Politics “Trumping” Science

Interesting article how the variables and unknowns introduced by “conservative” and Trump attacks on science can skew epidemiological models.

The article does not make this point but Trump’s predictions of an early end to the need for social distancing (a “miracle”) is a self-defeating prophecy. The more people who believe it, the less likely it is to happen.

23 thoughts on “Politics “Trumping” Science

  1. Modeling is limited in its usefulness by the assumptions in the models and the quality of data input.

    For any predictive model to be of any use, you need a pretty firm estimate of how many people are infected, how many have become immune and how many people will be exposed. We don’t have any of those.

    Testing for the virus is only done for those with significant symptoms and we have no reliable estimate of how many simply got sick and got well without reporting it or being tested.

    We don’t have the data to make very useful models using real numbers, so they have to get by with estimates. You can then hindcast the model to refine your estimates.

    Modeling on estimates of data is not inherently fraudulent, but failing to disclose the weakness of the predictions is. .

    Liked by 2 people

    1. “self-defeating prophecy“

      You make some good points and the comments coming out of trump’s ignorant mouth are making it even harder to effectively model as the data becomes skewed to the point of analyst having to guess at how much effect his ignorant comments contribute to the behavior of the cult who continue to believe him.

      I’ll reiterate my refrain of “he needs to SHFU” and fund the medical community to the highest level humanly possible.

      Liked by 2 people

      1. RE: “the comments coming out of trump’s ignorant mouth are making it even harder to effectively model as the data becomes skewed to the point of analyst having to guess at how much effect his ignorant comments contribute to the behavior of the cult who continue to believe him.”

        Not likely. Let’s not confuse correlation and causation. More to the point, the data we need (and don’t have) for better modeling is of a type that isn’t much affected by personal behavior, beliefs or politics. For example, either a person has or doesn’t have coronavirus antibodies (indicating immunity).

        Liked by 1 person

        1. RE: “You are clueless…..”

          Maybe so. So tell us, how do you imagine that the president’s public statements will alter statistical modelling of the epidemic?

          Like

          1. 3 semesters of study and a program you may struggle to get in to.

            Sometimes a google search will not provide sufficient information to either understand nor appreciate the complexity of a situation.

            The President also thinks that all the career professionals know less than he does and WE as a nation will/are paying the price.

            Liked by 2 people

          2. RE: “3 semesters of study and a program you may struggle to get in to.”

            I see. Well, if you ever choose to show us that you actually know what you’re talking about, you’re more than welcome to do it. Until then, the rest of us might as well assume that you don’t.

            Liked by 2 people

          3. …”the rest of us might as well assume that you don’t.”

            Speak for yourself. Don’t ASSUME all of us think he’s mistaken in any way. Just you and you little cadre of Trumpists.

            Like

          4. @Roberts

            Re; “So tell us, how do you imagine that the president’s public statements will alter statistical modelling of the epidemic?”

            Okay.

            When trying to model the spread of the pandemic one of the most significant variables is how many people will be infected by each infected person. If each person infects many the pandemic will explode. If each person infects 1 or less than 1 it will implode. Are you with me so far?

            That variable is determined by how people behave. Still with me?

            If people listen to the President and behave as if this is no big deal, no worse than a cold, easily cured by “game changer” drugs, etc then they will have more contact with other people, be less scrupulous about washing hands, coughing, etc. And that key variable will be a greater number,

            On the other hand, if they ignore Trump’s bullshit and listen to the scientists they will self quarantine, maintain social distance, avoid other people, wash carefully, etc. and that variable will be a lower number.

            So, a key driver of any prediction algorithm has to be data – or an assumption – on how people react to the President’s public statements.

            Got it now?

            Liked by 2 people

          5. RE: “Got it now?”

            No. Instead, I’m laughing. You are talking about the progress of the disease, not the validity of the statistical modeling, which is what I called Jimmie on. He, and you, seem to think that because people react to the president, the mortality rate and — more significantly — the infection rate will not be measured properly.

            Like

          6. @Roberts

            Laughing?

            Laugh your silly head off. Obviously not man enough to admit you asked a silly question with an obvious answer which I dutifully provided. And even though I deliberately put it in the simplest possible terms you totally did not understand it. I was NOT talking about the progress of the disease. I was taking about the variable required in statistical models to represent the Trump effect on effective social distancing.

            Liked by 2 people

    2. If you saw my post about Peggy Noonan’s WSJ piece, we can see there is still a major problem on tracking the disease. She showed many symptoms and was tested March 17. As of the article posting, March 26, she had not received any information.

