The signal-to-noise ratio is a measure of robustness, which can be used to identify the control factor settings that minimize the effect of noise on the response. Minitab calculates a separate signal-to-noise (S/N) ratio for each combination of control factor levels in the design. You can choose from different S/N ratios, depending on the goal of your experiment. In all cases, you want to maximize the S/N ratio.
The predicted values are based on main effects and selected interactions calculated from these signal-to-noise ratios.
The mean is the average response for each combination of control factor levels in a static Taguchi design.
The predicted values are based on main effects and selected interactions calculated from these means.
The slope is the rate of change for the response relative to the signal factor. For each factor combination in a dynamic response experiment, Minitab calculates the slope of the least squares fitted line for the signal-response data passing through the reference point (if specified).
The predicted values are based on main effects and selected interactions calculated from these slopes.
The standard deviation measures the variability in the response due to noise. Minitab calculates a separate standard deviation for each combination of control factor levels in the design.
For a static design, Minitab calculates the standard deviation around the mean of the responses. For a dynamic design, Minitab calculates the standard deviation around the regression line. The standard deviation is the square root of MSE, or mean square error.
The predicted values are based on main effects and selected interactions calculated from these standard deviations.
Minitab calculates the natural log of the standard deviation for each combination of control factor levels in the design.
The predicted values are based on main effects and selected interactions calculated from these transformed standard deviations.
Minitab uses the model and the factor settings to calculate the predicted values. Use the factor settings table to verify that you performed the analysis as you intended.
When you have multiple factor level combinations, you can determine which combinations are best by comparing predicted values.