James, Alan T.; Venables, William N. Matrix weighting of several regression coefficient vectors. (English) Zbl 0778.62048 Ann. Stat. 21, No. 2, 1093-1114 (1993). Summary: For small sample random effects models, results are derived which show in certain cases, and indicate in general, that an estimated random effects variance matrix may be used in the weight matrices without causing undue error in the empirically weighted mean. Exact error variances are derived mathematically for the empirically weighted mean for the two sample case in one and two dimensions.Simulation is used to determine errors for a practical example of six five-variate samples. For estimation of their mean, the differences between the samples are ancillary. The biases of the average and weighted mean estimators conditional on these ancillaries are illustrated in a diagram plotting values obtained by simulation. A curious range anomaly is illustrated which arises if random effects are ignored when present. MSC: 62H12 Estimation in multivariate analysis 62J10 Analysis of variance and covariance (ANOVA) 65C99 Probabilistic methods, stochastic differential equations Keywords:exact error variances; matrix weighting; estimated generalized least squares; residual maximum likelihood; moment estimator; conditional bias; cutoff function; efficiency; unbalanced data; small sample random effects models; estimated random effects variance matrix; weight matrices; empirically weighted mean; two sample case; biases; ancillaries; simulation; range anomaly × Cite Format Result Cite Review PDF Full Text: DOI