Overview of Predict Taguchi Results

Use Predict Taguchi Results to predict signal-to-noise ratios (S/N ratios) and other response characteristics for selected factor settings using the model from your Taguchi experiment. In order to predict results, you need to have analyzed the design using Analyze Taguchi Design.

For example, you can do the following:
  • Examine the response tables and main effects plots to identify the factors and settings that have the largest effect on the S/N ratio or standard deviation.
  • Choose several combinations of settings from the other factors.
  • From the prediction results, determine which combination of factor settings comes closest to the desired mean without significantly reducing the S/N ratio. You identify the best settings from the prediction results and then perform follow-up experiments with those settings to determine how well the prediction matches the observed result.

For example, an engineer for a golf equipment manufacturer has identified four controllable factors and one noise factor in a previous experiment. The engineer uses Predict Taguchi Results to determine which factor settings increase golf ball distance without increasing variability in the distance. Then, the engineer plans to conduct confirmation runs using the selected settings from the output to compare the observed results to the predicted results.

Where to find this analysis

Stat > DOE > Taguchi > Predict Taguchi Results

By using this site you agree to the use of cookies for analytics and personalized content.  Read our policy