[R] Order-restricted inference

Bruno L. Giordano bruno.giordano at music.mcgill.ca
Fri Jul 21 02:30:34 CEST 2006

I looked for R packages which focused on order-restricted statistical 
inference, but I could find only the isoreg() function.

I would need to test whether the means in my (repeated measures) data follow 
a given order, e.g. A<B=C<D.

I took a look at the monograph by Barlow et al. (1972) on this topic and 
found that for my case the null hypothesis is always A=B=C=D. This might be 
a naive question, but wouldn't it be more appropriate to use "at least one 
pair of means does not follow the alternative hypothesis order", or to use 
this latter hypothesis as the alternative and A<B=C<D as the null?

If R packages are not found for my purpose, can somebody please point me to 
some recent monographs on order-restricted inference?

Thank you,

More information about the R-help mailing list