss runs that show utilization. It is quite easy to figure out how much certain employees cost the health plan by matching dates and company chit chat. This is especially true when the group is in “no-man’s land” between say 100 – 300 employees and it is possible for the upper management to know everyone in the company. The other tell is when you know what drugs are being taken because it is very easy to figure out what they are being used for. The biggest problem with groups that size is that all it takes is one complicated pregnancy or car accident to blow the group out.

    While it is illegal to discriminate, you really start to understand how much more it costs to insure certain demographics. High tech companies made up of 25 year old educated single men are super inexpensive compared to female, blue collar, and older populations.

    • Thanks: Almost Missouri
  11. @Charles Erwin Wilson's phone
    Not if you properly deidentify? If you deviate from the narrative you will be punished. It's a certainly.

    Not if you properly deidentify?

    We could use some de-identity politics at the moment.

    Wait… I take that back. Not if it leads to President Biden.

    How’s the Draft Dukakis movement faring? “We could use someone born in the Thirties!”

  12. The obsession over the “relative risk” of different occupations is difficult to understand , since most people who get infected will never go to the hospital, therefor the information obtained from “data mining” will have little value.

    90% of those who get infected will never bother getting tested. The few who get tested are the elderly and obese who fear they will need hospitalization or special treatment to survive the Coronavirus.

    In the NY Metro area ~15,000 have died from CV in April alone and with a fatality rate of .5% we can estimate that 3 million were already infected 4 weeks ago and another 5 million have been infected over the last 4 weeks

    At least 8 million people in the NY Metro area have already been infected , which is 40% of the 20 million living here, To suggest we need to data mine to figure out which occupations are most at risk is absurd….almost everyone in The New York Metro Area has been exposed to the Coronavirus already.

    Obviously bus drivers in NY are exposed to CV positive passengers every hour they are driving , probably dozens of exposures per day….as are the clerks in the grocery stores and everyone taking the subway is exposed to dozens of CV positive people each minute they are riding on the subway.

    • Agree: botazefa
    • Replies: @Mr McKenna

    Obviously bus drivers in NY are exposed to CV positive passengers every hour they are driving , probably dozens of exposures per day….as are the clerks in the grocery stores and everyone taking the subway
     
    Finally got desperate enough for groceries that I used the parking-lot pickup service that my grocery offers. Place your order online and pay with your CC. Worked very well--there is a dedicated pickup zone separate from the main parking area and serviced by a separate [employees-only] entrance even. You don't even get out of your car, you just pop the trunk release.

    Two questions: One, why not make this standard for the duration? Granted there will always be some people who can't go this route, and they're the perfect customers for the grocery stores which can't go this route. For the rest of us, this seems like a great solution, and I describe it here because it's also likely to be great for the employees too. Provided we clear the stores of customers.

    And two, how on earth do you clean broccoli in the current environment? That stuff has the surface area of a football field.

    For bus drivers, there should be acrylic shields, which 1) will protect them from other things as well and 2) will permit them to open their side windows and avoid sharing the passengers' air. Win-win, as everyone likes to say.

  13. The Kaiser Foundation released data for its NorCal, SoCal, and Washington operations.

    4.3M NorCal Kaiser members.
    539 COVID Patients (.01%)
    379 Hospitalizations
    59 Deaths (.00001%) or 1 in 73000

    https://www.medrxiv.org/content/10.1101/2020.04.12.20062943v1.full.pdf

    • Replies: @Steve Sailer
    Thanks.
    , @Steve Sailer
    Thanks
  14. Anon[560] • Disclaimer says:

    The Census site has a lot of information on deidentification, and they have an office within the bureau that deals with it. I scanned through it once shortly after Cherry got census access (albeit only via Census staff researchers who have to implement certain steps firewalled off from Chetty’s group, and who are thus credited as coauthors). My impression as a non-mathematician was that there are myriad ways you can go wrong with this.

    And also, as your commenter mentioned, the actual data can make the job almost impossible, as was shown with the notorious AOL data release: somehow nobody thought that people would ego search, so you had things like “user 5775326” searching for “Robert Smithtanian” every week or so mixed among searches for “is my penis too small” and “how to hire a hitman.”

  15. we have the data in New York City, 63% of the deaths are over the age of 70
    https://covid19tracker.health.ny.gov/views/NYS-COVID19-Tracker/NYSDOHCOVID-19Tracker-Fatalities?%3Aembed=yes&%3Atoolbar=no&%3Atabs=n

    most people who have CV will never be tested, And even among those who are tested , most will never be hospitalized, it is unclear what we would learn from data mining the hospital data. We already know that the biggest risk factor is being old. 90% of those who are infected have mild symptoms and will never be tested. So data mining will only examine the 10% of CV cases which bothered to get tested, and these people are mostly elderly with severe symptoms and not typical cases.

    what do we expect to learn from data mining such a small subset of the CV cases ? we will not learn which occupations are at the highest risk from the confirmed cases, since the confirmed cases are those who are very sick which are mostly the elderly who are not employed. So we will not learn which occupations carry the biggest risk if we are restricted to analyzing elderly patients.

    In New York City we can estimate that at least 2 million New Yorkers were infected 3 weeks ago based on the 12,000 CV deaths. At this point we can project that ~50% of the NYC population has already contracted CV and the NYC death toll will reach 24,000 over the next 4 weeks if the fatality rate is .6% and 32,000 deaths if the fatality rate is .8%.

    if the death rate is actually only .2% then 6 million New Yorkers were infected 3 weeks ago and 90% of New Yorkers have already been infected with CV by today. If the fatality rate is less than ,2% then almost all New Yorkers have already been infected with CV and data mining would be pointless.

    • Replies: @Steve Sailer
    "what do we expect to learn from data mining such a small subset of the CV cases ? we will not learn which occupations are at the highest risk from the confirmed cases, since the confirmed cases are those who are very sick which are mostly the elderly who are not employed."

    It's called multiple regression. You adjust for age, sex, race, BMI, zip code, health problems, etc. A big HMO's database on its patients has all that information.

  16. @Travis
    we have the data in New York City, 63% of the deaths are over the age of 70
    https://covid19tracker.health.ny.gov/views/NYS-COVID19-Tracker/NYSDOHCOVID-19Tracker-Fatalities?%3Aembed=yes&%3Atoolbar=no&%3Atabs=n

    most people who have CV will never be tested, And even among those who are tested , most will never be hospitalized, it is unclear what we would learn from data mining the hospital data. We already know that the biggest risk factor is being old. 90% of those who are infected have mild symptoms and will never be tested. So data mining will only examine the 10% of CV cases which bothered to get tested, and these people are mostly elderly with severe symptoms and not typical cases.

    what do we expect to learn from data mining such a small subset of the CV cases ? we will not learn which occupations are at the highest risk from the confirmed cases, since the confirmed cases are those who are very sick which are mostly the elderly who are not employed. So we will not learn which occupations carry the biggest risk if we are restricted to analyzing elderly patients.

    In New York City we can estimate that at least 2 million New Yorkers were infected 3 weeks ago based on the 12,000 CV deaths. At this point we can project that ~50% of the NYC population has already contracted CV and the NYC death toll will reach 24,000 over the next 4 weeks if the fatality rate is .6% and 32,000 deaths if the fatality rate is .8%.

    if the death rate is actually only .2% then 6 million New Yorkers were infected 3 weeks ago and 90% of New Yorkers have already been infected with CV by today. If the fatality rate is less than ,2% then almost all New Yorkers have already been infected with CV and data mining would be pointless.

    “what do we expect to learn from data mining such a small subset of the CV cases ? we will not learn which occupations are at the highest risk from the confirmed cases, since the confirmed cases are those who are very sick which are mostly the elderly who are not employed.”

    It’s called multiple regression. You adjust for age, sex, race, BMI, zip code, health problems, etc. A big HMO’s database on its patients has all that information.

    • Replies: @Kratoklastes

    It’s called multiple regression. You adjust for age, sex, race, BMI, zip code, health problems, etc. A big HMO’s database on its patients has all that information.
     
    If it were that simple, there would not be a replicability crisis in medical quant research - specifically at the intersection of medicine and pharmacology (and also at the intersection of psych and pharma).

    Multiple regression might generate a better 'one-shot' fit, which sounds like something that would interest researchers chasing publications.

    Add a regressor, R² goes up: why not add another 20?

    However every parameter has its own uncertainty - and when model-wide uncertainty needs to be quantified, it becomes something that needs to be part of sensitivity analysis (if sensitivity analysis is to be worthy of the name).

    And people need to stop saying 'adjust for X₁ etc' as if that's what's being done in practice. What's reported as some or other risk 'adjusted for X₁' is just a naïve attempt at evaluating marginal effects.

    To paraphrase:

    https://www.dropbox.com/s/zn2l33rnusekojp/ODNSX4.png?dl=1

    The documentation for the R margins package includes quite a good discussion as to why that's the case.

    Quantifying the uncertainty contained in estimates of marginal effects is far more important.

