Mixed models and unbalanced data: wherefrom, whereat and whereto?
- 1 January 1988
- journal article
- research article
- Published by Taylor & Francis in Communications in Statistics - Theory and Methods
- Vol. 17 (4) , 935-968
- https://doi.org/10.1080/03610928808829667
Abstract
A brief history of the early years (1820-1947) of random effects models and the estimation of variance components is followed by a personal evaluation of M.L, REML and MINQUE estimation. A method is suggested for combining ML estimator obtained from subsets of a large data set, and comments are made on the need for simulation studies to assess the degree of approximation in using asymptotic properties of ML-type estimators as if they were exact for finite-sized unbalanced data sets.Keywords
This publication has 26 references indexed in Scilit:
- Asymptotic Behavior of Minque-Type Estimators of Variance ComponentsThe Annals of Statistics, 1976
- Comparisons of Designs and Estimation Procedures for Estimating Parameters in a Two-Stage Nested ProcessTechnometrics, 1967
- Maximum-likelihood estimation for the mixed analysis of variance modelBiometrika, 1967
- Theorems Concerning Eisenhart's Model IIThe Annals of Mathematical Statistics, 1961
- Properties of Model II--Type Analysis of Variance Tests, A: Optimum Nature of the $F$-Test for Model II in the Balanced CaseThe Annals of Mathematical Statistics, 1959
- A Note on Uniformly Best Unbiased Estimators for Variance ComponentsJournal of the American Statistical Association, 1956
- Estimation of Variance and Covariance ComponentsBiometrics, 1953
- The Use of the Analysis of Variance in Enumeration by SamplingJournal of the American Statistical Association, 1939
- The Estimation of Components of VarianceJournal of the Royal Statistical Society Series B: Statistical Methodology, 1939
- XV.—The Correlation between Relatives on the Supposition of Mendelian Inheritance.Transactions of the Royal Society of Edinburgh, 1919