Re: st: analysis of continuous gestational age

[Richard Goldstein <richgold@ix.netcom.com>]

There is *no* assumption in linear regression that the *data*
be normally distributed -- the estimates are unaffected by
this.  However, if you want to trust the confidence intervals
and/or the p-values, then the residuals must be normally
distributed.  There is no necessary relationship between the
distribution of the residuals and the distribution of the
data -- how else could one use "dummy" variables?

Rich Goldstein

Svend Juul wrote:
> Alo wrote:
>
> What about the idea that we can use linear regression even if the
> residuals are
> not normally distributed if we have a large dataset? Is there any basis
> for
> this?
>
> -----------------------------------------------------------------------
>
> No, not in the school I went to. You might be thinking of the fact that
> with
> large datasets even unimportant deviations from normality become
> significant, so you should not use significance testing to decide
> whether the deviation is important, but rather graphical inspection.
>
> Regardless of dataset size: Gestational age data are not from a normal
> distribution; they deviate a lot from that assumption.
>
> Svend
> __________________________________________
>
> Svend Juul
> Institut for Folkesundhed, Afdeling for Epidemiologi
> (Institute of Public Health, Department of Epidemiology)
> Vennelyst Boulevard 6
> DK-8000  Aarhus C, Denmark
> Phone: +45 8942 6090
> Home:  +45 8693 7796
> Email: sj@soci.au.dk
> __________________________________________ 
*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/