What is the standard deviation in a Taguchi design?

In a Taguchi design, the standard deviation is the variability in the response because of noise. Usually your goal is to choose factor levels that minimize the standard deviation. Minitab calculates a separate standard deviation for each combination of control factors in the design.

For example, a group of scientists wants to know how five control factors affect a plant's rate of growth. They consider two noise factors, temperature and humidity. The standard deviations for this example provide estimates of the variability in the plant growth because of temperature and humidity for each of the five control factors.

For a static design, Minitab calculates the standard deviation of the responses. For a Taguchi dynamic design, Minitab uses the square root of MSE, or mean square of the error. MSE is the square of the distances (residuals) divided by n – 2, where n is the number of observations.