Which reminds me that Paul & I have some follow-up modeling partly written. We should finish it.
-
-
Prikaži ovu nit
-
Okay
@psmaldino I am opening that folder now and staring at itPrikaži ovu nit
Kraj razgovora
Novi razgovor -
-
-
Its a great paper. If i may suggest for follow ups though to include more discussion of RR
-
The one Richard and I are working on goes in a different direction, but I did one follow-up with a couple of graduate students that addresses RRs more directly. https://royalsocietypublishing.org/doi/10.1098/rsos.191249 …
- Još 1 odgovor
Novi razgovor -
-
-
Keep in mind that measuring power in models with more than three variables or complex designs is essentially roulette. The power of what? We don't know.
@quantitudepod (episode 2) - Još 1 odgovor
Novi razgovor -
-
-
Fine paper, though depressing, but take a further step back: If the core problem is oversupply of Ph.D.s, model the evolution not of labs, but of Ph.D. programs. Deans & chairs lack disincentives for producing "zombie Ph.D.s", with limited chances of work in their field.
-
Two usable policy proposals: - Univs track 6-yr graduation rates, grad schools (or DoE!) should track 5-yr in-field and TT employment. Harder to hack than p-values, I hope. - Reward programs and students where Masters theses are replications.
Kraj razgovora
Novi razgovor -
-
-
Q: the paper says "confirmatory & disconfirmatory replications are published, possibly at lower rates" and "pos. and neg. replications communicated with probabilities cR+/cr-". But cR+ and cR+ are both 1. Does this mean all replications are published?pic.twitter.com/jNTzNs1MjK
- Još 2 druga odgovora
Novi razgovor -
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.