An optimal design uses the "best" group of design points, selected from reducing or augmenting the number of experimental runs in the original design. Minitab's optimal design capabilities can be used with general full factorial designs, response surface designs, and mixture designs. For more information, go to Overview for Select Optimal Design.
For example, a greenhouse manager wants to use a designed experiment to determine the ideal combination of temperature, light, soil, and water that maximizes the growth of hanging ferns. By using a designed experiment, the manager can evaluate fern growth in response to the various combinations of the factor levels. To do this, the manager creates a response surface design with four factors and two blocks (the blocks represent the two adjacent greenhouses), which results in a design with 30 points (that is, 30 factor level combinations). However, to save money and time, the manager decides to reduce the original design to 20 points, using D-optimality as the criterion to select the best set of design points. The manager can then evaluate fern growth based on only 20 experimental runs.