o2348diuu

4 ms·

1.
by o2348diuu 8y ago
Pearl's attention to the do conditionality -- i.e., P(Y|do(X)) versus P(Y|X) is interesting and important in a certain sense, but I'm not sure it's really resolved debates about causality in any practical sense. I don't
2.
by o2348diuu 8y ago
Like intellectual property and some other issues, I have often felt that the appearance in public discourse of arguments for two very opposite, extreme positions is a signal that something is really wrong with the current situation. That is
3.
by o2348diuu 8y ago
Not really disagreeing with you necessarily, but you can hack pre-registration too, in a way that becomes more tenable with smaller studies, which benefit less from pre-registration anyway. The tables are misleading too: of course p(h0=T|p&
4.
by o2348diuu 9y ago
Nim. I understand some reasons for this, as I didn't pay attention to it very much for awhile myself. I think it's only the last year or two that it's reached maturity. Its primary alternatives also have the advantage of huge
5.
by o2348diuu 9y ago
I completely understand the issue of not wanting spam of various sorts broadly defined, but I do think restricting it to academic addresses is a huge mistake. As an established researcher on the verge of leaving an academic institution, but
6.
by o2348diuu 9y ago
This piece hits close to home, as it's really clear to me I am going through a midlife crisis and have interesting parallels and contrasts with the author in terms of my career and next steps. What seems to me to be the heart of this p