Practical Use of Modified Maximum Likelihoods for Stratified Data

Stratified data arise in several settings, such as longitudinal studies or multicenter clinical trials. Between-strata heterogeneity is usually addressed by random effects models, but an alternative approach is given by fixed effects models, which treat the incidental nuisance parameters as fixed unknown quantities. This approach presents several advantages, like computational simplicity and robustness to confounding by strata. However, maximum likelihood estimates of the parameter of interest are typically affected by incidental parameter bias. A remedy to this is given by the elimination of stratum-specific parameters by exact or approximate conditioning. The latter solution is afforded by the modified profile likelihood, which is the method applied in this paper. The aim is to demonstrate how the theory of modified profile likelihoods provides convenient solutions to various inferential problems in this setting. Specific procedures are available for different kinds of response variables, and they are useful both for inferential purposes and as a diagnostic method for validating random effects models. Some examples with real data illustrate these points.

[1]  O. Barndorff-Nielsen.  On a formula for the distribution of the maximum likelihood estimator , 1983 .

[2]  M. Lesperance,et al.  Estimation efficiency in a binary mixed-effects model setting , 1996 .

[3]  D. Ruppert,et al.  Transformation and Weighting in Regression , 1988 .

[4]  T. Severini.  Likelihood Methods in Statistics , 2001 .

[5]  J. Neyman,et al.  Consistent Estimates Based on Partially Consistent Observations , 1948 .

[6]  M Palta,et al.  Analysis of longitudinal data with unmeasured confounders. , 1991, Biometrics.

[7]  D. Bates,et al.  Mixed-Effects Models in S and S-PLUS , 2001 .

[8]  P. Allison,et al.  7. Fixed-Effects Negative Binomial Regression Models , 2002 .

[9]  Brian D. Ripley,et al.  Modern Applied Statistics with S Fourth edition , 2002 .

[10]  J. Kalbfleisch,et al.  Between- and within-cluster covariate effects in the analysis of clustered data. , 1998, Biometrics.

[11]  T R Ten Have,et al.  An Empirical Comparison of Several Clustered Data Approaches Under Confounding Due to Cluster Effects in the Analysis of Complications of Coronary Angioplasty , 1999, Biometrics.

[12]  Nicola Sartori,et al.  Modified profile likelihoods in models with stratum nuisance parameters , 2003 .

[13]  Alessandra Rosalba Brazzale,et al.  Approximate Conditional Inference in Logistic and Loglinear Models , 1999 .

[14]  P. Heagerty,et al.  Misspecified maximum likelihood estimates and generalised linear mixed models , 2001 .

[15]  Youngjo Lee.  Fixed‐effect versus random‐effect models for evaluating therapeutic preferences , 2002, Statistics in Medicine.

[16]  P. McCullagh,et al.  Generalized Linear Models , 1992 .

[17]  J. Kalbfleisch,et al.  A Comparison of Cluster-Specific and Population-Averaged Approaches for Analyzing Correlated Binary Data , 1991 .

[18]  L. Skovgaard.  NONLINEAR MODELS FOR REPEATED MEASUREMENT DATA. , 1996 .

[19]  M Palta,et al.  Effect of omitted confounders on the analysis of correlated binary data. , 1997, Biometrics.

[20]  P. Diggle.  Analysis of Longitudinal Data , 1995 .

[21]  Luigi Pace,et al.  Principles of statistical inference : from a neo-Fisherian perspective , 1997 .

[22]  Nicola Sartori,et al.  Extending conditional likelihood in models for stratified binary data , 2003 .

[23]  Nicola Sartori,et al.  CONDITIONAL LIKELIHOOD INFERENCE IN GENERALIZED LINEAR MIXED MODELS , 2004 .

[24]  Applications of Likelihood Asymptotics for Nonlinear Regression in Herbicide Bioassays , 2000, Biometrics.

[25]  Herwig Friedl,et al.  Negative binomial loglinear mixed models , 2003 .