Comparison of the efficiency of nonparametric location tests based on grouped and individual data

Autor:	H Riedwyl, Jürg Hüsler
Rok vydání:	1989
Předmět:	Statistics and Probability Median test Wilcoxon signed-rank test Rank (linear algebra) Statistics Mann–Whitney U test Nonparametric statistics Sign test Statistics Probability and Uncertainty Grouped data Rank correlation Mathematics
Zdroj:	Statistics & Probability Letters. 7:287-291
ISSN:	0167-7152
DOI:	10.1016/0167-7152(89)90107-7
Popis:	The efficiency of rank tests for the location in the one and two sample case based on grouped data is compared with the Wilcoxon signed-rank test and with the Mann-Whitney-Wilcoxon test respectively. These efficiencies are calculated for some Box-Tiao densities and different numbers of groups. It shows that the rank tests with a grouping of the data into 8 to 14 groups is for large samples as efficient as the ungrouped rank tests, and in addition that they are more efficient than the ungrouped rank tests if the underlying density has large tails.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::7c468e846f2a9dd584b951da27a983e0 https://doi.org/10.1016/0167-7152(89)90107-7 Zobrazit plný text záznamu