{"title":"On the consistency of K-sign depth tests","authors":"Kevin Leckey, Mirko Jakubzik, Christine H. Müller","doi":"10.1016/j.ecosta.2023.10.002","DOIUrl":null,"url":null,"abstract":"The consistency of the so-called K-sign depth tests is considered. These tests are based on the K-sign depth, which originated from the simplicial regression depth, but is easier to compute. The K-sign depth tests use only the signs of residuals and are equivalent to the classical sign test for K=2. However, K-sign depth tests with K≥3 show a much better power than the classical sign tests in simulation studies. This property is attributed to the consistency of these tests for K=3. After deriving a general condition for consistency, it is shown that this condition is in particular satisfied for several relevant hypotheses in polynomial regression models.","PeriodicalId":54125,"journal":{"name":"Econometrics and Statistics","volume":"275 1","pages":"0"},"PeriodicalIF":2.0000,"publicationDate":"2023-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Econometrics and Statistics","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1016/j.ecosta.2023.10.002","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"ECONOMICS","Score":null,"Total":0}
引用次数: 0
Abstract
The consistency of the so-called K-sign depth tests is considered. These tests are based on the K-sign depth, which originated from the simplicial regression depth, but is easier to compute. The K-sign depth tests use only the signs of residuals and are equivalent to the classical sign test for K=2. However, K-sign depth tests with K≥3 show a much better power than the classical sign tests in simulation studies. This property is attributed to the consistency of these tests for K=3. After deriving a general condition for consistency, it is shown that this condition is in particular satisfied for several relevant hypotheses in polynomial regression models.
期刊介绍:
Econometrics and Statistics is the official journal of the networks Computational and Financial Econometrics and Computational and Methodological Statistics. It publishes research papers in all aspects of econometrics and statistics and comprises of the two sections Part A: Econometrics and Part B: Statistics.