Thursday Throughput: Solar Boil Edition

Avatar

Michael Siegel

Michael Siegel is an astronomer living in Pennsylvania. He is on Twitter, blogs at his own site, and has written a novel.

Related Post Roulette

22 Responses

  1. Avatar DensityDuck
    Ignored
    says:

    [ThTh10] don’t worry, the antivaxers are ready to go on retransmitting HPV. It’s hard work keeping viruses from going extinct, but they’re willing to do their part!

    Although one does have to consider what the “infection parties” are going to look like for that one.Report

  2. Avatar veronica d
    Ignored
    says:

    [ThT7] – A testable prediction from modern physics? How odd. I didn’t know we did those anymore.Report

  3. Avatar Mike Schilling
    Ignored
    says:

    C’mon, we all know ThTh11 is Crackerjacks.Report

  4. fillyjonk fillyjonk
    Ignored
    says:

    ThTh1: a case maybe of “when the only tool you have is a hammer, everything looks like a nail” but I’ve been reading some Bayesian statistics and I think a person could Bayes the heck out of those numbers to demonstrate that the false-positive rate for “sentinel injuries” being abuse in very small children is high and so maybe CPS (or whatever) shouldn’t be so quick on the draw, but….I suspect this is one of those cases where emotions runs high, and “protecting the children” even if it means breaking up a family where nothing is going wrong and subjecting innocent parents to one of the worst suspicions will just….continue.Report

    • Avatar Michael Siegel in reply to fillyjonk
      Ignored
      says:

      Someone did that. It depends a lot on what you assume the rate of child abuse is. But for, say, 2%, you get five times as many false positive as real ones.Report

    • Avatar Road Scholar in reply to fillyjonk
      Ignored
      says:

      As Michael notes, the base rate is key to that analysis. A bit of anecdata: around 35 years ago my nephew, a toddler, experienced a broken arm. CPS investigated and while the injury was never adequately explained in my vuew, my brother and SIL were cleared or at least there was insufficient evidence to proceed against them.

      Fast forward 30 years and it turns out SIL is schizophrenic af. Like hospitalized and heavily medicated. I’d always thought of her as weird, hyper-religious was what I saw, but it seems her mental illness was characterized by screaming fits of rage, including physical violence, that I wasn’t aware of. Well that changes everything about my nephew’s mysterious injury doesn’t it?

      Thankfully, she’s doing better now but I had no idea what my brother was going through. It was all hush-hush family secret stuff that was even hidden from me (Honestly, I’m not real close with my siblings).

      All of which is to say that figuring out what’s what, even when you’re relatively close to the situation, isn’t always easy. Trying to get a handle on population statistics seems fraught.Report

      • Avatar Oscar Gordon in reply to Road Scholar
        Ignored
        says:

        Honestly, obscuring the base rate is one of the great ways to lie with P&S. Be it abuse, or illness/poisoning*, or injury/death** rates.

        * Michaels post about RoundUp & NHL; any story about the supposed toxicity of any given substance that avoids talking about dose & L/D rates.

        ** The infamous gun control canard about a person being 21x more likely to be injured or killed with a firearm if there was one in the house completely ignored the base rate (along with all manner of confounding factors).Report

        • Avatar Road Scholar in reply to Oscar Gordon
          Ignored
          says:

          The worst sin is getting the correlation backwards. I don’t doubt that these “sentinel injuries” are present in a large fraction of confirmed cases of child abuse. But that still doesn’t tell us anything really about the likelihood in the other direction. And I don’t think it’s so much a matter of “obscuring” the base rate as just not knowing the base rate. And definitions come into play: one person’s abusive beating is another’s loving discipline. It’s not a clear binary.Report

          • Avatar Oscar Gordon in reply to Road Scholar
            Ignored
            says:

            If you can say that Y causes X to Z times more likely, then you have to have a base rate for X to occur. Maybe it’s just a SWAG at the base rate, but you still have to have something.

            And if you are not telling people what that base rate is, then I have to wonder why.

            Also, leaving the base rate out of the discussion helps to get correlation and causation backwardsReport

    • Avatar InMD in reply to fillyjonk
      Ignored
      says:

      The law/public policy when it comes right down to it is really bad at science and statistics. Frankly so is the voting public. It lends itself to all kinds of overreaction and the ability to put an agenda in a scientific gloss that the ultimate decision maker has no real business assessing.

      Maybe there’s merit to these techniques but sounds to me more like the crime lab forensic stuff that’s discredited every time someone halfway objective tests it but which nevertheles remains admissible in courtrooms across the land.Report

  5. Avatar Oscar Gordon
    Ignored
    says:

    ThTh6: I love stories like that, not just the guy escaping the poverty and making good, but discoveries that upset the apple cart.Report

  6. Avatar Jaybird
    Ignored
    says:

    ThTh9: Dude. That’s awesome.Report

    • Avatar George Turner in reply to Chip Daniels
      Ignored
      says:

      A tweet saying: (I have no idea how to make tweets show up right)

      The EU is planning to implement laws that make historic rebuilding illegal. “When new parts/elements are necessary, a project shall use contemporary design adding new value and/or use while respecting the existing ones.” EU is literally staffed by demons.

      I guess everything over there will be Soviet style Corbusier from now on, or glass strip malls.Report

  7. Avatar Mike Schilling
    Ignored
    says:

    [ThTh6] A guy who grew up in a trailer park, and earned a PhD without either mastering the obstacle course of grades, test scores and leadership activities or taking on crippling student loans.Report

  8. Avatar Oscar Gordon
    Ignored
    says:

    Science and Mathematics related: The Babylonians were doing Trig 1500 years before the Greeks supposedly discovered it.Report

    • Avatar veronica d in reply to Oscar Gordon
      Ignored
      says:

      I’m pretty sure we’ve discovered evidence the Babylonians had basic trig before this. I distinctly remember reading in one of my math history texts that we had base-60 sine and cosine tablets from Mesopotamia, which I expect were Babylonian.

      Maybe when I have some free time I’ll check, but it’s well understood that both the Egyptians and Babylonians had a pretty solid base of practical math well before the Greeks. The real advance of the Greeks was the idea of “pure” axiomatic math.Report

    • Avatar Michael Cain in reply to Oscar Gordon
      Ignored
      says:

      A friend sent me a piece the other day with someone’s clever geometric way to find roots of quadratic equations without the quadratic formula. That article finished up with a statement to the effect of see how much simpler this is, so this is how we should be teaching it. This one sneaks in the same idea.

      As I repeat, ad nauseam, there’s a reason things are taught the way they are: everything from Algebra 1 through the first semester of differential equations and linear algebra is the basic math sequence for engineers and a variety of sciences. We teach algebra the way we do because we need it that way later. We teach trigonometry the way we do because we need it that way later.Report

      • Avatar Mike Schilling in reply to Michael Cain
        Ignored
        says:

        E.g we have trig functions in radians and logs base e because that makes their derivatives simpler.Report

        • Avatar Michael Cain in reply to Mike Schilling
          Ignored
          says:

          I was thinking more that we teach trig based on angle — which seems to annoy the author of the original article — in order to get (in particular) sine and cosine functions that are periodic and defined everywhere (plus have a variety of other useful properties). Down the road a ways, replace the angle variable with a time variable and you get an enormously important tool in science and engineering. Replace angle with a two-dimensional position function and you get a critical tool for image compression.Report

Leave a Reply

Your email address will not be published. Required fields are marked *