DTK-package {DTK}R Documentation

Dunnett-Tukey-Kramer Pairwise Multiple Comparison Test Adjusted for Unequal Variances and Unequal Sample Sizes


Functions for conducting and plotting Dunnett's (1980) modified Tukey-Kramer pairwise multiple comparison test accounting for unequal variance and unequal sample sizes.


Package: DTK
Type: Package
Version: 3.5
Date: 2013-07-01
License: GPL version 2 or newer
LazyLoad: yes

This package was created to analyze multi-level one-way experimental designs. It is designed to handle vectorized observation and factor data where there are unequal sample sizes and population variance homogeneity can not be assumed. To conduct the Dunnett modified Tukey-Kramer test (a.k.a. the T3 Procedure), create two vectors: one for your observations and one for the factor level of each observation. The function, gl.unequal, provides a means to more conveniently produce a factor vector with unequal sample sizes. Next, use the DTK.test function to conduct the test and save the output as an object to input into the DTK.plot function, which produces a confidence interval plot for each of the pairwise comparisons. Lastly, the function TK.test conducts the original Tukey-Kramer test.


I would like to acknowledge the invaluable help of Professor Brent Burch in the Department of Mathematics and Statistics at Northern Arizona University and suggestions by multiple R- users: including M. Nunez, B. Roustan and S. Marshall.


Matthew K. Lau, Department of Biological Sciences, Northern Arizona University, AZ

Maintainer: Matthew K. Lau <mkl48@nau.edu>


Dunnett, C.W. (1980) Pairwise Multiple Comparisons in the Unequal Variance Case. Journal of the American Statistical Association. 75 (372): 796-800.

See Also

gl.unequal, DTK.test, DTK.plot, TK.test

[Package DTK version 3.5 Index]