Timeline for Is Wilcoxon Rank Sum test appropriate for comparing non-normal, imbalanced, heteroskedastic groups?
Current License: CC BY-SA 3.0
3 events
when toggle format | what | by | license | comment | |
---|---|---|---|---|---|
Feb 1, 2018 at 6:50 | comment | added | Bernhard | With data, that has large outliers as true values, you should make the decision of whether a mean or a median or both are relevant statistics depending on the subject, not on the easiest to perform test. Throwing a third hat in the ring, you could set up a permutation test for median, for mean, for 5% trimmed means or any other statistic that suites your substantial question best. I am not arguing in favour of a permutation test, just saying, you have the freedom to decide, which statistic should be tested based on the substance of your research. | |
Feb 1, 2018 at 5:28 | review | First posts | |||
Feb 1, 2018 at 8:33 | |||||
Feb 1, 2018 at 5:28 | history | asked | sk3w3d | CC BY-SA 3.0 |