Abstract
We aim to promote the use of the modified profile likelihood function for estimating the variance parameters of a GLMM in analogy to the REML criterion for linear mixed models. Our approach is based on both quasi-Monte Carlo integration and numerical quadrature, obtaining in either case simulation-free inferential results. We will illustrate our idea by applying it to regression models with binary responses or count data and independent clusters, covering also the case of two-part models. Two real data examples and three simulation studies support the use of the proposed solution as a natural extension of REML for GLMMs. An R package implementing the methodology is available online.
Similar content being viewed by others
References
Barndorff-Nielsen, O.E.: On a formula for the distribution of the maximum likelihood estimator. Biometrika 70, 343–365 (1983)
Beitler, P.J., Landis, J.R.: A mixed-effects model for categorical data. Biometrics 41, 991–1000 (1985)
Booth, J.G., Casella, G., Friedl, H., Hobert, J.P.: Negative binomial loglinear mixed models. Stat. Model. 3, 179–191 (2003)
Caffo, B., Griswold, M.: A user-friendly tutorial on link-probit-normal models. Am. Stat. 60, 139–145 (2006)
Clarkson, D.B., Zhan, Y.: Using spherical-radial quadrature to fit generalized linear mixed effects models. J. Comput. Graph. Stat. 11, 639–659 (2002)
Demidenko, E.: Mixed Models: Theory and Applications. Wiley, New York (2004)
Evans, M., Swartz, T.: Approximating Integrals via Monte Carlo and Deterministic Methods. Oxford University Press, Oxford (2000)
Fang, K.-T., Wang, Y.: Number-Theoretic Methods in Statistics. Chapman & Hall, London (1994)
Gonzalez, J., Tuerlinckx, F., De Boeck, P., Cools, R.: Numerical integration in logistic-normal models. Comput. Stat. Data Anal. 51, 1535–1548 (2006)
Jank, W.: Efficient simulated maximum likelihood with an application to online retailing. Stat. Comput. 16, 111–124 (2006)
Laird, N.M., Ware, J.H.: Random-effects models for longitudinal data. Biometrics 38, 963–974 (1982)
Lee, Y., Nelder, J.A., Pawitan, Y.: Generalized Linear Models with Random Effects: Unified Analysis via H-likelihood. Chapman & Hall/CRC Press, Boca Raton (2006)
Liao, J.G., Lipsitz, S.R.: A type of restricted maximum likelihood estimator of variance components in generalised linear mixed models. Biometrika 89, 401–409 (2002)
Liu, Q., Pierce, D.A.: A note on Gauss-Hermite quadrature. Biometrika 90, 624–629 (1994)
McCullagh, P., Tibshirani, R.: A simple method for the adjustment of profile likelihoods. J. R. Stat. Soc., Ser. B 52, 325–344 (1990)
McCulloch, C.E., Searle, R.E.: Generalized, Linear, and Mixed Models. Wiley, New York (2001)
McGilchrist, C.A.: Estimation in generalized mixed models. J. R. Stat. Soc., Ser. B 56, 61–69 (1994)
Molenberghs, G., Verbeke, G.: Models for Discrete Longitudinal Data. Springer, New York (2005)
Monahan, J.F.: Numerical Methods of Statistics. Cambridge University Press, Cambridge (2001)
Monahan, J., Genz, A.: Spherical-radial integration rules for Bayesian computation. J. Am. Stat. Assoc. 92, 664–674 (1997)
Nie, L.: Convergence rate of MLE in generalized linear and nonlinear mixed-effects models: theory and applications. J. Stat. Plann. Inference 137, 1787–1804 (2007)
Noh, M., Lee, Y.: REML estimation for binary data in GLMMs. J. Multivar. Anal. 98, 896–915 (2007)
Olsen, M.K., Schafer, J.L.: A two-part random-effects model for semicontinuous longitudinal data. J. Am. Stat. Assoc. 96, 730–745 (2001)
Pan, J., Thompson, R.: Gauss-Hermite quadrature approximation for estimation in generalised linear mixed models. Comput. Stat. 18, 57–78 (2003)
Pan, J., Thompson, R.: Quasi-Monte Carlo estimation in generalized linear mixed models. Comput. Stat. Data Anal. 51, 5765–5775 (2007)
Patterson, H.D., Thompson, R.: Recovery of inter-block information when block sizes are unequal. Biometrika 58, 545–554 (1971)
Pinheiro, J.C., Bates, D.M.: Approximation to the log-likelihood function in the nonlinear mixed-effects model. J. Comput. Graph. Stat. 4, 12–35 (1995)
R Development Core Team: R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna (2009)
Severini, T.A.: An approximation to the modified profile likelihood function. Biometrika 85, 403–411 (1998a)
Severini, T.A.: Likelihood functions for the elimination of nuisance parameters. Biometrika 85, 507–522 (1998b)
Severini, T.A.: Likelihood Methods in Statistics. Oxford University Press, Oxford (2000)
Skaug, H.J.: Automatic differentiation to facilitate maximum likelihood estimation in nonlinear random effects models. J. Comput. Graph. Stat. 11, 458–470 (2002)
Stiratelli, R., Laird, N.M., Ware, J.H.: Random effects models for serial observations with binary responses. Biometrics 40, 961–971 (1984)
Tooze, J.A., Grunwald, G.K., Jones, R.H.: Analysis of repeated measures data with clumping at zero. Stat. Methods Med. Res. 11, 341–355 (2002)
Train, K.: Discrete Choice Methods with Simulation. Cambridge University Press, New York (2003)
Waller, L.A., Zelterman, D.: Log-linear modeling with the negative binomial multinomial distribution. Biometrics 53, 971–982 (1997)
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Bellio, R., Brazzale, A.R. Restricted likelihood inference for generalized linear mixed models. Stat Comput 21, 173–183 (2011). https://doi.org/10.1007/s11222-009-9157-4
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11222-009-9157-4