Cookies help us deliver our services. By using our services, you agree to our use of cookies. More information

Hekimoğlu 2005 zfv

From Bioblast
Publications in the MiPMap
Hekimoğlu S (2005) Do robust methods identify outliers more reliably than conventional tests for outliers?. zfv 130(3):174-180.

» Open Access

Hekimoğlu S (2005) zfv

Abstract: In order to identify outliers, there are two approaches: the conventional tests for outliers and robust methods. Statisticians working with robust methods argue that their results are more reliable than the conventional tests for outliers. Which one of these approaches is more reliable? This question is investigated here in view of the problems caused by masking effects, swamping effects and leverage points and discussed by simulated linear regression models. The mean success rate is used to compare the two approaches. Summarizing, the robust methods can identify outliers at a rate of 22% more reliably than the conventional test for outliers in a simple regression.

Bioblast editor: Iglesias-Gonzalez J

Cited by

  • Iglesias-Gonzalez et al (2021) Proficiency test in mt-respiration: A necessary tool for reliable and reproducible results. MitoFit Preprints 2021 (in prep).

Labels:






MitoFit 2021 PT