by on 9 hours ago
In that state of affairs, the unequal variances version of the t-test may give extra dependable results. Rank transformations don't preserve variances, but variances are recomputed from samples after rank transformations. The Brown-Forsythe check has been suggested as an acceptable non-parametric equal to the F-check for equal variances. A extra powerful check is the Brunner-Munzel check, outperforming the Mann-Whitney U test in case of violated assumption of exchangeability. The Mann-Whitney U ...
7 views 0 likes