Jämför metoder
Granska de valda metoderna sida vid sida; rader som skiljer sig är markerade.
| Wilcoxon teckenrangtest× | Kruskal-Wallis H-test× | |
|---|---|---|
| Ämnesområde | Statistik | Statistik |
| Familj | Hypothesis test | Hypothesis test |
| Ursprungsår≠ | 1945 | 1952 |
| Upphovsperson≠ | Frank Wilcoxon | William Kruskal & W. Allen Wallis |
| Typ≠ | Nonparametric paired comparison | Nonparametric group comparison |
| Ursprungskälla≠ | Wilcoxon, F. (1945). Individual comparisons by ranking methods. Biometrics Bulletin, 1(6), 80–83. DOI ↗ | 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 ↗ |
| Alias≠ | Wilcoxon matched-pairs signed-rank test, signed-rank test, Wilcoxon İşaretli Sıra Testi | Kruskal-Wallis H test, one-way ANOVA on ranks, Kruskal-Wallis one-way analysis of variance, Kruskal-Wallis Testi |
| Närliggande≠ | 4 | 5 |
| Sammanfattning≠ | The Wilcoxon signed-rank test is the nonparametric alternative to the paired t-test, comparing two related measurements on the same subjects to decide whether their typical difference is zero. It was introduced by Frank Wilcoxon in 1945 and works on continuous or ordinal data without assuming normality. | 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. |
| ScholarGateDatamängd ↗ |
|
|