Uporedite metode
Pregledajte izabrane metode jednu pored druge; redovi koji se razlikuju su istaknuti.
| H-test po Kruskal-Valisu× | Permutation (Randomization) Test× | |
|---|---|---|
| Oblast | Statistika | Statistika |
| Porodica≠ | Hypothesis test | Regression model |
| Godina nastanka≠ | 1952 | 2005 |
| Tvorac≠ | William Kruskal & W. Allen Wallis | Good (2005); Edgington & Onghena (2007); resampling tradition |
| Tip≠ | Nonparametric group comparison | Nonparametric resampling test |
| Temeljni izvor≠ | Kruskal, W. H. & Wallis, W. A. (1952). Use of ranks in one-criterion variance analysis. Journal of the American Statistical Association, 47(260), 583–621. DOI ↗ | Good, P. (2005). Permutation, Parametric and Bootstrap Tests of Hypotheses (3rd ed.). Springer. ISBN: 978-0387202792 |
| Drugi nazivi | Kruskal-Wallis H test, one-way ANOVA on ranks, Kruskal-Wallis one-way analysis of variance, Kruskal-Wallis Testi | randomization test, exact permutation test, re-randomization test, Permütasyon Testi |
| Srodne | 5 | 5 |
| Sažetak≠ | The Kruskal-Wallis H test is a nonparametric hypothesis test that compares three or more independent groups to decide whether their distributions (typically their medians) differ. Introduced by William Kruskal and W. Allen Wallis in 1952, it works on ranks rather than raw values and is the distribution-free counterpart to one-way ANOVA. | The permutation test is a nonparametric resampling procedure that builds the sampling distribution of a test statistic directly from the data by repeatedly shuffling the group labels. Developed in the resampling tradition and treated systematically by Good (2005) and Edgington & Onghena (2007), it requires no parametric distributional assumption and yields an exact p-value. |
| ScholarGateSkup podataka ↗ |
|
|