Goodness of fit tests with misclassified data

K. F. Cheng, H. M. Hsueh, T. H. Chien

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

The most popular goodness of fit test for a multinomial distribution is the chi-square test. But this test is generally biased if observations are subject to misclassification. In this paper we shall discuss how to define a new test procedure when we have double sample data obtained from the true and fallible devices. An adjusted chi-square test based on the imputation method and the likelihood ratio test are considered. Asymptotically, these two procedures are equivalent. However, an example and simulation results show that the former procedure is not only computationally simpler but also more powerful under finite sample situations.

Original languageEnglish
Pages (from-to)1379-1393
Number of pages15
JournalCommunications in Statistics - Theory and Methods
Volume27
Issue number6
Publication statusPublished - 1998
Externally publishedYes

Fingerprint

Goodness of Fit Test
Chi-squared test
Multinomial Distribution
Misclassification
Imputation
Likelihood Ratio Test
Biased
Simulation

Keywords

  • Chi-square test
  • Double sampling
  • Goodness of fit
  • Imputation
  • Likelihood ratio test
  • Misclassification

ASJC Scopus subject areas

  • Safety, Risk, Reliability and Quality
  • Statistics and Probability

Cite this

Cheng, K. F., Hsueh, H. M., & Chien, T. H. (1998). Goodness of fit tests with misclassified data. Communications in Statistics - Theory and Methods, 27(6), 1379-1393.

Goodness of fit tests with misclassified data. / Cheng, K. F.; Hsueh, H. M.; Chien, T. H.

In: Communications in Statistics - Theory and Methods, Vol. 27, No. 6, 1998, p. 1379-1393.

Research output: Contribution to journalArticle

Cheng, KF, Hsueh, HM & Chien, TH 1998, 'Goodness of fit tests with misclassified data', Communications in Statistics - Theory and Methods, vol. 27, no. 6, pp. 1379-1393.
Cheng, K. F. ; Hsueh, H. M. ; Chien, T. H. / Goodness of fit tests with misclassified data. In: Communications in Statistics - Theory and Methods. 1998 ; Vol. 27, No. 6. pp. 1379-1393.
@article{6d9f6a6612f34a75ae558257348f8c3e,
title = "Goodness of fit tests with misclassified data",
abstract = "The most popular goodness of fit test for a multinomial distribution is the chi-square test. But this test is generally biased if observations are subject to misclassification. In this paper we shall discuss how to define a new test procedure when we have double sample data obtained from the true and fallible devices. An adjusted chi-square test based on the imputation method and the likelihood ratio test are considered. Asymptotically, these two procedures are equivalent. However, an example and simulation results show that the former procedure is not only computationally simpler but also more powerful under finite sample situations.",
keywords = "Chi-square test, Double sampling, Goodness of fit, Imputation, Likelihood ratio test, Misclassification",
author = "Cheng, {K. F.} and Hsueh, {H. M.} and Chien, {T. H.}",
year = "1998",
language = "English",
volume = "27",
pages = "1379--1393",
journal = "Communications in Statistics - Theory and Methods",
issn = "0361-0926",
publisher = "Taylor and Francis Ltd.",
number = "6",

}

TY - JOUR

T1 - Goodness of fit tests with misclassified data

AU - Cheng, K. F.

AU - Hsueh, H. M.

AU - Chien, T. H.

PY - 1998

Y1 - 1998

N2 - The most popular goodness of fit test for a multinomial distribution is the chi-square test. But this test is generally biased if observations are subject to misclassification. In this paper we shall discuss how to define a new test procedure when we have double sample data obtained from the true and fallible devices. An adjusted chi-square test based on the imputation method and the likelihood ratio test are considered. Asymptotically, these two procedures are equivalent. However, an example and simulation results show that the former procedure is not only computationally simpler but also more powerful under finite sample situations.

AB - The most popular goodness of fit test for a multinomial distribution is the chi-square test. But this test is generally biased if observations are subject to misclassification. In this paper we shall discuss how to define a new test procedure when we have double sample data obtained from the true and fallible devices. An adjusted chi-square test based on the imputation method and the likelihood ratio test are considered. Asymptotically, these two procedures are equivalent. However, an example and simulation results show that the former procedure is not only computationally simpler but also more powerful under finite sample situations.

KW - Chi-square test

KW - Double sampling

KW - Goodness of fit

KW - Imputation

KW - Likelihood ratio test

KW - Misclassification

UR - http://www.scopus.com/inward/record.url?scp=0040429989&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0040429989&partnerID=8YFLogxK

M3 - Article

VL - 27

SP - 1379

EP - 1393

JO - Communications in Statistics - Theory and Methods

JF - Communications in Statistics - Theory and Methods

SN - 0361-0926

IS - 6

ER -