1/ Say I show you casually this picture... And I am mention that it comes from somewhere around the Pacific. What would be your immediate reaction to it?
2/ I know, I know everything is better with a poll.
3/ Truth be told I haven't look at the methodology in detail, so I cannot comment on how many horrors I will find. But it is funny that the conclusion after watching that is:
4/ But you know, we have been asking ourselves this question. How on earth, NOBODY realizes that there are simpler explanations.
5/ I dont like this type of analysis because it usually is an assumption on top of an assumption, on top of another assumption, with an assumption at the top. But if you want to look at it, just follow the link. nber.org/system/files/w…
• • •
Missing some Tweet in this thread? You can try to
force a refresh
@Akustronique 1/ There is a general misunderstanding on what a CI with p=whatever means. When you say your CI is [x..y] with a p=1-alpha you are essentially saying... I will accept p chances (say 4 in 100) of being wrong that the actual value X is within the bounds defined by x and y ...
@Akustronique 2/ What most people dont understand is that any value within the interval x and y is fair game with uniform distribution. There is NO guarantee that the prior distribution (the thing you are measuring) would follow a gaussian distribution or any other.
@Akustronique 3/ Which also mean that given a wide enough confidence interval [x..y] you just cannot accept a point estimate at face value. Where those intervals cross the null-hypothesis in the crude there is no 'adjustment' to be done that wouldn't bias your sample and throw off the analysis
1/ A new "Masks RCT" has been doing the rounds lately with "The miracle of masks". Don't say more. What do you know when your CI cannot rule out no-effect or harm? Yeah, that's correct. You drop it in the garbage can where it belongs.
2/ I wouldn't even going to comment on it, a cursory look at the tables already show that whatever math you apply to it was going to be anything... but given that I got bombarded with "What do you think about the masks RCT" it forced me to comment.
3/ Looking at the tables is a good time saver. I just didn't have to read 90 pages where a single table would suffice to tell me everything I need to know.
Made a mistake in the equation in an earlier tweet of the thread, but the general result is the same. In the wake of waning immunity, it becomes much more relevant.
1/ While I agree here, with most of the critique; I cannot stop thinking that everybody is looking for the tree but it is missing the forest. In the process to find explanations of why this and that, confounders and else. The elephant in the room is there staring at us.
@Grady_Booch The immune system is pretty complex, you can stimulate like crazy certain parts of it (in this case the humoral immunity), but in the same process hamper and/or enhance other parts like the innate immunity (the ones that kids rely more and lose potency when we age).
@Grady_Booch At the same time, in the humoral immunity there are many moving parts, the most important is the realization that you have parts that behave like RAM and others that behave like hard drives. The antibodies are transient, they are RAM. The B-cells are like hard drives.
@Grady_Booch The antibodies generated during the acute phase are transient, but in the process the body builds B-Cells whose only purpose is to initiate the process of generation of antibodies the next time a new bug matching the fingerprint of something known arrives.