    Also... which marginal effect? The marginal effect at 'representative values' of the other parameters? At 'representative values' of the regressors? At 'representative values' of the output variable? That last one's a doozy, because people forget: in general, even linear models are not bijective.

    It invariably turns out that adding a regressor that improves R² - i.e. gives a better fit to the existing data - still causes more uncertainty. And when you chuck 25 or 30 variables into the 'explanatory' variables (as Petrelli et al did), the uncertainty in the model gets ridiculous and 'marginal effects' of one variable become a cloud.

    This could be investigated diligently and the results reported in data appendices, but that never happens. That's particularly egregious in generalised linear models, since regressors are supposed to be uncorrelated and so the convolutions required to systematic sensitivity analysis are straightforward (and computationally easy) - but it's never done.

    And even the 'regressors are supposed to be uncorrelated' question is not settled if investigators use tests that are inappropriate for the data type.

    .

    When a multiple regression is performed, there are conditions on the regressors that must be satisfied (for a linear model, it's the Gauss-Markov conditions).

    And underpinning that, is the meta-problem of testing if the conditions are satisfied.

    Learning to test if the relevant conditions are satisfied - by itself - involves enough material for a challenging, dense one-semester subject at (roughly) US first-year PhD level. It includes learning which tests (there are many) are relevant for different data types, and the shortcomings of tests when an inappropriate tests is used.

    Pick the wrong test and you might as well not bother.

    This was - almost certainly - the reason for the silly Odds Ratios for age groups in the Petrelli paper a few days ago. They used a test for collinearity that is known to be next-to-useless for categorical variables (VIF). (That paper still has the makings of an influential paper, if they fix the test they used for collinearity and present age-group data for to-ICU and deaths).

    A cynic might put that down to 'test-seeking' - selecting a test that gives the desired answer (in this case, 'no collinearity').

    I put it down to them simply not properly understanding what they're doing. Even in places where people are paid to be right (e.g., quant finance), very few people give UQ the respect it deserves.

    If the conditions on regressors are not satisfied, critical values for hypothesis tests differ from their asymptotic values. In other words, the value of the test associated with the desired significance level will be different from tabulated values.

    Worse, they differ in unpredictable ways, and are not ameliorated by increased sample size.

    The way to deal with this is to 'bootstrap' an estimate of the distribution of the hypothesis test statistic under the null hypothesis (by Monte Carlo, usually) for some models the same thing is required under some specific alternative because the distribution of the test changes.

    .

    People often use the term GIGO, to point out that data quality matters. That's just a bromide, because nobody disagrees with it.

    There's stuff being done between GI and GO: if that stuff's not done properly, then there's a risk of PIGO (perfection in, garbage out) - where the data can be flawless (in terms of its relevance, collection, filtering and so forth) but the technique involved in transforming inputs to outputs is flawed in important respects.

    inb4: "These people are public health experts." That's what they're referred to, but they have got almost every major public health issue wrong for the last 200 years - they even had to be dragged kicking and screaming when the issue was basic surgical hygeine (Semmelweiss) and basic epidemiology (Snow and the famous 'pump handle': when he did that, the establishment consensus was still 'miasma theory').

    Fauci's a 'public health expert' who got hetero-AIDS completely wrong (Oprah-level wrong), and changed his evaluation of covid19 from "The US has nothing to worry about" to "Existential threat: shut down the economy" in the space of two weeks.

    Ferguson's a public health expert who got H5N1 and vCJD completely wrong.
  17. Since I don’t click on WP links, will someone please tell me how funny the Four Reasons are?

    washingtonpost.com
    4 reasons coronavirus is hitting black communities so hard

    A Washington Post analysis of early data from jurisdictions across the country found that the novel …

    https://www.washingtonpost.com/politics/2020/04/10/4-reasons-coronavirus-is-hitting-black-communities-so-hard/

    • Replies: @Almost Missouri
    http://archive.is/https://www.washingtonpost.com/politics/2020/04/10/4-reasons-coronavirus-is-hitting-black-communities-so-hard/

    Spoiler Alert, they are:

    1. Higher rates of underlying health conditions, and less access to care
    2. Black Americans hold a lot of ‘essential’ jobs
    3. Insufficient information
    4. Housing disparities
     
    These can be translated as:

    1. Blacks sicker, lower life expectancy
    2. Blacks disproportionately in urban public sector jobs
    3. Blacks ignore laws, government
    4. Blacks more urban

    Needless to say, there is no attempt at quantification. I suspect most or all of the disparity, if it actually exists, can be ascribed to #4, which is just a backwards way of saying that "whites more rural" than other populations.

    Having looked at the CA data, I wonder if one adjusts for exposure that whites might do worse than others due to skewing older. The surprise in CA was how low the Hispanic rates were. This may also be due to CA Hispanics being more rural than average.
  18. @Hernan Pizzaro del Blanco
    The obsession over the “relative risk” of different occupations is difficult to understand , since most people who get infected will never go to the hospital, therefor the information obtained from “data mining” will have little value.

    90% of those who get infected will never bother getting tested. The few who get tested are the elderly and obese who fear they will need hospitalization or special treatment to survive the Coronavirus.

    In the NY Metro area ~15,000 have died from CV in April alone and with a fatality rate of .5% we can estimate that 3 million were already infected 4 weeks ago and another 5 million have been infected over the last 4 weeks

    At least 8 million people in the NY Metro area have already been infected , which is 40% of the 20 million living here, To suggest we need to data mine to figure out which occupations are most at risk is absurd....almost everyone in The New York Metro Area has been exposed to the Coronavirus already.

    Obviously bus drivers in NY are exposed to CV positive passengers every hour they are driving , probably dozens of exposures per day....as are the clerks in the grocery stores and everyone taking the subway is exposed to dozens of CV positive people each minute they are riding on the subway.

    Obviously bus drivers in NY are exposed to CV positive passengers every hour they are driving , probably dozens of exposures per day….as are the clerks in the grocery stores and everyone taking the subway

    Finally got desperate enough for groceries that I used the parking-lot pickup service that my grocery offers. Place your order online and pay with your CC. Worked very well–there is a dedicated pickup zone separate from the main parking area and serviced by a separate [employees-only] entrance even. You don’t even get out of your car, you just pop the trunk release.

    Two questions: One, why not make this standard for the duration? Granted there will always be some people who can’t go this route, and they’re the perfect customers for the grocery stores which can’t go this route. For the rest of us, this seems like a great solution, and I describe it here because it’s also likely to be great for the employees too. Provided we clear the stores of customers.

    And two, how on earth do you clean broccoli in the current environment? That stuff has the surface area of a football field.

    For bus drivers, there should be acrylic shields, which 1) will protect them from other things as well and 2) will permit them to open their side windows and avoid sharing the passengers’ air. Win-win, as everyone likes to say.

    • Replies: @Muggles
    >>Two questions: One, why not make this standard for the duration? Granted there will always be some people who can’t go this route, and they’re the perfect customers for the grocery stores which can’t go this route. For the rest of us, this seems like a great solution, and I describe it here because it’s also likely to be great for the employees too. Provided we clear the stores of customers.<<

    As to the first question about grocery stores providing pick up for those online ordering, it has been offered in my big city adjacent suburb by most major stores for a year or more. You now see aisles being clogged with mostly zoomers tugging large rolling bins using hand held tablets for their bin filling/shopping. Not just a pandemic thing. Also I think in the large city many stores have been doing this.

    Others have and still offer home delivery for a nominal fee.

    I can't say how widespread it is but this isn't new, several years or more in place.
  19. EXCLUSIVE: Dr. Kizzmekia Corbett, who is described by the National Institutes of Health (NIH) as the “lead” member of the U.S. government team racing to find a coronavirus vaccine, has engaged with online theories calling the pandemic a black “genocide” and condemned what she called “systematic oppression” by white people, a review of her social media posts by “Tucker Carlson Tonight” reveals.

    Corbett has also reposted a tweet urging Dr. Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases (NIAID), and Dr. Jerome Adams, the surgeon general, to “check” their “privilege.”

    She further slammed the White House coronavirus task force in February, saying it “is largely people (white men) he appointed to their positions as directors of blah blah institute. They are indebted to serve him NOT the people.”

    Her Twitter biography reads: “Virology. Vaccinology. Vagina-ology. Vino-ology. My tweets are my own. My science is the world’s.”

    Here’s Kizzimekia lecturing them white supremacists on what’s up with Virology.

    https://www.foxnews.com/politics/coronavirus-researcher-pandemic-genocide-blacks-nih

    Yep, our fate rests in her capable hands. And this is with Trump as president.

    • Replies: @Pericles

    Dr. Kizzmekia Corbett, who is described by the National Institutes of Health (NIH) as the “lead” member of the U.S. government team racing to find a coronavirus vaccine

     

    This time the damn racists won't be able to complain about the hagiographic movie being greenlit as we read this.

