Complete the following steps to determine the value to use for the number of blocks in a power and sample size analysis for a 2-level factorial design or a Plackett-Burman design.
- Start with one block and calculate how many replicates of the base design you need to get the power that you want.
- Add blocks based on the number of runs you have. You must be able to evenly divide the number of corner points times the number of replicates by the number of blocks.
- Rerun the power and sample size analysis to determine how the additional blocks affect your power. Including additional blocks in your model can reduce your power by taking away degrees of freedom from the error term. You can increase power by adding center points to your model.