Re: Inferred Correlation via Known Correlations?

Reef Fish wrote:
>
> Aren't you confusing the issue of "uncorrelated" and "independent"
> random variables? Random variables can be DEPENDENT and
uncorrelated.
>
No. The question and the steps in my argument all concern
correlation/correlated/uncorrelated considerations.
Dependence/independence doesn't enter the discussion. But indeed
random variables can be DEPENDENT and uncorrelated. The representation
such as X= 0.8 Y + ErrX, with Y and ErrX uncorrelated can be found
in Cox and Hinkleys well-known book: in general it is not true that Y
and ErrX are independent, but it is the "uncorrelated" result that is
needed.

David Jones


.

Relevant Pages

  • Independence between sums of variables
    ... I have a question in investigating the independence between sums of ... random variables. ... I guess that is always dependent of, ...
    (sci.stat.math)
  • Sum of two dependent random variables
    ... Is there any known literature on the PDF of the sum of two dependent ...
    (sci.stat.math)
  • Sum of random variables
    ... Let X and Y are independent random variables. ... How one can prove that there exists a random variable Z (which can be dependent with X) such that X + Z ~ Y by distribution? ... Prev by Date: ...
    (sci.stat.math)