✦ LIBER ✦

Non-ignorable missing covariates in generalized linear models

✍ Scribed by Stuart R. Lipsitz; Joseph G. Ibrahim; Ming-Hui Chen; Harriet Peterson

Publisher: John Wiley and Sons
Year: 1999
Tongue: English
Weight: 123 KB
Volume: 18
Category: Article
ISSN: 0277-6715
DOI: 10.1002/(sici)1097-0258(19990915/30)18:17/18<2435::aid-sim267>3.0.co;2-b

No coin nor oath required. For personal study only.

✦ Synopsis

We propose a likelihood method for estimating parameters in generalized linear models with missing covariates and a non-ignorable missing data mechanism. In this paper, we focus on one missing covariate. We use a logistic model for the probability that the covariate is missing, and allow this probability to depend on the incomplete covariate. We allow the covariates, including the incomplete covariate, to be either categorical or continuous. We propose an EM algorithm in this case. For a missing categorical covariate, we derive a closed form expression for the E-and M-steps of the EM algorithm for obtaining the maximum likelihood estimates (MLEs). For a missing continuous covariate, we use a Monte Carlo version of the EM algorithm to obtain the MLEs via the Gibbs sampler. The methodology is illustrated using an example from a breast cancer clinical trial in which time to disease progression is the outcome, and the incomplete covariate is a quality of life physical well-being score taken after the start of therapy. This score may be missing because the patients are sicker, so this covariate could be non-ignorably missing.

📜 SIMILAR VOLUMES

NON-RESPONSE MODELS FOR THE ANALYSIS OF

NON-RESPONSE MODELS FOR THE ANALYSIS OF NON-MONOTONE NON-IGNORABLE MISSING DATA

✍ JAMES M. ROBINS 📂 Article 📅 1997 🏛 John Wiley and Sons 🌐 English ⚖ 353 KB 👁 1 views

I introduce a new class of non-ignorable non-monotone missing data models. These models are useful for investigating the sensitivity of one's estimates to untestable assumptions about the missing data process. I use the new models to analyse data from a case-control study of the effect of radiation

NON-RESPONSE MODELS FOR THE ANALYSIS OF

NON-RESPONSE MODELS FOR THE ANALYSIS OF NON-MONOTONE IGNORABLE MISSING DATA

✍ JAMES M. ROBINS; RICHARD D. GILL 📂 Article 📅 1997 🏛 John Wiley and Sons 🌐 English ⚖ 332 KB 👁 1 views

We discuss a new class of ignorable non-monotone missing data models -the randomized monotone missingness (RMM) models. We argue that the RMM models represent the most general plausible physical mechanism for generating non-monotone ignorable data. We show that there exists ignorable missing data pr

Modeling continuous auxiliary covariate

Modeling continuous auxiliary covariate data in generalized linear mixed models using the kernel smoother

✍ Jianwei Chen; Chii-Dean Lin 📂 Article 📅 2010 🏛 John Wiley and Sons 🌐 English ⚖ 256 KB

Generalized linear modelling in geomorph

Generalized linear modelling in geomorphology

✍ Atkinson, Peter; Jiskoot, Hester; Massari, Remo; Murray, Tavi 📂 Article 📅 1998 🏛 John Wiley and Sons 🌐 English ⚖ 270 KB 👁 2 views

Generalized linear modelling (GLM) is a statistical technique used to model the relation between a response variable and a set of explanatory variables. GLM is similar to the well known multiple regression. However, GLM is a powerful technique for exploratory data analysis with many advantages over

Resampling Methods in Generalized Linear

Resampling Methods in Generalized Linear Models Useful in Environmetrics

✍ Herwig Friedl; Ernst Stadlober 📂 Article 📅 1997 🏛 John Wiley and Sons 🌐 English ⚖ 208 KB 👁 3 views

Generalized linear models are important tools for analysing relationships between binary, count or continuous response variables and predictors with ®xed eects. In this paper we present a survey on bootstrap methods based on (extended) quasi-likelihood assumptions. We discuss two approaches: one-ste

PAC learning in non-linear FIR models

✍ K. Najarian; G. A. Dumont; M. S. Davies; N. E. Heckman 📂 Article 📅 2001 🏛 John Wiley and Sons 🌐 English ⚖ 140 KB 👁 1 views