Effect size estimation for multiple post-hoc comparisons

[rsoto310790]

Dear all,

After performing one-way ANOVA on the diamonds data set, I use the pairwise.t.test function and I get this:

I don't trust the p_values here with big sample sizes. Is there any function to report the effect size for every possible comparison at once? I know I could do this for each comparison individually using cohen.d() but it'd be very useful to get the effect sizes for each comparison along side each adjusted p value.

Thanks!

[pete]

Hi,

See here for a good tutorial, in particular the TukeyHSD section. The UCLA site has a lot of good tutorials in a variety of topics.

Also, from a Bayesian perspective, you might look at the Multiple hypotheses section here