-
Hi Team, mse = (np.sum([np.var(arg, ddof=1) for arg in args] *
(nsamples_treatments - 1)) / (nobs - ntreatments)) Whereas the model comparison file uses the output of a repeated measures ANOVA test from the |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment
-
Hi @mps-hlx, This is intentional. We discuss this in the preprint, see end of section 3.2.1.
The TukeyHSD assumes independence of samples, which for repeated CV will result in CIs that are too large. Repeated measures ANOVA appropriately accounts for the experimental design, and results in more accurate CIs that are tighter, enabling researchers to find more statistically significant differences. Glad to have the opportunity to stress this more -- beyond that brief paragraph. It is important to use the right experimental design here, and with real practical implications! |
Beta Was this translation helpful? Give feedback.
Hi @mps-hlx,
This is intentional. We discuss this in the preprint, see end of section 3.2.1.
The TukeyHSD assumes independence of samples, which for repeated CV will result in CIs that are too large. Repeated measures ANOVA appropriately accounts for the …