Three models are evaluated on the same datasets. Why consider Friedman rather than independent-group Kruskal–Wallis?
StatisticsLevel: IntermediateMean and paired inference
Three models are evaluated on the same datasets. Why consider Friedman rather than independent-group Kruskal–Wallis?
Hints
- Every dataset evaluates every model.
- Rank within the shared block.
Answer
Datasets form repeated-measure blocks. Friedman ranks treatments within blocks, accounting for dataset difficulty; independent-group testing discards that matched design.
Source and licence: SciPy — statistical estimators and hypothesis tests · BSD-3-Clause · answer: Quantoros editorial