Dear All,
Where would I be able to find a description (with the appropriate ANOVA model and table) of a split-plot design with subsampling?
Just in case, let me briefly describe the design. Replications (random) are whole plots randomly assigned to receive one level of treatment A (fixed). Each plot is then subdivided into split-plots, where each split-plot is randomly assigned to receive one level of treatment B (fixed). And within each split-plot, a random sample is taken, constituting the observational units. So, the design looks like this (of course, there could be more levels of the treatments, more replicates, and more subsamples):
Split1 Split2
Rep1 Sub1 Sub2 Sub1 Sub2
Trt1 Rep2 Sub1 Sub2 Sub1 Sub2
Rep3 Sub1 Sub2 Sub1 Sub2
Rep1 Sub1 Sub2 Sub1 Sub2
Trt2 Rep2 Sub1 Sub2 Sub1 Sub2
Rep3 Sub1 Sub2 Sub1 Sub2
Taking the average of the subsamples yields a standard split-plot design. But without this averaging, the subsamples represent an additional random effect in this model and this is a nested effect. I think this design is sometimes used in the agricultural sciences.
Any pointers, hints, and suggestions most welcome!
--
Wolfgang Viechtbauer
Department of Methodology and Statistics
University of Maastricht, The Netherlands
http://www.wvbauer.com/
