Bias correction in random effects models with sparse binary responses

Abstract

Sparse correlated binary data are frequently encountered in many applications involving either rare event cases or small sample sizes. In this study, we consider correlated binary data and a logit random effects model framework. We discuss h-likelihood estimates and how the computational procedure is affected by sparseness. We propose an adjustment to the fitting process that involves the adaption of the regression calibration method to the estimation of random effects. Using this adjustment, we correct for the bias in the random effects estimates resulting in better properties for the fixed effects estimates of the model. This is supported by the results of the simulation study that was conducted under different sparseness levels. The proposed adjusted h-likelihood estimation approach is also used for the analysis of two real meta-analyses data sets.

Keywords

Sparse binary data h-likelihood calibration approach bias correction random intercept models

Get full access to this article

View all access options for this article.

References

Sweeting

Sutton

Lambert

. What to add to nothing? use and avoidance of continuity corrections in meta-analysis of sparse data. Stat Med 2004; 23: 1351–1375.

Simmonds

Higgins

PTJ

. A general framework for the use of logistic regression models in meta-analysis. Stat Method Med Res 2016; 25: 2858–2877.

Cole

Chu

Greenland

. Maximum likelihood, profile likelihood and penalized likelihood: A primer. Am J Epidemiol 2014; 179: 252–260.

Lee

Nelder

. Hierarchical generalized linear models (with discussion). J R Stat Soc B 1996; 58: 619–656.

Lee

Nelder

. Hierarchical generalized linear models: A synthesis of generalized linear models, random effect model and structured dispersion. Biometrika 2001; 88: 987–1006.

Lee

Nelder

. Double hierarchical generalized linear models (with discussion). Appl Stat 2006; 55: 139–185.

Lee

Nelder

Pawitan

. Generalized Linear Models with Random Effects. Unified Analysis via H-likelihood. Boca Raton, FL: Chapman and Hall, 2007.

Chernozhukov

Chetverikov

Demirer

, et al. Double/Debiased machine learning for treatment and structural parameters. Economet J 2018; 21: C1–C68.

Firth

. Bias reduction of maximum likelihood estimates. Biometrika 1993; 80: 27–38.

10.

Kosmidis

Firth

. Bias reduction in exponential family nonlinear models. Biometrika 2009; 96: 793–804.

11.

Cox

Reid

. Parameter orthogonality and approximate conditional inference. J R Stat Soc B 1987; 49: 1–39.

12.

Kosmidis

Guolo

Varin

. Improving the accuracy of likelihood-based inference in meta-analysis and meta-regression. Biometrika 2017; 104: 489–496.

13.

Kuss

. Statistical methods for meta-analyses including information from studies wihout any events - add nothing to nothing and succeed nevertheless. Stat Med 2015; 34: 1097–1116.

14.

Jeon

Hsu

Gorfine

. Bias correction method in the hierarchical likelihood approace to the analysis of multivariate survival data. Biostatistics 2011; 0: 1–14.

15.

Yun

Lee

. Comparison of hierarchical and marginal likelihood estimators for binary outcomes. Comput Stat Data Anal 2004; 45: 639–650.

16.

Noh

Lee

. REML estimation for binary data in GLMMs. J Multivar Anal 2007; 98: 896–915.

17.

Carroll

Ruppert

Stefanski

, et al. Measurement Error in Nonlinear Models. Boca Raton, FL: Chapman & Hall/CRC, 2006.

18.

Corroll

Stefanski

. Approximate quasilikelihood estimation in models with surrogate predictors. J Am Stat Assoc 1990; 85: 652–663.

19.

Prentice

. Covariate measurement errors and parameters estimation in failure time regression models. Biometrika 1982; 69: 331–342.

20.

Armstrong

. Measurement error in generalized linear models. Commun Stat, Ser B 1985; 14: 529–544.

21.

Wang

Hsu

Feng

, et al. Regression calibration in failure time regression. Biometrics 1997; 53: 131–145.