4 ms·
2-sigma events should vanish all the time. p < 5% just means that you see it (vanishment) happening less than 1-in-20 times. If you have hundreds of such events
by jsprogrammer 10y ago
2-sigma events should vanish all the time. p < 5% just means that you see it (vanishment) happening less than 1-in-20 times. If you have hundreds of such events, you'll have dozens of vanishings.
- yk 10y agoYes, exactly. My argument is the opposite, if you reason from evidence that vanishes from time to time, then you will have a very hard time to build a reasonable sound argument.
- stewbrew 10y agoIsn't this how science works? Even if "evidence" doesn't "vanish", your understanding of it could be invalidated. Truth is a temporary agreement.
- mamon 10y agoI'm no expert on this but isn't p < 5% riddiculously low threshold? I would rather set standard at p < 0.5% or even 0.05% This would probably solve "replicability crisis" in "soft" sciences like psychology.
- stdbrouw 10y agoThe problem is that then you will never find anything. There's always a trade-off between finding stuff that's not there and not finding stuff that really is out there.