Statistics is all about dealing with uncertainty, and yet, ironically, I shy away from teaching higher-level uncertainty. Like, how to effectively communicate that the answer we are certain about is "it's complicated and we don't really know".
-
-
Prikaži ovu nit
-
There are Bayesian ways of expressing model uncertainty, but these don't cover things like uncertainty about whether the model assumptions are met.
Prikaži ovu nit
Kraj razgovora
Novi razgovor -
-
-
Yes, about model assumptions, the difference between lack of power and lack of true signal, with an emphasis on visual inspection whenever possible. Also, about publication bias, and the need to be honest and upfront about any negative results and limitations.
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
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.