Multiple imputation is a method of estimating the variances of estimators that are constructed with some imputed data. We give an expression for the bias of the multiple-imputation variance estimator for data that are collected with a complex sample design. The bias may be sizable for certain...

Although welfare reform began in 1996 at the national level, Iowa was one of the earliest states to obtain a waiver to initiate the Iowa Family Investment Program (FIP) in 1993. To gain a better understanding of welfare recidivism, we use Iowa administrative quarterly data between October1993...

The use of segments of polynomials to approximate production surfaces and time-series trends is described and illustrated. These segmented curves are restricted to be continuous and have a continuous derivative(s) at the join points.

Prediction for the mixed model requires estimates of covariance matrices. There is often a direct estimate of the “within area” covariance matrix, and for survey samples this is an estimate of the sampling covariance matrix. The estimated covariance matrix may have large sampling variance,...