What is the primary advantage of a 'Split-plot design' in an experiment?
It allows for different levels of precision for different factors
It completely eliminates nuisance variables
It removes the need for randomization
It reduces the total number of experimental units to one