The question of how to analyze unbalanced hierarchical data generated from structural equation models has been a common problem for researchers and analysts. Among difficulties plaguing statistical modeling are estimation bias due to measurement error and the estimation of the effects of the individual's hierarchical social milieu. This paper presents empirical Bayes estimation by means of the EM algorithm in the context of unbalanced sampling designs. The EM algorithm is particularly useful when analytic expressions exist for the conditional distribution of the missing data given complete data and for the complete-data maximum likelihood estimates of the model parameters. One example is included for illustrating the specification of the model that incorporates variations at two levels of hierarchy. (Contains 2 tables, 1 figure, and 32 references.) (Author/SLD)