Permutation tests for linear models in R

An R package which uses permutation tests to obtain p-values for linear models. Standard R linear model functions have been modified to produce p-values obtained from permutation tests instead of from normal theory. These values are advantageous when the degrees of freedom for error is small or non-existent, as is the case with saturated experimental designs, or when the data is drawn from a non-normal population, or when there are apparent outliers. The package also supports ANOVA for polynomial models such as those used for response surfaces.