Instructions

This calculator estimates the minimum number of trials you will need to fit a specific model. It also estimates the total amount of work you can expect by adding 3 extra trials and 8 replicates to this minimum number.

It's best to add at least 3 extra trials in a real experiment.

You will also want to add at least 8 runs more than the minimum to judge the quality of your data.


(Note: If you're really budget limited you can often get away with 5 runs more than the minimum as long as you are careful to test your Sweet Spot before reporting it.)

An I-Optimal Design is most efficient.