    She does seem a bit crazy and aggressive though. Is the pressure bad? Better hope 'black don't crack'.
    , @Anon7
    Our president, Donald J. Trump, will never get reelected without more of the black vote and the Hispanic vote than Republican candidates usually get. He knows it, and in the public parts of the government, he constantly shows how he's advancing and helping black people. Just look at the recent celebration of truckers at the White House.

    https://www.youtube.com/watch?v=miImQAY8PiU

    This is pure politics, for consumption by the masses, and if it leads to Trump's reelection - as opposed to the reign of Joe Biden 46 - then I say fine. Since when has a government panel of scientists ever discovered anything scientific?

  20. @Santa Clara County
    The Kaiser Foundation released data for its NorCal, SoCal, and Washington operations.

    4.3M NorCal Kaiser members.
    539 COVID Patients (.01%)
    379 Hospitalizations
    59 Deaths (.00001%) or 1 in 73000


    https://www.medrxiv.org/content/10.1101/2020.04.12.20062943v1.full.pdf

    Thanks.

  21. @Santa Clara County
    The Kaiser Foundation released data for its NorCal, SoCal, and Washington operations.

    4.3M NorCal Kaiser members.
    539 COVID Patients (.01%)
    379 Hospitalizations
    59 Deaths (.00001%) or 1 in 73000


    https://www.medrxiv.org/content/10.1101/2020.04.12.20062943v1.full.pdf

    Thanks

  22. @Mr McKenna
    Since I don't click on WP links, will someone please tell me how funny the Four Reasons are?


    washingtonpost.com
    4 reasons coronavirus is hitting black communities so hard

    A Washington Post analysis of early data from jurisdictions across the country found that the novel ...

    https://www.washingtonpost.com/politics/2020/04/10/4-reasons-coronavirus-is-hitting-black-communities-so-hard/
     

    http://archive.is/https://www.washingtonpost.com/politics/2020/04/10/4-reasons-coronavirus-is-hitting-black-communities-so-hard/

    Spoiler Alert, they are:

    1. Higher rates of underlying health conditions, and less access to care
    2. Black Americans hold a lot of ‘essential’ jobs
    3. Insufficient information
    4. Housing disparities

    These can be translated as:

    1. Blacks sicker, lower life expectancy
    2. Blacks disproportionately in urban public sector jobs
    3. Blacks ignore laws, government
    4. Blacks more urban

    Needless to say, there is no attempt at quantification. I suspect most or all of the disparity, if it actually exists, can be ascribed to #4, which is just a backwards way of saying that “whites more rural” than other populations.

    Having looked at the CA data, I wonder if one adjusts for exposure that whites might do worse than others due to skewing older. The surprise in CA was how low the Hispanic rates were. This may also be due to CA Hispanics being more rural than average.

    • Thanks: Mr McKenna
    • Replies: @Meretricious
    I'd wager (everything I own) that the biggest cause is health habits.

    Less access to care? Care is free for negroes and immigrants in this country. And it's as close as the nearest ER.

    Every urban and university ER is clogged with these people using it as an outpatient clinic.

    , @Anon
    From seeing that Los Angeles birthday party video, I'd say that No. 3, ignoring rules and government, looms large. There's an instant sense among blacks that any assertion of control over them is "racis'."
  23. Maybe somebody could do something similar to help figure out how to get out of this medical and economic hole we’re in?

    There’s an easier solution, but it is a violation of 18 USC 2385.

    • Replies: @Anonymous
    18 U.S. Code § 2385. Advocating overthrow of Government

    U.S. Code


    Whoever knowingly or willfully advocates, abets, advises, or teaches the duty, necessity, desirability, or propriety of overthrowing or destroying the government of the United States or the government of any State, Territory, District or Possession thereof, or the government of any political subdivision therein, by force or violence, or by the assassination of any officer of any such government; or

    Whoever, with intent to cause the overthrow or destruction of any such government, prints, publishes, edits, issues, circulates, sells, distributes, or publicly displays any written or printed matter advocating, advising, or teaching the duty, necessity, desirability, or propriety of overthrowing or destroying any government in the United States by force or violence, or attempts to do so; or

    Whoever organizes or helps or attempts to organize any society, group, or assembly of persons who teach, advocate, or encourage the overthrow or destruction of any such government by force or violence; or becomes or is a member of, or affiliates with, any such society, group, or assembly of persons, knowing the purposes thereof—

    Shall be fined under this title or imprisoned not more than twenty years, or both, and shall be ineligible for employment by the United States or any department or agency thereof, for the five years next following his conviction.

    If two or more persons conspire to commit any offense named in this section, each shall be fined under this title or imprisoned not more than twenty years, or both, and shall be ineligible for employment by the United States or any department or agency thereof, for the five years next following his conviction.

    As used in this section, the terms “organizes” and “organize”, with respect to any society, group, or assembly of persons, include the recruiting of new members, the forming of new units, and the regrouping or expansion of existing clubs, classes, and other units of such society, group, or assembly of persons.

     

    I don't advocate it, and I'm too old and worn out to do anything about it if I wanted to. I do predict it will happen-long after I'm dead and buried.

    Anyone trying it now or in the near future would be squashed like a bug. We are dealing with an Empire with nuclear weapons and nuclear submarines and aircraft carriers, attack helicopters, and sizable numbers of elite special warfare types who are amongst the most well equipped and trained combat troops in history, even if our mainline forces are not particularly hardened as a group.

    The difference between what happened in 1861 and what will happen when the Empire implodes will primarily come down to there being one cohesive group splitting off from another cohesive group then and a multi-way, fractious split consisting of numerous tribal and particularist groups squabbling with all the others. When America was all whites and blacks, with most blacks in the South, the War of Northern Aggression was thinkable and winnable. In an "America" with numerous racial, religious and lifestyle groups none of which really like the others, it'll be a free for all, a veritable lumberjack match.

    Already the system has made mistakes. I've met several Chinese, Vietnamese and dot-Indian engineers and programmers working at the Honeywell "bomb plant", on the critical arming, fuzing and surety systems on US nuclear weapons. I bet that ninety-plus percent of them are completely loyal to the country they hold citizenship in. I've seen films of the workers filing into Pantex in Amarillo. A lot of them are mestizo or indio "Hispanics" presumably from Mexico, and I bet many have been here two, three, four, five generations and I have no question that 99 plus percent of them consider themselves 100 percent loyal Americans. Ask anyone in the nuke business if "ninety-nine percent surety" is okay in nuclear surety when it comes to mechanical, electrical, or software matters.

    But that's not the worst of it, since if anyone in power actually decides to deploy nukes internally things have already gone past the heretofore conceivable. Okay, consider that the Big Four biker outfits-never exactly paragons of criminal competence, they couldn't even whack Mick Jagger-have substantially penetrated the DMV and vital records departments of several states. Scientology and the ADL have somewhat competently penetrated various law enforcement agencies and I have it on fairly good authority the Scienos have several key assets in at least one of the big military contractors in Wichita. This kind of thing is going to get more and more common. It's inherent to diversity.

    I was an avid reader of Harold Covington's novels in which he did a pretty good job of going over the conditions needed to actually 'pull "it" off'- but where I disagreed with him then and still do is it won't be one little breakaway province vs. the rest of the world. It'll be a pool table ball break and the cueball will be some random, unpredictable event, much like today's coronavirus crisis.


    I'm guessing when it hits, within a year or two the current CONUS (along with parts of West Canada and maybe parts of Mexico) will wind up being four to seven different polities. Who gets what-where the borders are drawn-and who gets the nukes are going to be big questions.

  24. @Almost Missouri
    http://archive.is/https://www.washingtonpost.com/politics/2020/04/10/4-reasons-coronavirus-is-hitting-black-communities-so-hard/

    Spoiler Alert, they are:

    1. Higher rates of underlying health conditions, and less access to care
    2. Black Americans hold a lot of ‘essential’ jobs
    3. Insufficient information
    4. Housing disparities
     
    These can be translated as:

    1. Blacks sicker, lower life expectancy
    2. Blacks disproportionately in urban public sector jobs
    3. Blacks ignore laws, government
    4. Blacks more urban

    Needless to say, there is no attempt at quantification. I suspect most or all of the disparity, if it actually exists, can be ascribed to #4, which is just a backwards way of saying that "whites more rural" than other populations.

    Having looked at the CA data, I wonder if one adjusts for exposure that whites might do worse than others due to skewing older. The surprise in CA was how low the Hispanic rates were. This may also be due to CA Hispanics being more rural than average.

    I’d wager (everything I own) that the biggest cause is health habits.

    Less access to care? Care is free for negroes and immigrants in this country. And it’s as close as the nearest ER.

    Every urban and university ER is clogged with these people using it as an outpatient clinic.

    • Agree: Jim Don Bob
  25. @Mr McKenna

    EXCLUSIVE: Dr. Kizzmekia Corbett, who is described by the National Institutes of Health (NIH) as the "lead" member of the U.S. government team racing to find a coronavirus vaccine, has engaged with online theories calling the pandemic a black "genocide" and condemned what she called "systematic oppression" by white people, a review of her social media posts by "Tucker Carlson Tonight" reveals.

