Re: Statistical Indep and Corr of Normal RVs

Thank you for the reply,

The counter-example you gave sounds like the one I saw before.
Unfortunately, it does little (for me at least) to get some type of
insight as to what is actually going on. Not to discount the power of
the counter-example, but somehow, this canonical example that is always
trotted out seems a bit pathological to me. Is there an exact
mathematical (or statistical) condition one can state when two normal
rvs cannot be considered to have a bivariate distribution? A more
pedestrian question along these lines is to ask "If I don't
conditionally define the value of Y in terms of X using some type of
rule as in this counter-example, then does r=0 imply stat
independence?"

TA,

Matt

.