      Liked by 2 people

      1. Unless there’s been some auful SNAFU, Noonan’s test results will come back eventually. To Dr. Tabor’s point, however, knowing whether she tested positive or negative for coronavirus won’t tell her much, and won’t add much to our understanding of the virus. Its the data we don’t have, and which the original test doesn’t produce, that we really want.

        Like

        1. If she knew she was positive she could alert all those she has had contact with for the last few weeks. Who could then be more careful about their own social interactions.

          If what your saying that the results won’t tell her much or add to our understanding of the virus, then why bother. Just hospitalize the ones who can barely breathe, bury the ones who can’t and ignore the rest.

          I know there is some work on people with antibodies, but in order to find them, we need to test everybody and find those who test positive but show no symptoms after incubation period.

          Liked by 2 people

        2. RE: “If what your saying [is] that the results won’t tell her much or add to our understanding of the virus, then why bother?”

          The positive/negative test is not useless. It distinguishes between flu and Covid-29, for example. It is also all we had when the crisis started. Now we know more than we did, and that means we have new questions, some of them more fundamental, such as determining the infection rate.

          RE: “I know there is some work on people with antibodies, but in order to find them, we need to test everybody and find those who test positive but show no symptoms after incubation period.”

          We don’t need 100% testing, we need the right kind of testing combined with statistically significant results. Same as election polling.

          Like

    3. @Tabor

      “Modeling is limited in its usefulness by the assumptions in the models and the quality of data input.”

      Certainly true. But the point of this story is that it is made even more difficult by people like Trump working at cross purposes to the best available science as imperfect as it may be.

      A big part of Trump failure of leadership is the failure to get a better quality of data sooner. While other countries have used widespread testing to corral the virus, we have not. You can blame FDA or CDC or whomever you want, but Trump is the man in charge. Even worse he has lied about the availability of testing repeatedly. What purpose do you think those lies serve?

      Liked by 2 people

    4. I gather you did not read the article.It addresses the modeling weaknesses and does not say that the epidemiologists are failing to disclose or committing fraud by not doing so.

      It also points out that major, influential people like Ingraham and WSJ are casting doubts on the original Ferguson models that predicted 2.2 million US deaths 500,000 UK fatalities. They say he has changed his prediction without saying why.

      “Ferguson had not revised his projections in his testimony, which he made clear in interviews and Twitter. His earlier study had made clear the estimate of 500,000 deaths in Britain and 2.2 million in the United States projected what could happen if both took absolutely no action against the coronavirus. The new estimate of 20,000 deaths in Britain was a projected result now that Britain had implemented strict restrictions, which this week came to include a full lockdown.”

      Now this is truly “fake news” that can cause people to die. Ingraham, et. al., are responsible for trying to undermine what epidemiologists, scientists and medical professionals are saying and doing for pure partisan political reasons.

      Shame on them. Time for FOX to shut up. They did a lot of damage earlier and they are trying to do the same again.

      Not my opinion, that is a fact.

      Liked by 3 people

      1. @Len

        “ that is a fact.”

        Not to be patronizing, but you seem to get the essentials of the data and analytics problem with shit for brains spewing misinformation on a daily basis.

        I taught Research Methods to PHD student for a number of years and the damage being done by trump’s comments is yet to be determined, but there is NO question people will die that did not need to because of his arrogant stupidity.

        Liked by 2 people

          1. Of course one too many; my greater concern are the people continuing to interact with each other based on the Administration’s minimizing of the risks involved.

            Hell, even Fauci did it last night…

            Liked by 2 people

  2. There has been chatter about scientists not being the best determinants of how to proceed with the COVID-19 and its economic impacts.

    The FOX noise is playing the role of megaphone in disparaging Ferguson in this particular case. Disparaging by lying about why he changed his numbers, not his mind.

    The Challenger disaster in 1985 was in large part due to the pressure by the Reagan administration to ramp up the Star Wars program via a popular Space Shuttle operation.

    The failure of the “O” rings was predicated by Thiokol engineers who tried to stop the launch. But they were overridden by the political echelons at the top.

    Politics is flexible and a matter of compromises. Science is not about compromise. Not that it is always right, but in life and death situations, like a shuttle launch which was always very dangerous or a rampant pandemic that is currently very dangerous, I would suggest that science should prevail until proven wrong.

    And that is not even the case here. Science did not change its forecast. It altered the prediction based on the recommended actions that science put forth.

    So again I say, “FOX shut up and play music or something.”

    Liked by 3 people

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s