    Corbett has also reposted a tweet urging Dr. Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases (NIAID), and Dr. Jerome Adams, the surgeon general, to "check" their "privilege."

    She further slammed the White House coronavirus task force in February, saying it "is largely people (white men) he appointed to their positions as directors of blah blah institute. They are indebted to serve him NOT the people."

    Her Twitter biography reads: "Virology. Vaccinology. Vagina-ology. Vino-ology. My tweets are my own. My science is the world’s."
     

    Here's Kizzimekia lecturing them white supremacists on what's up with Virology.

    https://a57.foxnews.com/static.foxnews.com/foxnews.com/content/uploads/2020/04/1862/1048/1000.jpeg

    https://www.foxnews.com/politics/coronavirus-researcher-pandemic-genocide-blacks-nih

    Yep, our fate rests in her capable hands. And this is with Trump as president.

    https://a57.foxnews.com/static.foxnews.com/foxnews.com/content/uploads/2020/04/1862/1048/Kizzmekia-Corbett-WikiMedia-1.jpg

    Dr. Kizzmekia Corbett, who is described by the National Institutes of Health (NIH) as the “lead” member of the U.S. government team racing to find a coronavirus vaccine

    This time the damn racists won’t be able to complain about the hagiographic movie being greenlit as we read this.

    She does seem a bit crazy and aggressive though. Is the pressure bad? Better hope ‘black don’t crack’.

    • Replies: @Mr McKenna
    Good point! Nipping that one in the bud! Now how do we take it back even further so that racists won't claim she got her degrees and position through affirmative action? Once we have that settled, though, we may need to go back a step further--naah, let's just kill all the racists. AKA anyone we don't like. Easier that way.
  26. In the late 1970’s, I was a hospital administration manager, and was one of the first to get my hands on a “microcomputer” as they were called. I was asked to do some patient care analysis, and so I asked for patient information over a five year period. After waiting for months, I got a six inch pile of printouts, with one patient per line, diagnosis, admission date, discharge date, billing, etc.

    Since there was no other way to print out the information from the hospital’s “Big Iron” computer system, the entries came complete with patient name, address and social security number. Obviously, I understood what I had and I kept it locked up while I was using it, then turned it back in for disposal.

    Another era, not so long ago.

  27. @Mr McKenna

    EXCLUSIVE: Dr. Kizzmekia Corbett, who is described by the National Institutes of Health (NIH) as the "lead" member of the U.S. government team racing to find a coronavirus vaccine, has engaged with online theories calling the pandemic a black "genocide" and condemned what she called "systematic oppression" by white people, a review of her social media posts by "Tucker Carlson Tonight" reveals.

    Corbett has also reposted a tweet urging Dr. Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases (NIAID), and Dr. Jerome Adams, the surgeon general, to "check" their "privilege."

    She further slammed the White House coronavirus task force in February, saying it "is largely people (white men) he appointed to their positions as directors of blah blah institute. They are indebted to serve him NOT the people."

    Her Twitter biography reads: "Virology. Vaccinology. Vagina-ology. Vino-ology. My tweets are my own. My science is the world’s."
     

    Here's Kizzimekia lecturing them white supremacists on what's up with Virology.

    https://a57.foxnews.com/static.foxnews.com/foxnews.com/content/uploads/2020/04/1862/1048/1000.jpeg

    https://www.foxnews.com/politics/coronavirus-researcher-pandemic-genocide-blacks-nih

    Yep, our fate rests in her capable hands. And this is with Trump as president.

    https://a57.foxnews.com/static.foxnews.com/foxnews.com/content/uploads/2020/04/1862/1048/Kizzmekia-Corbett-WikiMedia-1.jpg

    Our president, Donald J. Trump, will never get reelected without more of the black vote and the Hispanic vote than Republican candidates usually get. He knows it, and in the public parts of the government, he constantly shows how he’s advancing and helping black people. Just look at the recent celebration of truckers at the White House.

    This is pure politics, for consumption by the masses, and if it leads to Trump’s reelection – as opposed to the reign of Joe Biden 46 – then I say fine. Since when has a government panel of scientists ever discovered anything scientific?

    • Replies: @Meretricious
    Agreed. When you put it that way. If the window dressing has the intended effect that is. If it doesn't, it's just greasing the slide a bit more.
    , @Mr McKenna
    Granted they have to have "people of color"--ideally black women--in figurehead positions nowadays, but can't they find any who aren't quite so consumed with racist hatred as this harridan? From her pronouncements she doesn't appear terribly bright, either. I know that finding modern-day negroes who are 1) bright and 2) fair-minded is a needle-in-a-haystack task, but are we not up to it?

    Maybe we should get more Indians to migrate:


    Do testicles make men more vulnerable to coronavirus?
    Study suggests testes harbour Covid-19 giving virus sanctuary from immune system

    Men are dying from coronavirus at twice the rate as women in the UK alone

    https://i.dailymail.co.uk/1s/2020/04/18/19/27265796-8225245-image-a-51_1587234568747.jpg

    The study, carried out by researchers in New York and Mumbai, followed 48 men and 20 women living in Mumbai who had been infected by coronavirus.

    While the women in the study took four days to clear the infection men took 50 per cent longer, needing six days.

    Dr Aditi Shastri, an oncologist at Montefiore Medical Center in the Bronx, carried out the study with her mother Jayanthi Shastri, a microbiologist at the Kasturba Hospital for Infectious Diseases in Mumbai.

    Their work was released ahead of publication on medical website MedRxiv. It has not been peer reviewed.

     

    (Typical reader comment: "No worries, my wife keeps mine in her purse.")
  28. This is one of these “facts” that Steve wants to be true, so it is. Kaiser can use its own data for research and for public health purposes without violating HIPAA. It can also hire people to analyze its data. It can also let, say, academic researchers analyze its data. It can also report results of this research (as long as the results don’t identify anyone). Read here https://www.cdc.gov/phlp/publications/topic/hipaa.html, for example. All of this happens constantly, every day.

    For avoidance of doubt, HHS clarified that it will not take enforcement action against entities which provide personal health information for purposes of assisting with public health. This is helpfully provided on the website Steve links in his post. https://www.hhs.gov/about/news/2020/04/02/ocr-announces-notification-of-enforcement-discretion.html

    Personally identifiable data can be used for research: https://www.hhs.gov/hipaa/for-professionals/special-topics/research/index.html

    Personally identifiable data can be used for public health purposes: https://www.hhs.gov/hipaa/for-professionals/special-topics/public-health/index.html

    So what’s all this about de-identified data? De-identified data can be released publicly, to anyone, say on a website. The people who market drugs, medical devices, and the like want data on their potential customers. Insurers have these data. Both sides want this transaction to occur, and de-identified data provide a way for the transaction to occur without risk to the insurer.

    HIPAA is a costly annoyance to many people who want to use healthcare data. Like most of these regulatory boondoggles, its main effect is to provide employment to a class of box-ticking bureaucrats and lawyers similar to HR bureaucrats. It is not, however, a barrier categorically preventing important research from being done.

    Jonathan Mason gives a good example above of what HIPAA actually forbids. Your health insurer can’t just tell friends, relatives, and advertising agencies about your syphillis.

  29. @Almost Missouri
    http://archive.is/https://www.washingtonpost.com/politics/2020/04/10/4-reasons-coronavirus-is-hitting-black-communities-so-hard/

    Spoiler Alert, they are:

    1. Higher rates of underlying health conditions, and less access to care
    2. Black Americans hold a lot of ‘essential’ jobs
    3. Insufficient information
    4. Housing disparities
     
    These can be translated as:

    1. Blacks sicker, lower life expectancy
    2. Blacks disproportionately in urban public sector jobs
    3. Blacks ignore laws, government
    4. Blacks more urban

    Needless to say, there is no attempt at quantification. I suspect most or all of the disparity, if it actually exists, can be ascribed to #4, which is just a backwards way of saying that "whites more rural" than other populations.

    Having looked at the CA data, I wonder if one adjusts for exposure that whites might do worse than others due to skewing older. The surprise in CA was how low the Hispanic rates were. This may also be due to CA Hispanics being more rural than average.

    From seeing that Los Angeles birthday party video, I’d say that No. 3, ignoring rules and government, looms large. There’s an instant sense among blacks that any assertion of control over them is “racis’.”

    • Replies: @Meretricious
    Even the Daily Mail is running breathless headlines about the racial disparity in cobid mortality. This will not stand!
  30. They won’t allow it to be presented without being adjusted for SES, race, sex and maybe obesity.

  31. @Mr McKenna

    Obviously bus drivers in NY are exposed to CV positive passengers every hour they are driving , probably dozens of exposures per day….as are the clerks in the grocery stores and everyone taking the subway
     
    Finally got desperate enough for groceries that I used the parking-lot pickup service that my grocery offers. Place your order online and pay with your CC. Worked very well--there is a dedicated pickup zone separate from the main parking area and serviced by a separate [employees-only] entrance even. You don't even get out of your car, you just pop the trunk release.

