It is not a requirement of regression that the predictors be normally

distributed. There is no need to test for it. David Greenberg, Sociology

Department, New York University

----- Original Message -----

From: Marc <

[hidden email]>

Date: Sunday, July 30, 2006 6:39 pm

Subject: Testing for normality (Gauss)

> Hi there.

> I'm not sure what test I have to use to test for normal (gauss)

> distributionon a number of variables that I want to use as

> predictors in a linear

> regression model.

>

> I've read something about the K-S test. Is that the correct test to

> use ?

>

> Bye,

> Marc.

>