Subscribe Bookmark RSS Feed

Post-hoc test with unequal variance ?

I would like to compare means. However, I have small number of replicates (<10) and the tests of variance equality (Levene's and Bartlett tests) reveal that i got unequal variances. I thus think that I can not use Tukey or Dunnett to compare means. In SPSS, there is a Games-Howell test... do you know if there is any think close to that on JMP ?