    Two questions: One, why not make this standard for the duration? Granted there will always be some people who can't go this route, and they're the perfect customers for the grocery stores which can't go this route. For the rest of us, this seems like a great solution, and I describe it here because it's also likely to be great for the employees too. Provided we clear the stores of customers.

    And two, how on earth do you clean broccoli in the current environment? That stuff has the surface area of a football field.

    For bus drivers, there should be acrylic shields, which 1) will protect them from other things as well and 2) will permit them to open their side windows and avoid sharing the passengers' air. Win-win, as everyone likes to say.

    >>Two questions: One, why not make this standard for the duration? Granted there will always be some people who can’t go this route, and they’re the perfect customers for the grocery stores which can’t go this route. For the rest of us, this seems like a great solution, and I describe it here because it’s also likely to be great for the employees too. Provided we clear the stores of customers.<<

    As to the first question about grocery stores providing pick up for those online ordering, it has been offered in my big city adjacent suburb by most major stores for a year or more. You now see aisles being clogged with mostly zoomers tugging large rolling bins using hand held tablets for their bin filling/shopping. Not just a pandemic thing. Also I think in the large city many stores have been doing this.

    Others have and still offer home delivery for a nominal fee.

    I can't say how widespread it is but this isn't new, several years or more in place.

  32. Related item possibly on topic.

    Most of us have or will receive mailings from the US Census Bureau telling us that we have to all go online using our handy mailed ID info and fill out an online Census document of some kind.

    They in their propaganda claim this is all “safe” and private due to some Public Law XYZ which is nonsense. Data identified in and online is also data online identified and never removed (probably). So what “guarantees” privacy? Some claim of illegal hacking? The government wouldn’t lie to us?

    Just ask the nisei about how private those 1940 census forms were. Thanks to the US Census they were all quickly identified and interned in concentration camps. So easy!

    Now the Census is constitutionally mandated to apportion Congress members by population. I’m okay with that. Here we live, there are two of us. Period. No names, occupations, etc. are mandated by the constitution. I am constantly encountering historical biographical information by scholars who have used prior detailed Census data (from the 19th century, for example) to verify that such and such subject was living in this particular location. Some privacy.

    Or do we just lie back, close our eyes, and think of Big Brother…

    • Replies: @eD
    OT, but I agree with this. The Census Bureau should do a count of how many people are in a particular location on a given day of the year, every ten years, for the constitutionally mandated purpose of apportionment.

    If other government agencies want to collect statistics and can get that added in their budget, they can and they do so now. A central federal statistical bureau, separate from the Census, would be fine for for this and such agencies exist in other countries. They could use random sampling, which would also help protect the anonymity of the respondents. But the purpose of the census is compromised by adding in all sorts of irrelevant data collection.
    , @Reg Cæsar

    I am constantly encountering historical biographical information by scholars who have used prior detailed Census data (from the 19th century, for example) to verify that such and such subject was living in this particular location. Some privacy.
     
    This is called genealogy. Your right to privacy ends with your life.

    The big jump in census information came with the 1845 New York census. (States used to have their own censuses.) The US Census Bureau followed suit in 1850. That every member of the family is listed, rather than just the head of household followed by chicken scratches in broad age-and-sex ranges, hardly seems damaging, especially after 170 years. It is a godsend to researchers, though.

    Federal law restricts release of individual information for 72 years after each census. 1950's will be opened up in 2022. Those (natives) making their first appearance will range from 72 to 81. This has been in force for decades. If you know of any cases where this has been abused to someone's detriment, do please clue us in. I haven't come across any in 21 years of research.

    (You can, if you wish, pay the Census Bureau about $50 to have your own line on the schedules released to you sooner. Some do this to prove residence or establish eligibility for some benefit or other.)

    In Canada and the UK, the interval is 99 years. A few years back, Canada threatened to close off access completely. Needless to say, family researchers were quite disturbed by this. Perhaps a centenarian or two sighed in relief...

    Asking how many refrigerators or whatever you own is a bit much-- it's essentially free market research for business. The Bureau sent us a questionnaire about homeschooling a year or two back. I don't think we filled it out, but even if we did, wasn't the decennial census and probably won't be released in 2090 or whenever. And we probably won't care if it is. I'll be in my 130s.

    , @Jonathan Mason

    Most of us have or will receive mailings from the US Census Bureau telling us that we have to all go online using our handy mailed ID info and fill out an online Census document of some kind.
     
    I received one of these. They claimed they had already sent me "several" forms in the mail, but that was not true, unless possibly they were disguised as junk mail, but I really doubt that.

    I did go online and fill out their silly questions as accurately as possible. For some reason I felt like putting is some wrong answers to throw them off, but I did my best to give true answers to ambiguous questions, rather than automatically go to prison for contempt of Census.

    In my area quite a few people have signs at the end of their driveway saying things like "intruders will be shot on sight", so I don't know how accurate the census will be.

    Like scanning and bagging your own groceries at Walmart, one feels that one should receive a discount for doing their work for them, but in real life it never seems to work that way.
  33. Speaking of “a class of box-ticking bureaucrats and lawyers similar to HR bureaucrats”, life here in the Peoples Republic is going swimmingly even though they are all home walking their dogs, and playing with their kids.

    The Mexicans show up to mow their grass, the construction guys continue to widen the interstate, FedEx and UPS trucks prowl the streets, the mail gets delivered, the trash gets picked up, there is food in the grocery stores, the power and the internet stay on, etc.

    So just how valuable are all these white collar jobs that aren’t getting done?

  34. @Anon7
    Our president, Donald J. Trump, will never get reelected without more of the black vote and the Hispanic vote than Republican candidates usually get. He knows it, and in the public parts of the government, he constantly shows how he's advancing and helping black people. Just look at the recent celebration of truckers at the White House.

    https://www.youtube.com/watch?v=miImQAY8PiU

    This is pure politics, for consumption by the masses, and if it leads to Trump's reelection - as opposed to the reign of Joe Biden 46 - then I say fine. Since when has a government panel of scientists ever discovered anything scientific?

    Agreed. When you put it that way. If the window dressing has the intended effect that is. If it doesn’t, it’s just greasing the slide a bit more.

  35. @Anon7
    Our president, Donald J. Trump, will never get reelected without more of the black vote and the Hispanic vote than Republican candidates usually get. He knows it, and in the public parts of the government, he constantly shows how he's advancing and helping black people. Just look at the recent celebration of truckers at the White House.

    https://www.youtube.com/watch?v=miImQAY8PiU

    This is pure politics, for consumption by the masses, and if it leads to Trump's reelection - as opposed to the reign of Joe Biden 46 - then I say fine. Since when has a government panel of scientists ever discovered anything scientific?

    Granted they have to have “people of color”–ideally black women–in figurehead positions nowadays, but can’t they find any who aren’t quite so consumed with racist hatred as this harridan? From her pronouncements she doesn’t appear terribly bright, either. I know that finding modern-day negroes who are 1) bright and 2) fair-minded is a needle-in-a-haystack task, but are we not up to it?

    Maybe we should get more Indians to migrate:

    Do testicles make men more vulnerable to coronavirus?
    Study suggests testes harbour Covid-19 giving virus sanctuary from immune system

    Men are dying from coronavirus at twice the rate as women in the UK alone

    The study, carried out by researchers in New York and Mumbai, followed 48 men and 20 women living in Mumbai who had been infected by coronavirus.

    While the women in the study took four days to clear the infection men took 50 per cent longer, needing six days.

    Dr Aditi Shastri, an oncologist at Montefiore Medical Center in the Bronx, carried out the study with her mother Jayanthi Shastri, a microbiologist at the Kasturba Hospital for Infectious Diseases in Mumbai.

    Their work was released ahead of publication on medical website MedRxiv. It has not been peer reviewed.

    (Typical reader comment: “No worries, my wife keeps mine in her purse.”)

    • Replies: @Anon7
    "I know that finding modern-day negroes who are 1) bright and 2) fair-minded is a needle-in-a-haystack task, but are we not up to it?"

    The problem is that every corporation and city, state and national government agency in the USA wants to hire that woman. And the corporations will pay most handsomely.

    "...can’t they find any who aren’t quite so consumed with racist hatred as this harridan?"

    No! They're all trained to do this starting in middle school. Most are sent to "poetry slam" summer camps where they are taught the rhetoric of victimization and race hatred. Believe me, I know whereof I speak.
  36. @Pericles

    Dr. Kizzmekia Corbett, who is described by the National Institutes of Health (NIH) as the “lead” member of the U.S. government team racing to find a coronavirus vaccine

     

    This time the damn racists won't be able to complain about the hagiographic movie being greenlit as we read this.

    She does seem a bit crazy and aggressive though. Is the pressure bad? Better hope 'black don't crack'.

