[R] Multiple comparison tests - Homogenous subsets (i.e. for Tukey)

Prof. Dr. Karl-Heinz Kunzelmann karl-heinz at kunzelmann.de
Tue Oct 23 15:13:42 CEST 2007


Dear all,

is there any possibility to generate a table of homogeneous subsets for 
multiple comparison tests as in SPSS with R? I am specially interested 
in Tukey's B test.

I have tried to find the answer myself and found an excellent 
description here:

http://www.ppsw.rug.nl/~sda/siv/Posthoc_print.pdf (-> Tab. 2)

Together with the information here ...

http://www.agr.kuleuven.ac.be/vakken/statisticsbyR/ANOVAbyRr/multiplecomp.htm

(Differences between Brands are significant at 5% level if the 
confidence interval around the estimation of the difference does not 
contain zero.)

.. I am already able to figure out the subsets "manually". But for 
larger groups the risk to make errors increases.

I found a description of the SPSS algorithm (here: 
http://www1.uni-hamburg.de/RRZ/Software/SPSS/Algorith.115/app10_post_hoc_tests.pdf 

)
  but as a dentist I cannot really figure out how to do it with R myself.

Any help or hint is appreciated very much.

Thanks a lot already in advance.

Karl-Heinz

-- 
Prof. Dr. Karl-Heinz Kunzelmann
Dept. Operative Dentistry
Ludwig-Maximilians-University Munich
Goethestr. 70
D-80336 München
Germany

Tel.: +49 - 89 - 51609346
homepage: www.kunzelmann.de
email:  karl-heinz at kunzelmann.de



More information about the R-help mailing list