Timeline for Distribution of a centered standardized sample
Current License: CC BY-SA 3.0
6 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Aug 24, 2013 at 15:27 | answer | added | JAW | timeline score: 1 | |
| Dec 1, 2012 at 4:01 | history | tweeted | twitter.com/#!/StackStats/status/274724982691622914 | ||
| Sep 27, 2012 at 18:36 | answer | added | Emmanuel Charpentier | timeline score: 0 | |
| Aug 19, 2012 at 17:22 | comment | added | Emmanuel Charpentier | Because if I know the distribution of ${(X-\mu_X)} \over \sigma$, (normal(0, 1), of course), I want to be able to use the distribution of ${(X-m)} \over s$, which is the propre prior distribution to use when these $X$ are themselves parameters in a Bayesian model. See an example of this use in Gelman's regression textbook (2007) about IRT models, when he uses this trick to get a better convergence of the IRT coefficients ; I'd like to have an explicit derivation instead of blind faith in BUGS sampling. PS : you're right $m \over s$ has a $t(\mu_x, 1, N-1)$ distribution. | |
| Aug 10, 2012 at 22:01 | comment | added | Michael R. Chernick | It is not m/s that has the student t distribution it is (m-μ$_x$)/s. Why are you interested in the joint distribution of the components of the Y vector? | |
| Aug 1, 2012 at 20:18 | history | asked | Emmanuel Charpentier | CC BY-SA 3.0 |