On Sat, 1 Feb 2003, David C. Howell wrote in part:

> This morning I sent out the message below. I still have basically
> the same question, but when I went back and looked at my resampling
> program, I discovered an error. When I corrected that, the
> distribution is normal. But I still don't know what is wrong with my
> argument that it shouldn't be normal.

Might it have to do with assuming values of X to be fixed, not random,
in the regression model?  The distribution of r is commonly derived, I
thought, on the assumption that both X and Y be normally distributed.
But if X be considered to have fixed values, and therefore not to be a
random variable, the usual distribution of r presumably would not apply.

I suspect your question arises because beta and r have the same *value*;
but to say that the *value* of beta is equal to the *value* of r is not
necessarily to say that the *distribution* of beta is the same as the
(usual) *distribution* of r, is it?

> *****************************************
> This morning's message [condensed -- DFB]:
>
> Yesterday I received a query asking about the sampling distribution of
> beta, the standardized regression coefficient.  < snip >
>
> Let's ... look at the sampling distribution of b  <snip>
>
> [References cited, and arguments presented, stating and/or proving the
> assertion that b is normally distributed.]
>
> BUT, suppose that we standardized our variables. [Then] the slope is
> equal to the standardized slope (beta) [which] is equal to the
> correlation coefficient. So [for] standardized variables ... r, b,
> and beta will all be numerically equal.
>
> [Comments on sampling distribution for r.]
>
> So now I have shown that the sampling distribution is skewed, though I
> began by quoting experts I respect saying that it is normal.  < snip >
>
> So where did I go wrong? The sampling distribution of b cannot be both
> normal and skewed, at the same time.

 -----------------------------------------------------------------------
 Donald F. Burrill                                            [EMAIL PROTECTED]
 56 Sebbins Pond Drive, Bedford, NH 03110                 (603) 626-0816

.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
.                  http://jse.stat.ncsu.edu/                    .
=================================================================

Reply via email to