Testing equality of variances of correlated variates with incomplete data on both responses

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

Testing the hypothesis of equality of the variances is considered when sampling from a bivariate normal distribution with means μ1 and μ2, variances and, and correlation coefficient ρ, where some observations on either of the variables are missing. The null distributions of the proposed statistics are obtained in terms of well-known distributions. The empirical powers of the statistics are computed for different values of and ρ, and are compared with those obtained from a statistic based on complete observations.

Original languageEnglish (US)
Pages (from-to)681-683
Number of pages3
JournalBiometrika
Volume66
Issue number3
DOIs
StatePublished - Dec 1 1979

Fingerprint

Incomplete Data
Equality
statistics
Statistics
Bivariate Normal Distribution
Testing
Null Distribution
Correlation coefficient
Statistic
Normal Distribution
testing
Normal distribution
Sampling
Observation
Incomplete data
sampling

All Science Journal Classification (ASJC) codes

  • Statistics and Probability
  • Mathematics(all)
  • Agricultural and Biological Sciences (miscellaneous)
  • Agricultural and Biological Sciences(all)
  • Statistics, Probability and Uncertainty
  • Applied Mathematics

Cite this

@article{d0a41d2d90f04140876018a96b193ed9,
title = "Testing equality of variances of correlated variates with incomplete data on both responses",
abstract = "Testing the hypothesis of equality of the variances is considered when sampling from a bivariate normal distribution with means μ1 and μ2, variances and, and correlation coefficient ρ, where some observations on either of the variables are missing. The null distributions of the proposed statistics are obtained in terms of well-known distributions. The empirical powers of the statistics are computed for different values of and ρ, and are compared with those obtained from a statistic based on complete observations.",
author = "Dinesh Bhoj",
year = "1979",
month = "12",
day = "1",
doi = "10.1093/biomet/66.3.681",
language = "English (US)",
volume = "66",
pages = "681--683",
journal = "Biometrika",
issn = "0006-3444",
publisher = "Oxford University Press",
number = "3",

}

Testing equality of variances of correlated variates with incomplete data on both responses. / Bhoj, Dinesh.

In: Biometrika, Vol. 66, No. 3, 01.12.1979, p. 681-683.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Testing equality of variances of correlated variates with incomplete data on both responses

AU - Bhoj, Dinesh

PY - 1979/12/1

Y1 - 1979/12/1

N2 - Testing the hypothesis of equality of the variances is considered when sampling from a bivariate normal distribution with means μ1 and μ2, variances and, and correlation coefficient ρ, where some observations on either of the variables are missing. The null distributions of the proposed statistics are obtained in terms of well-known distributions. The empirical powers of the statistics are computed for different values of and ρ, and are compared with those obtained from a statistic based on complete observations.

AB - Testing the hypothesis of equality of the variances is considered when sampling from a bivariate normal distribution with means μ1 and μ2, variances and, and correlation coefficient ρ, where some observations on either of the variables are missing. The null distributions of the proposed statistics are obtained in terms of well-known distributions. The empirical powers of the statistics are computed for different values of and ρ, and are compared with those obtained from a statistic based on complete observations.

UR - http://www.scopus.com/inward/record.url?scp=0018702938&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0018702938&partnerID=8YFLogxK

U2 - 10.1093/biomet/66.3.681

DO - 10.1093/biomet/66.3.681

M3 - Article

AN - SCOPUS:0018702938

VL - 66

SP - 681

EP - 683

JO - Biometrika

JF - Biometrika

SN - 0006-3444

IS - 3

ER -