Use Predict Taguchi Results to predict values for characteristics at specified factor settings. The predicted results can help you decide which factor settings lead to the best results for your process or product. You should conduct confirmation runs at the settings you choose and compare the actual results to the predictions.

For this example, the scientists include Variety, Light, Fertilizer and Water in the model. They do not include Spraying or the interaction terms because they were not significant in the regression/ANOVA results in Analyze Taguchi Design.

Several scientists are interested in examining the effect of five factors on the growth of young basil plants. There are five control factors, each with two levels. They are:

·    Variety of plant

·    Amount of light

·    Amount of fertilizer

·    Amount of water

·    Frequency of spraying

The scientists selected two noise factors, temperature and humidity, each with two levels, and combined them to make four noise conditions. Time was used as a signal factor, making this a dynamic response experiment. The response is leaf size measurement. The scientists also test the interaction between variety and fertilizer. The goal of the experiment is to determine which factor settings increase the plant's rate of growth (slope) without increasing the variability in growth.

Data: Basil2.MTW (available in the Sample Data folder).