Good stuff. If you're looking for ideas on blog posts, would you consider talking about confidence intervals in extrapolations? E.g. CIs of population size estimate from year 1 + what you would predict based on known birth/death rates for year 2.
-
-
-
It's my understanding that Andrew only reads comments on the blog page, not on twitter. These are tweeted automatically.
- 3 more replies
New conversation -
-
-
I went through an entire PhD program in statistics where I was taught exactly that! Statistics is a really messed up field right now and I am yet to see signs that collectively we actually have a decent clue as to what the hell we are actually doing.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
It depends on what you mean by "sure." Under the standard interpretation, the statement is correct. If you don't like that interpretation, then it's not correct. :-) Personally, I think the interpretation is fine.
-
Take a perfect die. The probability for a 6 is .167 exactly. If you roll your die three times and get three 6, you get a confint of [.29, 1]. You know, the true value is not in the interval. Does that count as a counter example or did I completely miss your point?
- 4 more replies
New conversation -
-
-
It is correct to say for Bayesian credible intervals though!
-
The same caveat in interpretation Andrew describes here applies equally well to credible intervals—it depends on your prior.
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.