deSigns of the Times, Or is Latin a Dead Language?
The previous article initiated a discussion on factorial experiments. The article concentrated on full factorial experiments where all combinations of every level of all variables are tested. For example, if four variables were to be evaluated at four levels, 256 experiments would have to be done. When continuous variables are studied, two level full factorial designs with center points can be used to minimize the number of experiments; for example, four variables can be studied at two levels and would require 16 experiments and five center points. Using only a fraction of the full factorial experiments can have further savings. This article begins a discussion on fractional factorial designs, but I'm going to come from a different direction.
The experimental factor of interest is the four different formulations. The response or test of interest is the dry time.