    Good point! Nipping that one in the bud! Now how do we take it back even further so that racists won’t claim she got her degrees and position through affirmative action? Once we have that settled, though, we may need to go back a step further–naah, let’s just kill all the racists. AKA anyone we don’t like. Easier that way.

  37. “stewardesses.”

    Don’t ever change, Steve.

  38. Anonymous[427] • Disclaimer says:
    @The Alarmist

    Maybe somebody could do something similar to help figure out how to get out of this medical and economic hole we’re in?
     
    There's an easier solution, but it is a violation of 18 USC 2385.

    18 U.S. Code § 2385. Advocating overthrow of Government

    U.S. Code

    Whoever knowingly or willfully advocates, abets, advises, or teaches the duty, necessity, desirability, or propriety of overthrowing or destroying the government of the United States or the government of any State, Territory, District or Possession thereof, or the government of any political subdivision therein, by force or violence, or by the assassination of any officer of any such government; or

    Whoever, with intent to cause the overthrow or destruction of any such government, prints, publishes, edits, issues, circulates, sells, distributes, or publicly displays any written or printed matter advocating, advising, or teaching the duty, necessity, desirability, or propriety of overthrowing or destroying any government in the United States by force or violence, or attempts to do so; or

    Whoever organizes or helps or attempts to organize any society, group, or assembly of persons who teach, advocate, or encourage the overthrow or destruction of any such government by force or violence; or becomes or is a member of, or affiliates with, any such society, group, or assembly of persons, knowing the purposes thereof—

    Shall be fined under this title or imprisoned not more than twenty years, or both, and shall be ineligible for employment by the United States or any department or agency thereof, for the five years next following his conviction.

    If two or more persons conspire to commit any offense named in this section, each shall be fined under this title or imprisoned not more than twenty years, or both, and shall be ineligible for employment by the United States or any department or agency thereof, for the five years next following his conviction.

    As used in this section, the terms “organizes” and “organize”, with respect to any society, group, or assembly of persons, include the recruiting of new members, the forming of new units, and the regrouping or expansion of existing clubs, classes, and other units of such society, group, or assembly of persons.

    I don’t advocate it, and I’m too old and worn out to do anything about it if I wanted to. I do predict it will happen-long after I’m dead and buried.

    Anyone trying it now or in the near future would be squashed like a bug. We are dealing with an Empire with nuclear weapons and nuclear submarines and aircraft carriers, attack helicopters, and sizable numbers of elite special warfare types who are amongst the most well equipped and trained combat troops in history, even if our mainline forces are not particularly hardened as a group.

    The difference between what happened in 1861 and what will happen when the Empire implodes will primarily come down to there being one cohesive group splitting off from another cohesive group then and a multi-way, fractious split consisting of numerous tribal and particularist groups squabbling with all the others. When America was all whites and blacks, with most blacks in the South, the War of Northern Aggression was thinkable and winnable. In an “America” with numerous racial, religious and lifestyle groups none of which really like the others, it’ll be a free for all, a veritable lumberjack match.

    Already the system has made mistakes. I’ve met several Chinese, Vietnamese and dot-Indian engineers and programmers working at the Honeywell “bomb plant”, on the critical arming, fuzing and surety systems on US nuclear weapons. I bet that ninety-plus percent of them are completely loyal to the country they hold citizenship in. I’ve seen films of the workers filing into Pantex in Amarillo. A lot of them are mestizo or indio “Hispanics” presumably from Mexico, and I bet many have been here two, three, four, five generations and I have no question that 99 plus percent of them consider themselves 100 percent loyal Americans. Ask anyone in the nuke business if “ninety-nine percent surety” is okay in nuclear surety when it comes to mechanical, electrical, or software matters.

    But that’s not the worst of it, since if anyone in power actually decides to deploy nukes internally things have already gone past the heretofore conceivable. Okay, consider that the Big Four biker outfits-never exactly paragons of criminal competence, they couldn’t even whack Mick Jagger-have substantially penetrated the DMV and vital records departments of several states. Scientology and the ADL have somewhat competently penetrated various law enforcement agencies and I have it on fairly good authority the Scienos have several key assets in at least one of the big military contractors in Wichita. This kind of thing is going to get more and more common. It’s inherent to diversity.

    I was an avid reader of Harold Covington’s novels in which he did a pretty good job of going over the conditions needed to actually ‘pull “it” off’- but where I disagreed with him then and still do is it won’t be one little breakaway province vs. the rest of the world. It’ll be a pool table ball break and the cueball will be some random, unpredictable event, much like today’s coronavirus crisis.

    I’m guessing when it hits, within a year or two the current CONUS (along with parts of West Canada and maybe parts of Mexico) will wind up being four to seven different polities. Who gets what-where the borders are drawn-and who gets the nukes are going to be big questions.

    • Replies: @The Alarmist
    They'll start branding these anti-lockdown rallies as white nationalist racist rallies, and that will wake more than a few people up.

    What amuses me is that you get all these green enviro types who say that we need to get world population back to a sustainable level ... I've heard the number 500 million bandied about, but never any serious discussion about what they intend to do with the other 8 billion souls ... but when the grim reaper appears in the form of COVID-19 to speed the process along, these are the same types who shriek, jump up on a chair, lift their skirts, and insist big gov do something to save them. You'd think they'd embrace their own deaths as doing their part for Mother Gaia.
  39. eD says:
    @Muggles
    Related item possibly on topic.

    Most of us have or will receive mailings from the US Census Bureau telling us that we have to all go online using our handy mailed ID info and fill out an online Census document of some kind.

    They in their propaganda claim this is all "safe" and private due to some Public Law XYZ which is nonsense. Data identified in and online is also data online identified and never removed (probably). So what "guarantees" privacy? Some claim of illegal hacking? The government wouldn't lie to us?

    Just ask the nisei about how private those 1940 census forms were. Thanks to the US Census they were all quickly identified and interned in concentration camps. So easy!

    Now the Census is constitutionally mandated to apportion Congress members by population. I'm okay with that. Here we live, there are two of us. Period. No names, occupations, etc. are mandated by the constitution. I am constantly encountering historical biographical information by scholars who have used prior detailed Census data (from the 19th century, for example) to verify that such and such subject was living in this particular location. Some privacy.

    Or do we just lie back, close our eyes, and think of Big Brother...

    OT, but I agree with this. The Census Bureau should do a count of how many people are in a particular location on a given day of the year, every ten years, for the constitutionally mandated purpose of apportionment.

    If other government agencies want to collect statistics and can get that added in their budget, they can and they do so now. A central federal statistical bureau, separate from the Census, would be fine for for this and such agencies exist in other countries. They could use random sampling, which would also help protect the anonymity of the respondents. But the purpose of the census is compromised by adding in all sorts of irrelevant data collection.

  40. In relation to the pandemic —

    the reality is that the more infected the better the analysis and treatment — individuation and rating anyone’s ID is not very helpful in such circumstances. If one takes this seriously, then we are well beyond Mary Mallon calculates.

  41. @Anonymous
    18 U.S. Code § 2385. Advocating overthrow of Government

    U.S. Code


    Whoever knowingly or willfully advocates, abets, advises, or teaches the duty, necessity, desirability, or propriety of overthrowing or destroying the government of the United States or the government of any State, Territory, District or Possession thereof, or the government of any political subdivision therein, by force or violence, or by the assassination of any officer of any such government; or

    Whoever, with intent to cause the overthrow or destruction of any such government, prints, publishes, edits, issues, circulates, sells, distributes, or publicly displays any written or printed matter advocating, advising, or teaching the duty, necessity, desirability, or propriety of overthrowing or destroying any government in the United States by force or violence, or attempts to do so; or

    Whoever organizes or helps or attempts to organize any society, group, or assembly of persons who teach, advocate, or encourage the overthrow or destruction of any such government by force or violence; or becomes or is a member of, or affiliates with, any such society, group, or assembly of persons, knowing the purposes thereof—

    Shall be fined under this title or imprisoned not more than twenty years, or both, and shall be ineligible for employment by the United States or any department or agency thereof, for the five years next following his conviction.

    If two or more persons conspire to commit any offense named in this section, each shall be fined under this title or imprisoned not more than twenty years, or both, and shall be ineligible for employment by the United States or any department or agency thereof, for the five years next following his conviction.

    As used in this section, the terms “organizes” and “organize”, with respect to any society, group, or assembly of persons, include the recruiting of new members, the forming of new units, and the regrouping or expansion of existing clubs, classes, and other units of such society, group, or assembly of persons.

     

    I don't advocate it, and I'm too old and worn out to do anything about it if I wanted to. I do predict it will happen-long after I'm dead and buried.

    Anyone trying it now or in the near future would be squashed like a bug. We are dealing with an Empire with nuclear weapons and nuclear submarines and aircraft carriers, attack helicopters, and sizable numbers of elite special warfare types who are amongst the most well equipped and trained combat troops in history, even if our mainline forces are not particularly hardened as a group.

