Methods Inf Med 2010; 49(05): 426-432
DOI: 10.1055/s-0038-1625133
Original Articles
Schattauer GmbH

Discussion of “Generalized Estimating Equations: Notes on the Choice of the Working Correlation Matrix”

J. Breitung
1   University of Bonn, Bonn, Germany
,
N. R. Chaganty
2   Old Dominion University, Norfolk, VA, USA
,
R. M. Daniel
3   London School of Hygiene and Tropical Medicine, London, UK
,
M. G. Kenward
3   London School of Hygiene and Tropical Medicine, London, UK
,
M. Lechner
4   University of St. Gallen, St. Gallen, Switzerland
,
P. Martus
5   Charité – Universitätsmedizin Berlin, Berlin, Germany
,
R. T. Sabo
6   Virginia Commonwealth University, Richmond, VA, USA
,
Y.-G. Wang
7   The University of Queensland, St. Lucia, Queensland, Australia
,
C. Zorn
8   Pennsylvania State University, University Park, PA, USA
› Institutsangaben
Weitere Informationen

Publikationsverlauf





Publikationsdatum:
20. Januar 2018 (online)

Summary

Objective: To discuss generalized estimating equations as an extension of generalized linear models by commenting on the paper of Ziegler and Vens “Generalized Estimating Equations: Notes on the Choice of the Working Correlation Matrix”.

Methods: Inviting an international group of experts to comment on this paper.

Results: Several perspectives have been taken by the discussants. Econometricians have established parallels to the generalized method of moments (GMM). Statisticians discussed model assumptions and the aspect of missing data. Applied statisticians commented on practical aspects in data analysis.

Conclusions: In general, careful modeling correlation is encouraged when considering estimation efficiency and other implications, and a comparison of choosing instruments in GMM and generalized estimating equations (GEE) would be worthwhile. Some theoretical drawbacks of GEE need to be further addressed and require careful analysis of data. This particularly applies to the situation when data are missing at random.

 
  • References

  • 1 Ziegler A, Vens M. Generalized Estimating Equations: Notes on the Choice of the Working Correlation Matrix. Methods Inf Med 2010; 49 (05) 421-425.
  • 2 Hayashi F. Econometrics. Princeton University Press; 2000
  • 3 Breitung J, Lechner M. Some GMM Estimation Methods and Specification Tests for Nonlinear Models. In: Matyas L, Sevestre P. (eds). The Econometrics of Panel Data. 2nd edition. Dordrecht: Kluwer; 1996. pp 583-612.
  • 4 Andrews DWK. Consistent Moment Selection Procedures for Generalized Methods of Moments Estimation. Econometrica 1999; 67: 543-564.
  • 5 Donald SG, Newey WK. Choosing the number of instruments. Econometrica 2001; 69: 1161-1192.
  • 6 Doran H, Schmidt P. GMM Estimators with Improved Finite Sample Properties Using Principal Components of the Weighting Matrix, with an Application to the Dynamic Panel Data Model. Journal of Econometrics 2006; 133: 387-409.
  • 7 Hall AR, Peixe FPM. A consistent method for the selection of relevant instruments. Econometric Reviews 2003; 3: 269-287.
  • 8 Hall AR, Inoue A, Jana K, Shin C. Information in generalized method of moments estimation and entropy-based moment selection. Journal of Econometrics 2007; 138: 488-512.
  • 9 Chaganty NR, Joe H. Range of correlation matrices for dependent Bernoulli random variables. Biometrika 2006; 93 (01) 197-206.
  • 10 Chaganty NR, Mav D. Estimation methods for analyzing longitudinal data occurring in biomedical research. In: Khattree R, Naik DN. (eds). Computational Methods in Biomedical Research. London, Boca Raton, FL: Chapman and Hall, CRC; 2007. 12 371-400.
  • 11 Sabo RT, Chaganty NR. What can go wrong when ignoring correlation bounds in the use of generalized estimating equations. Statistics in Medicine. 2010 DOI: 10.1002/sim.4013.
  • 12 Ashford JR, Sowden RR. Multi-variate probit analysis. Biometrics 1970; 26 (03) 535-546.
  • 13 Crowder M. On the use of working correlation matrix in using generalised linear models for repeated measures. Biometrika 1995; 82: 407-410.
  • 14 Chaganty NR, Joe H. Efficiency of generalized estimating equations for binary responses. Journal of the Royal Statistical Society, B 2004; 66 (04) 851-860.
  • 15 Lee Y, Nelder JA. Conditional and marginal models: Another view. Statistical Science 2004; 19 (02) 219-238.
  • 16 Daniel RM. On Aspects of Robustness and Sensitivity in Missing Data Methods. Unpublished PhD thesis. London School of Hygiene and Tropical Medicine, UK: 2009
  • 17 Liang K-Y, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika 1986; 73: 13-22.
  • 18 Molenberghs G, Kenward MG. Missing Data in Clinical Studies. Chichester: Wiley; 2007
  • 19 Rubin DB. Inference and missing data. Biometrika 1976; 63: 581-592.
  • 20 Seaman S, Copas A. J Doubly robust generalised estimating equations for longitudinal data. Statistics in Medicine 2009; 28: 937-955.
  • 21 Tsiatis AA. Semiparametric Theory and Missing Data. New York: Springer; 2006
  • 22 Bertschek I, Lechner M. Convenient estimators for the panel probit model. Journal of Econometrics 1998; 87: 329-371.
  • 23 Chamberlain G. Asymptotic efficiency in estimation with conditional moment restrictions. Journal of Econometrics 1987; 34: 305-334.
  • 24 Hansen L. Large sample properties of generalized methods of moments estimators. Econometrica 1982; 50: 1029-1055.
  • 25 Newey WK. Effcient estimation of models with conditional moment restrictions. In: Maddala G, Rao C, Vinod H. (eds.). Handbook of Statistics, vol 11, chap 16. North-Holland, Amsterdam: 1993
  • 26 Newey WK. Efficient instrumental variables estimation of nonlinear models. Econometrica 1990; 59: 809-837.
  • 27 Fitzmaurice GM. A caveat concerning independence estimating equations with multivariate binary data. Biometrics 1995; 51: 309-317.
  • 28 Wang Y-G, Hin L-Y. Modeling strategies in longitudinal data analysis: Covariate, variance function and correlation structure selection. Computational Statistics and Data Analysis. 2009 Doi: 10.1016/j.csda.2009.11.006.
  • 29 Hubbard AE, Ahern J, Fleischer N, Laan M, Lippman S, Jewell N, Bruckner T, Satariano W. To GEE or Not to GEE: Comparing Population Average and Mixed Models for Estimating the Associations Between Neighborhood Risk Factors and Health. Epidemiology 2010; 21: 467-474.