Has anyone written a function for Games-Howell post-hoc testing* in R? Google tells me that there was none as of 2005, but perhaps things have changed since then. Thanks, Drew ?*Or similar: I am looking for a post-hoc testing algorithm that will work with (slightly) unequal sample sizes and possibly unequal variance among samples. -- Andrew D. Steen, Ph.D. Center for Geomicrobiology, Aarhus University Ny Munkegade 114 8000 ?rhus C Denmark andrew.steen at biology.au.dk