Estimating the asymptotic relative efficiency of weighted rankings |
| |
| Authors: | Claudio Silva Dana Quade |
| |
| Institution: | 1. Facultad de Ciencia , Universidad de Santiago de Chile , Santiago, Chile;2. Department of Biostatistics , University of North Carolina , Chapel Hill, NC, 27514 |
| |
| Abstract: | Five tests for the hypothesis of no treatment effects in complete blocks designs are compared with respect to asymptotic relative efficiency: the ordinary two-way analysis of variance, the analysis of variance of ranks (Friedman, 1937), the method of ranking after alignment (Hodges and Lehmann, 1962), and a method of weighted rankings proposed by Quade (1972, 1979). This comparison required devising a procedure for estimating the asymptotic relative efficiencies of test statistics with asymp-totic chi-squared distributions. |
| |
| Keywords: | complete blocks designs Monte Carlo study two-way analysis of variance Friedman test ranking after alignment |
|
|
