Abstract
Testing of hypotheses in unbalanced incomplete block designs. The randomized complete block design enables an experimenter to reduce the error variance and obtain more precise estimates of treatment effects. In many experiments the size block must be small to obtain homogeneity of the experimental units, and increase the precision of the design. In this case it is possible to use randomized block design s in which every treatment it not present in every block, although if are presented according to some specified pattern. These designs are known as balanced o partially balanced incomplete block designs. However, it may happen by accident or some other reason, that the investigators is forced to use incomplete block designs, but unbalanced. The purpose of this work is to show a method that allows determining the functions and contrasts that are estimable and the associated hypotheses that are of general interest.
