permuANOVA {PermutationR}

permuANOVA

Description

Conducts permutational one-way and two-way ANOVAs. This code has been modified from Howell D. (2015)

Usage

permuANOVA(y,x,z, perm.type="unrestricted", reps=5000)

Arguments

y

Dependent variable, numeric vector

x

Independent variable, integer vector or factor up to 10 levels

z

(Optional) Independent variable, integer vector or factor up to 10 levels

perm.type

(Optional) Method of permutation for two-way designs. Either "unrestricted" or "restricted". Default is "unrestricted"

reps

Number of permutations, default is 5000.

Details

"unrestricted" computes main effect permutations in an unrestricted fashion; "restricted" restricts main effect permutations within levels of the other independent variable. In both cases, the interaction is computed in an unrestricted fashion

Value

For two-way designs: returns a data-frame containing p-values from permutation test for "Variable_x", "Variable_z" and interaction "x:z" For one-way designs: returns a data-frame containing p-values from permutation test for "Variable_x" only.

Author(s)

Osama Abdalla [e-mail]

References

Howell D. (2015). Permutation Tests for Factorial ANOVA Designs. Retrieved from: <https://www.uvm.edu/~statdhtx/StatPages/Permutation%20Anova/PermTestsAnova.html>

See Also

Package permuco function aovperm() and Package RVAideMemoire function perm.anova()

Examples

Data <- ToothGrowth
attach(ToothGrowth)
permuANOVA(y = len, x = supp, reps=99)

permuANOVA(y=len,x=supp,z=dose, perm.type="unrestricted", reps=99)

permuANOVA(y=len,x=supp,z=dose, perm.type="restricted", reps=99)