Ghent University Academic Bibliography

Advanced

Analyzing the sensitivity of generalized linear models to incomplete outcomes via the IDE algorithm

Stijn Vansteelandt UGent and Els Goetghebeur UGent (2001) JOURNAL OF COMPUTATIONAL AND GRAPHICAL STATISTICS. 10(4). p.656-672
abstract
Incomplete data models typically involve strong untestable assumptions about the missing data distribution. As inference may critically depend on them, the importance of sensitivity analysis is well recognized. Molenberghs, Kenward, and Goetghebeur proposed a formal frequentist approach to sensitivity analysis which distinguishes ignorance due to unintended incompleteness from imprecision due to finite sampling by design. They combine both sources of variation into uncertainty. This article develops estimation tools for ignorance and uncertainty concerning regression coefficients in a complete data model when some of the intended outcome values are missing. Exhaustive enumeration of all possible imputations for the missing data requires enormous computational resources. In contrast, when the boundary of the occupied region is of greatest interest, reasonable computational efficiency may be achieved via the imputation towards directional extremes (IDE) algorithm. This is a special imputation method designed to mark them boundary of the region by maximizing the, direction of change of the complete data estimator caused by perturbations to the imputed outcomes. For multi-dimensional parameters, a dimension reduction approach is considered. Additional insights are obtained by considering structures within the region, and by introducing external knowledge to narrow the boundary to useful proportions. Special properties hold for the generalized linear model. Examples from a Kenyan HIV study will illustrate the points.
Please use this url to cite or link to this publication:
author
organization
year
type
journalArticle (original)
publication status
published
subject
keyword
INFERENCE, REGRESSION, SEMIPARAMETRIC NONRESPONSE MODELS, enumeration, imputation, missing data, nonignorable, parametric regression analysis, BOUNDS, sensitivity analysis
journal title
JOURNAL OF COMPUTATIONAL AND GRAPHICAL STATISTICS
J. Comput. Graph. Stat.
volume
10
issue
4
pages
656 - 672
Web of Science type
Article
Web of Science id
000173007800004
JCR category
STATISTICS & PROBABILITY
JCR impact factor
0.568 (2001)
JCR rank
29/70 (2001)
JCR quartile
2 (2001)
ISSN
1061-8600
DOI
10.1198/106186001317243386
language
English
UGent publication?
yes
classification
A1
id
1894158
handle
http://hdl.handle.net/1854/LU-1894158
date created
2011-08-29 14:39:25
date last changed
2016-12-19 15:41:56
@article{1894158,
  abstract     = {Incomplete data models typically involve strong untestable assumptions about the missing data distribution. As inference may critically depend on them, the importance of sensitivity analysis is well recognized. Molenberghs, Kenward, and Goetghebeur proposed a formal frequentist approach to sensitivity analysis which distinguishes ignorance due to unintended incompleteness from imprecision due to finite sampling by design. They combine both sources of variation into uncertainty. This article develops estimation tools for ignorance and uncertainty concerning regression coefficients in a complete data model when some of the intended outcome values are missing. Exhaustive enumeration of all possible imputations for the missing data requires enormous computational resources. In contrast, when the boundary of the occupied region is of greatest interest, reasonable computational efficiency may be achieved via the imputation towards directional extremes (IDE) algorithm. This is a special imputation method designed to mark them boundary of the region by maximizing the, direction of change of the complete data estimator caused by perturbations to the imputed outcomes. For multi-dimensional parameters, a dimension reduction approach is considered. Additional insights are obtained by considering structures within the region, and by introducing external knowledge to narrow the boundary to useful proportions. Special properties hold for the generalized linear model. Examples from a Kenyan HIV study will illustrate the points.},
  author       = {Vansteelandt, Stijn and Goetghebeur, Els},
  issn         = {1061-8600},
  journal      = {JOURNAL OF COMPUTATIONAL AND GRAPHICAL STATISTICS},
  keyword      = {INFERENCE,REGRESSION,SEMIPARAMETRIC NONRESPONSE MODELS,enumeration,imputation,missing data,nonignorable,parametric regression analysis,BOUNDS,sensitivity analysis},
  language     = {eng},
  number       = {4},
  pages        = {656--672},
  title        = {Analyzing the sensitivity of generalized linear models to incomplete outcomes via the IDE algorithm},
  url          = {http://dx.doi.org/10.1198/106186001317243386},
  volume       = {10},
  year         = {2001},
}

Chicago
Vansteelandt, Stijn, and Els Goetghebeur. 2001. “Analyzing the Sensitivity of Generalized Linear Models to Incomplete Outcomes via the IDE Algorithm.” Journal of Computational and Graphical Statistics 10 (4): 656–672.
APA
Vansteelandt, S., & Goetghebeur, E. (2001). Analyzing the sensitivity of generalized linear models to incomplete outcomes via the IDE algorithm. JOURNAL OF COMPUTATIONAL AND GRAPHICAL STATISTICS, 10(4), 656–672.
Vancouver
1.
Vansteelandt S, Goetghebeur E. Analyzing the sensitivity of generalized linear models to incomplete outcomes via the IDE algorithm. JOURNAL OF COMPUTATIONAL AND GRAPHICAL STATISTICS. 2001;10(4):656–72.
MLA
Vansteelandt, Stijn, and Els Goetghebeur. “Analyzing the Sensitivity of Generalized Linear Models to Incomplete Outcomes via the IDE Algorithm.” JOURNAL OF COMPUTATIONAL AND GRAPHICAL STATISTICS 10.4 (2001): 656–672. Print.