5 ms·
> It's so frustrating. One micronewton per kilowatt. All that power going in, and so little force coming out. All it takes is some tiny noise effect to get tha
by gwern 10y ago
> It's so frustrating. One micronewton per kilowatt. All that power going in, and so little force coming out. All it takes is some tiny noise effect to get that much force. The effect is so weak and they're so near the noise threshold.
Yep. Reminds me of psi. You can set up mechanical radiation-decay powered RNGs with pre-set runs of numbers in pre-registered double-blind experiments with remote audit logs analyzed carefully with permutation tests and randomness test suites, and all of the huge effects claimed by regular people and psychics go away... but you still get on net a slight excess of bit flips in the psi direction. Is psi real? Seems highly unlikely. But what could possibly be left - publication bias, fraud, or what? No one knows. At some point, effects become subtle enough that you just can't believe studies finding it on net - there's some sort of epistemological 'noise threshold' where stuff is happening but you don't know why or can come up with causal explanations.
- Nokinside 10y agoThe real mystery is why it always happens to people like White. Smart engineer who gets his PhD in physics at age 43. How can experimental physicists and research groups who have the decade of more experience avoid these traps?
- andygates 10y agoI think they might be coming from a broader perspective: "If this is true then conservation of momentum and Noether's theorem are false, and we'd see that out in the wider universe: therefore this is a lab error and people who chase it obsessively are cranks."
- genmon 10y agoI can't tell whether you're referencing a movie or talking about some interesting fringe science. If the latter, do you have any recommendations of papers I should check out?
- austinjp 10y ago> you still get on net a slight excess of bit flips in the psi direction. Is that "net" extracted from all published experiments, i.e. in some meta syntheses? And is statistically significance calculated in each individual experiment? > Is psi real? Seems highly unlikely. But what could possibly be left - publication bias, fraud, or what? No one knows. Publication bias toward positive results should be detectable via methods such as funnel plots. If you've looked into these aspects it would make interesting reading.
- retreatguru 10y agoDean Radin has published a lot on this topic. He has an interesting list of selected publications 1. A number of meta analysis are listed there. His book "The Conscious Universe" discusses many of the meta analysis in detail. Facinating stuff. He addresses the file drawer effect in depth. [1] http://deanradin.com/evidence/evidence.htm http://deanradin.com/evidence/evidence.htm [2] http://www.goodreads.com/book/show/319012.The_Conscious_Universe http://www.goodreads.com/book/show/319012.The_Conscious_Univ...
- gwern 10y ago> Is that "net" extracted from all published experiments, i.e. in some meta syntheses? And is statistically significance calculated in each individual experiment? You can find excesses in both individual experiments and meta-analyses, IIRC. > Publication bias toward positive results should be detectable via methods such as funnel plots. Funnel plots have notoriously low statistical power, it would require only a little publication bias to produce excesses, and asymmetry in a funnel plot can be due to other reasons.
- austinjp 10y agoThanks for these comments, will make for interesting investigations.
- Confusion 10y agoNo one knows And no one needs to care. It shows that if psi exists, its practical value is zero, because it is weak and uncontrolled.