topic Re: Nonparametric equivalent of proc glm? in SAS Procedures

Nonparametric equivalent of proc glm?

deleted_user — Mon, 10 Nov 2008 10:16:06 GMT

Hello,

I want to use a nonparametric test (such as the Kolmogorov Smirnov test available in proc npar1way) to make comparisons between a large number of distributions. I know that you can use proc glm to make two-way comparisons between a reference distribution and a large set of other distributions, but this uses anova for comparison. Is there an alternative to proc glm for doing this type of analysis which uses a nonparametric test for comparison?

Thank you.

Re: Nonparametric equivalent of proc glm?

lvm — Wed, 10 Dec 2008 13:38:05 GMT

There have been many developments in the nonparametric analysis of factorials, including repeated measures and split plots, as well as analyses for very large number of treatments. I suggest you read the book by Brunner, Domhof and Langer (Nonparametric analysis of longitudinal data in factorial experiments). A lot of SAS code is included. Also, you can read the introductory article in the biological journal Phytopathology (2004. Nonparametric analysis of ordinal data in designed factorial experiments. Phytopathology 94: 33-43). Good luck.

Re: Nonparametric equivalent of proc glm?

deleted_user — Fri, 12 Dec 2008 11:59:26 GMT

Thank you. It looks a bit complicated but I'll have a go!