Nested ANOVA permutation test in R

A R function for the permutation test for onw-way nested design ANOVA.

This work is inspired by Dr. David C. Howell, University of Vermont. In Dr. Howell’s webpage “Permutation Tests for Factorial ANOVA Designs” (http://www.uvm.edu/~dhowell/StatPages/More_Stuff/Permutation%20Anova/PermTestsAnova.html [access on May 28, 2012]), he described the algorithm of permutation test for a one-way nested design ANOVA by using R language. I followed this algorithm and made a R function to do this test.

Example

dat <- read.csv(textConnection("
 trt , unit , obs
 1 , 1 , 11
 1 , 1 ,  9
 1 , 1 ,  9
 1 , 2 ,  8
 1 , 2 ,  7
 1 , 2 ,  6
 1 , 3 ,  8
 1 , 3 , 10
 1 , 3 , 11
 2 , 4 , 11
 2 , 4 ,  8
 2 , 4 ,  7
 2 , 5 , 10
 2 , 5 , 14
 2 , 5 , 12
 2 , 6 ,  9
 2 , 6 , 10
 2 , 6 ,  8
"))

## traditional nested ANOVA
mod.2 <- aov(
 obs ~ factor(trt) + Error(factor(unit)),
 data = dat
)
summary(mod.2) # alternative

## permutation nested ANOVA
nestedPermutationAnova(dat$obs, dat$trt, dat$unit, 499)

Logistic regression with package rms in R

The rms package by Frank E Harrell Jr provides a useful function lrm to apply logistic regression. Moreover, this function also provides many useful statistics, and a gooeness-of-fit test (le Cessie and Houwelingen test) could be applied by using the function residuals(a lrm model, "gof"). Read the official document of package rms for more details.