The problem of determining the estimability of linear parametric
functions in classification models is examined. The two estimability
problems which are specifically aimed at are: the problem of determining
whether the design matrix is of maximal rank (i. e. , all cell
expectations are estimable); and in the event that...
A classification scheme will group the blocks of an experimental
design model into distinct categories, resulting in a new design with
classification parameters instead of block parameters. If the original
design is an incomplete block design, some balance is still retained
in the classification design, allowing the theoretical development of...