    The difference between what happened in 1861 and what will happen when the Empire implodes will primarily come down to there being one cohesive group splitting off from another cohesive group then and a multi-way, fractious split consisting of numerous tribal and particularist groups squabbling with all the others. When America was all whites and blacks, with most blacks in the South, the War of Northern Aggression was thinkable and winnable. In an "America" with numerous racial, religious and lifestyle groups none of which really like the others, it'll be a free for all, a veritable lumberjack match.

    Already the system has made mistakes. I've met several Chinese, Vietnamese and dot-Indian engineers and programmers working at the Honeywell "bomb plant", on the critical arming, fuzing and surety systems on US nuclear weapons. I bet that ninety-plus percent of them are completely loyal to the country they hold citizenship in. I've seen films of the workers filing into Pantex in Amarillo. A lot of them are mestizo or indio "Hispanics" presumably from Mexico, and I bet many have been here two, three, four, five generations and I have no question that 99 plus percent of them consider themselves 100 percent loyal Americans. Ask anyone in the nuke business if "ninety-nine percent surety" is okay in nuclear surety when it comes to mechanical, electrical, or software matters.

    But that's not the worst of it, since if anyone in power actually decides to deploy nukes internally things have already gone past the heretofore conceivable. Okay, consider that the Big Four biker outfits-never exactly paragons of criminal competence, they couldn't even whack Mick Jagger-have substantially penetrated the DMV and vital records departments of several states. Scientology and the ADL have somewhat competently penetrated various law enforcement agencies and I have it on fairly good authority the Scienos have several key assets in at least one of the big military contractors in Wichita. This kind of thing is going to get more and more common. It's inherent to diversity.

    I was an avid reader of Harold Covington's novels in which he did a pretty good job of going over the conditions needed to actually 'pull "it" off'- but where I disagreed with him then and still do is it won't be one little breakaway province vs. the rest of the world. It'll be a pool table ball break and the cueball will be some random, unpredictable event, much like today's coronavirus crisis.


    I'm guessing when it hits, within a year or two the current CONUS (along with parts of West Canada and maybe parts of Mexico) will wind up being four to seven different polities. Who gets what-where the borders are drawn-and who gets the nukes are going to be big questions.

    They’ll start branding these anti-lockdown rallies as white nationalist racist rallies, and that will wake more than a few people up.

    What amuses me is that you get all these green enviro types who say that we need to get world population back to a sustainable level … I’ve heard the number 500 million bandied about, but never any serious discussion about what they intend to do with the other 8 billion souls … but when the grim reaper appears in the form of COVID-19 to speed the process along, these are the same types who shriek, jump up on a chair, lift their skirts, and insist big gov do something to save them. You’d think they’d embrace their own deaths as doing their part for Mother Gaia.

  42. @Muggles
    Related item possibly on topic.

    Most of us have or will receive mailings from the US Census Bureau telling us that we have to all go online using our handy mailed ID info and fill out an online Census document of some kind.

    They in their propaganda claim this is all "safe" and private due to some Public Law XYZ which is nonsense. Data identified in and online is also data online identified and never removed (probably). So what "guarantees" privacy? Some claim of illegal hacking? The government wouldn't lie to us?

    Just ask the nisei about how private those 1940 census forms were. Thanks to the US Census they were all quickly identified and interned in concentration camps. So easy!

    Now the Census is constitutionally mandated to apportion Congress members by population. I'm okay with that. Here we live, there are two of us. Period. No names, occupations, etc. are mandated by the constitution. I am constantly encountering historical biographical information by scholars who have used prior detailed Census data (from the 19th century, for example) to verify that such and such subject was living in this particular location. Some privacy.

    Or do we just lie back, close our eyes, and think of Big Brother...

    I am constantly encountering historical biographical information by scholars who have used prior detailed Census data (from the 19th century, for example) to verify that such and such subject was living in this particular location. Some privacy.

    This is called genealogy. Your right to privacy ends with your life.

    The big jump in census information came with the 1845 New York census. (States used to have their own censuses.) The US Census Bureau followed suit in 1850. That every member of the family is listed, rather than just the head of household followed by chicken scratches in broad age-and-sex ranges, hardly seems damaging, especially after 170 years. It is a godsend to researchers, though.

    Federal law restricts release of individual information for 72 years after each census. 1950’s will be opened up in 2022. Those (natives) making their first appearance will range from 72 to 81. This has been in force for decades. If you know of any cases where this has been abused to someone’s detriment, do please clue us in. I haven’t come across any in 21 years of research.

    (You can, if you wish, pay the Census Bureau about $50 to have your own line on the schedules released to you sooner. Some do this to prove residence or establish eligibility for some benefit or other.)

    In Canada and the UK, the interval is 99 years. A few years back, Canada threatened to close off access completely. Needless to say, family researchers were quite disturbed by this. Perhaps a centenarian or two sighed in relief…

    Asking how many refrigerators or whatever you own is a bit much– it’s essentially free market research for business. The Bureau sent us a questionnaire about homeschooling a year or two back. I don’t think we filled it out, but even if we did, wasn’t the decennial census and probably won’t be released in 2090 or whenever. And we probably won’t care if it is. I’ll be in my 130s.

  43. @Anon
    From seeing that Los Angeles birthday party video, I'd say that No. 3, ignoring rules and government, looms large. There's an instant sense among blacks that any assertion of control over them is "racis'."

    Even the Daily Mail is running breathless headlines about the racial disparity in cobid mortality. This will not stand!

  44. @Muggles
    Related item possibly on topic.

    Most of us have or will receive mailings from the US Census Bureau telling us that we have to all go online using our handy mailed ID info and fill out an online Census document of some kind.

    They in their propaganda claim this is all "safe" and private due to some Public Law XYZ which is nonsense. Data identified in and online is also data online identified and never removed (probably). So what "guarantees" privacy? Some claim of illegal hacking? The government wouldn't lie to us?

    Just ask the nisei about how private those 1940 census forms were. Thanks to the US Census they were all quickly identified and interned in concentration camps. So easy!

    Now the Census is constitutionally mandated to apportion Congress members by population. I'm okay with that. Here we live, there are two of us. Period. No names, occupations, etc. are mandated by the constitution. I am constantly encountering historical biographical information by scholars who have used prior detailed Census data (from the 19th century, for example) to verify that such and such subject was living in this particular location. Some privacy.

    Or do we just lie back, close our eyes, and think of Big Brother...

    Most of us have or will receive mailings from the US Census Bureau telling us that we have to all go online using our handy mailed ID info and fill out an online Census document of some kind.

    I received one of these. They claimed they had already sent me “several” forms in the mail, but that was not true, unless possibly they were disguised as junk mail, but I really doubt that.

    I did go online and fill out their silly questions as accurately as possible. For some reason I felt like putting is some wrong answers to throw them off, but I did my best to give true answers to ambiguous questions, rather than automatically go to prison for contempt of Census.

    In my area quite a few people have signs at the end of their driveway saying things like “intruders will be shot on sight”, so I don’t know how accurate the census will be.

    Like scanning and bagging your own groceries at Walmart, one feels that one should receive a discount for doing their work for them, but in real life it never seems to work that way.

    • Replies: @Anonymous
    I know several people who put complete horseshit down on their census, both to get more local apportionment and give people a chuckle decades later. I’ll let you know if any get backlash from it.
  45. I only scanned quickly, but here’s a front page, above-the-fold sub-headline and excerpt from my local newspaper this morning:

    “[XYZ County] Public Health officer also works for [ABC] Health”

    “[XYZ] County Public Health’s medical director, who spoke out against public release of details on cases at long-term care facilities, also serves as a top medical officer for [ABC] Health, which operates two local facilities reporting coronavirus cases.”

    Good grief! What happens when the public health folks and the private medical interest folks are the same people?

  46. @Mr McKenna
    Granted they have to have "people of color"--ideally black women--in figurehead positions nowadays, but can't they find any who aren't quite so consumed with racist hatred as this harridan? From her pronouncements she doesn't appear terribly bright, either. I know that finding modern-day negroes who are 1) bright and 2) fair-minded is a needle-in-a-haystack task, but are we not up to it?

    Maybe we should get more Indians to migrate:


    Do testicles make men more vulnerable to coronavirus?
    Study suggests testes harbour Covid-19 giving virus sanctuary from immune system

    Men are dying from coronavirus at twice the rate as women in the UK alone

    https://i.dailymail.co.uk/1s/2020/04/18/19/27265796-8225245-image-a-51_1587234568747.jpg

    The study, carried out by researchers in New York and Mumbai, followed 48 men and 20 women living in Mumbai who had been infected by coronavirus.

    While the women in the study took four days to clear the infection men took 50 per cent longer, needing six days.

    Dr Aditi Shastri, an oncologist at Montefiore Medical Center in the Bronx, carried out the study with her mother Jayanthi Shastri, a microbiologist at the Kasturba Hospital for Infectious Diseases in Mumbai.

    Their work was released ahead of publication on medical website MedRxiv. It has not been peer reviewed.

     

    (Typical reader comment: "No worries, my wife keeps mine in her purse.")

    “I know that finding modern-day negroes who are 1) bright and 2) fair-minded is a needle-in-a-haystack task, but are we not up to it?”

    The problem is that every corporation and city, state and national government agency in the USA wants to hire that woman. And the corporations will pay most handsomely.

    “…can’t they find any who aren’t quite so consumed with racist hatred as this harridan?”

    No! They’re all trained to do this starting in middle school. Most are sent to “poetry slam” summer camps where they are taught the rhetoric of victimization and race hatred. Believe me, I know whereof I speak.

  47. @Jonathan Mason

    Most of us have or will receive mailings from the US Census Bureau telling us that we have to all go online using our handy mailed ID info and fill out an online Census document of some kind.
     
    I received one of these. They claimed they had already sent me "several" forms in the mail, but that was not true, unless possibly they were disguised as junk mail, but I really doubt that.

    I did go online and fill out their silly questions as accurately as possible. For some reason I felt like putting is some wrong answers to throw them off, but I did my best to give true answers to ambiguous questions, rather than automatically go to prison for contempt of Census.

    In my area quite a few people have signs at the end of their driveway saying things like "intruders will be shot on sight", so I don't know how accurate the census will be.

    Like scanning and bagging your own groceries at Walmart, one feels that one should receive a discount for doing their work for them, but in real life it never seems to work that way.

    I know several people who put complete horseshit down on their census, both to get more local apportionment and give people a chuckle decades later. I’ll let you know if any get backlash from it.

  48. @Steve Sailer
    "what do we expect to learn from data mining such a small subset of the CV cases ? we will not learn which occupations are at the highest risk from the confirmed cases, since the confirmed cases are those who are very sick which are mostly the elderly who are not employed."

    It's called multiple regression. You adjust for age, sex, race, BMI, zip code, health problems, etc. A big HMO's database on its patients has all that information.

    It’s called multiple regression. You adjust for age, sex, race, BMI, zip code, health problems, etc. A big HMO’s database on its patients has all that information.

    If it were that simple, there would not be a replicability crisis in medical quant research – specifically at the intersection of medicine and pharmacology (and also at the intersection of psych and pharma).

    Multiple regression might generate a better ‘one-shot’ fit, which sounds like something that would interest researchers chasing publications.

    Add a regressor, R² goes up: why not add another 20?

    However every parameter has its own uncertainty – and when model-wide uncertainty needs to be quantified, it becomes something that needs to be part of sensitivity analysis (if sensitivity analysis is to be worthy of the name).

    And people need to stop saying ‘adjust for X₁ etc’ as if that’s what’s being done in practice. What’s reported as some or other risk ‘adjusted for X₁‘ is just a naïve attempt at evaluating marginal effects.

    To paraphrase:

    The documentation for the R margins package includes quite a good discussion as to why that’s the case.

    Quantifying the uncertainty contained in estimates of marginal effects is far more important.

    Also… which marginal effect? The marginal effect at ‘representative values’ of the other parameters? At ‘representative values’ of the regressors? At ‘representative values’ of the output variable? That last one’s a doozy, because people forget: in general, even linear models are not bijective.

    It invariably turns out that adding a regressor that improves R² – i.e. gives a better fit to the existing data – still causes more uncertainty. And when you chuck 25 or 30 variables into the ‘explanatory’ variables (as Petrelli et al did), the uncertainty in the model gets ridiculous and ‘marginal effects’ of one variable become a cloud.

    This could be investigated diligently and the results reported in data appendices, but that never happens. That’s particularly egregious in generalised linear models, since regressors are supposed to be uncorrelated and so the convolutions required to systematic sensitivity analysis are straightforward (and computationally easy) – but it’s never done.

    And even the ‘regressors are supposed to be uncorrelated‘ question is not settled if investigators use tests that are inappropriate for the data type.

    .

    When a multiple regression is performed, there are conditions on the regressors that must be satisfied (for a linear model, it’s the Gauss-Markov conditions).

    And underpinning that, is the meta-problem of testing if the conditions are satisfied.

    Learning to test if the relevant conditions are satisfied – by itself – involves enough material for a challenging, dense one-semester subject at (roughly) US first-year PhD level. It includes learning which tests (there are many) are relevant for different data types, and the shortcomings of tests when an inappropriate tests is used.

    Pick the wrong test and you might as well not bother.

    This was – almost certainly – the reason for the silly Odds Ratios for age groups in the Petrelli paper a few days ago. They used a test for collinearity that is known to be next-to-useless for categorical variables (VIF). (That paper still has the makings of an influential paper, if they fix the test they used for collinearity and present age-group data for to-ICU and deaths).

    A cynic might put that down to ‘test-seeking’ – selecting a test that gives the desired answer (in this case, ‘no collinearity’).

    I put it down to them simply not properly understanding what they’re doing. Even in places where people are paid to be right (e.g., quant finance), very few people give UQ the respect it deserves.

    If the conditions on regressors are not satisfied, critical values for hypothesis tests differ from their asymptotic values. In other words, the value of the test associated with the desired significance level will be different from tabulated values.

    Worse, they differ in unpredictable ways, and are not ameliorated by increased sample size.

    The way to deal with this is to ‘bootstrap’ an estimate of the distribution of the hypothesis test statistic under the null hypothesis (by Monte Carlo, usually) for some models the same thing is required under some specific alternative because the distribution of the test changes.

    .

    People often use the term GIGO, to point out that data quality matters. That’s just a bromide, because nobody disagrees with it.

    There’s stuff being done between GI and GO: if that stuff’s not done properly, then there’s a risk of PIGO (perfection in, garbage out) – where the data can be flawless (in terms of its relevance, collection, filtering and so forth) but the technique involved in transforming inputs to outputs is flawed in important respects.

    inb4: “These people are public health experts.” That’s what they’re referred to, but they have got almost every major public health issue wrong for the last 200 years – they even had to be dragged kicking and screaming when the issue was basic surgical hygeine (Semmelweiss) and basic epidemiology (Snow and the famous ‘pump handle’: when he did that, the establishment consensus was still ‘miasma theory’).

    Fauci’s a ‘public health expert’ who got hetero-AIDS completely wrong (Oprah-level wrong), and changed his evaluation of covid19 from “The US has nothing to worry about” to “Existential threat: shut down the economy” in the space of two weeks.

    Ferguson’s a public health expert who got H5N1 and vCJD completely wrong.

  49. Speaking of datamining: a recent publication makes clear that the strains of SARS-nCoV2 that infected different populations, are distinct clades and that the strain that affected Wuhan is a sibling of, not an ancestor of, the other main strain infecting Italy.

    Turns out that as far as can be established (and making some assumptions about mutation rates), there is some likelihood that the ancestor was doing the rounds – in Yunnan – about 1,000 miles from Wuhan… as early as mid-September 2019.

    The publication is dated April 8th – 3 days after I wrote this in a comment (I fucking nailed it – if this problem was Jesus and I was a centurion, it would have been impossible to get Jewboy down off the cross) ->

    If there was an undetected pathogen doing the rounds in Italy (or China, or both) that predates the genomes sequenced from people from the Wuhan wet market, it’s better to know about it than to not know about it.

    nextstrain’s ‘family tree’ for 2019nCoV is really interesting: there aren’t many nucleotide mutations at each node, and the preponderance of European genomes sequenced to date are in 1 or 2 clades that are genetically well-separated from the preponderance of Chinese clades.

    Given the related-ness of the European genomes, it’s not entirely out of the question that their parent node represents a different underlying genome that is not-very-different to the existing ‘root’ node. (Also worth knowing – but a days-long project in itself – would be examining the ‘Italian’ genomes and identifying which ones are actually taken from Italians, as opposed to Chinese people in Italy).

    [snip]…

    Not for nothin’: the very first two Chinese sequences (from samples in Wuhan, taken 2 days apart) are in different clades. Wuhan-Hu-1/2019 (sample date 26 Dec 2019) differs from Wuhan/IPBCAMS-WH-01/2019 (sample date 24 Dec 2019) by mutations in 3 nucleotides and 3 amino acids.

    Wuhan-Hu-1/2019 is assumed to be the ‘OG’ WuFlu because other nearby child nodes have identifiable nucleotide and AA changes relative to it, and don’t have the mutations found in Wuhan/IPBCAMS-WH-01/2019. Perfectly sensible.

    Under that assumption, an earlier variant that differed from ‘OG’ WuFlu by 1 nucleotide mutation, would be automatically classed as the root node of a descendant clade from OG WuFlu, rather than a sibling from a common ancestor. That’s a harder row to hoe.

    Reference:

    Phylogenetic network analysis of SARS-CoV-2 genomes

Comments are closed.

Subscribe to All Steve Sailer Comments via RSS
PastClassics
The unspoken statistical reality of urban crime over the last quarter century.
Which superpower is more threatened by its “extractive elites”?
How a Young Syndicate Lawyer from Chicago Earned a Fortune Looting the Property of the Japanese-Americans, then Lived...
Becker update V1.3.2