Misspecified maximum likelihood estimates and generalised linear mixed models

By Patrick J. Heagerty, B. Kurland. 2001

302 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/bdc2fbf90c34abaddf67624b5620a2b2ee12f581

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Statistics and Probability

Citations in the corpus

Modelling sparse generalized longitudinal observations with latent Gaussian processes (2008, methodology): “…on applying such models to repeated binary measurements, Pourahmadi (2000) for related aspects of covariance modeling, and Heagerty and Zeger (2000), Heagerty and Kurland (2001) and Chiou and Müller (2005) for discussions on limitations, modifications and feasibility of the underlying parametric…”
Misspecifying the Shape of a Random Effects Distribution: Why Getting It Wrong May Not Matter (2011, methodology): “Heagerty and Kurland (2001), in their Table 1, show vir- tually no impact on the asymptotic bias in logistic regression models with a wide variety of gammadistributed random effects.”
Conditional and Marginal Models: Another View (2004, background): “Such biases were discussed in detail by Neuhaus, Hauck and Kalbfleisch (1992) and Heagerty and Kurland (2001).”
Separating between‐ and within‐cluster covariate effects by using conditional and partitioning methods (2006, background): “An exception to this is Heagerty and Kurland (2001) in which unequal random-effects variance as a function of the covariate can bias the estimators of the regression parameters.”
Semiparametric estimation in general repeated measures problems (2006, background): “Here the variance component σ20 may be of interest in itself and may in some cases depend on components of X such as gender; see Heagerty and Kurland (2001) for an example.”
Intermediate and advanced topics in multilevel logistic regression analysis (2017, background): “Similarly, Heagerty and Kurland found that the asymptotic relative bias in estimated regression coefficients was small in nearly all of the scenarios that they examined [44] (see Table I of their paper).”
Robustness of linear mixed‐effects models to violations of distributional assumptions (2020, background): “For generalized linear mixed-effects models, the patterns have been more variable and fixed effect estimates can be biased in some cases (Grilli & Rampichini, 2015; Heagerty & Kurland, 2001).”