Can you please link that study? Is it the one where hospital employees were observed from more than a year ago? I vaguely remember there was one from Thailand that was relevant at the time.
My methods are probably lacking severely from the perspective of a statistician.
However taking an honest look at the data, I don't see any of them co…
Can you please link that study? Is it the one where hospital employees were observed from more than a year ago? I vaguely remember there was one from Thailand that was relevant at the time.
My methods are probably lacking severely from the perspective of a statistician.
However taking an honest look at the data, I don't see any of them coming to a different conclusion than I have. They could end up saying it's 500,000 or 1,500,000 damaged hearts and I might not even disagree.
The basic idea is that data is sufficiently random for symptoms to be represented equiproportionally. For diagnoses and other medical concepts it is different.
Can you please link that study? Is it the one where hospital employees were observed from more than a year ago? I vaguely remember there was one from Thailand that was relevant at the time.
My methods are probably lacking severely from the perspective of a statistician.
However taking an honest look at the data, I don't see any of them coming to a different conclusion than I have. They could end up saying it's 500,000 or 1,500,000 damaged hearts and I might not even disagree.
The basic idea is that data is sufficiently random for symptoms to be represented equiproportionally. For diagnoses and other medical concepts it is different.
I think Igor Chudov is a statistican. Link the post to him on his substack.
Thanks I will do that