Example of a static Taguchi design
main topic     interpreting results     session command     see also  

You manufacture golf balls and are working on a new design to maximize ball flight distance. You have identified four control factors, each with two levels:

·    Core material (liquid vs. tungsten)

·    Core diameter (118 vs. 156)

·    Number of dimples (392 vs. 422)

·    Cover thickness (.03 vs. .06)

You also want to test the interaction between core material and core diameter.   

The response is ball flight distance in feet. The noise factor is two types of golf clubs: driver and a 5-iron. You measure distance for each club type, resulting in two noise factor columns in the worksheet. Because your goal is to maximize flight distance, you select the larger-is-better signal-to-noise (S/N) ratio.

1    Open the worksheet GOLFBALL.MTW. The design and response data have been saved for you.

2    Choose Stat > DOE > Taguchi > Analyze Taguchi Design.

3    In Response data are in, enter Driver and Iron.

4    Click Analysis.

5    Under Fit linear model for, check Signal-to-noise ratios and Means. Click OK.

6    Click Terms.

7    Verify that A:Material, B:Diameter, C:Dimples, D:Thickness, and AB are in Selected Terms. Click OK.

8    Click Options.

9    Under Signal to Noise Ratio, choose Larger is better. Click OK in each dialog box.

Session window output

Graph window output

Interpreting the results

Each linear model analysis provides the coefficients for each factor at the low level, their p-values and an analysis of variance table. Use the results to determine whether the factors are significantly related to the response data and each factor's relative importance in the model.

The order of the coefficients by absolute value indicates the relative importance of each factor to the response; the factor with the biggest coefficient has the greatest impact. The sequential and adjusted sums of squares in the analysis of variance table also indicate the relative importance of each factor; the factor with the biggest sum of squares has the greatest impact. These results mirror the factor ranks in the response tables.

In this example, you generated results for S/N ratios and means. For S/N ratios, all the factors and the interaction terms are significant at an a-level of 0.10. For means, core material (p=0.045), core diameter (p=0.024), and the interaction of material with diameter (p=0.06) are significant because their p-values are less than 0.10.  However, because both factors are involved in the interaction, you need to understand the interaction before you can consider the effect of each factor individually.

The response tables show the average of each response characteristic (S/N ratios, means) for each level of each factor. The tables include ranks based on Delta statistics, which compare the relative magnitude of effects. The Delta statistic is the highest minus the lowest average for each factor. Minitab assigns ranks based on Delta values; rank 1 to the highest Delta value, rank 2 to the second highest, and so on. Use the level averages in the response tables to determine which level of each factor provides the best result.

In this example, the ranks indicate that core diameter has the greatest influence on both the S/N ratio and the mean. For S/N ratio, cover thickness has the next greatest influence, followed by core material and dimples. For means, core material has the next greatest influence, followed by dimples and cover thickness.

For this example, because your goal is to increase ball flight distance, you want factor levels that produce the highest mean. In Taguchi experiments, you always want to maximize the S/N ratio. The level averages in the response tables show that the S/N ratios and the means were maximized when the core was liquid, the core diameter was 118, there were 392 dimples, and the cover thickness was .06. Examining the main effects plots and interaction plots confirms these results. The interaction plot shows that, with the liquid core, the flight distance is maximized when the core diameter is 118.   

Based on these results, you should set the factors at:

Next, you may want to use Predict Results to see the predicted S/N ratios and means at these factor settings. See Example of predicting results.