An experiment is to be designed as a completely randomized design with a split-plot treatment assignment. Suppose the wholeplot treatment (A) has four levels and the split-plot treatment (B) has three levels. There are a total of 10 replications of the wholeplot treatment. Assume that both factors A and B have fixed levels.
a. Describe a method of randomizing the experimental units to the levels of factors A and B in this experiment.
b. Write a linear model for this experiment. Make sure to identify each of the terms in the model and list the range of values for all subscripts.
c. Construct an analysis of variance table for this experiment, including columns for sources of variation, degrees of freedom, and expected mean squares.