Ignorability and coarse data: Some biomedical examples

Research output: Contribution to journalReview article

50 Citations (Scopus)

Abstract

Heitjan and Rubin (1991, Annals of Statistics 19, 2244-2253) define data to be 'coarse' when one observes not the exact value of the data but only some set (a subset of the sample space) that contains the exact value. This definition covers a number of incomplete-data problems arising in biomedicine, including rounded, heaped, censored, and missing data. In analyzing coarse data, it is common to proceed as though the degree of coarseness is fixed in advance-in a word, to ignore the randomness in the coarsening mechanism. When coarsening is actually stochastic, however, inferences that ignore this randomness may be seriously misleading. Heitjan and Rubin (1991) have proposed a general model of data coarsening and established conditions under which it is appropriate to ignore the stochastic nature of the coarsening. The conditions are that the data be coarsened at random [a generalization of missing at random (Rubin, 1976, Biometrika 63, 581-592)] and that the parameters of the data and the coarsening process be distinct. This article presents detailed applications of the general model and the ignorability conditions to a variety of coarse-data problems arising in biomedical statistics. A reanalysis of the Stanford Heart Transplant Data (Crowley and Hu, 1977, Journal of the American Statistical Association 72, 27-36) reveals significant evidence that censoring of pretransplant survival times by transplantation was nonignorable, suggesting a greater benefit from cardiac transplantation than previous analyses had found.

Original languageEnglish (US)
Pages (from-to)1099-1109
Number of pages11
JournalBiometrics
Volume49
Issue number4
StatePublished - Dec 1 1993

Fingerprint

Ignorability
heart transplant
Coarsening
statistics
Heart Transplantation
Transplantation
Transplants
Statistics
Randomness
sampling
Missing at Random
Sample space
Survival Time
Incomplete Data
Censored Data
Censoring
Missing Data
Cardiac
Cover
Distinct

ASJC Scopus subject areas

  • Agricultural and Biological Sciences(all)
  • Agricultural and Biological Sciences (miscellaneous)
  • Applied Mathematics
  • Statistics and Probability
  • Public Health, Environmental and Occupational Health

Cite this

Ignorability and coarse data : Some biomedical examples. / Heitjan, D. F.

In: Biometrics, Vol. 49, No. 4, 01.12.1993, p. 1099-1109.

Research output: Contribution to journalReview article

@article{aab75511fff7436e84a547f014f797e9,
title = "Ignorability and coarse data: Some biomedical examples",
abstract = "Heitjan and Rubin (1991, Annals of Statistics 19, 2244-2253) define data to be 'coarse' when one observes not the exact value of the data but only some set (a subset of the sample space) that contains the exact value. This definition covers a number of incomplete-data problems arising in biomedicine, including rounded, heaped, censored, and missing data. In analyzing coarse data, it is common to proceed as though the degree of coarseness is fixed in advance-in a word, to ignore the randomness in the coarsening mechanism. When coarsening is actually stochastic, however, inferences that ignore this randomness may be seriously misleading. Heitjan and Rubin (1991) have proposed a general model of data coarsening and established conditions under which it is appropriate to ignore the stochastic nature of the coarsening. The conditions are that the data be coarsened at random [a generalization of missing at random (Rubin, 1976, Biometrika 63, 581-592)] and that the parameters of the data and the coarsening process be distinct. This article presents detailed applications of the general model and the ignorability conditions to a variety of coarse-data problems arising in biomedical statistics. A reanalysis of the Stanford Heart Transplant Data (Crowley and Hu, 1977, Journal of the American Statistical Association 72, 27-36) reveals significant evidence that censoring of pretransplant survival times by transplantation was nonignorable, suggesting a greater benefit from cardiac transplantation than previous analyses had found.",
author = "Heitjan, {D. F.}",
year = "1993",
month = "12",
day = "1",
language = "English (US)",
volume = "49",
pages = "1099--1109",
journal = "Biometrics",
issn = "0006-341X",
publisher = "Wiley-Blackwell",
number = "4",

}

TY - JOUR

T1 - Ignorability and coarse data

T2 - Some biomedical examples

AU - Heitjan, D. F.

PY - 1993/12/1

Y1 - 1993/12/1

N2 - Heitjan and Rubin (1991, Annals of Statistics 19, 2244-2253) define data to be 'coarse' when one observes not the exact value of the data but only some set (a subset of the sample space) that contains the exact value. This definition covers a number of incomplete-data problems arising in biomedicine, including rounded, heaped, censored, and missing data. In analyzing coarse data, it is common to proceed as though the degree of coarseness is fixed in advance-in a word, to ignore the randomness in the coarsening mechanism. When coarsening is actually stochastic, however, inferences that ignore this randomness may be seriously misleading. Heitjan and Rubin (1991) have proposed a general model of data coarsening and established conditions under which it is appropriate to ignore the stochastic nature of the coarsening. The conditions are that the data be coarsened at random [a generalization of missing at random (Rubin, 1976, Biometrika 63, 581-592)] and that the parameters of the data and the coarsening process be distinct. This article presents detailed applications of the general model and the ignorability conditions to a variety of coarse-data problems arising in biomedical statistics. A reanalysis of the Stanford Heart Transplant Data (Crowley and Hu, 1977, Journal of the American Statistical Association 72, 27-36) reveals significant evidence that censoring of pretransplant survival times by transplantation was nonignorable, suggesting a greater benefit from cardiac transplantation than previous analyses had found.

AB - Heitjan and Rubin (1991, Annals of Statistics 19, 2244-2253) define data to be 'coarse' when one observes not the exact value of the data but only some set (a subset of the sample space) that contains the exact value. This definition covers a number of incomplete-data problems arising in biomedicine, including rounded, heaped, censored, and missing data. In analyzing coarse data, it is common to proceed as though the degree of coarseness is fixed in advance-in a word, to ignore the randomness in the coarsening mechanism. When coarsening is actually stochastic, however, inferences that ignore this randomness may be seriously misleading. Heitjan and Rubin (1991) have proposed a general model of data coarsening and established conditions under which it is appropriate to ignore the stochastic nature of the coarsening. The conditions are that the data be coarsened at random [a generalization of missing at random (Rubin, 1976, Biometrika 63, 581-592)] and that the parameters of the data and the coarsening process be distinct. This article presents detailed applications of the general model and the ignorability conditions to a variety of coarse-data problems arising in biomedical statistics. A reanalysis of the Stanford Heart Transplant Data (Crowley and Hu, 1977, Journal of the American Statistical Association 72, 27-36) reveals significant evidence that censoring of pretransplant survival times by transplantation was nonignorable, suggesting a greater benefit from cardiac transplantation than previous analyses had found.

UR - http://www.scopus.com/inward/record.url?scp=0027756948&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0027756948&partnerID=8YFLogxK

M3 - Review article

C2 - 8117903

AN - SCOPUS:0027756948

VL - 49

SP - 1099

EP - 1109

JO - Biometrics

JF - Biometrics

SN - 0006-341X

IS - 4

ER -