Dennis Roberts wrote:
> in the case of CIs ... no, you are not sure at all that the range you got
> in your CI encompasses the parameter but, what are the odds that it does
> NOT? generally, fairly small.
You're slipping into Bayesian territory... I would say the answer
to your question is, "It depends", but more important, it doesn't
really matter. (It might beforehand when we are studying its
properties, but not afterwards when we've made the decision to use
it in practice.) If it did, we should be following Radford Neal's
suggestion that we all become Bayesians.
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
