Home > Statistics Every Writer Should Know > The Stats Board > Discusssion
GLM/Non parametric test
Message posted by Kim on February 6, 2000 at 12:00 AM (ET)
Can someone tell me if there is a non-parametric test to the general linear model?
READERS RESPOND:
(In chronological order. Most recent at the bottom.)
Re: GLM/Non parametric test
Message posted by Jack Tomsky on February 9, 2000 at 12:00 AM (ET)
A general approach is to rank all the data. Then replace the ranks with the corresponding percentiles from a standard normal distribution. At this stage, you can treat the data as if it met the usual normality assumptions and use the usual parametric methodology. Some people apply the analysis to just the ranks, without the normal-inverse transformation.
Your $5 contribution helps cover part the $500 annual cost of keeping this site online.