Analytical Techniques for Incomplete Block Experiments

Publisher Website
Google Scholar
Abstract
A general and computationally simple method for the maximum likelihood estimation of fixed effects in mixed models is illustrated in the analysis of an incomplete block nutrition experiment. The equivalence of the method to the standard analysis with recovery of interblock information is demonstrated for the general BIB [Balanced incomplete block] design. The variance-covariance matrix of the estimated treatment effects is derived, and some computational simplifications are shown. The method appears to be preferable to conventional analyses, particularly for unbalanced designs.

This publication has 0 references indexed in Scilit: