Re: Residual Plot Question


DickW wrote:
> >from Reef Fish Bob: "That's why in validating the behavior of the errors, on the assumptions
> >of stochastic independence or homoscedasticity, the only valid frame
> >of reference to plot the residuals against is the fitted y-hat; and
> >not Y or any of the X's."
>
> ========================================================================
>
> I fully agree with the above general conclusion when there are multiple
> X variables. However, I note that when there is only one X variable
> (bivariate linear regression), that fitted y-hat is a simple linear
> transform of X-observed. X-observed can therefore equally successfully
> be used, instead of y-hat, to show well-behaved residuals.
>
> DickW

What you pointed out is obviously true.

I already covered that special case of the simple regression in my
initial follow-up:

RF> There is no specific assumption of the
RF> independence of the errors on the observed Xs in a multiple
regression.
RF> In a simple regression, the order of the X is always the same as
the
RF> order of the fitted Y.

-- Reef Fish Bob.

.