Nonparametric Weighted-Signs Tests for Location
S.A. Kassam, John B. Thomas
Abstract
S.A. Kassam, John B. Thomas
Abstract
Strictly nonparametric weighted-signs tests for the hypothesis of symmetry against location-shift alternatives are considered. The tests are extensions of the sign test and use a conditional structure to achieve a fixed size for all measures of the hypothesis. It is shown that the performance of the weighted-signs tests can be much better than that of the simple sign test, and that it compares favorably with that of the Wilcoxon signed-rank test.
OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Strictly nonparametric weighted-signs tests for the hypothesis of symmetry against location-shift alternatives are considered. The tests are extensions of the sign test and use a conditional structure to achieve a fixed size for all measures of the hypothesis. It is shown that the performance of the weighted-signs tests can be much better than that of the simple sign test, and that it compares favorably with that of the Wilcoxon signed-rank test.
Key concepts: Sign test, Nonparametric statistics, Wilcoxon signed-rank test, Sign (mathematics), Mathematics, Simple (philosophy), Test (biology), Statistical hypothesis testing