Improving the accuracy of identifying the lognormal curve in the Johnson system

Multiple estimation methods provide discriminants to determine curves in the Johnson (1949) system, but few studies have been conducted to explore their accuracy. In this study, a simulation study was implemented to determine whether three discriminants can discriminate the types of Johnson curves c...

Full description

Saved in:
Bibliographic Details
Published inCommunications in statistics. Simulation and computation Vol. 49; no. 5; pp. 1321 - 1332
Main Authors Pan, Tianshu, Yin, Yue
Format Journal Article
LanguageEnglish
Published Philadelphia Taylor & Francis 03.05.2020
Taylor & Francis Ltd
Subjects
Identification methods
Receiver operating characteristic curves
The Johnson system
The lognormal curve
Online AccessGet full text

Cover

More Information
Summary:Multiple estimation methods provide discriminants to determine curves in the Johnson (1949) system, but few studies have been conducted to explore their accuracy. In this study, a simulation study was implemented to determine whether three discriminants can discriminate the types of Johnson curves correctly. The results show these discriminants failed to identify the lognormal curve in the system. The receiver operating characteristic curves and the Youden index (Youden 1950) were used to find the optimal cutoff points for these discriminants to identify the lognormal curve more precisely. A new discriminant was also proposed for the moment method (Johnson 1949) in this study.
ISSN:0361-0918
1532-4141
DOI:10.1080/03610918.2018.1494834