Efficiency of incomplete split-plot designs – a compromise between traditional split-plot designs and randomised complete block design

Summary

The paper shows, how incomplete split-plot designs can be constructed from -designs and how they can be analysed. The incomplete split-plot design can be regarded as both a practical and statistical compromise between traditional split-plot design and randomised complete block design. The efficiency of the design is compared to traditional split-plot design and randomised complete block design using data from 5 trials carried out using incomplete split-plot designs in Denmark through 2004-2007. The comparisons showed that the incomplete split-plot design were superior to traditional split-plot design in most cases – and in all cases when comparing differences among treatments in the whole-plot factor. The incomplete split-plot design was in many cases also better than the randomised complete block design.